pme i j international journal of production management and engineering special issue: advances in engineering networks de castro, r. a1, gimenez, g. a2, mula, j.b1, boza, a.b2, cuenca, ll.b3, gómez-gasquet, p.b4, and vicens-salort, e.b5 a grup de recerca en enginyeria de procés, producte i producció. universitat de girona, spain b centro de investigación en gestión e ingeniería de producción. universitat politècnica de valència, spain a1 rudi.castro@udg.edu, a2 gerusa.gimenez@udg.edu, b1 fmula@cigip.upv.es, b2 aboza@cigip.upv.es, b3 llcuenca@cigip.upv.es, b4 pgomez@cigip.upv.es, b5 evicens@cigip.upv.es 1. introduction this special issue of the international journal of production management and engineering is dedicated on the most recent and relevant research, theories and practices in industrial engineering and operations management presented at the 12th international conference on industrial engineering and industrial management and xxii congreso de ingeniería de organización cio18. the cio’18 was held on 12th and 13th july 2018 in girona and was organized by the universitat de girona (udg) and the asociación para el desarrollo de la ingeniería de organización (adingor) and with the collaboration of gradient and grepp (udg research groups). the assistance was 104 participants. talking about classifications we could say that they are 35% women and 65% men, and about affiliation to adingor, 60% belong to association. finally, there were 244 authors from 4 countries and they contributed with a total of 100 communications, which were presented in parallel sessions (59 papers), in poster session (33 papers) and 8 in doctoral symposium. nine of the best communications were selected to be published in this special issue. we gratefully acknowledge the authors and particularly the reviewers, whose valuable comments have improved the quality of the selected papers, which were extended (and again reviewed by pairs) after the conference in order to be published in this special issue. 2. overview of the papers the papers chosen for this special issue were selected for their quality and their scientific rigor. the topics are diverse, but all refer to problems related to the cio’s scope. the conference was a forum to disseminate, to all branches of academy and industry, information on the most recent and relevant research, theories and practices in industrial engineering and operations management. the cio 2018 conference motto was: “advances in engineering networks”. the mission was to promote links between researchers and practitioners from different branches, to enhance an interdisciplinary perspective of industrial engineering and management. the first is article is patentometric: monitoring the scientific and technological trends of additive manufacturing in medical applications by izaskun alvarez, enara zarrabeitia, rosa rio-belver, iñaki bildosola, itziar martinez de alegria. this is about the use of patent statistics to monitor what is the state of the inventive activity of additive manufacturing in medical applications. the next one is also related to basque country, based on analysis of technological knowledge flows in the basque country by javier gavilanes-trapote, ismael etxeberria-agiriano, ernesto cilleruelo, gaizka garechana, iñaki bildosola the paper is about the influence of a region in technological knowledge generation by patents data. third article is methodology to validate results from european projects: the c2net case study by raquel sanchis, beatriz andres, raúl poler. conhttps://doi.org/10.4995/ijpme.2019.11901 to cite this article: de castro, r., gimenez, g., mula, j., boza, a., cuenca, ll., gómez-gasquet, p., and vicens-salort, e. (2019). special issue: advances in engineering networks. international journal of production management and engineering, 7(si), 63-64. https://doi.org/10.4995/ijpme.2019.11901 http://polipapers.upv.es/index.php/ijpme 63int. j. prod. manag. eng. (2019) 7(special issue), 63-64creative commons attribution-noncommercial-noderivatives 4.0 international orcid.org/0000-0002-8099-8493 http://orcid.org/0000-0003-3371-8759 http://creativecommons.org/licenses/by-nc-nd/4.0/ cretely they present the optimiser module of this project which has been diffused in many publications. the next one is testing successful business model using through system dynamics by jaione ganzarain, maria ruiz and juan ignacio igartua is related to face the firm performance complexity with a systemic view in order to understand better the phenomena. the next one is also related to a model applied in a r&d center: a multicriteria decision model for the evaluation and selection of technologies in a r&d centre by rafael lizarralde, jaione ganzarain, ander olaran. a new approach linked to key factors as technology, economy, human and organisational factors is presented and applied in an r&d center in basque country the last article is related to a project to implement green technology in a big city: electrification of madrid fleet public transport company (emt): strategic analysis and implementation by jaime garcía hernanz, gustavo morales, gonzalo fernández sánchez, eduardo pilkington gonzález, teresa sánchez-chaparro. this is a study of a future strategic plan of the company applied in madrid in order to support that the electrification of the whole fleet is the best option in the long term. we hope you enjoy the reading of these papers and contribute to extend the knowledge of these topics in the area of industrial engineering. de castro, r. et al. 64 int. j. prod. manag. eng. (2019) 7(special issue), 63-64 creative commons attribution-noncommercial-noderivatives 4.0 international http://creativecommons.org/licenses/by-nc-nd/4.0/ titular capítulo http://doi.org/10.4995/ijpme.2019.10163 received: 2018y-05-16 accepted: 2019-07-18 to cite this article: hernadewita, rochmad, i., hendra, hermiyetti, yuliani, e.n.s. (2019). an analysis of implementation of taguchi method to improve production of pulp on hydrapulper milling. international journal of production management and engineering, 7(2), 125-131. https://doi.org/10.4995/ijpme.2019.10163 creative commons attribution-noncommercial-noderivatives 4.0 international int. j. prod. manag. eng. (2019) 7(2), 125-131 | 125 https://polipaper.upv.es/index.php/ijpme an analysis of implementation of taguchi method to improve production of pulp on hydrapulper milling. hernadewita a , rochmad, i. b , hendra c , hermiyetti d , yuliani, e.n.s. e a,b,e magister of industrial engineering, mercu buana university, jl. meruya selatan, jakarta, indonesia c mechanical engineering, university of bengkulu, jl. kandang limun, bengkulu, indonesia. d faculty of economic social science, bakrie university, jl. hr rasuna said, jakarta selatan, indonesia. a hernadewita@mercubuana.ac.id , a imb_rochmad@yahoo.co.id . abstract: taguchi method is one of a design of experimental (doe), by using statistical approach to optimize the process parameters and maintaining the minimum variability and also improve the quality of product. based on data characterisation, nominal is best in taguchi methods is suitable application in this study. its describe the procedures and steps that occur in doe to find an optimum quality parameter corresponding quality characterisation. nominal is the best applied in milling process of pulp on the hydrapulper with pulp freeness 650 canadian standard freeness (csf). the result is shown by orthogonal array, signal-to-noise (s/n) ratio and analysis of variances (anova). three factors cosidered in this study and namely the composition of pulp (waste paper), pulp consistency and milling time. the experiment will conducted after determination of each level and the appropriate orthogonal array was selected. after measuring of pulp freeness produced by the pulp milling on the hydrapulper, then signal-tonoise (s/n) ratio is calculated. as the conclussion, the factors and levels of optimum freeness obtained, pulp composition in level 1 (100%), pulp consistency at level 2 (8%) and milling time factor in level 2 (45 minutes). the result of experimental verification was interpreted in the conclusion. keywords: taguchi, doe, pulp, hydrapulper, nominal is the best, anova. 1 introduction in the globalization era, the competition among the business is very tight. one is shown by the development of materials input in industry. its indicated by arising on launching the new products with variety of brands into the market, either national and international. in such situation, the main problem needs to be solve by businesses in order to survive is by giving focused on satisfaction to customers compared to the other factors. there is no doubt that the main factor determining of customer satisfaction is the quality of product or services. organizations or companies which meet quality requirement of products and services will be satisfying the customer needs. in fullfilness of customer satisfaction, as one of the fiber cement company in indonesia, pt. bbi is committed to constantly improve their competitiveness and seize in a larger market, thus increasing in the quality of products and services as one of the determinants factor of customer satisfaction. as one of the flagship products of pt. bbi is a corrugated fiber cement, which is produced in sheet machine, line i to iv. which is one of main content are milleing pulp in hydrapulper with specific freeness of 650 canadian standard mailto:ahernadewita@mercubuana.ac.id mailto:imb_rochmad@yahoo.co.id hernadewita et al. 126 | int. j. prod. manag. eng. (2019) 7(2), 125-131 creative commons attribution-noncommercialnoncommercial-noderivatives 4.0 freeness (csf). the freeness of the pulp mill on hydrapulper (fatoki et al., 2015) currently very varietive, so that, it will have an impact to overall quality of fiber cement for corrugating roof. to improve the quality of the pulp production, so then, its conducted taguchi experimental design approach with characterisation of quality by nominal is the best, in achieving the freeness of milled pulp of 650 csf. the objectives in this study was to determine the factors that influence the quality of the pulp slurry freeness and examine of each level of optimal factor for the improvement of the quality of the pulp milling on the hydrapulper. 2 taguchi method this study is used an experimental design approach with the application of taguchi method. taguchi method is one of a new method in the field of engineering that objectives to improve the quality of products and processes and the mean time will costs and resources to a minimum (taguchi, 1993; kawamura, 2010; athreya and venkatesh, 2012; bellavendram 1995; dobrzañski, 2007; hassan et al., 2012; kamarudin et al., 2004; lajis et al., 2009; roy, 2010; verma et al., 2012; yadav et al., 2012). the objective of taguchi method is to meet the product robust against noise, as it is often referred as the robust design. robust design method, also called the taguchi method, pioneered by dr. genichi taguchi, greatly improves engineering productivity. by consciously considering the noise factors (environmental variation during the product’s usage, manufacturing variation, and component deterioration) and the cost of failure in the field the robust design method helps ensure customer satisfaction. robust design focuses on improving the fundamental function of the product or process, thus facilitating flexible designs and concurrent engineering. indeed, it is the most powerful method available to reduce product cost, improve quality, and simultaneously reduce development interval. as the definition of quality according to taguchi is loss received by the public since the product was shipped. taguchi’s philosophy of quality consists of four concepts (taguchi, 1993), as: 1. quality should be designed into the product and not just duing control by quality check. 2. the best quality is achieved by minimising the deviation from the target. 3. products must be designed to be robust against environmental factors that could not be controlled. 4. the cost of quality should be measured in a function of a certain standard deviation, so that the loss should be measured in the whole system. taguchi method characterise as off-line quality control, which means as preventive quality control in product design or production process before arriving at the shop floor level. off-line quality control is determined at the beginning of the life cycle of product improvement at the beginning of the product (to get right first time). taguchi contribution to quality are: 1. loss function: represents the loss produced by the people (producers and consumers) due to the quality produced. for producers with the cost of quality carried out by customers is their dissatisfaction or frustration of products purchased or used because of poor quality. 2. orthogonal array: used to design an experiment that efisisen and used to analyze experimental data. orthogonal array is used to determine the minimum number of experiments that can give as much information as possible all factors that influence the parameter. the most important part of the orthogonal array lies in the selection level combination of input variables for each experiment. 3. robustness: minimizing the sensitivity of the system to the sources of variation. an analysis of implementation of taguchi method to improve production of pulp on hydrapulper milling. creative commons attribution-noncommercialnoncommercial-noderivatives 4.0 int. j. prod. manag. eng. (2019) 7(2), 125-130 | 127 2.1 taguchi design of experiments taguchi experimental design is an assessment simultaneously to two or more factors (parameters) affecting the ability of the average or the variability of the combined results of the features of the product or process. some of the steps proposed by taguchi to experiment systematically, namely: a. formulation of the problem b. experimental purposes c. determination of the dependent variable d. identify the factors (independent variables) e. separation of control factors and noise factors f. specifies the number of levels and the level of each factor g. the calculation of degrees of freedom h. selection of an orthogonal matrix i. placement of the factors and the interaction space into an orthogonal array j. implementation of the experiment according to an orthogonal array table 2.1. determination of total level and level value factor. table 2.2. results of experiments with orthogonal array l27 (313). hernadewita et al. 128 | int. j. prod. manag. eng. (2019) 7(2), 125-131 creative commons attribution-noncommercialnoncommercial-noderivatives 4.0 k. analyzing the experimental data with anova, calculate the optimal quality prediction. l. the implementation of the verification experiment. 3 result based on taguchi experimental and the result to the orthogonal array, the data processed came out as follows table 3.1 to 3.8. to reach the intended target (nominal is the best), the determination of the optimal factor level is the result on the test that approach the pulp freeness on 650 csf. so that, the optimal table 3.1. factors interaction solutions a and b. table 3.2. response pulp freeness average of factors effect. table 3.3. analysis of variance combined pulp freeness average. table 3.4. percentage constribution. table 3.5. response s / n ratio pulp freeness of factors effect. table 3.6. factors interaction solutions. an analysis of implementation of taguchi method to improve production of pulp on hydrapulper milling. creative commons attribution-noncommercialnoncommercial-noderivatives 4.0 int. j. prod. manag. eng. (2019) 7(2), 125-130 | 129 combination of factors level are: a1 = waste paper 100%. b2 = pulp consistency 8%. c2 = milling time for 45 minutes. the percentage contribution from the table 3.4 shows that the factor c (milling time) was contributed to the most of average freeness of the pulp, which 55.205% followed by factor b (consistency) 20.084% and factor a (waste paper) 14.561%. to obtain the target of nominal is the best, combination of optimal factor level achieved in the average value of s/n ratio is the lowest level of each factor, indicating the smaller of the value closer to the target. the optimum level factors are: a1 = waste paper 100%. b2 = pulp consistency 8%. c2 = milling time for 45 minutes. 3.1 analysis of varians (anova) s/n ratio pulp freeness average the anova for s/n ratio pulp freeness average is shown of table 3.7 and table 3.8. the percentage contribution from the table 3.7 and 3.8, shows that the factor c (milling time) contributed to the most of average freeness of the pulp, which is 17.121% followed by interaction of factors axc (1) 0.381% and an average interaction axb (2) -3.535%. 4 conclusions and discussion as the definition of pulp freeness on milling process is spread/decomposition of pulp fibers after the milling process, the measuring standard of freeness on industry are generalise in three types, such as; canadian standard freeness (csf), schopper riegler (osr) and williams slowness (s). as the result from taguchi experimental, it shown that the freeness of pulp produced from hydrapulper milling prediction increasing the quality (see fig. 4.1). also, as the experimental verification data showed an increasing in the quality of milling process significantly and the freeness of the pulp milled on hydrapulper more stable (see fig. 4.2). 4.1 research limitations the accuracy of the results was strongly influenced by the choosen of the instruments. the table 3.7. analysis of variance combined s/n ratio pulp freeness. table 3.8. percent contribution of s/n ratio pulp freeness. table 4.1. interpretation of measurement results of pulp freeness average. hernadewita et al. 130 | int. j. prod. manag. eng. (2019) 7(2), 125-131 creative commons attribution-noncommercialnoncommercial-noderivatives 4.0 suitable measuring instruments will be determining the accuracy on the result of study. furthermore, reference to relevant support, with the access to reference study related will facilitate the next coming research in the same field and differences method and instruments. number of limitations in the this study, are: 1. the measurement scale on range was used a multiply of 10. so that, the measuring results are less accuracy on readings than compared to the results of calculations verification on taguchi experimental. 2. the lack of references on this study. in this study there was little impediment as a comparison. furthermore, the comparison was used cross references study in the sense of the pulp. the taguchi design experimental is considered into the factors that influence the quality characteristization, without considering to the interference factors. thus, the experiment was focused on the implementation of complexity and significant costs. as the engine noise is not included in this study, the effect of engine noise also can affect the accuracy and quality of experimental results. 5 conclusion and discussion 5.1 conclusion 1. the quality improvement of the pulp freeness produced from hydrapulper milling was shown by taguchi experimental design implementation with the character of nominal is the best (650 csf). the obtained factors and the level of each factors that influence the pulp freeness produced from hydrapulper milling processing, as follows: a. the composition of recycled paper (waste paper) at the level 1 (100%). b. pulp consistency at level 2 (8%). c. milling time at level 2 (45 minutes). 2. the use of recycled paper of cement bags (waste paper) is decreased the production cost of the fiber cement roofing. as the price of recycled paper of cement bags are generally cheaper than new paper that used as an imported, and at the same time maintaining appropriate quality standards. 5.2 discussion in addition to the conclusions, the recommendation on more accurate and detail of experimental results to quality improvement are as follows: figure 4.1. the current pulp freeness produced from hydrapulper milling. figure 4.2. the pulp freeness produced from hydrapulper milling on experimental verification. an analysis of implementation of taguchi method to improve production of pulp on hydrapulper milling. creative commons attribution-noncommercialnoncommercial-noderivatives 4.0 int. j. prod. manag. eng. (2019) 7(2), 125-130 | 131 1. carefully use a measuring tube with a scale size and clearly, will avoid errors in reading, so it will produce accurate information or data that used as the experimental results. 2. the use of taguchi experimental design can be developed, such as in engineering tests to find the best product quality in manufacturing line. and also improved the quality of product for fiber cement roofing and other products are affected by several factors and levels. references athreya, s., venkatesh, y.d. (2012). application of taguchi method for optimization of process parameters in improving the surface roughness of lathe facing operation. international refereed journal of engineering and science (irjes). 1(3), 13-19. bellavendram, n. (1995). quality by design: taguchi techniques for industrial experimentation. prentice hall, london. dobrzañski, l.a. (2007). application of taguchi method in the optimisation of filament winding of thermoplastic composites. international scientific journal, 28(3). published monthly as the organ of the committee of materials science of the polish academy of sciences. fatoki, j.g., omoniyi, t.e., onilude, m.a. (2015). design and fabrication of a hydrapulper for disintegrating disused tetra pack® beverage cartons. european international journal of science and technolog,4(6), 113-123. hassan, k., kumar, a., garg, m.p. (2012). experimental investigation of material removal rate in cnc turning using taguchi method. international journal of engineering research and applications, 2(2), 1581-1590. kamarudin, s., zahid, a., wan, k.s. (2004). the use of taguchi method in determining the optimum plastic injection moulding parameters for the production of a consumer product. jurnal mekanikal. disember 2004, bil.18, 98-110. kawamura, t. (2010). performance measures for robust design and its applications. 4th international workshop on reliable engineering computing (rec 2010). isbn: 978-981-08-5118-7. published by research publishing services. lajis, m.a., radzi, h.c.d., amin, a.k.m.n. (2009). the implementation of taguchi method on edm process of tungsten carbide. european journal of scientific research, 26(4), 609-617. roy, r.k. (2010). a primer on the taguchi method. 2nd edition. society of manufacturing engineers. dearborn: michigan. usa. selvaraj, d.p., chandramohan, p. (2010). optimization of surface roughness of aisi 304 austenitic stainless steel in dry turning operation using taguchi design method. journal of engineering science and technology, 5(3), 293-301. taguchi, g. (1993). taguchi on robust technology development: bringing quality engineering upstream. okhen associates. tokyo. japan. asme press new york. https://doi.org/10.1115/1.2905506 verma, j., agrawal, p., bajpai, l. (2012). turning parameter optimization for surface roughness of astm a242 type-1 alloys steel by taguchi method. international journal of advances in engineering & technology, 3(1), 255-261. yadav, u.k., narang, d., attri, p.s. (2012). experimental investigation and optimization of machining parameters for surface roughness in cnc turning by taguchi method. international journal of engineering research and applications, 2(4), 2060-2065. https://doi.org/10.1115/1.2905506 pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.12173 received: 2019-08-02 accepted: 2020-03-26 genetic algorithms for the scheduling in additive manufacturing castillo-rivera, s.a1, de antón, j.a2, del olmo, r.b, pajares, j.a3, lópez-paredes, a.a4 a insisoc research group (uic086), university of valladolid, paseo del cauce 59, 47011 valladolid, spain. b insisoc research group (uic086), university of burgos, c/ villadiego s/n. 09001, burgos, spain. a1 salvador.castillo@uva.es, a2 juan.anton@uva.es b rdelolmo@ubu.es, a3 pajares@insisoc.org, a4 adolfo@insisoc.org abstract: genetic algorithms (gas) are introduced to tackle the packing problem. the scheduling in additive manufacturing (am) is also dealt with to set up a managed market, called “lonja3d”. this will enable to determine an alternative tool through the combinatorial auctions, wherein the customers will be able to purchase the products at the best prices from the manufacturers. moreover, the manufacturers will be able to optimize the production capacity and to decrease the operating costs in each case. key words: scheduling, packing problem, genetic algorithm. 1. introduction nowadays, am has become a relevant technology in some manufacturing environment, due to it can be used in application such as customized production, (zhang et al., 2014). the fast development is proof that it is a suitable technology for manufacturing components and final products, (pour et al., 2016). the process that covers, from the model preparation phase to the prototype manufacturing, is complex and this is because of different departments may be involved. the performance of the 3d printing procedure can be summarized in figure 1. as shown, the first step is to carry out digital modelling, and it consists of building up a 3d model according to customer order. a software package often called cad (computer-aided design) is used to achieve this (ikonen et al., 1997). a file should be generated in the adequate format, it is usually taken as “stl” (standard triangle language) and the whole geometrical information should be included to set up the digital model, it is called the exportation. in the slice, the digital model must be converted into a list of 3d print out commands to execute the corresponding code (g-code), the connection allows to provide a list of the printer instructions through the usb port or copying the file in a memory card, which could be directly read by the printer. once this step is done, the placement of pieces on the build platform and the printing must be carried out; finally, the object/objects should be removed from the printing platform. to cite this article: castillo-rivera, s., de antón, j., del olmo, r., pajares, j., lópez-paredes, a. (2020). genetic algorithms for the scheduling in additive manufacturing. international journal of production management and engineering, 8(2), 59-63. https://doi.org/10.4995/ijpme.2020.12173 figure 1. 3d printing procedure flow chart. int. j. prod. manag. eng. (2020) 8(2), 59-63creative commons attribution-noncommercial-noderivatives 4.0 international 59 mailto:salvador.castillo@uva.es http://creativecommons.org/licenses/by-nc-nd/4.0/ this procedure impacts a company on planning and scheduling as well as the manufacturing process, (ahsan et al., 2015). otherwise, the current methods have not shown their suitability to address the requirement for a paradigm shift. consequently, it is necessary to obtain performances that are done through the implementation of am, (pour et al., 2016). am is being focussed on the manufacturing of many pieces production environment, the planning and scheduling of the corresponding items to be processed on the am machines. the arrangement of pieces on the build platform is another problem in the am scheduling, an adequate approach will allow to optimize the build jobs as well as the machine utilization. highly uncertain demand can cause a revision of production planning, being one of the main cost drivers because of adverse effects on inventory and labour levels (demirel et at. 2018). scheduling of 3d printers presents new problems linked to its production capability and it must cope to packing problem, which is presented as the numbers of pieces are put on the corresponding build platform surfaces. the planning of parts to be processed on 3d printers performs a fundamental role in reducing operational costs; this should provide a less price and increase the profitability of the companies. as a consequence, lonja3d is designed to purchase products made with 3d printing, which facilitates the organization and coordination of collaborative between the customers that will receive the bids from the manufacturers of 3d printing services. this market will allow customers to get better prices from the manufacturers. on the other hand, these last ones can optimize their installed production capacity and they can decrease operating costs in each case according to the technology. modern manufacturing environment has been featured by customer demands, raised product variety, demand for limited production cycles, etc. all of them highlight the relevance of adopting new technologies and systems to deal with this. different algorithms have been mainly used for planning and scheduling in am. another feature that presents an impact on scheduling is the packing problem. one of the first challenges that “lonja 3d” has to face, is to figure out a suitable arrangement of multiple items in different build platforms, minimizing the empty area. this presents a relevant interest due to it entails the reduction of the production costs. the complexity of the problem must be dealt with, taking into account the geometry of the objects and using a successful method of resolution. gas are one of the most known algorithms based on the principles of artificial intelligence that have found out its use in different branches of science. gas can be used to solve the two or three-dimensional packing problems. this work presents the preliminary results of an ongoing study. a theoretical framework of am and ga has been done, to shed light on the challenging that the 3d printing industry must face. it has been carried out because the authors are currently working in the design of a managed market called “lonja3d”. it should be used to purchase products made using 3d printing. in this market, the coordination and organization of offers will be eased between the customers that will receive the bids from the providers of 3d printing services. this “lonja3d” or market will allow customers to obtain better prices from the manufacturers. in addition to this, the manufacturers can optimize their installed production capacity and they can decrease operating costs in each case according to the technology. the work is organized as follows: section 2 presents the theoretical framework. section 3 deals with additive manufacturing and genetic algorithms. section 4 summarizes the results and conclusions. 2. theoretical framework process planning and scheduling are two of the most important functions in manufacturing systems that have shown a large influence on product design along with the considerations of timing aspects and technological. these are taken into account as two different functions in a manufacturing environment. however, a critical problem has been raised with the potential action of alternative machines, setups and processes for producing a specific part in a manufacturing shop has got. if scheduling and process planning are implemented independently, the schedule that is derived lacks flexibility and adaptability. consequently, for obtaining realistic and effective timetables both functions should be considered, (kim and egbelu, 1999). alternative plans allow allocation of tasks to other machines with further flexibility and it decreased the possibility of conflict between a job and a machine. int. j. prod. manag. eng. (2020) 8(2), 59-63 creative commons attribution-noncommercial-noderivatives 4.0 international castillo-rivera et al. 60 http://creativecommons.org/licenses/by-nc-nd/4.0/ the integration of the process planning and scheduling come from the class of most tricky combinatorial problems and it needs high efficient approaches for finding optimal solutions. therefore, its integration should be enhanced the efficiency as well as the performance of manufacturing systems (milošević et al., 2016). the integrating process planning and scheduling functions is a change of model in most am organizations. the alternative operations for a specific job explained that determined arrangements lead to an advantage in a manufacturing environment (wilhelm and shin, 1985; xirouchakis et al., 1998). alternative layouts are able to be used for: a) to determine disruption problems such as machine overloads and machine breakdowns, among others. b) reduction of process inventory. c) rising equipment utilization. li et al. (2017) presented a mathematical model to formulate the production planning in am problem. to figure out the optimum allocation of several pieces on a set of machines with different specifications such as unit time cost and processing speed, among others. the average production cost per volume of material was minimized. only one material was taken into account, nesting problems, and the due dates for fulfilling orders were not considered. two heuristic procedures were established called “bestfit” and “adapted best-fit” rules. a computational study was done to assess the performance of the heuristics. furthermore, the necessity of developing specific planning and scheduling was proven for am processes. chergui et al. (2018) presented the planning, nesting and scheduling problem in am to achieve the orders got from different customers by due dates. a mathematical formulation of the problem was proposed and a heuristic approach was also derived using python to work out it. experimental assessments were carried out, these showed the need for developing suitable am planning and scheduling methods for satisfying the technical and the organizational requirements of am. the scheduling and freedom of design are often limited due to the printing complex geometries are presented in am. the packing problem is also presented in mechanical design and manufacture, among others. araújo et al. (2015) introduced a classification of am build as well as a summary of existing benchmarks for the study of these problems. furthermore, it is discussed benchmarks to encourage research toward effective and efficient am to build volume packing, which should be integrated into am manufacturing planning methodologies. canellidis et al. (2006) described a pre-processing methodology that makes automatic the procedure of finding suitable fabrication orientations and packing arrangements. the methodology consisted of two separate stages as the orientation and the packing stage. in the beginning, each part was adequately oriented to obtain better surface quality and minimal projection area, lower build time, etc. the second stage took into account the projections of the parts on the fabrication platform. a ga along with a new improved placement rule was used to lead the associated 2d bin-packing problem. two sets of studies were employed to prove the performance of the approach, which involved simple nearly orthogonal-shaped parts as well as objects/parts. modern gas have shown to be trustworthy for finding optimal process schedules. lawrynowicz (2011) provided a survey of developments in building ga for advanced scheduling. a new approach was proposed to the distributed scheduling in industrial clusters using a modified ga. milošević et al. (2016) presented the state of the art review of gas for optimization of process planning, scheduling and their corresponding integration. hybrid approaches and different modifications were briefly studied. genetic components and strategies were also presented with some sample parts, which are frequently taken into account as testing ga performances. the capability of producing pieces with different geometries at the same time is a challenge that is presented in scheduling. as a consequence, fera et al. (2018) provided a mathematical model for an am/slm (selective laser melting) machine scheduling. the complexity of the model was nphard and the likely solutions should be figured out by metaheuristic algorithms such as gas. these solve sequential optimization problems by handling vectors. the proposed algorithms were assessed on a case composed of a thirty part number production plan with an elevated variability and complexity. 3. additive manufacturing and genetic algorithms most of the packing problems found in the specialist literature are two dimensional. according to the items to be placed, two main groups can be distinguished, those which present regular or irregular shapes. regular figures have shapes that are established by a few parameters such as rectangles, circles. on the other hand, the term irregular is applied to int. j. prod. manag. eng. (2020) 8(2), 59-63creative commons attribution-noncommercial-noderivatives 4.0 international genetic algorithms for the scheduling in additive manufacturing 61 http://creativecommons.org/licenses/by-nc-nd/4.0/ asymmetrical concave and convex shapes. it must be optimized the production capacity and the reduction of operating costs, establishing a production rate with the least space on the build platform, (hopper and turton, 1997). a first approach can be done through a layout space which the size could be determined i.e., rectangle packing problem. this scenario presents common features with a standard problem of how it can be the knapsack. the rectangles must be located into the layout of space and required to satisfy the constraints to derive certain optimal standards. a ga can be implemented according to the requirements established by the authors to derive suitable outcomes for “lonja 3d”. hence, it is considered a set of objects {1,...,n} each one j is characterized by a weight function pj and a value vj, being these positive integer numbers. the surface of the build platform should contain sets of objects which surface should not exceed p0 units, and besides the total number of them should be maximum. it should be assumed that pj ≤ p0 ∀j and ∑ n j=1pj > p0. the mathematical model should be set up as a variable xj associated with each set of objects j. xj = 1 if the set of objects j should be inserted or xj = 0, otherwise. as a result, an integer linear programming can be derived to implement the ga: ax xm ∑ n j=1 vj j (1) subject: x ≤p∑ n j=1 pj j 0 (2) 0 ≤ xj ≤ 1; ∀j ∈ 1, …, n xj ∈ z; ∀j ∈ 1, …, n the evolution process starts with the generation of an initial random population. the fitness of each member of the population is computed and probabilities are assigned to each individual. the reproducing population is formed (selection) by drawing with replacement to sample where each individual has a probability of surviving. a new population is generated from the reproducing population using crossover and mutation operators. after this, the algorithm returns to the fitness evaluation step. when convergence criteria are met the evolution stops, see figure 2. 4. results and conclusions 3d printing shows characteristics that other industries have not displayed. 3-d printing allows small quantities of customized lots to be produced at relatively low costs, which will significantly decrease the advantages of producing small good sizes in low-wage countries, through reduced need for workers (berman, 2012). based on the costs supplied by the different manufacturers for the printing of parts, “lonja3d” will develop at each moment the best feasible offers with the demands of the users. allowing to focus the order to the most efficient manufacturer for every technology. 3d printing manufacturing presents some restrictions as the reduced surface of the build platform and printing time, among others. this can be traceable throughout the several approaches as gas that enables to study for example the packaging problem. the geometry and the dimensions of the items implicated in the arrangement method enable to classify these problems. a ga has been presented to establish an adequate stage for setting up “lonja 3d”. the work provides suitable information for 3d printing industry scheduling. to expand the work, combinatorial auctions will be used to optimize the advantageous participation of the productive capacity of the manufacturers to the proper demand at each moment, also raising the competition between all the producers and improving the profits for the customers. these types of auctions have been successfully employed in fields as federal communications commission’s radio spectrum licenses. figure 2. flow chart of the genetic algorithm. int. j. prod. manag. eng. (2020) 8(2), 59-63 creative commons attribution-noncommercial-noderivatives 4.0 international castillo-rivera et al. 62 http://creativecommons.org/licenses/by-nc-nd/4.0/ acknowledgements this research has been partially financed by the project: “lonja de impresión 3d para la industria 4.0 y la empresa digital (lonja3d)” funded by the regional government of castile and leon and the european regional development fund (erdf, feder) with grant va049p17. 5 references ahsan, a., habib, a., khoda, b. (2015). resource based process planning for additive manufacturing. computer-aided design, 69, 112125. https://doi.org/10.1016/j.cad.2015.03.006 araújo, l., özcan, e., atkin, j., baumers, m., tuck, c., hague, r. (2015). toward better build volume packing in additive manufacturing: classification of existing problems and benchmarks. 26th annual international solid freeform fabrication symposium an additive manufacturing conference, 401-410. berman, b. (2012). 3-d printing: the new industrial revolution. business horizons, 55: 155-162. https://doi.org/10.1016/j.bushor.2011.11.003 canellidis, v., dedoussis, v., mantzouratos, n., sofianopoulou, s. (2006). preprocessing methodology for optimizing stereolithography apparatus build performance. computers in industry, 57, 424-436. https://doi.org/10.1016/j.compind.2006.02.004 chergui, a., hadj-hamoub, k., vignata, f. (2018). production scheduling and nesting in additive manufacturing. computers & industrial engineering, 126, 292-301. https://doi.org/10.1016/j.cie.2018.09.048 demirel, e., özelkan, e.c., lim, c. (2018). aggregate planning with flexibility requirements profile. international journal of production economics, 202, 45-58. https://doi.org/10.1016/j.ijpe.2018.05.001 fera, m., fruggiero, f., lambiase, a., macchiaroli, r., todisco, v. (2018). a modified genetic algorithm for time and cost optimization of an additive manufacturing single-machine scheduling. international journal of industrial engineering computations, 9, 423-438. https://doi.org/10.5267/j.ijiec.2018.1.001 hopper, e., turton, b. (1997). application of genetic algorithms to packing problems a review. proceedings of the 2nd online world conference on soft computing in engineering design and manufacturing, springer verlag, london, 279-288. https://doi.org/10.1007/978-1-4471-0427-8_30 ikonen, i., biles, w.e., kumar, a., wissel, j.c., ragade, r.k. (1997). a genetic algorithm for packing three-dimensional non-convex objects having cavities and holes. icga, 591-598. kim, k.h., egbelu, p.j. (1999). scheduling in a production environment with multiple process plans per job. international journal of production research, 37, 2725-2753. https://doi.org/10.1080/002075499190491 lawrynowicz, a. (2011). genetic algorithms for solving scheduling problems in manufacturing systems. foundations of management, 3(2), 7-26. https://doi.org/10.2478/v10238-012-0039-2 li, q., kucukkoc, i., zhang, d. (2017). production planning in additive manufacturing and 3d printing. computers and operations research, 83, 157-172. https://doi.org/10.1016/j.cor.2017.01.013 milošević, m., lukić, d., đurđev, m., vukman, j., antić, a. (2016). genetic algorithms in integrated process planning and scheduling–a state of the art review. proceedings in manufacturing systems, 11(2), 83-88. pour, m.a., zanardini, m., bacchetti, a., zanoni, s. (2016). additive manufacturing impacts on productions and logistics systems. ifac, 49(12), 1679-1684. https://doi.org/10.1016/j.ifacol.2016.07.822 wilhelm, w.e., shin, h.m. (1985). effectiveness of alternate operations in a flexible manufacturing system. international journal of production research, 23(1), 65-79. https://doi.org/10.1080/00207548508904691 xirouchakis, p., kiritsis, d., persson, j.g. (1998). a petri net technique for process planning cost estimation. annals of the cirp, 47(1), 427-430. https://doi.org/10.1016/s0007-8506(07)62867-4 zhang, y., bernard, a., gupta, r.k., harik, r. (2014). evaluating the design for additive manufacturing: a process planning perspective. procedia cirp, 21, 144-150. https://doi.org/10.1016/j.procir.2014.03.179 int. j. prod. manag. eng. (2020) 8(2), 59-63creative commons attribution-noncommercial-noderivatives 4.0 international genetic algorithms for the scheduling in additive manufacturing 63 https://doi.org/10.1016/j.cad.2015.03.006 https://doi.org/10.1016/j.bushor.2011.11.003 https://doi.org/10.1016/j.compind.2006.02.004 https://doi.org/10.1016/j.cie.2018.09.048 https://doi.org/10.1016/j.ijpe.2018.05.001 https://doi.org/10.5267/j.ijiec.2018.1.001 https://doi.org/10.1007/978-1-4471-0427-8_30 https://doi.org/10.1080/002075499190491 https://doi.org/10.2478/v10238-012-0039-2 https://doi.org/10.1016/j.cor.2017.01.013 https://doi.org/10.1016/j.ifacol.2016.07.822 https://doi.org/10.1080/00207548508904691 https://doi.org/10.1016/s0007-8506(07)62867-4 https://doi.org/10.1016/j.procir.2014.03.179 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering special issue: advances in engineering networks https://doi.org/10.4995/ijpme.2019.10789 received: 2018-10-09 accepted: 2019-01-24 patentometric: monitoring the scientific and technological trends of additive manufacturing in medical applications alvarez-meaza, i. a1*, zarrabeitia-bilbao, e. a2, rio-belver, r.m. b1, martinez de alegria, i. a3, and bildosola, i. a4 auniversity of the basque country, industrial organization and management engineering department. faculty of engineering in bilbao. spain. buniversity of the basque country, industrial organization and management engineering department. faculty of engineering in vitoria -gasteiz. spain a1 izaskun.alvarez@ehu.eus, a2 enara.zarrabeitia@ehu.eus, a3 itziar.martinezdealegria@ehu.eus, a4 inaki.bildosola@ehu.eus b1 rosamaria.rio@ehu.eus abstract: patents are a means of protecting inventions developed by firms, institutions or individuals, and they may be interpreted as indicators of invention. patents indicators convey information on the processes of inventive activities. therefore, patent statistics will assess science and technology (s&t) activities. besides, additive manufacturing (am) has become a revolutionary technology that is changing medical science. for this reason, the patent statistics will allow us to monitor what is the state of the inventive activity of am in medical applications. the database used in order to retrieve patent information is patseer and the data have been analyzed through the analytics package called quick stats. from the data obtained, it can be concluded that, additive manufacturing in medical applications is an emerging technology with huge market potential. undoubtedly, the core of invention is located in united states, followed by germany, united kingdom and china somewhat behind. firms are the main holders of legal rights, and the firm’s market value and the knowledge diffusion of technology are ensured by the technological diversity and the number of forward citations presented by patents. key words: additive manufacturing, medical applications, patents, ipc, forward citations. 1. introduction additive manufacturing (am) is defined as “the process of joining materials to make parts from 3d model data, usually layer upon layer, as opposed to subtractive manufacturing and formative manufacturing methodologies” (astm, 2015). this approach has revolutionised the manufacturing process, because it offers the possibility of manufacturing parts of any geometric complexity without using additional tools or machines (atzeni & salmi, 2012; galati & iuliano, 2018; gibson et al., 2010; hopkinson & dickens, 2006). synonyms found in the literature include additive processes, additive techniques, additive fabrication, rapid manufacturing, three-dimensional (3d) printing, rapid prototyping, layer manufacturing, additive layer manufacturing, direct digital manufacturing, freeform fabrication and so on (jin et al., 2017; mellor, et al., 2014). therefore, it is not surprising that the main advantages of additive manufacturing that jin et al. (2017) have synthesized from previous research works are: parts can be made easily on-demand for to cite this article: alvarez-meaza, i., zarrabeitia-bilbao, e., rio-belver, r.m., martinez de alegria, i., and bildosola, i. (2019). patentometric: monitoring the scientific and technological trends of additive manufacturing in medical applications. international journal of production management and engineering, 7(special issue), 65-72. https://doi.org/10.4995/ijpme.2019.10789 int. j. prod. manag. eng. (2019) 7(special issue), 65-72creative commons attribution-noncommercial-noderivatives 4.0 international 65 http://orcid.org/0000-0002-2110-0719 http://orcid.org/0000-0002-2347-3885 http://orcid.org/0000-0002-4244-9098 http://orcid.org/0000-0003-0731-3010 https://orcid.org/0000-0002-4964-9762 mailto:izaskun.alvarez@ehu.eus mailto:enara.zarrabeitia@ehu.eus mailto:itziar.martinezdealegria@ehu.eus mailto:rosamaria.rio@ehu.eus http://creativecommons.org/licenses/by-nc-nd/4.0/ customization and personalization, fast and free manufacturing can be realized with no time limits in 3-dimensional structure, the cost of manufacturing can be reduced significantly and am enables the environmental friendly product design. therefore, additive manufacturing has become a revolutionary technology that has managed to change both the production management systems and the research strategies in developing countries. applications for am are found in many fields including aerospace, architecture, medical devices or healthcare, consumer products, automotive, jewellery and defence applications (mellor et al., 2014; petrovic et al., 2011). if we highlight the applications in health sector, am enables the production or the creation of prostheses, pre-surgery planning tools, alignment jigs and surgical cutting templates. however, it is particularly necessary to emphasize the importance of 3d bioprinting technology that can be used in organ production and also creating live tissues that retain biological functions (rodriguez-salvador et al., 2017). taking into account the importance of am in human health, the study of scientific production regarding to this technology becomes critical. rodriguez et al. (rodriguez-salvador et al., 2017) described that the number of this studies are still lacking. in addition, the study of patents in a specific technology such as laser additive manufacturing made by zarrabeitia (zarrabeitia et al., 2017) concluded that the international patent classification (ipc) of patents with the highest number of patents is the a61, medical applications. in order to determine general trends of am technology following a qualitative process, the study of patents is a very appropriate method. the patent statistics have been used to define the path of the science assessing scientific and technology activities for a long time. patents provide a uniquely detailed source of information of inventive activity (oecd, 2009). therefore, this research work focuses on analyzing patents statistics of additive manufacturing technologies, which are classified as medical science. 2. objective the main objective of this research work is to achieve an overall perspective of the technological trends of additive manufacturing technologies applied in medical applications such as, prostheses and orthopaedic devices, among others. specifically, a worldwide patent study has carried out, in order to analyse the invention performance, the technological fields and the patent value. 3. methods the process to accomplish the main objective of the work starts in the planning stage, which stablished the identification of information sources, such as patent database sources. the database used in the study has been patseer, which is a complete online global patent database and research platform containing the world’s most comprehensive fulltext patent collection (sinha & pandurangi, 2016). after that, the specific queries for database are defined to collect information (see table 1). the definition of the appropriate query is very important due to the fact that the additive manufacturing technologies collect many synonyms terms and the medical applications cover a very large area of action that must be shortened. thus, the query of am were tested by collecting information from a technical report about am done by experts (gridlogics, 2014), and the medical applications has been focused through the international patent classification (ipc), that constitutes a first reference for identifying patent in a specific technical domain (oecd, 2009). ipc provides us the classification of patents and utility models according to the different areas of technology to which they pertain. thus, the interest of the research work has defined a61f2 ipc, related to filters implantable into blood vessels; prostheses, i.e. artificial substitutes or replacements for parts of the body; appliances for connecting them with the body; devices providing patency to, or preventing collapsing of, tubular structures of the body. the patents have been studied by families, because patent families are a way of working out patent indicators that are comparable across countries. a patent family comprises all patents protecting the same invention (oecd, 2009). the data have been analyzed through the analytics package of patseer called quick stats. besides, in order to identified what the main knowledge areas about additive manufacturing technologies applied in medical applications are, a clusterization analysis has been carried out using vosviewer software. int. j. prod. manag. eng. (2019) 7(special issue), 65-72 creative commons attribution-noncommercial-noderivatives 4.0 international alvarez-meaza et al. 66 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. results 4.1. monitoring inventive performance of the patents in order to reflect inventive performance, the earliest priority date of the patents is analysed because it can be considered the closest to the invention date. in the figure 1, it is shown that the 2015 year is the most productive in inventions related to am in medical services. it is useful the inventor’s country of residence to compile patent statistics aimed at reflecting inventive activity. it allow us analysing the market allocation strategy of companies. the main country is united state of america with 197 number of records, followed by germany (37), united kingdom (26) and china (26) somewhat behind (see figure 2). the holder of the legal rights and obligations on a patent application is the applicant; in the united states, that it is called the assignee. it can be an individual, a company, a university, a hospital or a government entity. regarding am technologies in medical applications, the holders with the most number of records are companies from europe, china and usa, followed by universities from china (see figure 3). 4.2. the identification of technological fields the information provided by the international patent classification (ipc) constitutes a first reference for table 1. main information of patent search. information data query tac:((3d or 3 d or 3-d or 3-dimension* or 3 dimension* or (three* w2 dimension) or desktop* or additive) wd2 (print* or fabricat* or manufactur*)) and (icgr:(a61f2*) or ic:( a61f2*) or cpc:( a61f2*)) and not (tacd:(stereoscopic* or oxidation product* or streaming interactive or nanoweb or nano web or nanofiber* or nanofibre* or nano fiber* or nano fibre* or nanometer fiber* or nanometer fibre* or nanometre fiber* or nanometre fibre* or non halogen or non-halogen or media access control or multi-wafer 3d cam cell or ((foof* or feed* or liquid*)w2 additive*) or seed culture or antibacteria* or 3-sigma or three sigma or rheolog* additive* or vibration isolator*)) database patseer timespan from 1994 to 2017 type single family of patents date of the search february 23, 2018 results 767 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 250 200 150 100 50 0 total n o. o f f am ili es 11 9 25 29 74 113 223 185 68 figure 1. earliest priority date. int. j. prod. manag. eng. (2019) 7(special issue), 65-72creative commons attribution-noncommercial-noderivatives 4.0 international patentometric: monitoring the scientific and technological trends of additive manufacturing in medical applications 67 http://creativecommons.org/licenses/by-nc-nd/4.0/ identifying patents in a specific technical domain (oecd, 2009). the main result of the analysis is related to the core ipc that the research study takes in account, a61f. but it is an interesting results that other technological fields are also linked to the additive manufacturing technology patents based on medical applications, such as, b33 additive manufacturing, b29 working of plastics, g06 computing (image data processing and electric digital data processing) and g05 control of regulating systems in general, among others (see figure 4). the number of technical classes attributed to a patent application has been used as an indicator for the scope, and hence for the value of the patent (oecd, 2009). lerner (lerner, 1994) finds a positive correlation between the firm’s market value and the average scope of its patents. 4.3. the patent value in this research, the patent value is analysed by forwards citations of the patent, the patent family size and the number of inventor in a patent. the prior art of the invention cited in a patent documents provides useful information about the diffusion of technologies. the forward citation as indicator of a patent value show the technological importance of the invention and the impact on further developments (oecd, 2009). the analysis of the data shows that most have about eight number of forward citations, but two of them have an important difference with the rest. specifically, figure 2. investor’s country. zhou huixing univ xi an jiaotong univ zhejiang univ pekin 3rd mat co ltd univ south china biomet mfg corp li peng shenzhen yiheping co materialise nv tech 18 16 14 12 10 8 6 4 2 0 total n o. o f f am ili es 5 6 6 7 8 8 9 10 11 16 figure 3. current assignee. int. j. prod. manag. eng. (2019) 7(special issue), 65-72 creative commons attribution-noncommercial-noderivatives 4.0 international alvarez-meaza et al. 68 http://creativecommons.org/licenses/by-nc-nd/4.0/ the us2010291401a1 with 40 citations and the us2014107628a1 with 34 citations (see figure 5). the value of patens is also associated with the geographical scope of patent protection; that is, with the patent family size it can approximate the cost to have protection in different jurisdictions and the sign of market potential of an invention (oecd, 2009). the biggest family have 30 members (see figure 6) the patent au2010284197b9, followed by 22 and 18. but, in general, the number of members is in the range of 6 to 12. the number of inventors may proxy the cost of the research behind the invention (oecd, 2009). in general, the average number of inventors is around 10, but the patent ru2612528c1 with 18 inventors stands out (see figure 7). figure 4. international patent classification (ipc). 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 us2010291401a1 no. of forward citations us2014107628a1 wo2007045000a2 cn101032430a kr20120088928a wo2010120990a1 cn103584930a us2014257518a1 cn103584931a wo2012113030a1 cn103860293a us2014172116a1 us8790408b2 cn103284815a cn203829093u us2015328005a1 wo2015023077a1 figure 5. most cited records. int. j. prod. manag. eng. (2019) 7(special issue), 65-72creative commons attribution-noncommercial-noderivatives 4.0 international patentometric: monitoring the scientific and technological trends of additive manufacturing in medical applications 69 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.4. visualization networks the global research activity in additive manufacturing technologies applied in medical applications is identified through knowledge areas that have been determined depending on the occurrence of keywords. this clusterization is carried out when the terms have a high co-occurrence, and the created clusters represent a singular concept related to additive manufacturing technologies applied in medical applications. relating to patents, the clusters were created by extracting keywords from the title. the co-occurring network of additive manufacturing technologies applied in medical applications of the 100 most cited keywords of patents generated by vosviewer is represented in figure 8. the intellectual landscape shows that “method”, “system”, “manufacture”, “implant” and “same” are the top keywords. clustering terms allows us to know the affinity between the keywords, and thus to know which are the thematic fields of the different terms. for example, “manufacture” is related to “am process”, “prosthetic ear”, “bone implant” and “implantable medical device”. in this process 27 clusters have been created, most of them small, and a total of 238 links, with the term “method” having a total of 65 links associated with it. it is therefore a very common term in patents relating to additive manufacturing. 5. conclusions additive manufacturing in medical applications are relatively recent technological developments that are in full growth. the enterprises are the main holders of the legal rights and the core invention activity is located in usa, which is way ahead of germany, uk and china. many patents cover different technological fields what strengthens the firm’s market value. the number of forward citations are clearly there to justify the diffusion of 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 jp2012514488a no. of family members kr20160091944a uss782286a wo2011080953a1 ca2928070a1 gb201504280d0 jp2018501845a wo2016094298a1 ca2946389a1 wo2010063993a2 wo2016026021a1 jp2004524090a ca2839960a1 ca2782117a1 us2015051886a1 ca2859662a1 au2010284197b9 figure 6. largest families. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 au2012220362a1 no. of inventors ca2934405a1 cn103501731a cn105380728a cn105640679b cn106943215a ep3270821a kr101821746b1 kr20140016908a ru2015104291a us2015223939a1 us2017057169a1 us2018042718a1 wo2012113030a1 wo2016154148a1 wo2017214432a1 us2016303804a1 kr20160139684a cn106618812a cn106626374a ru2621874c2 ru2637602c1 jp2017038918a wo2016200201a1 ru2612528c1 figure 7. max. no. of inventors. int. j. prod. manag. eng. (2019) 7(special issue), 65-72 creative commons attribution-noncommercial-noderivatives 4.0 international alvarez-meaza et al. 70 http://creativecommons.org/licenses/by-nc-nd/4.0/ these technologies. in regards to the cost to have protection in different jurisdictions and the cost of the invention, it can be concluded that most of the patents, except for a couple of them, usually follow a very similar trend. so, the cost is not a differentiating milestone. the keywords extracted from the title have allowed us to know which is the field of knowledge that the patents are working. the clustering process of the 100 most used keywords results in “method”, “system” and “manufacture” being the most frequent terms and with the highest number of links to other terms. finally, future works related with this study could take other topics of investigation, such as, the analysis of knowledge diffusion and the dynamics of technical change, the economic value of the inventions and the role of university in technological development, among others. references astm. (2015). iso / astm52900-15, standard terminology for additive manufacturing – general principles – terminology. west conshohocken, pa: astm international. atzeni, e., and salmi, a. (2012). economics of additive manufacturing for end-usable  metal parts. the international journal of advanced manufacturing technology, 62(9-12), 1147-1155. https://doi.org/10.1007/s00170-011-3878-1 galati, m., and iuliano, l. (2018). a literature review of powder-based electron beam melting focusing on numerical simulations. additive manufacturing, 19, 1-20. https://doi.org/10.1016/j.addma.2017.11.001 gibson, i., rosen, d. w., and stucker, b. (2010). additive manufacturing technologies. ny: springer. https://doi.org/10.1007/978-1-4419-1120-9 figure 8. the visualization map of the co-occurring keywords of patents. int. j. prod. manag. eng. (2019) 7(special issue), 65-72creative commons attribution-noncommercial-noderivatives 4.0 international patentometric: monitoring the scientific and technological trends of additive manufacturing in medical applications 71 https://doi.org/10.1007/s00170-011-3878-1 https://doi.org/10.1016/j.addma.2017.11.001 https://doi.org/10.1007/978-1-4419-1120-9 http://creativecommons.org/licenses/by-nc-nd/4.0/ gridlogics. (2014). 3d printing: an analysis of patenting activity around 3d-printing from 1990-current gridlogics technologies pvt ltd. hopkinson, n., and dickens, p. (2006). emerging rapid manufacturing processes. rapid manufacturing (pp. 55-80) john wiley & sons, ltd. https://doi.org/10.1002/0470033991.ch5 jin, y., ji, s., li, x., and yu, j. (2017). a scientometric review of hotspots and emerging trends in additive manufacturing. journal of manufacturing technology management, 28(1), 18-38. https://doi.org/10.1108/jmtm-12-2015-0114 lerner, j. (1994). the importance of patent scope: an empirical analysis. rand journal of economics, 25(2), 319-333. mellor, s., hao, l., and zhang, d. (2014). additive manufacturing: a framework for implementation. international journal of production economics, 149, 194-201. https://doi.org/10.1016/j.ijpe.2013.07.008 oecd. (2009). oecd patent statistics manual oecd. petrovic, v., haro, j. v., jordá, o., delgado, j., blasco, j. r., and portolés, l. (2011). additive layered manufacturing: sectors of industrial application shown through case studies. international journal of production research, 49(4), 1061-1079. https://doi.org/10.1080/00207540903479786 rodriguez-salvador, m., rio-belver, r. m., and garechana-anacabe, g. (2017). scientometric and patentometric analyses to determine the knowledge landscape in innovative technologies: the case of 3d bioprinting. plos one, 12(6), 1-22. https://doi.org/10.1371/journal.pone.0180375 sinha, m., and pandurangi, a. (2016). guide to practical patent searching and how to use patseer for patent search and analysis. india: gridlogics technologies pvt. ltd. zarrabeitia, e., bildosola, i., rio-belver, r. m., alvarez, i., and cilleruelo, e. (2017). laser additive manufacturing: a patent overview. lect. notes in managem. industrial engineering engineering digital transformation. springer. https://doi.org/10.1007/978-3-319-96005-0_23 int. j. prod. manag. eng. (2019) 7(special issue), 65-72 creative commons attribution-noncommercial-noderivatives 4.0 international alvarez-meaza et al. 72 https://doi.org/10.1002/0470033991.ch5 https://doi.org/10.1108/jmtm-12-2015-0114 https://doi.org/10.1016/j.ijpe.2013.07.008 https://doi.org/10.1080/00207540903479786 https://doi.org/10.1371/journal.pone.0180375 https://doi.org/10.1007/978-3-319-96005-0_23 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering special issue: advances in engineering networks https://doi.org/10.4995/ijpme.2019.11458 received: 2019-03-04 accepted: 2019-04-29 a multicriteria decision model for the evaluation and selection of technologies in a r&d centre lizarralde, r.a and ganzarain, j. b a mondragon unibertsitatea, loramendi, 4. apartado 23 20500 arrasate mondragón. spain b ik4-ideko, arriaga industrialdea 2, e-20870 elgoibar-gipuzkoa, spain a jganzarain@mondragon.edu, b rlizarralde@ideko.es abstract: r&d centres play a key role in the technology development process of industries, and therefore in their competitive strategies. they have responsibility in the identification, selection, acquisition, development and transfer of technology. among these activities, the successful selection of new technologies is becoming a highly critical and complex challenge in the technology management process. the problem of succeeding in the selection of new technologies is, from the methodological side, linked to heterogeneous key factors (technological, economic, human, and organisational). many approaches deal with it by means of multiple criteria decision making (mcdm) techniques and tools. nevertheless, most of the works are related to the selection of technologies in industrial cases and very few works have been found in the bibliography related to r&d institutions and, in particular, technological centres. a model for the evaluation and decision about one or several technologies based on the mives (modelo integrado de valor para evaluaciones de sostenibilidad) method is proposed. introducing the motivations for using this method, after a review of the most used mcdm methods, and describing the structure of the model and the preliminary key parameters and relations among them. the proposed model is oriented to it’s application in the manufacturing sector, observing the particularities of the sector in the selection of the critical factors related to technology, r&d centre and industry. key words: technology selection, r&d centre, mcdm, mives, technology management. 1. introduction the main objective of this research work is to develop a methodology for the evaluation and selection of new technologies in a r&d centre. the need of a special approach for the case of r&d centres is based on the position and role of these entities in the technology development process chain. the key premise is that the success in the selection, and further development of a technology, is not only dependant on the characteristics of the technology itself and on the capabilities of the centre, but also, in a significant proportion by the factors related to the industrial receptors of the technology and the relationships between the r&d centre and them. the proposed model tries to provide an integrated approach in which the critical factors of the technology, the r&d centre and its potential industrial customers (final receivers of the technology) and their interrelations, are evaluated. the model must provide the option of selecting among several candidate technologies, for its application during strategic planning processes and also for making the decision in the adoption of a single technology, to be applied in daily, non-systematic situations. this paper introduces the relevance of the addressed to cite this article: lizarralde, r., and ganzarain, j. (2019). a multicriteria decision model for the evaluation and selection of technologies in a r&d centre. international journal of production management and engineering, 7(special issue), 101-106. https://doi.org/10.4995/ijpme.2019.11458 int. j. prod. manag. eng. (2019) 7(special issue), 101-106creative commons attribution-noncommercial-noderivatives 4.0 international 101 https://orcid.org/0000-0002-6144-636x http://creativecommons.org/licenses/by-nc-nd/4.0/ problem, based on a deep literature review, makes a revision of the most applied decision making methods, and selects mives (modelo integrado de valor para evaluaciones de sostenibilidad) method as the basic tool for development. after this selection, the basic structure of the model is presented with the proposed application methodology. the key paramenters that configure the model are selected based on their significance in the literature review. further refinemet must be perfomed, as introduced in the methodology description. 2. technology development and r&d centres research, development and innovation activities are considered a key factor to ensure competitiveness and sustainable socioeconomic development (phaal et al., 1995). this significance is observed both in the most developed countries and in developing countries or emerging economies, (guan et al., 2009). the process of technology maturation is being conditioned by the speeding up of industrial and economical activities, requiring significantly shorter development times and maximum effectiveness and efficiency in the complete development chain, form the identification to the implementation (clausen et al., 2013). this concern is also noticeable in the public administration policies, in the case of european commission support programs (h2020), under the epigraph “valley of death”, fostering strategies and instruments to close the gap between research and market-society (de la concha, 2014). 2.1. the role of r&d centres r&d centres are present almost in all the stages of the technology development chain: starting with the fundamental research, in collaboration with universities, proceeding to the development and demonstration of technology suitability in labs and pre-industrial prototypes, up to the transfer to their industrial partners, making use of different mechanisms: collaboration projects, licensing, generation of start-ups, transfer of researchers. this position is key for the technology evaluation and selection strategies and activities of the centres, making necessary an integral evaluation of the environment factors to ensure the suitability and success potentials of a technology. in the method presented in this paper this leads to an integrated three areas evaluation. figure 1. basic scheme of interrelations in the technology selection model. 2.2. the manufacturing sector although the proposed selection model is intended to be applicable in any technological and application sector, manufacturing sector has been particularly observed in the selection of key parameters that characterize the scenario (represented in figure 1). the importance of manufacturing sector has been widely addressed in the last decades, playing a fundamental role in the generation, support and traction of economic and social development (dobbs et al., 2012). the direct relation between the economic development and the weight of the sector in the gdp of the countries has been reported, both for developed countries (kaldor, 1967) and developing countries (felipe et al., 2014). this prominence varies as the economic growth of a region evolves, from the quantitative to a more qualitative role. in developed mature countries, other economic sectors such as services take the leading position, being the upper limit of the manufacturing weight in the gdp establish in a band between 25 and 35 %. however, even in that situation manufacturing is key actor, providing the equipment and technologies needed to keep the leading position of the advanced economies, and sustaining challenges in social and environmental areas. public administrations are fostering policies, programmes and instruments to support an advanced manufacturing sector as the motor for the current and future social competitiveness and sustainability, positioning the technological development as the cornerstone of these strategies. 3. technology selection 3.1. technology management assumed the relevance of the technology management process for the competitiveness of int. j. prod. manag. eng. (2019) 7(special issue), 101-106 creative commons attribution-noncommercial-noderivatives 4.0 international lizarralde and ganzarain 102 http://creativecommons.org/licenses/by-nc-nd/4.0/ industries and societies, gregory (1995) formulated one of the most accepted definitions of the process: “technology management addresses the effective identification, selection, acquisition, development, exploitation and protection of technologies needed to maintain a market position and business performance in accordance with the company’s objectives”, setting up the five main activities that configure the process. figure 2. technology management process phases (gregory, 1995). moreover, the process of technology management is not understood as an isolated activity, but as an integrated process in the strategy of any company, in an integrated view with strategic and operative aspects (phaal et al., 1998). also in the phase of technology evaluation and selection, these integral criteria should be observed. from these five activities this work is focused in the technology selection as one of the most critical, in particular for a mature sector, such as manufacturing, and at the same time is one of the activities least systematically approached in r&d institutions. the literature review contains numerous works about industry and public administrations, but very few about r&d centres. 3.2. technology selection and mcdm methods technology selection is been a matter of numerous research works, resulting in families of methods, in some cases combining the identification and selection activities. table 1 collects some of the most representative families of methods and tools: table 1. technology selection methods. method description and references cost benefit analysis applied for selecting particular technologies in industry. technology pyramid value (tvp) (tipping et al., 1995), strategic technology assessment review (star) (mcgrath and macmillan, 2000), system wide benefits value analysis (swbva) (ordoobadi and mulvaney, 2001) , technology balance sheet (hartman, 1999) impact analysis very used in the evaluation of technological areas, within the strategies of public administrations and large industrial and technological corporations. among the tools used are cross-impact analysis, delphi (dalkey and helmer, 1963), screening and positioning models, integrated impact assessment, ethical technology assessment. analysis of scenarios widely adopted in different management fields. diffenbach (1981) developed a three-step approach: formulation, scenario compatibility and compatibility assessment. winebrake and creswick (2003) combined ahp and perspective based scenario analysis (pbsa). banuls and salmeron (2006) proposed the scenario based assessment model (sbam) which combines ahp, cross-impact method (cim), and delphi. roadmapping management and planning tool. developed by motorola to improve the alignment between technology and innovation (willard and mcclees, 1987). extended to other large companies such as phillips, royalmail, general motors, lockheed martin, erickson and british telecom. phaal, farrukh and probert (2000) estimated that 10% of manufacturing companies had used the technique. surveillance, monitoring and evaluation methods patent analysis is widely used as a parameter for monitoring the impact of a technology (slowinski et al., 2000, grimaldi eta al., 2015). rohrbeck et al. (2006) developed the technology radar, for the evaluation and selection of emerging technologies in three phases of progressive filtering. multiple criteria decision models (mcdm) applied when the key factors are heterogeneous in nature and dimension. analytical hierarchy process (ahp), (saaty, 1980) is the most extended, and enabler to develop adapted evolutions, such as anp (saaty, 1996) and mives (viñolas et al., 2009). other methods are dea (data envelopment analysis) (charnes et al., 1978), topsis (technique for order of preference by similarity to ideal solution) (chen et al., 1992), electre (elimination et choix traduisant la realite) (roy, 1968), promethee (preference ranking organization method of enrichment evaluation) (brans, 1982). int. j. prod. manag. eng. (2019) 7(special issue), 101-106creative commons attribution-noncommercial-noderivatives 4.0 international a multicriteria decision model for the evaluation and selection of technologies in a r&d centre 103 http://creativecommons.org/licenses/by-nc-nd/4.0/ the mives method, introduced in table 1, is one of the youngest mcmd methods. it was developed and applied specially in building sector. methodologically it is based on a combination of ahp and delphi, and one of its main attributes lays in the special value functions. these value functions, which can be defined individually for each key factor, provide a better, more accurate characterisation of each one of the factors, improving the power of the methodology. this characteristic and the novelty of applying it in manufacturing sector are the motivations to select it for this work. therefore, mives will be the basis for the development of the technology selection method, with special emphasis in the definition of the proper value functions. 4. the proposed methodology and model for evaluation and selection of technology the model proposed for the evaluation and selection of technologies in a r&d centre is based on the application of mives method. the methodology, following mives method, is structured in these steps: 1. selection of critical factors and construction of the decision or hierarchy tree. this activity is carried out in two steps: in the initial one, the factors are selected from the literature review, in the second those factors are refined by the experts working group created for the generation of the model. 2. creation of value functions to rate the critical factors in the lower level of the tree. 3. factors weighting: the relative importance of each factor is assigned in relation to the others in the same level. 4. evaluation of the alternatives (alternative technologies to be evaluated or “adopt-reject” in case of a single technology evaluation), obtaining the value index for each one. 5. sensitivity analysis to evaluate the reliability of the model. as mentioned, the model is based in ahp methodology, combined with working tools such as delphi for the refining and weighting of factors. for this purpose a two level human working group is configured: a core group for the detailed work composed by experts from ik4-ideko and geographically close r&d centres and industrial partners, and an extended group of international relevant contact persons, for some eventual contrasts. 4.1. preliminar results the preliminary version of the decision tree has been constructed. the selection of factors in this preliminary version is based on the extensive literature review about technology selection, especially in manufacturing sector, but not restricted to that sector. as mentioned, few works have been detected figure 3. hierarchy tree, based on the bibliography. int. j. prod. manag. eng. (2019) 7(special issue), 101-106 creative commons attribution-noncommercial-noderivatives 4.0 international lizarralde and ganzarain 104 http://creativecommons.org/licenses/by-nc-nd/4.0/ about r&d centres, but the papers related to industrial companies, universities and public institutions provided very useful information, with complementary approaches and criteria. the tree is structured in three areas: internal factors of the technologi(es) to be evaluated, characteristics of the r&d centre and characteristics of the potential industrial receivers and implementers of the technology, as shown in figure 3. 5. conclusions and further work technology selection is one of the key processes in technology management. numerous research works can be found about the development and application of techniques to approach technology selection in industry, with a variety of orien-tations: strategy, impact, value chain, collaborations, capabilities and skills. r&d centres, due to their positioning and role in the technology development chain, need to adopt an integral approach to the selection problem, including in their analysis not only their own internal factors but also the parameters related to the technology final receivers, their industrial partners or customers. a mcdm methodology (mives) has been selected to tackle the challenge of developing a method for the selection of technology in a r&d centre. preliminary schemes of the model and hierarchy tree have been defined, based on the literature review. in the ongoing development phase, this tree will be refined, detailed and completed by the work of an expert’s panel and finally the model will be completed with a quantification tool for the selection outputs. final validation of the methodology by appliying it to use cases will also be carried out. references banuls, v.a., and salmeron, j.l. (2006). a scenario-based assessment model sbam. technological forecasting and social change, 74(6), 750-762. https://doi.org/10.1016/j.techfore.2006.05.015 brans, j.p., and vincke, p. (1985). note – a preference ranking organisation method. management science, 31(6), 647-656. https://doi.org/10.1287/mnsc.31.6.647 charnes, a., cooper, w.w., lewin, a., seiford, l.m. (1994). data envelopment analysis: theory, methodology and applications. massachusetts: kluwer academic publishers. https://doi.org/10.1007/978-94-011-0637-5 chen, s., hwang, c., beckmann, m., and krelle, w. (1992). fuzzy multiple attribute decision making: methods and applications. new york: springer verlag. https://doi.org/10.1007/978-3-642-46768-4_5 clausen, t.h., korneliussen, t., and madsen, e.l. (2013). modes of innovation, resources and their influence on product innovation: empirical evidence from r&d active firms in norway. technovation, 33(6-7), 225-233. https://doi.org/10.1016/j.technovation.2013.02.002 dalkey, n. c., and helmer, o. (1963). an experimental application of the delphi method to the use of experts. management science, 9(3), 458-467. https://doi.org/10.1287/mnsc.9.3.458 de la concha, s. (2014). horizonte2020, de la visión general a las oportunidades. xx y xxi seminarios de asesores de proyectos de i+d de la unión europea. diffenbach, j. (1981). a compatibility approach to scenario evaluation. technological forecasting and social change, 19(2), 161-174. https://doi.org/10.1016/0040-1625(81)90013-5 dobbs, r., manyika, j., and roxburh, c. (2012). manufacturing the future: the next era of global growth and innovation. mckinsey global, (november). felipe, j., aashish, m., and changyong, r. (2014). manufacturing matters... but it’s the jobs that count. asian development bank working paper no. 420. gregory, m.j. (1995). technology management a process approach. proceedings of the institution of mechanical engineers, 209(5), 347356. https://doi.org/10.1243/pime_proc_1995_209_094_02 grimaldi, m., cricelli, l., giovanni, m. di, and rogo, f. (2015). the patent portfolio value analysis: a new framework to leverage patent information for strategic technology planning. technological forecasting and social change, 94, 286-302 https://doi.org/10.1016/j.techfore.2014.10.013 guan, j.c., yam, r.c.m., tang, e.p.y., lau, a.k.w. (2009). innovation strategy and performance during economic transition: evidences in beijing, china. research policy, 38(5), 802-812. https://doi.org/10.1016/j.respol.2008.12.009 hartmann, m.h. (1999). theory and practice of technological corporate assessment. journal of engineering and technology management, 17(4), 504-521. https://doi.org/10.1504/ijtm.1999.002730 int. j. prod. manag. eng. (2019) 7(special issue), 101-106creative commons attribution-noncommercial-noderivatives 4.0 international a multicriteria decision model for the evaluation and selection of technologies in a r&d centre 105 https://doi.org/10.1016/j.techfore.2006.05.015 https://doi.org/10.1287/mnsc.31.6.647 https://doi.org/10.1007/978-94-011-0637-5 https://doi.org/10.1007/978-3-642-46768-4_5 https://doi.org/10.1016/j.technovation.2013.02.002 https://doi.org/10.1287/mnsc.9.3.458 https://doi.org/10.1016/0040-1625(81)90013-5 https://doi.org/10.1243/pime_proc_1995_209_094_02 https://doi.org/10.1016/j.techfore.2014.10.013 https://doi.org/10.1016/j.respol.2008.12.009 https://doi.org/10.1504/ijtm.1999.002730 http://creativecommons.org/licenses/by-nc-nd/4.0/ kaldor, n. (1967). strategic factors in economic development. ithaca: new york state school of industrial and labor relations, cornell university. mcgrath, r.g., macmillan, i.c. (2000). assessing technology projects using real options reasoning. research-technology management, 43(4), 35-49. https://doi.org/10.1080/08956308.2000.11671367 ordoobadi, s.m., mulvaney, n.j. (2001). development of a justification tool for advanced manufacturing technologies-value analysis. journal of engineering and technology management, 18(2),157-184. https://doi.org/10.1016/s0923-4748(01)00033-9 phaal, r., paterson, c.j., and probert, d.r. (1998). technology management in manufacturing business: process and practical assessment. technovation, 18(8-9), 541-589. https://doi.org/10.1016/s0166-4972(98)00026-1 phaal, r., farrukh, c.j.p., and probert, d.r. (2001). a framework for supporting the management of technological innovation. engineering department, university of cambridge, 1-14. rohrbeck, r., heuer, j., and arnold, h. (2006). the technology radar an instrument of technology intelligence and innovation strategy. icmit 2006 proceedings 2006 ieee international conference on management of innovation and technology, 2(september 2017), 978-983. https://doi.org/10.1109/icmit.2006.262368 saaty, t.l. (1980). the analytic hierarchy process. new york: mcgraw-hill. https://doi.org/10.21236/ada214804 saaty, t.l. (1996). decision making with dependence and feedback, the analytic hierarchy process, (rs publications: pittsburgh, pa). slowinski, g., stanton, s.a., tao, j.c., miller, w., and mcconnell, d.p. (2000). acquiring external technologies, res. technol. manag., 43(5), 29-35. https://doi.org/10.1080/08956308.2000.11671378 tipping, j.w., zeffren, e., fusfeld, a.r. (1995). assessing the value of your technology, res. technol. manag., vol. 38(5), 22-39. https://doi.org/10.1080/08956308.1995.11674292 viñolas, b., cortés, f., marques, a., josa, a., and aguado, a. (2009). mives: modelo integrado de valor para evaluaciones de sostenibilidad. ii congrés internacional de mesura i modelització de la sostenibilitat, 1-24. willyard, c.h., mcclees, c.w. (1987). motorola’s technology roadmap process, res. manag., 30(5), 13-19. https://doi.org/10.1080/0034 5334.1987.11757057 winebrake, j.j., creswick, b.p. (2003). the future of hydrogen fueling systems for transportation: an application of perspectivebased scenario analysis using the analytic hierarchy process, technological forecasting and social change, 70(4), 359-384. https://doi.org/10.1016/s0040-1625(01)00189-5 int. j. prod. manag. eng. (2019) 7(special issue), 101-106 creative commons attribution-noncommercial-noderivatives 4.0 international lizarralde and ganzarain 106 https://doi.org/10.1080/08956308.2000.11671367 https://doi.org/10.1016/s0923-4748(01)00033-9 https://doi.org/10.1016/s0166-4972(98)00026-1 https://doi.org/10.1109/icmit.2006.262368 https://doi.org/10.21236/ada214804 https://doi.org/10.1080/08956308.2000.11671378 https://doi.org/10.1080/08956308.1995.11674292 https://doi.org/10.1080/00345334.1987.11757057 https://doi.org/10.1080/00345334.1987.11757057 https://doi.org/10.1016/s0040-1625(01)00189-5 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.11619 received: 2019-04-06 accepted: 2019-11-23 use of the abc curve in medicine line balancing: a case study at a brazilian pharmaceutical distribution center oliveira, n.p. a1*, prado, d.g.o. b1, jesus, r.h.g.b2 a université de bordeaux, france. b technological federal university of paraná, brazil. a1 nathanpeixot@yahoo.com.br, b1 danigoprado@hotmail.com b2 romulohenriquegomes@hotmail.com abstract: within the logistics, especially in distribution centers, there is still use of manual activities, with great possibility of gaining productivity. this work aims to improve efficiency in the supply and separation stages by balancing drug lines using the abc classification. for this, a case study was carried out in a large brazilian retail pharmaceutical network, in which products of higher output were relocated in more strategic areas. ergonomic gains, reduction of required manpower, greater assertiveness, and savings of two hours of work (25%) were the main results achieved. the study also made it possible to disseminate an accessible, automated and effective tool for line balancing on large scale inventories. key words: abc curve, line balancing, logistics. 1. introduction even today, in the national logistics of large companies, the use of manpower is used to carry out innumerable activities. because of this, it has the opportunity to influence the productivity, bringing numerous improvements. special attention is given to the supply and separation process in a distribution center where, due to the lack of organization and because there are products with a high demand variation –as the example of flu remedies or sunscreens that are sold seasonally– they end up loosing considerable efficiency. this work intends to carry out a classification of the products according to their outputs, in order to better organize them in the supply and separation lines at a distribution center of an eminent brazilian pharmaceutical retailer network. for this, the methodology of the case study will be used, where the curves (a, b or c, according to the abc curve concept) and locations of each product will be defined initially, and in a second moment, they will be rearranged. the obtained results were not only the gain of efficiency, but the improvement in work ergonomics, since more accessed products were positioned strategically for the operators’ use. the contribution of this study is due to the use of a simple, accessible and effective tool for line balancing in large scale warehouses with the premise of pareto analysis or abc curve. in addition, the work proved to be valuable when applied successfully in a highly complex sector such as medicines, with high diversification and quantity of existing products. 2. theoretical foundation in this section will be addressed the key business of the article, the logistics area. where it will be limited and deepened to the distribution center field of study. finally, will be seen the the pareto analysis, the instrument used in the case study. to cite this article: oliveira, n.p., prado, d.g.o., jesus, r.h.g. (2020). use of the abc curve in medicine line balancing: a case study at a brazilian pharmaceutical distribution center. international journal of production management and engineering, 8(1), 13-19. https://doi.org/10.4995/ijpme.2020.11619 int. j. prod. manag. eng. (2020) 8(1), 13-19creative commons attribution-noncommercial-noderivatives 4.0 international 13 https://orcid.org/0000-0002-1082-8499 https://orcid.org/0000-0003-2802-6037 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2.1. business logistics according to the author ballou (2013), logistics can be defined as the process of transit and storage of materials, so as to allow them to be drained from a starting point to their final destination. in addition, information flows that ensure satisfactory levels of service at relevant costs are included. regarding its objective, the author defines as the need to deliver objects or services in the physical, temporal and geographical condition desired by the client. the big challenge, however, is to balance its supply and demand in order to avoid logistical costs. these tend to vary around 19% to 22% of net sales, however, it can undergo strong variations in the market (ballou, 2013). these costs can be divided into operational and administrative costs, in which the first is directly linked to distribution and the second is not (lancioni, 1991). viana (2008) cites as improvement of the logistic efficiency the reduction of the paths traveled by the goods and a better use of its volumetric capacity. 2.2. distribution center to a large extent, the distribution center is divided into input, stock, and output processes. and for this, it needs to be managed administratively bureaucratic and control steps and operationally physical stages (moura, 1997). in figure 1 the following summary is presented. 1-receipt 2-unpacking and inspection / unitization 3-send to stock 4-stock 5-orders separation 6,7-checkout and packing 8-expedition receipt control stock control expedition control suppliers company users input stock output physical activities bureaucratic procedures flow of materials documentation relationship organizational division figure 1. distribution center (source: adapted from moura (1997)). after being previously identified, it can be described its three types of processes as follows: a) input processes: viana (2008) describes how the step of receiving goods and other correlated activities such as unpacking, checking, returning or authorizing the storage of products, as can be seen in figure 2. it connects directly with the purchase and payment stages of suppliers; figure 2. input processes (source: author). b) stock processes: viana (2008) calls storage or stock, the process of maintaining materials, adapting to their physical constitution and conditioning them to a stock management until its use is necessary. ballou (2013) points out that raw material, semi-finished, finished materials, spares and administrative materials participate in this stage. figure 3 shows the main stock processes. among their main functions, gu et al. (2007) highlight the equilibrium in the transit of materials in the supply chain, reducing the risks of variability caused by demand (seasonality) and production. in addition, it enables the combination of products, making orders more flexible and reducing transportation costs. ballou (2013) estimates that warehousing costs in the industry revolve in the range of 12 to 40% of logistics costs. as a way to enhance efficiency in inventory management, the products are unitized on pallets or containers and stored at fixed or random addresses. in the first case, it is a bay or shelf for a particular product. its negative point is the possibility of emptying due to the inexistence of products and the necessity of maximum inventory volume. for those of variable location, occupancy is gained because it uses average inventory and management is lost, since the same product can occupy different positions in the stock (ibidem). according to ballou (2006), after being stored, the picking process or orders separation is guaranteed by an intensive labor use, with a high level of product mansonicity. that is why they are constantly aim of improvement, giving scope for considerable int. j. prod. manag. eng. (2020) 8(1), 13-19 creative commons attribution-noncommercial-noderivatives 4.0 international oliveira et al. 14 http://creativecommons.org/licenses/by-nc-nd/4.0/ operational improvements and productivity gains. among them, it can be cited the handling by: orders: the way the incoming orders are handled has the opportunity to reduce costs, lead time and workforce. as an example, it can be generated a picking list based on the purchase order; sequencing or sorting: is the arrangement of a set of products based on transpotations routes so they can be efficiently separated. this will also save time, avoiding going back after products to pick up goods. altghout, sequencing items based on the selling process may require the cooperation of sales staff and customers so that it could be possible to relate items in the most likely order of occurrence. alternatively, powerpul softwares can be used to sequence the items into efficient separation lists, saving time or setbacks in the line; zones: process in which operators will be selected for a delimited area of the stock; order splitting: derived from the zoning process, happens when the stock is in more than one location; lots: combining the products of more than one order at a single time; intercalation: separation and storage take place at the same time and pattern setting: labor performance models used as benchmarks for productivity measurement. c) output processes: regardless of the type of separation used, the conference or checkout stage will check for errors and trigger its correction. next, the product is packed, proceeding to the expedition, where the documentation is provided and the trucks are loaded (pedreira, 2006). the loading stage is influenced by customer orders, nature of the loaded material (solids, liquids, gases), conditioning (on pallets, in bulk), type of dock (high, low), type of loading (pallets, platforms) and transport vehicle (mauro, 2009). figure 4 shows the main exit processes. 2.3. abc curve due to the high complexity of the production system of managing hundreds or thousands of products (vollman et al., 2005). and according to van kampen et al. (2012) and aktunc et al. (2019), when it comes to inventory organization, the analysis of pareto or abc curve is one of the most prevalent and widely used prioritization methodology. it takes into account that the lack of different products generates different levels of impact due to their degree of importance (slack et al., 2015). according to the authors, products with high output generate greater dissatisfaction due to their lack, as well as those with high added value concentrate 3send to stock 5orders separation4stock figure 3. stock processes (source: author). figure 4. output processes (source: author). int. j. prod. manag. eng. (2020) 8(1), 13-19creative commons attribution-noncommercial-noderivatives 4.0 international use of the abc curve in medicine line balancing: a case study at a brazilian pharmaceutical distribution center 15 http://creativecommons.org/licenses/by-nc-nd/4.0/ a great financial impact, besides generating high storage cost. thus, a common practice is the use of the concept of “value movement”, which corresponds to the product between the output quantity of the commodity and its unit value. in this way, large movements of value must generate doubled care. the italian economist vilfredo pareto, in his analysis, found that 20% of the products had the capacity to contain 80% of the total value stored, being this study also known as 80/20 rule (corrêa and corrêa, 2012). in figure 5 it is possible to observe its representation. % value group a : 20% of the items represents 80% of the income group b : 30% of the items represents 15% of the income group c : 50% of the items represents 5% of the income % items 1005020 80 95 100 figure 5. abc curve (source: author). 3. case study 3.1. contextualisation based on the logistic distribution center of one of the country’s largest pharmaceutical retail chains, the present work focused on analyzing its inventory processes. more precisely, the stage of drug supply and separation, lacking a better balance of product lines, since the larger outputs should occupy more strategic positions in the structure where they are stored (flow racks). for this, we used the pareto analysis, or abc curve, to define the order of priority to occupy better positions in the layout of the company distribution center. 3.2. a,b,c products definition after receiving and checking the products in the distribution center, those classified as medicines are transported to their respective sector. as shown in figure 6, they are firstly stored on shelves –storage area in blue–, being fed and separated in flow racks –green area. finally, they are transported by mats to the conference or checkout area –in yellow– followed by the packing and expedition. send to stock stock and orders separation checkout and packing figure 6. medicine sector (source: author). once the area of study is defined as the green one in figure 6, the region decomposes into 2 corridors (line 1 and 2) of flow racks and, through the middle, a mat is passed. behind the flow racks, the products are supplied in bin boxes (figure 7 on the left) and separated by pick by light (figure 7 on the right). figure 7. stock and orders separation (source: author). taking into account that the flow rack has 7 levels of height, it was defined that in the more extreme (first and seventh floor) would be placed products c curve, since they are ergonomically more difficult to be accessed. in the second and sixth floors, because of the median access, those of curve b would be placed. and finally, in floors 3, 4 and 5 (of easy access) would be placed products of high turn, curve a. counting all positions of a, b and c curves, a total of respectively 1609, 1076 and 2084 addresses or products were reached (each product occupies an address). thus, the 1609 most accessed products (34% of the total) will be classified as curve a; the int. j. prod. manag. eng. (2020) 8(1), 13-19 creative commons attribution-noncommercial-noderivatives 4.0 international oliveira et al. 16 http://creativecommons.org/licenses/by-nc-nd/4.0/ following 1076 (23% of the total) as curve b and the remaining 2084 (44% of the total) as curve c. only the criterion of number of accesses per product was used and no use was made of its unit value, as it was explained in the section of theoretical foundation. this is due to the fact that high valueadded products are addressed in another sector of the warehouse called “expensives”. therefore, the price of the commodity is not a relevant factor and will not be analyzed. in this way, in order to stratify the information that will be relevant to the study, a report from the integrated logistics system (wms) was used, in which it could be found informations such as: code and description of the products found at each flow rack address, the level or height at which they are found and the number of accesses during the last 3 months before the analysis. a fragment of the result is in table 2. 4. theme discussion and results after analysis, it was possible to draw the study case abc curve as can be seen in figure 8. 34% 57% 100% % products % accesses 87% 98% 100% figure 8. study case abc curve (source: author). it was found that 1609 (34%) most accessed products, classified as curve a, concentrated 87% of the accesses. in addition, 1076 products which were denominated as curve b (57% accumulated) represented 98% of the accesses. and finally, 2084 products were titled as curve c (100% cumulative), consisting of 100% of the cases. knowing the abc classification of the products and where they were addressed (table 2), those data were inputed into microsoft excel®. using its fuction “vlookup” and using the flow rack address (example: 001-001) as the primary key, it was possible to get all the information about the product curve. figure 9 shows the representation of a flow table 2. wms report of products accesses. (source: author). floors flow rack product description accesses % % accumulated curve 4 012-024 2810 buscopan 5482 0,21 0,21 a 1 029-101 319015 deocil sl 10cp 5409 0,20 0,41 a 7 023-167 335479 indapamida 30 cp 5369 0,20 0,62 a 4 012-104 220850 ibuprof. 600 20cp 5080 0,19 0,81 a . . . . . . . . . . . . . . . . . . . . . . . . figure 9. mapping flow rack structures in excel® (source: author). int. j. prod. manag. eng. (2020) 8(1), 13-19creative commons attribution-noncommercial-noderivatives 4.0 international use of the abc curve in medicine line balancing: a case study at a brazilian pharmaceutical distribution center 17 http://creativecommons.org/licenses/by-nc-nd/4.0/ rack station, while figure 10 shows a mosaic with the two medicine lines. lastly, the exchange of poositions between products was established to obey a correct line balancing (curve a at levels 3, 4 and 5, curve b at levels 2 and 6 and curve c at levels 1 and 7). this barter was treated as a project and held by stock operators on a sunday (the normal work operation is from monday to saturday). after rearrangement, the result of a flow rack station could be observed in figure 11. in the same it is observed that addresses in gray are those used to return empty boxes, from the separation to the supply. and those in dark gray are characteristic of double picking, that is, products of high output that, for that reason, occupy two addresses of flow rack. it is also noted that the result has greatly influenced the ergonomics of suppliers and separators, making work less unhealthy, faster, more assertive and requiring a smaller amount of labor. gains were accounted for as a two-hour reduction in the time needed to supply and separate needed products in an 8-hour work day (25% reduction). these data were obtained by reducing the average working time during 4 working weeks where there were no significant changes in the demand or seasonality of the analyzed products. 5. final considerations with the objective of improving the productivity in the supply and separation of a logistics distribution center, the present work was able to deliver to the scientific community and area professionals, a tool able to prioritize products by number of accesses, to classify them according to the abc curve and perform a line balancing in large warehouses with a great quantity of products. figure 10. flow rack structures mosaic in excel® (source: author). figure 11. balanced station (source: author). int. j. prod. manag. eng. (2020) 8(1), 13-19 creative commons attribution-noncommercial-noderivatives 4.0 international oliveira et al. 18 http://creativecommons.org/licenses/by-nc-nd/4.0/ in addition, the technique proved to be simple to execute in a widely used software, microsoft excel®, with fast and expressive gains. the study was limited to a specific sector and could be extended to other areas. in this way, it is possible, for example, to perform line balancing at the products storage area, where they are placed in shelves as soon as they are received and checked. the classification of the products with the greatest number of accesses and the correct storage in the most accessible places would probably provide benefits, such as the use independence of forklifts or platforms to reach them, reduction in supply time, reduction in the number of failures, among others. references aktunc, e. a., basaran, m., ari, g., irican, m., gungor, s. (2019). inventory control through abc/xyz analysis. industrial engineering in the big data era, 175–187. https://doi.org/10.1007/978-3-030-03317-0_15 ballou, r. h. (2006). gerenciamento da cadeia de suprimentos: logística empresarial. 5. ed. rio grande do sul: bookman. ballou, r. h. (2013). logística empresarial: transportes, administração de materiais e distribuição física. são paulo: atlas. corrêa, h. l., corrêa, c. a. (2012). administração de produção e operações: manufatura e serviços uma abordagem estratégica. 3. ed. são paulo: atlas. gu, j., goetschalckx, m., mcginnis, l. f. (2007). research on warehouse operation: a comprehensive review. european journal of operational research, 177, 1-21. https://doi.org/10.1016/j.ejor.2006.02.025 lancioni, r. (1991). distribution cost accounting in international logistics. international journal of physical distribution & logistics management, 21(8), 12-16. https://doi.org/10.1108/eum0000000000399 mauro, v. m. (2009). análise do impacto da aplicação da filosofia lean em armazéns e centros de distribuição: o caso de um centro de distribuição de peças automotivas. dissertação de mestrado – programa de pós graduação em engenharia civil, universidade federal de santa catarina, florianópolis. moura, r. a. (1997). manual de logística: armazenagem e distribuição física. 3. ed. são paulo: imam. pedreira, l. n. (2006). proposta para um sistema de controle de armazéns (wcs) com aplicação em uma empresa de pequeno porte. dissertação de mestrado – programa de pós graduação em gerência de produção, pontifícia universidade católica do rio de janeiro, rio de janeiro. slack, n., brandon-jones, a., johnston, r. (2015). administração da produção. 4. ed. são paulo: atlas. van kampen, t. j., akkerman, r., van donk, d. p. (2012). sku classification: a literature review and conceptual framework. international journal of operations & production management, 32(7), 850–876. https://doi.org/10.1108/01443571211250112 viana, j. j. (2008). administração de materiais: um enfoque prático. são paulo: atlas. vollmann, t. e., berry, w. l., whybark, d. c., jacobs, f. r. (2005). manufacturing planning and control for supply chain management. new york, ny: mcgraw-hill/irwin. int. j. prod. manag. eng. (2020) 8(1), 13-19creative commons attribution-noncommercial-noderivatives 4.0 international use of the abc curve in medicine line balancing: a case study at a brazilian pharmaceutical distribution center 19 https://doi.org/10.1007/978-3-030-03317-0_15 https://doi.org/10.1016/j.ejor.2006.02.025 https://doi.org/10.1108/eum0000000000399 https://doi.org/10.1108/01443571211250112 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2019.12035 received: 2019-01-07 accepted: 2019-07-24 performance measurement in judo: main kpis, cluster categorization and causal relationships uriarte marcos, s.a1, rodríguez-rodríguez, r.a2, uriarte marcos, m.b, alfaro-saiz, j.j.a3 a universitat politècnica de valència, valencia, spain. b facultad de psicología y educación. universidad de deusto, avda. universidades, 24. 48007. bilbao, spain. a1 sugoi66@hotmail.com, a2 raurodro@cigip.upv.es b maier.uriarte@deusto.es, a3 jalfaro@cigip.upv.es abstract: performance measurement in judo usually focuses on some kpis whose values indicate the final performance of the athlete. this paper deals with firstly identifying which these main key performance indicators (kpis) in judo are. once this is done, the kpis are classified into four different clusters: physical training, specific training, psychology and lifestyle. then, it moves into analyzing possible quantitative techniques to identify cause-effect relationships between kpis in order to link not only the impact of the judo kpis with the achievement of the judoka’s strategic objectives but also to identify both the relative and the global importance of each of these kpis. finally, it points out the analytic network technique as the one that could be ideally applied in this context and offers future research actions. key words: judo, kpis, performance measurement, causal relationships, anp. 1. introduction performance measurement is widely applied within industry in order to provide decision-makers with information about the situation of some specific important variables, key performance indicators (kpis), which will mainly lead to check whether the target value (usually a range value) associated to these kpis is being reached or not. widening the approach, performance measurement has been also applied to service organisations such as hospitals, tourism, governments, etc. in many of these cases, performance measurement has gone a step further, forming a structure of not only kpis but also strategic objectives, where the achievement of the latter is linked to the real values measured by the former. such an evolved structure represents the so-called performance measurement systems (pms) (folan and browne, 2005). however, the existence of these pms does not imply that there are still organisations that use only kpis to manage their performance. this fact is more frequent in sport organisations, as they rely only on the kpis not only to control the evolution of achieved values but also to make more strategic decisions. these organisations, from a performance measurement viewpoint, count with lots of measures of their kpis overtime and they rely on making their decisions on these historic values. this research focuses on performance measurement applied to sport management and, more concretively, to judo. the main goal is to highlight the main kpis associated to judo, categorizing and grouping them into clusters for finally presenting an approach via multi-criteria techniques to link these kpis with the defined strategic objectives, quantifying and ranking the importance of the kpis and of the clusters to achieve the strategic objectives. to cite this article: uriarte marcos, s., rodríguez-rodríguez, r., uriarte marcos, m., alfaro-saiz, j.j. (2019). performance measurement in judo: main kpis, cluster categorization and causal relationships. international journal of production management and engineering, 7(2), 145-150. https://doi.org/10.4995/ijpme.2019.12035 int. j. prod. manag. eng. (2019) 7(2), 145-150creative commons attribution-noncommercial-noderivatives 4.0 international 145 http://creativecommons.org/licenses/by-nc-nd/4.0/ then, this paper tackles this line of research, and it is structured as follows: section 2 presents a literature review on the main kpis for judo performance; in section 3 these kpis are classified into clusters; section 4 offers a classification of different quantitative techniques that could be applied to identify cause-effect relationships between the kpis and the the objectives of the judokas. finally, section 5 presents the main conclusions and future research work are highlighted. 2. literature review in the scientific literature, it is possible to find multitude of kpis used to measure performance in sport. bringing the discussion to the judo field, there is a crescent trend to publish in academic journals. then, peset et al. (2013) carried out a bibliometric analysis whose main findings were that there were 383 papers and scientific reviews published in both the science citation index and the social science citation index during the time period of 1950 and 2011. these papers were published in 162 different journals, in 78 categories of the web of science and the publication of judo papers was mainly reduced to journals mainly referred to do with both sport science and sport medicine. alike other sport disciplines, judo is starting to gain a position within the scientific literature, constituting an interesting potential growing field. from these conclusions, it is possible to affirm that, regarding judo and its scientific position, the application of techniques/approaches of other disciplines would lead to advances and interesting applications. next, this research provides a resume of the main kpis found and their area of application regarding judo scientific literature: coordination. perrin et al. (2002) identified coordination as a key variable for performance in judo, stating that the training and improvement in coordination will lead to develop sensorimotor adaptabilities, as revealed their experiments with postural skills of elite jodoist. strength. hassmann et al. (2010) mentions strength as the most important variable to train and improve, as it concludes from their experiments with several olympic judo athletes. in this sense, ache-dias et al. (2012) compares the handgrip strength performance between judokas and non-judokas, discovering that judokas were not stronger than non-judokas but they had a higher level of resistance to fatigue. speed. almansba et al. (2008) categorises the importance of the speed in judo through a comparative study of speed consisting in the number of throws between lighter and heavier judo categories. heart rate. thun et al. (2015) carried out a study to determine how circadian rhythms and sleep affect to performance, concluding that the time of the day affects greatly to the desynchronisation of circadian rhythms. on the other hand, houvenaeghel et al. (2005) observed different heart rates during training of judokas to establish whether low rate levels where associated only with low intensity exercises trying to identify if the different heart rates were due either to a good physical condition or a bad execution. aerobic and anaerobic fitness. franchini et al. (2014) assessed the effect of high-intensity intermittent training to improve judo-training results. technical and tactic preparation. bocioaca (2014) demonstrated that the optimization of the technical training lead to the basis for the tactical preparation, which paves the way for preparing judo medium-long term training planning. in this sense, franchini et al. (2015) carried out a complete revision relating both technique and tactic in judo. age. franchini et al. (2012) carried out a study to compare the motion-time performance of judo fighters with different ages. weight. escobar-molina et al. (2016) highlighted the weight as one of the main differential factors in judo, fostering loss weight to compete in a lower category as a competitive advantage. however, calvo rico et al. (2018) alerts against extreme weight losing. focus and concentration level. grosu et al. (2014) evaluated the role of the cognitive process of attention for judo and alpine ski athletes, concluding that there is a relationship between both focused and the perception of attention. stress. arnold et al. (2018) studied the competitive stress process. they concluded that social support moderates the relationship between organizational stressing elements and subjective performance. int. j. prod. manag. eng. (2019) 7(2), 145-150 creative commons attribution-noncommercial-noderivatives 4.0 international uriarte marcos et al. 146 http://creativecommons.org/licenses/by-nc-nd/4.0/ motivation. gillet et al. (2010) explored the role of the coaches and the effect on athletes’ motivation. activation level. cohen-zada et al. (2017) studied the psychological momentum of different judo athletes by gender. they concluded that men judo fighters’ performance is influenced in a higher degree by the psychological momentum than women judo fighters. nutrition. the focus here is to evaluate nutritional supplements and their effect on performance. then, ramezani et al. (2019) carried out a systematic review to assess the effect of glutamine supplementation on athletes’ performance. on the other hand, sousa et al. (2016) assessed whether using nutritional supplements has to do with nutritional inadequacy coming from food. finally, hung et al. (2010) assessed the effect of supplementation of b-hydroxy and b-methylbutyrate (hmb) on reducing body fat during energy restriction in female judo fighters. they found that hmb during energy restriction may help to decrease body fat, having no relevant effect on lean body mass or performance in female judo athletes. dual career. muñoz-bullón et al. (2017) studied the effect of sports participation on high education academic performance, finding a direct and positive relationship. sleep. knowles et al. (2018) studied the effect of inadequate sleep on resistance training, concluding that there is a direct and negative relationship between them. on the other hand, rosenbloom and grossman (2018) affirmed that to take short naps in any scenario would have a positive impact on performance. 3. cluster aggregation once the main kpis that affect performance in judo found in the scientific literature have been previously highlighted, it is time to classify them into common groups or clusters. this will facilitate a posterior application of multi-criteria techniques and, more concretively, of the analytic network process. this will be further developed in the next point of the paper. then, table 1 presents both the formed clusters and the kpis that constitute each of them. table 1. clusters formed with the main kpis. cluster kpi physical training coordination strength speed heart rate aerobic and anaerobic fitness specific training technical and tactic preparation age weight psychology focus and concentration level stress motivation activation level lifestyle nutrition dual career sleep then, we propose 4 clusters: physical training, specific training, psychology and lifestyle. these 4 clusters have been built based on not only on the extensive international experience in judo of 2 of the authors (having participated in the olympic games and being european champion) but also on the main characteristics of the kpis. each of these clusters is constituted by, at least, 3 kpis found to be important regarding judo performance. the next question is: are there kpis from all these to be more important than others regarding performance? it is obvious that the answer to this question is yes. then, the next work to develop is to find out which ones are more important and under what circumstances. in this sense, it is necessary to consider these kpis not only individually but also globally. this is due to the fact that it is reasonable to expect that exist relationships between and among these kpis, which should be taken into account, as the improvement of one of these kpis will surely affect to other/s kpis. for example, it could be expected that the improvement of a lifestyle kpi such as sleep or nutrition will have a positive impact on the judoka’s performance. then, it would be interesting to apply some sort of technique to establish such kpis relationships. additionally, it is possible to think of a further scenario in which, besides finding relationships between and among kpis, it is possible to quantify the impact that has got each one of these on the athlete strategic performance, understood the latter as the specific formulation of objectives such as to int. j. prod. manag. eng. (2019) 7(2), 145-150creative commons attribution-noncommercial-noderivatives 4.0 international performance measurement in judo: main kpis, cluster categorization and causal relationships 147 http://creativecommons.org/licenses/by-nc-nd/4.0/ become national champion, to win the classification for the olympic games etc. this approach would offer a ranked list of the kpis, from a higher to a lower relative importance regarding their impact on achieving the defined objectives. such a list would constitute an additional information for judo fighters, who would know what factors (kpis) they should potentiate first in order to achieve their specific objectives. then, the next section presents some quantitative techniques that could be used to identify cause-effect relationships. 4. quantitative techniques to identify causal relationships there are many technique that could be applied to both identify and quantify causal relationships between and among kpis and the defined strategic objectives. table 2 highlights some of these techniques, classifying them according to the data they use (historical or based on experience), the type of the relationships found between the variables (dyadic or multiple) and the capacity of finding interrelationships between and among the variables. as it can be seen in table 2, there are many techniques candidates to be applied to solve this problem however the authors recommend a multi-criteria approach, as these techniques offer the possibility of not using historical data. this is of great importance as the judokas can obtain historical data for each one of the kpis mentioned previously, but some problems arise when trying to correlate them with the evolution of the performance objectives defined. in this sense, kpis in judo usually gather data with a daily frequency; i.e. times in successive running series, weight lifted, weight of the athlete, etc. on the other hand, the frequency of the strategic objectives is yearly (become national champion) or even wider (4 years: participate in olympic games). this make necessary to aggregate the data from the judo kpis to the yearly frequency before to carry out the correlation process, which augments the difficulty of the process. in addition, it will be necessary to carefully assess the correlation results to avoid interpretation mistakes; for example, if it results that speed turns out to be the most important kpi for the objective of participate in the olympic games the decision-makers (coaches, athlete) will need to decide what it the relation of speed with the others judo kpis and their reciprocal relative importance to achieve all the objectives. then, the multi-criteria analytic network process (saaty, 1996) offers the possibility of carrying out all these operations at once. the main limitation of anp when compared with statistical techniques in our context is that it is based on subjective comparisons carried out by an expert group. on the other hand, correlation analysis takes into account only historic data, being therefore an objective technique. as stated above, the application of anp would provide results to identify and quantify to what extent the different judo kpis are contributing to achieve defined strategic objectives. the anp is a technique that, depending on the size of its components, it may turn to be of quite high complexity. in the context of this problem, anp should be used indicating which the inputs or clusters of judo kpis are and which the alternatives or performance strategic objectives are. then, the model of the problem as an anp network should include the next elements: inputs. the four clusters of judo kpis above presented. physical training, specific training, psychology and lifestyle. alternatives. the strategic objectives of the judokas. it is important to keep in mind that this problem is unidirectional, from the inputs to the alternatives. in other words, we aim to quantify the relationships from the inputs (at both the clusters and the kpis table 2. clusters formed with the main kpis. technique/s data type of relationships interrelationships correlation analysis historical data (from kpis) dyadic not captured partial least squares historical data (from kpis) multiple captured factorial analysis historical data (from kpis) multiple captured principal component analysis historical data (from kpis) multiple captured structural equation modeling historical data (from kpis) multiple captured analytic hierarchical process based on experience/subjective judgement pairs not captured analytic network process based on experience/subjective judgement pairs captured int. j. prod. manag. eng. (2019) 7(2), 145-150 creative commons attribution-noncommercial-noderivatives 4.0 international uriarte marcos et al. 148 http://creativecommons.org/licenses/by-nc-nd/4.0/ levels) to the alternatives. then, the first step is to define the relationships matrix, which relates the direct relationships between all the different variables. when such a relationship exits, the intersection cell between the two variables will get a 1 value, otherwise it will get a 0 value. once this has been done the next steps are: the expert group carries out the comparisons between nodes and clusters. to compute the unweighted, weight and limit matrices. to interpret the obtained results. 5. conclusions and future research work this paper has presented which the main kpis for measuring judo performance are. then, these kpis have been grouped into four clusters: physical training, specific training, psychology and lifestyle. next, it has point out the problematic of obtaining the relative importance of these kpis and clusters when measuring their importance in achieving strategic objectives of the judokas. then, different techniques that could be applied to solve this matter have been presented and classified, recommending the usage of the anp technique as the tool to solve this problematic. at this point, the next research work should start completing all the necessary anp steps, obtaining then results to identify which the main judo kpis that affect performance are. references ache dias, j., wentz, m., külkamp, w., mattos, d., goethel, m., borges júnior, n. (2012). is the handgrip strength performance better in judokas than in non-judokas. science and sports, 27(3), 9-14. https://doi.org/10.1016/j.scispo.2011.10.005 almansb, r., franchini, e., sterkowicz s., imamurad, r.t., calmet, m., ahmaidi, s. (2008). a comparative study of speed expressed by the number of throws between heavier and lighter categories in judo. science & sports, 23, 186-188. https://doi.org/10.1016/j. scispo.2007.10.014 arnolda, r., edwards, t., rees, t. (2018). organizational stressors, social support, and implications for subjective performance in high-level sport. psychology of sport & exercise, 39, 204-212. https://doi.org/10.1016/j.psychsport.2018.08.010 bocioaca, l. (2014). technical and tactical optimization factors in judo. procedia-social and behavioral sciences, 117, 389-394. https://doi.org/10.1016/j.sbspro.2014.02.233 calvo-rico, b., fernandes-monteiro, l., aznar-laín, s., garcía-garcía, j.m. (2018). comparación de las variaciones de la composición corporal entre judokas y luchadores hidratados vs. deshidratados. / comparison of body composition variations between hydrated and dehydrated judokas and wrestler. retos: nuevas perspectivas de educación física, deporte y recreación, 2041(33). cohen-zada, d. krumer, a., shtudinerd, z. (2017). psychological momentum and gender. journal of economic behavior & organization, 135, 66-81. https://doi.org/10.1016/j.jebo.2017.01.009 escobar-molina, r., nikolaidis, p., torres-luque, g., garatachea, n., hernández-garcía, r. (2016). physical and physiological characteristics of judo athletes: an update. sports, 4(1), 10-20. https://doi.org/10.3390/sports4010020 folan, p.,browne, j. (2005). a review of performance measurement: towards performance management. computers in industry, 56, 663680. https://doi.org/10.1016/j.compind.2005.03.001 franchini, e., julio, u.f., panissa, v.l.g., calmet, m., del vecchio, f.b., miarka, b. (2012). a comparison of time-motion performance between age groups in judo matches. journal of sports sciences, 30(9), 899-905. https://doi.org/10.18002/rama.v5i1.139 franchini, e., sterkowicz-przybycien, k., takito, m.y. (2014). anthropometrical profile of judo athletes: comparative analysis between weight categories. international journal of morphology, 32(1), 36-42. https://doi.org/10.4067/s0717-95022014000100007 franchini, e., miarka, b., julio, u. f., calmet, m., del vecchio, f.b. (2015). técnica y táctica en judo: una revisión. revista de artes marciales asiáticas, 5(1), 86-91. https://doi.org/10.18002/rama.v5i1.139 gillet, n., vallerand, r.j., amoura, s., baldes, b. (2010). influence of coaches’ autonomy support on athletes’ motivation and sport performance: a test of the hierarchical model of intrinsic and extrinsic motivation. psychology of sport and exercise, 11, 155-161. https://doi.org/10.1016/j.psychsport.2009.10.004 grosu, e.f., vlad, t. moraru, c.e., motică, f. (2014). the role of attention in preparation of athletes in perceptual-motor discipline. procediasocial and behavioral sciences, 159, 151-155. https://doi.org/10.1016/j.sbspro.2014.12.348 hassmann, m., buchegger, m., stollberg, k.-p., sever, a., sabo, a. (2010). motion analysis of performance tests using a pulling force device (pfd) simulating a judo throw. procedia engineering, 2, 3329-3334. https://doi.org/10.1016/j.proeng.2010.04.153 int. j. prod. manag. eng. (2019) 7(2), 145-150creative commons attribution-noncommercial-noderivatives 4.0 international performance measurement in judo: main kpis, cluster categorization and causal relationships 149 https://doi.org/10.1016/j.scispo.2011.10.005 https://doi.org/10.1016/j.scispo.2007.10.014 https://doi.org/10.1016/j.scispo.2007.10.014 https://doi.org/10.1016/j.psychsport.2018.08.010 https://doi.org/10.1016/j.sbspro.2014.02.233 https://doi.org/10.1016/j.jebo.2017.01.009 https://doi.org/10.3390/sports4010020 https://doi.org/10.1016/j.compind.2005.03.001 https://doi.org/10.4067/s0717-95022014000100007 https://doi.org/10.18002/rama.v5i1.139 https://doi.org/10.1016/j.psychsport.2009.10.004 https://doi.org/10.1016/j.sbspro.2014.12.348 https://doi.org/10.1016/j.proeng.2010.04.153 http://creativecommons.org/licenses/by-nc-nd/4.0/ houvenaeghel, m., bizzari, c., giallurachis, d., demelas, j.m. (2005). continuous recording of heart rate during specific exercises of judo. science & sports, 20(1) 27-32. https://doi.org/10.1016/j.scispo.2004.05.009 hung, w., liu, t., chen, c., chang, c. (2010). effect of β-hydroxy-β-methylbutyrate supplementation during energy restriction in female judo athletes. j exerc sci fit, 8(1), 50-53. https://doi.org/10.1016/s1728-869x(10)60007-x knowles, o.e., drinkwater, e.j., urwin, c., lamona, s., aisbetta, b. (2018). inadequate sleep and muscle strength: implications for resistance training. journal of science and medicine in sport, 21, 959-968. https://doi.org/10.1016/j.jsams.2018.01.012 muñoz-bullón, f., sanchez-bueno, m.j., vos-saz, a. (2017). the influence of sports participation on academic performance among students in higher education. sport management review, 20, 365-378. https://doi.org/10.1016/j.smr.2016.10.006 perrin, p., deviterne, d., hugel, f., perrot, c. (2002). judo, better than dance, develops sensorimotor adaptabilities involved in balance control. gait and posture 15, 187-194. https://doi.org/10.1016/s0966-6362(01)00149-7 peset, f., ferrer-sapena, a., villamón, m., millán gonzález,, l., toca-herrera, j.-l., aleixandre-benavent, r. (2013). scientific literature analysis of judo in web of science. arch. budo, sciences of martial arts, 9(2), 81-91. https://doi.org/10.12659/aob.883883 ramezani, a., rayyani, e., bahreini, m., mansoori, a. (2019). the effect of glutamine supplementation on athletic performance, body composition, and immune function: a systematic review and a meta-analysis of clinical trials. clinical nutrition 38, 1076-1091. https://doi.org/10.1016/j.clnu.2018.05.001 rosenbloom, t., grossman, e. (2018). assessment of performance impairment after short naps with and without sleep inertia. transportation research, 52, 1-13. https://doi.org/10.1016/j.trf.2017.10.001 saaty, (1996). the analytic network process: decision making with dependence and feedback, rws publications, pittsburgh, pa. sousa, m., fernandes, m., carvalho, p., soares, j., moreira, p., teixeira, v.h. (2016). nutritional supplements use in highperformance athletes is related with lower nutritional inadequacy from food. journal of sport and health science, 5, 368-374. https://doi.org/10.1016/j.jshs.2015.01.006 thun, e., bjorvatn, b., flo, e., harris, a., pallesen, s. (2015). sleep, circadian rhythms and athletic performance. sleep medicine reviews, 23, 1-9. https://doi.org/10.1016/j.smrv.2014.11.003 int. j. prod. manag. eng. (2019) 7(2), 145-150 creative commons attribution-noncommercial-noderivatives 4.0 international uriarte marcos et al. 150 https://doi.org/10.1016/j.scispo.2004.05.009 https://doi.org/10.1016/s1728-869x(10)60007-x https://doi.org/10.1016/j.jsams.2018.01.012 https://doi.org/10.1016/j.smr.2016.10.006 https://doi.org/10.1016/s0966-6362(01)00149-7 https://doi.org/10.12659/aob.883883 https://doi.org/10.1016/j.clnu.2018.05.001 https://doi.org/10.1016/j.trf.2017.10.001 https://doi.org/10.1016/j.jshs.2015.01.006 https://doi.org/10.1016/j.smrv.2014.11.003 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering special issue: advances in engineering networks https://doi.org/10.4995/ijpme.2019.10572 received: 2018-07-27 accepted: 2019-01-22 analysis of technological knowledge flows in the basque country gavilanes-trapote, j. a*, etxeberria-agiriano, i. b, cilleruelo, e. c, and garechana, g. d a industrial organization and management engineering department. faculty of engineering vitoria-gasteiz. university of the basque country (upv/ehu). nieves cano 12, 01006 vitoria-gasteiz. spain b languages and computer systems department. faculty of engineering vitoria-gasteiz. upv/ehu c industrial organization and management engineering department. faculty of engineering bilbao. upv/ehu d industrial organization and management engineering department. faculty of economics and business bilbao. upv/ehu a javier.gavilanes@ehu.eus, b ismael.etxeberria@ehu.eus, c ernesto.cilleruelo@ehu.eus, d gaizka.garechana@ehu.eus abstract: knowledge flow of technology is important for continuous growth and extension of science. patent data analysis has facilitated this knowledge acquisition. the available patent information crosses borders, corresponds and interacts with new inventions to give new strength and dimension to the technology. therefore, the patent citation information functions as a key indicator of the knowledge flow providing relevant information. it can be identified to which extent a region is a relevant technological knowledge generator to other regions. as an illustrative case, we present a study to determine the role played by the basque country region as a generator of technological innovation during the period 1991-2011. key words: patent, citation, technology, diffusion, innovation. 1. introduction knowledge is an intangible strength that has a definite economic importance when it is well utilized and commercialized. knowledge spillover is something that occurs, imaginable but difficult to measure effectively. the tendency to locate knowledge in certain areas together with the effects derived from agglomeration economies are responsible for the strong concentration process and regional specialization observed in a growing way in the economy (krugman, 1995). from the interaction between infrastructures and the built environment, accessible natural resources, the institutional endowment and the knowledge and skills available in the territory, localized capacities are developed, difficult to imitate and of cumulative nature, which lead to competitive advantages of the territory (malmberg and maskell, 1997). the competitiveness of the regions depends to a large extent on the characteristics of their “innovation and development network”, the analysis of the knowledge transfers being an output of competitiveness of the region. there exists a debate about the suitability of considering patents as a proxy for technological knowledge, which has highlighted both its limitations and its benefits. in principle, we could consider that the optimal measure of technological knowledge is given by the number of innovations, which in turn are understood as those novelties that have come to be marketed. the main limitation of this measure is determined by the almost total unavailability of data, although the annual surveys of eustat (basque institute of statistics) include questions about the “economic impact of product innovations on turnover”. even so, to cite this article: gavilanes-trapote, j., etxeberria-agiriano, i., cilleruelo, e., and garechana, g. (2019). analysis of technological knowledge flows in the basque country. international journal of production management and engineering, 7(special issue), 73-79. https://doi.org/10.4995/ijpme.2019.10572 int. j. prod. manag. eng. (2019) 7(special issue), 73-79creative commons attribution-noncommercial-noderivatives 4.0 international 73 https://orcid.org/0000-0001-6773-0815 http://orcid.org/0000-0001-6146-3891 https://orcid.org/0000-0001-5691-9778 https://orcid.org/0000-0002-1913-3239 http://creativecommons.org/licenses/by-nc-nd/4.0/ this measure has a series of drawbacks. for example, data come from surveys sensitive to the response rate and to the interpretation of the term innovation by the companies, together with the average life cycle of the products of the companies consulted. on the other hand, patents and their evaluation process are objective. in addition, being official documents increases their rigor with less typographical errors than in other types of documents. another disadvantage is that the introduction of a new product in the market takes place in the last phase of the innovation process, which may be very far from the moment in which the r&d effort is made (schmoch, 2003). however, in the case of patents, this lapse of time is much shorter. another considerable advantage of patents is that they provide a wide temporal, geographical and technological scope as a measure of technological innovation. patent information is collected since the mid-nineteenth century, in most countries of the world and practically covers all technological fields. a remarkable exception is software, which is usually protected through property rights, and can only be patented if integrated into a product or production process, oecd (2009). particular advantage of great importance for the present work lies in the accessibility of patent information, highlighting three key aspects: the level of structure presented by patent documents, information technologies and the role played by patent offices. the degree of structuring presented by patent documents and the advancement in information technologies have made possible to create patent databases that are easy to use and allow quick retrieval of documents searched at each moment. on the other hand, the resources that patent offices invest in creating, maintaining and disseminating these databases, bringing them closer to users, are fundamental in this process. in this regard, in recent years, most patent offices have created various databases available to all users through the internet. finally, regarding the cost of information, it is interesting to distinguish between the services supplied by patent offices and private entities. the former are making great efforts to disseminate patent information and reduce the cost of their services, some of which are free. besides, private entities are mainly characterized by offering services adapted to their customers’ needs, like the creation of specialized databases in sectors such as pharmaceutical and chemical. in short, patents are far from being a perfect measure of technological output and its application in products and services, but so far we consider they are the best and most complete measure available for the analysis of the technological knowledge of a region. patents provide a detailed description of how inventions have been made at the earlier state of the art, thus constituting a reliable measure of the transfer of technological knowledge (oecd, 2009; higham et al., 2017). patent citations indicate the use of prior inventions, which makes it possible to identify the influence of a particular invention or set of inventions and mapping its diffusion in the economy (acs et al., 2002; you et al., 2017). citations from other patents or non-patent literature (npl) or knowledge between science and technology are useful in quantifying the level of knowledge transfer between organizations, geographical regions and/or technology sectors (gavilanes-trapote et al., 2015, 2017). there are basically two types of citation. on the one hand, patent references are citations of relevant technology previously protected by other patents applied for anywhere in the world, at any time and in any language (backward and forward citations, technological knowledge flows). conversely, references classified as npl are scientific publications, conference proceedings, books, database guides, technical manuals, standard descriptions, and so on. the potential offered by patent citation measurements for policy elaboration is immense. the main methods utilized in literature on innovation are three: i) the measurement of flows or the effects of propagation of knowledge (for example, jaffe and others, 2000); ii) measurement of the quality of the patent (for example, harhoff et al., 2004); and iii) the strategic behavior of the companies (for example, podolny et al., 1996). the present work will focus on analyzing the knowledge flow between patents. the degree of technological accumulation is defined as the frequency with which a society cites its own previous investigations. the identification of the self-citation (applicant or holder) has important implications, among other things for the study of the propagation effect of technological knowledge. it can be assumed that citations of patents belonging to the same owner represent transmissions of knowledge that are mostly internalized, while citations of patents of “others” are closer to the pure notion of the diffusion effect. int. j. prod. manag. eng. (2019) 7(special issue), 73-79 creative commons attribution-noncommercial-noderivatives 4.0 international gavilanes-trapote et al. 74 http://creativecommons.org/licenses/by-nc-nd/4.0/ a customary measure of the degree of technological accumulativeness of a region is the sum of retrospective citations of patents of that same region among the total number of patents it has, at a given time. according to malerba and orsenigo (1995), the degree of technological accumulation affects the extent to which innovative leaders build a competitive advantage over followers and the preservation of leadership in the future. 2. objectives the present work has as main objective to determine the role played by the basque country as a generator or receiver of technological knowledge in the world. for this purpose, countries receiving and emitting technological knowledge in the studied region will be identified. finally, the technological accumulation of the region will be calculated. 3. methods the sample includes all patents from january 1, 1992 to december 31, 2011. there are several reference dates for selecting patents over a period of time. the most important are the application date and the grant date. patent studies can be found where both dates are used indistinctly. in this study, the date of application has been chosen because it is the closest date to the innovative activity. the disadvantage of using this date as an element of selection is that it does not guarantee that the patent has been granted. therefore, it is necessary to identify which patents have been granted and eliminate the remaining patents from the sample. the use of this date presents another inconvenience due to the waiting time from the time a patent is applied for until its concession is published (up to 6 years), making it impossible to work with a recent sample without suffering a significant truncation effect. the final sample contains 3,503 patents. the databases used to access the patent information in the sample are invenes (http://invenes.oepm. es/invenesweb/faces/busquedainternet.jsp), of the spanish patent and trademark office (spto) and patstat (https://www.epo.org/searching-forpatents/business/patstat.html), of the european patent office (epo). the patstat database contains all fields including patent citations, but lacks the possibility of discriminating spanish patents by province. then again, invenes allows one to search by provinces, but it lacks access to certain important fields such as forward or backward patents. the followed process can be described as follows. all patents of the autonomous community of the basque country (capv for its spanish acronym) requested between the years 1992-2011 were downloaded. queried fields were those of province “prov” and publication number “npub”. this publication number uniquely identifies the patents requested in spain and is equivalent to the field “publn_nr” of the patstat database. in this field, duplicates were found due to the fact that throughout the process of querying for a patent in the spanish office various documents are being generated for the same registry. after the elimination of these duplicate records, the patstat database was accessed to extract all the fields necessary for the study of the capv patents, see table 1. the patstat database is a payment service but allows a free trial version for 2 months. sql queries are carried out online through a graphical interface, see figures 1 and 2. table 1. bibliographic fields downloaded from the patstat database. source: self elaboration. field description appln_auth nationality of the office from where the patent was requested appln_id number assigned by the epo to each patent for its univocal identification han_name name of patent applicants person_address address of applicants person_ctry_code nationality of applicants person_id number that uniquely identifies each applicant prior_earliest_date priority date or date of the first request publn_nr number that identifies the patent in the national office the epo website provides a manual to facilitate the formulation of queries in sql language. once the search is done, the selected records can be downloaded in “.csv” (comma separated values) format. query results were downloaded into “.csv” files for later importation into the vantage point (https://www.thevantagepoint.com) text mining int. j. prod. manag. eng. (2019) 7(special issue), 73-79creative commons attribution-noncommercial-noderivatives 4.0 international analysis of technological knowledge flows in the basque country 75 http://invenes.oepm.es/invenesweb/faces/busquedainternet.jsp http://invenes.oepm.es/invenesweb/faces/busquedainternet.jsp https://www.epo.org/searching-for-patents/business/patstat.html https://www.epo.org/searching-for-patents/business/patstat.html https://www.thevantagepoint.com http://creativecommons.org/licenses/by-nc-nd/4.0/ software. through this program data from different imported files were merged through the key field “appln_id”. subsequently, they were checked for duplicate records. this task, like that of the merged, is not complicated since the database contains a field that uniquely identifies each patent and the program has a specific function to perform these tasks. once all this information is merged into a single file we move on to the arduous and tedious task of data cleaning. the cleaning stage aims at solving one of the most notable problems in bibliographic databases: errors and lack of data consistency. for both database producers and researchers downloading data for scientific purposes, the lack of standardization and errors suppose the loss of information that forces the development of corrective systems, almost always personalized, that guarantee the rigor of the research, so dependent on the quality of the data (gálvez and moya-anegón, 2007). out of the fields used for the present study, “har_ name” has been the one that has required the most cleaning work. the normalization of the organizations and researchers applying for patents contained in this field requires facing two main problems: homonymy (two applicants with the same name) and synonymy (different variants of name referring to a single applicant). homonymy is presented only in the names of the researchers. organizations do not present this problem because they are obliged to have a unique denomination to allow them joining the mercantile registry. the solution to this bias is to extract information from the context. in the case of patents, it could be resolved by analyzing the researcher address in field “person_address”. the problem in our database is that this field is only completed in 8% of the records, which makes it far too complicated to affirm that this homonymy bias does not affect any record of the study. figure 1. graphical patstat database interface. querying window. source: self captured. figure 2. graphical patstat database interface. results window. source: self captured. int. j. prod. manag. eng. (2019) 7(special issue), 73-79 creative commons attribution-noncommercial-noderivatives 4.0 international gavilanes-trapote et al. 76 http://creativecommons.org/licenses/by-nc-nd/4.0/ the synonymy can be the source of serious problems in any type of bibliographic recount. variations of the names of applicants can be classified as follows: invalid variations: mainly caused by spelling, phonetic or typographical errors; incorrect use of capitals, nicknames or abbreviations; or accentuation problems. valid variations: caused by the exchange of word order; separation of words; use or absence of punctuation marks; use of initials or absence of any part of the name, see examples in figure 3. zigor corporacion,s.a. zigor sa azcoitia arreche jose miguel azcoitia arreche, jose, miguel azcoitia arteche, jose figure 3. examples of applicants transcribed in different ways. source: self elaboration. the solution to synonymy, unlike homonymy, is simple but costly in terms of resources. the vantage point program has grouping algorithms (fuzzy clustering) that try to solve this problem. in practice, it ends up being necessary to review applicants’ names one by one and manually perform the clustering through an interface provided by the program within the “list cleanup...” command. the result of this work is the thesaurus, a file that contains these groups. with data selected, correctly merged and cleaned, we are in a position to perform the analyses. to determine to which extent the basque country region is a recipient of technological knowledge generated in other countries, we will look at the “person_ctry_code” field in the backward patent citations. as the “person_ctry_code” citations field only provides information about the country and not about the region to which the patent belongs, a manual treatment, document by document, identifying the location of each applicant company or organization has been necessary, having to discard patents where only researchers are listed, as we cannot know if they belong to the capv or to another community. to determine to which extent the basque country region generates the technological knowledge used by other countries as a basis for their innovations, we will look at the “person_ctry_code” field of forward patent citations. finally, to determinate the degree of technological accumulation we shall calculate what percentage of forward citations of capv are self-citations. 4. results first of all, we will answer the question: “where do we get the knowledge of patents in the basque country?” to do this, it is necessary to analyze the nationality of the backward citations. 81% of patents in the basque country present one or more such citations, accounting for a total of 12,752 backward citations. 22% out of these have no nationality, reducing the number of citations to 9,947. finally, the total number of nationalities to be analyzed is 12,810. this number is higher than the number of citations due to the fact that the chosen count type is complete and more than one of these patent citations has several applicants with different nationalities. figure 4 shows how the united states, with 29%, is the country that contributes most technological knowledge to patents in the basque country. second, germany appears with 17%, followed by france and spain, with 9% each. figure 5 shows only patent citations of spanish nationals (1,194) discriminating against those corresponding to the basque country. the final number of citations of spanish nationality with an organization as applicant is 538, being 61% (324) organizations located in the region of the basque country. us 29% de 17%fr 9% es 9% jp 7% gb 5% it 4% ch 3% ca 2% others 15% figure 4. nationality of patents supporting inventions in the basque country, forward patent citations. int. j. prod. manag. eng. (2019) 7(special issue), 73-79creative commons attribution-noncommercial-noderivatives 4.0 international analysis of technological knowledge flows in the basque country 77 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 5. percentage of backward citations of national origin belonging to the basque country secondly, we will answer the question: “where is the knowledge of patents in the basque country transferred to?” to this end, the nationalities of forward citations are analyzed. only 20% of patents in the basque country are cited by other patents, receiving a total of 1,563 citations. only 4% out of these have no nationality assigned to any of their applicants, so the final number of subsequent appointments is 1,496. figure 6 shows how spain, with 46%, is the main country that receives technological knowledge from patents in the basque country. second, germany appears with 10% and then, in a smaller percentage, other countries such as france, the united states and italy, among others. figure 7 shows all the forward citations of spanish patents (688) discriminating those corresponding to the basque country. the final number of forward citations of spanish nationality with an organization as applicant is 378, being 61% (231) of companies located in the basque country region. these percentages are the same as those mentioned above, so the basque country maintains the selfcitation level in the region. es 46% de 10% fr 6% us 5% it 4% gb 4% ch 3% jp 3% nl 2% others 17% figure 6. nationality of patents that are relying on basque inventions to develop their technological knowledge, forward patent citations figure 7. nationality of patents that are relying on basque inventions to develop their technological knowledge, forward patent citations finally, to calculate the degree of technological accumulation in the basque country, we have 1,496 backward citations with nationality that have received basque patents, with 231 from basque companies. this implies at least a 15% degree of technological accumulation, that is, almost 1 out of 6 patents serves as a support to generate new technological knowledge in the region. 5. conclusion the present work identifies which regions or applicants have contributed with their innovations to the development of new patents, defining the flow of technological knowledge from a geographical or authoring point of view of the inventions. this information allows us to identify the regions that have contributed with the most technological knowledge, which will help in the design of regional policies. through the origin of the backward citations, we can determine which countries contribute the most to the technological development of studied patents. in the case of the basque country, the united states generates the greatest flow of technological knowledge. in second place comes germany, followed by france and spain. this knowledge serves to develop innovations in the studied region. on the contrary, through the origin of the forward citations we can determine how the region of study shares its technological knowledge with other regions. in the case of the basque country, it shares almost half of its knowledge with the region of spain (46%), followed by germany (10%). these data clearly indicate that the role played by the basque country region within spanish territory is mainly to generate technological knowledge as a base for future innovations. int. j. prod. manag. eng. (2019) 7(special issue), 73-79 creative commons attribution-noncommercial-noderivatives 4.0 international gavilanes-trapote et al. 78 http://creativecommons.org/licenses/by-nc-nd/4.0/ the degree of technological accumulation in the basque country is relatively high (15%). this means that 15% of the innovative knowledge generated in the region is used to generate new knowledge. this level of internal transfer helps innovative leaders to build a competitive advantage over their followers and maintain their leadership in the future. results from this study indicate the basque country is an important region from the point of view of knowledge generation within the spanish territory. acknowledgements the authors want to thank the technicians of the patent offices for resolving all the questions during the extraction and exploitation of the data, especially josé maría roncero from the spanish patent and trademark office (spto). references acs, z.j., anselin, l., and varga, a. (2002). patents and innovation counts as measures of regional production of new knowledge res. policy, 31(7), 1069-1085. https://doi.org/10.1016/s0048-7333(01)00184-6 balland, p., and rigby, d. (2017). the geography of complex knowledge. economic geography, 93(1), 1-23. https://doi.org/10.1080/0 0130095.2016.1205947 galvez, c., and moya-anegon, f. (2007). standardizing formats of corporate source data. scientometrics, 70(1), 3-26. https://doi.org/10.1007/s11192-007-0101-0 gavilanes-trapote, j., cilleruelo-carrasco, e., etxeberria-agiriano, i., garechana, g., rodríguez andara, a. (2019). qualitative patents evaluation through the analysis of their citations. case of the technological sectors in the basque country. in: ortiz, á., andrés romano, c., poler, r., garcía-sabater, j.p. (eds) engineering digital transformation. lecture notes in management and industrial engineering. springer, cham. https://doi.org/10.1007/978-3-319-96005-0_28 gavilanes-trapote, j., río-belver, r. m., cilleruelo, e., garechana g., and larruscain j. (2015). patent overlay maps: spain and the basque country. international journal of technology management (ijtm), 69(3/4), 261. https://doi.org/10.1504/ijtm.2015.072976 harhoff, d., scherer, f., and vopel, k. (2004). citations, family size, opposition and the value of patent rights (vol 32, pg 1343, 2003). research policy, 33(2), 363-364. https://doi.org/10.1016/j.respol.2003.10.001 higham, k.w., governale, m., jaffe, a.b., and zülicke, u. (2017) fame and obsolescence: disentangling growth and aging dynamics of patent citations. phys. rev. e 95, 042309. https://doi.org/10.1103/physreve.95.042309 jaffe, a., trajtenberg, m., and fogarty, m. (2000). knowledge spillovers and patent citations: evidence from a survey of inventors. american economic review, 90(2), 215-218. https://doi.org/10.1257/aer.90.2.215 krugman, p. (1995). development, geography, and economic theory. cambridge-massachusetts: the mit press. malerba, f., and orsenigo, l. (1995). schumpeterian patterns of innovation. cambridge journal of economics, 19(1), 47-65. https://doi.org/10.1093/oxfordjournals.cje.a035308 malmberg, a., and maskell, p. (1997). towards an explanation of regional specialization and industry agglomeration. european planning studies, 5(1), 25-41. https://doi.org/10.1080/09654319708720382 murata, y., nakajima, r., okamoto, r., and tamura, r. (2014). localized knowledge spillovers and patent citations: a distance-based approach. review of economics and statistics, 96(5), 967-985. https://doi.org/10.1162/rest_a_00422 organisation for economic co–operation and development (oecd). (2009). oecd patent statistics manual oecd. podolny, j., stuart, t., and hannan, m. (1996). networks, knowledge, and niches: competition in the worldwide semiconductor industry, 1984-1991. american journal of sociology, 102(3), 659-689. https://doi.org/10.1086/230994 schmoch, u. (2003). service marks as novel innovation indicator. research evaluation, 12(2), 149-156. https://doi. org/10.3152/147154403781776708 simmie, j. (2003). innovation and urban regions as national and international nodes for the transfer and sharing of knowledge. regional studies, 37(6-7), 607-620. https://doi.org/10.1080/0034340032000108714 stek, p.e., and van geenhuizen, m.s. (2016). the influence of international research interaction on national innovation performance: a bibliometric approach. technological forecasting and social change, 110, 61-70. https://doi.org/10.1016/j.techfore.2015.09.017 thompson, p., and fox-kean, m. (2005). patent citations and the geography of knowledge spill-overs: a reassessment. american economic review, 95(1), 450-460. https://doi.org/10.1257/0002828053828509 thompsori, p. (2006). patent citations and the geography of knowledge spillovers: evidence from inventorand examiner-added citations. review of economics and statistics, 88(2), 383-388. https://doi.org/10.1162/rest.88.2.383 you, h., li, m., hipel, k.w., jiang, j., ge, b., and duan, h. (2017) development trend forecasting for coherent light generator technology based on patent citation network analysis. scientometrics 111(1), 297-315. https://doi.org/10.1007/s11192-017-2252-y int. j. prod. manag. eng. (2019) 7(special issue), 73-79creative commons attribution-noncommercial-noderivatives 4.0 international analysis of technological knowledge flows in the basque country 79 https://doi.org/10.1016/s0048-7333(01)00184-6 https://doi.org/10.1080/00130095.2016.1205947 https://doi.org/10.1080/00130095.2016.1205947 https://doi.org/10.1007/s11192-007-0101-0 https://doi.org/10.1007/978-3-319-96005-0_28 https://doi.org/10.1504/ijtm.2015.072976 https://doi.org/10.1016/j.respol.2003.10.001 https://doi.org/10.1103/physreve.95.042309 https://doi.org/10.1257/aer.90.2.215 https://doi.org/10.1093/oxfordjournals.cje.a035308 https://doi.org/10.1080/09654319708720382 https://doi.org/10.1162/rest_a_00422 https://doi.org/10.1086/230994 https://doi.org/10.3152/147154403781776708 https://doi.org/10.3152/147154403781776708 https://doi.org/10.1080/0034340032000108714 https://doi.org/10.1016/j.techfore.2015.09.017 https://doi.org/10.1257/0002828053828509 https://doi.org/10.1162/rest.88.2.383 https://doi.org/10.1007/s11192-017-2252-y http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.13683 received: 2020-05-13 accepted: 2020-12-27 identifying and classifying attributes of packaging for customer satisfaction a kano model approach dash. s.k. balaji institute of telecom & management, sri balaji university, pune, india. prof.sanjit@gmail.com abstract: the packaging industry in india is predicted to grow at 18% annually. in recent years packaging becomes a potential marketing tool. the marketer should design the packaging of high quality from customer perspective. as the research in the area of packaging is very few, study of quality attributes of packaging is the need of the hour and inevitable. an empirical research was conducted by applying kano model. the researcher is interested to find out the perception of the customers on 22 quality attributes of packaging. 500 respondents which were selected randomly were asked about their experience of packing on everyday commodities through a well-structured questionnaire. the classification of attribute as must-be quality, one-dimensional quality, attractive quality, indifferent quality and reverse quality was done by three methods. marketer should make a note of it and prioritise the attributes for customer satisfaction. key words: attributes, classification, kano model, packaging, satisfaction. 1. introduction the packaging industry in india is predicted to grow at 18% annually, with flexible packaging growing at 25% and rigid packaging at 15%. packaging is the fifth largest sector in india’s economy and is one of the highest growth sectors in the country. according to the packaging industry association of india (piai), the sector is growing at 22% to 25% per annum. a study by trade and commerce trade association assocham and global consulting firm ey revealed that the packaging industry in india is anticipated to reach $73.6bn by the 2020 financial year (fy2020), due to india’s growing population and income levels. in recent years packaging becomes a potential marketing tool. it’s considered as one of the important p of marketing mix. it’s considered as “silent salesman” and “five second commercial”. the marketer should design the packaging of high quality from customer perspective. as the research in the area of packaging is very few, study of quality attributes of packaging is the need of the hour and inevitable. an empirical research was conducted by applying kano model. the researcher is interested to find out the perception of the customers on 22 quality attributes of packaging. 500 respondents which were selected randomly were asked about their experience of packing on everyday commodities through a wellstructured questionnaire. three approaches to kano model are used to categorize the quality attributes as must-be quality, one-dimensional quality, attractive quality, indifferent quality and reverse quality. marketer could classify and prioritise the attributes into 8 must-be, 6 one-dimensional, 5 attractive, 2 indifferent and 1 reverse quality.marketer should make a note of it and prioritise the attributes for customer satisfaction to cite this article: dash. s.k. (2021). identifying and classifying attributes of packaging for customer satisfaction a kano model approach. international journal of production management and engineering, 9(1), 57-64. https://doi.org/10.4995/ijpme.2021.13683 int. j. prod. manag. eng. (2021) 9(1), 57-64creative commons attribution-noncommercial-noderivatives 4.0 international 57 https://orcid.org/0000-0002-4152-4703 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. packaging – a glimpse packaging is considered to be one of the important element of marketing. now its gaining lot of importance whereas during 1950s it saw considered as indifferent. in this present era it gains lot of importance as it is providing different dimensions like functional, technical, informative and visual. customers want solutions to their problem. so, packaging should not be considered as a mere container rather it should be considered as one of the important “p” in the marketing mix. as per the research, purchase decision is generally taken by the customers within seconds. so it acts like a “silent salesman”. generally the product quality is ascertained by packaging. so packaging should be good from functional, technical, informative and visual. 3. theory of kano model kano model was developed by prof noriaki kano in 1984 which classify the attributes of a product or service into five categories. the categories are as shown in figure 1. 3.1. must-be quality (m): these are all essential attributes of a product. if these attributes are there, they may not influence the customer to go for the product, but if it’s not there, definitely the customers will be dissatisfied which ultimately results into rejection of the offer. 3.2. one-dimensional quality (o): these are the attributes responsible for lot of satisfaction because of its availability and creates lots of dissatisfaction because of its non-availability or when it is not fulfilled by the marketer. 3.3. attractive quality (a): these are the attributes generally delights the customers. availability of these attributes creates lot of satisfaction but non availability of these attributes do not create any dissatisfaction. these are the attributes used to differentiate the products from the competitors and creates a competitive advantage for the marketer. 3.4. indifferent quality (i): availability or non-availability of these attributes are not going to have any impact on customer satisfaction and dissatisfaction. customers are indifferent to these attributes. marketers should avoid these attributes. 3.5. reverse quality (r): lower the fulfilment of these attributes higher the satisfaction and vice versa. 4. methodology empirical investigation was undertaken to assess the quality attributes of packaging. 500 respondents were selected through random sampling. questionnaire was administered to find out the importance of different quality attributes as well as their category as per the kano model. ! customer satisfaction customer dissatisfaction requirement unfulfilled requirement fulfilled must be one-dimensional attractive indifferent reverse figure 1. kano model. int. j. prod. manag. eng. (2021) 9(1), 57-64 creative commons attribution-noncommercial-noderivatives 4.0 international dash 58 https://en.wikipedia.org/wiki/noriaki_kano http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.1. dimensions of attributes in packaging the author tries to study the packaging from four dimension. the dimensions are functional, technical, informative and visual. a. functional: it contains all the attributes pertaining to the functional aspects of packaging. 8 out of 22 attributes such as easy to grip, easy to open, easy to empty completely, easy to throw in the waste, fit in storage spaces, userfriendly, weight and additional functions were studied under this dimension. b. technical: it contains all the attribute pertaining to technical aspect of packaging. 5 out of 22 attributes such as hygienic, leakage proof, protection, recyclable material and resealability were studied under this dimension. c. informative: it contains all the attribute pertaining to communication aspects of packaging. 5 out of 22 attributes such as date of manufacturing, declaration of contents, instructions, url and customer care number were studied under this dimension. d. visual: it contains all the attributes pertaining to visual and branding aspects of packaging. 4 out of 22 attributes such as aesthetically appealing, appearance, brand name and symbols were studied under this dimension. kano model is used to assess the attributes that influence customer’s purchase decision and classify customer requirements to enhance performance of the product/service. following steps were undertaken to administer the kano model for assessing and classifying the attributes of packaging. step-i it is vital to identify and classify the key attributes of packaging. lots of attributes of packing which are crucial from customer satisfaction point of view were identified from different literature and reviewed. these are reflected in the questionnaire. step-ii surveying were done by administering questionnaire to 500 respondents in pune city, india. the respondents were finalised by random sampling method. the questionnaire contains 3 parts namely part “a”, part “b” and part “c”. part “a” contains all demographic information. part “b” contains kano model questionnaire and part “c” contains attribute rating scale where respondents were asked to rate the importance of attribute on a scale of 1 to 10,1 being least important and 10 being most important. kano model questionnaire is unique in nature and contains a pair of questions namely functional and dysfunctional for each attribute to ascertain its category. functional questions were designed in a positive way and dysfunctional questions were designed in a negative way. both functional and dysfunctional question has five options namely (a) like, (b) must be, (c) neutral, (d) live with and (e) dislike. the respondents were asked to choose one option each from functional and dysfunctional question. in this research we have taken a total of 22 questions pertaining to 4 dimensions of the packaging. a sample of both functional and dysfunctional question which are used in the questionnaire is illustrated below in table 1. table 1. example of functional and dysfunctional question. functional question response 1 a. packaging offers additional functions 1. like 2. must be 3. neutral 4. live with 5. dislike dysfunctional question response 1b. packaging does not offer additional functions 1. like 2. must be 3. neutral 4. live with 5. dislike step-iii a test run of questionnaire was done to avoid the confusion of respondents. when we found some confusion, the questions were revised and tested again. we have used kano evaluation table (table 2) to categories the response of individual respondent into different category. an attribute is classified as “must be (m)”, if the response is “must be” or “neutral” or “live with” for a functional question and “dislike” for a dysfunctional question. an attribute is classified as “one -dimensional (o)”, if the response is “like it” for a functional question and “dislike” for a dysfunctional question. an attribute is classified as “attractive (a)”, if the response is int. j. prod. manag. eng. (2021) 9(1), 57-64creative commons attribution-noncommercial-noderivatives 4.0 international identifying and classifying attributes of packaging for customer satisfaction a kano model approach 59 http://creativecommons.org/licenses/by-nc-nd/4.0/ “like” for a functional question and “must be” or “neutral” or “live with” for a dysfunctional question. an attribute is classified as “indifferent (i)”, if the responses is “must be” or “neutral” or “live with” for a functional question and “must be” or “neutral” or “live with” for a dysfunctional question, an attribute is classified as “reverse (r)”, if the response is “dislike” for a functional question and “like” or “must be” or “neutral” or “live with” for a dysfunctional question. an attribute also classified as “reverse (r)”, if the response is “must be” or “neutral” or “live with” for a functional question and “like” for a dysfunctional question. an attribute is classified as “questionable (q)”, if the response is “like” for both functional question as well as dysfunctional question. an attribute is also classified as “questionable (q)”, if the response is “dislike” for both functional question as well as dysfunctional question. step-iv based on the response given by the respondents, the classification of attributes were done by following methods (1) frequency-based attributes classification method: this method of classification is based on the frequency of response. classification of a particular attribute is based on the maximum frequency of response (m, o, a, i, r, q). (2) comparison-based attribute classification method: this method of classification suggests that for an attribute if summation of the frequency of m, o, a is greater than the summation of the frequency of i, r, q, then the attribute is classified among m, o, a which is having highest frequency amongst them. if summation of the frequency of i, r, q is greater than the summation of the frequency of m, o, a then the attribute is classified among i,r,q which is having highest frequency amongst them. if summation of m, o, a and summation of i, r, q are same, then the attribute is classified based on the priority order defined by matzler et al. (1996) i:e m > o > a > i. (3) index-based attribute classification method: this method suggests two indices namely satisfaction index and dissatisfaction index. satisfaction index (si) = (a+o)/ (a+o+m+i) which varies from 0 to 1 and dissatisfaction index (di) = (m+o) / (a+o+m+i)*(-1) which varies from -1 to 0. the satisfaction index and dissatisfaction index of 22 attributes are plotted in a diagram to get an overview. different attributes were classified based on satisfaction and dissatisfaction index as described in table 3. table 3. index based attribute classification. satisfaction index (si) dissatisfaction index(di) classification < 0.5 ≥ 0.5 must-be ≥ 0.5 ≥ 0.5 one-dimensional ≥ 0.5 < 0.5 attractive < 0.5 < 0.5 indifferent step-v category strength (cs) and total strength (ts) are the two measurements of attributes introduced by lee and newcomb in 1997. cs is the difference of the percentage of response between highest category and next highest category. example: suppose for an attribute “o” is the highest category having 45.5% and “a” is the next highest category having 25.5%. then cs=45.5%-25.5%=20. table 2. kano evaluation table. customer response functional like must be neutral live with dislike d ys fu nc tio na l like question (q) reverse (r) reverse (r) reverse (r) reverse (r) must be attractive (a) indifferent (i) indifferent (i) indifferent (i) reverse (r) neutral attractive (a) indifferent (i) indifferent (i) indifferent (i) reverse (r) live with attractive (a) indifferent (i) indifferent (i) indifferent (i) reverse (r) dislike one-dimensional (o) must be (m) must be (m) must be (m) question (q) int. j. prod. manag. eng. (2021) 9(1), 57-64 creative commons attribution-noncommercial-noderivatives 4.0 international dash 60 http://creativecommons.org/licenses/by-nc-nd/4.0/ ts is the total percentage of response in the three category like must-be (m), one-dimensional (o) and attractive (a). example: suppose for an attribute “m” is 25%, “o” is 35% and “a” is 10%. then ts=25%+35%+10%=70. 5. analysis of result twenty two packaging attributes were identified by summarizing relevant literatures and by taking reality into consideration which contributes towards customer satisfaction. they broadly have 4 dimensions namely functional which contains 8 attributes, technical which contains 5 attributes, informative which contains 5 attributes and visual which contains 4 attributes. four dimensions with 22 attributes shown in table 4. based on the response mentioned in table 4, category, satisfaction index (si), dissatisfaction index (di), category strength (cs) and total strength (ts) are estimated and presented in table 5. the category of attributes were estimated by frequency, comparison and index based method and overall category of 22 attributes were decided for 22 attributes. all the attributes were found to be in the same category in all the three methods except weight which is falling in reverse category in frequency and comparison based method but falls in indifferent category in index based method. as out of three it falls in reverse category in two methods, so overall category will be reverse only. out of 22, 8 attributes like easy to open, hygienic, leakage proof, protection, date of manufacturing, declaration of contents, instructions, appearance are in must be, 6 attributes like easy to grip, easy to empty completely, easy to throw in the waste, user friendly, communicates quality, symbols are in one dimensional, 5 attributes like fit in storage spaces, recyclable material, resealability, customer care number, aesthetically appealing are in attractive, 2 attributes like additional functions and url are in indifferent and 1 attribute like weight is in reverse category. attribute strength of all 22 attributes were table 4. dimensions and attributes of packaging. dimension assessed attributes a o m i r q total functional easy to grip 91 260 121 28 0 0 500 easy to open 26 222 230 22 0 0 500 easy to empty completely 42 245 172 41 0 0 500 easy to throw in the waste 115 289 54 42 0 0 500 fit in storage spaces 212 205 14 69 0 0 500 user-friendly 46 271 162 21 0 0 500 weight 17 23 41 79 340 0 500 additional functions 240 2 5 253 0 0 500 technical hygienic 34 182 261 23 0 0 500 leakage proof 28 123 332 17 0 0 500 protection 25 145 251 79 0 0 500 recyclable material 205 148 55 92 0 0 500 resealability 298 137 22 43 0 0 500 informative customer care number 287 142 25 46 0 0 500 date of manufacturing 17 193 269 21 0 0 500 declaration of contents 11 140 300 49 0 0 500 instructions 57 174 210 59 0 0 500 url 231 7 3 259 0 0 500 visual aesthetically appealing 228 44 9 219 0 0 500 appearance 57 145 231 67 0 0 500 communicates quality 137 158 103 102 0 0 500 symbols 49 221 176 54 0 0 500 int. j. prod. manag. eng. (2021) 9(1), 57-64creative commons attribution-noncommercial-noderivatives 4.0 international identifying and classifying attributes of packaging for customer satisfaction a kano model approach 61 http://creativecommons.org/licenses/by-nc-nd/4.0/ calculated by summating all the response(in a scale of 0 to 10) of 500 respondents and dividing 500 are also presented in table 5. attributes like leakage proof (9.74) and date of manufacturing (9.64) are most important in must be category. attributes like user friendly (8/91) and easy to grip (8.59) are most important in one-dimensional category. attributes like recyclable material (7.63) and resealability (7.44) are most important in attractive category. both the attributes of indifferent category like additional functions (5.25) and url (5.09) are having low strength. attribute weight (1.81) is having the lowest strength which falls in reverse category. 6. conclusion packaging plays a vital role in marketing. it creates the first impression in the retail outlet. it also affects the customer’s perception about the quality of the product after purchase. here the attributes are classified using kano model. broadly, packaging attributes are divided into 4 dimensions i.e. functional, technical, informative and visual. functional dimension contains 8 attributes predominantly one dimensional which indicates that if it will be complied, the satisfaction will increase and if not satisfaction will go down. technical dimension contains 5 attributes predominantly of must be category which indicates that it has to be fulfilled otherwise customer will defect the product and go to competitor products. informative dimension contains 5 attributes, majority of which are of must be category which has to be fulfilled at first priority otherwise it will hamper the sales of the product. visual dimension contains 4 attributes predominantly of one dimensional category presence of which enhance customer satisfaction and absence will lead to customer dissatisfaction. as priority order defined by matzler et al. (1996) i.e. m > o > a > i., i will be a and a will be o and o will be m subsequently over the product life cycle as reported by kano (2001). by using this model, marketer can prioritise the attributes and try to fulfil all must be quality attributes specifically table 5. estimation of category, sc, dc, cs, ts and attibute strength of attibutes of packaging. dimension assessed attributes category si di cs ts attribute strength frequencybased comparisionbased indexbased overall functional easy to grip o o o o 0.702 -0.762 27.8 94.4 8.59 easy to open m m m m 0.496 -0.904 1.6 95.6 9.01 easy to empty completely o o o o 0.574 -0.834 14.6 91.8 8.71 easy to throw in the waste o o o o 0.808 -0.686 34.8 91.6 8.02 fit in storage spaces a a a a 0.834 -0.438 1.4 86.2 7.09 user-friendly o o o o 0.634 -0.866 21.8 95.8 8.91 weight r r i r 0.25 -0.4 52.2 16.2 1.81 additional functions i i i i 0.484 -0.014 2.6 49.4 5.25 technical hygienic m m m m 0.432 -0.886 15.8 95.4 9.11 leakage proof m m m m 0.302 -0.91 41.8 96.6 9.74 protection m m m m 0.34 -0.792 21.2 84.2 9.56 recyclable material a a a a 0.706 -0.406 11.4 81.6 7.63 resealability a a a a 0.87 -0.318 32.2 91.4 7.44 informative customer care number a a a a 0.858 -0.334 29 90.8 7.05 date of manufacturing m m m m 0.42 -0.924 15.2 95.8 9.64 declaration of contents m m m m 0.302 -0.88 32 90.2 9.41 instructions m m m m 0.462 -0.768 7.2 88.2 8.74 url i i i i 0.476 -0.02 5.6 48.2 5.09 visual aesthetically appealing a a a a 0.544 -0.106 1.8 56.2 6.02 appearance m m m m 0.404 -0.752 17.2 86.6 8.61 communicates quality o o o o 0.59 -0.522 4.2 79.6 7.29 symbols o o o o 0.54 -0.794 9 89.2 8.26 int. j. prod. manag. eng. (2021) 9(1), 57-64 creative commons attribution-noncommercial-noderivatives 4.0 international dash 62 http://creativecommons.org/licenses/by-nc-nd/4.0/ easy to open, hygienic, leakage proof, protection, date of manufacturing and declaration of contents. packing attributes should be competitive enough in one dimensional category such as easy to grip, easy to empty completely and userfriendly. attractive category like recyclable material and resealability should be given importance in packaging to delight the customers. marketer should not invest in additional functions as it’s found to be indifferent. weight found to be reverse category, so marketer should try to minimise the weight as much as possible. so its concluded that packaging plays a vital element in the marketing mix and the attributes should be considered judiciously based on the priority. references bakhitar, a.,hannan, a., basit, a., ahmad, j.(2015). prioritization of value based services of software by using ahp and fuzzy kano model. international conference on computational and social sciences, 8, 2527. basfirinci, c., mitra, a. (2015). a cross cultural investigation of airlines service quality through integration of servqual and the kano model. journal of air transport management, 42(1), 239-48. https://doi.org/10.1016/j.jairtraman.2014.11.005 berger, c., blauth, r., boger, d., bolster, c., burchill, g., dumouchel, w., pouliot, f., richter, r., rubinoff, a., shen, d., timko, m., walden, d. (1993). kano’s methods for understanding customer-defined quality. the center for quality of management journal, 2(4), 2–36. brown, g.h. (1950). measuring consumer attitudes towards products. journal of marketing, 14(5), 691-98. https://doi. org/10.1177/002224295001400505 chaudha, a., jain, r., singh, a.r., mishra, p.k. (2011). integration of kano’s model into quality function deployment (qfd). journal advice manufacture technology, 53, 689–698. https://doi.org/10.1007/s00170-010-2867-0 cole, r.e. (2001). from continuous improvement to continuous innovation. quality management journal, 8(4), 7-21. https://doi.org/10.10 80/10686967.2001.11918977 dash, s.k. (2019). application of kano model in identifying attributes. a case study on school bus services. international journal of management studies, 6(1), 31-37. https://doi.org/10.18843/ijms/v6i1(3)/03 dziuba, s.t., śron, b. (2014). fam-fmc system as an alternative element of the software used in a grain and flour milling enterprise. production engineering archives, 4(3),29-31. https://doi.org/10.30657/pea.2014.04.08 ernzer, m., kopp, k.(2003). application of kano method to life cycle design. ieee proceedings of eco design: third international symposium on environmentally conscious de-sign and inverse manufacturing, tokyo japan, december 8-11, 383-389. https://doi.org/10.1109/ecodim.2003.1322697 feigenbaum, a.v. (1991).total quality control. mcgraw-hill. fundin, a., nilsson, l. (2003). using kano’s theory of attractive quality to better understand customer satisfaction with e-services. asian journal on quality, 4(2), 32-49. https://doi.org/10.1108/15982688200300018 friman, m., edvardsson, b. (2003). a content analysis of complaints and compliments. managing service quality, 13(1), 20-26. https://doi. org/10.1108/09604520310456681 garvin, d.a. (1987). competing on the eight dimensions of quality. harvard business review, 65(6), 101-109. hanan, m., karp, p. (1989). customer satisfaction, how to maximise, measure and market your company’s “ultimate product”. amacom. herzberg, f., bernard, m., snyderman, b.b. (1959). the motivation to work. john wiley and sons. hoch, s.j., ha, y.w. (1986). consumer learning: advertising and the ambiguity of product experience. journal of consumer research, 13, 221-33. https://doi.org/10.1086/209062 johnson, m.d., nilsson, l. (2003). the importance of reliability and customization from goods to services. quality management journal, 10(1), 8-19. https://doi.org/10.1080/10686967.2003.11919049 kano, n., seraku, n., takahashi, f., tsuji, s. (1984). attractive quality and must-be quality. journal of the japanese society for quality control, 41, 39-48. kapalle, p.k, lehmann, d.r. (1995). the effects of advertised and observed quality on expectations about new product quality. journal of marketing research, 32(8), 280-90. https://doi.org/10.1177/002224379503200304 lee, m.c., newcomb, j.f. (1997). applying the kano methodology to meet customer requirements: nasa’s microgravity science program. quality management journal, 4(3), 95-110. https://doi.org/10.1080/10686967.1997.11918805 löfgren, m. (2005). winning at the first and second moments of truth: an exploratory study. journal of service theory and practice, 15(1), 102-15. https://doi.org/10.1108/09604520510575290 löfgren, m., witell, l. (2005). kano’s theory of attractive quality and packaging. quality management journal, 12(3), 7-20. https://doi.org/10.1080/10686967.2005.11919257 int. j. prod. manag. eng. (2021) 9(1), 57-64creative commons attribution-noncommercial-noderivatives 4.0 international identifying and classifying attributes of packaging for customer satisfaction a kano model approach 63 https://doi.org/10.1016/j.jairtraman.2014.11.005 https://doi.org/10.1177/002224295001400505 https://doi.org/10.1177/002224295001400505 https://doi.org/10.1007/s00170-010-2867-0 https://doi.org/10.1080/10686967.2001.11918977 https://doi.org/10.1080/10686967.2001.11918977 https://doi.org/10.18843/ijms/v6i1(3)/03 https://doi.org/10.30657/pea.2014.04.08 https://doi.org/10.1109/ecodim.2003.1322697 https://doi.org/10.1108/15982688200300018 https://doi.org/10.1108/09604520310456681 https://doi.org/10.1108/09604520310456681 https://doi.org/10.1086/209062 https://doi.org/10.1080/10686967.2003.11919049 https://doi.org/10.1177/002224379503200304 https://doi.org/10.1080/10686967.1997.11918805 https://doi.org/10.1108/09604520510575290 https://doi.org/10.1080/10686967.2005.11919257 http://creativecommons.org/licenses/by-nc-nd/4.0/ matzler, k., hinterhuber, h.h., bailom, f., sauerwein, e. (1996). how to delight your customers. journal of product & brand management, 5(2), 6-18. https://doi.org/10.1108/10610429610119469 miarka, d., żukowska, j., siwek, a., nowacka,a., nowak, d. (2015). microbial hazards reduction during creamy cream cheese production. production engineering archives, 6(1), 39-44. https://doi.org/10.30657/pea.2015.06.10 nelson, p. (1970), information and consumer behaviour. journal of political economy, 78, 311-29. https://doi.org/10.1086/259630 nilsson-witell, l, fundin, a. (2005). dynamics of service attributes: a test of kano’s theory of attractive quality. international journal of service industry management, 16(2), 152-168. https://doi.org/10.1108/09564230510592289 parasuraman, a. (1997). reflections on gaining competitive advantage through customer value. academy of marketing science journal, 25(2), 154-61. https://doi.org/10.1007/bf02894351 parasuraman, a., colby, c.l. (2001). techno-ready marketing. free press. qiting, p., uno, n., kubota, y. (2013). kano model analysis of customer needs and satisfaction at the shanghai disneyland. in proceedings of the 5th intl congress of the intl association of societies of design research, tokyo, japan. http://design-cu.jp/iasdr2013/papers/1835-1b.pdf accessed on january 2021. sauerwein, e., bailom, f., matzler, k., hinterhuber, h.h. (1996). the kano model: how to delight your customers. volume i of the ix. international working seminar on production economics, innsbruck/igls/austria, february 19-23 1996, pp. 313-327. https://is.muni. cz/el/econ/podzim2009/mph_mar2/um/9899067/the_kano_model_-_how_to_delight_your_customers.pdf shewhart, w.a. (1931). economic control of quality of manufactured product. d. van nostrand company, inc. underwood, r.l., klein, n.m. (2002). packaging as brand communication: effects of product pictures on consumer responses to the package and brand. journal of marketing theory and practice, 10(4), 58-68. https://doi.org/10.1080/10696679.2002.11501926 underwood, r.l. klein, n.m., burke, r.r. (2001). packaging communication: attentional effects of product imagery. journal of product & brand management, 10(7), 403-22. https://doi.org/10.1108/10610420110410531 watson, g.h. (2003), “customer focus and competitiveness”, in stephens, k.s. (ed.), six sigma and related studies in the quality disciplines, asq quality press, milwaukee, wi. williams, d. (2020). the future of the packaging industry in india. packaging gateway. https://packaging-gateway.com/features/futurepackaging-industry-in-india accessed on january 2021. williams,h., wikström,f., löfgren.m. (2008). a life cycle perspective on environmental effects of customer focused packaging development.” journal of cleaner production, 16(7), 853-859. https://doi.org/10.1016/j.jclepro.2007.05.006 woodruff, r.b. (1997). customer value: the next source for competitive advantage. journal of academy of marketing science, 25(2), 13953. https://doi.org/10.1007/bf02894350 zeithaml, v.a. (1988). consumer perceptions of price, quality, and value: a means-end model and synthesis of evidence. journal of marketing, 52, 2-22. https://doi.org/10.1177/002224298805200302 int. j. prod. manag. eng. (2021) 9(1), 57-64 creative commons attribution-noncommercial-noderivatives 4.0 international dash 64 https://doi.org/10.1108/10610429610119469 https://doi.org/10.30657/pea.2015.06.10 https://doi.org/10.1086/259630 https://doi.org/10.1108/09564230510592289 https://doi.org/10.1007/bf02894351 http://design-cu.jp/iasdr2013/papers/1835-1b.pdf https://is.muni.cz/el/econ/podzim2009/mph_mar2/um/9899067/the_kano_model_-_how_to_delight_your_customers.pdf https://is.muni.cz/el/econ/podzim2009/mph_mar2/um/9899067/the_kano_model_-_how_to_delight_your_customers.pdf https://doi.org/10.1080/10696679.2002.11501926 https://doi.org/10.1108/10610420110410531 https://packaging-gateway.com/features/future-packaging-industry-in-india https://packaging-gateway.com/features/future-packaging-industry-in-india https://doi.org/10.1016/j.jclepro.2007.05.006 http://woodruff https://doi.org/10.1007/bf02894350 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2019.10911 received: 2019-01-22 accepted: 2019-07-29 developing a novel based productivity model by investigating potential bounds of production plant hussain, z. sarhad university of science and information technology, peshawar, pakistan. zahid.btech@suit.edu.pk abstract: productivity level is based on reliability impression which is the primary aspect of the automatic assembly line for continuous production. productivity forecasting is a professional tool helping to enhance production system and attain the client petition by using precise model. due to mechanisms complexity of assembly lines, analysis of failure factors contributes a significant role for investigating potential bounds that require analytical approach to compare the current and proposed model of productivity effects. the issues related to the production losses need additional space for improvement of the productivity model which may not present a close comparison between the current and proposed productivity rate. the main purpose of this paper is to develop a novel based productivity model that will predict alternatives for the availability of assembly line workspaces pertain to an automobile tire manufacturing plant. for investigating the potential bounds of the productivity losses, dmaic and pace techniques were used. it was revealed that the novel productivity model yielded better results of 3.358% errors showing its accuracy as compared to real productivity level at different workspaces. key words: manufacturing system, automatic assembly lines, tire productivity, dmaic. 1. introduction productivity is a crucial indicator to point out the performance in a manufacturing unit. there are some approaches to precise productivity(gharfalkar, ali, & hillier, 2018). it can be articulated using different insights, prototypical variables, speculative outline and financing procedures(edgar & pistikopoulos, 2018). from the outline of the operations research, there are three different viewpoints about productivity extent which are costeffective, engineering, and manufacturing(singh, singh, & sharma, 2018). a high rate of production means the involvement of extra manpower which yields high productivity. according to cost-effective productivity perception productivity is the ratio of outputs to the inputs as expressed in relation 1 (s. k. gupta, gupta, & dhamija, 2019). figure 1 illustrates the characterization of productivity in which the input resources are transformed into desired outputs. productivity goodsproduced input used= (1) manufacturing and consumption are the two major concerns of the productivity. figure 1. cost-effective system value adds by transforming inputs to outputs. to cite this article: hussain, z. (2019). developing a novel based productivity model by investigating potential bounds of production plant. international journal of production management and engineering, 7(2), 151-159. https://doi.org/10.4995/ijpme.2019.10911 int. j. prod. manag. eng. (2019) 7(2), 151-159creative commons attribution-noncommercial-noderivatives 4.0 international 151 https://orcid.org/0000-0002-5671-7030 http://creativecommons.org/licenses/by-nc-nd/4.0/ during the production process input and output are transformed however consumption is coverging on the practice of processing plants and organization (giovannetti & piga, 2017). however, productivity in the cost-effective aspect is centering on variables like quality, quantity and cost of the final good produced (nawanir, fernando, & teong, 2018). the engineering perspective of productivity requires more impost due to the involvement of industrial equipment and technological processes (fettermann, cavalcante, almeida, & tortorella, 2018). productivity in this case, is based on the best techniques to the extent of the overall equipment effectiveness. overall equipment effectiveness (oee) is extensively recognized and consistently used as a quantifying tool for measuring the production capacity of single equipment in industries (hussain, 2018). owing to advancements in the technological process, production systems and developments are converting into more complex and automated assimilation which seriously concerned with equipment effectiveness followed by innumerable practices of human control (dresch, veit, lima, lacerda, & collatto, 2019). the analysis and modeling of productivity for such complex systems are becoming a competitive contest for concerned engineers as well as researchers in academic spheres (yan et al., 2018). expression 2 is considered to present oee as a mathematical model on the basis of availability level (a), performance (p) and quality level (q): oee=(a)×(p)×(q) (2) according to oee model, the availability level (a) is crucial for productivity which reflects the diminishing of production time while performance (p) reflects the actual production time delays and the quality level (q) reflects the satisfactory quality values (yazdi, azizi, & hashemipour, 2019). the manufacturing viewpoint of productivity emphases and concentrating on the level of production of a specific workspace within the industry which follows the constraints of quantity requirements and the time of producing goods (xu, 2017). the study of the productivity model of different workspaces is imperative in modern assembly lines as it enables the estimation of a certain manufacturing system (morales méndez & rodríguez, 2017). however modern industries use fully automated production lines because of the tough industrial competition, rapid demands and complication in products characteristics (bauerdick, helfert, petruschke, sossenheimer, & abele, 2018). the need of applied research about modeling for such complex automated processing lines and workspaces is extremely imperative to be revealed (hussain, 2019). regarding the role of the productivity model in processing units, expression 3 is normally used for measuring productivity in automated production lines. productivity (q) produced timefor production(t) parts (n) = (3) this model determines the level of productivity for the amount of the goods produced and the corresponding time consumed to produce those goods in a specific workspace and processing assembly line (manitz, 2008). besides establishing productively modeling, an ordinary level of reliability has already been industrialized. considering the operating time of the product (top ), auxiliary time for loading a piece in machine area (tax ) with number of working spaces (x), mean time to repair (tmr ), rate of the failure of a workspace (λw ), rate of the failure of transportation (λ(t)) and (λ(cs)) as rate of the failure of transportation system of the whole unit and finally (λ(d)) as the failure frequency of defective parts then, the model takes in account a sequential linear part as shown in equation 4. t x t t x 1 1 1 productivity (q ) ( ) ( )( ) op ax mr w t dcs ar # m m m m = + + + + + ^ ^ h h" , (4) the productivity model under study using system availability seems not to be precise to present authentic productivity rate due to limitations of some of the productivity constraints. therefore, those potential bounds are necessary to be examined before to discover a true productivity model. 2. methodology to explore an accurate productivity model, the required potential bounds are investigated using dmaic approach which is a famous data-driven quality strategy for improving the manufacturing processes and pace matrix technique used for prioritizing various production task. dmaic approach addresses on the quality issues for refining progressions which is fundamental fragment of the six sigma quality initiative (neha gupta, 2013). dmaic approach establishes five consistent levels including define, measure, analyze, improve, and int. j. prod. manag. eng. (2019) 7(2), 151-159 creative commons attribution-noncommercial-noderivatives 4.0 international hussain, z. 152 http://creativecommons.org/licenses/by-nc-nd/4.0/ control. each level in the process is essential to achieve the optimum expected outcomes (“study and analysis the wastage reduction of fluorescent powder in cfl 23 w in philips pvt ltd mohali, using six sigma methodology,” 2016). on the other hand, pace prioritization matrix which was formed by karen martin group, is a lean engineering tool established to explore the productivity parameters for increasing the production efficiency (sin et al., 2014). to improve the productivity model using system availability using lean as well as sigma techniques, pace matrix is introduced during the improvement stage of dmaic approach to identify the required factors. the dmaic methodology is based on defining the issues related to the enhancement of the production parameters for presenting an effective productivity model. this methodology is shown in figure 2 in the form of a flowchart. figure 2. dmaic approach flowchart for exploring productivity parameters define: to state and outline the most negative aspect of the model, the productivity model in machinecontrolled and production line with the accessibility of productivity is needed. however, based on the two productivity equations discussed above, it should be clear enough to point out the product comparison. measure: for measuring the difference and analyzing the comparison between the current and desired productivity parameters related to actual and reliability approaches, both the productivity equations are used through operational order. after applying the equations (6) and (7) the computation will show the nonconformity results of the productivity model by comparing the current system reliability and actual productivity performance. analyze: this step helps to determine the gap between the current and goal performance through contributing opportunity to analyze the difference between the model of productivity and availability for which loss diagram has been considered. by the implementation of loss diagram the factors necessary for the productivity model which may be left ignored are re-analyzed for process improvement (v. gupta, jain, meena, & dangayach, 2018). improve: for improving the effectiveness of productivity model parameters, pace matrix technique is implemented to determine the significance of critical parameters (zhang et al., 2018). the output is categorized using different spaces of pace criteria. priority (highest anticipated value of productivity parameters), action (slightly lower strength with relatively restrained influence of productivity parameters), consider (next to p and a implementation, reviewing the process to discover the difficulties occurred during consideration of the model parameters with great influence) and eliminate (those factors which do not create difficulties during model formulation). control: the control stage keeps the potential parameter of the productivity model from the development level and settles the required potential bounds for further improvement. present research work has been performed in united rubber industries dealing with the production of automobile tire and tube. all the necessary parameters for developing productivity model, different failures and productivity rates along with production losses were considered in the tire curing assembly hall. 3. results and discussion data assortment: for data analysis, three different arrangements of data sets were considered. the first set is the real data consisting of total working shifts, assembly time and tire produced. for this purpose, the data of two months (march and april 2018 with 60 working shifts) has been collected. during this period, total production of tire grade 205/55, r14 w16 was recorded as 7909 with 18000 minutes as production time. figure 3 presents the production of tires in 60 working shifts for the specified period. the next two data sets are related to technical and reliability concerns, however, the applied data to methodology stated above for determination of research’s results are described in tables 1 to 4. table 3 data values shows the output of tire produced in tire curing assembly hall. the result with dmaic approach are discussed below. int. j. prod. manag. eng. (2019) 7(2), 151-159creative commons attribution-noncommercial-noderivatives 4.0 international developing a novel based productivity model by investigating potential bounds of production plant 153 http://creativecommons.org/licenses/by-nc-nd/4.0/ define: with accordance to equation (1) productivity is the ratio of goods produced to the input used which follows that as per industrial constraints, productivity may likely be greater than methodological restrictions that is quantity of the goods and producing time, however, the real productivity is based on equation (3). implementing the reliability concept, the basic equation of productivity has been improved and presented in equation (4). measure: matching both the equations (3) and (4), the real productivity is based on set of additional methodical data applied on above equations for distinguishing the correctness of productivity model with the actual productivity. the detail of this necessary data is revealed in tables 1–4. total production time (t, min) as expressed in table 3 is computed as follows: production time (t)=top+tax=25 min+0.4 min=25.4 min (5) after setting the production time which is equal to the auxiliary time for loading a piece in the machine area, the actual productivity is determined through equation (3) while the findings are presented in table 3. finally, the changeover of data presented in tables 1 and 2 into equation (4) the required productivity model in average is presented. figure 4 shows the comparison of both fallouts of the actual and model productivity as per findings of the table 4. analyze: figure 4 reveals that the productivity model results through sole reliability are much progressive as compare to real productivity. subsequently, the main purpose of the model of productivity in the tire curing assembly unit is to predict the productivity rate with specific and accurate results, therefore the model understudy needs to be enhanced. further analysis is based on failure rate at different stages which are listed down and their values are calculated for the determination of productivity losses in tire curing assembly hall using the following additional concerns. potential productivity (pp ): the difference between the definite and most effective (optimal) productivity close. it is considered as unique productivity losses due to the lack of idle time and may be determined using the following expression. pp t x 1 op = ^ h (6) table 2. data description of reliability listing for tire curing assembly line. description workspace (x) failure frequency/minute (λoa) workspaces failure frequency (x, λoa) 1-4 0.0025 5-8 0.0150 9-12 0.010 13-16 0.0047 17-20 0.0080 21-24 0.0020 25-28 0.0030 29-32 0.0080 33-36 0.0090 37-40 0.0080 failure frequency of all workspaces oai 1 40 m = | 0.0702 average failure frequency of all workspaces 0.00702 rate of failure of transportation system of whole unit (λ(cs)) 0.0009 rate of failure of transportation system (λ(t)) 0.0004 failure rate of defective parts produced (λ(d)) 0.7445 mean time to repair (mtr), 1 minute table 1. data description of tire curing assembly line. description data values tire curing time (machine output), top (min) 25 auxiliary time for loading a piece in machine area, tax (min) 0.4 number of working spaces (x) 40 adjustment factor recommended for machine output time for bottleneck parameter of workspace (ƒad ) 1.30 int. j. prod. manag. eng. (2019) 7(2), 151-159 creative commons attribution-noncommercial-noderivatives 4.0 international hussain, z. 154 http://creativecommons.org/licenses/by-nc-nd/4.0/ recurring productivity (pr): this type of productivity presents the hidden losses of processing time without taking in account the downtime of the machine, however, in actual practice routine based machinery breakdowns and maintenance activities are the part of the normal production system. it is also noticeable that during normal production flow if a certain workspace starts malfunctioning due to any sort of mechanism failure then the whole assembly line is affected resulting in production time delay issues. recurring productivity is determined by using the following expression. figure 3. production of tires against available working shifts in tire curing assembly hall. figure 4. graphical comparison of real and model productivity. pr t x t 1 op ax = +^ h (7) bottleneck bound productivity (pb): it concentrates on the unfavorable circumstances which mostly exist at assembly line and lead to create other significant economic losses. therefore, the entire production process may be carefully examined to identify a bottleneck for minimizing its influence. taking in account the adjustment factor recommended for machine output time for bottleneck parameter of the workspace (ƒad), it may be determined as follows. pb t x t fd 1 op ax = + +^ h (8) workspace failure level productivity (pw): it concerns with reliability aspect of the workspaces related machines, taking into account the meantime to repair in case a breakdown occurs. these losses associated with this bound are computed as follows. pw t x t fd m 1 1 1 ( )op ax tr oai 1 40# m = + + + = ^ _h i| (9) transportation system failure level productivity (pt): expression (10) is used to determine the transportation system failure level productivity. due to different levels of reliability of each workspaces machines, the failures regarding transportation have to be determined to present a true picture of real productivity. p t x t fd m t 1 1 1 ( )op ax tr oai t1 40# m m = + + + + = ^ _h i| (10) productivity of transportation system failure of whole unit (ptw ): in case that the entire existing production assembly hall start malfunctioning then, the productivity effect of transportation structure can vary across other industrial sections. hence, the expression (11) is utilized in such circumstances. table 3. detail of actual productivity for tire curing assembly line. workspaces (x) 4 8 12 16 20 24 28 32 36 40 total parts produced (n) 0.8636 1.244 1.295 1.346 1.295 1.206 1.117 1.016 0.9144 0.8636 production time (t, min) 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 productivity (q, parts/min) 0.034 0.049 0.051 0.053 0.051 0.047 0.044 0.040 0.036 0.034 table 4. contrast of productivity model with actual productivity for tire curing assembly line. workspaces (x) 4 8 12 16 20 24 28 32 36 40 qac, (parts/min) 0.034 0.049 0.051 0.053 0.051 0.047 0.044 0.040 0.036 0.034 qmod, (parts/min) 0.117 0.181 0.218 0.239 0.252 0.258 0.260 0.260 0.258 0.256 int. j. prod. manag. eng. (2019) 7(2), 151-159creative commons attribution-noncommercial-noderivatives 4.0 international developing a novel based productivity model by investigating potential bounds of production plant 155 http://creativecommons.org/licenses/by-nc-nd/4.0/ p t x t fd m 1 1 1 ( ) tw op ax tr oai t cs1 40# m m m = + + + + + = ^ _h i| (11) defective parts productivity (pd): production defects explain defective goods produced through normal production. this happen in almost every sort of manufacturing unit that diminishes the rate of real productivity. it comprises argued fragments followed by those parts which may be reworked and considered as product quality forfeiture and may be determined using expression (12). p t x t fd m d 1 1 1 ( )op ax tr oai t cs d1 40# m m m m = + + + + + + = ^ _h i| (12) actual or unique productivity (pac): it is calculated using expression (13) after considering all the associated productivity losses due to issues concerned with management planning and scheduling that seriously affect the production schedules. actualproductivity (p ) production time(t) total partsproduced(n) ac = (13) productivity comparison: to compare the real productivity in tire curing assembly line with productivity model, the percentage error based strategy is conducted using the expression (14). (%)error p p p 100 ac d ac #= (14) the main purpose of comparison of both the productivities is to prove that the productivity model understudy would yield better and precise results likely to be (< 10%) errors when compared to the real productivity level. obviously the model productivity rate for different workspace failure levels showed a percentage error of 3.358%. henceforth it is confirmed and endorsed that the model of productivity level presents the maximum truthful fallouts that satisfy the criteria of (< 10%). for productivity bound’s improvement, the productivity losses are presented through productivity losses diagram developed on the basis of productivity equations (09-14) and shown in figure 5 while the corresponding fallouts are presented in table 5. the productivity losses parameter has been demonstrated in mathematical computation form in which l1 represents the productivity losses associated with auxiliary time showing 0.624 tires/min and this is the factor that has the higher impact of losses of the productivity in the final assembly. l2 is concerned with the bottleneck machining time and gives 0.150 tires/min, l3 is the loss of 0.054 tires/min owing to the workspace failure frequency, l4 is the influence of loss related to the controlling system and contributes 0.000647 tires/min, l5 is the lowermost contributor of the losses due to failure of the transportation system of the whole unit, l6 is 0.315555 tires/min that is the second uppermost of productivity losses of caused by various defects in raw material and defective parts produced and preceding is l7 which is due to the unexpected factor like power failure and contributes 0.014753 tires/min. improve: after conducting the analysis, the results highlighted various productivity-related issues that could be improved based on the group discussion held with the plant management team of united rubber industries. the group comprised of individuals involved in actual productivity of tire curing assembly line and the productivity model researchers. a maintainable improvement process is carefully selected using pace prioritization matrix as shown in figure 6 which is the outcome of the group discussion. the defective tire parameter is desired to be enhanced owing to its higher influence on productivity losses and for the reason that it is easy to improve as compared to the others productivity losses. control: control is the concluding juncture to establish the measures and examines identified potential productivity bounds for improvement. these bounds are suggested to be implemented in a prospect growth as a vigorous productivity models related to assembly lines. table 5. dissimilarity of productivity model with real one for tire curing assembly line. productivity losses due to tax l1 pp pr 1.600-0.975 0.624 parts/min losses due bottleneck bound productivity l2 pr pb 0.975-0.824 0.150 parts/min workspace failure level productivity losses l3 pb -pw 0.824-0.770 0.054 parts/min transportation system failure level productivity losses l4 pw -pt 0.770-0.769 0.000647 parts/min losses due to failure of transportation system of whole unit l5 ptptw 0.769-0.769 0.000288 parts/min defective parts productivity l6 ptw -pd 0.769-0.454 0.315555 parts/min actual or unique productivity l7 pd -pac 0.454-0.439 0.014753 parts/min int. j. prod. manag. eng. (2019) 7(2), 151-159 creative commons attribution-noncommercial-noderivatives 4.0 international hussain, z. 156 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. conclusion attributable to enhance productivity necessities in modern processing industries, the role of productivity model approach becomes more widespread for probable prediction and decision making. in this work, dmaic and pace techniques were applied for considering productivity model analysis in united rubber industries ltd with various productivity bounds that were computed to highlight their role for possible improvement in productivity rate. through this productivity model, developed for tire curing assembly line of the captioned unit, the output could be thoroughly and visually analyzed with an effective outlook. it is also has been verified that novel productivity model yielded much better results of 3.358% errors in productivity estimating which shows its accuracy as compared to real productivity level at different workspaces. current productivity model approach may be helpful for automatic manufacturing units based on continuous assembly figure 5. productivity losses diagram for tire curing assembly line. figure 6. pace prioritization matrix for productivity model. int. j. prod. manag. eng. (2019) 7(2), 151-159creative commons attribution-noncommercial-noderivatives 4.0 international developing a novel based productivity model by investigating potential bounds of production plant 157 http://creativecommons.org/licenses/by-nc-nd/4.0/ line facilities using the dmaic and pace analysis for investigating the potential manufacturing bounds that could improve their current manufacturing system. acknowledgements this research work has been expedited by sarhad university of science and information technology, peshawar, pakistan and uniter rubber industries. i am really thankful to my supporting colleagues from voronezh state university of engineering and technology, russian federation who provided awareness and proficiency that helped in completing the research. references bauerdick, c.j.h., helfert, m., petruschke, l., sossenheimer, j., & abele, e. (2018). an automated procedure for workpiece quality monitoring based on machine drive-based signals in machine tools. procedia cirp, 72, 357-362. https://doi.org/10.1016/j.procir.2018.03.245 dresch, a., veit, d.r., lima, p.n. de, lacerda, d.p., & collatto, d.c. (2019). inducing brazilian manufacturing smes productivity with lean tools. international journal of productivity and performance management, 68(1), 69-87. https://doi.org/10.1108/ijppm-10-2017-0248 edgar, t.f., & pistikopoulos, e.n. (2018). smart manufacturing and energy systems. computers and chemical engineering, 114, 130-144. https://doi.org/10.1016/j.compchemeng.2017.10.027 fettermann, d.c., cavalcante, c.g.s., almeida, t.d. de, & tortorella, g.l. (2018). how does industry 4.0  contribute to operations management? journal of industrial and production engineering, 35(4), 255-268. https://doi.org/10.1080/21681015.2018.1462863 gharfalkar, m., ali, z., & hillier, g. (2018). measuring resource efficiency and resource effectiveness in manufacturing. international journal of productivity and performance management, 67(9), 1854-1881. https://doi.org/10.1108/ijppm-11-2017-0282 giovannetti, e., & piga, c.a. (2017). the contrasting effects of active and passive cooperation on innovation and productivity: evidence from british local innovation networks. international journal of production economics, 187, 102-112. https://doi.org/10.1016/j. ijpe.2017.02.013 gupta, s.k., gupta, s., & dhamija, p. (2019). an empirical study on productivity analysis of indian leather industry. benchmarking, 26(3), 815-835. https://doi.org/10.1108/bij-06-2018-0156 gupta, v., jain, r., meena, m.l., & dangayach, g.s. (2018). six-sigma application in tire-manufacturing company: a case study. journal of industrial engineering international, 14(3), 511-520. https://doi.org/10.1007/s40092-017-0234-6 hussain, z. (2018). envisaging maintenance costs of hydraulic press in the ceramic industry with regression modelling. international journal of scientific & engineering research, 9(10), 1921-1928. retrieved from http://www.ijser.org hussain, z. (2019). statistical analyses of productivity model parameters for process improvement. advances in science and technology research journal, 13(2), 157-167. https://doi.org/10.12913/22998624/106240 manitz, m. (2008). queueing-model based analysis of assembly lines with finite buffers and general service times. computers and operations research, 35(8), 2520-2536. https://doi.org/10.1016/j.cor.2006.12.016 morales méndez, j.d., & rodriguez, r.s. (2017). total productive maintenance (tpm) as a tool for improving productivity: a case study of application in the bottleneck of an auto-parts machining line. international journal of advanced manufacturing technology, 92(1–4), 1013-1026. https://doi.org/10.1007/s00170-017-0052-4 nawanir, g., fernando, y., & teong, l.k. (2018). a second-order model of lean manufacturing implementation to leverage production line productivity with the importance-performance map analysis. global business review, 19(3_suppl), s114-s129. https://doi. org/10.1177/0972150918757843 neha gupta, n.g. (2013). an application of dmaic methodology for increasing the yarn quality in textile industry. iosr journal of mechanical and civil engineering, 6(1), 50-65 https://doi.org/10.9790/1684-0615065 sin, t.c., usubamatov, r., hamzas, m.f.b.m.a., wai, l.k., yao, t.k., & bahari, m.s. (2014). parameters investigation of mathematical model of productivity for automated line with availability by dmaic methodology. journal of applied mathematics, article id 206717. https:// doi.org/10.1155/2014/206717 singh, j., singh, h., & sharma, v. (2018). success of tpm concept in a manufacturing unit – a case study. international journal of productivity and performance management, 67(3), 536-549. https://doi.org/10.1108/ijppm-01-2017-0003 study and analysis the wastage reduction of fluorescent powder in cfl 23 w in philips pvt ltd mohali, using six sigma methodology. (2016). international journal of science and research (ijsr). https://doi.org/10.21275/v5i2.nov161472 xu, x. (2017). machine tool 4.0 for the new era of manufacturing. international journal of advanced manufacturing technology, 92(5-8), 1893-1900 https://doi.org/10.1007/s00170-017-0300-7 int. j. prod. manag. eng. (2019) 7(2), 151-159 creative commons attribution-noncommercial-noderivatives 4.0 international hussain, z. 158 https://doi.org/10.1016/j.procir.2018.03.245 https://doi.org/10.1108/ijppm-10-2017-0248 https://doi.org/10.1016/j.compchemeng.2017.10.027 https://doi.org/10.1080/21681015.2018.1462863 https://doi.org/10.1108/ijppm-11-2017-0282 https://doi.org/10.1016/j.ijpe.2017.02.013 https://doi.org/10.1016/j.ijpe.2017.02.013 https://doi.org/10.1108/bij-06-2018-0156 https://doi.org/10.1007/s40092-017-0234-6 http://www.ijser.org https://doi.org/10.12913/22998624/106240 https://doi.org/10.1016/j.cor.2006.12.016 https://doi.org/10.1007/s00170-017-0052-4 https://doi.org/10.1177/0972150918757843 https://doi.org/10.1177/0972150918757843 https://doi.org/10.9790/1684-0615065 https://doi.org/10.1155/2014/206717 https://doi.org/10.1155/2014/206717 https://doi.org/10.1108/ijppm-01-2017-0003 https://doi.org/10.21275/v5i2.nov161472 https://doi.org/10.1007/s00170-017-0300-7 http://creativecommons.org/licenses/by-nc-nd/4.0/ yan, w., lin, s., kafka, o.l., yu, c., liu, z., lian, y., … liu, w.k. (2018). modeling process-structure-property relationships for additive manufacturing. frontiers of mechanical engineering, 13(4), 482-492 https://doi.org/10.1007/s11465-018-0505-y yazdi, p.g., azizi, a., & hashemipour, m. (2019). a hybrid methodology for validation of optimization solutions effects on manufacturing sustainability with time study and simulation approach for smes. sustainability (switzerland), 11(5), 1454. https://doi.org/10.3390/ su11051454 zhang, d., sun, s., qiu, d., gibson, m. a., dargusch, m. s., brandt, m., … easton, m. (2018). metal alloys for fusion-based additive manufacturing. advanced engineering materials, 20, 1700952. https://doi.org/10.1002/adem.201700952 int. j. prod. manag. eng. (2019) 7(2), 151-159creative commons attribution-noncommercial-noderivatives 4.0 international developing a novel based productivity model by investigating potential bounds of production plant 159 https://doi.org/10.1007/s11465-018-0505-y https://doi.org/10.3390/su11051454 https://doi.org/10.3390/su11051454 https://doi.org/10.1002/adem.201700952 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.12254 received: 2019-08-24 accepted: 2020-11-16 green productivity: waste reduction with green value stream mapping. a case study of leather production prayugo, j.a, zhong, l.x.b a tianjin university of science and technology, china. a1 jovanotianjinuniversity@gmail.com, b lixz@tust.edu.cn abstract: this study aims to identify and reduce waste in pt rajapaksi adya perkasa manufacturing company. the analysis uses the concept of green productivity with green value stream mapping as the analysis model. through the gvsm current state, this study calculates the amount of green waste in the production line including seven green waste. the result of this study shows an improvement to minimize the waste that occurs through the gvsm future state. gvsm technique increases the green productivity value from 1.12 to 1.81. key words: green productivity, green waste, green value stream mapping, economic indicator, environmental indicators. 1. introduction globalization has an impact on the development of the world economy. an open economy has created a competitive atmosphere among industries not only at the national level but also at the international level (saxena et al., 2007). the industry must be able to survive and have a strategy in winning the global competition. increasing production is one of the right steps. manufacturers face continuous demand so producers need to pay attention to the environment as a rule that is enforced throughout the world (gungor & gupta, 1999). environmental problems from the impact of products that are continuously rarely noticed by companies. indonesia, especially east java, which is one of the largest leather material industries in indonesia, has problems with poor environmental management. many production processes are not based on environmentally friendly concepts. the remainder of the skin cutting is not used properly, besides being sold, it is also disposed of freely into the environment. disposal of raw materials is detrimental to the company in terms of the economy and can cause landfill. more directed environmental management is needed to reduce its impact on the environment. overcoming environmental problems, elkington (1997) presents a challenge to achieving sustainability as an unprecedented source of commercial opportunities for competitive companies, one of which is through increasing environmental efficiency. the underlying assumption is that financial success can be made consistent with the compliance of ethics, environment, and society (dobers & wolff, 2000; mohanty & deshmukh, 1998; stead & stead, 2000). in the concept of environmental efficiency, green productivity is one part of this effectiveness. green productivity is an effort to protect and improve the quality of the environment (hur, 2004). green productivity is an alternative solution in evaluating and improving the environment by adjusting financial analysis (singgih et al., 2010). green productivity is to cite this article: prayugo, j., zhong, l.x. (2021). green productivity: waste reduction with green value stream mapping. a case study of leather production. international journal of production management and engineering, 9(1), 47-55. https://doi.org/10.4995/ijpme.2021.12254 int. j. prod. manag. eng. (2021) 9(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international 47 http://creativecommons.org/licenses/by-nc-nd/4.0/ a strategy that is not only a commitment to protect the future of the environment, preserving natural resources, but also building a future factory towards sustainable development (sheng et al., 2005). green productivity must be increasingly important and quick to do (brandt, 2007; corbett & klassen, 2006; dills & stone, 2007; stead & stead, 2000) given that as population increases, and the economy develops, ecosystems and resources of the planet experience external challenges usual (de burgos & cespedes, 2001; esty & winston, 2009; hart, 1995; kleindorfer et al., 2005; mohanty & deshmukh, 1998). the production system, which supplies increasing demand for goods, is associated with adverse environmental impacts (frosch & gallopoulos, 1989). gaur et al., research (2011) shows that green productivity can know the development of declining productivity and neglect the environment. this can be used as a reference to find alternative solutions for future improvements. by setting standards for assessment to be ratified by the government. in addition, providing solutions by renewing environmentally friendly technology. with the green productivity approach, companies benefit from minimizing the use of energy resources so that production processes are more effective and increase profits. increasing productivity alone will not continue if you do not pay attention to environmental security issues. integrating productivity improvements by paying attention to the environment is a challenge in the current industrial era. even though such things are needed to survive in increasing global competition. while productivity provides a framework for continuous improvement, environmental protection provides the basis for sustainable development. based on the description of the above problems, the purpose of this research is to increase productivity by paying attention to the environment through green productivity, as well as providing alternative solutions to solve economic and environmental problems so that it can improve and develop sustainable industries. 2. literature review 2.1. green productivity green productivity is an understanding that a healthy environment and a strong economy can foster an interdependent competitive business climate. the potential for using green productivity integrates environmental protection into business operations as a platform of increasing productivity (asia productivity organization, 2008). green productivity measurement is a measurement tool used to measure the performance of green productivity implementation. it could be said as a strategy to increase productivity and environmental protection, by analyzing productivity and environmental performance separately (findiastuti, et al., 2011). asia productivity organization (2008) explain that life cycle cost is the entire product life cycle starting from extraction, use and disposal of products by evaluating the environmental burden of a product. this approach comes from europe and is considered as the development of environmental friendly products. conventionally, the increased in productivity is focused on cost effectiveness through cost reduction. with the emergence of the “quality” impulse, productivity is measured by comparing the benefits obtained from the program (output) with the quality of the resources used in the program (input). green productivity index = selling price/life cycle cost environmental impact adopting gp ratio, gandhi (findiastuti, et al., 2011) justify ‘environmental impact’ by weighting the environmental indicators of solid waste generation (swg), gas waste generation (gwg), water consumption (wc) as the following: green productivity index = selling price/life cycle cost w1swg+w2gwg+w3wc 2.2. green waste besides lean waste, breet wills (balinski & grantham, 2013) states that the presence of green waste that can be measured and systematic. green waste is categorized into seven namely energy, water, material, waste, transportation, emissions, and biodiversity. wills explained in detail every green waste on the following: 1. energy refers to the source of electricity generation and production. energy is believed to be a waste when it is overused and it becomes a dirty source. int. j. prod. manag. eng. (2021) 9(1), 47-55 creative commons attribution-noncommercial-noderivatives 4.0 international prayugo & zhong 48 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. water as a limited source. excessive use of water is said to be wasteful, in financial terms it is seen from the cost of water consumption and contaminated water. 3. material waste that comes from defect products. 4. waste that comes from green waste disposal. 5. emissions that are formed from waste dumps, recycling, and combustion. 6. transportation waste, namely how much transportation is used for people, materials, supplies, and finished products. 7. biodiversity waste includes disruption to the surrounding environment such as tree logging or garbage accumulation. 2.3. environmental indicators findastuti (2011) says that eco-efficiency of production is related to the ability to produce goods and services by minimizing the causes of environmental damage. there are two indicators to find environmental efficiency, namely economic indicators and environmental indicators. environmental indicators are the denominator of eco-efficiency ratios. environmental indicators at all levels are related to the environmental themes of each unit depending on the product, process, or service. marizkaa et al. (2015) explains that the green productivity index is calculated by the environmental impact formula, a model of increasing green productivity to ensure the quality of production and reduce environmental impacts simultaneously. according to asia productivity organization says that the environmental impact from the production. the process is defined as the accumulation of the production process of leather material, which is the three environmental variables including solid waste generation (swg), gas waste generator (gw) and water consumption (wc) as described in the equation: ei = a(swg) + b(gwg) + c(wc) + d(lc) ei = (0.375×gwg) + (0.25×wc) + + (0.125×swg) + (0.25×lc) note: ei: environmental impact, gwg: gas waste generator, wc: water consumption, swg: solid waste generator, lc: soil pollution. 2.4. economic indicator findastuti (2011) sais that eco-efficiency of production is related to the ability to produce goods and services by minimizing the causes of environmental damage. there are two indicators in finding efficient environments, which are economic indicator and environmental indicator. economic indicator is numerator in environmental efficiency ratio. micro level economic indicators are about the value of products or services, such as: net sales, production per year, gross added value. 3. methodology this research is a qualitative study with a case study approach. this research was conducted on the small and medium scale leather industry (ikm), namely pt. rajapaksi adyaperkasa east java, indonesia. data collection is used by the method of observation, interviews and documentation. this study focuses on knowing the green productivity process (green productivity). the measurement of green productivity comes from the comparison of economic indicators and environmental indicators. economic indicators derived from productivity and environmental indicators are obtained from the green value stream mapping. green value stream mapping as a model minimizes economic impacts by maintaining indicators of economic growth (marizkaa et al., 2015). breet wills (balinski & grantham, 2013) said the green value streaming mapping procedure is very similar to value streaming mapping, namely the initial structure of supplier diagrams, customers, value stream activities, and information flow are identical. but the difference lies in green waste, namely energy, water, materials, waste emissions, transportation, and biodiversity. breet wills (balinski & grantham, 2013) continues that there are several steps in the green flow map, namely: 1. identification of inputs and outputs, 2. measuring recycling, 3. classification of each input and output as biological or technical, 4. assessing the effect of the material on environment and society, and 5. elimination of materials that have a negative impact on the environment. int. j. prod. manag. eng. (2021) 9(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international green productivity: waste reduction with green value stream mapping. a case study of leather production 49 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. result and discussion this study focuses on knowing the green productivity process at pt. rajapaksi adyaperkasa. the measurement of green productivity comes from the comparison between economic indicators and environmental indicators. economic indicators derived from productivity, while environmental indicators are obtained from the green value stream mapping. 4.1. current state maps current state maps is one of the bases of continuous improvement. all information about the current state map is now collected based on the method suggested by rother and shook (1999). in current state maps, economic indicators and environmental indicators are explained. economic indicators for analyzing circumstances about process inputs, production processes, and output processes and the initial productivity of industry in this case leather industry and the use of resources that will be counted as environmental indicators. 4.1.1. input input is obtained from the labor force, production process, product, and inventory. the production process consists of cutting, preparation, stitching, and assembling. the workforce is 3,000 employees and 250 office staffs. there are 3 product categories namely woman, man, baby, and unisex under the brands of santica, airmax, disney, hasbro, cars, pokemon, and marvel. the company produces 54,000 pairs of shoes per month. the production process can be described in the following sections: 1.1.1. production process the production process starts by receiving pre-orders from consumers, choosing the desired shoe model. then from the company offers about certain models as additional shoes, if it does not agree, the model is kept as a database, and if agreed, the development division works on the model desired by the customer. this division uses sophisticated machines so that the process can be guaranteed. the next process is making shoes, and when the shoes are ready, the company sends shoes to the customer. the order flow diagram can be seen in figure 1. the process of production unit consists of 4 stages, namely cutting, preparation, stitching, and assembling. first, cutting process is trimming shoe materials into certain shapes which are divided into upper side (outer hood), lining (inside of shoe), insole (instock, eva, texon), strap (replacement strap), famp (front part of shoe), quarter (middle body), back counter (the back side of shoe consisting of back tabs and variations), outer and inner tongue, and toe cap. the cutting process uses trimming machines. figure 1. order scheme of shoes. the second process is preparation, this process only consists of two times namely interlining and lining using a sewing machine. the goal is to provide a coating so that the feet don’t get scratched and feel warm. after that, matching each shoe-making component makes it easier for the assembling stage to assemble each shoe. this division prepares 20-40 components that must be paired into a pair of shoes. the third stage is the stitching stage that contains shoe components sewing process. the process begins with skiving/scrapping which aims to smoothen the component. each component is sewn in a zigzag pattern, then the variations are placed to order. the sewing process uses 1 needle or 2 needles. next, the shoes are returned to the foam collar section and trimmed by the stitches. next, the strap insertion process is done to attach the rope by sewing “tongue” int. j. prod. manag. eng. (2021) 9(1), 47-55 creative commons attribution-noncommercial-noderivatives 4.0 international prayugo & zhong 50 http://creativecommons.org/licenses/by-nc-nd/4.0/ part. after the stitches meet the standard required, there is a shoe cleaning process. before entering the final stage, there is a quality control team and molding process to make sure that the shoe shape is firm. the last step is assembling, which is the stage of assembling the shoes. this stage consists of three parts. the first stage, shoes must be checked first by quality control to prevent damaged items. shoes enter texon, upper, and shoe lase preparations process that contains applying laxes taxon and upper layers to the shoes, then the goods are put into the oven. the next process is using the tupap machine to make the shoes supple and firm. shoes are drawn sampling by the “workers” to sew upper sampling and use a cutting machine to form the back of the foot. the second assembling stage consists of a vacuum heater (leather), there is a front-part grinding process and the side-part grinding process. next, scrapping the upper and bottom part of the shoes. shoes are given adhesive material for upper (subordinate shoes) and outsole (rubber shoes). then the shoes are put into the oven again and applying the adhesive material to the upper and outsole. shoes are put in the oven and apply the second adhesive material to the upper and outsole to be assembled together. the shoe enters the universal press machine to press the shoes so that it can be more adhesive and the shoes are inserted into the excessive glue removal machine and heating machine. the final assembling stage is a finishing process consisting of shoes inserted into a cooling machine. then the shoe lasts releasing process and cleaning the rope, insock, as a whole. before being packaged, goods are controlled in advance by a quality control team. the goal is to ensure the goods are in good quality and avoid defective items. if the item is defective, it will be repaired immediately. if the damage is severe then it needs to be returned to preorder items. when the item is perfect, the item is ready to be packaged through the packaging process. the production process chart of pt. rajapaksi adyaperkasa as follow (figure 2). this production process requires costs in the process. the total production cost for 79,370 unit per month is rp. 9,741,228,000, that consists of: table 1. cost of production process current state maps. no cost of total (in rupiah) 1 worker salary 1573 workers (@3,580,000) 152 office staff (@5,000,000) 6,816,500,000 6,056,050,000 760,000,000 2 material 1,157,000,000 3 raw-material 1,237,000,000 4 electricity 530,728,000 4.1.2. output output of each stage has different targets. cutting target consists of 10-12 pairs/hour (leather), 2000 pairs/day (upper), 300 pairs/hour, while preparation is 70 pairs/hour (all component), for stitching consists of 33 people produce locally 23 pairs/hour and expert 85 pairs/hour and assembling stages which are 30 pairs/hour. from the entire process, the total number of shoes produced was 79,370 consisting of standard products and defects of 2920 pairs. total sales of shoes for 77,950×rp. 300,000 is rp. 23,385,000,000. 4.1.3. productivity to get the high productivity level, it is necessary to compare inputs and outputs in each production process. calculations are calculated as the ratio of product sales revenue /total production cost. the total production cost for 79,370 units per month is rp. 9,741,228,000. total sales of shoes for 77,950×rp. 300,000 is rp. 23,385,000,000 rp. 23,385,000,000 =2.40 rp. 9,741,228,000 productivity level below 50% indicates low productivity. santica brand shoes at the lowest productivity level, because the productivity percentage is only 40% of the total sales of santica shoes. meanwhile, the productivity level included in the high productivity category should have the percentage of 50% above. the highest productivity level is 80%, namely airmax shoes. figure 2. production process. int. j. prod. manag. eng. (2021) 9(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international green productivity: waste reduction with green value stream mapping. a case study of leather production 51 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.1.4. environmental impact indicator according to the asia productivity organization, the environmental impact of production is the process of accumulation of the production process of leather material, among others, the three weights of environmental variables including solid waste generator (swg), gas waste generator (gwg) and water consumption (wc) as described in the equation: table 2. waste production of current state maps. green waste rajapaksi adyaperkasa company energy 1503 kwh water 3843.5 liter material 4571.5 kg garbage 230.5 kg transport 95 km emission 21.81 kg bioversity 1,001 ha total production of leather shoes is 54,000 pairs of shoes. the assumption is that producer produces 700 per one lead. so that for a production base of 1 ton of shoes at least 80 times the lead. gas waste generator (gwg) is 21.81 kg per machine. land consumption (lc) 4571.5 kg. water consumption is 3843.5 liters because the density of water reaches 1 kg/liter. solid waste generator (swg) 230.5 kg is waste originating from dirty leather material and material that is not in accordance with the standard. from the data generated from the process, the environmental impact can be calculated in part as follows: ei= (0.375×21.81) + (0.25×3843.5) + (0.125×230.50) + (0.25×4571.5) = 8.17875 + 960.875 + 28.8125 + 1142.875 = 2140.74125 kg or 2.14 ton based on the results of the impact of economic indicators and environmental indicators, then calculated with green waste in accordance with current conditions (current stage), then the following comparison is obtained: gpi = 2.40: 2.14 = 1.12 based on the data described above, current state maps can be seen in figure 3 below as follows: figure 3. current state maps. int. j. prod. manag. eng. (2021) 9(1), 47-55 creative commons attribution-noncommercial-noderivatives 4.0 international prayugo & zhong 52 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.2. future state maps future state maps situation is nothing more than an implementation plan that describes the tools needed in the lean process to eliminate waste and where (in what process) the tool is needed in the value flow of a product. the process of mapping future conditions develops productivity and reduces the amount of waste that can affect the level of productivity of the company. meanwhile, a complete value stream to map perfect future conditions must try to define lean equipment to reduce both. this, of course, follows an efficient procedure in which it tries to answer a set of questions that have been prepared, therefore, produces a mapping of future conditions that will facilitate the reduction or at least eliminate various types of waste in the current manufacturing structure. the application of future state maps illustrates an improvement in every production in manufacturing. explanation of future state maps is a change in economic and environmental indicators. 4.2.1. input the total production cost of 79,370 unit per month is rp. 7,357,228,000, consists of: table 3. cost of production process future state maps. no cost of total (in rupiah) 1 worker salary 1145 workers (@3,580,000) 118 office staff (@5,000,000) 4,692,500,000 4,099,100,000 592,900,000 2 material 997,000,000 3 raw-material 1,237,000,000 4 electricity 430,728,000 based on the table data above, there was a reduction in workforce of 428 people for workers and 34 people for office staff. the use of materials and electricity also has a reduction so that high productivity occurs in the production process. 4.2.2. output the total number of shoes produced is 79,370 consisting of standard products and defects of 2920 pairs. total sales of shoes for 77,950×rp. 300,000 is rp. 23,385,000. 4.2.3. productivity to get the high productivity level, it is necessary to compare inputs and outputs in each production process. calculations are calculated as the ratio of product sales revenue/total production cost. the total production cost for 79,370 units per month is rp. 7,357,228,000. total sales of shoes for 77,950×rp. 300,000 is rp. 23,385,000,000. rp. 23,385,000,000 =3.17 rp. 7,357,228,000 4.2.4. environmental impact indicator meanwhile, economic impact indicators are described as follows: table 4. waste production of future state maps. green waste pt.rajapaksi adyaperkasa energy 1089 kwh water 3527.5 liter material 3407 kg garbage 206 kg transport 80 km emisi 12.38 kg bioversity 1,001 ha total production of leather shoes is 54,000 pairs of shoes. the assumption is that producer produces 700 per one lead. so that for a production base of 1 ton of shoes at least 80 times the lead. gas waste generator (gwg) is 12.38 kg per machine. land consumption (lc) 3407 kg. the use of water in production is 3527,5 liters because the density of the water reaches 1 kg/liter. solid waste generators (swg) 206 kg is a waste originated from the dirty leather material and material that is not in accordance with the standard. from the data generated from the process, the environmental impact can be calculated in part as follows: ei= (0.375×12.38) + (0.25×3527,5) + (0.125×206) + (0.25×3407) = 4.6425 + 881.875 + 25.75 + 851.75 = 1765.0175 = 1.75 ton gpi = 3.17: 1.75 = 1.81 based on the data described above, future state maps can be seen in figure 4 below as follows: 5. conclusion this study analyzes green productivity based on economic indicators and environmental indicators. the economic indicator is obtained based on the input and the result of the output ratio is 2.40. while environmental indicators are calculated by int. j. prod. manag. eng. (2021) 9(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international green productivity: waste reduction with green value stream mapping. a case study of leather production 53 http://creativecommons.org/licenses/by-nc-nd/4.0/ the environment index formula is 2.14. so, green productivity index produces 1.12. green productivity techniques are able to provide solutions to improve pollution (balist, 2016). this study uses the green value streaming mapping technique to evaluate alternatives to improve green productivity. the current state of green value streaming mapping is based on the results of green productivity. the researchers give suggestions through the future state green value streaming mapping table. the gvsm technique can increase the value of green productivity from 1.12 to 1.81. increasing in value of green productivity is a good sign of the value of green productivity. the same idea with primary research (2015) that the greater the epi value, the better the environmental performance that has been applied in the company. deviation (pi) can be said to be good if the percentage is positive and bigger. references asia productivity organization.(2008). green productivity. lynn johannson:canada. balinski, k.d., grantham, k. (2013). quantifying the carbon footprint of lean waste. engineering, 5(1), 80-91. balist, j., sargazi, e., hoveidi, h., faryadi, s. (2016). environmental management system and green productivity (ems_gp) implementation in kurdistan cement plant. international journal of business and management invention, 5(4). brandt, d. (2007). a world gone green. industrial engineer, 39(9), 29-33. corbett, c.j., klassen, r.d. (2006). extending the horizons: environmental excellence as key to improving operations. manufacturing & service operations management, 8(1), 5-22. https://doi.org/10.1287/msom.1060.0095 de burgos jiménez, j., cespedes lorente, j.j. (2001). environmental performance as an operations objective. international journal of operations & production management, 21(12), 1553-1572. https://doi.org/10.1108/01443570110410900 dills, j., stone, h. (2007). environmentally conscious technologies. surface mount technology, 21(6), 14-19. dobers, p., wolff, r. (2000). competing with “soft” issues-from managing the environment to sustainable business strategies. business strategy and the environment, 9(3), 143-150. https://doi.org/10.1002/(sici)1099-0836(200005/06)9:3%3c143::aidbse239%3e3.0.co;2-c elkington, j. (1997). partnerships from cannibals with forks: the triple bottom line of 21st-century business. environmental quality management, 8(1), 37-51. https://doi.org/10.1002/tqem.3310080106 esty, d.c., winston, a.s. (2009). green to gold: how smart companies use environmental strategy to innovate, create value, and build competitive advantage. wiley, chichester, uk. figure 4. future state maps of pre processed rubber. int. j. prod. manag. eng. (2021) 9(1), 47-55 creative commons attribution-noncommercial-noderivatives 4.0 international prayugo & zhong 54 https://doi.org/10.1287/msom.1060.0095 https://doi.org/10.1108/01443570110410900 https://doi.org/10.1002/(sici)1099-0836(200005/06)9:3%3c143::aid-bse239%3e3.0.co;2-c https://doi.org/10.1002/(sici)1099-0836(200005/06)9:3%3c143::aid-bse239%3e3.0.co;2-c https://doi.org/10.1002/tqem.3310080106 http://creativecommons.org/licenses/by-nc-nd/4.0/ findiastuti, w., anityasari, m., singgih, m.l. (2011). green productivity index: do different terms measure the same things. proceeding of industrial engineering and service science, hal. 20-21. frosch, r.a., gallopoulos, n.e. (1989) strategies for manufacturing. scientific american, 261, 144-152. https://doi.org/10.1038/ scientificamerican0989-144 gaur, s., matta, g., singh, v. (2011). impotance and role of green productivity in industries: a review. environment conversation journal, issn 0972-3099. gungor, a., gupta, s.m. (1999). issues in environmentally conscious manufacturing and product recovery: a survey. computers & industrial engineering, 36(4), 811-853. https://doi.org/10.1016/s0360-8352(99)00167-9 hart, s.l. (1995). a natural-resource-based view of the firm. academy of management review, 20(4), 986-1014. https://doi.org/10.5465/amr.1995.9512280033 hur, t., kim, i., yamamoto, r. 2004. measurement of green productivity and its improvement. j. clean. product., 12(7), 673-683. https://doi.org/10.1016/j.jclepro.2003.08.004 kleindorfer, p.r., singhal, k., van wassenhove, l.n. (2005). sustainable operations management. production and operations management, 14(4), 482-492. https://doi.org/10.1111/j.1937-5956.2005.tb00235.x marizkaa, d.a., djatnab, t., arkeman, y. (2015). a model of green value stream mapping for rubber based automotive products. scientific journal of ppi-ukm issn, 2356-2536. mohanty, r.p., deshmukh, s.g. (1998). managing green productivity: some strategic directions. production planning & control, 9(7), 624633. https://doi.org/10.1080/095372898233614 rother, m., shook, j. (1999). learning to see. cambridge, ma: lean enterprise institute. saxena, a.k., bhardwaj, k.d., sinha, k.k. (2007). sustainable growth through green productivity: a case of edible oil industry in india. international energy journal, 4(1). sheng, t.l., shamsudin m.z., ling, l.c. (2005). sustainable development with green productivity in manufacturing. issm 2005, ieee international symposium on semiconductor manufacturing, 2005., san jose, ca, 2005, pp. 267-270. https://doi.org/10.1109/ issm.2005.1513353 singgih, m.l.,suef, m., putra, c.a. (2010). waste reduction with green productivity approach for increasing productivity (case study: pt. indopherin jaya). in the 14th asia pacific regional meeting of international foundation for production research, melaka, 7-10 december 2010. stead, j.g., stead, e. (2000). eco-enterprise strategy: standing for sustainability. journal of business ethics, 24(4), 313-329. https://doi.org/10.1023/a:1006188725928 int. j. prod. manag. eng. (2021) 9(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international green productivity: waste reduction with green value stream mapping. a case study of leather production 55 https://doi.org/10.1038/scientificamerican0989-144 https://doi.org/10.1038/scientificamerican0989-144 https://doi.org/10.1016/s0360-8352(99)00167-9 https://doi.org/10.5465/amr.1995.9512280033 https://doi.org/10.1016/j.jclepro.2003.08.004 https://doi.org/10.1111/j.1937-5956.2005.tb00235.x https://doi.org/10.1080/095372898233614 https://doi.org/10.1109/issm.2005.1513353 https://doi.org/10.1109/issm.2005.1513353 https://doi.org/10.1023/a:1006188725928 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.14483 received: 2020-10-13 accepted: 2020-12-30 validation of production system throughput potential and simulation experiment design standridge, c.a*, wynne, m.b a school of engineering, grand valley state university, 301 fulton avenue west, grand rapids, mi 49504, usa. b yanfeng automotive interiors. 1600 washington street, holland, mi 49423, usa. a standric@gvsu.edu, b mjw.michaelwynne@gmail.com abstract: the throughput potential of a production system must be designed and validated before implementation. design includes creating product flow by setting the takt time consistent with meeting customer demand per time period and the average cycle time at each workstation being less than the takt time. creating product flow implies that the average waiting time preceding each workstation is no greater than the takt time. kingman’s equation for the average waiting time can be solved for the variation component given the utilization and the cycle time. the variation component consists of the variation in the demand and the variation in the cycle time. given the variation in demand, the maximum allowable variation in cycle time to create flow can be determined. throughput potential validation is often performed using discrete event simulation modeling and experimentation. if the variation in cycle time at every workstation is small enough to create flow, then a deterministic simulation experiment can be used. an industrial example concerning a tier-1 automotive supplier with two possible production systems designs and various levels of variation in demand assumed is used to demonstrate the effectiveness of throughput validation using deterministic discrete event simulation modeling and experimentation. key words: throughput potential validation, kingman’s equation, discrete event simulation. 1. introduction validating that a production system can meet customer throughput requirements before implementation has long been an important goal. ferrin, muller, and muthler (2005) discuss how discrete event simulation (des) is uniquely able to support achieving this goal by finding a very good solution that meets system design and operation requirements before implementation. marvel and standridge (2009) discuss and illustrate an enhanced lean improvement process that uses des to validate a proposed future state of a production system before implementation. one particular class of production systems of interest produces a single part type, or perhaps a small family of part types, for delivery as a subassembly to another business, as opposed to a finished product for delivery to a consumer. this can result from a supplier fulfilling a contract for a specified number of parts per day or week with little or no variation in demand. for example, a tier-1 automotive supplier contracts to build a specific number of door handles per week for an automobile manufacturer. a typical structure for such a system is shown in figure 1. a main cell completes production of the subassembly to be delivered to the customer. the main cell receives subassemblies from one or more feeder cells. a feeder cell may receive a subassembly from another feeder cell. the cells may not be adjacent within a facility. a worker, called a runner, moves subassemblies among the feeder and to cite this article: standridge, c., wynne, m. (2021). validation of production system throughput potential and simulation experiment design. international journal of production management and engineering, 9(1), 15-23. https://doi.org/10.4995/ijpme.2021.14483 int. j. prod. manag. eng. (2021) 9(1), 15-23creative commons attribution-noncommercial-noderivatives 4.0 international 15 http://creativecommons.org/licenses/by-nc-nd/4.0/ main cells on a fixed route. the maximum time to complete movement through the route once is a specified constant. figure 1. typical structure of a business-to-business production system (source: authors’ drawing). each cell consists of one or more workstations. work is performed either automatically by a machine or according to a standard work specification by a human. thus, cycle time variance is often small. since the variation in cycle times is small and the variation in demand is also small, the question arises as to whether these variances can be ignored in formulating the des model and experiment. if possible to do so, a deterministic model and experiment offers advantages over a stochastic model and experiment. simulation results are not statistical estimates but constant values. thus, the analysis of such results is much simplified. only one replicate is required per combination of input values, greatly reducing computational time. considering a deterministic model and experiment seems reasonable since pritsker (1989) points out that many industrial models, around 30%, have no random quantities as they are used to evaluate operating procedures which are complex, multivariate and contain non-trivial algorithms. for example, kleijnen and standridge (1988) describe the analysis of a flexible manufacturing system using deterministic des. for each of three operations, either a machine only capable of performing that operation or a flexible machine capable of performing all three operations was chosen. a regression meta-model related lead time in the system (dependent variable) to the number of each of the four types of machines (independent variables). lead time depended only on the number of machines performing operation two and the number of flexible machines. thus, the effectiveness of deterministic simulation was shown. what is missing is a systematic way to evaluate whether deterministic des can be used. a way to determine the maximum variation is developed which is consistent with the lean manufacturing idea of creating flow. in addition, the des modeling approach for systems with the structure shown in figure 1 is presented. application to an existing production system is shown. the application considers two alternative configurations of workstations as well as no variance, a small variance, and a large variance in demand. 2. background reducing model and experiment complexity without compromising the validity required to support decision making is a methodological issue that is still being addressed. this work contributes to this ongoing effort in the context of production systems. askin and standridge (1993) provide an overview, including a review of methods. for example, jayaraman and gunal (1997) discuss the complexities of automotive powertrain manufacturing resulting from the need to assemble hundreds of components which are produced by separate systems or suppliers and then integrated. how des models and experiments are used to address this complexity is described and illustrated. more recently, pinheiro et al. (2019) used des to compare the performance of push, pull, and pull with a conwip control in the operation of a production system. they found that the pull system yielded the best performance with respect to meeting customer demand and minimizing lead time. zupan and herakovic (2015) discuss how simulation is used to improve a production line after line balancing using traditional methods has been performed. as seen in these examples, des models most often contain elements represented by random variables such as cycle times and times between arrivals. simulation experiments must be designed and multiple replicates, usually between 10 and 30, executed for each of the many combinations of input parameter values as discussed by law (2014) and kleijnen (2015). in addition, statistical analysis of the simulation results must be performed, the complexity of which depends on the experiment design. for example, atalan and dönmez (2020) report the design of a large-scale, full factorial simulation experiment that was used to reduce patient waiting time in an emergency room by about int. j. prod. manag. eng. (2021) 9(1), 15-23 creative commons attribution-noncommercial-noderivatives 4.0 international standridge and wynne 16 http://creativecommons.org/licenses/by-nc-nd/4.0/ 75% and increase the number of patients seen by about 10%. aggregate modeling is one way of reducing model complexity. khan and standridge (2019) discuss one possibility for aggregate modeling in the context of a production system. the over 100 products produced by the system are combined into one aggregate product with routing between workstations chosen at random. the model is shown to be valid and effective in estimating the single parameter of a conwip flow control system (spearman, et al. 1990). computational efficiency related to aggregation methods is discussed by tribastone and vandin (2018). schruben (1983) developed the event graph modeling technique to address the complexity of modeling and simulating extremely long production lines such as those found in the semi-conductor industry. instead of modeling individual parts moving from workstation to workstation, the sequence of events that changes the state of the system is modeled. such events include the beginning and end of operations at a station that change station status from idle to busy and back again as well as counting the number of parts at each station. thus, the number of parts in the system and in any subarea of the system can be directly determined. the average lead time can be computed using little’s law (little, 1961). this approach was shown to greatly reduce simulation execution times. as discussed above, deterministic simulation experiments are less complex than stochastic experiments. for example, dagkakis et al. (2019) demonstrate the effectiveness of deterministic simulation in analyzing a production system where the experiment uses an optimization algorithm to search the solution space to maximize throughput. this approach optimally assigns cross-trained operators in an assembly line considering the various levels of cross-training of each operator and the demand placed on the line. improvement in throughput versus assigning workers to minimizes bottlenecks and wip-levels was shown. using optimization algorithms to search a simulation experiment solution space typically causes many combinations of model parameter values to be evaluated. thus, it is important to use deterministic simulation to minimize execution time. what is missing in the published literature is a systematic way to decide when deterministic simulation can be used. while uriarte et al. (2020) provide a comprehensive review of the joint use of lean and des, they do not identify reduction in simulation experiment complexity as a research gap. as reduction in variance is a primary goal of lean (tapping et al. 2002), the opportunity to use deterministic simulation experiments when jointly applying simulation and lean would seem to exist. simularly, mourtzis (2019) provides a review of the use of simulation in the design and operation of production systems particularly with regard to factory 4.0. however, the use of deterministic simulation and complexity reduction in simulation experimentation are not discussed. in the same vein, sanchez et al. (2020) discuss the design of simulation experiments. while acknowledging the possibility of both deterministic and stochastic experiments, they do not discuss criteria for choosing between them. finally, puvanasvaran et al. (2020) discuss a simulation application for determining the throughput of a production systems with significant material movement delays. however, they assume that a stochastic simulation experiment is needed without assessing the possibility of employing a deterministic one. thus, an approach for determining when deterministic simulation is appropriate for a commonly occurring type of production system as developed here is another, needed step toward reducing des modeling and experimentation complexity. 3. methods the simulation modeling approach for a production system with the structure shown in figure 1 is given in figure 2. the model consists of one process for the main lines and for each of the feeder cell as well as a process for the runner. the processes do not directly communicate with each other. instead they share a set of state variables that record the number of the various work-in-process elements, such as subassemblies, in the system. each feeder cell process produces one or more work-in-process elements and may need one or more work-in-process elements produced by another feeder process to begin working. the main process needs one or more work-in-process elements to begin working and produces one item to finish goods inventory. the runner process moves the work-in-process elements among the feeder lines and the main line. int. j. prod. manag. eng. (2021) 9(1), 15-23creative commons attribution-noncommercial-noderivatives 4.0 international validation of production system throughput potential and simulation experiment design 17 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 2. modeling approach for a constant rate production system (source: authors’ drawing). an entity arrives to each feeder cell process and the main process at an average time between arrivals equal to the takt time. the runner arrives to the runner process once, at the beginning of the simulation. the takt time is the pace at which work must be done in order to meet the production target for a given time period and is shown mathematically in equation 1 (tapping et al., 2002). takt time= work time available (1) number of units required thus, the average cycle time at each station in each line cannot exceed the takt time. this implies that the average time between initiation of parts on each line is the takt time. this also implies that the average waiting time for processing in the queue at each station cannot exceed the takt time as waiting can be viewed as one more processing step. the average waiting time is given by kingman’s equation (kingman, 1961). (2) each quantity in equation 2 is described in table 1. table 1. quantities in kingman’s equation (source: authors). quantity description ltq average waiting time in the queue preceding a workstation cv coefficient of variation = standard deviation average cv2 squared coefficient of variation tba time between arrivals ct cycle time m workstation utilization = percent busy time = ct / tba note that the maximum value for ltq that is consistent with using deterministic simulation is the takt time. in addition, the cycle time can be rewritten as the product of the utilization and the time between arrivals. these lead to equation 3. (3) solving for v yields equation 4. (4) equation 4 means that in order for deterministic simulation to be used the variation at each workstation can be no greater than a value that is a function of the utilization alone. the difficulty is that the variation, v, consists of two parts: the variation in the time between arrivals and the variation in the cycle time. the strategy will be to assume a value for the variance of the time between arrivals to the first station in the main cell or a feeder cell. as discussed above, this variance should be small when the item produced is to be delivered to another business. the variance could be zero if the demand per week is a constant value specified by a contract. as standridge and heltne (2000) discovered, even weekly order sizes placed by another business only have a small random component, no more than 20% of the average order size (cv = 0.2). with the variance of the demand assumed and demand expressed as the time between arrivals, the maximum variance in the cycle time can be computed given the workstation utilization as shown in equation 5. (5) the maximum allowed value for the variance of the cycle time, expressed as the squared coefficient of variation, is illustrated for the case where the variance of the time between arrivals is low (cvtba = 0.0, 0.2) as well as for the case where the variance of the time between arrivals is high (cvtba = 1.0). as previously discussed, the former corresponds to production meeting the demand of another business. the latter corresponds to meeting consumer demand, the practical worse case variance equal to the variance of an exponential distribution as discussed by hopp and spearman (2011). the results are shown graphically in figure 3 and in int. j. prod. manag. eng. (2021) 9(1), 15-23 creative commons attribution-noncommercial-noderivatives 4.0 international standridge and wynne 18 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 2. note that negative values of cv2ct indicate that it is not feasible to use deterministic simulation for a workstation with the corresponding utilization and cvtba combination. figure 3. maximum allowable cycle time variation (source: author generated). table 2. maximum allowable cycle time variation (source: author generated). utilization cv2ct with cvtba = 0.0 cv2 ct with cvtba = 0.2 cv2 ct with cvtba = 1.0 50% 4.00 3.96 3.00 55% 2.98 2.94 1.98 60% 2.22 2.18 1.22 65% 1.66 1.62 0.66 70% 1.22 1.18 0.22 73% 1.01 0.97 0.01 75% 0.89 0.85 -0.11 80% 0.63 0.59 -0.38 85% 0.42 0.38 -0.58 90% 0.25 0.21 -0.75 95% 0.11 0.07 -0.89 99% 0.02 -0.02 -0.98 while the variance of the time between arrivals to the first workstation in sequence is assumed, the variance of the time between arrivals to the next and following workstations must be determined. this issue is discussed by hopp and spearman (2011). the variance in the time between arrivals to the following workstation can be assumed to be equal to the variance in the time between departures from the current workstation, which is a function of the variance in the time between arrivals, the variance in the cycle time, and the utilization as given in equation 6. (6) note that even in the case where the variance of the time between arrivals is zero, the variance in the time between departures will be greater than zero if the variance in the cycle time is greater than zero. based on equations 5 and 6, the observed variance in the cycle time at each station must be less than the maximum allowed in order to support the use of deterministic simulation. 4. results and example consider an existing parts manufacturing system as shown in figure 4 with the structure given in figure 1. figure 4. structure of parts manufacturing system (source: authors’ drawing). the assembly cell produces a final product which consists of the subassemblies produced by pt and hpw, which are carried to the assembly cell by the runner. mag provides a subassembly to hpw. mag and hpw are located near enough to each other to allow the direct transfer of the subassembly produced by mag. takt time is computed as shown in table 3. recall that the takt time is the quotient of the available work time per day, 1305 minutes, and the demand per day, 840 parts. the supplier uses a 10% allowance to control for unforeseen events such as breakdowns, thus lowering the takt time from 93.2 seconds to 83.9 seconds per part. the average cycle time and observed coefficient of variation were estimated from 10 observations of the cycle time for each work element at each workstation. these were collected by manual observation. int. j. prod. manag. eng. (2021) 9(1), 15-23creative commons attribution-noncommercial-noderivatives 4.0 international validation of production system throughput potential and simulation experiment design 19 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 3. takt time determination (source: authors’ data collection). item value units demand / day 840 parts work time / shift 435 min shifts / day 3 work time / day 1305 min takt time 1.55 min takt time 93.2 sec allowance 10% takt time with allowance 83.9 sec table 4 shows the utilization of each workstation in each cell along with the cycle time average and squared coefficient of variation computed from the collected data and the takt time. in addition, the maximum squared coefficient of variation allowed for deterministic simulation computed using equations 5 and 6 is shown for three values of the coefficient of variation of the time between arrivals to the first station in the cell: cvtba= 0.0, 0.2, and 1.0. for every station and each cvtba value, the squared coefficient of variation of the cycle time (cv2) is much less, by orders of magnitude, than the maximum allowed squared coefficient of variation (max cv2). thus, deterministic simulation is appropriate regardless of the variation in the time between arrivals. thus, the simulation model has the structure shown in figure 5, which follows from figure 2. figure 5. simulation model structure for the parts manufacturing system (source: authors’ drawing). there are five processes, one for each of the cells and one for the runner. the assembly process waits for a subassembly from the both the pt and hpw processes before proceeding to produce one unit of finished goods inventory. the pt and mag processes use readily available raw material which is not modeled. the hpw process uses one subassembly from the mag process. the runner’s route is from assembly to pt to hpw and back to assembly. the runner completes one loop in 15 minutes. five minutes are allocated for each of the three stations: assembly, pt, and hpw which includes walking between stations, delivering subassemblies, and picking up subassemblies to move. thus, a station must have a wip inventory of at least 11 subassemblies (= 15 minutes / 83.9 seconds and rounded up) to avoid starvation between arrivals of table 4. variation analysis for deterministic simulation (source: authors’ analysis). cell station utilization cycle time average cv2 max cv2 cvtba 0.0 0.2 1.0 pt pt1 0.48 40.42 0.011 4.47 4.43 3.47 pt2 0.65 54.90 0.0048 1.61 1.59 1.04 hpw hpw 0.71 59.85 0.010 1.13 1.09 0.13 mag mag 0.53 44.82 0.011 3.26 3.22 2.26 assembly ml1a 0.70 58.80 0.0057 1.22 1.18 0.22 ml1b 0.81 67.61 0.0053 0.59 0.58 0.24 ml2 0.65 54.89 0.073 1.59 1.56 1.01 ml3 0.81 68.27 0.0022 0.56 0.55 0.22 ml4 0.91 76.10 0.0094 0.22 0.21 0.04 eol 0.69 57.65 0.0070 1.32 1.30 0.79 audit 0.52 43.72 0.017 3.52 3.49 2.80 int. j. prod. manag. eng. (2021) 9(1), 15-23 creative commons attribution-noncommercial-noderivatives 4.0 international standridge and wynne 20 http://creativecommons.org/licenses/by-nc-nd/4.0/ the runner. however, the subassemblies are organized into totes with capacity 8. at least two totes, 16 subassemblies, are needed to avoid starvation. this is the initial value of the wip_pt and wip_hpw at the assembly station as well as wip_mag which is shared between the mag and hpw stations. furthermore, at the beginning of the simulation, there is one tote in wip_pt at the pt station and one tote in wip_hpw at the hpw station. thus, the simulation begins with a total wip inventory of 64 subassemblies. the key results from the deterministic simulation are as follows: 1. the utilization reported by the simulation for each workstation is exactly the same as shown in table 4. 2. the number of finished units produced in one week (5 days) is 4662, which exceeds the demand of 4200 parts. 3. the average wip inventory in the entire production system is 75 subassemblies and the maximum 76 subassemblies. an alternative configuration of the production system has been proposed under which all cells are co-located so that the runner is not needed. in addition, the work elements for the stations comprising the assembly cell are consolidated such that the audit station is no longer needed. further, the work elements for the pt cell are improved such that only one workstation is needed. the utilization and variation analysis for this configuration are shown in table 5. note that while deterministic simulation is feasible for cvtba = 0.0 and 0.2, it is not feasible for cvtba= 1.0. the maximum allowed value of cv2ct for pt is less than zero and cv2ct for ml4 is greater than the maximum allowed value. the assembly cell must have enough wip to avoid starvation. in the proposed configuration, one tote of each type of subassembly, those from pt and hpw should be sufficient. in the same way, the hpw cell needs one tote of subassemblies from the mag cell. thus, the simulation begins with a total wip inventory of 24 subassemblies. the key results from the deterministic simulation are as follows: 1. the utilization reported by the simulation for each workstation is exactly the same as shown in table 5. 2. the number of finished units produced in one week (5 days) is 4666, which exceeds the demand of 4200 parts. 3. the average wip inventory in the entire production system is 33 subassemblies and the maximum 34 subassemblies. 5. discussion kingman’s equation has been widely used to analyze production systems as discussed in hopp and spearman (2011) and standridge (2019). the variation term in kingman’s equation has two components: the variation in cycle time and in the time between arrivals (demand). previous work is extended by determining the level of variation in table 5. variation analysis for deterministic simulation – consolidated workstations (source: authors’ analysis). cell station utilization cycle time average cv2 max cv2 cvtba 0.0 0.2 1.0 pt pt 0.78 65.60 0.17 0.71 0.67 -0.29 hpw hpw 0.71 59.85 0.010 1.13 1.09 0.13 mag mag 0.53 44.82 0.011 3.26 3.22 2.26 assembly ml1a 0.72 60.18 0.09 1.10 1.06 0.10 ml1b 0.82 69.12 0.07 0.48 0.46 0.15 ml2 0.84 70.22 0.05 0.43 0.42 0.13 ml3 0.85 70.96 0.05 0.39 0.38 0.11 ml4 0.86 72.14 0.11 0.30 0.29 0.04 eol 0.84 70.63 0.08 0.39 0.38 0.10 int. j. prod. manag. eng. (2021) 9(1), 15-23creative commons attribution-noncommercial-noderivatives 4.0 international validation of production system throughput potential and simulation experiment design 21 http://creativecommons.org/licenses/by-nc-nd/4.0/ cycle time that is consistent with average waiting time preceding a workstation being less than the takt time given the variation in the time between arrivals. this condition indicates that flow in the production system has been achieved, one of the chief goals of lean. furthermore, this is the condition under which deterministic simulation can be used to validate production throughput potential. the practical worst case, cvtba = 1, models demand coming from consumers as opposed to another business. as seen in figure 3 and table 2 for this case, the maximum allowed squared coefficient of variation is positive for workstations with utilizations of 73% and less. thus, low utilization workstations can be modeled as deterministic if the variation in the cycle time is low enough which is likely the case for a utilization of 65% or less. to model demand coming from another business, cvtba = 0.0 or 0.2 was used. in this case, workstations with utilizations up to 90% and reasonable small variation can be modeled using deterministic simulation as can workstations with utilizations between 90% and 99% having very small cycle time variation. also note that the maximum allowed variation declines in a non-linear fashion as the utilization increases. the use of deterministic simulation versus stochastic simulation greatly simplifies experimentation and results analysis as well as reducing computer execution time. the example demonstrates the effectiveness of this approach. the model is validated by comparing the utilization of each workstation computed from gather data with that computed by the simulation. these quantities were found to be the same for each workstation for each configuration. in addition, each configuration was shown to be capable of meeting the throughput target set by the customer with near constant work-in-process inventory as the average and maximum values differ by only one. the proposed reconfiguration results in reducing the wip by over 50% as well as eliminating the runner and two workstations. thus, it is preferred. 6. conclusions criteria for using deterministic simulation of production systems versus stochastic simulation are established. this adds to the existing literature both of the use of deterministic simulation and to the application of kingman’s equation to production systems. the criteria are straightforward to apply requiring only the comparison of a quantity that is a function of workstation utilization to the known or estimated variation in the cycle time at each workstation. both are expressed as the squared coefficient of variation. references atalan, a., dönmez, c.c. (2020). optimizing experimental simulation design for the emergency departments. brazilian journal of operations & production management, 17(4), e2020854. https://doi.org/10.14488/bjopm.2020.026 askin, r.g., standridge, c.r. (1993). modeling and analysis of manufacturing systems. new york: john wiley and sons. dagkakis, g., rotondo, a., heavey, c. (2019). embedding optimization with deterministic discrete event simulation for assignment of cross-trained operators: an assembly line case study. computers and operations research, 111, 99-115. https://doi.org/10.1016/j. cor.2019.06.008 ferrin, d.m., miller m.j., muthler d. (2005). lean sigma and simulation, so what’s the correlation?, in proceedings of the 2005 winter simulation conference, ieee, usa. retrieved july 22, 2020 from: https://informs-sim.org/wsc05papers/249.pdf hopp, w.j., spearman, m.l. (2011). factory physics: foundations of manufacturing management, 3rd ed. long grove, il: waveland press. jayaraman, a., gunal, a.k. (1997). applications of discrete event simulation in the design of automotive powertrain manufacturing systems”. in proceedings of the 1997 winter simulation conference, ieee, usa. https://doi.org/10.1145/268437.268620 khan, s., standridge, c.r. (2019). aggregate simulation modeling with application to setting the conwip limit in an hmlv cell. international journal of industrial engineering computation, 10(2), 149-160. https://doi.org/10.5267/j.ijiec.2018.10.002 kingman, j.f.c. (1961). the single server queue in heavy traffic. mathematical proceedings of the cambridge philosophical society, 57(4), 902. https://doi.org/10.1017/s0305004100036094 kleijnen, j.p.c. (2015). design and analysis of simulation experiments. new york: springer. https://doi.org/10.1007/978-3-319-18087-8 kleijnen, j.p.c., standridge, c.r. (1988). experimental design and regression analysis in simulation: an fms case study. european journal of operations research, 33, 257-261. https://doi.org/10.1016/0377-2217(88)90168-3 law, a.m. (2014). simulation modeling and analysis, 5th ed. new york: mcgraw-hill. int. j. prod. manag. eng. (2021) 9(1), 15-23 creative commons attribution-noncommercial-noderivatives 4.0 international standridge and wynne 22 https://doi.org/10.14488/bjopm.2020.026 https://doi.org/10.1016/j.cor.2019.06.008 https://doi.org/10.1016/j.cor.2019.06.008 https://informs-sim.org/wsc05papers/249.pdf https://doi.org/10.1145/268437.268620 https://doi.org/10.5267/j.ijiec.2018.10.002 https://doi.org/10.1017/s0305004100036094 https://doi.org/10.1007/978-3-319-18087-8 https://doi.org/10.1016/0377-2217(88)90168-3 http://creativecommons.org/licenses/by-nc-nd/4.0/ little, j.d.c. (1961). a proof for the queuing formula: l = λw. operations research, 9(3), 383-387. https://doi.org/10.1287/opre.9.3.383 marvel, j.h., standridge, c.r. (2009). a simulation enhanced lean design process. journal of industrial engineering and management, 2(1), 90-113. https://doi.org/10.3926/jiem.2009.v2n1.p90-113 mourtzis, d. (2019) simulation in the design and operation of manufacturing systems: state of the art and new trends. international journal of production research, 58(7), 1927-1949. https://doi.org/10.1080/00207543.2019.1636321 pinheiro, n.m.g, cleto, m.g., zattar, i.c., muller, s.i.m.g. (2019). performance evaluation of pulled, pushed and hybrid production through simulation: a case study. brazilian journal of operations & production management, 16, 685-697. https://doi.org/10.14488/bjopm.2019.v16.n4.a pritsker, a.a.b. (1989). why simulation works. in proceedings of the 1989 winter simulation conference, ieee, usa. https://doi.org/10.1145/76738.76739 puvanasvaran, p., teoh, y.s., ito, k. (2020). novel availability and performance ratio for internal transportation and manufacturing processes in job shop company. journal of industrial engineering and management, 13(1), 1-17. https://doi.org/10.3926/jiem.2755 sanchez, s.m., sanchez, p.j., wan, h. (2020). work smarter, not harder: a tutorial on designing and conducting simulation experiments. in proceedings of the 2020 winter simulation conference, ieee, usa. retrieved december 23, 2020 from https://informs-sim.org/ wsc20papers/135.pdf schruben, l. (1983). simulation modeling with event graphs. communications of the a.c.m., 26(11). https://doi.org/10.1145/182.358460 spearman, m.l., woodruff, d.l., hopp, w.j. (1990). conwip: a pull alternative to kanban, international journal of production research, 28(5), 879-894. https://doi.org/10.1080/00207549008942761 standridge, c.r. (2019). introduction to production: philosophies, flow, and analysis. allendale michigan: grand valley state university libraries. retrieved july 22, 2020 from: https://scholarworks.gvsu.edu/books/22/ tapping, d., luyster, t., shuker, t. (2002). value stream management. boca raton: crc press. https://doi.org/10.4324/9781482278163 tribastone, m., vandin, a. (2018). speeding up stochastic and deterministic simulation by aggregation: an advanced tutorial. . in proceedings of the 2018 winter simulation conference, ieee, usa. https://doi.org/10.1109/wsc.2018.8632364 uriarte, a.g., ng, a.h.c., moris, m.u. (2020). bringing together lean and simulation: a comprehensive review, international journal of production research, 58(1), 87-117. https://doi.org/10.1080/00207543.2019.1643512 zupan, h., herakovic. n. (2015). production line balancing with discrete event simulation: a case study”, ifac-papersonline, 48(3), 23052311. https://doi.org/10.1016/j.ifacol.2015.06.431 int. j. prod. manag. eng. (2021) 9(1), 15-23creative commons attribution-noncommercial-noderivatives 4.0 international validation of production system throughput potential and simulation experiment design 23 https://doi.org/10.1287/opre.9.3.383 https://doi.org/10.3926/jiem.2009.v2n1.p90-113 https://doi.org/10.1080/00207543.2019.1636321 https://doi.org/10.14488/bjopm.2019.v16.n4.a https://doi.org/10.1145/76738.76739 https://doi.org/10.3926/jiem.2755 https://informs-sim.org/wsc20papers/135.pdf https://informs-sim.org/wsc20papers/135.pdf https://doi.org/10.1145/182.358460 https://doi.org/10.1080/00207549008942761 https://scholarworks.gvsu.edu/books/22 https://doi.org/10.4324/9781482278163 https://doi.org/10.1109/wsc.2018.8632364 https://doi.org/10.1080/00207543.2019.1643512 https://doi.org/10.1016/j.ifacol.2015.06.431 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2019.11435 received: 2019-04-02 accepted: 2019-05-20 modelling inventory management: separate issues for construction and application luchko, m.r. a1*, lukanovska, i.r. a2, ratynskyi, v. b a department of accounting and control in public administration, ternopil national economic university, ternopil, ukraine. b department of accounting and taxation, ternopil ivan puluj national technical university, ternopil, ukraine. a1 m_luchko@ukr.net, a2 i.lukanovska@tneu.edu.ua b ratvadim@gmail.com abstract: the article presents the model of inventory management for forestry enterprises. it allows us to make calculations of the efficiency of the use of inventories, helps minimize the cost of transportation and storage and to avoid fines for lack of inventory in the presence of demand. the suggested model takes into account the dependence of the price of the received reserves, warehouse costs, the cost of holding stocks, the amount of storage costs and penalties for customer stocks that have been unloaded on time. the model will allow us to avoid delays in supply, determine the size and guarantee optimal inventory levels. it is proved that rational inventory management enables companies to calculate the optimal amount of reserve ordering, and the time interval between such orders. putting this model into practice will allow for managing of the logistical and warehouse costs, determining the risk of not receiving or reducing the company’s profits due to the excess inventory costs, and understanding of the efficiency of turnover of inventories. purpose. the aim of the study is to construct an economic and mathematical model of inventory management of the enterprise, taking into account the dependence of the shipping cost from the supplier to the warehouse, the cost of holding the stocks, the amount of storage costs and penalties for non-shipped products. methodology. approbation of theoretical developments was carried out in the application package statistica and the use of the module marspline – an integral element of technology data mining results. the theoretical contribution. the main dependencies influencing the formation of the value of stocks were discovered on the basis of this and a regression model of the optimal size of stocks was constructed with a fairly precise approximation; calculations were made of the efficiency of the use of inventories, minimizing the total costs associated with delivery, storage and fines for the absence of stocks at availability of demand for them. practical implications. we have constructed, investigated and tested the economic and mathematical model of the optimal size of stocks, which takes into account the dependence of the shipping cost from the supplier to the warehouse, the cost of holding the stocks, the amount of storage costs and penalties for non-shipped products. key words: trade enterprises methods, modelling, inventory economic and mathematical models, optimization, spline surfaces, random and independent values, data mining results, marspline, statistica package. 1. introduction forestry enterprises in ukraine operate under conditions of high competition, uncertainty, and risk. this is due to the political risk of the state and the instability of tax legislation. therefore, it is of great importance for the sustainable functioning of these companies that forestry enterprises be effectively managed. in order to make effective management decisions, modern scientific approaches and practicetested techniques, namely: forecasting, econometric and mathematical modeling of economic objects, production processes and individual economic to cite this article: luchko, m.r., lukanovska, i.r., ratynskyi, v. (2019). modelling inventory management: separate issues for construction and application. international journal of production management and engineering, 7(2), 117-124. https://doi.org/10.4995/ijpme.2019.11435 int. j. prod. manag. eng. (2019) 7(2), 117-124creative commons attribution-noncommercial-noderivatives 4.0 international 117 http://orcid.org/0000-0001-6499-4188 http://orcid.org/0000-0001-5591-6487 http://orcid.org/0000-0001-9283-6371 http://creativecommons.org/licenses/by-nc-nd/4.0/ situations, should be utilized. in the process of modeling the economic situations associated with the use of reserves, in practice, a number of tasks should be solved (table 1). table 1. the content of solving particular tasks of inventory management optimization problems. no tasks content 1 the task of distribution of stocks arises when a certain set of economic activities (work) that should be implemented by the small amount of resources 2. the task of inventory management they are in search of the best values of reserve levels and order sizes to meet the permanent production process 3. the task of forecasting placing reserves associated with the definition of the optimal number and location of new objects, taking into account their interaction with existing objects and interactions among themselves modeling of inventory management makes it possible to design a rational scheme of continuous and effective provision of material resources. for its implementation, the following tasks are required: study of accounting of the current level of inventories based on information flows of data; inventory analysis to determine the current level of stocks that will ensure continuity of activities; analysis of current programs in order to calculate the size of reserves; determining the intervals of time necessary to carry out the production process. sectoral and technological features of forestry enterprises have a great influence on the management system. this is explained by the organization of activities and complex logistics. it relies on production processes, product sales, commodity flow and cash flows. forestry companies hold in question the optimal batch of ordering stocks of raw materials or goods. this is because the supply of a considerably large quantity of stocks or a surplus, greatly increases the cost of their logistics and storage, and may reduce their commodity value, and hence reduce the sale price. otherwise, when the supply of reserves is less than the optimal demand or orders available, there will be a deficit of sales, that is, the enterprise will not be able to fulfill the contractual obligations in time, and will not receive a profit and may be punished by penal sanctions. at the same time, it should also be taken into account that a significant part of the logistical, warehouse and conservation costs are constant, do not vary depending on the volumes of work, and therefore affect the financial performance of the activity. consider the solution inventory management problem in woodworking enterprises. with the help of a mathematical model that allows us to calculate the efficiency of the use of inventories, we can minimize the total costs associated with their delivery, and decrease storage and fines in the absence of stocks in the presence of demand for them. in response to the premise of our study, we propose the following hypotheses or assumptions that will be verified experimentally for a possible solution of the problem. first, there is a connection between the amount of shipping costs, storage costs, and penalties that are associated with the lack of wood when necessary. secondly, during certain periods, wood-processing companies can reimburse the cost of storing inventory to avoid imposing penalties for non-grassland timber products. 2. literature review questions that are subject to research within the framework of this article have been the subject of many scientific investigations, depending on the content of the optimization tasks of inventory management. an increasing interest is emerging towards optimization modeling of stock management processes, it is recognized as an effective apparatus for formalizing scientific and technical problems in various fields of knowledge. paying tribute to many scholars who have made a significant contribution to solving theoretical and practical problems in the specified range of issues, in the framework of this study, pay attention only to those that, in our opinion, have a direct impact on the subject of research. shygun (2008) conducted an analysis of the content of scientific articles and reports on economic modeling in ukraine, presented in professional editions and scientific conferences with the use of the bibliometric method, systematization of articles and reports on the issues of open issues, and the tendencies of economic modeling development were revealed. int. j. prod. manag. eng. (2019) 7(2), 117-124 creative commons attribution-noncommercial-noderivatives 4.0 international luchko et al. 118 http://creativecommons.org/licenses/by-nc-nd/4.0/ atamanchuk and pasenchenko (2016) investigated the problem of choosing the optimal inventory management policy. they developed an algorithm for optimizing the parameters of the inventory management system, disclosed the features of efficient inventory management by the receipt of such (optimal) stock size, at which costs for their maintenance and maintenance would be minimal, and the number of stocks − sufficient for the stable operation of the business entity. an economical and mathematical model that allows us to choose the optimal policy for inventory management in an unstable economy is presented in the paper, as well as numerical calculations have been carried out demonstrating the application of this model at the garment industry enterprise (andrushkiv et al., 2012; bukan & kenіgsberg, 2007; kaplan & norton, 2006; kovács & kot, 2017; popovic et al., 2017; robert, 1992). the time point at which the delivery of the order should be designated, which allows maintaining stocks at the optimal level, while simultaneously reducing the cost of storing the stock and the loss from the shortage of goods, is determined. demand for goods in the presented model is considered as a random variable with a normal probability distribution. isonin and lagotsky (2012) studied the problem of ensuring the optimal balance between minimizing investment in stocks on the one hand and maximizing the level of service users of the enterprise for a continuous production process – on the other. it is indicated that many theoretical and methodological aspects of inventory management, taking into account the stochastic nature of demand, remain little studied and poorly covered in the economic literature. kushnirenko and ralle (2015) reviewed the main approaches to the analysis of commodity and material inventory in the field of production and warehouse services and explored the methods of modeling the processes of material inventory management and their impact on the operation of the enterprise. they proposed the rationing of commodity inventories proposed by various methods such as: experimental and statistical method, method of technical and economic calculations, economic and mathematical modeling, etc. pasenchenko and trubnikova (2011) reviewed inventory management issues at a trading enterprise. they define the main indicators of the activity of a trading enterprise within the framework of the concept of balanced development of the company, analyze their classification and the main factors that influence their change and can be called managers and used to improve the work of the enterprise. inventories stocks are related to objects according to a dan dzo (2015) at the enterprise level that require significant investment. so, therefore, they are one of the factors that determine the policy of an enterprise and affect its level of liquidity, and for trading companies for profit. shraybfeder (2005) noted that efficient inventory management allows you to: serve the buyer well, to ensure return on investment and to eliminate dead stocks and surpluses. according to takha (2005) classical inventory management models and their level control are based on the fact that inventory management is a complex set of measures aimed at providing the highest possible level of service to customers with minimum current costs associated with holding stocks. in practice, inventory management is limited to two main questions: when to replenish stock and in what quantity? the classic model of stock management is the wilson model, which allows you to calculate the optimal amount of the lot and the time period of ordering. min and zhou (2002) pointed out that over the years, most firms have focused on the overall effectiveness and efficiency of individual business functions. however, as a new way of doing business, businesses began to realize the strategic importance of planning, controlling and designing the supply chain as a whole. in an effort to help firms capture the synergy between inter-functional and inter-organizational integration and supply chain coordination and subsequently make better decisions on the supply chain, they drew attention to the synthesis of previous efforts to model supply chains. franca, jones, richards and carlson (2010) drew attention to elements of fuzzy logic for optimizing, monitoring and controlling the process of executing orders in the supply chain of the global retail firm. they introduced a model for optimizing order fulfillment, which improves the integration of the supply chain and cooperation between supply chain partners through effective monitoring and control of supply chain variables. this model examines the int. j. prod. manag. eng. (2019) 7(2), 117-124creative commons attribution-noncommercial-noderivatives 4.0 international modelling inventory management: separate issues for construction and application 119 http://creativecommons.org/licenses/by-nc-nd/4.0/ critical requirements of the customer at the stage of supply chain development, making it a useful model by differentiating customers and separating supply channels. melo, nickel and da gama (2006) and takha (2005) focused on strategic planning of supply networks based on mathematical modelling, which simultaneously covers many practical aspects of network design problems: dynamic planning horizon, overall supply chain network, external material supply, opportunities of product inventory, product distribution, object configuration, capital availability for investments and storage restrictions. seuring (2013) summarized the research of quantitative models of forward chain supply. he drew attention to the social and environmental problems of the supply chain and noted that on the modeling side there are three dominant approaches: the equilibrium model, the multicriteria process of decision making and the analytical hierarchical process. only limited empirical studies were conducted by this time. beamon (1998) has proven that over the years, theorists and practitioners have primarily studied various processes in the supply chain of production individually. at the same time, the development and analysis of the supply chain as a whole has been neglected. this focus is largely due to rising production costs, resource base reductions in production bases, shortening the product life cycle, leveling the playing field in production, and the globalization of market economies. biswas and narahari (2004) investigated the possibility of stochastic models, models of mathematical programming, heuristic methods and modeling in the supply chain simulation. because different problems for making decisions in supply networks require different approaches to modeling and solving problems, there is a need for a unified approach to supply chain modeling so that any necessary solution can be created in a fast and flexible way. in this paper, they have developed a decision support system for desscom (support for solutions for supply chains using object simulation), which enables strategic, tactical and operational making decision in supply chains. scientific investigations of issues related to the management of material inventories and their distribution on the basis of the application of economic and mathematical modeling remain the requirements of time. taking into account the above and sectoral features of woodworking enterprises, one can conclude that research is incomplete in this direction and the need to find a model that would allow them to manage their inventories in order to minimize unproductive costs. 3. research methods the modelling of inventory management is closely linked with such components of the theory of management of economic systems as the information theory of hierarchical systems; target and oriented planning; the theory of project management; the theory of contracts that investigate problems in an uncertain environment. through the system analysis, we analyzed the activities of woodworking enterprises, identified the main areas that require further research. since the main task of these companies is profit, we used software packages statistica, and the module marspline with module data mining results. this enabled to calculate regression model in multidimensional space and build a spline surfaceplanned dependencies through the basic functions. the obtained regression model established the relationships between variables with a rather precise approximation. 4. research results the development of a mathematical model of enterprise inventory management will contribute to increasing the efficiency of inventory turnover. let’s assume that there is a stock of wood (t) of time periods in the warehouse. indicate the amount of wood stock (z0) meters cubic at the beginning of the period. demand (vy) on timber forms the amount of their delivery from the warehouse in (y) period. in each separate period demand (vy) is a random and independent value with probability density fy(vy). price of shipment volume (m3) ordered at the beginning of the wood period from the supplier to the warehouse until the end (y)th of the period is (qy). the cost of maintaining a certain amount of timber int. j. prod. manag. eng. (2019) 7(2), 117-124 creative commons attribution-noncommercial-noderivatives 4.0 international luchko et al. 120 http://creativecommons.org/licenses/by-nc-nd/4.0/ in the (y) – period is (gy), and the price of placing the stock is proportional to its volume at the end of the period is equal (gyz). the amount of expenses for delivery, storage, and penalties related to the lack of wood, if necessary, depends on the proper receipt of the timber for the warehouse business entity. let’s notice (ky) the volume of (m 3) of the ordered wood at (y) – period. hence, the stock (zy) of timber for the end (y)thof the period will be (1): z z k vy y i y y i y 0 1 1 = + = = | | (1) the number of undisturbed wy of wood to maintain optimum functioning of the composition zy ≥ 0 due to its absence at warehouses until the end of the y-th period will have the form (2): w v k z w 0y y i y y y i y 1 0 1 $= = = ^ h| | (2) thus, if during a certain (y-th) period a wood processing enterprise will reimburse the cost of storing inventory (zy≥0), there will be no penalties for non-donation (wy = 0) and vice versa. this situation can be represented as follows (3): when kx z v g z when z w w 0 0 >yi y y y y y i y y y y y 11 0 $ r + = = = d n (|| (3) the total cost of the warehouse (c) of woodworking enterprise for (t) periods can be written down as: x z k vc g k y y i y y i yt y y y y t 0 1 11 1 = + + = = = = d n| || | (4) this formula (4) is a function of random and independent values vy. it is necessary to calculate the volume of timber that has already been delivered and the demand for it to substantiate the volume (m3) of the order of wood at the beginning of the next period. thus, it is not possible to set all values of t for ky at the same time. summarizing the above, we note that in the conditions of dynamic development of computer technologies, management decisions regarding orders in the planning period are taken on the basis of the results of the reporting period. this implies the following: for any (n)-th period we have k*n = k * n (zn). we will follow the use of this model at the wood processing enterprise ltd “hansakom-west”. let’s calculate ztheor by the data on the remnants of wood in warehouses and storerooms at the beginning of the reporting period, the volume of timber products that left the period, and their volume, which passes from the warehouses in the y-th period, which is determined by vy demand for goods (the initial data is given in table 2). in figure 1, we show the quadratic surface on the basis of data in table 2: . . . . . . z v k v vk k 39771 4135 7 5882 3 1036 0 0005 0 0005 0 0001 theor 2 2 $ $= + + + + (5) table 2. stock of wood at the end of the y-th period, thousand uah. period, month demand, ν received, k value, ztheor 1. 18256 16890 2156 2. 18749 17450 1759 3. 17256 18652 1632 4. 18413 18509 2587 5. 19785 21687 2374 6. 21059 21059 1960 7. 19760 19760 3658 8. 17694 16751 2520 9. 18340 17996 2136 10. 19761 19761 1825 11. 19800 19800 957 12. 16543 15266 1653 figure 1. the quadratic surface of the theoretical stock (ztheor) according to the table 2. as a result, we obtain the estimated values of theoretical data volumes (m3) of inventories, comparing with which real volumes we will find variants of immobilization of funds invested in production reserves. int. j. prod. manag. eng. (2019) 7(2), 117-124creative commons attribution-noncommercial-noderivatives 4.0 international modelling inventory management: separate issues for construction and application 121 http://creativecommons.org/licenses/by-nc-nd/4.0/ hansakom-west ltd (ukraine) uses the method of paying for the storage of reserve stocks of wood in warehouses in the amount of five-day volume of use of timber (0.03% from the cost of materials). now, it is necessary to calculate the cost of storing the reserve volume of the raw forest (table 3) and to construct the surface of the smallest squares (figure 2). table 3. amount of payments for storage of timber, thousand uah. period, month value, ztheor cost of storage, g the amount of payment for storage, x 1. 3522 0.0598 210.96 2. 3056 0.0647 198.02 3. 3026 0.0408 123.47 4. 2683 0.0656 176.20 5. 4276 0.0627 268.11 6. 1965 0.0496 97.64 7. 2368 0.0591 140.17 8. 3463 0.0526 182.46 9. 2480 0.0437 108.45 10. 2652 0.0481 127.63 11. 1860 0.0463 86.21 12. 2930 0.0535 157.03 figure 2. the surface of the smallest squares according to table 3. ztheor = 2766.8 – 49620.8 · g + 17.67 · x (6) 0 500 1000 1500 2000 2500 3000 3500 4000 4500 1 2 3 4 5 6 7 8 9 10 11 12 z theor z actual figure 3. dynamics of actual and theoretical data on the turnover of wood stocks, thousand uah. source: it was formed by the author indicator r2=0.988 indicates the adequacy and accuracy of the proposed model. let’s introduce the value of ztheor and zactual graphically, which will help compare the results obtained during computations with the actual results in figure 3. calculate the amount of payments for the theoretical volume (m3) of the forest raw material for comparison with the real volume (m3). results of calculation of theoretical and actual costs borne by a woodworking enterprise in the process of keeping the timber will be reflected in figure 4. figure 4. dynamics of expenses for storage of the theoretical and actual volume of wood, thousand uah. source: authors table 4. determination of the cost of warehouse for the storage of the actual volume of wood by periods, thousand uah. period, month amount of payment for storage, x received, k cost of delivery, q total cost of warehouse, c 1 210.96 16890 0. 1254 383.54 2 198.02 17450 0.1332 396.59 3 123.47 18652 0.1410 414.86 4 176.20 18509 0.1394 411.37 5 268.11 21687 0.2185 492.88 6 97.64 21059 0.2663 478.61 7 140.17 19760 0.1982 449.09 8 182.46 16751 0.1687 380.70 9 108.45 17996 0.1410 409.36 10 127.63 19782 0.1865 451.73 11 86.21 19800 0.1935 462.58 12 157.03 15266 0. 1478 346.95 int. j. prod. manag. eng. (2019) 7(2), 117-124 creative commons attribution-noncommercial-noderivatives 4.0 international luchko et al. 122 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 5. quadratic surface according to table 4. . . . . . . c x q x xq q 484 7 1 8 199 97 0 005 0 083 4 77– – – 2 2 $ $= + + + (7) thus, the carried out calculations allow to make optimal managerial decisions regarding the volume and structure of production stocks, accordingly adjusting the work of warehouse areas. 5. discussion of the results the conducted research has made it possible to identify the main factors influencing the management of stocks of woodworking enterprises in order to minimize unproductive costs. the solution of the proposed model in the application package statistica and the use of the module marspline – an integral element of the technology data mining results and the resulting regression equations allow to determine the general dependencies. this enabled through the basic functions calculate regression model in multidimensional space and build a spline surface and planned dependencies. the obtained regression model established the relationships between variables with a rather precise approximation. the constructed generalized economic and mathematical model of inventory management allows minimizing unproductive costs. the model can be adapted for woodworking enterprises, as well as others, which are characterized by similar technological features of activity. in addition, it should be noted that, in our opinion, the study conducted allowed to form new scientific problems of great theoretical and practical importance and could become the subject of further scientific research. these, in the first place, should include: optimization of management of financial capital of economic entities of different forms of ownership; intensification of the use of blockade and artificial intelligence technologies in conditions of uncertainty of logistics of stocks and management of them. the possible direction of further research on this problem is also to take into account in this model the future value of money and the effect on it of the inflationary effect of the depreciation of money. in our opinion, this requires additional substantiation and changes in the individual components of the calculations. 6. conclusions the management of stocks under uncertainty should be as much as possible protected from risk operations in order to prevent unproductive costs. therefore, the transparency of the data is a requirement of time. in this case, the issue of not only obtaining additional funds for inventory management, but also their efficient and well-considered use becomes a matter of particular importance. in addition, the developed organizational, methodological and methodical tools can be used to solve a wide range of more narrow and specific topical scientific and theoretical applied tasks related to the construction of models of business entities. references a dan dzo. (2015). inventory control. available at: http://www.adandzo.com/consulting/82/ accessed 22 august 2014. agarwal, a., shankar, r., tiwari, m.k. (2006). modeling the metrics of lean, agile and leagile supply chain: an anp-based approach. european journal of operational research, 173(1), 211-225. https://doi.org/10.1016/j.ejor.2004.12.005 agarwal, a., shankar, r., tiwari, m.k. (2007). modeling agility of supply chain. industrial marketing management, 36(4), 443-457. https:// doi.org/10.1016/j.indmarman.2005.12.004 andrushkiv, b.m., vovk, y.y., vovk, i.p. (2012). resursonomika: teoretychni ta prykladni aspekty [resourcenomy: theoretical and applied aspects]. ternohraf, ternopil, ukraine. int. j. prod. manag. eng. (2019) 7(2), 117-124creative commons attribution-noncommercial-noderivatives 4.0 international modelling inventory management: separate issues for construction and application 123 http://www.adandzo.com/con https://doi.org/10.1016/j.ejor.2004.12.005 https://doi.org/10.1016/j.indmarman.2005.12.004 https://doi.org/10.1016/j.indmarman.2005.12.004 http://creativecommons.org/licenses/by-nc-nd/4.0/ atamanchuk, yu.s., pasenchenko, yu.a. (2016). modeling of enterprise’s inventory management with allowance for uncertain demand. actual problems of economics and management, 10, 115-123. beale, e.m. (1955). on minimizing a convex function subject to linear inequalities.  journal of the royal statistical society: series b (methodological), 17(2), 173-184. https://doi.org/10.1111/j.2517-6161.1955.tb00191.x beamon, b. m. (1998). supply chain design and analysis:: models and methods. international journal of production economics, 55(3), 281294. https://doi.org/10.1016/s0925-5273(98)00079-6 biswas, s., narahari, y. (2004). object oriented modeling and decision support for supply chains.  european journal of operational research, 153(3), 704-726. https://doi.org/10.1016/s0377-2217(02)00806-8 bukan, d., kenіgsberg, e. (2007). nauchnoe upravlenie zapasami [scientific inventory management]. moscow: nauka, russia. ermoliev, y. m., norkin, v.i. (1997). on nonsmooth and discontinuous problems of stochastic systems optimization. european journal of operational research, 101(2), 230-244. https://doi.org/10.1016/s0377-2217(96)00395-5 franca, r.b., jones, e.c., richards, c.n., carlson, j.p. (2010). multi-objective stochastic supply chain modeling to evaluate tradeoffs between profit and quality. international journal of production economics, 127(2), 292-299. https://doi.org/10.1016/j.ijpe.2009.09.005 isonin, i.v., lagotsky, t.ya. (2012). the problem of stochastic modeling in the management of production reserves. scientific herald of nltu of ukraine, 22(5), 351-356. kaplan, r.s., norton, d.p. (2006). alignment: using the balanced scorecard to create corporate synergies. harvard business press. kovács, g., kot, s. (2017). economic and social effects of novel supply chain concepts and virtual enterprises. journal of international studies, 10(1), 237-254. https://doi.org/10.14254/2071-8330.2017/10-1/17 kushnirenko o.s., ralle, n.v. (2015). simulation of inventory management. visnyk knutd, 2(85), 69-73. luchko, m.r. (2016). contract theory: selected issues of accounting and analysis. socio-economic problems and the state, 2(15), 71-77. luchko, m.r. (2018). analytical procedures in the process of inventory management audit. in proceedings iii international conf. accounting, taxation and control: theory and methodology [ternopil, june 26, 2018]. ternopil: tneu, ukraine, 95-97. melo, m.t., nickel, s., da gama, f.s. (2006). dynamic multi-commodity capacitated facility location: a mathematical modeling framework for strategic supply chain planning. computers & operations research, 33(1), 181-208. https://doi.org/10.1016/j.cor.2004.07.005 min, h., zhou, g. (2002). supply chain modeling: past, present and future.  computers & industrial engineering,  43(1-2), 231-249. https://doi.org/10.1016/s0360-8352(02)00066-9 mula, j., poler, r., garcía-sabater, j.p., lario, f.c. (2006). models for production planning under uncertainty: a review. international journal of production economics, 103(1), 271-285. https://doi.org/10.1016/j.ijpe.2005.09.001 pasenchenko, yu.a, trubnikova, o.i. (2011). modeling of commercial enterprise reserves management in the system of balanced indicators. investments: practice and experience, 15, 38-41. perea, e., grossmann, i., ydstie, e., tahmassebi, t. (2000). dynamic modeling and classical control theory for supply chain management. computers & chemical engineering, 24(2-7), 1143-1149. https://doi.org/10.1016/s0098-1354(00)00495-6 popovic, t., kraslawski, a., barbosa-póvoa, a., carvalho, a. (2017). quantitative indicators for social sustainability assessment of society and product responsibility aspects in supply chains. journal of international studies, 10(4), 9-36. https://doi.org/10.14254/20718330.2017/10-4/1 robert, s. (1992). the balanced scorecard: measures that drive performance. harvard business review, 70(1), 71-79. seuring, s. (2013). a review of modeling approaches for sustainable supply chain management. decision support systems, 54(4), 15131520. https://doi.org/10.1016/j.dss.2012.05.053 shraybfeder, d. (2005). efficient inventory control. alpina business books, moscow, russia. shygun m.m. (2008). economic modeling in dissertation research: bibliometric analysis. visnyk zhdtu, 1(43), 115-127. swaminathan, j.m., smith, s.f., sadeh, n.m. (1998). modeling supply chain dynamics: a multiagent approach. decision sciences, 29(3), 607-632. https://doi.org/10.1111/j.1540-5915.1998.tb01356.x takha, k.a. (2005). introduction to operations research. williams, moscow, russia. int. j. prod. manag. eng. (2019) 7(2), 117-124 creative commons attribution-noncommercial-noderivatives 4.0 international luchko et al. 124 https://doi.org/10.1111/j.2517-6161.1955.tb00191.x https://doi.org/10.1016/s0925-5273(98)00079-6 https://doi.org/10.1016/s0377-2217(02)00806-8 https://doi.org/10.1016/s0377-2217(96)00395-5 https://doi.org/10.1016/j.ijpe.2009.09.005 https://doi.org/10.14254/2071-8330.2017/10-1/17 https://doi.org/10.1016/j.cor.2004.07.005 https://doi.org/10.1016/s0360-8352(02)00066-9 https://doi.org/10.1016/j.ijpe.2005.09.001 https://doi.org/10.1016/s0098-1354(00)00495-6 https://doi.org/10.14254/2071-8330.2017/10-4/1 https://doi.org/10.14254/2071-8330.2017/10-4/1 https://doi.org/10.1016/j.dss.2012.05.053 https://doi.org/10.1111/j.1540-5915.1998.tb01356.x http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.12737 received: 2019-11-18 accepted: 2020-04-09 demand driven mrp – the need to standardise an implementation process orue, a. *a1, lizarralde, a. a2, kortabarria, a. a3 adpto. de organización industrial. escuela politécnica superior de mondragon unibertsitatea. c/ loramendi 4, 20500 mondragón (spain). a1* aorue@mondragon.edu, a2 alizarralde@mondragon.edu, a3 akortabarriai@mondragon.edu abstract: since the creation of the demand-driven material requirement planning (ddmrp) model, numerous studies have analysed the methodology’s significant impact on different organisations. several successful cases and research studies into ddmrp have demonstrated that the methodology is beneficial to organisations because it increases their service level and stock adjustments; however, there is a dearth of literature regarding the steps necessary to implement this model successfully. this document delivers a systematic review of the literature based on the work done by kitchenham (2004) with the aim of analysing studies that investigate the standardization of the process of implementing the model. once the lack of research has been demonstrated, a possible line of future research can be outlined to standardise the implementation process of the ddmrp model to achieve its full potential. key words: demand driven mrp, ddmrp, process standardisation. 1. introduction market competition has caused an evolution in industrial operations with continual growth in the number of catalogue references, as well as reduced serial sizes and product life on the market (figure 1) (de la calle et al., 2017; stevenson et al., 2005). these circumstances mean that managing industrial operations requires a point of view different from that of the client in terms of what creates an excellent company (gupta & boyd, 2008). within this paradigm, a variety of methodologies have emerged to respond to this problem, including the theory of constraints (toc), lean manufacturing, quickresponse-manufacturing (qrm) and demand-driven material requirement planning (ddmrp). the mentioned methodologies describe the steps to follow during the implementation phases without detailing the specific instructions the specific instructions are decided on a case-by-case basis according to the knowledge of the planner, which may lead to problems in implementing the methodology correctly (pretorius, 2014). to cite this article: orue, a., lizarralde, a., kortabarria, a. (2020). demand driven mrp – the need to standardise an implementation process. international journal of production management and engineering, 8(2), 65-73. https://doi.org/10.4995/ijpme.2020.12737 figure 1. evolution of production systems (koren, 2010). int. j. prod. manag. eng. (2020) 8(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international 65 https://orcid.org/0000-0002-3824-8816 https://orcid.org/0000-0002-1390-9817?lang=en https://orcid.org/0000-0001-5298-0942 mailto:aorue@mondragon.edu http://creativecommons.org/licenses/by-nc-nd/4.0/ many aspects of the ddmrp methodology are subjective and depend on the judgment of the planner (lee & rim, 2019). these aspects include deciding where the buffer should be strategically positioned and choosing the percentage and variability of lead time, the type of buffer profile and the frequency of dynamic buffer readjustments (velasco acosta et al., 2019). in addition, the red zone buffer is calculated based on the subjectivity of the planner implementing the methodology because the planner can choose values based on industry experience (lee & rim, 2019). to ensure that this subjectivity exists, all ddmrp implementations carried out by the mondragon unibertsitatea team of researchers have been reviewed and analysed for comparative purposes. the analysis focused on discovering significant differences in executing the phases of the ddmrp model. the analysis and comparison were carried out through semi-structured interviews with four researchers from mondragon unibertsitatea. in these interviews, they were asked about the implementation phases of the ddmrp methodology, and their responses have been introduced into the specific instructions they follow to define the different parameters within the implementation phases. for example, each of the references defines the policy to be followed differently. depending on the crossed abc methodology and how the policies of each reference are decided, the decouple points, the batch, the buffer profile and the rotation objectives can be different and are different in this particular case. it can be concluded that each of the researchers selected the ‘how’ based on their own experience and expertise. this conclusion allows for specific differences at the time of ddmrp’s implementation based on the person’s implicit subjectivity when making decisions. to avoid these issues of subjectivity, a standard implementation process provides multiple benefits for organisations. for example, ramakumar (2004) has demonstrated that standardising a business process can be profitable for an organisation. swaminathan (2001) further indicated that process standardisation delivers tremendous benefits to organisations. fomin & lyytinen (2000) discussed a successful case study of a standardised process, providing a list of advantages of standardisation for companies and clients. münstermann and weitzel (2008) presented a bibliographic review of process standardisation enumerating several benefits, including these more remarkable benefits: reduction in implementation time, lower implementation costs, fewer possibilities for error, and improved quality of the process. the current paper presents a systematic literature review based on the work done by kitchenham (2004) with the aim of analysing studies that investigate the standardization of the implementation process of the ddmrp model. in case there are no studies, possible future lines of research will be opened in order to investigate the implementation process of the ddmrp. this standardisation will ensure that the potential of the methodology is fully exploited. 2. literature review this article begins by introducing basic ddmrp concepts, beginning with the reason the ddmrp methodology was developed and the problems it solves. the five phases that must be followed to implement the methodology are then explained. subsequently, a literature review is presented to define the words ‘standard’ and ‘process’. once the necessary terminology is defined, the research objectives and the methodology to be used are illustrated. we then present the results of the research before outlining our conclusions and recommending future lines of research. 2.1. demand driven material requirement planning traditional production planning and control systems, as well as material requirement planning (mrp), just in time (jit) and toc, lack the functionality to respond to new scenarios (ptak & smith, 2016). the traditional mrp push approach has several shortcomings in environments with changing or unpredictable demands. tools based on the pull philosophy, such as jit and toc, also have int. j. prod. manag. eng. (2020) 8(2), 65-73 creative commons attribution-noncommercial-noderivatives 4.0 international orue et al. 66 http://creativecommons.org/licenses/by-nc-nd/4.0/ inadequacies in implementing a demand-driven strategy due to their lack of planning and inventory control tools (ptak & smith, 2016). when a company uses a traditional production planning and control system, their inventory level has a bimodal distribution that alternates from too high to too low. this change in distribution results in a high-cost inventory level and a low service level (figure 2) (ptak & smith, 2016). figure 2. bimodal inventory distribution (ptak & smith, 2016, p. 11). to respond to this problem, ptak and smith (2016) introduced a new methodology known as ddmrp. ddmrp is based on mrp, jit and toc, and it incorporates new concepts for managing inventory. with the ddmrp, companies are better positioned to respond to variability in demand by adjusting inventory levels while maintaining and even increasing their service level. the ddmrp is composed of five steps (figure 3). the first three steps determinate the initial configuration and evolution of the ddmrp. steps four and five define the operational aspects of the methodology, planning and execution. figure 3. the five phases of ddmrp (ptak & smith, 2016, p. 53). phase 1: strategic positioning of the inventory. including inventory in all parts of the supply chain to meet the changing demand of the market is a waste of an organisation’s resources. on the other hand, eliminating the inventory completely endangers the supply chain and, therefore, the organisation (ptak & smith, 2016). phase 2: profile types and buffer levels. the second step of the methodology is to define the quantity of protection at the decoupling points. maintaining too many inventory levels requires an excess of money invested, materials and capacity and additional space to store this inventory, in addition to a risk of obsolescence of the inventory. on the other hand, having too little inventory can lead to lost sales and expensive urgent orders (ptak & smith, 2016). the ddmrp methodology has three buffer zones, each of which has a specific function, as can be seen in figure 4. to be able to size each of the zones, factors like minimum order quantity, average daily consumption and decoupled maturation period are used (ptak & smith, 2016). figure 4. buffer zones (ptak & smith, 2016, p. 98). phase 3: dynamic adjustments. companies and their supply chains must be prepared to adapt to ever-changing markets to offer the best customer service. this requires the use of dynamic buffers to adapt to the new requirements. to achieve this goal, the ddmrp methodology provides dynamic adjustments based on operational parameters, changes in the market and future planned or known events (ptak & smith, 2016). phase 4: demand-driven planning. this is the part that generates the supply orders (purchase orders, production orders and transfer orders). the ddmrp methodology uses a net flow equation for buffer replacement that generates the supply order recommendation signal in terms of time and quantity. in addition, this equation gives the net flow position of each buffer, which is calculated daily at all decoupled points (ptak & smith, 2016). int. j. prod. manag. eng. (2020) 8(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international demand driven mrp – the need to standardise an implementation process 67 http://creativecommons.org/licenses/by-nc-nd/4.0/ when the net flow position value is entered in the refuelling zone, the ddmrp generates and recommends a supply order. in terms of the colour code, the value is represented in the yellow buffer zone with a supply order amount reaching the top of the green dimension. phase 5: visible and collaborative execution. the ddmrp methodology distinguishes between planning and execution. the planning stage includes generating supply order requirements using the net flow position, and it ends when the recommendations are approved and become open ministerial orders. the execution stage includes the management of these open supply orders to protect and to promote the flow of inventory. ddmrp incorporates different colour-coded alerts to provide visibility and to prioritise orders. the alerts draw attention to critical situations that require attention. in this way, the company can prioritise orders correctly according to the state of the available buffer rather than relying solely on the delivery date (ptak & smith, 2016). 2.2. standard processes to establish standard processes, one must first define a ‘standard’. jang and lee (1998) define standardisation as the degree to which work rules, policies and operating procedures are formalised and followed. the international organization for standardization (iso) guide states that standards are documents established by consensus and approved by a recognised organisation. these standards provide rules, guidelines or characteristics for activities or their results to achieve the optimum degree of order in any given context (iso, 2005). defining ‘process’ involves multiple concepts. the european foundation for quality management model defines a process as a sequence of activities that add value while producing a specific product or service based on certain contributions. the international organization for standardization iso 9000 defines a process as a set of interrelated or interactive activities that transform inputs into outputs (iso, 2005). an example of the standardisation of an implementation process is the work carried out by lizarralde et al. (2020). they created a systematic implementation process of the first two steps of toc to enhance operative performance in drum-bufferrope (dbr) implementation (lizarralde et al., 2020). 3. research design and methodology 3.1. research objective the objective of this research was to determine whether any researcher or planner has systematised the implementation process of ddmrp methodology, thus ensuring its full potential is achieved. to do so, a systematic review of the literature was conducted, which demonstrated a lack of research in this area. subequently, a possible line of future research could be outlined to standardise the implementation process of the ddmrp model to achieve its full potential. 3.2. methodology to deepen the implementation process of the ddmrp methodology, a systematic review of the literature has been carried out. a systematic literature review is a means of identifying, assessing and interpreting all available research that is relevant to a particular research question, thematic area or phenomenon of interest (kitchenham, 2004). in order to carry out a systematic review properly, it is necessary to define a research strategy (kitchenham, 2004). this study uses a strategy based on kitchenham’s (2004) and is detailed below (figure 5): figure 5. systematic literature review methodology based on kitchenham (2004) 3.2.1. planning the review this research was motivated by the possibility that the ddmrp methodology was not used to its full potential in its implementations by researchers at int. j. prod. manag. eng. (2020) 8(2), 65-73 creative commons attribution-noncommercial-noderivatives 4.0 international orue et al. 68 http://creativecommons.org/licenses/by-nc-nd/4.0/ mondragon unibertsitatea. the aim of this literature review is to analyse the existing research in the field of the ddmrp methodology and, more specifically, in the implementation process to determine the research gaps in the field. when planning the literature review, it is necessary to define a protocol that specifies the methods used to conduct a specific systematic review. a defined protocol is needed to reduce the effect of researchers’ bias (kitchenham, 2004). the defined protocol of this study includes the following components: keywords to carry out the literature review: the articles focus exclusively on the ddmrp methodology, so the following words were entered in the search engine: ‘demand-driven mrp’ or ‘ddmrp’. this choice of keywords was meant to guarantee that ddmrp would be the main theme of the article. to refine the search and to focus on approaches related to the systematisation or standardisation of the model’s implementation process, the following keywords were added: ‘process’, ‘standard’, ‘systematic’, ‘implementation’ and ‘benefit’. the sources to identify primary studies: the databases chosen for the research were emerald and sciencedirect, which include research into operations, organisational management and social sciences. scopus and web of science (wos) were also used to guarantee investigation of the entire field, as these are two of the largest available databases of citations and abstracts from peer-reviewed literature and include the main publishers of indexed operations and administration (for example, emerald, elsevier, and springer). american production and inventory control society (apics) operations planning articles were also considered. this is because apics is one of the most important global associations involved in operations management and has strong ties to the most important companies in the world. select the exclusion and inclusion criteria for the studies: the criteria used to select and evaluate the articles included: (1) exclusive focus on the ddmrp methodology, (2) inclusion of no other methodology, (3) publication in an academic journal or conference, (4) not being written for a terminal degree or master’s degree, (5) the chosen articles included some case studies, both simulated and real period of publication: the delimited publication period was from 2011 -2019 to produce a detailed outline of the ddmrp model since its creation (the year of the publication of the first book). study quality assessment: both quantitative and qualitative documents were considered for this research. the criteria for evaluating the quality of the selected journals were established by using indicators such as, journal citation report (jcr) and scimago journal rank (sjr). 3.2.2. conduct the review in order to find as many primary studies related to the ddmrp methodology as possible, an unbiased search strategy was defined. there are not many studies concerning the ddmrp methodology, so the first decision was to perform a search with the terms ddmrp and demand-driven mrp. the remaining keywords were then added. the bibliographic references of the resulting articles were also taken into account to find more studies. table 1 provides an example of the results returned. table 1. example of the results returned. keyword wos results scopus results ddmrp 12 16 demand driven mrp 31 30 demand driven mrp process 8 14 ddmrp process 3 3 ddmrp standard 0 1 demand driven mrp standard 0 0 all articles related to the defined search were collected, duplicates were eliminated and the first stage of content control was performed by reading titles and abstracts. articles that did not meet the requirements were excluded. following this process, 16 articles remained, and these were carefully reviewed. the frequency of works on this topic over the years is illustrated in graphic 1. int. j. prod. manag. eng. (2020) 8(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international demand driven mrp – the need to standardise an implementation process 69 http://creativecommons.org/licenses/by-nc-nd/4.0/ graphic 1. ddmrp papers frequency 2011–2019. 4. field work results of the 16 documents selected, three articles (favaretto & marin, 2018; smith & smith, 2013; trojan et al., 2019) seek to demonstrate the need for new production controls and planning systems meeting the needs of today’s changing paradigm. six of the articles discuss the quantitative benefits of the ddmrp methodology compared to traditional models, such as mrp or mrpii (ihme & stratton, 2015; miclo et al., 2016, 2019; shofa et al., 2018; table 2. the variables defined to measure the impacts of the results. research area representative articles results new innovative methodology of ddmrp (smith & smith, 2013) the authors explain the ddmrp model and discuss how to move from a push model to a pull model that positions the inventory. (favaretto & marin, 2018) the authors explain the different production planning and control models of the last 100 years. they also explain the context in which the ddmrp model was created as well as its fundamental characteristics. (trojan et al., 2019) the objectives of this publication are to extend knowledge of demand-driven supply logistics using the ddmrp methodology in the specific context of industry 4.0 and to verify this processed theoretical knowledge through a case study. ddmrp benefits, theoretical or simulated (ihme & stratton, 2015) the authors evaluate the potential benefits of the ddmrp model using simulated data from a company that produces printing inks. the results of the simulation across 28 sample products showed how the aggregation and formalised signalling system reduced high and low inventory alerts by 45% and stockouts by 95%. (miclo, fontanili, lauras, lamothe, & milian, 2016) this article compares the mrpii model with the ddmrp model through a case study using the discrete event simulation approach. ddmrp appears to outperform mrp ii in all situations because it allows the same level of on time delivery with less work capital (10% less in general) and less anxiety. (shofa & widyarto, 2017) this article evaluates and compares the mrp model and the ddmrp model in terms of the systems’ inventory levels. the evaluation is based on a simulation using data from an indonesian automotive company. ddmrp reduces the lead-time from fifty-two to three days (94% reduced) and shifts the inventory level for the three parts to the effective stock. because of this, ddmrp is more effective than mrp. (shofa et al., 2018) through a simulation of discrete events, the authors compare the mrp model with the ddmrp model in cases of uncertain demand and long maturation periods. ddmrp improves the inventory level from 106,852 pieces per day to 95,284 pieces per day (11% reduction) and makes inventory stable. because of this, ddmrp is more effective at production planning than mrp. (table 2, continue in the next page) int. j. prod. manag. eng. (2020) 8(2), 65-73 creative commons attribution-noncommercial-noderivatives 4.0 international orue et al. 70 http://creativecommons.org/licenses/by-nc-nd/4.0/ shofa & widyarto, 2017; velasco acosta et al., 2019). to report these advantages, those authors use theoretical calculations and simulations of discrete events as well as real data from various companies. related to the implementation process, another article proposes a dynamic adjustment of the decoupled lead time, taking into account lead time variability (dessevre et al., 2019). two of the studies analyse the changes implemented and the qualitative and quantitative results obtained by several companies following their conversion from traditional or classical models to the ddmrp model (bahu et al., 2019; kortabarria et al., 2018). an additional three articles (jiang & rim, 2017, 2016; lee & rim, 2019) introduce mathematical models to define the positioning of inventory, depending on the circumstances of different organisations. related to the implementation process, another article proposes a dynamic adjustment of the decoupled lead time, taking into account lead time variability (dessevre research area representative articles results ddmrp benefits, theoretical or simulated (velasco acosta et al., 2019) this article evaluates the applicability of ddmrp in a complex manufacturing situation in terms of customer satisfaction and stock levels. the evaluation is based on a simulation of the ddmrp model using discrete event software. the results were a 41% reduction in lead-time and an 18% reduction in stock levels. (romain miclo et al., 2019) the authors explain and explore the ddmrp model. they also evaluate the model’s effectiveness compared with two other methodologies (mmrp ii and kanban/ lean production) using a series of structured computer simulation experiments. the results indicate that ddmrp does represent a superior approach. benefits of ddmrp in a real case study (kortabarria et al., 2018) this article analyses the quantitative and qualitative results of an industrial company’s shift from mrp to ddmrp. the results strongly show that, by using ddmrp, the company increased visibility in the supply chain. ddmrp also reduced the inventory level (52.53% reduction), while material consumption was increased (8.7%). all these results were achieved while maintaining a high level of service. (bahu et al., 2019) this article describes the operation and limits of the ddmrp model. through a study of 30 real cases, it also discusses the reasons that companies using a push model should implement the aforementioned methodology. implementation process for ddmrp methodology (jiang & rim, 2016) the authors created a mathematical model to position and to quantify the work in progress. the paper provides a systematic solution process to determine the best location of the buffer in the make-to-order manufacturing process to minimise the total inventory cost. (jiang & rim, 2017) the authors created a mathematical model to position the work in progress, the quantity of work for on-demand orders and random processing times. the paper addresses the problem of defining the stations to hold work-inprocess inventory to reduce the production lead-time. (lee & rim, 2019) this article proposes a new stock formula for safety stock, which comes from the ddmrp replenishment guidelines. the defined safety stock formula eliminates subjectivity when calculating the safety stock of ddmrp. (dessevre et al., 2019) this article defines a dynamic adjustment of the decoupled lead time, taking into account lead time variability. the results show that the dynamic adjustment of buffer sizes reduces stock while ensuring a good quality of service. definition of a process map of demand-driven adaptative enterprise (martin et al., 2018) this article discusses how the ddmrp model has evolved toward the demand-driven adaptive enterprise (ddae). this article proposes a cartography of the processes of adaptive companies driven by demand. (table 2, continue from the previous page) int. j. prod. manag. eng. (2020) 8(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international demand driven mrp – the need to standardise an implementation process 71 http://creativecommons.org/licenses/by-nc-nd/4.0/ et al., 2019). the final article (martin et al., 2018) describes the evolution of the ddmrp model toward the demand-driven adaptive enterprise (ddae). this model involves a complete set of business rules, from the strategic level to the execution level. in summary, the ddmrp methodology is attracting interest in the scientific field, specifically in the industrial operations area. the number of articles on this topic has increased considerably over the years (20111-2019). however, this literature did not find any studies that have investigated the standardization of the process of implementing the ddmrp model. 5. conclusions and future research the ddmrp methodology represents a significant advance in production planning and control systems that is capable of responding to the needs of the new paradigm. though it offers multiple benefits for organisations, the steps required to implement this promising methodology remain unclear. along with this literature review, we analysed the ddmrp implementations carried out by a team of mondragon unibertsitatea researchers. in all these cases, positive results were obtained in terms of increasing the visibility and the flow of materials; however, significant differences existed in implementing the methodology. after carrying out the literature review, we have found no evidence of a standardised implementation process for ddmrp that could maximise its potential. therefore, to improve the ddmrp methodology, we invite other authors to continue researching and defining a standardised implementation process. references bahu, b., bironneau, l., hovelaque, v. (2019). compréhension du ddmrp et de son adoption: premiers éléments empiriques. logistique & management, 1–13. https://doi.org/10.1080/12507970.2018.1547130 de la calle, a., grus, m.e., álvarez, e. (2017). value creation through demand and supply chains: evidences from spanish companies. direccion y organizacion, 61, 4–11. dessevre, g., martin, g., pellerin, r., lamothe, j., lauras, m. (2019). decoupled lead time in finite capacity flowshop: a feedback loop approach. 2019 international conference on industrial engineering and systems management (iesm), 1–6. https://doi.org/10.1109/iesm45758.2019.8948198 favaretto, d., marin, a. (2018). an empirical comparison study between ddmrp and mrp in material management. ssrn electronic journal. https://doi.org/10.2139/ssrn.3305114 fomin, v., lyytinen, k. (2000). how to distribute a cake before cutting it into pieces: alice in wonderland or radio engineers' gang in the nordic countries?. in jakobs, k. (ed.), information technology standards and standardization: a global perspective (pp. 222-239). igi global. https://doi.org/10.4018/978-1-878289-70-4.ch014 gupta, m.c., boyd, l.h. (2008). theory of constraints: a theory for operations management. international journal of operations & production management, 28(10), 991–1012. https://doi.org/10.1108/01443570810903122 ihme, m., stratton, r. (2015). evaluating demand driven mrp: a case based simulated study. in international conference of the european operations management association (1–10). retrieved from http://irep.ntu.ac.uk/id/eprint/26668/ iso, i. s. o. (2005). 9000-quality management systems—fundamentals and vocabulary. quality management systems, berlin: beuth verlag gmbh, 22–25. jang, y., lee, j. (1998). factors influencing the success of management consulting projects. international journal of project management, 16(2), 67–72. https://doi.org/10.1016/s0263-7863(97)00005-7 jiang, j., rim, s.c.c. (2017). strategic wip inventory positioning for make-to-order production with stochastic processing times. mathematical problems in engineering, 2017, 1–7. https://doi.org/10.1155/2017/8635979 jiang, j., rim, s. c. (2016). strategic inventory positioning in bom with multiple parents using asr lead time. mathematical problems in engineering, 2016, 1–9. https://doi.org/10.1155/2016/9328371 kitchenham, b. (2004). procedures for performing systematic reviews. keele, uk, keele university. koren, y. (2010). the global manufacturing revolution: product-process-business integration and reconfigurable systems (vol. 80). john wiley & sons. https://doi.org/10.1002/9780470618813 kortabarria, a., apaolaza, u., lizarralde, a., amorrortu, i. (2018). material management without forecasting: from mrp to demand driven mrp. journal of industrial engineering and management, 11(4), 632. https://doi.org/10.3926/jiem.2654 lee, c.j., rim, s.c. (2019). a mathematical safety stock model for ddmrp inventory replenishment. mathematical problems in engineering, 2019, 1–10. https://doi.org/10.1155/2019/6496309 int. j. prod. manag. eng. (2020) 8(2), 65-73 creative commons attribution-noncommercial-noderivatives 4.0 international orue et al. 72 https://doi.org/10.1080/12507970.2018.1547130 https://doi.org/10.1109/iesm45758.2019.8948198 https://doi.org/10.2139/ssrn.3305114 https://doi.org/10.4018/978-1-878289-70-4.ch014 https://doi.org/10.1108/01443570810903122 http://irep.ntu.ac.uk/id/eprint/26668 https://doi.org/10.1016/s0263-7863(97)00005-7 https://doi.org/10.1155/2017/8635979 https://doi.org/10.1155/2016/9328371 https://doi.org/10.1002/9780470618813 https://doi.org/10.3926/jiem.2654 https://doi.org/10.1155/2019/6496309 http://creativecommons.org/licenses/by-nc-nd/4.0/ lizarralde, a., apaolaza, u., mediavilla, m. (2020). a strategic approach for bottleneck identification in make-to-order environments: a drum-buffer-rope action research based case study. journal of industrial engineering and management, 13(1), 18–37. https://doi.org/10.3926/jiem.2868 martin, g., baptiste, p., lamothe, j., miclo, r., lauras, m., & albi, m. (2018). a process map for the demand driven adaptive enterprise model: towards an explicit cartography 3. the cartography: how to represent the complete demand driven adaptive. in 7 th international conference on information systems, logistics and supply chain (664–672). retrieved from https://hal-mines-albi.archives-ouvertes.fr/hal-01883504/ miclo, r., fontanili, f., lauras, m., lamothe, j., milian, b. (2016a). an empirical comparison of mrpii and demand-driven mrp. ifacpapersonline, 49(12), 1725–1730. https://doi.org/10.1016/j.ifacol.2016.07.831 miclo, r., fontanili, f., lauras, m., lamothe, j., milian, b. (2016b). mrp vs. demand-driven mrp: towards an objective comparison. proceedings of 2015 international conference on industrial engineering and systems management, ieee iesm 2015, (october), 1072– 1080. https://doi.org/10.1109/iesm.2015.7380288 miclo, r., lauras, m., fontanili, f., lamothe, j., melnyk, s.a. (2019). demand driven mrp: assessment of a new approach to materials management. international journal of production research, 57(1), 166–181. https://doi.org/10.1080/00207543.2018.1464230 münstermann, b., weitzel, t. (2008). what is process standardization? (vol. 64). retrieved from http://aisel.aisnet.org/confirm2008/64 pretorius, p. (2014). introducing in-between decision points to toc ’ s five focusing steps. international journal of production research, (november), 37–41. https://doi.org/10.1080/00207543.2013.836612 ptak, c.a., smith, c. (2016). demand driven material requirements planning (ddmrp). industrial press, incorporated. ramakumar, a., quality, b.c. (2004), u. (n.d.). process standardization proves profitable: the demand for higher product quality has resulted in an increasingly complex value chain. bnp media. retrieved from https://scholar.google.es/scholar?hl=es&as_sdt=0%2c5&q=ramakumar+y+cooper+2004&btng= shofa, m.j., moeis, a.o., restiana, n. (2018). effective production planning for purchased part under long lead time and uncertain demand: mrp vs demand-driven mrp. in iop conference series: materials science and engineering (vol. 337, p. 012055). iop publishing. https://doi.org/10.1088/1757-899x/337/1/012055 shofa, m.j., widyarto, w.o. (2017). effective production control in an automotive industry: mrp vs. demand-driven mrp. in aip conference proceedings (vol. 1855, p. 020004). aip publishing llc. https://doi.org/10.1097/01.ju.0000143904.17666.0b smith, b.d., smith, c. (2013, november 1). becoming demand-driven: how to change from push and promote to position and pull. strategic finance, 95(november), 37–45. retrieved from https://go.galegroup.com/ps/i.do?p=aone&sw=w&u=googlescholar&v=2.1&it=r&id =gale%7ca349741171&sid=googlescholar&asid=052f118f stevenson, m., hendry, l.c., kingsman, b.g. (2005). a review of production planning and control: the applicability of key concepts to the maketo-order industry. international journal of production research, 43(5), 869–898. https://doi.org/10.1080/0020754042000298520 swaminathan, j.m. (2001). enabling customization using standardized operations. california management review, 43(3), 125–135. https://doi.org/10.2307/41166092 trojan, j., pekarčíková, m., kliment, m., trebuňa, p. (2019). demand driven material requirements planning. some methodical and practical comments. management and production engineering review, 10(2), 50–59. https://doi.org/10.24425/mper.2019.129568 velasco acosta, a.p., mascle, c., baptiste, p. (2019). applicability of demand-driven mrp in a complex manufacturing environment. international journal of production research, 0(0), 1–13. https://doi.org/10.1080/00207543.2019.1650978 int. j. prod. manag. eng. (2020) 8(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international demand driven mrp – the need to standardise an implementation process 73 https://doi.org/10.3926/jiem.2868 https://hal-mines-albi.archives-ouvertes.fr/hal-01883504 https://doi.org/10.1016/j.ifacol.2016.07.831 https://doi.org/10.1109/iesm.2015.7380288 https://doi.org/10.1080/00207543.2018.1464230 http://aisel.aisnet.org/confirm2008/64 https://doi.org/10.1080/00207543.2013.836612 https://scholar.google.es/scholar?hl=es&as_sdt=0%2c5&q=ramakumar+y+cooper+2004&btng= https://doi.org/10.1088/1757-899x/337/1/012055 https://doi.org/10.1097/01.ju.0000143904.17666.0b https://go.galegroup.com/ps/i.do?p=aone&sw=w&u=googlescholar&v=2.1&it=r&id=gale%7ca349741171&sid=goo https://go.galegroup.com/ps/i.do?p=aone&sw=w&u=googlescholar&v=2.1&it=r&id=gale%7ca349741171&sid=goo https://doi.org/10.1080/0020754042000298520 https://doi.org/10.2307/41166092 https://doi.org/10.24425/mper.2019.129568 https://doi.org/10.1080/00207543.2019.1650978 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering special issue: advances in engineering networks https://doi.org/10.4995/ijpme.2019.10807 received: 2018-10-11 accepted: 2019-05-30 testing successful business model using system dynamics ganzarain j. a1, ruiz, m. a2, igartua, j.i. a3 a mechanical and manufacturing department, mondragon university, loramendi 420500 mondragón, spain. a1 jganzarain@mondragon.edu, a2 mruiz@mondragon.edu, a3 jigartua@mondragon.edu abstract: in our increasingly globalised economy, managing continuous change whilst remaining competitive and dynamic has become a central issue for firms in the industrial sector. one of the elements for obtaining this competitiveness is the value creation model of the firm. the most important challenges in firms are characterised by dynamic complexity which makes it difficult to understand factors in their context. consequently management and decision making is hindered (antunes et al., 2011). business models are characterised by complexity and dynamism. performance of the firm is a complex topic determined by the large amount of variables that can be involved in the system, and the different effects that influence the system in the short and long term. due to this complexity a systemic view is required, that is, an holistic view of the whole system. such a systemic view enables managers to make decisions based on evidence rather than intuition and personal experiences, as they understand how the whole system works. thus, the main aim of this research is to use an empirical tool such as system dynamics (sd), to support and sustain firms in the identification of new constructs related to their business model (bm). key words: business model, business model design, firm performance, system dynamics, decision making. 1. introduction in this paper we establish and build a system dynamics (sd) model which enables companies to take most effectively their value creation model. in addition, this facilitates the analysis of the consequences of their decisions and their impact on competitiveness. the aim is to support and sustain companies in the identification of new constructs related to the design of the business model (bm) and how it is measured is integrately linked to their long-term financial performance. the bm theoretical construct can be represented as an analysis unit of the creation of value for the sustainable competitiveness of the company. the bm is a new way of making coherent strategy business based on different mechanisms of economic, social and sustainable relationship between companies, suppliers, partners and customers. in this paper the novelty and efficiency have been selected as the main specific components composing the business model design (bmd) (miller, 1996) and the main goal is to analyse how these two components are affecting on the performance of the firm. this model is the tool used to guide and help companies to take decisions about their current bmd. the objective of the article was to establish an sd simulation model as a useful tool to explain decision makers how they can redesign their business model taking into account the theory originally proposed by amit and zott (2001). this technique enables the understanding and dynamic representation of each of the variables that compose the whole business model system. to cite this article: ganzarain j., ruiz, m., and igartua, j.i. (2019). testing successful business model using system dynamics. international journal of production management and engineering, 7(special issue), 91-100. https://doi.org/10.4995/ijpme.2019.10807 int. j. prod. manag. eng. (2019) 7(special issue), 91-100creative commons attribution-noncommercial-noderivatives 4.0 international 91 https://orcid.org/0000-0002-6144-636x https://orcid.org/0000-0002-4209-1487 https://orcid.org/0000-0002-7039-4629 mailto:jganzarain@mondragon.edu mailto:mruiz@mondragon.edu mailto:jigartua@mondragon.edu http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. success of business models: business model design and the performance of the firm according to the literature, a successful bm takes place when two aspects are met: 1) there is a gap between the necessity and existing offers in the market and 2) the company has the resources necessary to fill this goal. bm pioneers understand what customers want and have the ability to fulfill them (teece, 2010). in summary, in this investigation we analyze the success of the bm and the two main components, the novelty and efficiency as mentioned before. 2.1. business model design according to zott and amit (2007), the design of a company’s business model, which focuses on the issues of novelty and efficiency, is associated with the performance of companies. with respect to business model design (bmd) issues, there are different approaches to the components that make up the bmd (zott and amit, 2007). miller (1996) highlights innovation and efficiency as the main themes due to their influence on the final performance of the company. their decision is accepted for the study of the bm adopted by companies because innovation and efficiency show multiple solutions to create new value under uncertainty. efficiency and innovation are not totally independent and exclusive, any bmd can focus on novelty and focus on efficiency at the same time. according to amit and zott (2001), there are four different categories which represent main factors that can increase the total value created by firms: novelty, efficiency, lock-in, and complementarities. as we mentioned earlier, and referring to miller (1996) in an attempt to understand the novelty and efficiency of the bm and how these two components influence the performance of the company, we will focus on these specific components of bmd, figure 1. in this article we identify two critical dimensions of bmd, which are called “efficiency-centered” and “novelty-centered” design themes. the origin is on the theory of innovation which is based on the transaction cost perspective (milgrom, 1992) and in schurnpeter’s (1934). we analyse the impact of novelty and efficiency bmd themes on the performance of the firm. efficiency-centered bmd is focused on reducing transaction costs for all transaction participants, while novelty-centered bmd refers to new ways of conducting economic exchanges among several stakeholders (zott and amit, 2007). 2.2. business model outcomes and firm’s performance literature regarding the bm and its effect in economic performance is based on both real economic performance and perceived economic performance. the real economic performance show quantitative evidence of the impact of bmi on firm performance (zott and amit, 2007, demil and lecocq, 2010, nair et al., 2013, giesen et al., 2007, weill et al., 2005), while the perceived economic performance takes into account managers’ perception of bm impacts on the economic performance of companies. according to zott and amit (2007) the business models of multiple innovative firms were classified based on whether they are efficiency-centered or novelty-centered, concluding that the design of a business model focused on novelty has a great impact on the financial performance. the research points out that the options of the companies to support innovation paths have repercussions on the success of the designs of the business models. according to zott and amit these two categories represent variables that can enhance the total value created by firms (zott and amit, 2007). in particular they refer the following specific kinds of business solutions: value novelty lock-in complementarities efficiency new transaction structures new transactional content new participants, etc search costs selection range simplicity speed scale economies, etc between products and services for customers between on-line and off-line assets betweeen technologies between activities switching costs • loyalty programs • dominant design • trust • customization positive network externalities figure 1. sources of value creation in firms (amit and zott, 2001). int. j. prod. manag. eng. (2019) 7(special issue), 91-100 creative commons attribution-noncommercial-noderivatives 4.0 international ganzarain et al. 92 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2.2.1. novelty the basis of novelty-centered bmd is identification and definition of new ways of arranging economic exchanges, by connecting different independent parties, linking transaction participants in new ways, or designing new transaction mechanisms (zott and amit, 2007). the novelty could be measured taking into account different items like (zott and amit, 2007): i) the new business model offers new combinations of products, services and information, ii) the new business model brings together new participants. therefore, according to zott and amit (2007), we confirm that there is a positive effect of noveltycentered bmd on the performance of a company. 2.2.2. efficiency another solution to create value for companies is to imitate rather than innovate: creating different solutions to established firms, but in a more efficient way (aldrich, 1999, zott, 2003). to analyse the performance implications of efficiency-centered bm, we build on transaction cost perspective, which refers to the design of economic transactions (milgrom, 1992). efficiency-centered design refers to the key indicators that companies may identify to achieve transaction efficiency through their business models. the main pillar of a business model focused on efficiency is the reduction of transaction costs. there is an influencing direct relationship between the design of transactions and firm performance. so, based on the results of the study carried out by zott and amit (2007), a main positive effect is expected from the bmd focused on efficiency in the performance of a firm. the efficiency could be measured taking into account different items like (zott and amit, 2007): i) inventory costs for participants in the business model are reduced, ii) marketing and sales costs, transaction processing costs and communication costs. 2.2.3. firm performance the bmd can be an important factor in the performance of the firm. in addition, there is an increasingly interest in the field of how bm typology outperform others (weill et al., 2005). the competitiveness of the companies is directly affected by the design of the business model (andreini and bettinelli, 2017). thus, several authors confirm the relationship between business model and the competitive advantage of the firm. “the competitive advantage explains how the firm will do better than its rivals, and doing it better, by definition, means being different” (magretta, 2002). overall, studies focuses on measuring the impact of bm on firm’s performance and innovativeness: i) the design of bm, impacts on the level of sales exceeding the objectives., ii) the design of bm, impacts on market share and sales level objectives, iii) the design of bm, impacts on the profitability. therefore, considering the positive effect of noveltycentered bmd, it is confirmed a positive effect of novelty-centered bmd on the performance of a firm (zott and amit, 2007). consequently, there exist a positive main effect of efficiency-centered bmd on the performance of a firm. 3. system dynamics for business model decision making the most important occurrences, systems and challenges in firms are characterised by dynamic complexity (heterogeneous agents, behaviours and rules). dynamic complexity makes it difficult to understand these factors in their context, and consequently management and decision making is hindered (antunes & respício 2008; janssen et al., 2015). one of the factors which drives this complexity is the lack of consistency to define bms. how to define a consistant organizational strategy is a complex topic determined by the large amount of variables that could be involved in the system, and the different effects that influence the system in the short and long term. such complexity requires a systemic view, that is, a holistic view of the whole system. this systemic view enables the understanding of the interrelationships between the different constructs by which bms are composed. these interrelationships are not linear, they are circular, defined by feedback loops. this requires systems thinking to be understood. the short and long term perspective of bms helps to predict and define more consistent organisational strategies. both the qualitative and quantitative analysis of the variables (measurable and non measurable) are required for obtaining the most int. j. prod. manag. eng. (2019) 7(special issue), 91-100creative commons attribution-noncommercial-noderivatives 4.0 international testing successful business model using system dynamics 93 http://creativecommons.org/licenses/by-nc-nd/4.0/ accurate understanding of the complex issue of successful bm design. such a systemic view enables managers to make decisions based on evidence rather than intuitions and personal experiences, as they understand how the whole system works. decision making is one of the most vital processes to achieve the objectives of a firm, however many decisions are made without evidence based management. to compound the problem, future decision makers are often not able to take advantage of the experience of their predecessors (schalk et al., 2013). it is clear that the most effective firms of the future will be those which make decisions focused on evidence based predictions (degregorio, 1999). when decisions are made with no evidence, ineffective practices and experiences in the workplace are dominant. the increasingly competitive nature of the global economy has left many firms searching for new strategies to build capacity and sustainable competitive advantage. key to achieving this result is an effective decision making process. computational tools show great potential to assist decision makers, due to the speed and efficiency with which they are able to identify emergent behaviours (antunes & respício 2008; janssen et al., 2015). with the purpose of developing a more effective decision making processes, it is necessary to: (i) to conceptualise the system of causes and effects related to business model design in order to achieve a full understanding, and (ii) to model it using simulation to facilitate the interactive manipulation and generation of real scenarios. the final aim of (i) conceptualisation and, (ii) modelling is: to define a systemic and real perspective of the business model so as to facilitate a learning process for more effective decision making, and thus to fostering competitivity and improving firm sustainability. 4. methodology this research uses modelling to understand the impact of the success of the business model on firm performance using an sd model. firstly, the gap in the literature was defined (problem identification) to provide a clear focus for the subsequent phase. secondly, input collection was undertaken, variable identification and input collection was done. once this phase was completed, the simulation model was built, and finally policy testing was done. the process was iterative based on sterman (2002) which is standardized and the most commonly used methodology for modelling. figure 2. methodology for simulation (sterman, 2000). modelling is essentially creative, and is not a standardised process (morecroft, 2015). at the same time, however, it is a disciplined, scientific and rigorous procedure that involves observing dynamic phenomena in the real world, analysing hypotheses, collecting data and improving the model to obtain a better understanding of the issue of analysis. modelling is iterative, it begins with a concern about dynamics (performance over time) and preliminary ideas about feedback structure (morecroft, 2015). some examples of modelling processes are presented by morecroft (2015), sterman (2000), and warren (2002). the purpose is not to create a perfect model that replicates the real world situation in every detail. it is to engage in a learning process using the model as a tool for research, clarification, and discovery. the real value of modelling becomes evident when models are used to support organisational redesign. the final goal should be to design management policies and organisational structures that lead to greater success (sterman, 2000). the structure of modelling is based on two different points: i) hypotheses about the physical and institutional environment, and ii) hypotheses about the decision processes of the agents who act in those structures (sterman, 2000). a description of these hypotheses is set out below. the physical and institutional environment of a model includes the model boundary and stock and flow structures of people, material, money, information, and so forth that characterise the system. int. j. prod. manag. eng. (2019) 7(special issue), 91-100 creative commons attribution-noncommercial-noderivatives 4.0 international ganzarain et al. 94 http://creativecommons.org/licenses/by-nc-nd/4.0/ one of the examples that showed this environment in the literature, was presented by sterman (2000), who used forrester’s (1969) urban dynamics to understand why america’s large cities continued to decay despite massive amounts of aid and numerous renewal programs. factors describing the physical and institutional setting were included in the model, such as, size, quality of the housing stock, and attributes of population. decision processes refer to the decision rules that determine the behaviour of the agents in the system. these rules are represented through behavioural hypotheses. these hypotheses of a simulation model describe the way in which the system evolves over time. the most important value of simulation is to identify both observed behaviours and future possible circumstances (sterman, 2002). the behavioural hypotheses of a simulation model describe the way in which people respond to different situations. again, sterman (2000) used the urban dynamics model as an example, which included decision rules, governing migration and construction. essentially, the rule was to mark up the wholesale cost of the goods and the mark up was gradually reduced until the goods were sold. thus, it is not enough to model a particular decision. modellers must also detect and represent “the guiding policy” that yields the stream of decisions (forrester, 1961). each detail and characteristic in the model related to stocks and flows creates a decision point, and the modeller must specify accurately the decision rule determining the variable of analysis (sterman, 2000). 4.1. input collection the principal variables in our model are: i) business model design, the core of this research, ii) performancethe variable to be measured iii) noveltyone of the principal links between bmd and performance, and iv) efficiencya second connection between bmd and performance. in table 1 presents the different variables that compose the model and their functions. the variables are categorized in three different groups: i) literature about “success of business models”, ii) variables selected for measurement in this research and iii) variables which belong to an example of an sd to analyse the impact of business models (“impacto de los planes de negocio”). table 1. methodology for simulation (sterman, 2000). variable functions number of sales belongs to the construct “success of business models” perceived image belongs to the construct “success of business models” costs is used to define efficiency new products refers to the general launch of new products, services, and information new participants is connected to novelty group i efficiency refers to inventory, marketing and sales. key variable for performance novelty key variable for performance group ii competition (garcía, 2017) is based on the model “impacto de los planes de negocio” conjunction (garcía, 2017) represents the impact of the external factors group iii 4.2. conceptual model development the next step of the process was “conceptual model development”. in this phase the causal loop diagrams were drawn. challenges to be analysed with sd are represented through feedback and causal loop diagrams (cld). they are a standard code to represent the structure of the issue of analysis. when an element of a system indirectly influences itself it is called a feedback loop or causal loop. more explicitly, a feedback loop is a closed sequence of causes and effects, that is, a closed path of action and information (richardson & pugh, 1981). such diagrams are useful to analyse relationships that are difficult to describe and understand because of the circular character of the system. in addition, these diagrams can show the cause and effect circularities (kirkwood,1998). causal loop diagrams (clds) are relevant effective tools to represent the feedback structure of systems. they are particularly useful for: (i) quickly defining hypotheses about the causes of dynamics, (ii) obtaining the mental models of individuals or teams, and (iii) communicating the important feedback to be considered in the problem (kirkwood,1998). they consist of variables connected by arrows denoting the causal influences between the variables. variables are related by causal links, shown by arrows. int. j. prod. manag. eng. (2019) 7(special issue), 91-100creative commons attribution-noncommercial-noderivatives 4.0 international testing successful business model using system dynamics 95 http://creativecommons.org/licenses/by-nc-nd/4.0/ a positive link means that if the cause increases, the effect increases above what it would otherwise have been, and if the cause decreases, the effect decreases below what it would otherwise have been. conversely, a negative link means that if the cause increases, the effect decreases below what it would otherwise have been, and if the cause decreases, the effect increases above what it would otherwise have been. 4.3. computational model development after conceptualisation, the model is transferred to a computational model. stock and flow diagrams are used for computerisation of models under system dynamics. stocks are accumulations. they characterise the state of the system sions and actions are based. stocks give systems inertia and provide them with memory. stocks create delays by accumulating the difference between inflow to a process and their outflow. by decoupling rates of flow, stocks are the source of disequilibrium dynamics in systems (sterman, 2000). three examples of stock are as follows: the inventory of a manufacturing firm is the stock of product in its warehouses, the number of people employed by a business is a stock, or the balance of checking account is a stock. the characteristic particularities of stock and flow structures are the following: (i) stocks are represented by rectangles, (ii) inflows represented by a pipe pointing into the stock, (iii) outflows are represented by pipes pointing out of the stock, (iv) valves are reported by two inward pointing triangles and control the flows by opening or closing them, and (v) clouds represent the sources and sinks for the flows. stocks and flows are the elements used in the computational model of this research. stocks for the variables that must be measured, such as customers and performance. flows for representing the variability of the system using the variables which change the measured stocks, such increase in performance. the activities that define the modelling process are: (1) articulating the problem to be addressed, (2) formulating a dynamic hypothesis or theory about the causes of the problem, (3) developing a simulation model to test the dynamic hypothesis (formulation of the simulation model), (4) testing the model until it suits the objectives of the modeller and (5) designing and evaluating policies for improvement. the iterative steps for modelling are defined by sterman (2000). figure 2. conceptual model (own source) int. j. prod. manag. eng. (2019) 7(special issue), 91-100 creative commons attribution-noncommercial-noderivatives 4.0 international ganzarain et al. 96 http://creativecommons.org/licenses/by-nc-nd/4.0/ 1. problem articulation: this is the most relevant step and identifies the issue of concern, time frame, boundary and scope of factors involved. during this phase reference modes and time horizon should be defined. -reference modes definition: a set of graphs and other descriptive data showing the development of the problem over time. -time horizon definition: the period of time to be analysed. it should start as far back in history as necessary to show how the problem emerged and describe its symptoms. 2. dynamic hypothesis: this is the hypothesis the modeller defines to represent the problem and focuses on specific structures. this hypothesis characterises the problem in terms of the underlying feedback loops and stock and flow structure of the system. it is not static, it is temporary and prone to revision. it is related to discussion of the problem and theories associated with causes of the problem (morecroft, 2015). 3. formulation: in most cases it is very difficult or almost impossible to conduct real world experiments that show the faults in a dynamic hypothesis. for this reason accurate and detailed simulation is vital. in this stage it is understood that causal loops, stock and flow diagrams and general policy structure are already defined. causal loops are defined as the maps showing causal links among variables and contain arrows linking causes, effects, and stock and flow tracks. 4. testing: every equation defined in the previous stages must be reviewed for dimensional consistency. in this step sensitivity of model behaviour and policy recommendations need to be evaluated in order to reduce uncertainty. during this stage a comparison of the simulated behaviour of the model to the real behaviour should be done. policy instructions must be checked and models should be tested under extreme conditions. 5. policy formulation and validation: policy design is much more than changing the values of parameters involved in the model. rather it is based on the creation of entirely new strategies, structures and decision rules. policy design not only is based on value change of parameters, but it also combines the creation of entirely new strategies, structures and decision rules (e.g. changing feedback loops, eliminating time delays, defining new decision processes). according to morecroft (2015), the principal interest of policy formulation is improving organisational activity. this question is directly linked to what-ifs. policies can be tested through simulation (morecroft, 2015). yao et al. (2018) figure 3. computational model (own source). int. j. prod. manag. eng. (2019) 7(special issue), 91-100creative commons attribution-noncommercial-noderivatives 4.0 international testing successful business model using system dynamics 97 http://creativecommons.org/licenses/by-nc-nd/4.0/ used system dynamics to explore the influences of different recycling scenarios in china. the simulation is composed of 2 stocks, 2 flows and 10 variables. the key variables as mentioned before are “business model design”, “novelty”, “efficiency”, and “performance”. the rest of the factors are variables defined for each of the constructs. the main aim of this model is to measure performance and understand the influence of “novelty” and “efficiency” on the system. performance is measured, whereas the rest of the variables are direct influencers on bdm. these directly affect performance levels of the firm. 4.4. validation the aim of system dynamics model validation is to verify the validity of the structure of the model. once the structure is validated, behaviour accuracy (of the model) and reproduction of real behaviour through the model is guaranteed. direct structure tests were applied for the validation of the model (barlas, 1996). direct structure tests can be divided into empirical or theoretical. empirical tests involve comparing the model structure with information (quantitative or qualitative) extracted from the real system being modelled. on the other hand, theoretical tests are focused on comparing the model structure with generalised knowledge about the system that exists in the literature. in this project two hypotheses were defined according to the literature (theoretical tests), and then these hypotheses were compared to real situations (empirical tests). h1: the higher the novelty level, the higher performance level. h2: the higher the efficiency level, the higher performance level. the objective was to compare in the same screen the impact of novelty and efficiency, on performance. the value of novelty was reduced in order to visualise its influence. as it was expected, performance lever was lowered. blue oscillation represents the standard model and green represents the scenario in which novelty was reduced. novelty directly influences bmd, and bmd has a direct impact on performance of the firm. brown oscillation represents the scenario in which efficiency was reduced. as expected, an efficiency reduce implies a performance decrease. figure 4. scenario simulation (own source). table 2. variables and their values int. j. prod. manag. eng. (2019) 7(special issue), 91-100 creative commons attribution-noncommercial-noderivatives 4.0 international ganzarain et al. 98 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5. conclusions this research was useful to test the effect of novelty and efficiency, considered key variables, for firm performance. their direct and positive impact on competitiveness has been demonstrated. such an empirical tool can thus be beneficial for decision makers, enabling the acquisition of knowledge, leading to more effective decisions. this research has contributed to the definition of a standard process that could be followed for the successful design of a business model. the steps to be followed are: i) define the strategical goal, ii). identify the key variables that will compose the constructs of the business model. ii) draw a conceptual map with systemic view of the whole business model must be defined, iv) simulate the conceptual model, v) test the hypotheses (understood as the levers and goals to be implemented in the firm), vi) design policies in order to make more effective decisions and define a specific plan for the subsequent decision makers. one of the principal future lines is the aim of testing the tool with a real database. in this way, firms will acquire knowledge about successful business models, as well as key variables for innovation and measurement of outcomes. finally, the authors plan to combine different hypotheses to obtain more real world scenarios. this would take the form of an interactive application that could be used by non experts in simulation, and thus facilitate the use of the tool by managers. references aldrich, h.e. (1999). organizations evolving sage, c.a., thousand oaks. amit, r., zott, c. (2001). value creation in e-business. strategic management journal, 22(6-7), 493-520. https://doi.org/10.1002/smj.187 andreini, d., and bettinelli, c. (2007). business model innovation: from systematic literature review to future research directions, springer. antunes, l., respicio, a., balsa, j., and coelho, h. (2011). gaming and simulations: concepts, methodologies, tools and applications, igi global. degregorio, g. (1999). enterprise‐wide requirements & decision management. incose international symposium, 9, 644-650. https://doi.org/10.1002/j.2334-5837.1999.tb00221.x demil, b., and lecocq, x. (2010). business model evolution: in search of dynamic consistency. long range planning, 43, 227-246. https://doi.org/10.1016/j.lrp.2010.02.004 forrester, j. w. (1997). industrial dynamics. journal of the operational research society, 48(10), 1037-1041. https://doi.org/10.1057/palgrave.jors.2600946 garcía, j. m. (2017). teoría y ejercicios prácticos de dinámica de sistemas para vensim ple, spain, juan martin garcía. giesen, e., berman, s.j., bell, r., and blitz, a. 2007. paths to success: three ways to innovate your business model. ibm institute for business value, 23, 436-438. kirkwood, c. c. (1998). system dynamics methods. (college of business arizona state university usa, ed.), system dynamics methods. magretta, j. (2002). why business models matter. harvard business review, 80, 86-87. milgrom, p. r. (1992). economics, organization and management, nj, prentice-hall. miller, d. (1996). configurations revisited. strategic management j., 17, 505-512. https://doi.org/10.1002/(sici)10970266(199607)17:7%3c505::aid-smj852%3e3.0.co;2-i morecroft, j. (2015). strategic modelling and business dynamics: a feedback systems approach, john wiley & sons., ed. https://doi.org/10.1002/9781119176831 nair, s., paulose, h., palacios, m., and tafur, j. (2013). service orientation: effectuating business model innovation. the service industries journal, 33, 958-975. https://doi.org/10.1080/02642069.2013.746670 richardson, g. p., and pugh iii, a. i. (1981). introduction to system dynamics modeling with dynamo. or: productivity press. schalk, r., timmerman, v., and van den heuvel, s. (2013). how strategic considerations influence decision making on e-hrm applications. human resource management review, 23(1), 84-92. https://doi.org/10.1016/j.hrmr.2012.06.008 sterman, j.d. (2000). business dynamics: systems thinking and modeling for a complex world. irwin-mcgraw-hill. sterman, j.d. (2002). system dynamics: systems thinking and modeling for a complex world. in proceedings of the esd internal symposium. teece, d.j. 2010. business models, business strategy and innovation. long range planning, 43, 172-194. https://doi.org/10.1016/j.lrp.2009.07.003 int. j. prod. manag. eng. (2019) 7(special issue), 91-100creative commons attribution-noncommercial-noderivatives 4.0 international testing successful business model using system dynamics 99 https://doi.org/10.1002/smj.187 https://doi.org/10.1002/j.2334-5837.1999.tb00221.x https://doi.org/10.1016/j.lrp.2010.02.004 https://doi.org/10.1057/palgrave.jors.2600946 https://doi.org/10.1002/(sici)1097-0266(199607)17:7%3c505::aid-smj852%3e3.0.co;2-i https://doi.org/10.1002/(sici)1097-0266(199607)17:7%3c505::aid-smj852%3e3.0.co;2-i https://doi.org/10.1002/9781119176831 https://doi.org/10.1080/02642069.2013.746670 https://doi.org/10.1016/j.hrmr.2012.06.008 https://doi.org/10.1016/j.lrp.2009.07.003 http://creativecommons.org/licenses/by-nc-nd/4.0/ warren, k. (2002). competitive strategy dynamics. wiley. weill, p.m., malone, w.t., d’urso, v.t., herman, g., and woerner, s. (2005). do some business models perform better than others?, a study of the 1000 largest us firms. sloan school of management, massachusetts institute of technology, mit sloan school of management , mit centre for coordination science . yao, l., liu, t., chen, x., mahdi, m., and ni, j. (2018). an integrated method of life-cycle assessment and system dynamics for waste mobile phone management and recycling in china. journal of cleaner production, 187, 852-862. https://doi.org/10.1016/j.jclepro.2018.03.195 zott, c. 2003. dynamic capabilities and the emergence of intra-industry differential firm performance: insights from a simulation study. strategic management journal, 24(2), 97-125. https://doi.org/10.1002/smj.288 zott, c., and amit, r. 2007. business model design and the performance of entrepreneurial firms. organization science, 18(2), 181-199. https://doi.org/10.1287/orsc.1060.0232 int. j. prod. manag. eng. (2019) 7(special issue), 91-100 creative commons attribution-noncommercial-noderivatives 4.0 international ganzarain et al. 100 https://doi.org/10.1016/j.jclepro.2018.03.195 https://doi.org/10.1002/smj.288 https://doi.org/10.1287/orsc.1060.0232 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.10834 received: 2018-10-18 accepted: 2020-01-04 development of a biking index for measuring mediterranean cities mobility ros-mcdonnell, l.a1, de-la-fuente, m.v.a2*, ros-mcdonnell, d.b, cardós, m.c a research goup “industrial engineering”, technical university of cartagena, spain. b research goup “project+city”, technical university of cartagena, spain. c dpto. organización de empresas, universidad politécnica de valencia, spain. a1 lorenzo.ros@upct.es, a2*marivi.fuente@upct.es, b diego.ros@upct.es, c mcardos@doe.upv.es abstract: the european union, its member states and local authorities have been working for long time on the design of solutions for future sustainable mobility. the promotion of a sustainable and affordable urban transport contemplates the bicycle as a mean of transport. the reasons for analysing the cycling mobility in urban areas, has its origin in the confrontation with motorized vehicles, as a sustainable response to the environment. in this context of sustainable mobility, the research team has studied the use of bicycles in mediterranean cities, specifically in coastal tourist areas. the present work shows the development of a mobility index oriented to the bicycle, transport that competes with the private vehicle. by means of a survey methodology, the research group proceeded to collect field data and the subsequent analysis of them, for the development of a mobility index adapted to bicycle mobility, and with possibilities to adapt to urban environments. key words: urban mobility; sustainable mobility; bicycle in cities; biking index. 1. introduction mobility is crucial for the socioeconomic growth of urban areas; its positive effects must be confronted with the negative effects that this growth of mobility has brought during the last 30 years. in this sense, the constant increase of motorization indices in cities and countries annul any argument that seeks to minimize their use (goldman & gorham, 2006; holden, 2007). likewise, the difficulty of parking in the urban area along with traffic congestion reduces the efficiency of private cars, equaling in this aspect to the public transport. in addition, the environmental costs (noise, pollution, etc.) of its use begin to affect the conscience of many drivers. for this reason, the eu, its member states and local authorities have been working for some time on the design of solutions for future sustainable mobility (akerman et al., 2000; banister, 2008; eu commision, 2013). the promotion of a sustainable and affordable urban transport contemplates the transport of the bicycle (dgpi, 2010; eu commision, 2013). in this context of sustainable mobility, the research team has studied the use of bicycles in a mediterranean city, both in the urban centre and in the touristic seaside area. the reasons for analysing to cite this article: ros-mcdonnell, l., de-la-fuente, m.v., ros-mcdonnell, d., cardós, m. (2020). development of a biking index for measuring mediterranean cities mobility. international journal of production management and engineering, 8(1), 21-29. https://doi.org/10.4995/ijpme.2020.10834 int. j. prod. manag. eng. (2020) 8(1), 21-29creative commons attribution-noncommercial-noderivatives 4.0 international 21 mailto:lorenzo.ros@upct.es mailto:marivi.fuente@upct.es mailto:diego.ros@upct.es mailto:mcardos@doe.upv.es http://creativecommons.org/licenses/by-nc-nd/4.0/ the cycling mobility in urban areas, has its origin in the confrontation with motorized vehicles, as a sustainable response to the environment. the needs of the roadways for motorized vehicles have been widely studied (e.g. nijkamp et al., 1996), the vehicles that circulate through each street have been counted (e.g. muñuzuri et al., 2000; ros et al. 2018) , the waiting times at the traffic lights have been analysed, the width of the roadway is being worked on, speeds and other variables (rodriguez & alonzo, 2005), in such a way that it allows to evaluate the road network and act in the points where it is required, but this analysis of the road system does not take into account other forms of mobility such as the bicycle, as much as urban planners arrive to work with certain design parameters that work better or worse, without stopping to analyse in any case, the reasons of the users of bicycles to choose some streets or others, urban networks or others, affecting other urban mobility systems. 2. cycling mobility in urban areas there are many ways to consider urban mobility. in the majority of european countries, mobility discussions focus on promoting the shift from motorized to non-motorized vehicles for short trips or to promote walking or cycling as a healthy leisure activity (dgpi, 2010; eu commission, 2013). the bicycle is a flexible means of transport for urban and interurban trips, as well as other uses (sport, leisure, tourism). in addition, it does not pollute, it is silent, fast in small and medium distances, economical, easy to use and beneficial for health, economy and the improvement of environmental quality. therefore, the creation of a framework that allows the increase of cycling in its different aspects (sports, recreational, daily transport) making it more secure and properly combined with public transport systems, could ensure citizens mobility and accessibility easier (ecf, 2017). no wonder and worldwide, more than 1000 million people use this method on their daily journeys, and around 30% of the european population uses it regularly, but 73% consider that the bicycle should enjoy preferential treatment in front of the automobile (sanz, 1997; miralles & cebollada, 2003; santos & rivas, 2008; dgpi, 2010; geosp, 2017). this european development, promoted since the 70s and 80s, has been due to several factors: agreement between the different political groups or leaving the bicycle outside political controversy, and promoting its use as a means of transport. support to biking user groups, and stable and permanent participation with associations, companies and public administrations related to mobility in the city. take advantage of the capital of technical knowledge and people who have been promoting favorable changes for sustainable mobility and the use of bicycles as a means of transport. promotion of the process of changing infrastructures, the necessary services and of changing the culture and mobility habits of the population. planning process of temporary actions, integrating the bicycle into the general mobility plans and the urban planning of the city. figure 1. use of the bicycle in europe (eurobarometer, 2014). as a result of these actions, and according to eurobarometer (2014), more and more europeans are using the bicycle as a means of transport in everyday life, although the preferred transport continues to be the car. the european average of the use of bicycle in its daily mobility is 8%. but spain has not yet reached those levels, its percentage is 3%. far are countries like holland, with 36%; denmark, with 23%; hungary, with 22%; sweden, which meets 17%; finland, 14%; and belgium, 13%. int. j. prod. manag. eng. (2020) 8(1), 21-29 creative commons attribution-noncommercial-noderivatives 4.0 international ros-mcdonnell et al. 22 http://creativecommons.org/licenses/by-nc-nd/4.0/ besides that, according to the barometro de la bicicleta 2017 (geosp, 2017) made by the spanish network of cities for the bicycle, in spain there has been an increase in its use and the intensity with which it is used with respect to previous years (2011 and 2014). the main results of this barometer (geosp, 2017) are: half of spanish population are bicycle users with some frequency, and 10% use it every day. a quarter of the users who work or study use the bike as a means of transport to go to their jobs or the study centers. the use given in large municipalities and small is different, since in large cities people use it more to go to work or to their study center, while in towns, especially in the smaller ones, its use is for short trips or sports. there is still a long way to reach the european levels of daily bicycle use. but what are the reasons why the bicycle is not used in spain? the main reason, according to barometer 2017, is because they do not want or do not need them. in this motive they fit some like: it does not have bicycle, it prefers to walk or the car, lack of time or, simply, that it does not like. because they can not use the bike, either due to health problems, age, lack of physical shape or because the orography does not allow to use it. due to lack of facilities: afraid, the municipality is not adapted for the use of the bicycle, that there is a lot of traffic, or it has no place to store it. in spain, the social potential of the bicycle is greater than imagined. around 66% of the population is in favour of alternative mobility policies to those that have been granting hegemony to the automobile (miralles & cebollada, 2003; santos & rivas, 2008; geosp, 2017). the bicycle represents a more than valid means of transport to promote sustainable mobility and reduce the daily conflicts of urban traffic. annually, the dgt publishes the report of the annual barometer of the bicycle in spain. these reports reflect an increase of 0.53% of cyclists per year, reflecting a positive image of cycling, both as a vehicle and as a healthy leisure offer, as well as the main advantages of its use and the reasons for its non-use in the city (geosp, 2017). table 1. modal share of bicycle, on total trips, in spanish cities (geosp, 2107). sevilla 6% zaragoza 3% vitoria 2-3% san sebastian 2-3% sabadell 2% lerida 2% barcelona 1.33% madrid 0.3% granada 0.25% malaga 0.2% intermodality, as reflected in table 1, is a pending issue in most spanish cities. taking into account the bicycle in relation to other transport (bus, tram, subway, car, etc.) is a proof of the structural integration of the bicycle in urban planning and mobility. 3. proposal of a biking index the objective of this work is to study the citizens’ mobility using means such as cycling in mediterranean cities, considering: the importance of the mobility of citizens in their immediate surroundings. obtain an index that could be used by local authorities. help city planning by understanding the conditions of biking mobility compared to other cities or urban areas. assist city planning by identifying useful aspects for biking mobility as well as future recommendations. next section will show a brief summary of existing bikeability indexes, as well as the work developed by the research team. 3.1. measuring bikeability: literature review the development of bikeability indexes during the last decade has shown that cycling has received less attention than walking in the scientific literature. different cycling measures began in order to improving citizens health and public health aspects, but now its also focus is on neighborhoods and city planning and design. a growing body of research has explored how the built environment influences int. j. prod. manag. eng. (2020) 8(1), 21-29creative commons attribution-noncommercial-noderivatives 4.0 international development of a biking index for measuring mediterranean cities mobility 23 http://creativecommons.org/licenses/by-nc-nd/4.0/ physical activity, with findings that people who live in more walkable neighbourhoods walk more, have lower rates of obesity and chronic disease, and travel less by car (ewing et al., 2003 & 2006; pucher & dijkstra, 2003; forsyth et al, 2007; flowerdew et al., 2008; dill, 2009; frank et al., 2009). the existing focus on walking is justifiable given that it is the most common form of leisure-time physical activity, with few barriers and no cost. however bicycle travel, being faster and more efficient while nearly as accessible and economical, is a more reasonable substitute for automobile travel when trip distances exceed 1 km (european commission, 1999). in this sense, cycling is not only an underused transportation mode in developed countries. the utility of cycling for transportation has been recognized in model cities such as copenhagen and amsterdam, and but eu and american countries want to promote a shift to active transportation for trips of moderate distance, beyond distances suitable to walking. among the variety of indexes to measure active accessibility, bike score® stands out, developed by winters et al. (2013, 2016), which combines environmental characteristics with between and within-city variability in cycling behavior, and based on the concepts of infrastructure, safety, topography and climatology. its peculiarity resides in that it has been exclusively developed for north american cities, clearly different from european cities. krenn et al. (2015) propose a new bikeability index developed for mid-sized european cities and based on gis, that seeks to measure the bicyclefriendliness, based on infrastructure, green areas and topography. for the calculation of this index they define 200m cell buffers, but it presents severe limitations in the update of the digital maps. an international index, the copenhagenize index (2011), gives cities points for their efforts towards reestablishing the bicycle as a feasible, accepted and practical form of transport. the authors consider three parameters (streetscape, culture and ambition) that cover different factors of the city and the bicycle: infrastructures, facilities, traffic calmimg, safety, modal share, politics, urban planning, etc. developed since 2011, the copenhagenize index aims to point out the most important bicycle-friendly cities. although it only considers capital nations and cities with more than 600,000 inhabitants, not facilitating the calculation of the index to small cities. it should not be forgotten indexes specifically designed for cities and regions. in this sense, it is important to note the study conducted by har-tanto et al. (2017), in a dutch city region, mea-suring bikeability in a tod context, with the objective of consolidating and facilitating combined means of transport, especially bicycle-train, the most common combination of travel for work in the netherlands. in another study, mesa & barajas (2013) developed the bikeability index for cali city in mexico. this biking index is based on four factors (slope, environmental quality, quality of infrastructure and personal safety), and was used to evaluate connectivity between major zones that generate and attract cycling trips, in order to show that the potential impacts of proposed cycling investments are in areas with low bikeability. finally, to point out a recent spanish study conducted for a mediterranean city (sanchez, 2016), but that only contemplates the infrastructure of the bike lane, and avoiding factors such as environment, traffic conditions, or personal safety. as a summary of the literature review made for this work, note that most authors who have studied this topic have developed their indexes based on the two groups of variables next described. some factors will encourage bicycle ridership: safer bicycle routes, better lighting, etc., while some factors were identified as obstacles to cycling: inclement weather, reduced bicycle security, crime, fear about personal safety and lack of bicycle lanes (botma, 1995; hydén et al., 1998; bulkeley & betsill, 2005; jensen, 2007; marqués et al., 2015; vale et al., 2016; winters et al., 2016; gutierrez et al., 2017; hartanto et al., 2017). the proposal of a new index for mediterranean cities, is carried out because most existing indexes obviate the size and characteristics of the city, focusing on the possibility of moving around the city between traffic but do not contemplate studying small cities, and as lowry et al. (2012) propose, this new index wants to assess existing infrastructures, identify the problematic sections, inform against the high non-connectivity level of the bicycle network, and how to enhance biking mobility, and not only for leisure and sports. int. j. prod. manag. eng. (2020) 8(1), 21-29 creative commons attribution-noncommercial-noderivatives 4.0 international ros-mcdonnell et al. 24 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.2. defining a biking index in a mediterranean city the present work shows the development of a mobility index oriented to the bicycle, transport that competes with the private vehicle, especially for mediterranean cities, coastal and flat cities, with a strong tourist component throughout the year due to its good weather, and with compact urban centers and narrow streets. according to the study of krambeck (2006), and by means of a survey methodology, the research group proceeded to collect field data and the subsequent analysis of them, for the development of an index adapted to bicycle mobility, also for urban centres and tourist areas. the biking index is based on several variables: type of bike lane, lane cleaning, parking, illumination and signals, obstacles, lane layout, danger and accidents, etc.; includying the number of users (cyclists) and the lane’s lenght. the combination of these variables (qualitative and quantitative) allows the calculation of an index that will reflect the status of bicycle mobility, firstly in tourist areas (environment initially studied), but later extended to biking mobility in the city center. 3.2.1. research methodology in a first phase, based on the work of krambeck (2006), the research team designed a questionnaire in order to collect all the peculiarities of the bicycle lane and its use in order to reflect all the information related to the mobility of cyclists in the mediterranean city,differentiating the tourist area from the urban area. likewise, the two areas selected for the present study were clearly delimited (figure 2 and 3). in the tourist area the bike lane was divided into 38 equidistant segments (of 500 m), but in the urban area the bike lane network was divided in 58 segments of 100-300 m of length, due to the irregular lane layout in the city centre. these divisions of the bike lane in segments will facilitate the data collection. in a second phase, the research team conducted the surveys for the data collection, during two different periods: june-july: for the tourist area, because of there is enough population in this area to study the infrastructure under analysis, the cycle path and its use by cyclists. march-may: for the urban area, according to labor periods and also on easter holidays. information was collected from two random statistical samples (one for each area under study and its segments), taking data from each segment in the morning and afternoon, and distinguishing between right and left lanes (if possible). figure 2. bike lane developed along la manga area (marked as red line). figure 3. bike network developed in urban area of cartagena. existing lane is marked as red line. proyected lane is marked as green line. proposed lane is marked as blue line. int. j. prod. manag. eng. (2020) 8(1), 21-29creative commons attribution-noncommercial-noderivatives 4.0 international development of a biking index for measuring mediterranean cities mobility 25 http://creativecommons.org/licenses/by-nc-nd/4.0/ in a third phase, the collected data were analysed, and the analysis of the defined variables allowed the development of the biking index (see tables 2 and 3). in these tables the research team show the collected data (parameters of the bike lane network and the number of cyclists on it) and the biking index for a selection of segments of the two areas studied. 3.2.2. analysis of results as remarked in the previous section, the research team has considered a set of variables (type of bike lane, parking, illumination and signals, obstacles, lane layout, accidents, etc.) for the biking index calculation. the set of variables was grouped into six parameters to characterized the bike lane network: parameter 1: characteristics of the segment for mobility (conflicts with other modes of transport, bike lane conditions, maintenance). parameter 2: mobility and urban road crossing (type of bike lane, quality of crossing points). parameter 3: obstructions in mobility segments (congestion due to use, temporary or permanent obstructions). parameter 4: safety in mobility (bike lane safety, theft and crime, security against other modes of transport). parameter 5: signaling and lighting of the bike lane. parameter 6: connection and distribution (use of the lane to reach the destination, parking, other facilities). the research team shows a first approximation of the biking index (bi) for mediterranean cities, the authors have considered that all the parameters have the same weight in the calculation of the index, which also includes as calculation factors the number of cyclists that use the lane segment and the length of the segment studied. bi = αp + βc + γl were: p: average of the parameters of the segment c: number of cyclists in the segment l: length of the studied segment. α, β, γ: coefficients associated with the variables the value obtained for each of the segments of the bike lane varies from 1 to 5, indicating this value the level of use and state of the infrastructure. the results obtained are shown in tables 2 and 3, for each segment of the studied areas studied. table 2. collected data along the segments of the tourist area. results for the biking index. a-4 a-5 a-12 a-16 a-19 a-20 37 48 a-53 a-65 a-71 parameter 1 3 5 3 4 3 5 2 4 2 1 4 parameter 2 2 1 1 1 1 1 3 3 1 1 3 parameter 3 5 5 3 4 5 5 5 5 3 4 5 parameter 4 2 4 3 3 4 5 3 4 1 2 5 parameter 5 2 3 3 3 3 3 3 4 2 2 2 parameter 6 3 3 3 3 4 4 5 4 3 2 5 n° cyclists 1 2 1 3 3 0 4 5 1 2 5 biking index 2.7 3.6 3 3.3 3.6 3.9 3.4 4.9 2.3 2.4 4.3 table 3. collected data along the segments of the urban area. results for the biking index. 3 5 9 15 25 26 32 41 48 54 57 parameter 1 4 4 4 5 3 4 3 5 5 5 5 parameter 2 4 3 3 4 4 3 2 4 4 4 4 parameter 3 3 3 3 4 3 3 2 4 3 4 3 parameter 4 4 4 4 4 3 4 3 4 4 5 4 parameter 5 5 3 4 4 2 2 3 3 3 3 4 parameter 6 4 2 3 2 2 4 1 3 2 2 26 n° cyclists 3 3 6 15 6 6 7 12 3 5 4 biking index 1.96 3.21 3.62 4.04 3.46 2.86 2.68 4.24 4.25 4.36 3.67 int. j. prod. manag. eng. (2020) 8(1), 21-29 creative commons attribution-noncommercial-noderivatives 4.0 international ros-mcdonnell et al. 26 http://creativecommons.org/licenses/by-nc-nd/4.0/ finally, the bi value has been calculated for each area of the city, obtaining the following biking index values: urban area: bi= 3.54 tourist area: bi= 3.14 the calculated biking index for each segment determines its status and its use, as we can see in tables 2 and 3, most sections have a biking index value between 2 and 4, which shows the deteriorated aspect, the lack of maintenance and repairing, and the misuse of the bike path (figures 4 to 7). only in few sections, where the lane is unfolded (l/ r), segregated from the road and well signposted, the value of the index exceeds 4. these results are very important for the responsible manager of the bike lane network. this information will entail different actions (as explained in next section) from the local administration, which has been recently warned, in order to improve the cycle path for daily use in the city and the next tourist season. figure 4. example of deteriorated bike lane, and lack of maintenance (tourist area) figure 5. example of permanent obstacles in the bike lane (tourist area) figure 6. example of conflicts with pedestrians (urban area) figure 7. example of permanent obstacles and bad signals in the bike lane (urban area) int. j. prod. manag. eng. (2020) 8(1), 21-29creative commons attribution-noncommercial-noderivatives 4.0 international development of a biking index for measuring mediterranean cities mobility 27 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. conclusions a great number of spanish cities, (dgpi, 2010, geosp, 2017) deal with the bicycle issue with great sensitivity, adapting its cyclist lane network around the urban core and within it, which enable a comfortable and functional circulation, choosing the bike as a vehicle to move. a key to this good functioning is the cyclist network master planning (mobility studies, user surveys), achieving a branched network, with connected segments, and reducing the isolated lanes to zero. the proposal of a new index for mediterranean cities, is carried out because most existing indexes are defined for big cities and focused on the possibility of cyclying betwwen daily traffic congestion. this new index wants to assess existing infrastructures, identify the problematic sections, inform against the high non-connectivity level of the bicycle network, and how to enhance biking mobility, and not only for leisure and sports. the analysis of the studied network (infrastructure in la manga del mar menor and the urban área of cartagena) allows to identify the disparity of situations found along the bicycle lane (existence or not of cycle path, different types of them, etc.) together with an overpopulation of the geographical area in the spring and summer seasons, during the realization of the survey. the following results and conclusions found after the visual analysis and data collection were considered: need for connection of the different stretches of bike lanes, uniformity of the different types of bike lanes, need for investments to maintain the bike lane (high level of deterioration in a large number of areas). if the local and/or regional administrations decide to bet on the use of the bicycle and its connection with urban buses and commuter trains, the abovementioned needs could improve and increase bicycle mobility in the municipality, and establishing the connection of the different urban centers with the beach area. this biking index will allow local authorities to take realistic decisions about: safer and more adequate infrastructures in cities and neighbourhoods, promote safety and education for bike riders and other citizens, importance of intermodality, facilitating the use of bicycles together with other means of transport, prevention of robberies and aggressions against cyclists. references akerman, j., banister, d., dreborg, k., nijkamp, p., schleicher-tappeser, r., stead, d., steen p. (2000). european transport policy and sustainable mobility. ed. routledge, taylor & francis group. https://doi.org/10.4324/9780203857816 banister, d. (2008). the sustainable mobility paradigm . transport policy, 15(2), 73-80. https://doi.org/10.1016/j.tranpol.2007.10.005 botma, h. (1995). method to determine level of service for bicycle paths and pedestrian-bicycle paths. transportation research record: journal of the transportation research board, 1502, 38-44. bulkeley, h., betsill, m.m. (2005). rethinking sustainable cities: multilevel governance and the urban politics of climate change. environmental politics,14(1), 42-63. https://doi.org/10.1080/0964401042000310178 copenhagenize index (2011). https://copenhagenizeindex.eu dgpi (2010). fomento del transporte en bicibleta. dirección general de políticas interiores – transporte y turismo, parlamento europeo. dill, j. (2009). bicycling for transportation and health: the role of infrastructure. journal of public health policy, 30, 95-110. https://doi.org/10.1057/jphp.2008.56 ecf (2017). eu cycling strategy. recommendations for delivering green growth and an effective mobility in 2030. https://ecf.com/eu_cycling_strategy eurobarometer (2014). attitudes on issues related to eu transport policy. european commission (1999). cycling: the way ahead for towns and cities (european commission, brussels). european commission (2013). civitas initiative, https://www.civitas-initiative.eu. ewing, r., schmid, t., killingsworth, r., zlot, a., & raudenbush, s. (2003). relationship between urban sprawl and physical activity obesity and morbidity. american journal of health promotion, 181, 47–57. https://doi.org/10.4278/0890-1171-18.1.47 int. j. prod. manag. eng. (2020) 8(1), 21-29 creative commons attribution-noncommercial-noderivatives 4.0 international ros-mcdonnell et al. 28 https://doi.org/10.4324/9780203857816 https://doi.org/10.1016/j.tranpol.2007.10.005 https://doi.org/10.1080/0964401042000310178 https://copenhagenizeindex.eu https://doi.org/10.1057/jphp.2008.56 https://ecf.com/eu_cycling_strategy http://www.civitas-initiative.eu https://doi.org/10.4278/0890-1171-18.1.47 http://creativecommons.org/licenses/by-nc-nd/4.0/ ewing, r., handy, s., brownson, r., clemente, o., winston, e. (2006). identifying and measuring urban design qualities related to walkability. journal of physical activity and health, 3, 223-240. https://doi.org/10.1123/jpah.3.s1.s223 flowerdew, r., manley, d., sabel, c. e. (2008). neighbourhood effects on health: does it matter where you draw the boundaries? social science and medicine, 666, 1241-1255. https://doi.org/10.1016/j.socscimed.2007.11.042 forsyth, a., oakes, j., schmitz k., hears, m. (2007). does residential density increasing walking and other physical activity? urban studies, 44, 679-697. https://doi.org/10.1080/00420980601184729 frank, l., sallis, j., saelens, b., leary, l., cain, k., conwa, t., hess, p. (2009) the development of a walkability index: application to the neighborhood quality of life study. british journal of sports medicine, 44, 924-933. https://doi.org/10.1136/bjsm.2009.058701 geosp (2017). barómetro de la bicicleta en españa. informe de resultados 2017. goldman, t., gorham, r. (2006). sustainable urban transport: four innovative directions. technology in society, 28(1-2), 261-273. https://doi.org/10.1016/j.techsoc.2005.10.007 gutierrez, c., gu, s., karam, l., thomas, t. (2017). measuring and evaluating bikeability in san francisco. in urbanst 164: sustainable cities, 3-29. hartanto, k., grigolon, a., maarseveen, m., brussel, m. (2017). developing a bikeability index in the context of transit-oriented development (tod). in: 15th international conference on computers in urban planning and urban management (cupum), adelaide (australia). holden, e. (2007). achieving sustainable mobility: everyday and leisure-time travel in the eu. ed. ashgate publishing. hydén, c., nilsson, a., risser, r. (1998). how to enhance walking and cycling instead of shorter car trips and to make these modes safer. in: institutionen för trafikteknuk, lunds tekniska högskola, n° 165. jensen, s.u. (2007). pedestrian and bicyclist level on roadway segments. transportation research record: journal of the transportation research board, 2031, 43-51. https://doi.org/10.3141/2031-06 pucher, j., dijkstra, p. (2003) promoting safe walking and cycling to improve public health: lessons from the netherlands and germany. american journal of public health, 939, 1509-1516. https://doi.org/10.2105/ajph.93.9.1509 krambeck, h.v. (2006). the global walkability index. mit. krenn, p.j., oja, p., titze, s. (2015). developing a bikeability index to score the biking friendliness of urban environments. open journal of civil engineering, 5, 451-459. https://doi.org/10.4236/ojce.2015.54045 lowry, m. b., callister, d., gresham, m., moore, b. (2012). assessment of communitywide bikeability with bicycle level of service. transportation research record, journal of the transportation reserach board, 2314, 41-48. https://doi.org/10.3141/2314-06 marqués, r., hernández-herrador, v., calvo-salazar, m., garcía-cebrián, j. a. (2015). how infrastructure can promote cycling in cities: lessons from seville. research in transportation economics, 53. https://doi.org/10.1016/j.retrec.2015.10.017 mesa, v. g., barajas, d. e. p. (2013). cali bikeability index map : a tool for evaluating public investment and future needs. journal of transport geography, 4(1), 5-8. miralles, c., cebollada, a. (2003). movilidad y transporte. opciones políticas para la ciudad. fundación alternativas. muñuzuri, j., cortés, p., onieva, l., guadix, j. (2000). modelling freight delivery flows: missing link of urban transport analysis. journal of urban planning and development, 135(3), september 2000. https://doi.org/10.1061/(asce)up.1943-5444.0000011 nijkamp, p., reggiani, a., tritapepe, t. (1996). modelling inter-urban transport flows in italy: a comparison between neural network analysis and logit analysis. transportation research part c: emerging technologies, 4(6), 323-338. https://doi.org/10.1016/s0968-090x(96)00017-4 rodriguez, g., alonzo, l. (2005). carreteras. ed. upm. ros-mcdonnell, l., de-la-fuente, m.v., ros-mcdonnell, d., cardós, m. (2018). analysis of freight distribution flows in an urban fuctional area. cities, 79, 159-168. https://doi.org/10.1016/j.cities.2018.03.005 sanchez, m. (2016) metodología para la evaluación de la ciclabilidad de una red urbana de carriles bici. aplicación en la ciudad de málaga. xvii congreso nacional de tecnologías de información geográfica (málaga) santos, l., de las rivas, j.l. (2008). ciudades con atributos: conectividad, accesibilidad y movilidad. revista ciudades, 11, 13-32. sanz, a. (1997). movilidad y accesibilidad: un escollo para la sostenibilidad urbana. in http://habitat.aq.upm.es/cs/p3/a013.html. vale, d., saraiva, m., pereira, m. (2016). active accessibility: a review of operational measures of walking and cycling accessibility. the journal of transport and land use, 9, 209-235. https://doi.org/10.5198/jtlu.2015.593 winters, m., brauer, m., setton, e., teschke, k. (2013). mapping bikeability: a spatial tool to support sustainable travel. environment and planning b: planning and design, 40(5), 865-883. https://doi.org/10.1068/b38185 winters, m., teschke, k., brauer, m., fuller, d. (2016). bike score: associations between urban bikeability and cycling behaviour in 24 cities. international journal of behavioral nutrition and physical activity, 13(18), 1-10. https://doi.org/10.1186/s12966-016-0339-0 int. j. prod. manag. eng. (2020) 8(1), 21-29creative commons attribution-noncommercial-noderivatives 4.0 international development of a biking index for measuring mediterranean cities mobility 29 https://doi.org/10.1123/jpah.3.s1.s223 https://doi.org/10.1016/j.socscimed.2007.11.042 https://doi.org/10.1080/00420980601184729 https://doi.org/10.1136/bjsm.2009.058701 https://doi.org/10.1016/j.techsoc.2005.10.007 https://doi.org/10.3141/2031-06 https://doi.org/10.2105/ajph.93.9.1509 https://doi.org/10.4236/ojce.2015.54045 https://doi.org/10.3141/2314-06 https://doi.org/10.1016/j.retrec.2015.10.017 https://doi.org/10.1061/(asce)up.1943-5444.0000011 https://doi.org/10.1016/s0968-090x(96)00017-4 https://doi.org/10.1016/j.cities.2018.03.005 http://habitat.aq.upm.es/cs/p3/a013.html https://doi.org/10.5198/jtlu.2015.593 https://doi.org/10.1068/b38185 https://doi.org/10.1186/s12966-016-0339-0 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering special issue: advances in engineering networks https://doi.org/10.4995/ijpme.2019.11513 received: 2019-03-14 accepted: 2019-05-18 electrification of madrid fleet public transport company (emt-madrid): strategic analysis and implementation garcía hernanz j. a1, morales-alonso, g. a2, fernández sánchez, g.b1, pilkington gonzález, e.b2, and sánchez chaparro, t. a3 a dpto. de ingeniería de organización, administración de empresas y estadística. etsi industriales. universidad politécnica de madrid. c/ josé gutiérrez abascal, 2, 28006 madrid. b empresa municipal de transportes de madrid (emt). dpto. de estrategia, procesos e innovación. c/ cerro de la plata, 4, 28007 madrid. a1 jaime.garcia.hernanz@alumnos.upm.es, a2 gustavo.morales@upm.es, b1 gonzalo.fernandez@emtmadrid.es, b2 eduardo.pilkington@emtmadrid.es, a3 teresa.sanchez@upm.es abstract: madrid public transport company (emt-madrid) is a property of the madrid city council, and it provides the public buses service in the whole city. madrid, as most of the big cities in the world, is facing problems related to high levels of urban pollution, which directly affects the health and life quality of their inhabitants. emt, having a fleet of around 2000 buses, has an impact in the mentioned problem and in the global warming. with the strategic plan 2017-2020, many new buses will be acquired, resulting in a fleet of natural gas, hybrid and electric vehicles by the end of 2020. the present study has the goal of being the cornerstone of a future strategic plan of the company. to this end, both external and internal analyses of the company have been conducted, which support that the electrification of the whole fleet is the best option in the long term. furthermore, a benchmarking of the state of the public transport in other 25 cities and the technology used in them has been conducted. last, a model that allows replicability of this strategic assessment is proposed, in order to help other transport companies and city councils to decide which transport fleet is the best to implement in their cities depending on their necessities and resources. key words: sustainable transport, electric buses, air quality, emissions, benchmarking. 1. introduction the presence of high concentrations of air pollutants in the city of madrid has led to traffic restrictions in the city up to three days during december 2018 (ayuntamiento de madrid, 2019). this problem is not exclusive from madrid, as many other large cities of the world are frequently exceeding the pollutants concentration level that the world health organization considers safe to breathe (world health organization, 2005). the pollutants with the highest impact (due to its concentrations in the urban air and its harmfulness) are particulate matter (pm), nitrogen oxides (nox) and ozone (o3). these compounds cause an increase in mortality and numerous cardiovascular and respiratory system diseases. there exists criticism both among the scientific community and public institutions on the attribution of pollutants solely to the exhaust of combustion engines. for instance, the german federal environmental agency supports that brakes are responsible (in modern cars of euro 6 type) to cite this article: garcía hernanz j., morales-alonso, g., fernández sánchez, g., pilkington gonzález, e., and sánchez chaparro, t. (2019). electrification of madrid fleet public transport company (emt-madrid): strategic analysis and implementation. international journal of production management and engineering, 7(special issue), 107-116. https://doi.org/10.4995/ijpme.2019.11513 int. j. prod. manag. eng. (2019) 7(special issue), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international 107 https://orcid.org/0000-0002-3243-2719 http://orcid.org/0000-0001-5753-495x https://orcid.org/0000-0003-3444-1501 http://creativecommons.org/licenses/by-nc-nd/4.0/ for four times more particle emission pollutants than the combustion engine. likewise, grigoratos and martini (2014) pose that brake wear can sum up to 21% of pm10 traffic-related emissions. airparif (2017) reports that, while pm2.5 can be attributed to exhaust emissions, pm10 “particles include a significant fraction related to tire-wear, brake-wear, road abrasion and dust suspension”. regardless of whether pollutants come from the engine exhausts or from other traffic sources, it is clear that private traffic is directly responsible for two of them, pm and nox, and indirectly responsible of o3. regarding the city of madrid, reducing these emissions in the city is in the air quality plan of the city (ayuntamiento de madrid, 2017). each kind of engines produces different pollutants. diesel buses produces high quantities of co2, nox and pm. natural gas buses slightly increase the co2, reduce the nox and practically vanish the pm emissions. electric and hydrogen powered buses have zero emissions. in the field of electric buses, there are also different ways to charge them: a direct classic connector to the electric network in a charging station (which can be fast or slow depending on the power), a pantograph (automatic mechanical arm that connects with the bus from above) and magnetic induction (a magnetic field charges the battery from below). also, trolleybuses can use the tram lines to charge their batteries. emt-madrid (empresa municipal de transportes de madrid) is the public company that operates the whole public buses network in the city of madrid, having a fleet of around 2000 units. the current strategic plan of the company for the period of 2017-2020 has the aim of replacing all their diesel buses for low emissions or zero emissions buses by the end of 2020 (empresa municipal de transportes de madrid, 2017). this ecofriendly or green fleet includes compressed natural gas (cng), hybrid and electric buses. emt-madrid currently has around 1000 diesel, 940 cng, 40 hybrid and 20 electric buses. by the end of 2020, emt-madrid will have acquired close to 80 new electric buses and 940 new cng buses, achieving a full green fleet. the present study aims to be the seed of the future strategic plan of the company, on its way to the 0 emissions fleet. this procurement is well aligned with the proposal of ocampo and ocampo (2015), “developing initiatives that simultaneously enhance customer satisfaction and community development by addressing environmental concerns on toxic substance, ghg emissions and air emissions is fundamentally important to increase revenue and profit.”. moreover, vernadat (2014) states that successful management delves with making sure that the processes are executed in such a way that business objectives are achieved in an efficient, effective and economic way. the three main goals of this research are: strategic analysis of the company (internal and external), focusing in the fleet, infrastructure, human resources and everything related with the electrification. benchmarking of the situation of the public transport in other cities and their plans and strategies. creation of a model that can guide and support other transport companies and city councils of the world to decide which transport fleet is the best to implement in their cities depending on their necessities and resources. the rest of the article is organized as follows, similarly as in ros-mcdonnell et al., (2018). next section presents the theoretical framework, while section 3 presents the methodology used. section 4 presents the results obtained in the strategic analysis and benchmarking, which includes the strategic lines that the company should implement. in section 4, the decision tree is also presented, and finally, in the last section, conclusions, limitations and avenues for further research are gathered. 2. theoretical framework 2.1. global warming global warming is a huge problem that affects the whole planet and it is mainly caused by the greenhouse effect. this effect, according to kyoto protocol (1998), is produced by the greenhouse effect gases: carbon dioxide (co2), methane (ch4), nitrogen dioxide (no2), hydrofluorocarbons (hfcs), perfluorocarbons (pfcs) and sulphur hexafluoride (sf6), emitted in sectors as energy (including fuel consumption in transport), industry, agriculture and waste (united nations, 1998). these gases reflect part of the heat that the earth emits to space, causing an increase of the average temperature of the planet. according to the study “long-term climate change: projections, commitments and irreversibility”, int. j. prod. manag. eng. (2019) 7(special issue), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international garcía hernanz et al. 108 http://creativecommons.org/licenses/by-nc-nd/4.0/ consequences of global warming include more hot and cold extreme temperatures, changes in atmospheric circulation, in the ocean and in the water cycle, reduction of the arctic ice extent and climate change (cambridge university press, 2013). 2.2. urban growth and its consequences norman foster in the “future is now forum madrid 2017” explained that society is migrating from rural to urban areas to an unprecedented level, and estimated that by 2050, 75% of the world population would live in large cities. in addition, the private car will be no longer used, and more sustainable and innovative transport systems will be implemented (norman foster, 2017). enrique dans and gildo seisdedos, in their study “upgrading urban mobility”, analyze the current situation of unsustainability in large cities. the solutions proposed so far, from restrictive measures to incentives for models based in sustainable transport systems have been, in most cases, insufficient. the alternatives available to the citizen are not enough to consider transport models not based on the use intensive private vehicle, even for the cities with the best public transport. it is estimated that 60% of the world population will live in cities in 2030. in that year, if the current trend continues, the world’s car fleet will double its amount from 1.2 to 2.4 billion cars. a change to a public and shared transport is needed (dans and seisdedos, 2016). 2.3. emissions and their origin the most important pollutants according to the world health organization are the following: particulate matter (pm): pollution by particulate material occurs with the entry of suspended particles into the atmosphere. the main difference between these particles in their way of affecting the organism is that pm10 particles (diameter equal to or less than 10 μm) can penetrate to the lower respiratory tract, while pm2.5 (diameter equal to or less than 2.5 μm), being smaller, can be breathed, reaching the gas exchange zones of the lung. if the particles are ultrafine, that is, less than 100 nanometers, they can also get into the bloodstream (world health organization, 2005). ozone (o3): the polluting ozone is not the one found in the upper atmosphere, but it is the one that is produced by secondary reactions of other pollutants and is at ground level, so can be breathed by people. due to its high oxidizing power over living materials and tissues, ozone can cause premature aging and deterioration of the lungs, as well as irritations in the respiratory system, cough, asthma, headaches and alterations of the immune system (world health organization, 2005). nitrogen dioxide (no2): the effects of no2 on health are an important inflammation in the respiratory tract, decreased development of lung function and an increase in bronchitis symptoms in children with asthma (world health organization, 2005). sulfur dioxide (so2): so2 affects the respiratory system and pulmonary functions, producing heart disease, cough, mucous secretion, asthma, chronic bronchitis and an increase in mortality, in addition to increasing the risk of contracting respiratory infections. so2 also causes eye irritation to exposed people (world health organization, 2005). table 1 shows a comparison between the maximum values allowed in spain according to the royal decree in force with the values defined by the table 1. comparison maximum concentration allowed by the eu and by the who. pollutant measurement period who (2005) 2008/50/ec rd 102/2011 times it can be exceeded rd 102/2011 pm2.5 24 hours 25 µg/m 3 cannot be exceed 1 year 10 µg/m3 20 µg/m3 cannot be exceed pm10 24 hours 50 µg/m 3 50 µg/m3 35 times/year 1 year 20 µg/m3 40 µg/m3 cannot be exceed o3 8 hours 100 µg/m 3 120 µg/m3 25 times/year no2 1 hour 200 µg/m 3 200 µg/m3 18 times/year 1 year 40 µg/m3 40 µg/m3 cannot be exceed so2 10 minutes 500 µg/m 3 1 hour 350 µg/m3 24 times/year 24 hours 20 µg/m3 125 µg/m3 3 times/year int. j. prod. manag. eng. (2019) 7(special issue), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international electrification of madrid fleet public transport company (emt-madrid): strategic analysis and implementation 109 http://creativecommons.org/licenses/by-nc-nd/4.0/ european union and the values defined by the world health organization (boletín oficial del estado, 2011) (world health organization, 2005). as can be seen in the table 1, the european regulations are less restrictive than the maximum recommended values of the world health organization (who). according to the ministry of energy, in spain, transport is the sector with the higher emissions, having in 2016 a 42% of the energy demand of the country (ministerio de energía, turismo y agenda digital. gobierno de españa, 2016). 2.4. research question in light of the negative effects of air pollutants over the health of big cities inhabitants, public institutions have the responsibility of taking measures against the origin of these pollutants. in that vein, public transport policies are a powerful leverage to improve air quality. in the authors’ opinion, the strategic plan of emtmadrid can be taken as a brave initiative, and therefore suggests the following research question of this study: “what are the main recommendations that can be extracted from the strategic plan of emtmadrid in order to be followed in other cities facing similar problems?” 3. methodology in order to shed light on the abovementioned research question, a two-fold methodology is proposed. 3.1. strategic analysis firstly, the methods used for the strategic analysis of the company were the model of the competitive forces (michael e. porter, 1998) and the pestel analysis (political, economic, social, technological, ecological and legal factors) for the external part (effects from the environment). on the other hand analysis of the significant documents, interviews with the stakeholders and visits to the different sections of the company for the internal part of the analysis. with these factors, a swot matrix (strengths, opportunities, weaknesses and threats) was created and used to decide the strategic lines of attack, defence, reorientation and survival. 3.2. benchmark and decision tree the second method used is a benchmarking analysis (see the process followed in figure 1), that aims at identifying the actions taken by other cities, to analyze those from emt-madrid under a new perspective. to that end, 25 cities were selected, with the criteria of (i) having similar characteristics with madrid or (ii) being particularly innovative in the conception of their public transport. next, the indicators for conducting the comparison were selected. then, the cities were analyzed in light of all the available documentation. finally, recommendations of measures to be implemented by emt-madrid are summarized. the indicators in use for the benchmarking analysis focus on three streams. (i) the city, (ii) the inhabitants’ characteristics and (iii) the city public transport system. regarding the first stream, the surface of these cities, the average traffic speed, the modal split and the average annual levels of pollutants in the air have been compared. secondly, the focus is made on the inhabitants, in which the number of inhabitants, population density, minimum and average salary, comparing it with the price of a bus ticket and the trips that are made per year in each means of transport have been examined. that third stream deals with public transport systems, especially buses, the number of bus lines, the number of bus stops, the average speed of bus lines, the kilometers traveled by buses per year, the number of buses and the breakdown of the fleet according to the criteria of fuel or electrical technology and the kilometers of lines of each type of public transport in the city. figure 1. methodology used for the benchmarking. int. j. prod. manag. eng. (2019) 7(special issue), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international garcía hernanz et al. 110 http://creativecommons.org/licenses/by-nc-nd/4.0/ the complete list of the cities used for the benchmarking analysis is listed in table 2. lastly, a decision tree has been crafted upon the results obtained from the benchmarking, in order to allow for replicability of the strategic decisions of emt-madrid in other cities. table 2. list of the cities used for the benchmarking analysis. 1. barcelona 14. gothenburg 2. valladolid 15. genève 3. bilbao 16. amsterdam 4. san sebastián 17. tokyo 5. berlin 18. beijing 6. bremen 19. shanghai 7. rome 20. honk kong 8. milan 21. sao paulo 9. brussels 22. tel-aviv 10. paris 23. chattanooga 11. lyon 24. nashville 12. london 25. san francisco 13. stockholm 4. results 4.1. strategic analysis the result of the strategic analysis was a swot matrix that describes the most important internal and external factors of the company, focusing in the electric transition. in figure 2 the results of the internal analysis (strengths and weaknesses) are shown with the most important found points in documentation analysis, interviews and visits. the results of the external analysis (opportunities and threats) are also shown in figure 2 with the main pestel factors. with the swot analysis, 4 strategic lines have been defined: attack (take advantage of the opportunities with the strengths), defence (face the threats with the strengths), reorientation (correct weaknesses with the opportunities) and survival (face the threats that affect the weaknesses). electrification of madrid fleet public transport company (emt-madrud): strategic analysis and implementation. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain and external factors of the company, focusing in the electric transition. in figure 2 the results of the internal analysis (strengths and weaknesses) are shown with the most important found points in documentation analysis, interviews and visits. the results of the external analysis (opportunities and threats) are also shown in figure 2 with the main pestel factors. with the swot analysis, 4 strategic lines have been defined: attack (take advantage of the opportunities with the strengths), defence (face the threats with the strengths), reorientation (correct weaknesses with the opportunities) and survival (face the threats that affect the weaknesses). 4.2 benchmark the goal of the benchmarking analysis was to review the situation of the public transport in other cities of the world and their plans for the future and strategies to achieve them. figure 2. swot analysis of the emt figure 2. swot analysis of the emt. int. j. prod. manag. eng. (2019) 7(special issue), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international electrification of madrid fleet public transport company (emt-madrid): strategic analysis and implementation 111 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.2. benchmark the goal of the benchmarking analysis was to review the situation of the public transport in other cities of the world and their plans for the future and strategies to achieve them. 25 cities have been selected for this analysis, choosing with the criteria of similarities with madrid (population, pollution, economy, fleet…) or innovation in the transport (special buses, new systems, alternative management solutions…). a heat matrix has been created, showing graphically the different fleets that each city is using nowadays (diesel, gas, hybrid, electric and hydrogen), see figure 3, and its associated legend. buses cities hydrogen electric hybrid gnc/glp/ biogas diesel/ biodiesel san francisco 0 5 4 0 0 amsterdam 0 5 0 0 5 london 1 3 5 0 5 bilbao 0 2 3 0 5 san sebastián 0 2 3 0 5 hong kong 0 2 3 0 5 beijing 0 2 3 0 5 madrid 0 2 2 5 5 paris 0 2 0 5 5 rome 0 2 0 0 5 shanghai 0 2 0 0 5 nashville 0 2 0 0 5 barcelona 0 1 4 5 5 berlin 1 1 0 0 5 milano 1 1 0 0 5 genève 0 1 2 0 5 gothenburg 0 1 1 5 5 sao paulo 0 1 0 5 5 tokyo 0 1 0 3 5 tel-aviv 0 1 0 0 5 bremen 0 1 0 0 5 lyon 0 1 0 0 5 chattanooga 0 1 0 0 5 valladolid 0 0 2 5 5 stockholm 0 0 1 4 5 brussels 0 0 0 0 5 0 1 2 3 4 5 0% <0,5% <5% <15% <35% >35% figure 3. heat map on the fleets used in each city. higher numbers (darker colors) represent higher percentages of the fleet. the order of the cities in the table can be read as higher or lower sustainable fleet, being those more sustainable located in the upper part of the table. another heat matrix (see figure 4) has also been created to represent the opportunity electric chargers of each city (trolleybuses, magnetic induction and pantograph). the order of the cities in this table also shows those that are better equipped to leverage sustainable mobility, being the best those located in the upper part of the table. opportunity chargers cities induction pantograph trolleybuses london 1 0 0 madrid 1 0 0 berlin 1 0 0 stockholm 1 1 0 amsterdam 0 2 0 shanghai 0 2 0 barcelona 0 1 0 genève 0 1 5 gothenburg 0 1 0 tel-aviv 0 1 0 san francisco 0 0 5 milano 0 0 3 rome 0 0 2 bilbao 0 0 0 san sebastián 0 0 0 hong kong 0 0 0 beijing 0 0 0 paris 0 0 0 nashville 0 0 0 sao paulo 0 0 0 tokyo 0 0 0 bremen 0 0 0 lyon 0 0 0 chattanooga 0 0 0 valladolid 0 1 0 brussels 0 0 0 0 1 2 3 4 5 0% <0,5% <5% <15% <35% >35% figure 4. heat map of electric infrastructure. most of the companies have the highest percent of their fleet formed by diesel vehicles, because it has int. j. prod. manag. eng. (2019) 7(special issue), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international garcía hernanz et al. 112 http://creativecommons.org/licenses/by-nc-nd/4.0/ been the traditional fuel used by buses. but it exists a general tendency to electrify the fleets in the most relevant cities of the world to fight against the air quality problems. many actions and implementations in the public transport field have been found in this analysis, also similarities and correlations between cities. with this information, combined with the strategic analysis of the emt-madrid, the decision tree was made. 4.3. decision tree the decision tree is a model that can guide and support other transport companies and city councils of the world to decide which transport fleet is the best to implement in their cities depending on their necessities and resources. the parameters that this model combines are buses, infrastructure, emissions and costs and investments. part of the outcome of the decision tree (zoom to the first questions) can be seen in figure 5. the full tree is provided as an annex to the paper. in this model, it has been considered that the objective of the cities is to obtain the least polluting fleet within the economic possibilities of the city and trying to reuse the existing infrastructure. also, that the priority of the cities is the reduction of nox and of the particles before co2. with options of emissions and investments of the same order of magnitude, preference is given to the option with the simplest logistics. in this model, hydrogen buses have been considered as the best option, since the logistics of their use is similar to diesel and natural gas buses, and hydrogen vehicles do not have any type of emissions. the problem of hydrogen is its difficult access for most cities, so the first question is if the city has the possibility of easy access to hydrogen at low cost. the second best option is the trolleybuses, since they can work in electric mode for long periods of time, as they obtain electricity directly from the network that can charge their battery to operate when they need to travel in places without tram infrastructure. in the cases in which they must operate during very long periods of time outside the electric line, they should be hybrids, preferably of cng. the main problem with trolleybuses is their infrastructure, since it goes through the surface of all or a large part of the line and has a very high cost and visual impact. for this reason, the second question is if the city already has easy access to inexpensive hydrogen? does the city have a tram network? hydrogen bus … no yes no possibility of high investments? trolleybus yes no electric bus easy access to natural gas? yes notrolleybuses leaving the network for long periods? yes no hybrid trolleybus electric trolleybus yes … … figure 5. first questions of the decision tree. int. j. prod. manag. eng. (2019) 7(special issue), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international electrification of madrid fleet public transport company (emt-madrid): strategic analysis and implementation 113 http://creativecommons.org/licenses/by-nc-nd/4.0/ the trams infrastructure, because, if not, it is not a good alternative. if this infrastructure already exists, its use should be prioritized. electric vehicles are the best option after hydrogen vehicles and trolleybuses, as they do not have emissions. for its proper functioning, a large investment must be made in the recharging infrastructure and buses so that they have long duration batteries. for this reason, the question is asked whether a high investment in vehicles and infrastructure can be made. the next question, within the electric buses, is whether the necessary time of operation before returning to the operational centre to recharge is very long. if this time is longer than the duration of the batteries, opportunity charging systems are required in addition to the charge in the operational centre. within systems of opportunity charge there are magnetic induction and pantograph. magnetic induction systems require a higher investment, but the infrastructure is more protected and generates less visual impact. inside the pantographs, there are the exterior ones, placed in the tower that charge the bus or pantographs placed in the bus that are connected to the charging tower. those placed in the tower have a higher infrastructure cost, while those placed in buses have a higher cost of the vehicle. the same happens in the operational centers, where pantographs or plug chargers can be placed, depending on the charging time that the bus needs before starting the service. if it is not possible to invest in electric buses, the solution to reduce local emissions is compressed natural gas buses. if the city can also have electric chargers, the best option is the gnc plug-in hybrid buses. if it is possible to invest in cng vehicles, but not in electrical infrastructure, the least polluting vehicles are the gnc non-pluggable hybrid vehicles, although they cost more than cng vehicles. if the city cannot invest in cng vehicles, but in electric chargers, the best option would be the hybrid diesel plug-in buses, because of their possibility of operating in electric mode recharging their batteries. if the city cannot make any type of investment in infrastructure, the best option is the hybrid diesel non-plug buses, since they have less consumption and emissions. 5. conclusions 5.1. implications for public institutions the result of the analysis supports the proposal of switching to a 100% electric fleet at emt-madrid. however, the large investments and change costs required call for a gradual implementation. to this end, new operational centers need to be built. alternatively, adaptation of the existing ones and, in some cases, implementing opportunity charging structures, such as pantographs and magnetic induction chargers on the bus lines, will be needed. the employees of the company should be trained to work with this new electrical technology, especially the mechanics, since electric engines and batteries are completely different to diesel and natural gas engines. these needs have been detected for the city of madrid, but can however, be of use for other cities facing similar problems. moreover, a decision tree that can guide and support other transport companies and city councils of the world to decide which transport fleet is the best to implement in their cities depending on their necessities and resources has been created. this study is the detonator of a process that will reshape the emission of pollutants within the city of madrid. insights for the development of a new strategic plan have been given, which conclude with the need of creating a 0 emissions fleet for the company in the near future. this will influence private users, which will move to this type of mobility as well. to this end, attention should be paid to the three key points to focus in the implementation of the electric buses, namely the fleet, the infrastructure and the human resources. they should be perfectly coordinated during all the transition to ensure a proper change. investing in sustainable mobility is investing in the future, improving health and life quality of the citizens, and building a better city and a better world for future generations. int. j. prod. manag. eng. (2019) 7(special issue), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international garcía hernanz et al. 114 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5.2. limitations and avenues for further research this study is not without limitations. first, the final recommendations are extracted from a limited number of observations. second, it is not a longitudinal study, and for this reason, the analyses have been made for a particular moment in time. third, the cost analysis has not been included. these limitations open, in fact, possibilities for further research. for instance, an economic study of each type of electric fleet and associated infrastructure could be conducted, with particular emphasis on investment requirements and cost of operations, maintenance and training of workers. likewise, an implementation plan for the electric fleet in madrid, with simulation of different scenarios could be performed, with the aim of assessing which bus lines should be prioritized for the use of new electric buses. references airparif (paris regional air observatory) (2017) air quality in paris region. summary 2016. https://www.airparif.asso.fr/_pdf/publications/ bilan-2016-anglais170830.pdf retrieved march 11th, 2019. ayuntamiento de madrid. (2017). plan de calidad de aire de la ciudad de madrid y cambio climático. ayuntamiento de madrid, (2019). episodios de contaminación. https://www.madrid.es/portales/munimadrid/es/inicio/medidas-especialesde-movilidad/protocolo-de-contaminacion/episodios-de-contaminacion/ retrieved march 11th, 2019. collins, m., knutti, r., arblaster, j., dufresne, j.l., fichefet, t., friedlingstein, p., gao, x., gutowski, w.j., johns, t., krinner, g., shongwe, m., tebaldi, c., weaver, a.j., and wehner, m. 2013: longterm climate change: projections, commitments and irreversibility. in: climate change 2013: the physical science basis. contribution of working group i to the fifth assessment report of the intergovernmental panel on climate change [stocker, t.f., d. qin, g. k. plattner, m. tignor, s.k. allen, j. boschung, a. nauels, y. xia, v. bex and p.m. midgley (eds.)]. cambridge university press, cambridge, united kingdom and new york, ny, usa. dans, e., and seisdedos, g. (2016). upgrading urban mobility. los retos de la movilidad urbana. ie business school. empresa municipal de transportes de madrid (emt). (2017). plan estratégico cerca 2017-2020. foster, n. (1st june 2017). “future is now” forum, madrid. grigoratos, t., and martini, g. (2014). brake wear particle emissions: a review. environmental science and pollution research international, 22(4), 2491-2504. ministerio de la presidencia. gobierno de españa. (29th january 2011). boletín oficial del estado. real decreto 102/2011, de 28 de enero, relativo a la mejora de la calidad del aire. ministerio de energía, turismo y agenda digital. gobierno de españa. (2016). la energía en españa. porter, m.e. (1998). competitive advantage: creating and sustaining superior performance. ed. free press. ocampo, l., and ocampo, c. o. (2015) ‘a robust evaluation of sustainability initiatives with analytic network process (anp)’, international journal of production management and engineering, 3(2), 123. https://doi.org/10.4995/ijpme.2015.3595. ros-mcdonnell, l., de-la-fuente-aragon, m.v., ros-mcdonnell, d., and cardós carboneras, m. (2018) ‘designing an environmental zone in a mediterranean city to support city logistics’, international journal of production management and engineering, 6(1), 1. https://doi.org/10.4995/ijpme.2018.8771. vernadat, f. (2014) ‘enterprise modeling in the context of enterprise engineering: state of the art and outlook’, international journal of production management and engineering, 2(2), 57. https://doi.org/10.4995/ijpme.2014.2326. united nations. (1998). kyoto protocol to the united nations framework convention on climate change. world health organization (who). (2005). air quality guidelines global update 2005. int. j. prod. manag. eng. (2019) 7(special issue), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international electrification of madrid fleet public transport company (emt-madrid): strategic analysis and implementation 115 https://www.airparif.asso.fr/_pdf/publications/bilan-2016-anglais170830.pdf https://www.airparif.asso.fr/_pdf/publications/bilan-2016-anglais170830.pdf https://www.madrid.es/portales/munimadrid/es/inicio/medidas-especiales-de-movilidad/protocolo-de-contaminacion/episodios-de-contaminacion https://www.madrid.es/portales/munimadrid/es/inicio/medidas-especiales-de-movilidad/protocolo-de-contaminacion/episodios-de-contaminacion https://doi.org/10.4995/ijpme.2015.3595 https://doi.org/10.4995/ijpme.2018.8771 https://doi.org/10.4995/ijpme.2014.2326 http://creativecommons.org/licenses/by-nc-nd/4.0/ e as y ac ce ss to in ex pe ns iv e hy dr og en ? d oe s o pe ra ti ng fo r lo ng pe ri od s? th e ci ty ha ve a tr am ne tw or k? h yd ro ge n b us n o p os si bi lit y of h ig h in ve st m en ts ? t ro lle yb us y e s n o e le ct ri c b us e as y ac ce ss to na tu ra l g as ? y e s n o y e s n o t ro lle yb us es le av in g th e ne tw or k fo r lo ng pe ri od s ? y e s n o n o h yb ri d t ro lle yb us e le ct ri c t ro lle yb us p os si bi lit y of in ve st in g in el ec tr ic in fr as tr ut ur e p os si bi lit y of in ve st in g in e le ct ri c in fr as tr ut ur e n o h yb ri d b us h yb ri d c ha rg ea bl e b us h yb ri d g n c c ha rg ea bl e b us c ha rg in g in op er at io na l c en te r o pp or tu ni ty ch ar gi ng g n c b us y e s n o y e s y e s y e s n o m ag ne ti c in du cc ti on p an to gr ap h ch ar gi ng y e s n o p os si bi lit y of h ig h in ve st m en ts ? annex complete version of the decision tree int. j. prod. manag. eng. (2019) 7(special issue), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international garcía hernanz et al. 116 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2019.10812 received 2018-10-13 accepted: 2019-01-20 profiles of human capital and strategic technological decisions on turbulence environment pérez, d.a1, sáiz-bárcena, l.a2, manzanedo, m.a.a3 and pérez, a.a4 adepartment of civil engineering. area of business organisation. higher polytechnic school university of burgos. campus río vena (edificio a), avda. cantabria s/n 09006-burgos, spain a1 dperezmiguel@gmail.com, a2 lsaiz@ubu.es, a3 mmanz@ubu.es, a4 arrturopm@gmail.com abstract: the objective of this article is to examine the absorptive capacity in the technology industry and aspires to recognize how firms can manage their strategic decisions in the turbulence contexts. in particular, we examine how organizations can strengthen their organizational contexts in order to absorb knowledge. from the knowledge management literature, this investigation extends our perception of the relationship between the human capital profiles (organization, research and development unit, and recent incorporations) and technological decision-making. through the sepi foundation, a balanced panel of 1,220 spanish industrial companies has used that answer to the survey of business strategies (sbs) for a threeyear period, which signifies a total of 3,660 cases. the principal finding is the presence of high levels of human resources to understand a decision efficiency process. it also highlights its relationship to the firm’s technological committee. these contributions are notable for both researchers and practitioners. it could be stimulating to expand the study to the association between human capital profiles and other strategic technological decisions, as the preparation of an innovation plan or the measurement of innovation performance. key words: knowledge management, absorptive capacity, turbulence environment, human capital, innovation. 1. introduction the dynamism in the modern environments complicates, in large measure, the maintenance of the competitiveness. organizations that do not possess the necessary source of knowledge to innovate resort to external sources (spithoven et al., 2011; chesbrough, 2012) and must resolve how to renew their knowledge with the purpose of updating their innovative potential (wang and han, 2011). in this sense, absorptive capacity is highly connected with the firm’s competitive environment. this is because, during knowledge exploration, the organization interacts with its environment and brings in potentially useful knowledge (aribi and dupouët, 2016). to the extent, the environment is more dynamic and unstable the validity of useful knowledge will be less and the organization will bear greater difficulties to take advantage. in the first place, being able to find it quickly and, secondly, being able to assimilate it properly. a firm’s absorptive capacity does not simply depend on the organization’s direct interface with the external environment (cohen and levinthal, 1990). these authors suggest that it also depends on knowledge transfers across and within subunits. during the transformation, in order to gain access to the necessary complementary knowledge in the environment, firms must set specific channels to external knowledge sources (aribi and dupouët, to cite this article: pérez, d., sáiz-bárcena, l., manzanedo, m.a. and pérez, a. (2019). profiles of human capital and strategic technological decisions on turbulence environment. international journal of production management and engineering, 7(1), 39-47. https://doi.org/10.4995/ijpme.2019.10812 int. j. prod. manag. eng. (2019) 7(1), 39-47creative commons attribution-noncommercial-noderivatives 4.0 international 39 mailto:dperezmiguel@gmail.com mailto:lsaiz@ubu.es http://creativecommons.org/licenses/by-nc-nd/4.0/ 2016). the environmental turbulence is a dynamic, unpredictable, expanding, fluctuating environment (khandwalla, 1977). under high rate of market turbulence, orientation towards learning facilitates firms to increase organizational innovativeness and performance (baba et al., 2017). therefore, the role of human capital is critical to organizations in turbulence contexts. there is a lack of research on the absorptive capacity literature (cohen and levinthal, 1990; zahra and george, 2002; todorova and durisin, 2007; volberda et al., 2010; roberts et al., 2012) in the turbulence conditions. furthermore, researchers of absorptive capacity recommend it, such as sun (2010), datta (2012), aribi and dupouët (2016). to date, few studies explained absorptive capacity and its effects on a company in dynamics markets. for example, according to the study conducted by santoro and gopalakrishnan (2015), it shows the most technologically uncertain environments are associated positively with the implementation of external knowledge, as a component of the absorptive capacity. to fill the aforementioned gaps, the present study seeks to answer the following three questions. the primary purpose of this research is the discovery of the relationship between the human capital and the strategic technological decisions in turbulence environment. the second motivation is driven by the next question: ‘‘the superior levels of human capital in the technology industry explains the presence of technological committee in the firm’’. the third intention is the discovery of the human capital profile more related with this internal capacity to absorb knowledge. the article presents the following structure. first, the authors emphasize especially on the link among human capital, absorptive capacity and technological decisions, providing an appropriate working framework. second, the empirical work is presented. third, the debate is focused on the human capital’s profiles in the technological and turbulence situations. finally, the more relevant implications and conclusions complete this document. 2. human capital and technological decisions employees perceive knowledge as the accumulation of successes and failures; without the existence of errors, the possibility of learning is also limited (garcía et al., 2014). while the learning is an accumulative process without end, the valuable absorptive capacity depends on the establishment of goals. the firm’s absorptive capacity ceases when the objective is satisfied and thus more learning is not necessary (pérez-miguel et al., 2017). knowledge development through the learning process is fundamental to the emergence of new production techniques (tortorella et al., 2015). furthermore, during turbulent organizational learning predict innovativeness (baba et al., 2017). when the complexity of knowledge increases, as in situations of turbulent and dynamic environments, more qualified employees are needed to enable the knowledge management processes. when the company has more people, related, you might exploit more absorptive capacity. a firm’s absorptive capacity is not simply the sum of the absorptive capacities of its employees (cohen and levinthal, 1990). the number of people, in principle, does not have to transcend to the results of the absorptive capacity. the organizations aware of it providing the means to manage their potential. in fact, they can decide on the employees training so that they are able to communicate new knowledge of the different external sources, as well as to recognize intangible ones. the companies can also form a managers committee to administer it, as an area and specific tasks. as well as interview workers, highly qualified human resources, where the value of the executive team is increased because the information which is absorbing them is more valuable. to integrate this information, and share it, the organization will provide knowledge to better decisions. furthermore, as the technicians are focused, in many cases, with technological issues, this absorbed knowledge, similarly, will allow the best technological solutions. to recognize the foundations of a company’s absorptive capacity, we center the form of communication between the environment and the firm, as well as among the subunits of the organization, and also on the character and distribution of expertise within the organization (cohen and levinthal, 1990). in this point, successful company decisions depend on knowledge transfer. thus, tacit knowledge and face-to-face interactions help to dissipate the ambiguity of codified knowledge (grandinetti and tabacco, 2015). this kind of knowledge allows companies to engage in a number of innovation projects. additionally, the effects of human capital int. j. prod. manag. eng. (2019) 7(1), 39-47 creative commons attribution-noncommercial-noderivatives 4.0 international pérez, d., sáiz-bárcena, l., manzanedo, m.a. and pérez, a. 40 http://creativecommons.org/licenses/by-nc-nd/4.0/ on innovation performance varies depending on the levels of technology (buenechea-elberdin et al., 2017). the deficiency of a developed technological base is likely to have difficulties to absorb the technologies (tsai and wang, 2009). in a study of cases, aribi and dupouët (2016) reveal the coordination role of a technology committee or small group of managers specifically gathered for the follow-up of the innovation activities. in particular, groups are more probable to apply an individual member’s distributed knowledge when a majority of teammates are aware of the individual’s expertise (baumann & bonner, 2013), and that teams are more likely to incorporate their distributed information when expertise knowledge is concentrated in one member than when it is disseminated homogeneously among members (mell et al., 2014). thus, when top managers work nearly with engineers or other qualified employees, they understand the knowledge needs and adopt better technological decisions. therefore, the percentage of engineers in the workforce can be associated with the business strategic decisions and, therefore, greater link with the technologies management. innovation performance needs the valuable role of doctors and research technicians. these professionals are trained for complex activities and, apparently, have to be more associated to take complicated decisions, thus, their presence can be more associated with turbulence contexts because organizations require workers with greater autonomy. these functions imply that the qualified workers’ absorptive capacity is out of control. therefore, the greater number of highly qualified personnel, the company interest will be greater for manage appropriately this absorptive capacity to integrate it in the organizational decision structure. in knowledge area, the informational resources may add to the quality of team decision making and innovation (van knippenberg and mell, 2016). additionally, the organizations use their positions of technological vigilance and management to form their capabilities, and subsequently, to enhance innovation outcomes (pérez-miguel et al., 2018). based on previous literature, we present the first hypothesis: h1: high levels of human capital above the industry average in the firm, which belongs to a technological industry and competes in the turbulence environment, are significant and positively related with the presence of the firm’s technology committee. assimilation capability represents the organization’s routines and processes that permit it to analyze, process, interpret and understand the information obtained from external sources (szulanski, 1996). along with the exploration capacity, it is associated with the potential absorptive capacity (jansen et al., 2005). zahra and george (2002:188) indicate that the “organizational capabilities of knowledge acquisition, assimilation, transformation, and exploitation build on each other to yield absorptive capacity, a dynamic capability that influences the firm’s ability to create and deploy the knowledge necessary to build other organizational capabilities”. in particular, these researchers denote to the key component of knowledge assimilation is the understanding and the specific roles are interpretation, comprehension, and learning. thus, the elements of knowledge assimilation, as a cognitive process, are conditioned by the mind of every individual. firms that want to adjust to a fluctuating world, employ those aspects that affect the absorptive capacity that is identified to influence in the breadth of knowledge inside the organization, preferably related to diversity, and those that establish links across the boundaries of the company (van winkelen and mckenzie, 2008). the literature of absorptive capacity is based on the existence of research and development (r&d) internal activities (cohen and levinthal, 1990; todorova and durisin, 2007; volberda et al., 2010) which take advantage of the external sources of knowledge to improve the innovation. despite the dependence on external partners, organizations maintain its investment in r&d with the aim of generating new knowledge internally and to build the absorptive capacity for the follow-up of activities outside of its limits (dahlander and gann, 2010). therefore, the role of internal r&d function within the firm is crucial to the effective absorption process. companies with this competition look constantly to the outside to capture knowledge that can influence in the r&d projects. innovation highlights properly the specific role of doctors and research technicians. these groups are trained for reaching research activities and, presumably, have to be more associated to act implicitly in favor of a higher absorptive capacity. for example, technological vigilance functions int. j. prod. manag. eng. (2019) 7(1), 39-47creative commons attribution-noncommercial-noderivatives 4.0 international profiles of human capital and strategic technological decisions on turbulence environment 41 http://creativecommons.org/licenses/by-nc-nd/4.0/ and knowledge exploration. thus, the accumulated experience of workers assigned to this r&d function will be valuable and will result in higher absorption levels. furthermore, the accumulated knowledge of qualified workers will result in higher absorption levels. therefore, the greater number of highly qualified personnel in the r&d unit, the company interest will be greater to manage appropriately the absorptive capacity. based on previous literature, we present the second hypothesis: h2: high levels of human capital above the industry average in the r&d unit are significant and positively related with the presence of the firm’s technology management. the literature on knowledge management considers the prior knowledge as an outstanding antecedent of the absorptive capacity at the organizational level (cohen and levinthal, 1990; zahra and george, 2002; volberda et al., 2010; batarseh et al., 2017) and the diversity of these knowledge of the absorptive capacity at the individual level (lowik et al., 2017), which are achieved through the professional experiences, education and life experiences. the capacity to assimilate knowledge is a function of the richness of the pre-existing knowledge structure (cohen and levinthal, 1990). the previous knowledge related influence the absorptive capacity and are fed at the same time to absorb knowledge (nonaka and takeuchi, 1995). in the highly qualified personnel may be more present the bisociative and associative cognitive styles (lowik et al., 2017), that are critical to develop a strong capacity to codification of knowledge that allows them: (1) to absorb important knowledge from outside the company; and (2) to transform the knowledge absorbed so that it can be used, recodified or tacit, in the activities and internal services (grandinetti, 2011). the absorption is virtually a necessary part of any company’s start-up phase when teams exploit their social capital to mitigate any knowledge gap that organizations frequently experience at birth (grandinetti, 2016). in the other side, buenecheaelberdin et al. (2017) strengthen that the effect of qualified personnel, learning and entrepreneurship on innovation varies depending on the levels of technology. when companies lack a sufficiently developed technological base, they are likely to have difficulties to absorb the technologies of recruitment market (tsai and wang, 2009). the learning outcome is greater when the object of learning is related to what is already known, consequently, learning is more difficult in novel domains (cohen and levinthal, 1990). the transformative learning process enhances assimilation and transformative capabilities (rezaei and darwish, 2016). in this sense, knowledge assimilation is highly dependent on the complexity of the knowledge (batarseh et al., 2017). as the complexity levels increases, more qualified workers are required to ease this assimilation process. the higher skill levels or years of experience, so related, the higher prior knowledge levels (buenechea-elberdin et al., 2017). the active role of the top management can facilitate the knowledge assimilation and integration. thus, we are interested in when managers are working with human resource practices to achieve the above situations. that is when organizations are improving their levels of training and experience, among others, through the recruitment of highly qualified personnel. taking into account these deliberations, the subsequent hypothesis can be specified as: h3: the inputs of superior human capital is significantly and positively associated with the firm’s technology management. for contributing to the research, figure 1 displays the planned model and we can see the organization’s human resources in a technological industry to explicate the firm’s technological decisions in a turbulence environment. figure 1. model of human capital profiles for the firm’s strategic technological decisions (source: own elaboration). int. j. prod. manag. eng. (2019) 7(1), 39-47 creative commons attribution-noncommercial-noderivatives 4.0 international pérez, d., sáiz-bárcena, l., manzanedo, m.a. and pérez, a. 42 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3. research methodology 3.1. data and sample with the goal of contrasting the earlier planned hypotheses, an empirical study was arranged on the basis of the spanish industrial companies that answered to the survey of business strategies (sbs). the sbs has been realized by the sepi foundation in relationship with the spanish ministry of science and technology and aspires to investigate the strategic progression of spanish manufacturing firms. this survey shows an unbalanced panel since some entities conclude to provide data while others persist to do so every year. especially, one of the points that distinguish the sbs from other data sources is its explicit objective of generating information with a panel structure (sáiz et al., 2018). all the data integrated into the sbs is exposed to corroboration and reliability controls. to carry out this study, a sample is included of a balanced panel of 5,566 companies. there were counted exclusively those firms that have consistently replied in the exploration period, 2010-2015 to establish a sample of the complete panel of firms. accordingly, 1,220 organizations for a three-year period was utilized, which signifies a total of 3,660 cases. the turbulence environment accounted for 694 cases and the medium and high technology sectors involved a total of 226 observations. the panel distribution is shown in table 1. with a confidence level of 95%, the sampling error was 0.245. table 1. panel distribution for spanish firms in the medium and high technology sectors (2010-2015). medium and high technology sectors turbulence context stable context chemical industry 61 204 agricultural and industrial machines 51 171 computer, electronic and optical products 14 46 machinery and electrical apparatus 28 113 motor vehicles 54 132 other transport material 18 45 total 226 731 3.2. variables the variables employed insights from prior literature and follow principally diaz-diaz and de saá (2014) and sáiz et al. (2018). this section defines how concepts were determined from the sbs questionnaire. the dependent variable is the technological committee. this measure takes the value 1 if the company replies affirmatively about the existence of a firm’s technology committee, and 0 if it does not. explanatory variables. the human capital is measured by three variables. the first variable includes the superior levels of engineers and graduates in the r&d unit. the following constructs reflect the firm’s superior levels of engineers and graduates and the recent incorporation of these qualified workers. selection variable: turbulence environment, which takes the value of 1 if the firm answers affirmatively to the question about expanding markets, and 0 if it does not. control variables. these constructs are: size calculated as the logarithm of the number of workers. age, as measured the number of years since the company’s founding and the firm’s return on assets. location calculated as a dummy variable, which takes the value of 1 if the company responds affirmatively to the question about national or international markets. 3.3. results our dependent variable is dichotomous or binary. for this cause, to test the supposition, a regression binomial logical model examines the relationship between the human capital measure and the firm’s technological committee. the planned model was assessed using the econometric package spss (statistical product and service solutions) version 24 for windows. the researchers select this statistical program for its applicability and ease in handling. the statistical, expose in table 2 to table 5, highlights the suitability and good fit of the regressions completed to estimate the model for the technological industry. first, the chi-square statistic is significant (p<0.001) and denotes that the model achieves a significant increase compared with the null model. second, the r2 of cox and snell and the r2 of nagelkerke, as guidance indicators, inform on the goodness of fit. in this instance, these statistics get high values of 0.449 and 0.599 respectively in table 3. third, the model can be differentiated int. j. prod. manag. eng. (2019) 7(1), 39-47creative commons attribution-noncommercial-noderivatives 4.0 international profiles of human capital and strategic technological decisions on turbulence environment 43 http://creativecommons.org/licenses/by-nc-nd/4.0/ among the different selection groups. the correct classification percentage of observations is high for the technology committee presence, for example, more than 86% in table 3. concerning control variables, results show that the organization’s size has positive and significant effects on the dependent variable. moreover, the age, roa and geographical location have no significant effects on the technological committee in the turbulence environment. with the intention of contrasting hypothesis h1 associated with the first measure of human capital, specifically, the availability of engineers and graduates in the company above the industry average. the results of table 2 indicate a positive and strong significant influence of the human capital (organization) on firm’s technological management (p<0.001; wald, 19.701) in technological industry and turbulence environmental, in order to test h1. with the intention of testing hypothesis h2 related with the second measure of human capital, in particular, the presence of engineers and graduates in the r&d unit above the industry average. table 3 reveals a positive and strong significant influence of the human capital (r&d unit) on the dependent variable (p<0.001; wald, 65.157) in technological industry and turbulence context, in order to test h2. finally, with the aim of testing hypothesis h3 related to the third measure of human capital, especially, the recent incorporation of engineers and graduates in the organization. the results of table 4 denote a positive and weak significant effect of the human capital (inputs) on firm’s technological management (p<0.1; wald, 3.002) in technological industry and turbulence environment, in order to test h3. to drive general conclusions, in table 5, we depict the evolution of the significance of the three human capital profiles together. we can see that the human capital of the r&d unit (p<0.001; wald, 56.489) globally concentred the explication despite the variables entry. furthermore, when this regression obtains the best results in all statistics examined, such as the highest r2 of nagelkerke (0.610) or the minor logarithm of the likelihood (167.854) which, in turn, indicate a better adjustment of the model. in synthesis, according to the empirical results showed for the observations of our complete panel of spanish industrial companies, in particular, for the technological industry in the turbulence environment, measures of human capital, such as elements in the assimilation phase of the absorptive capacity, are robustly related to the organizational ability to adopt new strategic decisions (technological committee). and it is resolved that the superior level of the human capital is relevant to understand the firm’s strategic table 3. binomial logical regression of human capital (r&d unit) on the firm’s technological committee. variables b (wald) (constant) -1.656 (1.809) size 1.604**** (15.374) age 0.010 (0.958) roa -222.377 (0.088) location 0.300 (0.203) human capital (r&d unit) 3.353**** (65.157) r2: cox and snell 0.449 r2: nagelkerke 0.599 chi-quadrate 129.14**** -2 logarithm of the likelihood 171.119 n 226 % correct division 86.6% **** p < 0.001; *** p < 0.01; ** p < 0.05;* p < 0.1 table 2. binomial logical regression of human capital (organization) on the firm’s technological committee. variables b (wald) (constant) -3.079*** (9.378) size 1.449**** (19.406) age 0.009 (1.621) roa -534.915 (1.011) location -0.415 (0.589) human capital (organization) 1.434**** (19.701) r2: cox and snell 0.215 r2: nagelkerke 0.287 chi-quadrate 53.925**** -2 logarithm of the likelihood 254.676 n 226 % correct division 72.2% **** p <0.001; *** p < 0.01; ** p < 0.05;* p < 0.1 int. j. prod. manag. eng. (2019) 7(1), 39-47 creative commons attribution-noncommercial-noderivatives 4.0 international pérez, d., sáiz-bárcena, l., manzanedo, m.a. and pérez, a. 44 http://creativecommons.org/licenses/by-nc-nd/4.0/ technological decisions. finally, it is concluded that the most important profile is the human capital of the r&d unit. second, the firm’s human capital and, third, the inputs of new human capital. the contributions of this work should be noticed as it is founded on panel data models and are coherent with prior studies, associated with the cooperation activities and absorption process (sáiz et al., 2018), the technology alliance diversity (lucena and roper, 2016), the r&d function (sánchez et al., 2013), the firm’s innovation (segarra-ciprés et al., 2012), the size of the company (revilla and fernández, 2012) and the technological knowledge assets (díaz-díaz et al., 2006). 4. conclusion and limitations this research exposes relevant empirical consequences that make a principal support to explain the issue about the presence of technological committee in turbulence environments. when the companies have an average above the industry of engineers and graduates, above all, in the r&d unit, assume that they have the adequate level of human resources to organize around the technology committee, and optimize their processes of knowledge assimilation. with regard to the use of engineers and graduates, the paper findings confirm that the relationship between superior levels of the human capital and firm’s technological committee in technological sectors on turbulence environmental, as stated h1, h2, and h3. from the knowledge management literature, the absorptive capacity contributes to valuable explications to the technological decisions and arises as one of the issues most relevant to the r&d function. this capacity is determined by the needs of knowledge to be absorbed and can be used to adopt investment decisions in r&d or alliances. in addition, a good knowledge assimilation structure is critical to complete the knowledge absorption. this research finds that those companies that maintain levels in qualified workers (engineers and graduates) higher than the industry average in the r&d unit or in the firm, or newly incorporated, are more likely to deploy the technology committee, which involves an internal capacity to absorb and supports the knowledge absorption process. in other words, in the dynamic and technological contexts, this investigation reveals that firms that assimilate more knowledge are more prepared to manage it. table 4. binomial logical regression of human capital (inputs) on the firm’s technological committee. variables b (wald) (constant) -2.351** (5.027) size 1.059*** (10.967) age 0.009 (1.490) roa -642.608 (1.980) location 0.252 (0.235) human capital (inputs) 0.533* (3.002) r2: cox and snell 0.147 r2: nagelkerke 0.196 chi-quadrate 35.442**** -2 logarithm of the likelihood 273.159 n 226 % correct division 63.2% **** p < 0.001; *** p < 0.01; ** p < 0.05;* p < 0.1 table 5. binomial logical regression of human capital profiles on the firm’s technological committee. variables b (wald) (constant) -1.876 (1.797) size 1.732**** (14.452) age 0.010 (1.177) roa -116.315 (0.019) location 0.188 (0.079) human capital (organization) 0.768* (3.261) human capital (r&d unit) 3.203**** (56.489) human capital (inputs) -0.013 (0.001) r2: cox and snell 0.457 r2: nagelkerke 0.610 chi-quadrate 132.41**** -2 logarithm of the likelihood 167.854 n 226 % correct division 84.8% **** p < 0.001; *** p < 0.01; ** p < 0.05;* p < 0.1 int. j. prod. manag. eng. (2019) 7(1), 39-47creative commons attribution-noncommercial-noderivatives 4.0 international profiles of human capital and strategic technological decisions on turbulence environment 45 http://creativecommons.org/licenses/by-nc-nd/4.0/ concerning limitations, it would be stimulating to expand the work to the relation of firm’s technological committee and other questions related with the planning of an innovation strategy and the follow-up of innovation performance, examining the entire construct of firm’s technological management. furthermore, the election of external sources of knowledge represents an essential decision for the company’s innovation about the knowledge assimilation process. future research direction should also aim to investigate the profiles of human capital in the entire process of absorption to the innovation performance. there may be further evidence of the relationship among superior levels of human capital, agreements of cooperation and quantity of innovations, which are not incorporated in this study, and could be interesting to assess in future investigation. acknowledgments the authors want to acknowledge the practical support of the sepi foundation and university of burgos to facilitate access to the sbs. references aribi, a., dupouët, o. (2016). absorptive capacity: a non-linear process. knowledge management research & practice, 14(1), 15-26. https://doi.org/10.1057/kmrp.2014.17 baba, m., mahmood, r., halipah, a. (2017). the moderating role of environmental turbulence on the relationship between organizational learning and firm innovativeness. international journal of management research & review, 7(2), 148-159. batarseh, f., usher, j., daspit, j. (2017). absorptive capacity in virtual teams: examining the influence on diversity and innovation. journal of knowledge management, 21(6), 1342-1361. https://doi.org/10.1108/jkm-06-2016-0221 baumann, m.r., bonner, b.l. (2013). member awareness of expertise, information sharing, information weighting, and group decision making. small group research, 44, 532-562. https://doi.org/10.1177/1046496413494415 buenechea-elberdin, m., sáenz, j., kianto, a. (2017). exploring the role of human capital, renewal capital and entrepreneurial capital in innovation performance in high-tech and low-tech firms. knowledge management research & practice, 15(3), 369-379. https://doi.org/10.1057/s41275-017-0069-3 chesbrough, h. (2012). open innovation. where we’ve been and where we’re going. research technology management, 55(4), 20-27. https://doi.org/10.5437/08956308x5504085 cohen, m.w., levinthal, d.a. (1990). absorptive capacity: a new perspective on learning and innovation. administrative science quarterly, 35(1), 128-152. https://doi.org/10.2307/2393553 dahlander, l., gann, d.m. (2010). how open is innovation? research policy, 39(6), 699-709. https://doi.org/10.1016/j.respol.2010.01.013 datta, a. (2012). it-based knowledge capability and commercialization of innovations: modeling the impacts of ambidexterity and absorptive capacity. international journal of knowledge management, 8(3), 83-97. https://doi.org/10.4018/jkm.2012070105 díaz-díaz, n.l., de saá, p. (2014). the interaction between external and internal knowledge sources: an open innovation view. journal of knowledge management, 18(2), 430-446. https://doi.org/10.1108/jkm-07-2013-0257 díaz-díaz, n.l., aguilar, i., de saá, p. (2006). technological knowledge assets in industrial firms. r&d management, 36(2), 189-203. https://doi.org/10.1111/j.1467-9310.2006.00425.x garcía, j., maldonado, a., alvarado, a., rivera, d. (2014). human critical success factors for kaizen and its impacts in industrial performance. the international journal of advanced manufacturing technology, 70, 2187-2198. https://doi.org/10.1007/s00170-013-5445-4 grandinetti, r. (2011). local/global interfaces within industrial districts: an italian case study. the learning organization, 18(4), 301-312. https://doi.org/10.1108/09696471111132513. grandinetti, r., tabacco, r. (2015). a return to spatial proximity: combining global suppliers with local subcontractors. international journal of globalisation and small business, 7(2), 139-161. https://doi.org/10.1504/ijgsb.2015.071189 grandinetti, r. (2016). absorptive capacity and knowledge management in small and medium enterprises. knowledge management research & practice, 14(2), 159-168. https://doi.org/10.1057/kmrp.2016.2 jansen, j., van den bosch, f.,volberda, h. (2005). managing potential and realized absorptive capacity: how do organizational antecedents matter? academy of management journal, 48(6), 999-1015. https://doi.org/10.5465/amj.2005.19573106 khandwalla, p.n. the design of organizations (vol. 260). new york: harcourt brace jovanovich, 1977. lowik, s., kraaijenbrink, j., groen, a. (2017). antecedents and effects of individual absorptive capacity: a micro-foundational perspective on open innovation. journal of knowledge management, 21(6), 1319-1341. https://doi.org/10.1108/jkm-09-2016-0410 lucena, a., roper, s. (2016). absorptive capacity and ambidexterity in r&d: linking technology alliance diversity and firm innovation. european management review, 13(3), 159-178. https://doi.org/10.1111/emre.12074 int. j. prod. manag. eng. (2019) 7(1), 39-47 creative commons attribution-noncommercial-noderivatives 4.0 international pérez, d., sáiz-bárcena, l., manzanedo, m.a. and pérez, a. 46 https://doi.org/10.1057/kmrp.2014.17 https://doi.org/10.1108/jkm-06-2016-0221 https://doi.org/10.1177/1046496413494415 https://doi.org/10.1057/s41275-017-0069-3 https://doi.org/10.5437/08956308x5504085 https://doi.org/10.2307/2393553 https://doi.org/10.1016/j.respol.2010.01.013 https://doi.org/10.4018/jkm.2012070105 https://doi.org/10.1108/jkm-07-2013-0257 https://doi.org/10.1111/j.1467-9310.2006.00425.x https://doi.org/10.1007/s00170-013-5445-4 https://doi.org/10.1108/09696471111132513 https://doi.org/10.1504/ijgsb.2015.071189 https://doi.org/10.1057/kmrp.2016.2 https://doi.org/10.5465/amj.2005.19573106 https://doi.org/10.1108/jkm-09-2016-0410 https://doi.org/10.1111/emre.12074 http://creativecommons.org/licenses/by-nc-nd/4.0/ mell, j.n., van knippenberg, d., van ginkel, w.p. (2014). the catalyst effect: the impact of transactive memory system structure on team performance. academy of management journal, 57, 1154-1173. https://doi.org/10.5465/amj.2012.0589 nonaka, i., takeuchi, h. (1995). the knowledge-creating company: how japanese companies create the dynamics of innovation. new york: oxford university press. rezaei, m., darwish, t.k. (2016). antecedents of absorptive capacity: a new model for developing learning processes. the learning organization, 23(1), 77-91. https://doi.org/10.1108/tlo-04-2015-0026 pérez-miguel, d., sáiz bárcena, l., manzanedo, m.a. (2017). the knowledge absorptive capacity to improve the cooperation and innovation in the firm. 11th international conference on industrial engineering and industrial management. xxi congreso de ingeniería de organización. revilla, a.j., fernández, z. (2012). the relation between firm size and r&d productivity in different technological regimes. technovation, 32(11), 609-623. https://doi.org/10.1016/j.technovation.2012.06.004 roberts, n., galluch, p., dinger, m., grover, v. (2012). absorptive capacity and information systems research: review, synthesis, and directions for future research. mis quarterly, 36(2), 625-648. https://doi.org/10.2307/41703470 sáiz, l., pérez-miguel, d., manzanedo-del campo, m.a. (2018). the knowledge absorptive capacity to improve the cooperation and innovation in the firm. journal of industrial engineering and management, 11(2), 78-89. https://doi.org/10.3926/jiem.2505 sánchez, p., rosell, j., garcía, j.m. (2013). innovation as a driver of absorptive capacity from foreign direct investment in spanish manufacturing firms. procedia social and behavioral sciences, 75, 236-245. https://doi.org/10.1016/j.sbspro.2013.04.027 santoro, m.d., gopalakrishnan, s. (2015). alliances with universities how firms absorb and apply knowledge. international journal of knowledge management studies, 6(2), 175-199. https://doi.org/10.1504/ijkms.2015.071768 segarra-ciprés, m., bou-llusar, j.c., roca-puig, v. (2012). exploring and exploiting external knowledge: the effect of sector and firm technological intensity. innovation-management policy & practice, 14, 192-206. https://doi.org/10.5172/impp.2012.14.2.203 spithoven, a., clarysse, b., knockaert, m. (2011). building absorptive capacity to organise inbound open innovation in traditional industries. technovation, 31(1), 10-21. https://doi.org/10.1016/j.technovation.2010.10.003 sun, p. (2010). five critical knowledge management organizational themes. journal of knowledge management, 14(4), 507-523. https://doi.org/10.1108/13673271011059491 szulanski, g. (1996). exploring internal stickiness: impediments to the transfer of best practice within the firm. strategic management journal, 17(s2), 27-43. https://doi.org/10.1002/smj.4250171105 todorova, g., durisin, b. (2007). absorptive capacity: valuing a reconceptualization. academy of management review, 32(3), 774-786. https://doi.org/10.5465/amr.2007.25275513 tortorella, g.l., marodin, g.a., fogliatto, f.s., miorando, r. (2015). learning organisation and human resources management practices: an exploratory research in medium-sized enterprises undergoing a lean implementation. international journal of production research, 53(13), 3989-4000. https://doi.org/10.1080/00207543.2014.980462 tsai, k.h., wang, j.c. (2009). external technology sourcing and innovation performance in lmt sectors: an analysis based on the taiwanese technological innovation survey. research policy, 38(3), 518-526. https://doi.org/10.1016/j.respol.2008.10.007 van knippenberg, d., mell, j.n. (2016). past, present, and potential future of team diversity research: from compositional diversity to emergent diversity. organizational behavior and human decision processes, 136, 135-145. https://doi.org/10.1016/j.obhdp.2016.05.007 van winkelen, c., mckenzie, j. (2008). integrating individual and organisational learning initiatives: working across knowledge management and human resource management functional boundaries. the electronic journal of knowledge management, 5(4), 527-538. volberda, h.w., foss, n.j., lyles, m.a. (2010). absorbing the concept of absorptive capacity: how to realize its potential in the organization field. organization science, 21(4), 931-951. https://doi.org/10.1287/orsc.1090.0503 wang, c., han, y. (2011). linking properties of knowledge with innovation performance: the moderate role of absorptive capacity. journal of knowledge management, 15(5), 802-819. https://doi.org/10.1108/13673271111174339 zahra, s.a., george, g. (2002). absorptive capacity: a review, reconceptualization, and extension. academy of management review, 27(2), 185-203. https://doi.org/10.5465/amr.2002.6587995 int. j. prod. manag. eng. (2019) 7(1), 39-47creative commons attribution-noncommercial-noderivatives 4.0 international profiles of human capital and strategic technological decisions on turbulence environment 47 https://doi.org/10.5465/amj.2012.0589 https://doi.org/10.1108/tlo-04-2015-0026 https://doi.org/10.1016/j.technovation.2012.06.004 https://doi.org/10.2307/41703470 https://doi.org/10.3926/jiem.2505 https://doi.org/10.1016/j.sbspro.2013.04.027 https://doi.org/10.1504/ijkms.2015.071768 https://doi.org/10.5172/impp.2012.14.2.203 https://doi.org/10.1016/j.technovation.2010.10.003 https://doi.org/10.1108/13673271011059491 https://doi.org/10.1002/smj.4250171105 https://doi.org/10.5465/amr.2007.25275513 https://doi.org/10.1080/00207543.2014.980462 https://doi.org/10.1016/j.respol.2008.10.007 https://doi.org/10.1016/j.obhdp.2016.05.007 https://doi.org/10.1287/orsc.1090.0503 https://doi.org/10.1108/13673271111174339 https://doi.org/10.5465/amr.2002.6587995 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.13984 received: 2020-07-07 accepted: 2021-01-04 analysis and modelling of value added tax revenues on imports: some issues of application in ukraine luchko, m. a*, drozd, i. b, plutytska, k. c, ruska, r. d, vovk, i. e a department of accounting and control in public administration, ternopil national economic university, ternopil, ukraine. b educational and scientific center for public administration and financial control, institute of postgraduate education, taras shevchenko national university of kyiv, kyiv, ukraine. c department of criminal law and justice, zaporizhzhia national university, zaporizhzhia, ukraine. d department of applied mathematics, ternopil national economic university, ternopil, 46009, ukraine. e department of innovation and services management, ternopil ivan puluj national technical university, ternopil, ukraine. a m_luchko@ukr.net, b drozdirina@ukr.net, d r_ruslana@ukr.net, e vovk.ira.2010@gmail.com abstract: the aim of the article is to study the issues of analysis, modeling with the purpose of forecasting the payment of value added tax (vat) on goods, works and services imported as imports into the customs territory of ukraine. the reliability and validity of the planned vat rate depend on the assessment of the status, forecast, seasonality and trends of economic and social development. the purpose of the work is to analyze and systematize the methodology for modeling vat revenues from imports, justify the use of the econometric method and develop an adequate arima model. it application is possible in the long term as well as smaller periods of time, which is relevant for monitoring and control of tax revenues. the study revealed the main factors influencing the application of the arima model when modeling vat revenues from imports. the resulting regression model in statistica linked the variables with an accurate approximation. key words: forecasting, value added tax, tax revenue, arima modelling, regression. 1. introduction as has already been pointed out, considerable attention is paid to tax forecasting and modeling of tax revenues. in our opinion, tax forecasting, tax revenue modeling, and analysis of their results should be explored with consideration of such functions as part of management, process, and category. it has to be a continuous process, some sort of a well-balanced system. such a systematic approach means that tax revenue forecasting, modeling, has a purpose, a goal function, and consists of manageable interrelated elements. it is accepted that the purpose of tax forecasting consists of two parts: first, to understand how the future development of certain fiscal indicators depends on the future changes in the bases of these revenue sources, to assess the «budgetary elasticity» of these indicators, taking into account the expected changes in the policy of their collection; second, to understand how the change in revenue sources will additionally indirectly affect the behavior of the tax base, including its influence on the size of macroeconomic indicators. it is well known that ukraine’s economy is heavily dependent on imports. this is clearly evidenced by to cite this article: luchko, m., drozd, i., plutytska, k., ruska, r., vovk, i. (2021). analysis and modelling of value added tax revenues on imports: some issues of application in ukraine. international journal of production management and engineering, 9(1), 37-46. https://doi.org/10.4995/ijpme.2021.13984 int. j. prod. manag. eng. (2021) 9(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international 37 https://orcid.org/0000-0001-6499-4188 https://orcid.org/0000-0002-1930-2522 https://orcid.org/0000-0002-0280-0453 https://orcid.org/0000-0002-1854-9734 https://orcid.org/0000-0002-4617-516x http://creativecommons.org/licenses/by-nc-nd/4.0/ the data in table 1 and figure 1. in some years, the amount of imports reaches 50 percent of nominal gdp, which significantly complicates the functioning of the economy. 0 500000 1000000 1500000 2000000 2500000 3000000 3500000 4000000 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 m ill io n u a h national gdp and import of goods in ukraine, milllion uah nominal gdp per year import of goods and services 0 500000 1000000 1500000 2000000 2500000 3000000 3500000 4000000 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 m ill io n u a h national gdp and import of goods in ukraine, milllion uah nominal gdp per year import of goods and services figure 1. ratio of nominal gdp and imports in ukraine, million uah (source: minfin, 2019). vat is one of the main sources of the state budget. unlike many european countries, administering this tax in ukraine is problematic. this is due to many factors and the main ones are the state of the economy, the exports-imports ratio, electronic administering and the creation of vat debt, numerous tax exemptions, benefits and an extremely low level of business culture. these problems certainly complicate the vat revenue forecast. table 1. import of ukraine from 2005 to 2018 (uah million). year national gdp import of goods, works and services import quota, % of gdp 2005 441452 223555 50.6 2006 544153 269200 49.5 2007 720731 364373 50.6 2008 948056 520588 54.9 2009 913345 438860 48.0 2010 1082569 580944 53.7 2011 1316600 779028 59.2 2012 1408889 835394 59.3 2013 1454931 805662 55.4 2014 1566728 834133 53.2 2015 1979458 1084016 54.8 2016 2383182 1323127 55.5 2017 2982920 1618749 54.3 2018 3558706 1914893 53.8 import of goods, works and services into the customs territory of ukraine is subject to vat under the provisions of the tax code of ukraine. vat liabilities on import already exist by the time of submitting customs declaration for customs clearance. yet vat on import would still have to be paid before or on the day of the customs declaration submission. usually, the import of goods into the customs territory of ukraine is subject to 20% vat rate. a 7% tax rate is applied only to the import of medicine and medical devices provided that: (1) the medicine is authorized for production and use in ukraine; is listed in the state register of medical products; (2) medical devices are authorized for marketing and/or putting into service and use in ukraine; are listed in the state register or meet the requirements of the relevant technical regulations (confirmed by a document of conformity). failure to meet the conditions results in an application of a 20% import tax rate. thus, in practice, we have two tax rates, which will undoubtedly make adjustments to the vat payment and its administration. in order to forecast vat on import, the most important is to develop the appropriate methodology for vat revenues estimating for countries in transition, in particular for ukraine. there are four basic tax revenues estimating methodologies: the actual rate method, the elasticity method, the simulation method, and the trend and auto-correlation method. it is quite difficult to determine the best methodology to be applied in ukraine. all methods have their advantages and disadvantages. 2. literature review for the article goals to be achieved, they are necessary to be structured in the following areas: general questions and provisions, values of tax analysis and forecasting, methods of regression analysis models, use of arima model. many authors have addressed the problem of vat modeling, including impact on main indicators of well-being (bilan et al., 2020; danilov, 1999; ebrill, et al., 2002; louzis, 2019; luchko et al., 2019; mishchuk and grishnova, 2015; ngoc huy, 2018; shahnazarian et al., 2017). it is considered that the most widely discussed topics in the economic scientific literature on vat include topics such as administering the tax, problems of its payment, tax evasion, vat reimbursement in eu, vat refunds int. j. prod. manag. eng. (2021) 9(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international luchko et al. 38 http://creativecommons.org/licenses/by-nc-nd/4.0/ and fraud, vat assessment, etc. the topic of vat revenue modeling, though extremely important, is not widely explored. according to the data, it has been little discussed in the current literature on public finances. there is reason to believe that vat is considered as one of the most stable taxes. accordingly, it is relatively easy to forecast based on simple methods. the modern vat (ebrill, keen, bodin and summers, 2001) is one of the most interesting research papers on vat. the content of the work shows that it thoroughly investigated the nature of vat, analyzed the income from this tax, its rates and benefits, administration and audit of vat and vat refunds. the study proposes a summary of vat efficiency through the «efficiency ratio», which is calculated as vat revenue gdp ratio divided by the standard vat rate. since errors are possible in the calculation of gdp, the more appropriate baseline is aggregate consumption, which is the ideal vat base. the c-efficiency ratio is the vat revenue consumption ratio divided by the standard tax rate. if the c-efficiency ratio is 100%, then the vat system works very well. in ukraine, however, this ratio is low. jenkins, kuo and shukla (2000) have indicated that tax analysis and revenue forecasting are crucial in ensuring the stability of the state’s tax policy and budgeting. for a timely and effective analysis of the revenue aspects of a country’s fiscal policy, it is increasingly important to rely on internal tax policy forecasting rather than to rely solely on tax experts from the outside. this makes it possible to predict and analyze the impact of fiscal policy on the economy and to assess the effects of tax measures on revenues, with the ultimate aim of ensuring a healthy fiscal situation in the economy. the following broad functions are necessary to be performed: (a) monitoring of revenue collection; (b) an assessment of the economic, structural and revenue aspects of tax policy; (c) analysis of tax expenditures; (d) assessing the impact of non-tax economic policies; (e) forecasting future tax revenues. as noted by bogetic and hassan (1993), vat receipts usually depend on three groups of factors: the structural aspects of tax (rates, bases, thresholds, etc.), the amount of taxable activity, and tax discipline. in general, vat revenue can be modeled as a function of r (ι, α), where ι is the tax collection variables (rates, bases, thresholds, etc.) and α is the economic environment variables (which affect the tax base and compliance with the collection rules). next, the authors distinguish between taxable and non-observable tax variables. researcher stock (2001) believes that historical empirical patterns should be used to predict time series. at the same time, it is expedient to be guided by a theoretical understanding of economic processes, which provides the basis for the creation of future forecasts. economic forecasts are used in a wide range of activities, including the definition of monetary and budgetary policies, state and local budgets, financial management and financial engineering. key elements of economic forecasting include the choice of one or several forecasting models that are appropriate for addressing the problem, as well as estimation and reporting of uncertainty associated with the forecast, and protection against model instability. the main purpose of the study conducted by gamboa and sophia (2002) is developing a tax forecasting model. existing tax forecasting methodologies have shown that the approach to tax elasticity using the regression procedure is more efficient, and is considered to be the best forecasting methodology. thus, a tax elasticity approach is used for the purposes of this study. structural models for unified equation prediction have been identified and evaluated. both linear and double logarithmic functions were applied in an attempt to use the ordinary least squares method as a regression procedure. giesecke and tran (2012) point out that the modeling of payment on exported and imported goods in different countries, combined vat compliance measures are valuable for identifying problematic areas of its payment. they are also important for meaningful cross-country comparisons and for the shortest possible time for vat compliance. a comprehensive and general basis for calculating vat compliance rates at both the economy and industry level is presented. this makes it possible to set multiple vat rates, exemptions, registration rates, refund restrictions, informal activities, taxation of domestic non-residents and undeclared imports. andrejovska and pulikova (2018) point out that, as a tool for assessing the macroeconomic impact of alternative tax policies in the country, taxes are often weakened by restrictions on the measurement of tax revenues. the article investigates the quantitative impact of selected macroeconomic indicators (gross domestic product, employment rate, public int. j. prod. manag. eng. (2021) 9(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international analysis and modelling of value added tax revenues on imports: some issues of application in ukraine 39 http://creativecommons.org/licenses/by-nc-nd/4.0/ debt, foreign direct investment, effective tax rate, statutory tax rate) on the total amount of tax revenues taking into account the competitiveness of 28 eu member states. the methods of three regression analysis models were used: merger models, fixed effects models, and random effects models. the hypothesis that gross domestic product has the greatest impact on tax revenues was tested. the strongest correlation was observed between tax revenues and employment rates, as well as foreign direct investment and gross domestic product, as was confirmed by the analysis. sabaj and kahveci (2018) propose the introduction of new forecasting models and the use of forecast combinations for albania, with forecasting errors exceeding the average. the results of the evaluation show that the influence of internal and external factors on the forecasting of tax revenues creates a significant improvement in the accuracy of tax revenues. the estimates and combinations of forecasts in this paper result in fewer errors than the official forecasts, suggesting that revising the tax forecasting methodology can increase the accuracy of forecasts for emerging market economies. rudzkis and maciulaityte (2007) examined the forecast of budget revenues using econometric models in lithuania. it is established that the set of applied models should be reduced to very simple models in short time series. therefore, the regression analysis is proposed to be conducted in two stages: econometric modeling being the first stage, and the algorithms for forecasting the tax base being the second. cross-estimation was used to evaluate the accuracy of these algorithms. the work of legeida and sologoub (2003) analyzes modeling and forecasting of vat payment in ukraine. unlike most vat administering countries, it is problematic to administer this tax in ukraine. of paramount concern are vat debt, numerous tax exemptions, and extremely low vat compliance. these issues certainly complicate the forecast of vat revenues. that is why the main purpose of the work was to test different methodologies for forecasting vat revenues. using an effective rate approach, the authors found out that actual vat revenues constituted less than a half of potential vat revenues. the econometric method establishes a strong empirical long-term relationship between vat revenues and its base. soto-ferrari et al. (2019) examined the capabilities and efficiency of arima models, forecasting capabilities, and developed procedures to improve such forecasting. according to büyükşahin and ertekin (2019), there are different methods for predicting time series that use linear and nonlinear models separately or a combination of both. studies show that combining linear and nonlinear models can be effective for improving forecasting efficiency. a new hybrid method of autoregressive integrated moving average (arima), an artificial neural network (ann), which operates in a more general framework, is proposed. it is stated that the experimental results showed that the strategies of decomposition of the initial data and the combination of linear and nonlinear models in the process of hybridization are key factors for predicting the efficiency of the methods. tiao (2001) indicates that time series data in business, economics, environment, medicine, and other scientific fields tend to exhibit patterns such as trends, seasonal fluctuations, irregular cycles, and periodic level or variability shifts. the task of analyzing such series is often to extrapolate the dynamic picture to predict future observations, as well as to evaluate the effect of known exogenous interventions and to detect unexpected interventions. the imf (2001) outlines four main revenue forecasting approaches. the first approach is based on the actual rate. to calculate tax revenue, we first calculate the actual rate by dividing the tax amount by the estimated tax base. usually, the actual tax rate is below the statutory rate. the difference between the two may be explained by tax exemptions and taxpayer discipline problems. next, we estimate tax revenue by multiplying the estimated tax base for the next period by the actual tax rate for the current period. the second approach is based on flexibility. it is based on establishing a stable empirical relationship between the growth of tax revenues and the corresponding increase in the tax base, known as “elasticity”. the increase in tax revenues is obtained as a product of the projected increase in the tax base for elasticity. it also takes into account the estimated value of the impact of changes in the tax structure, its administration and tax discipline. the third one is an econometric approach. it can be used to calculate revenue forecasts based on general equilibrium model or micro modeling based on sample tax amounts. int. j. prod. manag. eng. (2021) 9(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international luchko et al. 40 http://creativecommons.org/licenses/by-nc-nd/4.0/ the fourth is a trend-based and autocorrelation approach. past trends can be used to forecast, along with specific information about each source of revenue. empirical approaches often use arima models, in particular for forecasting. the objective of the article is to investigate the possibilities of building a model of the dynamics of monthly import vat revenues and its analysis. standard white noise errors are analyzed and the relative error of the model is determined. the significance of the parameters by the student’s t-test was checked. in our opinion, this will enable us to check the equality of the averages in the two samples. to develop the main idea of our study, we propose the following hypothetic assumptions that we will experimentally test for a possible solution to the problem. the first assumption is whether vat on imports can act as a tool for macroeconomic impact on the formation of budget revenues. the second one is to find out if parameter importance is confirmed when using the arima model to forecast vat on imports. 3. research methods let us check the assumptions made. the results of the study confirm that the import quota is expressed as a percentage of the volume of imports to gdp, which characterizes the country’s dependence on world markets for goods and services. according to table 1 in ukraine, it is about 50%, which indicates the dependence of the country’s economy on imports. in this respect, it should be borne in mind that vat on imports is a tool for macroeconomic impact on state budget revenues. it is safe to say that on the one hand, imports of goods have a negative impact on the development of domestic economy and on the other hand, they stimulate it and significantly improve the budget. imports vat payment statistical data is presented in table 2. according to table 2, we will analyze the analytical model of calculating vat revenues from imported goods to ukraine. specification of the model (analytical form of econometric model) is written in equation (1). yt=b0+b1t+u (1) where yt – the level of a series of dynamics at time t=1,2,...,55. consider the dynamics of actual monthly vat revenues from imported goods, works and services into the territory of ukraine as a time series (figure 2). there is reason to believe that it represents fluctuations around some level. let us check the assumption made. it seems likely to use regression and moving average methods. essentially, the regression method is based on the construction of a line that “on average” deviates the least from the array of values that specify the behavior of the baseline. mathematically it is described by the equation (2): yt=φ1yt-1+φ2yt-2+ …+φpyp+εt (2) where yt – the value of y at time t; φi – equation coefficients (i=1,2…p); p – autoregression order; εt – random variable. 0 5000 10000 15000 20000 25000 30000 35000 0 10 20 30 40 50 60 v a t period figure 2. dynamics of monthly actual vat revenues from imported goods, works and services to ukraine according to the table 2. int. j. prod. manag. eng. (2021) 9(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international analysis and modelling of value added tax revenues on imports: some issues of application in ukraine 41 http://creativecommons.org/licenses/by-nc-nd/4.0/ at the same time, the moving average method is that each element of the series is prone to the total effect of previous errors: yt=ω1εt-1+ω2εt-2+ …+ωqεt-q+εt (3) where yt – the value of y at time t; ωi – equation coefficients (j=1,2…q); q – moving average order; εt – random variable. for further study, we use the arima model, which combines the two methods and looks like: y t p i 1 ϕ i y t 1 q j 0 ω j ε t 1 const (4) testing the input data of the monthly vat receipts from the imported goods, works and services into the territory of ukraine, we transformed the time series table 3. simulation results in statistica. model: (0,0,1)(0,0,1) conversions: ln(x) seasonal lag: 12 approxim. number.: 55 initial ss=5275.1 final ss=2.3758 (0.0450%) ms=0.04569 parameters (p/ps-autoregression, q/qs-moving aver.); selection: p<0.05 parameter parameter asymptote. stat. f. asymptote. t(52) p const. 9.684210 0.075863 127.6533 0.000000 q(1) -0.639558 0.093089 -6.8704 0.000000 qs(1) -0.579764 0.119670 -4.8447 0.000012 model: (0,0,2)(0,0,1) conversions: ln(x) seasonal lag: 12 approxim. number.: 55 initial ss=5275.1 final ss=1.8078 (0.0343%) ms=0.03545 parameters (p/ps-auto-regression, q/qsmoving aver.); selection: p<0.05 const. 9.591779 0.113452 84.54483 0.000000 q(1) -0.863142 0.144147 -5.98791 0.000000 q(2) -0.511931 0.127891 -4.00286 0.000203 qs(1) -0.600854 0.124562 -4.82373 0.000013 model: (1,1,1)(0,2,1) conversations: d(1),2*d(1) seasonal lag: 12 approxim. number.: 55 initial ss=1723e6 final ss=4112e5 (23.86%) ms=8567e3 parameters (p/ps-auto-regression, q/qsmoving aver.); selection: p<0.05 const. 41.06193 38.31173 1.07178 0.289176 p(1) -0.32089 0.16237 -1.97627 0.053886 q(1) 0.94344 0.08058 11.70816 0.000000 qs(1) -0.68889 0.13429 -5.12980 0.000005 model: (1,1,1)(0,1,1) conversations: d(1),d(1) seasonal lag: 12 approxim. number. 55 initial ss=1612e6 final ss=4279e5 (26.54%) ms=7377e3 parameters (p/ps-auto-regression, q/qsmoving aver.); selection: p<0.05 const. 19.32306 19.50255 0.99080 0.325900 p(1) -0.45805 0.12184 -3.75936 0.000398 q(1) 0.95106 0.04868 19.53620 0.000000 qs(1) -0.19042 0.13032 -1.46121 0.149355 table 2. dynamics of monthly vat receipts from imports, mln uah. years 2015 2016 2017 2018 2019 months vat vat vat vat vat 1 6814,2 9531,0 15291,5 22033,0 20972,4 2 10779,1 13969,9 18674,6 21459,9 23778,4 3 12294,6 16203,0 22280,0 21865,3 24863,4 4 11482,4 14565,3 19961,5 20995,9 22992,7 5 9637,7 13064,3 19190,7 22828,4 24847,0 6 10429,7 13284,8 19394,6 21387,5 21155,2 7 12578,9 14836,5 19924,9 25215,5 26056,7 8 12071,9 16622,4 21022,3 26251,1 9 12673,6 16247,0 21111,7 26534,9 10 12715,3 16818,8 23712,4 31038,1 11 13545,7 17257,4 24722,0 29474,9 12 13741,2 19052,9 25303,9 26292,8 source: state tax service of ukraine (2020). int. j. prod. manag. eng. (2021) 9(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international luchko et al. 42 http://creativecommons.org/licenses/by-nc-nd/4.0/ and conducted the research using arima modeling. as a result, the following models were obtained (table 3). to select the appropriate model, consider the simulation results in the statistics program (table 3). in the first two models, we log the time series and get: the initial sum of squares of ss residuals is the same, the finite one is smaller in the second model, and the mean square of ms residuals in the second model is also smaller. so, in terms of residuals, the second model is preferable, but in our belief we will check the significance of the parameters by the student’s t-test. to this end, we propose two hypotheses: h0– the model parameters are zero and the alternative hypothesis h1– not all parameters are zero. for each parameter φi and ωjtdefines are defined as the ratio of the regression coefficient taken to the module to its standard error. let’s check the assumptions made. the calculated value is compared with with tcrit=2.006 the significance level α=0.01 and the number of degrees of freedom df = 53. comparing the value of tcalcul and tcrit for each of the parameters obtained, confirms the hypothesis of the significance of all parameters in both models, so we accept the alternative hypothesis. consider the last indicator p. the closer it is to zero, the better the result, the closer or equal to one the parameter is insignificant, so all the parameters are significant. models (1,1,1) (0,2,1) and (1,1,1) (0,1,1) are created using the differences of lags 2 and 1, respectively. analyzing the sum of squared residuals by these models we can say that they are quite large and the series have not been smoothed out. as a consequence, in the student test, in the model (1,1,1) (0,2,1), the first two parameters and in the model (1,1,1) (0,1,1), the first and fourth are not statistically significant, that is, they can be neglected in further research. this, in turn, will lead to incorrect forecasts, so we reject these models. since the models (0,0,1) (0,0,1) and (0,0,2) (0,0,1) according to preliminary tests are suitable for further use, we analyze the residuals (table 4). according to the results, the standard white noise errors in model (0,0,1) (0,0,1) have one emission, while in model (0,0,2) (0,0,1) there are two. the conducted research confirms that we should give preference to the first model, and confirms our initial assumption about the non-stationary nature of the time series, and that all the transformations made are correct. to achieve this goal, we compare the forecasting results of these models (table 5). these forecasts are presented graphically in figure 3. check the assumptions made and calculate the relative error on the models. to do this, take the actual value for july 2019 y55=26056.7, the predicted value for the model arima (0,0,1) (0,0,1) is 25135.49, then: 26056.7 25135.49 2605.7 0.035 table 4. modelling results standard errors statistica white noise estimates. lag autocorrelation function vat on imports: arpss (0,0,1) (0,0,1) balances autocorrelation function vat on imports: arpss (0,0,2)(0,0,1) balances autocorrelation statistical error ljung–box q p autocorrelation statistical error ljung–box q p 1 0.256711 0.131244 3.82590 0.050475 -0.032985 0.131244 0.06316 0.801564 2 0.591859 0.130023 24.54628 0.000005 0.188356 0.130023 2.16172 0.339317 3 0.333622 0.128790 31.25662 0.000001 0.403724 0.128790 11.98833 0.007430 4 0.485453 0.127546 45.74306 0.000000 0.210263 0.127546 14.70596 0.005358 5 0.333841 0.126289 52.73098 0.000000 0.191661 0.126289 17.00918 0.004489 6 0.269011 0.125020 57.36098 0.000000 0.134199 0.125020 18.16140 0.005851 7 0.335951 0.123738 64.73236 0.000000 0.253217 0.123738 22.34917 0.002216 8 0.193429 0.122442 67.22800 0.000000 0.077519 0.122442 22.75000 0.003709 9 0.253762 0.121132 71.61669 0.000000 0.154776 0.121132 24.38262 0.003745 10 0.125740 0.119808 72.71815 0.000000 0.106300 0.119808 25.16984 0.005045 11 0.171745 0.118470 74.81975 0.000000 0.086523 0.118470 25.70322 0.007199 12 0.043976 0.117116 74.96075 0.000000 0.054752 0.117116 25.92178 0.011034 13 0.094175 0.115746 75.62275 0.000000 0.044603 0.115746 26.07028 0.016667 14 0.005324 0.114360 75.62492 0.000000 0.028680 0.114360 26.13318 0.024938 15 0.064706 0.112956 75.95307 0.000000 0.045239 0.112956 26.29358 0.035101 int. j. prod. manag. eng. (2021) 9(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international analysis and modelling of value added tax revenues on imports: some issues of application in ukraine 43 http://creativecommons.org/licenses/by-nc-nd/4.0/ according to the arima model (0,0,2) (0,0,1), the forecast value is 26119.60, then 26056.7 26119.6 2605.7 0.002 it is obvious that in the first case an error is approximately 4%, whereas in the second case it is only 0.2%. since the standard errors have only one emission in the arima (0,0,1) (0,0,1) model, it is safe to say that this model should be used, even though the relative error in it is greater. using this model, we will make a quarterly forecast for the following years (figure 4). it is evident that the final seasonal factors show that the first quarters of 2020-2021 will be characterized by steady import vat revenues, with a slight increase in 2022. at the same time, from 2023 to 2028 a steady decrease in revenues is observed. the second quarter of 2020 is characterized by the same situation, as was in 2019. in 2021 there will be a partial decrease, and starting with 2022, a partial increase is expected. a steady increase in revenues is expected starting from 2023. the third quarters of these years will be characterized by stability and slight growth. the fourth quarters from 2020 to 2023 will be shaped by steady revenues, and from 2023 to 2028 there will be a slight decline. final si differences with extremes and final si differences without extremes characterize forecasts of surplus receipts (errors) and those without surpluses. final si differences without extremes have some differences from final seasonal factors. final si differences with extremes characterize significant emissions, in particular in the first quarter of 2021 and 2027, a sharp decrease in revenues or their termination, which gives a rather pessimistic forecast, and in 2025 and 2028 a sharp increase, making an optimistic forecast. such an increase is still possible in the fourth quarter of 2020, in the table 5. modeling results — forecasting by models using statistica. no observation forecasts; model: (0,0,1) (0,0,1) seasonal lag: 12 initial.: vat imported from the territory of ukraine goods initial: 1 final: 55 forecasts; model: (0,0,2) (0,0,1) seasonal lag: 12 (t_3) initial.: vat imported from the territory of ukraine goods initial: 1 final: 55 forecast lower 90% upper 90% forecast lower 90% upper 90% 56 25489.09 24048.62 30176.79 25666.46 24650.59 31321.53 57 20078.61 13083.21 30814.33 21746.53 14317.07 33031.31 58 21466.87 13987.81 32944.88 20437.28 13057.20 31988.67 59 20360.96 13267.20 31247.66 19332.74 12351.51 30259.83 60 19340.07 12601.98 29680.90 18349.16 11723.12 28720.33 61 27163.60 17699.79 41687.56 27689.91 17690.84 43340.57 62 18691.34 17179.27 28685.31 17774.05 11355.68 27820.15 63 19479.83 12693.05 29895.40 18571.21 11864.98 29067.88 64 18664.97 15162.09 28644.84 17765.61 11350.29 27806.94 65 18888.85 12307.97 28988.44 17959.61 11474.24 28110.60 66 17708.07 11538.57 27176.31 16822.19 10747.55 26330.29 67 18965.95 12358.20 29106.75 18036.79 11523.55 28231.40 figure 3. vat import forecast, mln. int. j. prod. manag. eng. (2021) 9(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international luchko et al. 44 http://creativecommons.org/licenses/by-nc-nd/4.0/ third quarter of 2021 and in the second quarter of 2026. in general, it should be noted that vat receipts from imports will not be spasmodic, but rather uniform. 4. conclusion and discussion of the results we do not claim the absolute scientific exclusivity of our thoughts. the study revealed the main factors of influence on the application of the arima model when modeling vat revenues from imports in the short and long term. the resulting regression model in statistica software linked the variables with a fairly accurate approximation. in answer to the questions and hypotheses suggested, we would like to point out that the study has shown that vat on import has a significant impact on budget revenues, and therefore, choosing the best methodology for vat modeling for a transition country will have a significant impact on budget and gdp indicators as a whole. we believe it expedient to use the arima model. a generalized economic and mathematical model makes it possible to minimize losses and to forecast budget revenues in terms of vat on imports. the model can be adapted to forecast revenues from other taxes and fees. in addition, it should be noted that, in our opinion, the performed research has enabled us to set new scientific problems that are of great theoretical and practical importance and may be the subject of further scientific research. first of all, they should include: optimization of tax administration; application of the possibilities of block-chain and artificial intelligence technologies in conditions of uncertain tax payment. possible direction for further research on this problem is also taking into account in the given model the future value of money and its impact on the inflationary effect due to its depreciation. in our opinion, this requires additional justifications and changes to the individual components of the calculations made. references andrejovska, a., pulikova, v. (2018). tax revenues in the context of economic determinants. montenegrin journal of economics, 14(1), 133-141. https://doi.org/10.14254/1800-5845/2018.14-1.10 figure 4. quarterly vat import forecast chart, mln. int. j. prod. manag. eng. (2021) 9(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international analysis and modelling of value added tax revenues on imports: some issues of application in ukraine 45 https://doi.org/10.14254/1800-5845/2018.14-1.10 http://creativecommons.org/licenses/by-nc-nd/4.0/ bilan, y., mishchuk, h., samoliuk, n., yurchyk, h. (2020). impact of income distribution on social and economic well-being of the state. sustainability, 12(1), 429. https://doi.org/10.3390/su12010429 bogetic, z., hassan, f.m. (1993). determinants of value-added tax revenue: a cross-section analysis (vol. 1203). world bank publications. büyükşahin, ü.ç., ertekin, ş. (2019). improving forecasting accuracy of time series data using a new arima-ann hybrid method and empirical mode decomposition. neurocomputing, 361, 151-163. https://doi.org/10.1016/j.neucom.2019.05.099. danilov, o. (1999). methodological approach to the economic and mathematical modelling of tax policy. scientific works ufei, 3(6), 21. ebrill, l.p., keen, m., bodin, j.p., summers, v.p. (2001). the modern vat. international monetary fund. https://doi.org/ 10.5089/9781589060265.071 ebrill, l., keen, m., bodin, j.p., summers, v. (2002). the allure of the value-added tax. finance & development, 39(2), 44-44. gamboa, a.m., sophia, j. (2002). development of tax forecasting models: corporate and individual income taxes (no. 2002-06). pids discussion paper series. giesecke, j., tran, n.h. (2012). a general framework for measuring vat compliance rates. applied economics, 44(15), 1867-1889. https://doi.org/10.1080/00036846.2011.554382. imf. (2001). manual on fiscal transparency. retrieved march 23, 2020 from, https://www.imf.org. jenkins, g.p., kuo, c.y., shukla, g. (2000). tax analysis and revenue forecasting. cambridge, massachusetts: harvard institute for international development, harvard university. retrieved from: https://cri-world.com/publications/qed_dp_169.pdf. leal, t., pérez, j.j., tujula, m., vidal, j.p. (2008). fiscal forecasting: lessons from the literature and challenges. fiscal studies, 29(3), 347386. https://doi.org/10.1111/j.1475-5890.2008.00078.x legeida, n., sologoub, d. (2003). modeling value added tax (vat) revenues in a transition economy: case of ukraine. institute for economic research and policy consulting working paper, 22, 1-21. louzis, d.p. (2019). steady state modelling and macroeconomic forecasting quality. journal of applied econometrics, 34(2), 285-314. https://doi.org/10.1002/jae.2657 luchko, m., lew, g., ruska, r., vovk, i. (2019). modelling the optimal size of investment portfolio in a non-state pension fund. journal of international studies, 12(1), 239-252. https://doi.org/10.14254/2071-8330.2019/12-1/16 minfin ministry of finance of ukraine. (2019). exports and imports of ukraine. retrieved from official web site of the ministry of finance of ukraine https://index.minfin.com.ua/ua/economy/gdp/eximp/ mishchuk, h., grishnova, o. (2015). empirical study of the comfort of living and working environment-ukraine and europe: comparative assessment. journal of international studies, 8(1), 67-80. https://doi.org/10.14254/2071-8330.2015/8-1/6 ngoc huy, d. (2018). tax rates effects on the risk level of listed viet nam real estate firms during global economic crisis 2007-2009. economics, management and sustainability, 3(2), 29-41. https://doi.org/10.14254/jems.2018.3-2.3 rudzkis, r., maciulaityte, e. (2007). econometrical modelling of profit tax revenue. nonlinear analysis: modelling and control, 12(1), 95-112. https://doi.org/10.15388/na.2007.12.1.14724 sabaj, e., kahveci, m. (2018). forecasting tax revenues in an emerging economy: the case of albania (no. 84404). university library of munich, germany. shahnazarian, h., solberger, m., spånberg, e. (2017). forecasting and analysing corporate tax revenues in sweden using bayesian var models. finnish economic papers, 28(1), 50-74. retrieved from https://www.taloustieteellinenyhdistys.fi/wp-content/uploads/2017/10/ fep_1_17_4_spanberg.pdf soto-ferrari, m., chams-anturi, o., escorcia-caballero, j. p., hussain, n., & khan, m. (2019, september). evaluation of bottom-up and top-down strategies for aggregated forecasts: state space models and arima applications. in international conference on computational logistics (pp. 413-427). springer, cham. https://doi.org/10.1007/978-3-030-31140-7_26 state tax service of ukraine. (2020). income of taxes and fees. retrieved from official web site of the state tax service of ukraine: https://tax.gov.ua/diyalnist-/pokazniki-roboti/nadhodjennya-podatkiv-i-zboriv--obovyaz/nadhodjennya-podatkiv-i-zboriv/. stock, j.h. (2001). time series: economic forecasting. international encyclopaedia of the social & behavioural sciences. 15721-15724. https://doi.org/10.1016/b0-08-043076-7/00526-x tiao, g.c. (2001). time series: arima methods. international encyclopedia of the social & behavioural sciences, 15704-15709. https://doi.org/10.1016/b0-08-043076-7/00520-9 int. j. prod. manag. eng. (2021) 9(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international luchko et al. 46 https://doi.org/10.3390/su12010429 https://doi.org/10.1016/j.neucom.2019.05.099 https://doi.org/10.5089/9781589060265.071 https://doi.org/10.5089/9781589060265.071 https://doi.org/10.1080/00036846.2011.554382 http://www.imf.org https://cri-world.com/publications/qed_dp_169.pdf https://doi.org/10.1111/j.1475-5890.2008.00078.x https://doi.org/10.1002/jae.2657 https://doi.org/10.14254/2071-8330.2019/12-1/16 https://index.minfin.com.ua/ua/economy/gdp/eximp https://doi.org/10.14254/2071-8330.2015/8-1/6 https://doi.org/10.14254/jems.2018.3-2.3 https://doi.org/10.15388/na.2007.12.1.14724 https://www.taloustieteellinenyhdistys.fi/wp-content/uploads/2017/10/fep_1_17_4_spanberg.pdf https://www.taloustieteellinenyhdistys.fi/wp-content/uploads/2017/10/fep_1_17_4_spanberg.pdf https://doi.org/10.1007/978-3-030-31140-7_26 https://tax.gov.ua/diyalnist-/pokazniki-roboti/nadhodjennya-podatkiv-i-zboriv--obovyaz/nadhodjennya-podatkiv-i-zboriv/ https://doi.org/10.1016/b0-08-043076-7/00526-x https://doi.org/10.1016/b0-08-043076-7/00520-9 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.10817 received: 2018-10-14 accepted: 2019-11-23 empirical study of the business growth strategy related to the added value by intellectual capital alcalde-delgado, r. a1*, sáiz-barcena l. b1, olmo, r.b2, alonso de armiño, c.b3 a department of economics and business administration. area of business organisation. faculty of economics and business studies, university of burgos. plaza infanta doña elena, s/n, 09001 burgos, spain. b department of civil engineering. area of business organisation. higher polytechnic school, university of burgos. campus río vena (edificio a): avda. cantabria s/n 09006-burgos, spain. a1 roberto.alcalde.delgado@gmail.com, b1 lsaiz@ubu.es b2 rdelolmo@ubu.es, b3 caap@ubu.es abstract: four consecutive years of more than a thousand spanish companies from different economic sectors are analyzed to determine the influence of intellectual capital on the business growth strategy. one of the purposes of this work is to establish a classification criterion of the strategic behaviour of a company linked to the growth of three factors: the demand of the sector, the sales of the company and the financial sustainability of the company. another purpose is to develop and validate an appropriate classification of where the value added by human intellectual capital is structurally concentrated and used according to the strategic behaviour, growth and sector of the company. interesting conclusions are drawn about the strategic behaviour of the company and its intangible capital, as well as a different method for classifying companies according to their growth, which helps predict business profitability. key words: value added intellectual coefficient (vaictm), sustainable growth rate (sgr), sales growth, demand growth, business strategy. 1. introduction intellectual capital, as measured by the method vaic™, is positively associated with the profitability of companies, since it allows the performance of a company to be measured in the efficient use of capital, resources and intellectual capacity (greco et al., 2014) (gupta and tarikasingh, 2015) (salazar and villegas, 2019). nowadays, empirical research conducted on the vaic™ model has been focused on specific activity sectors without taking into account the strategic behavior of companies. therefore, there is a gap in the state of the art, when studying the value added by intellectual capital for each state of strategic behaviour. strategic behavior is related to the use of the company’s resources, in combination with core competences, to achieve better competitive positions in the sector that are the source of sustainable competitive advantage. a company may not, however, have the resources it requires to develop a source of sustainable competitive advantage; it has to develop a business strategy (kabue and kilika, 2016). the execution of the business strategy involves changes in the organization of the company, in to cite this article: alcalde-delgado, r., sáiz-barcena l., olmo, r., alonso de armiño, c. (2020). empirical study of the business growth strategy related to the added value by intellectual capital. international journal of production management and engineering, 8(1), 1-11. https://doi.org/10.4995/ijpme.2020.10817 int. j. prod. manag. eng. (2020) 8(1), 1-11creative commons attribution-noncommercial-noderivatives 4.0 international 1 https://orcid.org/0000-0002-2019-7926 https://orcid.org/0000-0003-1785-5198 mailto:roberto.alcalde.delgado@gmail.com mailto:lsaiz@ubu.es http://creativecommons.org/licenses/by-nc-nd/4.0/ the human and technical resources, as well as the behaviour of relations inside and outside the company, in order to improve the competitive position. the enhancement of the competitive position is associated with a cycle of growth of the company, and therefore, the improvement of the company’s indicators such as sales, finance, employees, etc., (peñate santana, 2013) (davidsson, achtenhagen and naldi, 2010). the companies’ performance are positively related to structural capital (sc), human capital (hc), and relational capital (rc). it should be stressed that rc has the greatest influence on financial performance indicators, as well as the sustainable growth rate (sgr) (xu, and wang, 2018) (firer and mitchell williams, 2003). this work will make available significant practical contributions related to the following questions: how can we classify the strategic behaviour of companies? are there different values in the components of the vaic™ model for companies between activity sectors or between differences in strategic behaviour? in order to answer the first research question, the authors propose three parameters (growth in sales, growth in finances, and growth of the economic sector of the company), which are linked to the strategic behaviour of the companies (schwab et al., 2019) and allow a method of classifying the companies to be created. to answer the second research question, an empirical study has been conducted for a wide and varied sample of companies. this study makes it possible to find certain relationships between strategic behaviour, the sector of activity and the value added by the intellectual capital of companies. 2. classification of companies the use of company classifications in all fields of research to create order in the field and facilitate theorization deserves careful and explicit consideration by researchers. companies are classified according to size, business orientation, industry and business models. without a certain level of consensus on the classification of objects within a field of research, the accumulation of knowledge and meta-analysis are impeded and theorization is forced to be on a large scale. (lambert, 2015). applying lambert’s methodology (2015), the purpose of the proposed classification design is to obtain a classification of companies according to their strategic behaviour. the characteristics that best define this purpose are the growth of the company (mahdjour, 2015). in order to determine the classification criteria, the variables and the combination of the same are determined so as to allow different categories of company growth to be established. these rules define the procedure for classifying companies into different categories that are related to strategic behavior. 2.1. growth of the company for a company, the meaning of the term growth can be approached from several points of view, such as the following: as a view of life-cycle and linear interpretation, involving several stages (slow, rapid, moderate and decreasing growth) where only the companies that best adapt to change survive. as an internal growth of optimization managed by managers to allow business growth and a financial vision of external growth, in which the greater the amount of resources committed, the greater the growth. as a microeconomic approach, the growth corresponds to the adjustment of the company while also looking for a balance between the company’s effectiveness and evolution and, in the long term, of efficiency. as an increase in the physical resources of the company, which seeks to establish its optimal dimension, below or above which a company may or may not be competitive. as a dynamic approach, asking the company to rethink its activities and assess its available resources. as a neoclassical approach to business growth, it proposes the drivers of economic development to the entrepreneurial and managerial capacity, as it discovers the opportunities of the environment before others who are not able to perceive them. what makes companies different is their resources. int. j. prod. manag. eng. (2020) 8(1), 1-11 creative commons attribution-noncommercial-noderivatives 4.0 international alcalde-delgado et al. 2 http://creativecommons.org/licenses/by-nc-nd/4.0/ as a modern evolutionary theory of growth, the factors that contribute to it are identified in order to explain the temporal evolution of a company. it considers that the real limit of growth is human behaviour and its inertia, originating in the driven behaviour or existing routines in the company. on the other hand, the strategic approach to the growth of the company is posed with an approach whereby the company is a portfolio of products and businesses, in which its growth would be based on these parameters (products and businesses). therefore, growth means different things: aggregate growth in products and services that consume energy and materials, growth in profits, growth in trade, growth in disparity in consumption, wealth, and income, growth in underand un-employment (ashford, 2016). futhermore, the problem is to define, in each company, what is meant by growth (álvarez, 2008) and to define how to use the different drivers (human capital, strategy, hrm, innovation, and capabilities) to increase the growth (demir et al., 2017). the strategic approach recognizes the existence of two main strategic directions for growth: internal or organic growth and external growth. both directions create value through internal business processes and differ in the sources used to achieve the said growth (guerras and navas, 2015). every company at some point goes through two challenges they have to overcome: the financial fragility that arises when the company grows more than its ability to finance growth, or the market fragility, which is when the company grows less than the market. these extremes have to be avoided because, in either case, the company can disappear. consequently, companies can be in four different situations (sallenave, 1991), which are: the company can follow the growth of demand and maintain its competitive position, which is a situation of balanced growth. the company can sustain a growth much higher than that of its sector of economic activity. the company cannot follow the growth of its sector and is losing market share. the company has a capacity for growth below the growth of demand. therefore, the proper management of a company to create sustainable growth without financial difficulties implies having a balanced financial structure to make important decisions, while generating value for shareholders (kaplan and norton, 2004) (magreta, 2001). therefore, to determine the classification criteria, the growth in sales, the growth of the demand and the accounting concept of sustainable growth are used. the combination of these variables allows different categories of the growth of the company to be established. these rules define the procedure for classifying companies according to their strategic behaviour. 2.2. sustainable growth rate higgins defined for the first time the accounting concept of sustainable growth (sgr or g*) as the rate of increase in assets and sales that a company can financially support, that is, it reflects the rate at which a company’s sales can grow without financial stress, as long as profit, debt and profit sharing ratios remain constant (higgins, 1977) (amouzesh et al., 2011). a company’s real sales growth rate (gs) is a percentage that measures the growth over a period of time of the sales value of a company that has one or more businesses. the growth rate of a company’s market demand (gd) is a percentage that measures the growth over a period of time by the market demand of the company’s businesses prorated by the amount of sales of each business. the condition of balanced sustainable growth implies that if demand grows, the company must have the same growth in sales in order not to lose market share (gd = gs). consequently, the increase in sales will require an increase in the revolving fund and production capacity, which means a growth in assets (ga), as expressed in equation 1. ga= (assets1 assets0) / assets0 (1) this growth in assets (ga) implies an equal growth in liabilities. in order for liability growth (gl) to be balanced, the ratio between debt and equity (e) must be kept constant, that is, the same debt ratio. int. j. prod. manag. eng. (2020) 8(1), 1-11creative commons attribution-noncommercial-noderivatives 4.0 international empirical study of the business growth strategy related to the added value by intellectual capital 3 http://creativecommons.org/licenses/by-nc-nd/4.0/ otherwise, it would increase indebtedness and thus the risk of financial bankruptcy, as well as a greater dependence on creditors. capital growth (gp) can be expressed as a function of the return on equity (roe) and the target percentage allocated to dividend payments (d), as shown in equation 2. g e d net profit roe d 1 1p 0 0= = ^ ^h h (2) therefore, the balance in the growth ratio between debt and equity implies equal growth of assets, liabilities, capital and debt, as shown in equation 3 (higgins, 2007; sallenave, 1991). g*=ga=gp=roe0(1-d) (3) the total equilibrium will be fulfilled when the balanced growth rate is equal to all growth rates (gd=gs=ga=gp). when this condition is satisfied, the balanced growth rate is called the sustainable growth rate (g*) and the calculation of this value can be seen in equation 3 (álvarez, 2008). therefore, sustainable growth (g*) means the highest growth in sales volume that the company can achieve, maintaining its accounting ratios of assets and liabilities (equity and external funds). it is very difficult to achieve this balance, although the company may approach it after successive temporary adjustments (garcía muñiz, 2011; sallenave, 1991). 2.3. equilibrium growth from the point of view of growth, there are three equilibrium situations for analyzing business management: the equilibrium of commercial management (gd = gs), financial management (g* = gs), and product/business portfolio design (g* = gd) (sallenave, 2002, 1991). if the growth of sales of businesses/products exceeds the growth of the market, this means that the company is gaining market share, and would otherwise be losing market share. therefore, in the case of equality in both types of growth, the commercial management of the company behaves in a balanced way with respect to the market. financial equilibrium is most desirable, since it means that financial management will not suffer tensions and, at the same time, financial resources are used. however, companies can either unbalance their financial situation, increasing the debt ratio, or waste their financial resources by not putting them to generate value. there are therefore two types of cases: profitability imbalance due to under-utilised financial surplus, i.e., an excess of funds that are not used, but which can be used for other purposes, such as financial restructuring of the company, reserves, etc., in order to bring variety. in this case, sustainable growth is higher than sales growth. imbalance of growth with financial deficit, which implies that indebtedness is required for its growth to be viable. in this case, the growth in sales is greater than its sustainable growth. a well-designed product/business portfolio is essential to achieve sustainable growth in excess of market growth. otherwise, it will be said that the design of the portfolio is unbalanced and the positioning of the products/businesses is inadequate. 2.4. strategy behaviour of the firm according to its state of growth according to the state of growth (sustainable growth, sales growth, market demand growth), as can be seen in figure 1, six conditions of non-equilibrium can be considered for companies, corresponding to their strategy behaviour (table 1) (sallenave, 1991, 2002; godet, 1994; socolich mansilla, 2007; álvarez, 2008). 2.4.1. expansive company (gs > g* > gd) expansive companies are those that are expanding beyond available financial resources, exposing themselves to financial stress due to increased indebtedness. at the same time, their sales are growing above the growth of market demand. they are successful companies, with correct commercial management and an adequate design of their business/product portfolio, even if they are getting into debt. therefore, they are desirable companies for investors or for the entry of new shareholders’ capital. 2.4.2. dominant company (g* > gs > gd) these are companies with satisfactory financial and commercial management and an adequate design of the business/product portfolio. int. j. prod. manag. eng. (2020) 8(1), 1-11 creative commons attribution-noncommercial-noderivatives 4.0 international alcalde-delgado et al. 4 http://creativecommons.org/licenses/by-nc-nd/4.0/ consequently, these companies increase their market share and accumulate unused financial resources. due to this excellent competitive position of the dominant company, it is recommended to approach strategies of diversification of the activity or financial restructuring, in order to take advantage of the financial resources generated. 2.4.3. shrinking company (g* > gd > gs) in this situation, we find companies that, despite having an excess of unused financial resources, are losing market share. they are companies with a conservative financial management and an adequate design of the business/product portfolio, but with problems of commercial myopia, since their commercial management is erroneous. if the sectorial rivalry is low, it is appropriate to invest in the company to recover its competitive position. however, if the sectorial rivalry is high, it is advisable for the company to apply strategies to restructure the product/business portfolio, in order to invest in those in which it can more easily improve its competitive position. 2.4.4. company in decline (gd > gs > g*) these companies are in decline as they lose market share while their financial situation deteriorates, therefore these companies require a significant change in their commercial and financial management as well as a restructuring of their product/business portfolio. figure 1. strategy behavior of the firm according to its state of growth. table 1. classification of strategy behavior according to its state of growth. zone classification business/product portfolio financial management portfolio design commercial management 1 expansive gs > g * > gd non-equilibrium of growth well designed gains market share 2 dominant g* > gs > gd non-equilibrium of profitability 3 shrinking g* > gd > gs loses market share 4 restructuring gs > gd > g * non-equilibrium of growth badly designed gains market share 5 in decline gd > gs > g * loses market share 6 unfocuse gd > g * > gs non-equilibrium of profitability int. j. prod. manag. eng. (2020) 8(1), 1-11creative commons attribution-noncommercial-noderivatives 4.0 international empirical study of the business growth strategy related to the added value by intellectual capital 5 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2.4.5. company for restructuring (gs > gd > g*) these companies are in decline as they lose market share while their financial situation deteriorates, so they do not have sufficient resources to invest and improve their competitive position, which significantly increases the likelihood of bankruptcy. they are companies that manage commercially and financially incorrectly and have an inadequate design of the business/product portfolio, so it is advisable to deeply restructure their product/business portfolio. 2.4.6. unfocused company (gd > g* > gs) this corresponds to companies that have sufficient financial resources to invest, while their growth in sales is lower than the growth of the market. this may be the case of companies in decline, which have worsened their financial situation and therefore a partial or total disinvestment in some products/ businesses is recommended in order to adjust to the financial possibilities. this is also the case of shrinking companies that follow a negative evolution due to the fact that they have not resolved their commercial management problems. they are companies that have an adequate conservative financial management but maintain an erroneous commercial management, which means an inadequate design of the business/ product portfolio. 3. the value added by intellectual capital (vaictm) the value contributed by intellectual capital is related to business growth (ghanei and ramezani kheibari, 2015), while the efficient use of intellectual capital reinforces the positive relationship between growth opportunities and financial performance (sardo and serrasqueiro, 2018). therefore, intellectual capital management is the most important factor in company improvement (firer and mitchell williams, 2003). the vaic model makes it possible to measure a company’s performance in the efficient use of capital, resources and intellectual capacity, although it does not measure the stock of intellectual capital at the company’s disposal (greco et al., 2014; gupta et al., 2015). efficiency is related to the value added (va) by the use of resources. the va is considered to be the difference between sales revenue (out) and supplier expenses for the purchase of materials, components and services used in sales, not including personnel expenses (in) (pulic, 2008). therefore, the va is the sum of two elements, the cost of human capital (hc) plus earnings before interest, taxes, depreciation and amortization (ebitda) or structural capital (sc), as indicated in equation 4 (pulic, 2000; ulum et al., 2014; pulic, 2008). va = out–in = hc+sc = hc+ebitda (4) the main components of this model are three (figure 2): human capital efficiency (hce), structural capital efficiency (sce) and capital employed efficiency (cee). figure 2. vaic (source: pulic, 2008). int. j. prod. manag. eng. (2020) 8(1), 1-11 creative commons attribution-noncommercial-noderivatives 4.0 international alcalde-delgado et al. 6 http://creativecommons.org/licenses/by-nc-nd/4.0/ human capital efficiency (hce) is the ratio showing the amount of value added created by each monetary unit spent on workers, i.e., wages. structural capital efficiency (sce) is the ratio that indicates the amount of structural capital (sc) needed to create a unit of value added and measures how successful it is in creating value; it is also called marginal ebitda (pew tan et al., 2007). capital employed efficiency (cee) is a measure of the appropriate use of physical capital; it is a ratio or coefficient that measures the value added per unit of physical capital employed (pulic, 1998). the vaic model provides a standardized and consistent basis for measurement, since data is obtained from audited financial reports rather than subjective evaluations, such as questionnaires (shiu, 2006), while other models are not able to provide a comparison between companies (firer and mitchell williams, 2003; maditinos et al., 2011). the measure of intellectual capital provided by the vaic model has a significant and positive influence on revenue growth, profitability (roe return on equity, roa return on asset, operating, ros return on sale, operating profitability) and companies’ sustainable growth (mukherjee and sen, 2019; smriti and das, 2018; xu and wang, 2018; sardo and serrasqueiro, 2018; kai wah chu et al., 2011; gupta, 2015; maditinos et al., 2011; joshi et al., 2013; najafizadeh and fordoei, 2014; zia et al., 2014; gan and saleh, 2008). approximately 50% of a company’s market value is not reflected in the accounts, as they do not include the positive correlation between intellectual capital and the market value of firms, and also between intellectual capital and financial improvement (chen et al., 2005). overall, evidence has been found to suggest that intellectual capital, as measured by the vaic, is positively associated with corporate profitability (kai wah chu et al., 2011; gupta, 2015; najafizadeh and fordoei, 2014; hajeb et al., 2015). there are different studies with different results for structural capital efficiency (sce), some with a positive relationship, others a negative one and yet others with a non-existent one for sce and profitability. however, for hce and cee, the majority relationship with profitability is positive (haris, et al., 2019). nevertheless, the human capital efficiency (hce) component is the one that presents the greatest difference between companies from different sectors (svanadze, 2015). 4. empirical analysis this work has been carried out in a non-probability convenience sample, that is, the sample is composed of companies that facilitate their measurement and are accessible or favorable. the sample is based on 1,379 companies (73% are public limited companies and 27% are limited liability companies) that have the accounting data for the annual period between 2009 and 2013. however, the data for the four years comprising the time period from 2010 to 2013 will be used, since the data for the year 2009 are used as the basis for the percentage growth for the year 2010. in total, the sample is 5,516 financial statements from 1,379 companies, as shown in table 2. table 2. sample used in the empirical analysis. sector of economic activity samples strategy behaviour: (state of growth of the company) samples 1. restaurants and hotels 124 1. expansive 357 2. construction 316 2. dominant 760 3. distribution &sales 1344 3. shrinking 1828 4. agriculture 56 4. restructuring 586 5. manufacturer 2376 5. in decline 586 6. energy and water 176 6. unfocused 1091 7. information technology 236 8. others 888 total 5516 total 5516 int. j. prod. manag. eng. (2020) 8(1), 1-11creative commons attribution-noncommercial-noderivatives 4.0 international empirical study of the business growth strategy related to the added value by intellectual capital 7 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.1. business profitability grouped by sector of economic activity and strategic behaviour for the selected sample, calculations are made of the company’s profitability ratios, specifically the following are used: roe (return on equity), roa (return on assets), operating profitability and ros (return on sales). in figure 3, a large variability of these profitability indicators can be observed within each sector, especially with regard to roe. this variability for companies within each sector may be related to differences in the strategic positioning of each of them, which motivates different growths. for this reason, the graph in figure 4 has been made, in which a greater concentration of the values of the profitability indicators is observed, when these values are grouped according to the states of growth instead of the company’s sector of economic activity. in this figure 4, negative results of profitability (roe, roa and ros) can be observed for companies in decline; while companies in a restructuring situation have profitability close to zero. on the other hand, positive returns correspond to all dominant, expansive and shrinking companies, and it is the dominant companies that have the highest values, followed by contracting companies. therefore, the six different states used in this study to classify companies are adequate to predict business profitability and thus assess business management performance. since the established classification of companies can determine the strategic behaviour and therefore the goodness of business management, it also seems correct to assume that this classification can be used to predict the change in value added by intellectual capital. thus, the six different states of strategic behaviour used in this study to classify companies are adequate to predict business profitability and thus assess business management performance. to determine these six categories, accounting data has been used to determine the sales growth, sustainable growth and market demand growth. 4.2. value added by intellectual capital grouped by sector of economic activity and strategic behaviour when companies are grouped by activity sectors (figure 5), the lowest values in human capital efficiency indicators are found in the construction and catering sectors, while the highest values correspond to energy and information technology sectors. the sector with the greatest added value of intangible capital is that corresponding to information technologies followed by the energy sector. in figure 3. profitability of the sample by industrial activity. agriculture; construction; distribution & sales; energy & water; manufacturer; others; restaurant & hotels; information technology. figure 4. profitability of the sample by strategic behavior. dominant; in decline; unfocused; expansive; shrinking; restructuring. int. j. prod. manag. eng. (2020) 8(1), 1-11 creative commons attribution-noncommercial-noderivatives 4.0 international alcalde-delgado et al. 8 http://creativecommons.org/licenses/by-nc-nd/4.0/ view of the current strategy behaviour , companies in a conquering business situation have the highest values of intellectual capital, following by expansive companies (figure 6). 5. conclusion this work makes an outstanding contribution to the state of the art of the classification of companies, since it establishes a new codification based on their strategic behaviour in the market. a structured method has been followed to obtain this new classification of the companies that relates the strategic behavior with the variables of growth in sales, sustainable growth and growth of the market demand. the combination of these variables has allowed six company categories (expansive, dominant, shrinking, restructuring, in decline, unfocused) to be established. an empirical study has been used, with a large sample of spanish companies, to verify that there is a relationship between profitability and the classification obtained from the companies’ strategic behavior in the market. therefore, the six categories obtained are not only suitable for classifying the evolution of companies in their market, but also their profitability. in addition, it has been highlighted that the usual classification of companies according to their sector of activity is not adequate for this purpose; nor have other company classifications been found in the state of the art that would allow this relationship to be obtained. on the other hand, the model for measuring the value added by intellectual capital (vaic) has been used for each of these categories of companies, obtaining relationships between some components of intellectual capital, strategic behaviour and corporate profitability. one result of the study is that declining companies have the lowest human and structural capital efficiency values, even though their employed capital efficiency is the highest. meanwhile, the companies that are in the process of restructuring and out of focus are those that have the lowest values of intellectual capital or vaic. the companies that are in a conquering business situation are the ones with higher values of the indicators of intellectual capital. it can also be concluded that it is companies in the energy, water and information technology sectors figure 5. hce, sce, cee, vaic by sector of economic activity. agriculture; construction; distribution & sales; energy & water; manufacturer; others; restaurant & hotels; information technology. figure 6. hce, sce, cee, vaic by state of growth. dominant; in decline; unfocused; expansive; shrinking; restructuring. int. j. prod. manag. eng. (2020) 8(1), 1-11creative commons attribution-noncommercial-noderivatives 4.0 international empirical study of the business growth strategy related to the added value by intellectual capital 9 http://creativecommons.org/licenses/by-nc-nd/4.0/ that add the most value to their intellectual capital processes, which is particularly relevant in terms of human capital. another conclusion is that the value added by intellectual capital (vaic) is higher in companies with sustainable growth higher than market growth (companies in business situation: conquering, explosive and shrinking). these companies have greater human and structural capital efficiencies than others, with technology and energy companies having the greatest added value of intellectual capital. consequently, this study combines data from companies in different sectors with indicators of intellectual capital (human, structural, relational) to establish new correlations between these intellectual components and the results of business management, from three points of view: financial management, product portfolio management, and commercial management. therefore, the new classification of companies obtained in this work is a starting point for new scientific studies on intellectual capital and the strategic behaviour of companies in the market. 6. acknowledgements the authors would like to acknowledge the grants t-cue doctoral prize for research on topics directly related to business needs, the grants of the phd extraordinary award and the support of the university of burgos to facilitate access to data gathering in relation to the work presented in the paper. references alcalde, r, manzanedo m a, sáiz-bárcena l (2016). empirical evidence for a classification of firm performance according to a sustainable growth rate model. book of proceedings international joint conference cio-icieom-iie-aim (ijc 2016). álvarez, h.f. (2008). la estructura de financiamiento y el crecimiento de las pequeñas y medianas empresas. https://doi. org/10.13140/2.1.2627.8084 amouzesh, n., moeinfar, z. and mousavi, z. (2011). sustainable growth rate and firm performance: evidence from iran stock exchange. international journal of business and social science, 2(23), 249-255. ashford, n. (2016). de-[constructing] growth. sustainability, 8(11), 1140. https://doi.org/10.3390/su8111140 chen, m.-c., cheng, s.-j., hwang, y. (2005). an empirical investigation of the relationship between intellectual capital and firms’ market value and financial performance. journal of intellectual capital, 6(2),159-176. https://doi.org/10.1108/14691930510592771 davidsson, p., achtenhagen, l., naldi, l. (2010). small firm growth. foundations and trends in entrepreneurship, 6(2), 69-166. https://doi.org/10.1561/0300000029 demir, r., wennberg, k., mckelvie, a. (2017). the strategic management of high-growth firms: a review and theoretical conceptualization. long range planning, 50(4), 431-456. https://doi.org/10.1016/j.lrp.2016.09.004 firer, s., mitchell williams, s. (2003). intellectual capital and traditional measures of corporate performance. journal of intellectual capital, 4(3), 348-360. https://doi.org/10.1108/14691930310487806 gan, k., saleh, z. (2008). intellectual capital and corporate performance of technology-intensive companies: malaysia evidence. asian journal of business and accounting, 1(1), 113-130. garcía muñiz, l.h. (2011). evidencia empírica de la evaluación del modelo de crecimiento sostenible de higgins en empresas mexicanas durante el periodo 1998-2008. (doctoral thesis). instituto politécnico nacional, méxico d.f. ghanei, h. ramezani kheibari, a. (2015) a survey of the relationship between intellectual capital and firm performance criteria. journal of scientific research and development, 2(7), 280-284. godet, m. (1994). from anticipation to action: a handbook of strategic prospective. paris, francia: unesco. greco, g., ferramosca, s., allegrini, m. (2014). exploring intellectual capital in family firms. an empirical investigation. international journal of learning and intellectual capital, 11(2), 91-106. https://doi.org/10.1504/ijlic.2014.060802 guerras, l.m., navas, e. (2015). la dirección estratégica de la empresa. teoría y aplicaciones. navarra, españa: aranzadi. gupta, m. (2015). intellectual capital & firm profitability: an empirical study on the it sector listed in nse. global journal of multidisciplinary studies, 4(4), 105-108. hajeb, h.r., moghaddam, a.g., alipour, s. (2015). an investigation of intellectual capital impact on companies’ market value and financial performace. academic journal of accounting and economic researches, 4(3), 131-139. haris, m., yao, h., tariq, g., malik, a., javaid, h. m. (2019). intellectual capital performance and profitability of banks: evidence from pakistan. journal of risk and financial management, 12(2), 56. https://doi.org/10.3390/jrfm12020056 higgins, r.c. (1977). how much growth can a firm afford?. financial management, 6(3), 7-16. https://doi.org/10.2307/3665251 int. j. prod. manag. eng. (2020) 8(1), 1-11 creative commons attribution-noncommercial-noderivatives 4.0 international alcalde-delgado et al. 10 https://doi.org/10.13140/2.1.2627.8084 https://doi.org/10.13140/2.1.2627.8084 https://doi.org/10.3390/su8111140 https://doi.org/10.1108/14691930510592771 https://doi.org/10.1561/0300000029 https://doi.org/10.1016/j.lrp.2016.09.004 https://doi.org/10.1108/14691930310487806 https://doi.org/10.1504/ijlic.2014.060802 https://doi.org/10.3390/jrfm12020056 https://doi.org/10.2307/3665251 http://creativecommons.org/licenses/by-nc-nd/4.0/ higgins, r.c. (2007). analysis of financial management. new york: mcgraw-hill. joshi, m., cahill, d., sidhu, j., kansal, m. (2013) intellectual capital and financial performance: an evaluation of the australian financial sector. journal of intellectual capital, 14(2), 264-285. https://doi.org/10.1108/14691931311323887 kabue, l. w., kilika, j. m. (2016). firm resources, core competencies and sustainable competitive advantage: an integrative theoretical framework. journal of management and strategy, 7(1), 98-108. https://doi.org/10.5430/jms.v7n1p98 kai wah chu, s., hang chan, k., wu, w.w. (2011). charting intellectual capital performance of the gateway to china. journal of intellectual capital, 12(2), 249-276. https://doi.org/10.1108/14691931111123412 kaplan, r.s. and norton, d.p. (2004). strategy maps: converting intangible assets into tangible outcomes. boston: harvard business school press. lambert, s. (2015). the importance of classification to business model research. journal of business models, 3(1),49-61 maditinos, d., chatzoudes, d., tsairidis, c.; theriou, g. (2011). the impact of intellectual capital on firms’ market value and financial performance. journal of intellectual capital, 12(1), 132-151. https://doi.org/10.1108/14691931111097944 magreta, j. (2001). la administración en la nueva economía. nuevas perspectivas. mexico: oxford university press. mahdjour, s. (2015). set up for growth? an exploratory analysis of the relationship of growth intention and business models. manchester: the international society for professional innovation management (ispim). https://doi.org/10.1142/s1363919615400095 mukherjee, t., sen, s. s. (2019). intellectual capital and corporate sustainable growth: the indian evidence. the east asian journal of business management (eajbm), 9(2), 5-15. https://doi.org/10.13106/jbees.2019.vol9.no2.5 najafizadeh, s.a., fordoei, e.j. (2014). investigation the effect of intellectual capital on market value and financial performance of listed companies in tehran stock exchange. global journal of management studies and researches, 1(3), 143-150. peñate santana, y. (2013). factores determinantes del crecimiento de las pymes: el papel de las capacidades dinámicas y la orientación emprendedora. (doctoral thesis). university of oviedo, spain. pew tan, h., plowman, d. and hancock, p. (2007). intellectual capital and financial returns of companies. journal of intellectual capital, 8(1), 76-95. https://doi.org/10.1108/14691930710715079 pulic, a. (1998). measuring the performance of intellectual potential in knowledge economy. paper presented at the 2nd mcmaster word congress on measuring and managing intellectual capital, mcmaster university, hamilton. pulic, a. (2000). vaic™–an accounting tool for ic management. international journal of technology management, 20(5-8), 702-714. https://doi.org/10.1504/ijtm.2000.002891 pulic, a. (2004). intellectual capital-does it create or destroy value?. measuring business excellence, 8(1), 62-68. https://doi. org/10.1108/13683040410524757 pulic, a. (2008). the principles of intellectual capital efficiency-a brief description. inspired by knowledge in organisations, 182 (1), 67-93. salazar, b. c., villegas, e. (2019). intellectual capital of industrial sector mexican companies measured through the vaic™ model. journal of administrative science, 1(1), 1-9. sallenave, j.-p. (1991). gerencia y planeación estratégica. bogotá, colombia: norma. sallenave, j.p. (2002). gerencia integral. bogotá, colombia: norma. sardo, f., serrasqueiro, z. (2018). intellectual capital, growth opportunities, and financial performance in european firms: dynamic panel data analysis. journal of intellectual capital, 19(4), 747-767. https://doi.org/10.1108/jic-07-2017-0099 schwab, l., gold, s., reiner, g. (2019). exploring financial sustainability of smes during periods of production growth: a simulation study. international journal of production economics, 212, 8-18. https://doi.org/10.1016/j.ijpe.2018.12.023 shiu, h. (2006), the application of the value added intellectual coefficient to measure corporate performance: evidence from technological firms, international journal of management, 23(2), 356-365. socolich mansilla, m.i. (2007). desempeño financiero y creación de valor en empresas chilenas fusionadas. (doctoral thesis). universidad de chile, chile. smriti, n., das, n. (2018). the impact of intellectual capital on firm performance: a study of indian firms listed in cospi. journal of intellectual capital, 19(5), 935-964. https://doi.org/10.1108/jic-11-2017-0156 svanadze, s., kowalewska, m. (2015). the measurement of intellectual capital by vaic method example of wig20. online journal of applied knowledge management, 3 (2), 36-44. ulum, i., ghozali, i. and purwanto, a. (2014). intellectual capital performance of indonesian banking sector: a modified vaic (m-vaic) perspective. asian journal of finance & accounting, 6(2), 103-123. https://doi.org/10.5296/ajfa.v6i2.5246 xu, j., wang, b. (2018). intellectual capital, financial performance and companies’ sustainable growth: evidence from the korean manufacturing industry. sustainability, 10(12), 4651. https://doi.org/10.3390/su10124651 zia, m., muhammad, h., arbab, s., shahzad, a., bilal, s. (2014). vaic and firm performance: banking sector of pakistan. information and knowledge management, 4(2), 100-107. int. j. prod. manag. eng. (2020) 8(1), 1-11creative commons attribution-noncommercial-noderivatives 4.0 international empirical study of the business growth strategy related to the added value by intellectual capital 11 https://doi.org/10.1108/14691931311323887 https://doi.org/10.5430/jms.v7n1p98 https://doi.org/10.1108/14691931111123412 https://doi.org/10.1108/14691931111097944 https://doi.org/10.1142/s1363919615400095 https://doi.org/10.13106/jbees.2019.vol9.no2.5 https://doi.org/10.1108/14691930710715079 https://doi.org/10.1504/ijtm.2000.002891 https://doi.org/10.1108/13683040410524757 https://doi.org/10.1108/13683040410524757 https://doi.org/10.1108/jic-07-2017-0099 https://doi.org/10.1016/j.ijpe.2018.12.023 https://doi.org/10.1108/jic-11-2017-0156 https://doi.org/10.5296/ajfa.v6i2.5246 https://doi.org/10.3390/su10124651 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering special issue: advances in engineering networks https://doi.org/10.4995/ijpme.2019.10795 received: 2018-10-10 accepted: 2019-05-01 methodology to validate results from european research projects: the c2net case study sanchis, r.a1*, andrés, b.a2, and poler, r.a3 aresearch centre on production management and engineering (cigip). universitat politècnica de valència. calle alarcón, 1 03801 alcoy (alicante), spain. a1 rsanchis@cigip.upv.es, a2 bandres@cigip.upv.es, a3 rpoler@cigip.upv.es abstract: one of the main priorities of the european commission is the utilisation of european projects results in further research activities, or in developing, creating and marketing a product or process. for this reason, it is critical to test and validate european projects results before implementing them in real scenarios. in this paper, a general validation methodology addressed to the assessment of technological results has been defined. this general methodology offers the foundations to define specific validation methodologies to validate particular results of different research projects. as an example, the general methodology has been applied to define a specific one for the validation of an optimiser developed within the european research project: cloud collaborative manufacturing networks (c2net) to guarantee the proper operation of the research results and facilitate their later real implementation and exploitation. key words: optimisation, c2net, validation, methodology, optimiser. 1. introduction in the global economy, there is an increasing interest in new organisational structures to flexible enough respond to market changes and at the same time to perform collaborative projects (andrés et al., 2015). enterprises, especially european smalland medium-sized ones (smes), do not have access to advanced management systems and collaborative tools due to their restricted resources. smes manufacturing value chains are distributed and dependent on complex information and material flows requiring new approaches to reduce the complexity of manufacturing management systems. motivated by this situation the european project: “cloud collaborative manufacturing networks” (c2net) was born. c2net is a european-funded project whose main goal is to create cloud-enabled tools to support the smes supply network optimisation of manufacturing and logistics assets based on collaborative demand, production and delivery plans (andrés et al., 2016). c2net has the characteristics of an “industry cloud,” in which groups of companies within a sector come together to leverage enhanced technologies and share best practices in a pooledcost environment. c2net develops solutions to help smes optimize their manufacturing and logistic supply chains by reducing the complexity currently surrounding manufacturing management systems. moreover, it offers a platform on which products, processes and logistical data can be securely stored and shared in the cloud (black, 2017). sanchis et al., (2018) offers a brief description of the main four exploitable results (figure 1): (i) the data collection framework (dcf) in charge of gathering the necessary data for the other c2net modules (agostinho et al., 2016 and mohammed to cite this article: sanchis, r., andrés, b., and poler, r. (2019). methodology to validate results from european research projects: the c2net case study. international journal of production management and engineering, 7(special issue), 81-90. https://doi.org/10.4995/ijpme.2019.10795 int. j. prod. manag. eng. (2019) 7(special issue), 81-90creative commons attribution-noncommercial-noderivatives 4.0 international 81 http://creativecommons.org/licenses/by-nc-nd/4.0/ et al., 2017); (ii) the optimizer (opt) whose main goal is the optimisation of several planning problems such as aggregate planning problems (ap), material requirement planning (mrp) problems, … (sanchis et al., 2018); (iii) the collaboration tools (cot) to provide the necessary means to collaboratively solve agility issues in the supply chain (benaben et al., 2016) and (iv) the c2net platform that hosts all the previous c2net modules (ramis-ferrer et al., 2016). the dcf main objective is to enable the continuous data collection from both legacy and internet of things (iot) systems. data is collected from supply network partners through dedicated middleware at the company side and stored at the c2net cloud platform in a reference form for proper sharing among remaining c2net components, to enable enlarged partners’ collaboration and optimisation of resources. the opt module solves planning problems and maximises the efficiency of supply network planning activities by computing production, replenishment and delivery plans to achieve shorter delivery times, faster speed and better consistency of schedules, better use of productive resources and more energy savings. the cot module includes mainly six functional components: (i) the knowledge base in charge of structuring the collected data and information through the dcf, (ii) the modelling service in charge of formalizing the collaborative situation, (iii) the detection service dedicated to monitor the collaboration and detect if there is any unwished situation, (iv) the adaptation service, which suggests resolution processes in case of deviation (v) the assessment service in charge of the evaluation of the deviations, and (vi) the orchestration service in charge of supporting the design and orchestration of the collaborative processes. and finally, the c2net cloud-based platform that hosts the previous c2net main three modules dcf, opt and cot; and allows a secure and user control access to the c2net features. all these results have been tested in advance to their implementation in real companies. however, tests should be planned and structured in a proper way to record the results of such assessment and verify that such results accomplish with the expected objectives. this task often is performed on the fly without formal procedures defined. for this reason, the main objective of this paper is to define a general methodology to assess research results, more specifically technological ones, before their real implementation to guide researchers and developers in this activity. moreover, an application of this methodology to the c2net opt is shown as an example. this paper is organised as follows. section 2 is focused on the c2net opt module characterisation as the validation methodology described in this paper is applied to this result. section 3 describes the main phases to test and validate the research and technological results before implementing them in real scenarios. based on this, section 4 shows the application of the general validation methodology to assess the c2net optimiser module. finally, section 5 provides the main conclusions. figure 1. c2net project main results. int. j. prod. manag. eng. (2019) 7(special issue), 81-90 creative commons attribution-noncommercial-noderivatives 4.0 international sanchis et al. 82 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. the c2net optimizer module the opt module provides advanced optimisation algorithms for single and collaborative computation of production, replenishment and delivery plans with the aim of optimising the use of manufacturing and logistics assets of the supply network from a holistic point of view. the opt provides decision-makers with a set of tools to easily manage the decision rules and to re-calculate alternative plans in real time, increasing the efficiency of the supply network by the global optimisation of operations plans and schedules. the c2net opt is composed of four main components (figure 2): (i) the optimisation algorithms, a set of algorithms, to support manufacturing networks in the optimisation of manufacturing and logistics processes. currently, the c2net opt module has 48 different optimisation algorithms addressed to solve planning problems such as production sequencing, goods delivery, material requirement planning… (ii) the solver manager in charge of managing algorithms, and as such, it allows to create, edit, categorize and delete algorithms and objective functions. moreover, it makes available to the components (iii) and (iv) (opc and poma), the set of methods that allow them to validate if an optimisation problem can be solved or which is the most appropriate algorithm to be applied depending on some criteria such as gap, solving time, etc.; (iii) the optimisation problem configurator (opc) in charge of managing the optimisation problems and proposing solutions to optimise manufacturing and logistics plans based on the input data available; and (iv) the poma (processes optimisation of manufacturing assets) manager that controls the launching of optimisation problems defined in the opc. the opt module interacts with other modules developed within the project. one of this modules is the data collection framework (dcf) from which the opt gathers the necessary input data to perform the optimisation and also returns the results to the dcf to be displayed to the users. the necessary input data to perform the optimisation come from different resources and each company has these input data available using different terminology. for this reason, the dcf hosts the standardised data model, called stables. the objective of the stables, besides storing the input data for optimisation, is standardizing them to use always the same term for the same concept. in the example shown in figure 3, there are 3 companies that use different concepts (inventory, stock, product stored…) to denote the number of products available in the inventory. these three concepts have the only name in the stables optimisation algorithms solver manager opc poma algorithms selection algorithms registration calculation /computation objective function launching opt module input data and structure output and structure other c2net modules figure 2. components and relationships among the opt module components (based on sanchis et al., 2018). int. j. prod. manag. eng. (2019) 7(special issue), 81-90creative commons attribution-noncommercial-noderivatives 4.0 international methodology to validate results from european research projects: the c2net case study 83 http://creativecommons.org/licenses/by-nc-nd/4.0/ of “availabilityamount” and the optimisation algorithms use this concept for managing the data. in the same way, the optimisation’s results are stored in the plan data model, also called ptables that is hosted in the dcf. following with the example of figure 3, one of the results of the optimisation performed is focused on planning the number of products to be manufactured in each period. this result is coined in the ptables as “normaloperationamount”, while the enterprises receive this piece of information with their own terminology that is “production” for company 1, “products manufactured” for company 2 and “units” for company 3. therefore, both models: stables and ptables have twofold objectives: (i) to store the input (stables) and output (ptables) data sets and (ii) to standardise concepts to facilitate the optimisations while companies provide the data and receive the optimisation information in their “own language”. 3. general validation methodology one of the main priorities of the european commission is the utilisation of european projects results in further research activities other than those covered by the action concerned, in developing, creating and marketing a product or process, in creating and providing a service, or in standardisation activities (european helpdesk, 2018). therefore, in order to be able to exploit the results of european projects, it is required to test and validate that the results achieved are suitable and appropriate before implementing them in real scenarios. one of the most common risks detected in european projects consists of difficulties in the implementation of the project results in real companies through piloting activities as they are not mature enough. for this reason, it is vital to test the viability and the correct running of the results before performing real tests in real companies. for this reason, the c2net project defines two types of validation methodologies. one that is based on the validation of the project results from a technical, technological and scientific point of view and once that this methodology is fully completed, then the validation methodology is applied to industrial companies. some phases of both validation methodologies are performed in parallel, however, it is important to implement results in the industry once that have been already fully proven. this paper describes the first of the validation methodologies as, although projects usually validate their results prior to real implementation, they do not usually define standardized procedures for this type of validations. the phases of this methodology have been defined as generalist as possible and for this reason it is considered a meta-methodology, i.e., a high level methodology. taking this into account, it is worth title 4 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain the opt module interacts with other modules developed within the project. one of this modules is the data collection framework (dcf) from which the opt gathers the necessary input data to perform the optimisation and also returns the results to the dcf to be displayed to the users. the necessary input data to perform the optimisation come from different resources and each company has these input data available using different terminology. for this reason, the dcf hosts the standardised data model, called stables. the objective of the stables, besides storing the input data for optimisation, is standardizing them to use always the same term for the same concept. in the example shown in figure 3, there are 3 companies that use different concepts (inventory, stock, product stored…) to denote the number of products available in the inventory. these three concepts have the only name in the stables of “availabilityamount” and the optimisation algorithms use this concept for managing the data. in the same way, the optimisation’s results are stored in the plan data model, also called ptables that is hosted in the dcf. following with the example of figure 3, one of the results of the optimisation performed is focused on planning the number of products to be manufactured in each period. this result is coined in the ptables as “normaloperationamount”, while the enterprises receive this piece of information with their own terminology that is “production” for company 1, “products manufactured” for company 2 and “units” for company 3. therefore, both models: stables and ptables have twofold objectives: (i) to store the input (stables) and output (ptables) data sets and (ii) to standardise concepts to facilitate the optimisations while companies provide the data and receive the optimisation information in their “own language”. 3 general validation methodology one of the main priorities of the european commission is the utilisation of european projects results in further research activities other than those covered by the action concerned, in developing, creating and marketing a product or process, in creating and providing a service, or in standardisation activities (european helpdesk, 2018). therefore, in order to be able to exploit the results of european projects, it is required to test and validate that the results achieved are suita figure 3. input and output data management in c2net. companies’ terms optimisation availability amount stables normaloper ationamount ptables production products manufactured units company 1 inventory company 3 company 2 stock product stored company 1 company 3 company 2 in pu t d at a o ut pu t d at a figure 3. input and output data management in c2net. int. j. prod. manag. eng. (2019) 7(special issue), 81-90 creative commons attribution-noncommercial-noderivatives 4.0 international sanchis et al. 84 http://creativecommons.org/licenses/by-nc-nd/4.0/ mentioning that this high level methodology has been applied to validate specifically the c2net opt module, that is a technological result. in other cases, where the results are of different nature, the methodology possibly needs some adaptation. the value of this high level methodology is based on the fact that the research results of a research project are usually developed by researchers who are experts in developing research solutions but are less familiar with validating and testing such results. that is why this methodology provides a high level perspective to support novel, junior, or even senior researchers with no previous experience in the validation phases, since they are more focused on the development than on assessing and putting the results on place. the general phases defined to be followed in order to validate technological results are descried as follows: phase 1. check the input and output data consistency in order to assess the correct operation of a technological result, it is necessary that the result is nurtured from the appropriate input information. this means that the necessary input data should be easily accessible and with a reasonable cost. other aspects to be analysed are the stability, uniformity, reliability of data in the long term. one of most common risk when a technological result is evaluated is related to the fact that enterprises wishes a specific output information but they are not able to provide the necessary input data to obtain the desired output information. in other cases, the information exists, but it is very costly to obtain it and its cost exceeds the benefits that the output information will provide to the enterprise. therefore, it is important, in the first phases of a technological result development, to understand perfectly what the company needs but also which information the company may provide and if it is economical viable to obtain it. in this phase it is also recommendable to analyse that the company has data records over time and the information is reliable, that it is standard and uniformly stored in the company’s information systems. it is very common to find gaps of information during some specific periods, or sometimes, the information is registered in different units, terms… what complicates the identification and collection of the input data that the technological result needs for running properly. the same applies for the output data in the sense that the solutions that the technical result offers should be in line with what the company was expecting. in this case, the communication and collaboration between the companies’ users and the researchers and developers is essential during the phases of the technological result development. moreover, in this phase it should be analysed that the outcomes offered by the technological result are easily understandable and consistent with the company’s information systems. phase 2. create realistic data sets. before performing real tests in companies, it is recommendable to create data sets similar to the ones that the company manages. in this way, researchers are simulating the normal operation of the technological result but with fake data in order not to focus efforts on obtaining real data that sometimes is complex and time-consuming. it is important to simulate real scenarios to study the behaviour of the technological result. otherwise, the test will not be representative. in case that the technological result does not need input data for operating, realistic conditions under which the result has to work, should be simulated. phase 3. identify the aspects to be analysed depending on the type of technological result, the aspects to be assessed will be different. for example, if the technological result is related to a videoconference system that allows synchronous communication between two parties, the test will be focused, among other aspects, on the speed, sound, image quality… it is important to define which aspects should be analysed to guarantee the proper functioning of the result. moreover, it may happen that some aspects are key for the correct operation of the technological result and therefore their validation should be a priority, while other aspects can be classified as recommendable, and in this case, its validation will be useful but not critical. besides the definition of the aspects to be analysed, in this phase the target values for each aspect should be also set up. following with the videoconference example, it is necessary not only to define that the image quality should be analysed but also to stablish the target values for each characteristic (e.g. number of pixels: 3840×2160 pixels; luminance: 550 cd/m²; fps: 100 hz…). phase 4. monitor the computation/running of the technological result this phase is focused on the proper functioning and running of the technological result. here, it is important to check that during the technological result int. j. prod. manag. eng. (2019) 7(special issue), 81-90creative commons attribution-noncommercial-noderivatives 4.0 international methodology to validate results from european research projects: the c2net case study 85 http://creativecommons.org/licenses/by-nc-nd/4.0/ running, the target values for each characteristic and aspect (defined in phase 3) are achieved. moreover, it is also recommendable to perform tests under different circumstances (e.g. if the technological result needs internet connection, it should be advisable to test it with both cable and wireless connection) in order to check the proper running of the technological results in different situations. phase 5. analyse the output data and results the technological result could have different purposes. for this reason, this phase could vary depending on the function of the result. anyhow, most of the research and technological results offer output information. if this is the case, it is also necessary to assess that the information obtained is the desired one and also congruent. for example, if the technological result that we are analysing is a iot (internet of things) hub, whose main purpose is to collect data from different physical resources (machines, vehicles…), this phase should focus on the correctness of the information that the hub has gathered. the information gathered should be exactly what the user was expecting and presented and organized as the user wishes. on the contrary, the causes by which the technological result does not offer the output information according to the users’ requirements should be investigated. but this phase is also focused on the analysis of the congruency of the output information. for example, if the user is expecting that the technological result offers information about when to order raw material to fulfil a specific order that should be delivered before 4th october, and the result provides a later ordering date, this solution is not consistent as the order of raw material should be done before 4th october. therefore, when testing, it is recommendable to test under the conditions in which we know the different potential solutions to analyse the congruency of the output information. phase 6. propose solutions for undesired situations undesired situations consist of any situation in which the technological result, or the input or output data present any deviation with regards to the expected ones. when the behaviour of the technological result is not the expected, it is advisable to propose solutions in order to return to the normal status of operation. following with the example of the ordering date of raw material, if the solution is not feasible, the causes should be investigated and, once that the reason why the technological result is not offering the correct information, this situation should be corrected. in this example, the cause was related to a wrong definition of the calendar and this was why the order date was defined after the delivery date. sometimes, it is complex to find out the reason that causes the undesired situation. for this reason, in this phase, it is advisable to form multidisciplinary groups, formed even by end users, to detect the cause of the undesired situation as soon as possible, whereas the solution to the undesired situation is normally fixed by researchers and developers. phase 7. identify new opportunities sometimes when a problem arises, and during the search for the solution, improvements and new opportunities are identified to enhance the technological result. innovation arises for many reasons, but one of them is during the search for solutions to existing problems. therefore, during the development of solutions to solve undesired situations, developers should be also receptive to new ideas that could improve the technological result, or even though, could promote the development of another new and innovative technological result. as aforementioned, the methodology offers a set of systematic procedures as guidelines oriented to support researchers and developers who have to validate technological results. depending on the nature of the results, the phases to be followed should need some adaptation, but these general phases serves as first attempt to develop more specific validation procedures. 4. specific validation methodology of the c2net optimiser module following the phases and guidelines defined in the methodology of section 3, this section shows the specific methodology that is created within the european project c2net in order to validate the c2net opt before implementing and validating it in real scenarios. figure 4 shows the alignment between the phases of the general validation methodology and the particular steps defined to validate the c2net opt. as it can be seen in this figure, not all the general phases are covered by the steps of the specific methodology. this is due to the fact that phase 4, that consists of monitoring the computation/running of the technological result, is int. j. prod. manag. eng. (2019) 7(special issue), 81-90 creative commons attribution-noncommercial-noderivatives 4.0 international sanchis et al. 86 http://creativecommons.org/licenses/by-nc-nd/4.0/ intrinsically developed in steps 3 and 4, since if the gap is identified (see step 3) and the solving time is determined (see step 4) is because the optimisation has been performed. however, it would have been advisable to include a new step, before step 3, focused on analysing the smooth running of the c2net opt. the same occurs with phase 6, as in each of the steps defined in the specific validation methodology, if problems are detected, they are solved in each of the steps where they are found. there is not a separate step in which all the problems detected are analysed for proposing the most adequate solutions but, in all the steps, when a problem is identified, developers also focus on solving such a problem. however, it is worth mentioning that it would also have been advisable to do it in a single step to encourage the collaboration in the problems solving process what fosters innovation. for this reason, as aforementioned, the general validation methodology has been defined as generalist as possible and it can be customized depending on the developers’ requirements and/or the nature of the technological result to be assessed. the work performed by the opt consists of defining an optimisation problem. to compute the optimisation, it is necessary a set of input data sets (stables) and a suitable algorithm from the set of optimisation algorithms developed and managed title 8 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain for this reason, as aforementioned, the general validation methodology has been defined as vand/or the nature of the technological result to be assessed. the work performed by the opt consists of defining an optimisation problem. to compute the optimisation, it is necessary a set of input data sets (stables) and a suitable algorithm from the set of optimisation algorithms developed and managed in the solver manager. the poma manager is in charge of the optimisation and providing the set of outputs/results (ptables), as it is shown in figure 5. the validation methodology addressed to the c2net opt consists of 8 main steps: step 1. check that optimisation algorithms are consistent with stables. the input data needed to perform the optimisation through the computation of the algorithms needs to be aligned with the needs of input information required by such optimisation algorithms. for this reason, the input data hosted in the stables and the necessary input information that needs the optimisation algorithms to perform the optimisation is mapped to detect any inconsistency. table 1 shows a small example of this mapping. step 2. create realistic data sets for the validation of the optimisation algorithms. three size types of input data sets are created: small (sds), medium (mds) and large (lds) data sets. each of them is created considering realistic input values that offer realistic output results: sds are used in a first stage for the computation of the optimisation algorithms and are composed by the minimum amount of input data sets necessary to generate feasible, logical, and figure 4. alignment between the general validation methodology and the specific one applied for the validation of the c2net opt. phase 1. check the input and output data consistency phase 2. create realistic data sets phase 3. identify the aspects to be analysed phase 4. monitor the computation/running of the technological result phase 7. identify new opportunities phase 5. analyse the output data and results phase 6. propose solutions for undesired situations general validation methodology opt validation methodology steps 1 and 5 step 2 step 6 steps 3 and 4 steps 7 and 8 figure 5. work performed by the c2net opt. c2net input data stables optimisation problem definition opc suitable optimisation algorithm solver manager optimisation poma manager c2net output data ptables company output data company database company input data company database figure 4. alignment between the general validation methodology and the specific one applied for the validation of the c2net opt. title 8 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain for this reason, as aforementioned, the general validation methodology has been defined as vand/or the nature of the technological result to be assessed. the work performed by the opt consists of defining an optimisation problem. to compute the optimisation, it is necessary a set of input data sets (stables) and a suitable algorithm from the set of optimisation algorithms developed and managed in the solver manager. the poma manager is in charge of the optimisation and providing the set of outputs/results (ptables), as it is shown in figure 5. the validation methodology addressed to the c2net opt consists of 8 main steps: step 1. check that optimisation algorithms are consistent with stables. the input data needed to perform the optimisation through the computation of the algorithms needs to be aligned with the needs of input information required by such optimisation algorithms. for this reason, the input data hosted in the stables and the necessary input information that needs the optimisation algorithms to perform the optimisation is mapped to detect any inconsistency. table 1 shows a small example of this mapping. step 2. create realistic data sets for the validation of the optimisation algorithms. three size types of input data sets are created: small (sds), medium (mds) and large (lds) data sets. each of them is created considering realistic input values that offer realistic output results: sds are used in a first stage for the computation of the optimisation algorithms and are composed by the minimum amount of input data sets necessary to generate feasible, logical, and figure 4. alignment between the general validation methodology and the specific one applied for the validation of the c2net opt. phase 1. check the input and output data consistency phase 2. create realistic data sets phase 3. identify the aspects to be analysed phase 4. monitor the computation/running of the technological result phase 7. identify new opportunities phase 5. analyse the output data and results phase 6. propose solutions for undesired situations general validation methodology opt validation methodology steps 1 and 5 step 2 step 6 steps 3 and 4 steps 7 and 8 figure 5. work performed by the c2net opt. c2net input data stables optimisation problem definition opc suitable optimisation algorithm solver manager optimisation poma manager c2net output data ptables company output data company database company input data company database figure 5. work performed by the c2net opt. int. j. prod. manag. eng. (2019) 7(special issue), 81-90creative commons attribution-noncommercial-noderivatives 4.0 international methodology to validate results from european research projects: the c2net case study 87 http://creativecommons.org/licenses/by-nc-nd/4.0/ in the solver manager. the poma manager is in charge of the optimisation and providing the set of outputs/results (ptables), as it is shown in figure 5. the validation methodology addressed to the c2net opt consists of 8 main steps: step 1. check that optimisation algorithms are consistent with stables. the input data needed to perform the optimisation through the computation of the algorithms needs to be aligned with the needs of input information required by such optimisation algorithms. for this reason, the input data hosted in the stables and the necessary input information that needs the optimisation algorithms to perform the optimisation is mapped to detect any inconsistency. table 1 shows a small example of this mapping. step 2. create realistic data sets for the validation of the optimisation algorithms. three size types of input data sets are created: small (sds), medium (mds) and large (lds) data sets. each of them is created considering realistic input values that offer realistic output results: sds are used in a first stage for the computation of the optimisation algorithms and are composed by the minimum amount of input data sets necessary to generate feasible, logical, and valid solutions. they are input data sets that are simple enough to allow developers and researchers to detect any potential problem and challenge related to the development of the algorithms and its resolution. subsequently, mds and lds are created through the extension of the sds. similarly to sds, mds and lds are also input data sets that generate feasible and logical solutions. mds include, on average, 11-40 products and periods and are used to test and validate if the solving time used to offer the results to the enterprises is reasonable (see step 4), and also to obtain the gap (see step 3). lds comprise more than 40 products and periods, and are large enough to validate the algorithms with amounts of data which are similar to the ones that will be used by the companies when solving their plans through the computation of the developed algorithms. step 3. identify the gap for the algorithms using realistic data sets. the gap is the difference between the result given by the optimal solution of a plan and the result provided by the algorithm. the realistic data sets are large enough to represent a consistent amount of input data so that the gap computed for the algorithms is representative. the main drawback when computing the gap is that for some of the developed algorithms there are no optimal algorithms or models, therefore the optimal solution is not known, hence the gap cannot be accurately computed. not all algorithms are optimiser, which means that for the heuristic and metaheuristic algorithms, the gap is difficult to find out. step 4. determine the solving time of the optimisation algorithms. the solving time is the time required to solve an optimisation problem in a successful scenario, i.e., in the case where a feasible solution is found. realistic data sets are used as input data to identify the solving time. the solving time is one of the parameters that characterises the optimisation algorithm to be selected when the solver manager module needs to solve a specific plan. step 5. generate and unify the ptables. stables correspond to the input data tables required by the optimisation algorithms in order to compute plans (defined optimisation problems), while plan tables (ptables) are the tables for the results or output data. ptables contain the results of the plans after table 1. analysis of the consistency of the input data needed by the optimisation algorithms and the stables. algorithmid type1 table field consistent2 1 id part partid ü 1 id part availabilityamount ü 1 id part availabilitycost ü 1 id part availabilitymaximumamount ü … … … … … 10 id production productionid ü 10 id production productiondate ü 10 id production_part productionid ü … … … … … 1. id: input data; od: output data. 2. consistent between the stables and optimisation algorithms. int. j. prod. manag. eng. (2019) 7(special issue), 81-90 creative commons attribution-noncommercial-noderivatives 4.0 international sanchis et al. 88 http://creativecommons.org/licenses/by-nc-nd/4.0/ the execution of the optimisation algorithms. the ptables are built following the same logic as the used in the stables. the output data required by the enterprise should be aligned with the results offered by the optimisation algorithms and hosted in the ptables. table 2 shows an example of the consistency of the optimisation algorithms with the ptables. step 6. check the validity and feasibility of the results. after the computation of the optimisation, researchers and developers should check if the results offered by the optimisation algorithms (ptables) are valid and feasible. according to the inputs (stables), the output values given by the optimisation algorithms are analysed in order to identify potential infeasibilities or whether the solution provides a bizarre result that cannot be accepted by the company. an example of lack of validity and feasibility would be when the input data shows that the company has enough resources and capacity to produce a specific product, but the order to manufacture such products is delayed. in this case, the results are not validated and the optimisation algorithms definition should be reviewed to detect the issue that is causing the unsuitable solution. step 7. modify current functions or develop new ones to make the optimisation algorithms more efficient and fast. new functions or extensions of the already developed algorithms are designed in order to (i) make algorithms more efficient and reduce the solving time; (ii) generate feasible solutions when infeasible results were previously generated; (iii) reduce the gap of the algorithm; or (iv) include new restrictions that the companies require for their plans resolution. step 8. identify potential new requirements and develop new algorithms. in case companies define a new plan (optimisation problem) that was not previously considered, and moreover, no algorithms are available to compute the plan, new algorithms should be developed. to do so, researchers and developers will seek on previous developed algorithms in order to identify similarities. not starting from scratch saves time. in case there are no algorithms that can be adjusted (e.g., by adding new restrictions), developers will look in the literature for suitable developed algorithms that can be used/ adapted to solve the plan defined by the company. if no algorithms are found, new algorithms need to be defined. 5. conclusions before implementing the results developed within research projects in real companies, it is very important to validate them to guarantee that the results are mature enough to be applied in real scenarios. however, despite the importance of this fact, there are few research projects that define formal procedures that guide this prior validation. for this reason, a general validation methodology addressed to the assessment of technological results developed within european research projects has been defined. this methodology has been set up as generalist as possible to be applied to a wide range of results. however, depending on the nature of the table 2. analysis of the consistency of the output data required by the enterprise and the ptables. algorithmid type1 table field consistent2 1 od s_mrp_a partid ü 1 od s_mrp_a periodid ü 1 od s_mrp_a availabilityamount ü 1 od s_mrp_a batchamount ü … … … … … 10 od m_psc_a productionid ü 10 od m_psc_a personid ü 10 od m_psc_a periodid ü 10 od m_psc_a labourid ü 10 od m_psc_a operationtime ü … … … … … 1 id: input data; od: output data 2 consistent between the ptables and optimisation algorithms int. j. prod. manag. eng. (2019) 7(special issue), 81-90creative commons attribution-noncommercial-noderivatives 4.0 international methodology to validate results from european research projects: the c2net case study 89 http://creativecommons.org/licenses/by-nc-nd/4.0/ research results, maybe this methodology needs some adaptation. the general methodology has offered the foundations to define the specific validation methodology used to assess a result generated within the european project: c2net, that is the opt module. this application shows that the general methodology could be customized depending on the researchers and developers’ requirements and also the characteristics of the results to be validated. therefore, the general methodology offers broad guidelines to support researchers in the definition of specific validation methodologies to validate research results and guarantee the creation and feasibility of a commercial product beyond the project. acknowledgements the research leading to these results is in the frame of the “cloud collaborative manufacturing networks” (c2net) project which has received funding from the european union’s horizon 2020 research and innovation programme under grant agreement no. 63690. references agostinho, c., pereira, j., lorre, j.p., ghimire, s., and benazzouz, y. (2016). a distributed middleware solution for continuous data collection in manufacturing environments. in enterprise interoperability in the digitized and networked factory of the future, 159-166. andres, b., saari, l., lauras, m., and eizaguirre, f. (2016). optimisation algorithms for collaborative manufacturing and logistics processes. in enterprise interoperability in the digitized and networked factory of the future, 167-173. andres, b., sanchis, r., and poler, r. (2016). a cloud platform to support collaboration in supply networks. international journal of production management and engineering, 4(1), 5-13. https://doi.org/10.4995/ijpme.2016.4418 benaben, f., jiang, z., wang, t., lamothe, j., agostinho, c., ghimire, s., melo, r., benazzouz, y., and lorré, j.p. (2016). c2net project. tools for agile collaboration. in enterprise interoperability in the digitized and networked factory of the future, 174-180. black, d. (2017). collaborative cloud under development for manufacturers in europe. available at: https://www.enterprisetech.com/ 2017/04/13/collaborative-cloud-development-smes-manufacturers-europe/ european helpdesk. (2018). available at: https://www.iprhelpdesk.eu/glossary/exploitation mohammed, w.m., ferrer, b.r., martinez, j.l., sanchis, r., andres, b., and agostinho, c. (2017). a multi-agent approach for processing industrial enterprise data. in international conference on engineering, technology and innovation (ice/itmc), 1209-1215. https://doi.org/10.1109/ice.2017.8280018 ramis-ferrer, b., de juan-marín, r., miedes, e., nieto, a., martínez lastra, j.l., and peña-ortiz, r. (2016). towards a cloud-based platform for enabling supply chain collaboration. in enterprise interoperability in the digitized and networked factory of the future, 152-158. sanchis, r., andrés, b., poler, r., mula, j., and díaz-madroñero, m. (2018). the c2net optimisation solution. dirección y organización, 64, 36-41. int. j. prod. manag. eng. (2019) 7(special issue), 81-90 creative commons attribution-noncommercial-noderivatives 4.0 international sanchis et al. 90 https://doi.org/10.4995/ijpme.2016.4418 https://www.enterprisetech.com/2017/04/13/collaborative-cloud-development-smes-manufacturers-europe/ https://www.enterprisetech.com/2017/04/13/collaborative-cloud-development-smes-manufacturers-europe/ https://www.iprhelpdesk.eu/glossary/exploitation https://doi.org/10.1109/ice.2017.8280018 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.12271 received: 2019-08-27 accepted: 2020-05-28 understanding and representation of organizational training programs and their evaluation ruiz, m. a1*, igartua, j.i.b, mindeguia, m.c, orobengoa, m.a2 a isea s.coop. (mondragon corporation) goiru, 7-20500 mondragón, spain. b mechanical and manufacturing department, mondragon university, loramendi 4-20500 mondragón, spain. c piarres larzabal kolegioa, 3 eugène corre etorbidea, 64500 ziburu, france. a1 mruiz@isea.eus, b jigartua@mondragon.edu c maitane.mindegia@plk.eus, a2 morobengoa@isea.eus abstract: the evaluation of the organizational training is a necessary strategy to guarantee the quality of the training activities in organizations. this research had as an initial objective the development of the first phase of a project funded by the eit raw materials in the “call for kava education projects”. it officially started on january 2019. the main aim of this research was to define a training program for workers of a specific industrial sector and to evaluate the impact of the skill acquisition of workers through the training program. this paper presents the initial part of the project, the authors were part of the team at that stage. this phase was helpful to obtain the resulting conceptual model from the analysis of the variables involved in an effective learning process. the research tool used for variable identification were three group model building (gmb) sessions with the partners of the project. the resulting model of this paper was helpful to represent through systems thinking the phases of a learning process and its evaluation. key words: organizational training, training, training evaluation, training program, group model building, systems thinking. 1. introduction human intangible resources are essential, as they fit perfectly the necessary conditions to generate competitive advantage. these resources are valuable, scarce, and difficult to be imitated or replaced. the training processes within this human based paradigm are considered key, as the evolution of the company on its own could devalue the worker competencies. frequently, the competencies that generated competitive advantage in the past, are not useful tools in the present to solve the current necessities of the organization. consequently, knowledge and skills of workers are devalued, and it is necessary to apply human resource management policies to ensure the durability of human labour in the company. this durability which is dependent on factors such as experience, skills, abilities, or capacity to adapt are the elements that bring to the company the sustainable competitive advantage necessary to be successful (del valle and castillo, 2005). as a result, training must be understood as a vital part for the definition of their organization rather than a function or area of the organization (sarramona et al., 1994); in addition, this definition must be related to the rest of the departments for the development and success of the organization (gasalla, 2003). thus, the final purpose of training is to empower the individual to conduct properly an specific task or work resulting from a technological change, to look for new ways of organizing, new conditions, to cite this article: ruiz, m., igartua, j.i., mindeguia, m., orobengoa, m. (2020). understanding and representation of organizational training programs and their evaluation. international journal of production management and engineering, 8(2), 99-109. https://doi.org/10.4995/ijpme.2020.12271 int. j. prod. manag. eng. (2020) 8(2), 99-109creative commons attribution-noncommercial-noderivatives 4.0 international 99 https://orcid.org/0000-0001-6900-4382 mailto:jganzarain@mondragon.edu mailto:morobengoa@isea.eus http://creativecommons.org/licenses/by-nc-nd/4.0/ new tasks to do or to delete the existent deficiencies. acquisition of new knowledge, tools and techniques is an alive activity that must be feed continuously during the training programs (bassi et al., 2002). organizations need a set of methods to canalize and transfer the expected knowledge for the development of their work. in this way, workers will be able to develop their work in an efficient and profitable way. organizations may turn training into a constant and integrated habit in the daily activities of the workers. as a result, training and work must be merged rather than compatible (del valle and castillo, 2005). due to the impact of training programs for the competitiveness of the organizations, it could be stated that the main aim of this research was to analyse and represent using systems thinking philosophy, the process of evaluation of training programs to define them in a more effective and profitable way. 2. objective the objective of the evaluation of training programs, as a generic activity, is to give value to learning, and its outcome as a whole process. this objective is the factor that promotes the selection of one type of evaluation or another. the functions and objectives of an evaluation process could be different. these functions are connected to the understanding and meaning of the learning process to be analysed, to the agents that are going to learn the skills needed, and to the role this learning process acquires for the society or the institution. thus, for this research the main aim is to analyse and represent using systems thinking, the process of evaluation of training programs to define them in a more effective and profitable way. systems thinking is defined as a skill set to better understand the deep roots of complex behaviours to better predict them, and ultimately, adjust their outcomes. systems thinking can be viewed as a system, a system of thinking about systems. and systems thinking consists of three kinds of things: elements (characteristics), interconnections (the way these characteristics relate to/or feedback into each other), and a function or purpose. though not all systems have an obvious goal or objective, systems thinking does (arnold and wade, 2015). after the analysis of this paradigm, it could be stated that the specific objectives of this research are: 1. identification of the connection between the variables that represent the phases of the learning process, and the evaluation of the process, using a conceptual model. 2. representation of the connection between the learning process and the phases needed for evaluation of the process using systems thinking. 3. methodology the methodology followed for the definition of the model was focused on three main steps: i) literature review and identification of the challenge to be addressed, ii), identification of significant variables, and iii) definition of the conceptual model as a final result. figure 1. methodology followed in this study. 3.1. literature review and identification of the challenge to be addressed organizational training is one of the most important human resource strategies that organizations are dealing with. in such a current changing and competitive environment, training becomes a factor of excellence and a key to obtain successful organizations (herrero, 2000). in recent years, organizations are investing importantly in training programs. however, they are not investing in a function directly connected to these programs to guarantee their quality: training evaluation. int. j. prod. manag. eng. (2020) 8(2), 99-109 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 100 http://creativecommons.org/licenses/by-nc-nd/4.0/ training evaluation in the organizations could be defined as “the analysis of the total value of a system, a training program, or a training course from both financial and social perspectives” (kenneydonnelly, 1976; aragón et al., 2003). the evaluation of the impact and profitability of training is one of the modalities of evaluation of the training that occurs in organizations. the impact of training is understood as the repercussions that carrying out training actions entails for the organization, in terms of response to the training needs, of solving problems, and contributing to the achievement of strategic objectives that the organization has defined. thus, the impact of training is based on the changes resulted from the learning and the transfer to the workplace generated in both, the department or area of the trained person, and the whole organization (herrero, 2000; aragón et al., 2003). as a systematic process for developing needed workplace knowledge and expertise, instructional systems design requires an evaluation component to determine if the training program achieved its intended goal—if it did what it purported to do. however, evaluation, the last phase of the addie (analysis, design, develop, implement, evaluate) model, is often overlooked when organizations create and implement training programs. strictly speaking, the larger view of evaluation may not be treated as a separate phase during the process. it is indeed an ongoing effort throughout all phases of the addie process (hannum & hansen, 1989; wang & wilcox, 2006) and culminating at the last phase. a number of reasons have been noted for organizations failing to conduct systematic evaluations. first, many training professionals either do not believe in evaluation or do not possess the mind-set necessary to conduct evaluation (swanson, 2005). others do not wish to evaluate their training programs because of the lack of confidence in whether their programs add value to, or have impact on organizations (spitzer, 1999). lack of evaluation in training was also attributed to the lack of resources and expertise, as well as lack of an organization culture that supports such efforts (desimone et al., 2002; moller, benscoter, & rohrer-murphy, 2000). even for limited efforts in training evaluation, most are retrospective in nature (brown & gerhardt, 2002; wang & wang, 2005). a study of a group of instructional design practitioners indicated that 89.5% of those conduct end-ofcourse evaluation, 71% evaluate learning; however, only 44% use acceptable techniques for measuring achievement. yet merely 20% of those surveyed correctly identified methods for results evaluation (moller & mallin, 1996; wang and wilcox, 2006). brown & gerhardt (2002) concluded that companies expend even less effort in evaluating the instructional design process (nadler & nadler, 2012; wang & wilcox, 2006). the success of training depends on the correct execution of all steps of the process: previous analysis of training needs, development and implementation of an adequate training plan and evaluation (pineda, 1995: 33; gómez-mejía et al., 1996: 253; solé & mirabet, 1997: 35, 63; aragón et al., 2003). however, despite the significance of both the training needs analysis, which influences the development, application and evaluation of training (mcgehee & thayer, 1961; agnaia, 1996; gray & hall, 1997; al-khayyat, 1998; legare, 1999; dickenson & blundell, 2000; holton, 2000; selmer, 2000) and the plan development and implementation stage where the training characteristics are established and put into practice (buckley & caple, 1991; goldstein, 1993; foot & hook, 1996; bee & bee, 1997; frazis et al., 1998, 2000), they are not studied enough to ensure the success. the model proposed in this paper goes into the evaluation of training in greater depth because this stage is the least studied of the whole process, following the objective pursued by aragón et al. (2003). 3.1.1. training evaluation models the existing literature proposes different models for carrying out training evaluation (kirkpatrick, 1997; kirkpatrick & kirkpatrick, 2016; phillips & philips, 2016; hamblin, 1974; tannenbaum & woods, 1992; kaufman & keller, 1994; holton, 1996; pineda, 1998). the one that kirkpatrick developed at the end of the 1950s, called the ‘model of four levels’ or ‘kirkpatrick’s model, can be highlighted. this is the most widely used by organizations and the most widely referenced in studies about this subject due to its simple and practical ideas (plant & ryan, 1992; oberman, 1996; alliger et al., 1997; phillips & philips, 2016; o’neill, 1998; aragón et al., 2003). training evaluation is the measurement of a training program’s success or failure with regard to content and design, changes in learners, and organizational payoffs. the evaluation techniques used to assess these depend on the evaluation model chosen, as four different models have been proposed. the first int. j. prod. manag. eng. (2020) 8(2), 99-109creative commons attribution-noncommercial-noderivatives 4.0 international understanding and representation of organizational training programs and their evaluation 101 http://creativecommons.org/licenses/by-nc-nd/4.0/ model, kirkpatrick’s fourdimensional measurement typology (i.e., reactions, learning, behavior, results), is perhaps the simplest method for understanding training evaluation and the most frequently cited technique. in this model, learning is measured during training and refers to attitudinal, cognitive, and behavioral learning. behavior refers to on-the-job performance and, thus, is measured after training. additionally, reactions to training are related to learning, learning is related to behavior, and behavior is related to results (alvarez et al., 2004). in the second model, tannenbaum et al. (1993) expanded on kirkpatrick’s typology by adding posttraining attitudes and dividing behavior into two outcomes for evaluation: training performance and transfer performance. in their model, reactions to training and posttraining attitudes are not related to any other target of evaluation. however, learning is related to training performance, training performance is related to transfer performance, and transfer performance is related to results (philips, 2012; álvarez et al., 2004). in the third evaluation strategy, holton (1996) included three evaluation targets: learning, transfer, and results. reactions are not a part of holton’s model because reactions are not considered a primary outcome of training; rather, reactions are defined as a mediating and/or moderating variable between trainees’ motivation to learn and actual learning. in this model, learning is related to transfer and transfer is related to results. in addition, holton argued for an integration of evaluation and effectiveness. as a result, in his model particular effectiveness variables are outlined as important for measurement when evaluating training outcomes (álvarez et al., 2004). the fourth and final evaluation strategy was provided by kraiger (2002). this model emphasizes three multidimensional target areas for evaluation: training content and design (i.e., design, delivery, and validity of training), changes in learners (i.e., affective, cognitive, and behavioral) and organizational payoffs (i.e., transfer climate, job performance, and results). reactions are considered a measurement technique for determining how effective training content and design were for the tasks to be learned. kraiger asserted that reaction measures are not related to changes in learners or organizational payoffs but that changes in learners are related to organizational payoffs (álvarez et al., 2004). 3.2. definition of the conceptual model three group model building (gmb) sessions were developed for the definition of the conceptual model. gmb is a form of causal modelling based on systems thinking. its main strength is its insistence on feedback loops. the different structures within an organization are defined through variables and causal relationships (hoppenbrouwers and rouwette, 2013). based on this logic, the resulting model of this study has four main groups: i) assessment typology, ii) assessment impact, iii) striker, and iv) initial diagnosis assessment. 3.2.1. variable selection the inputs used for the selection of the variables that compose the conceptual model were two: i) variables identified in the literature review, and ii) variables identified through the gmb sessions. in the following table 1 are shown the narratives and cycles identified in the gmb sessions. the narratives highlighted in red refer to the ones that were repeated in the three sessions, so they were considered relevant. moreover, the colour of the four circles surrounding them corresponds to the cycles of our final conceptual model. the variable surrounded by a blue circle define the narrative behind the “initial diagnosis assessment” feedback loop of the final conceptual model. the variable surrounded by a green circle define the narrative behind the “striker” feedback loop of the final conceptual model and taken from the gmb session. the variable surrounded by a black circle defines the narrative behind the “assessment typology” feedback loop of the final conceptual model and taken from the gmb session. finally, the variables surrounded by a red circle defines the narrative behind the “assessment impact”. this coloured circles will be used for the trazability of the identification of the final feedback loops in both table 1, and image 1. moreover, explanation of each of the variables mentioned in the gms sessions and included in the table and the text is presented in appendix i. consequently, it could be said that the most relevant narratives and feedback loops extracted from the gmb sessions were: i) identification of real learning needs according to industrial needs, ii) stakeholder satisfaction, iii) learning engagement, iv) design of the course and methodology, and v) willingness of the learner. int. j. prod. manag. eng. (2020) 8(2), 99-109 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 102 http://creativecommons.org/licenses/by-nc-nd/4.0/ the participants in the session defined identification of real learning needs as the key variable for “matching industrial needs”, which undoubtedly is directly related to the impact that training could have in the industrial needs fullfilment. this identification is considered the first step for the definition of the real necessity of the client. work situation analysis refers to the study needed prior to the planning of the course to analyze who will be the people that need to take the course, and which are the topics to be hold. identification of learners facilitates and conditions the definition of “number of jobs required” and “language ability”. the number of jobs required influences the job topology, and at the same time, this typology influences the number of jobs. on the other hand, the competence oriented methodology of learning enables the definition of specific jobs, and tailored training. this tailored training helps to increase the trainee interest and satisfaction level. it is understood that tailored training refers to a personalized course depending on the needs of the learners and their company. in addition, stakeholder satisfaction, the second cycle, was contemplated as a central part in the sessions. this satisfaction will turn into the striker that will engage the learner to take another training program. the narrative behind this loop is related to cost, when cost of both training and final product are increased, stakeholder satisfaction is decreased. on the other hand, quality of the final product, competitive advantage, and employability positively influence stakeholder satisfaction. the participants in the sessions defined the confidence in the training system as the consequence of stakeholder satisfaction. due to this confidence, recognition on the training system is increased, which increases competitive advantage, helps in the identification of new fields of skills, and enables the meet of governmental rules to get the fund. finally, worker ability to perform his work is considered a key variable for the quality of the final product, and employability. on the other hand, learning engagement is considered as an initial situation variable. in the loops based on learning engagement, this engagement is dependent on the positive competition between the trainees, content up to date, new training methodologies, and relevance of the means. these four factors are defined as direct influencers of both learning engagement and the increase of trainee interest satisfaction. training environment, number of equipment, and number of professionals available influence the relevance of the means. when we mention relevance of the means we consider the whole group of resources needed for an efficient training course. training environment refers to behavior and attitude between the learners, if they recognize the importance of the training, and the training course is developed fluently. table 1. narratives identified in the gmb sessions and their linkage to the final conceptual model. int. j. prod. manag. eng. (2020) 8(2), 99-109creative commons attribution-noncommercial-noderivatives 4.0 international understanding and representation of organizational training programs and their evaluation 103 http://creativecommons.org/licenses/by-nc-nd/4.0/ design of the courses is related to the assessment typology, depending on the arquitecture of the course, the most appropriated assessment method will be one or another. referring to the influence of the design of the courses, the participants in the sessions defined “knowledge transfer” as the central variable to be analyzed in the session. the key variable directly influencing the challenge was the design and methodology used for the definition of the course. the type of design chosen for knowledge transfer is dependent on trainer abilities, contents selected for the course, age of the public, previous knowledge of the learners about the topic to be hold, duration of the course, and the use of interactive tools. on the other hand, the type of methodology selected will influence another factors such as the dynamicity of the course, and the easy accessibility of the learners to the contents and tools. willingness was directly connected to strker variables that will engage the learner to take the course. when speaking about the influence of the willingness of the learner for knowledge transfer, learner behavior was defined as the central pillar. depending on the behavior taken by the learner the knowledge transfer will be done or not. at the same time, this behavior will be dependent on the same variables that influenced the design and methodology used for the courses. the age of the public will be decisive due to the importance participants gave to the fact of previous knowledge of the learners. selection of the profile referred to the study needed to choose workers that really need the training course to apply the skills acquired in their workplaces. finally, pedagogical objectives are used to define the competencies needed to achieve through the courses. accessibility as mentioned before describes the way to obtain data, and how to access the information. after this analysis of the variables from the gmb, in the image 2 below could be seen the variables identified in the literature. the ones highlighted in orange refer to the narratives that appear in both gmb sessions and literature. in addition, the coloured circles correspond to the feedback loops that define our final conceptual model. red circle refers to “striker” loop, green circle to “initial diagnosis assessment”, black to “assessment typology”, blue refers to “assessment impact”. 4. results the identification of groups used for the definition of the conceptual model is based on the principal phases of the learning process presented by alsina and rodriguez (2001). 1. assessment typology/ training assessment: this loop explains how the training evaluation is done. interaction with students is a direct way to do it. the results of this interaction are i) an adaptation of the teaching process to define adequately the skills that are needed to learned, and ii) the diagnosis of the obstacles to adapt the learning strategies to obtain more effective learning processes. 2. assessment impact: this loop refers to the effect of the assessment on the principal phases of the learning process. these variables help to the obtaining of the objective (effective training). three different variables were identified as key for this outcome obtaining, the satis faction level of the learner with the program, the effectivity of the program to acquire new skills, and the transfer of the learning competencies to the workplace. 3. striker: this loop refers to the phase in which learners have achieved new skills, and their self steem is higher. so, their wish to improve is higher, and they are opened to learn and accept more complex challenges. 4. initial diagnosis assessment: this loop refers to that phase of “wish to improve” mentioned in the previous loop, and its effects. this wish results in a higher autonomy level of the learner, and more identification with the learning process. at the same time, this loop is focused on the strategies defined for the learning proces 5. conclusions the increasing interest that organizations have been showing in the last decades in employees and practices related to human resource management, figure 2. variables identified in the literature. int. j. prod. manag. eng. (2020) 8(2), 99-109 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 104 http://creativecommons.org/licenses/by-nc-nd/4.0/ especially training, can be explained by the general acceptance of the fact that human resources and organizational knowledge are, at present, two of the main sources of sustainable competitive advantages for the organization. however, the significant role of training in the organization is not supported by an adequate level of investment, mainly due to the lack of knowledge about the contribution of training programs to the achievement of organizational outcomes. as a systematic process for developing needed workplace knowledge and expertise, instructional systems design requires an evaluation component to determine if the training program achieved its intended goal. however, evaluation, the phase considered as the most important one for the definition of effective training programs, is often overlooked when organizations create and implement training programs. several reasons have been identified for organizations failing to arrange systematic evaluations (lack of confidence, lack of resources and expertise, lack of organizational culture, etc.). this study was based on the identification of variables that influence training effectiveness and evaluation. for this identification the input sources used were literature on the one hand, and group model building (gmb) sessions on the other hand. this identification of variables was needed for the definition of a conceptual model in which variables were interconnected defining feedback loops, and focused on the logic of systems thinking. this final model represents the process of training evaluation with four main feedback loops: i) assessment typology, ii) assessment impact, iii) striker, and iv) initial diagnosis assessment. each of them encompasses all the key variables and narratives considered relevant in both literature and gmbs. as a future line, it was contemplated the transformation of the conceptual model into computational through the use of system dynamics. this computational model will be helpful for the interactive representation of the conceptual model, the effective communication of the conclusions of the model to the management board of every organization interested in the implementation of effective training programs. moreover, this model will facilitate the simulation of different scenarios to extract patterns related to the behaviour of the variables that influence the effectiveness of training programs and their evaluation. references agnaia, a.a. (1996). assessment of management training needs and selection for training: the case of libyan companies. international journal of manpower, 17(3), 31-52. https://doi.org/10.1108/01437729610119504 figure 3. whole conceptual model. int. j. prod. manag. eng. (2020) 8(2), 99-109creative commons attribution-noncommercial-noderivatives 4.0 international understanding and representation of organizational training programs and their evaluation 105 https://doi.org/10.1108/01437729610119504 http://creativecommons.org/licenses/by-nc-nd/4.0/ al-khayyat, r. (1998). training and development needs assessment: a practical model for partner institutes. journal of european industrial training, 22(1), 18-28. https://doi.org/10.1108/03090599810197658 alliger, g.m., tannenbaum, s.i., bennett, w., traver, h., shotland, a. (1997). a metaanalysis of the relations among training criteria. personnel psychology, 50, 341-58. https://doi.org/10.1111/j.1744-6570.1997.tb00911.x alsina, m.i.b., rodriguez, f.a.c. (2001). estrategias de evaluación de los aprendizajes centradas en el proceso. revista española de pedagogía ,25-48. alvarez, k., salas, e., garofano, c.m. (2004). an integrated model of training evaluation and effectiveness. human resource development review, 3(4), 385-416. https://doi.org/10.1177/1534484304270820 aragón-sánchez, a., barba-aragón, i., sanz-valle, r. (2003). effects of training on business results. the international journal of human resource management, 14(6), 956-980. https://doi.org/10.1080/0958519032000106164 arnold, r.d., wade, j.p. (2015). a definition of systems thinking: a systems approach. procedia computer science, 44(2015), 669-678. https://doi.org/10.1016/j.procs.2015.03.050 bassi, l.j., ludwig, j., mcmurrer, d.p., van buren, m. (2002). profiting from learning: firm-level effects of training investments and market implications. singapore management review, 24(3), 61-80. bee, f., bee, r. (1997). training needs analysis and evaluation. london: institute of personnel and development. brown, k.g., gerhardt, m.w. (2002). formative evaluation: an integrative practice model and case study. personnel psychology, 55, 951983. https://doi.org/10.1111/j.1744-6570.2002.tb00137.x buckley, r., caple, j. (1991). la formación: teoría and práctica. madrid: díaz de santos. del valle, i.d., castillo, m.á.s. (2005). problemas y propuestas de medición de la formación en la empresa/problems and measuring proposals for company training. cuadernos de estudios empresariales, 15, 27. desimone, r.l., werner, j.m., harris, d.m. (2002). human resource development. cincinnati, oh: south western. dickenson, p., blundell, b. (2000). transferring quality management experience to the russian aerospace industry. total quality management, 11(3), 319-27. https://doi.org/10.1080/0954412006838 foot, m., hook, c. (1996). introducing human resource management. singapore: longman. frazis, h., gittleman, m., horrigan, m., joyce, m. (1998). results from the 1995 survey of employer-provided training. monthly labor review, 121(6), 3-13. gasalla, j.m. (2003). marketing de la formación de directivos: el nuevo directivo en la cultura del aprendizaje. madrid: pirámide. goldstein, i.l. (1993). training in organizations. pacific grove, ca: brooks/cole. gray, g.r., hall, m.e. (1997).training practices in state government agencies. public personnel management, 26(2), 187-203. https://doi.org/10.1177/009102609702600203 hamblin, a.c. (1974). evaluation and control of training. maidenhead: mcgraw-hill. hannum,w., hansen, c. (1989). instructional systems development in large organizations. englewood cliffs, nj: educational technology publications. herrero, p.p. (2000). evaluación del impacto de la formación de las organizaciones. educar, (27), 119-133. holton, e.f. (2000). large-scale performance-driven training needs assessment. public personnel management, 29(2), 249-67. https://doi.org/10.1177/009102600002900207 hoppenbrouwers, s., rouwette, e. (2012). a dialogue game for analysing group model building: framing collaborative modelling and its facilitation. international journal of organisational design and engineering, 2(1), 19-40. https://doi.org/10.1504/ijode.2012.045905 kaufman, r., keller, j.m. (1994). levels of evaluation: beyond kirkpatrick. human resource development quarterly, 5(4), 371-80. https://doi.org/10.1002/hrdq.3920050408 kenney, j., donnelly, p. (1976). manpower training and development. londres: harrap. kirkpatrick, d. (1997). revisando las grandes ideas. training & development digest, september: 28-36. kirkpatrick, j.d., kirkpatrick, w.k. (2016). kirkpatrick’s four levels of training evaluation. association for talent development. kraiger, k., ford, j.k., salas, e. (1993). application of cognitive, skill-based, and affective theories of learning outcomes to new methods of training evaluation. journal of applied psychology, 78, 311-328. https://doi.org/10.1037/0021-9010.78.2.311 legare, t.l. (1999). defining training roles and responsibilities at partners healthcare system. national productivity review, 19(1), 5-13. https://doi.org/10.1002/npr.4040190103 mcgehee, r., thayere, j. (1961) training in business and industry. new york: john wiley. moller, l., mallin, p. (1996). evaluation practices of instructional designers and organizational support or barriers. performance improvement quarterly, 9(4), 82-92. https://doi.org/10.1111/j.1937-8327.1996.tb00740.x nadler, z., nadler, l. (2012). designing training programs. routledge. oberman, g. (1996). an approach for measuring safety training effectiveness. occupational health & safety, 65(12), 48-58. int. j. prod. manag. eng. (2020) 8(2), 99-109 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 106 https://doi.org/10.1108/03090599810197658 https://doi.org/10.1111/j.1744-6570.1997.tb00911.x https://doi.org/10.1177/1534484304270820 https://doi.org/10.1080/0958519032000106164 https://doi.org/10.1016/j.procs.2015.03.050 https://doi.org/10.1111/j.1744-6570.2002.tb00137.x https://doi.org/10.1080/0954412006838 https://doi.org/10.1177/009102609702600203 https://doi.org/10.1177/009102600002900207 https://doi.org/10.1504/ijode.2012.045905 https://doi.org/10.1002/hrdq.3920050408 https://doi.org/10.1037/0021-9010.78.2.311 https://doi.org/10.1002/npr.4040190103 https://doi.org/10.1111/j.1937-8327.1996.tb00740.x http://creativecommons.org/licenses/by-nc-nd/4.0/ o’neill, m. (1998). cómo enfocar la evaluación de la formación. info-line (american society for training and development), 1(special issue), 1-18. phillips, j.j. (2012). return on investment in training and performance improvement programs. routledge. https://doi.org/10.4324/9780080516257 phillips, j.j., phillips, p.p. (2016). handbook of training evaluation and measurement methods. routledge. https://doi.org/10.4324/9781315757230 pineda, p. (1998). el reto de evaluar la formación en la empresa: herramientas and soluciones. capital humano, 111, 32-6. plant, r.a., ryan, r.j. (1992).training evaluation: a procedure for validating an organizations investment in training. journal of european industrial training, 16(10), 22-38. https://doi.org/10.1108/03090599210021720 sarramona, j., colom, a, vázquez, g (1994). estrategias de formación en la empresa. narcea. selmer, j. (2000). a quantitative needs assessment technique for cross-cultural work adjustment training. human resource development quarterly, 11(3), 269-282. https://doi.org/10.1002/1532-1096(200023)11:3%3c269::aid-hrdq5%3e3.0.co;2-6 spitzer, d.r. (1999). embracing evaluation. training, 36(6), 42-47. swanson, r.a. (2005). evaluation, a state of mind. advances in developing human resources, 7(1), 16-21. https://doi.org/10.1177/1523422304272078 tannenbaum, s.i., woods, s.b. (1992). determining a strategy for evaluating training: operating within organizational constraints. human resource planning, 15(2), 63-82. wang, g.g., wilcox, d. (2006). training evaluation: knowing more than is practiced. advances in developing human resources, 8(4), 528539. https://doi.org/10.1177/1523422306293007 appendix i this appendix is included with the aim of explaining the variables identified during the gmb sessions. after the individual variable definition of the participants, we made groups according to the similarities of the variables, the groups were: i) training typology (variables related to the characteristics of the training), ii) prior analysis for definition of the course (variables related to the planning of the course), iii) commitment of the learner (variables related to the behavior of the learner and their predisposition for learning), iv) new learner generation characteristics (this refers to the characteristics of the new generation of learners), v) trainer tools (this refers to the available techniques and tools for the trainer), vi) resources (available human and physical resources), vii) government and investment (this describes governmental and investment issues), viii) prior knowledge of the learner (variables related to the knowledge learners have when the course starts), ix) job detail (variables related to the need of specific jobs, and the number of jobs required), x) design/methodology (variables related to the methodology used and the design of the courses), xi) learner behavior (variables related to the willingness of the learner to learn and attend the course), xii) needs identification (variables related to the study they have to do before defining the contents of the course), xiii) contents (variables related to the definition of the contents of the course), xiv) evaluation (variables related to the phase of evaluation and feedback after the course), and xv) expert (variables related to the specific characteristics of the experts). in the following tables, each group and the corresponding variables to each group are explained. training typology tailored training: it refers to personalized training, adapted training courses to the needs of the learner and their companies. competence oriented training: it refers to the identification of the training modules, knowledge and needed skills. training flexibility: it refers to the opportunity the training course gives to adapt to it. training duration, schedule: it refers to the duration of the course, and its schedule. new training methodology: it refers to the importance of choosing the training system that best fits our objectives. prior analysis for definition of the course work situation analysis: it refers to the relevance of an analysis prior to the definition of the course. definition of needed skills: it refers to the definition of the skills needed according to the industrial needs. identification of industrials needs: it refers to the real necessity of the companies in order to make them more competitive. working learning time management: it refers to the management of working and learning equilibrium. int. j. prod. manag. eng. (2020) 8(2), 99-109creative commons attribution-noncommercial-noderivatives 4.0 international understanding and representation of organizational training programs and their evaluation 107 https://doi.org/10.4324/9780080516257 https://doi.org/10.4324/9781315757230 https://doi.org/10.1108/03090599210021720 https://doi.org/10.1002/1532-1096(200023)11:3%3c269::aid-hrdq5%3e3.0.co;2-6 https://doi.org/10.1177/1523422304272078 https://doi.org/10.1177/1523422306293007 http://creativecommons.org/licenses/by-nc-nd/4.0/ commitment of the learner to increase the trainee interest (pleasure) satisfaction: it refers to the consequence of positive levers that increase trainee interest. learning engagement: it refers to the engagement shown by the learners, their commitment to the course. new learner generation characteristics new generation: it refers to the new generation of learners, the new trend appeared in the last years. evolution of brain research: it is related to biotechnology, analysis of the brain to help people and get more information. new habits of students: it refers to the new habits of this new generation of learners. evolution of technology: it refers to the evolution, and the changes suffered in the field of technology. worker ability(skills) to perform his work: it refers to the the abilities of the workers, and how they impact the final performance and productivity. new fields of skills: it refers to the new fields of skills appeared together when the evolution of technology and new learners. knowledge retention: it refers to the fact of not losing what they learn, how to retain that knowledge in order to facilitate its application. trainer tools manage to generate positive competition between trainees: it refers to climate, how to generate competitiveness between the learners to facilitate and create positive tension. feedback: it refers to the necessary feedback when the courses are ended, the evaluation of the efficiency of the course. to have simple tools to assess the match: it refers to the tools used by the trainer, the participants defined simple tools as the most efficient ones. content up to date: it refers to the fact updating contents, these courses must be a live and dynamic. oblivion curve: it refers to the point in which learners start to forget what they learned, hoe to make longer this horizon of knowledge retaining. resources number of equipment: it refers to the available physical resources. training location: it refers to place where the training is going to be hold. number of trainers available: it refers to the available human resources to give the training courses. government and investment meet governmental rules to get the fund: it refers to the french government as the principal source of funding of any training program, and how to meet their rules. cost: it refers to the cost of both the course, and the internal costs of the companies that take part in the training courses. prior knowledge of the learner original acquired skills: it refers to the basis of the students, which is the starting point from the point of view of prior knowledge. language ability: it refers to the language selected fro the training, the learners must be able to follow those courses in the corresponding language. job detail specific jobs: it refers to the topics that must be hold in the courses. number of jobs required: it refers to the number of different workplaces that will appear taking into account the necessity of the company. design/ methodology age: makes reference to the age of the learner, depending on that some factors related to their behaviour will change. duration: sets out the duration of the course, the participants considered that these courses must not be long to be effective and attractive. interactive tools: refers to the tools used in the courses, the necessity of being interactive was mentioned. dynamicity: explains the necessity of dynamic courses and contents to be updated and adapted to the real necessities of the market. access: makes reference to the accessibility of the learner to the course and the contents. balance theory and practical part: describes the necessity of a balanced distribution of theoretical and practical part. int. j. prod. manag. eng. (2020) 8(2), 99-109 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 108 https://www.linguee.es/ingles-espanol/traduccion/source+of+funding.html http://creativecommons.org/licenses/by-nc-nd/4.0/ learner behaviour knowledge transfer: sets out the central part of this gmb session, the transfer of basic knowledge from specialists to non specialists. knowledge retention: explains the concept of how to retain knowledge and to avoid its lost. needs identification pedagogical objectives: makes reference to the objectives defined as pedagogical, the expected results to be achieved after the course. company policies: refers to the strategy and type of decisions made by company, their philosophy. contents market evolution: explains the changes happened in the market and the society, changes that must be faced by the company to adapt to the client and be competitive. not deep technical explanation: makes reference to the type of explanation that must be given in the courses. according to the participants these explanations do not need to be very technical. selection of the profile: refers to the criteria that must be taken into account to select the profile of the future learners. turnover: sets out the amount of money taken by a business in a particular period. budget: refers to an estimate of income and expenditure for a set period of time updated: explains the necessity of continuously updating contents and methodologies to be more effective and competitive. evaluation long term know how: explains the concept of acquiring usable and long term knowledge. expert knowledge: refers to knowledge the expert has, which is based on their previous knowledge basis. availability: refers to the availability of the expert, if he or she is frequently available to respond to the doubts and questions of the learner. int. j. prod. manag. eng. (2020) 8(2), 99-109creative commons attribution-noncommercial-noderivatives 4.0 international understanding and representation of organizational training programs and their evaluation 109 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.13765 received: 2020-05-06 accepted: 2020-09-04 a systematic literature review of total quality management (tqm) implementation in the organization permana, a.a*, purba, h.h.b, rizkiyah, n.d.c a industrial engineering department, mercu buana university, jl. meruya selatan no. 1, jakarta 11610, indonesia. a1 adi.permana_ap@yahoo.com, a2 humiras.hardi@mercubuana.ac.id, a3 nyimas.desy@mercubuana.ac.id abstract: in today’s market situation and complex business environment, organization must be able to deliver the customer’s requirement and the expectations which are critical to the satisfaction such as high product quality, faster delivery and competitive cost. organization need to apply a comprehensive concept and method on managing those requirements. the concept of total quality management (tqm) is considered as one of a popular concept used to manage the quality of product and services comprehensively. this research is to observe is this concept and method still relevant to be use and effectively improved the business performance as well as customer satisfaction. it is a systematic literature review to the literatures from many industry sectors that were collected and reviewed in detail. the result show that this concept is still being used by many organizations around the world and its successfully help the organization to improve their competitiveness, business growth and the sustainability as well as increase employee’s morale. key words: tqm, business improvement, business competitiveness. 1. introduction in the dynamic market situation and a competitiveness as well as a highly changeable and complex business environment (eltawy & gallear, 2017). organizations are continuing facing the changes, globalization, fast technological advances, competition, disruptive business models, emerging new markets, where its constantly changes – are the challenges for the organizations both for big or small size organization (žitkienė & deksnys, 2018). organization must be ready to adapt to this situation that required a very demanding in product or service quality, faster delivery and a competitive price. organization must re-think their focus on a new model is not on how much they are making but on how well they are meeting customer’s requirements (george & weimerskirch, 1998). organizations around the world have used quality strategically to win customers (oakland, 2003). customer needs is continuing to evolve in line with the diversification of lifestyles, and high quality and functionality are expected of every product (kaname, 2003). the quality perception is different between one and another, including the definition of the quality itself as described in the table 1 (zairi, 1991). no matter what is the definition, quality must be delivered to the customer and it must be maintained to satisfy the customer and sustain the business. organization need to apply a comprehensive concept on how they can maintain the level of satisfaction of their customer by delivering best quality of their products and services. the organization need to cite this article: permana, a., purba, h.h., rizkiyah, n.d. (2021). a systematic literature review of total quality management (tqm) implementation in the organization. international journal of production management and engineering, 9(1), 25-36. https://doi.org/10.4995/ijpme.2021.13765 int. j. prod. manag. eng. (2021) 9(1), 25-36creative commons attribution-noncommercial-noderivatives 4.0 international 25 http://creativecommons.org/licenses/by-nc-nd/4.0/ to establish a strategy that will focus to improve business activities to overcome the competition as well as improved the competitive advantage (kantardjieva, 2015). organization need to consider on total quality management concept (tqm) to overcome those challenges. tqm is widely used by many organization around the world and its successfully applied and beneficial the organization significantly. it is clearly recognized that the very important aspect of rebuilding japanese quality after world war ii is the use of the concept quality management (dahlgaard et al., 2002). the evolution timeline of total quality management has breakdown into four stages (dahlgaard et al., 2002). it categorized as follows: (1) quality inspection, (2) quality control, (3) quality assurance, and (4) total quality management. it start in 1910 in ford motor company where it started employed inspector to check the product quality. about 20 years later, in 1920-1930s, a statistical process control is being used as well as sampling product to be inspect thanks to the method developed by shewhart & dodge-roming. in this era, control chart is used to distinguish two types of variation in the process, random variation and assigned variation. the third stage of quality assurance is started about 1950s later to guarantee the product by providing such a confidence to assure that the product or service will meeting customers’ expectation. the final stage of tqm was started in 1980s, tqm are clearly involves an understanding and implementation concept of quality management, in every aspect and area of organization. since then, deming, juran and crosby have had big contribution towards the continuous development of the tqm until the present. tqm approach is focusing to improve effectiveness of the processes and the responsiveness in meeting the customer requirement as part of organization excellence goals in achieving customer satisfaction (ramlawati & putra, 2018). to ensure the effectiveness of the concept implementation in the organization, all of the components must be work together as a team. each part, each activity, each person in the organization affects and is in turn affected by others (oakland, 2003). tqm as a philosophy for modern competitiveness and discussed the various contributions in the area of quality management which have propelled its prominence to today’s levels of competitiveness (zairi, 1991). tqm implementation can be divide into 3 stages (fitriani, 2019): (1) preparation stage, (2) planning stage and (3) execution stage. those 3 stages must be implemented well with the commitment from management and employees involvement and others factor such as training and communication (kumar & shanmuganathan, 2019). table 1. definition of quality. transcendent definition this means quality is hard to define, sometimes it’s blurred. it’s like i can’t define it, but i know when i see it product based definition this means quality is about measured how good is the product. its about how does the product works. user based definition this means quality is about how product fulfill the customer needs and their expectations manufacturing based definition this means quality is how the product meeting the specification. value based definition this means quality is about comparison level between costs and price. what cost/price will be acceptable in this case. table 2. the development of tqm. quality inspection (1910s) quality is started by doing a basic sorting of product quality and identify the problem source quality control (1920s) in this period, the process is controlled by the use of data analysis and statistic control approach, and improved in quality planning quality assurance (1950s) this is a period where quality planning and audits is performed as well as the use of quality risk assessment through fmea tqm (1980s) managing the quality comprehensively started with management commitment, employee involmenet, focus to customer and continuous improvement culture to improve quality int. j. prod. manag. eng. (2021) 9(1), 25-36 creative commons attribution-noncommercial-noderivatives 4.0 international permana et al. 26 http://creativecommons.org/licenses/by-nc-nd/4.0/ tqm must be viewed as a longterm process, the vision to continually looking into the future, not only have to meet short term target, organization must plan for their existence in the future (rogers, 2013). 2. methodology this study is considering of 4 stages as described on the figure 1. it consist of collecting of the predefined papers, journal and some textbooks, did some quick review and made the shortlisted papers to continue in dept review and classified it into 4 grouping. preliminary research with defined related keyword: 108 articles quick review on the papers to identify appropriate papers: 50 articles comprehensive review of the shortlisted articles summarize the result, evaluated and classified into 4 group application of tqm the impact of tqm the review of tqm the relationship of tqm figure 1. study framework. 3. result and discussion in depth review to about 50 papers was completely done. an evaluation is considering the research object as well as the result of each paper. the complete list of the paper had been reviewed are shown in the table 3. there are 4 groups of papers were classified; application of tqm, the impact of tqm, review of tqm and relationship of tqm. the application of tqm in the organization was appear in some papers with multiple organization sector are using tqm succesfuly. kumar & shanmuganathan, (2019) developed a standard reliable instrument that can be used to measure the tqm implementation in an automotive component manufacturing company. the result shown that tqm has significantly influences the organizational commitment as well as the overall performance. in thailand, sivalai & rojniruttikul (2018) did a measurement the greatest factor affecting to the tqm in rail way company, the results determined that organizational culture give the biggest impact. bunglowala & asthana (2016) found that tqm is can be implemented well in education sector, they concluded that the teaching and learning procedure are more effective and its improved the overall quality. arifin (2016) did a research in financial sector (banking) where he concluded that tqm is significantly improved the overall performance of the company. talib & rahman (2015) did an observation in service industry to study the tqm application, beside the benefit of tqm, they also observed that lack of communication and lack of management commitment were the top barrier in tqm implementation. the impact of tqm are showed by many researchers such as santos et al., (2019) did a research in brazillian electricity distribution company where he observed that implementation of tqm is helped the company in improving the order scheduling by 12% and reduce unproductive visit by 22%. sari & firdaus (2018) made another research of tqm implementation in sme (small medium enterprise) sector in indonesia, where they concluded that tqm implementation can improve the competitive advantage of the sme. similarly, nugroho & nurcahyo (2018) also did a research in sme in indonesia where they concluded that tqm implementation can improve the financial performance of sme. sabet et al., (2014) did a study in uk manufacture company an they concluded that tqm is successfully improve the overall performance and it can be combined with six sigma to improve product quality. it also the same with the result of observation in nigeria where tqm is improve the customer satisfaction mercy & taiye, (2015). in hotel industry, tqm also significantly improve the organization competitive advantadge as it confirmed by research study by yeng et al., (2018) in malaysian’s hotel. in indonesia, dewi et al., (2015) did an observation in education sector where they conclude that tqm has a significant effect on student satisfaction. while sukardi (2016) worked with dept. store in int. j. prod. manag. eng. (2021) 9(1), 25-36creative commons attribution-noncommercial-noderivatives 4.0 international a systematic literature review of total quality management (tqm) implementation in the organization 27 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 3. existing list of articles reviewed. nr paper identity research object result 1 santos et al., 2019 electricity distribution company tqm implementation result: there was a reduction of 12% in the “orders after deadline” (oad) and 22% in the “unproductive visits” (uvi) indicators 2 busu, 2019 energy sector tqm model starts with staff motivation, given the significant result between the staff motivation and the organization’s performance 3 garcia-alcaraz et al., 2019 manufacturing sector concluded that hr plays an important role in the tqm implementation process, which guarantees to obtain the operational benefits that are measured through the quality indices in a product, a continuous process improvement, reduction of costs, and waste and better labor security for employees 4 antunes et al., 2018 education sector (school) tqm is allows to obtain a significant competitive edge including in the education sector 5 sousa-mendes et al., 2016 medical device (sme) the findings of this study are very important in providing insights for medical device smes as they can evaluate their current practices and re-allocate reasonable resources to improve their tqm performance 6 vukomanovic et al., 2014 construction industry the study observe that effective efqm (tqm) implementation is requires an approach that manages the enabler and result criteria separately 7 suarez-barraza & ablanedo-rosas, 2014 manufacturing industry 5s practice proved to be a significant driver for a tqm philosophy and the success of tqm implementation 8 hasan et al., 2018 education sector (school) tqm implementation is successfully solved the problem related to the primary education system in bangladesh 9 bigliardi & galati, 2014 r&d environment a successful implementation of the tqm in r&d area can be achieved only through the effective balance between two perspectives: the customer focus and the product quality 10 yeng et al., 2018 hotel tqm implementation is improve the knowledge of employees in quality management and the organization competitive advantage 11 nugroho & nurcahyo, 2018 sme sector tqm implementation is impacted to company’s financial result thanks to improvement in quality 12 yang & tsai, 2014 business sector tqm strategy and tools can be applied in the investment management industry. 13 sainis et al., 2017 sme sector the quantitative survey results showed that the quality culture and performance appraisal elements, were the most valued elements for the implementation of tqm 14 small et al., 2017 electrical sector the implementation of tqm improved productivity increases of 2 to 3%, which also increased as the project progressed with proactive management involvement 15 mitreva et al., 2016 food sector the study showed that the application of tqm in the processing food company is increased effectiveness and efficiency 16 gómez-lópez et al., 2016 manufacturing sector the results show that the most important motivations that drive firms to implement the tqm through the efqm are internal motivations 17 kiruthiga, 2016 construction sector it observed some factors that affecting to the tqm implementation as well as the concept acceptance, they are: lack of knowledge, employee’s motivation and the culture 18 kim, 2016 manufacturing sector tqm can be started with good leadership, focus on customer and better process management. an improvement in quality will increase customer satisfaction 19 alanazi, 2020 business sector the study observed the importance for managers and practitioners in implementation of successful tqm practices to achieve the main goal. 20 kuo, 2016 manufacturing sector the results shown that tqm implementation is contributed to a positive effect on customer capital 21 nicolaou & kentas, 2017 healthcare industry tqm failure reasons are a considerable issue and have to be studied in depth. 22 prajogo & brown, 2004 business sector the study findings show the strong links between tqm practices and quality performance of the firm 23 balasubramanian, 2016 healthcare industry the result shown some barriers in tqm implementation are the social culture, mentality & leadership style (table 1 continued on the next page) int. j. prod. manag. eng. (2021) 9(1), 25-36 creative commons attribution-noncommercial-noderivatives 4.0 international permana et al. 28 http://creativecommons.org/licenses/by-nc-nd/4.0/ nr paper identity research object result 24 sader et al., 2017 manufacturing sector the implementation of tqm is contributed by awareness and requirement of industry 4.0 where the technologies and big data approach are improved quality assurance and quality control 25 steiber & alänge, 2013 technology the tqm concept is required to be updated, the brand ‘tqm’ is still associated with quality and continuous improvement. a re-branding strategy might therefore be necessary since tqm now is broadly used in many areas 26 sadikoglu & olcay, 2014 manufacturing sector this study has shown that different tqm practices significantly affect different performance outcomes. primary obstacles were lack of employee involvement, awareness and commitment of the employees, inappropriate firm structure, and lack of the resources 27 sukdeo et al., 2017 manufacturing sector the tqm implementation is significantly contributed to organization performance, improved product quality, employee and customer satisfaction as well as supplier performance 28 morath & doluschitz, 2009 food industry the study observed positive correlation between the fulfilment of the requirements of tqm and the financial success of the companies could be verified statistically. the company also tend to use tqm in terms of improving the quality issues happened in the company 29 tervonen et al., 2009 manufacturing sector the study observed that rather than merely imitating tqm procedures, companies must focusing their efforts on creating a culture within which these procedures can thrive 30 kumar et al., 2009 business sector the findings reported that the expected tqm-oriented characteristics of pms and pms are perceived appropriate and the traditional ones are not felt appropriate 31 haffar et al., 2013 manufacturing sector the findings from this study demonstrate that the group and adhocracy culture types have a positive impact on the successful implementation of tqm 32 sari & firdaus, 2018 sme sector the research concludes that tqm implementation can improve the competitive advantage of sme 33 dewi et al., 2015 education sector (school) the result concluded that tqm is have a positive correlation and affecting significantly to the student satisfaction level 34 houston, 2007 education sector (school) the study resulted that tqm is not fully matched with the substance of higher education due to ist complexity. however, it more fruitful to explore the development of locally appropriate systemic approaches to improving quality in and of higher education. 35 phan et al., 2019 manufacturing sector the result concluded that there are positive correlation between tqm implementation and jit production concept on improving the performance 36 sweis et al., 2019 business sector the result shown where tqm implementation is improve customer satisfaction that leads to increasing in customer order/demand, it means that will increase sales and profit 37 mercy & taiye, 2015 service industry the result concluded that tqm implementation is directly impacted in increasing of customer satisfaction 38 v. kumar et al., 2016 pharmaceutical industry tqm is a holistic methodology towards the general change of an organisation. the theory of tqm has tended to be fruitful in all fields provided that the management has enough potential to actualize it 39 musenze & thomas, 2020 local government the study observed that a valid and reliable instrument has been developed and recommended for use in the assessment of tqm practices in the service sector with explicit focus on local government 40 neyestani & juanzon, 2016 construction sector the result of the study is defined the performance measurement (kpi) to evaluate the tqm implementation linked to the balance score card approach 41 ngambi & nkemkiafu, 2015 manufacturing sector the result concluded that good tqm implementation will increase the performance of the organization especially improve the financial performance 42 tesfaye & kitaw, 2017 manufacturing sector the result concluded that the tqm implementation and jit will enable the companies to produce a higher quality of product, lower cost and faster lead time 43 sutrisno, 2019 sme sector the result concluded that tqm are significantly improved operational performance of the company, increase customer satisfaction as well as substantially increased product quality. in the end, it will improve an organization competitive advantage and business sustainability (table 1 continued on the next page) (table 1 continued from the previous page) int. j. prod. manag. eng. (2021) 9(1), 25-36creative commons attribution-noncommercial-noderivatives 4.0 international a systematic literature review of total quality management (tqm) implementation in the organization 29 http://creativecommons.org/licenses/by-nc-nd/4.0/ indonesia to determine the impact of tqm where he found that tqm is significantly improved the customer satisfaction and he also realize that there are some key factors in implemented tqm which are focus to customer, quality obsession, team work and employee involvement. houston, (2007) did some research in new zealand higher education and concluded tqm is not fully matched with the substance of higher education due to its complexity. however, it more fruitful to explore the development of locally appropriate systemic approaches to improving quality in and of higher education. some researcher did a comprehensive review and evaluation on the tqm implementation in the organization where most of them were observed that tqm is successfully improved the organization performance and its competitive advantages. sweis et al. (2019) did an observation in some organization with different sector in jordan, they observed that the result shown where tqm implementation is improve customer satisfaction that leads to increasing in customer order/demand, it means that will increase sales and profit. kumar et al. (2016) did a same study, they focus on one of the pharmaceutical organization in india, they found that tqm is a holistic methodology towards the general change of an organization, the theory of tqm has tended to be fruitful in all fields provided that the management has enough potential to actualize it. similarly in indonesia, sutrisno (2019) did study in one of sme and concluded that tqm are significantly improved operational performance of the company, increase customer satisfaction as well as substantially increased product quality. these improvement will increase an organization competitive advantage and business sustainability. the last grouping was study the relationship of tqm with organization vision in some industries. kantardjieva (2015) did a research in some industries sector in greece where she found the relationship between the tqm and the strategic management, in this process, she found that the quality is a key success factor, so the business is must focused on the implementation of a quality programs such as tqm. in india construction firm, kiruthiga (2016) observe the relationship of lack knowledge, lack motivation and culture issues as the top barrier in implementing of tqm where it was given an input to the management nr paper identity research object result 44 lawrence & mccollough, 2004 education sector the result indicates that while many universities have been successfully applying tqm in support and administrative functions, tqm has not migrated into the classroom to any significant extent at many of these institutions 45 sukardi, 2016 dept. store tqm is significantly improved customer satisfaction with following contributor variables: focus to customer, focus on quality, team work and employees involvement 46 green, 2006 business sector what six-sigma process brings to tqm is a methodology for disciplined quality management. six-sigma strengthens tqm efforts through a strategic approach that emphasizes strong executive involvement, bottom-line accountability, extensive practical training, and personnel devoted to getting worthwhile improvement projects carried out 47 sila & walczak, 2017 manufacturing sector based on the ann results, we can also conclude that the contextual factors for obtaining iso certification and implementing tqm have a positive effect on f&m results 48 benzaquen et al., 2019 manufacturing sector the average values have shown that, in practice, the best results in the eight tqm success factors are valid in the organizations of the private health sector that have implemented a qms 49 mensah et al., 2012 business sector the study concluded that 3 factors that critical in implementing tpm successfully are: (1) top management commitment, (2) empowerment and involvement of employees, resource availability, (3) competition and increased customer awareness, and a well-functioning quality network 50 sabet et al., 2014 manufacturing sector the study observed that implementing tqm in company has generally been successful, however, it can be combined with other concept such as six sigma to improve the product quality (table 1 continued from the previous page) int. j. prod. manag. eng. (2021) 9(1), 25-36 creative commons attribution-noncommercial-noderivatives 4.0 international permana et al. 30 http://creativecommons.org/licenses/by-nc-nd/4.0/ to create the correct strategy in ensuring the tqm is implemented well in the organization. the source of literature mapping is as describe in the figure 2. according to the papers mapping on figure 2, there are 50 papers reviewed in total, 35 papers are published in the last 5 years (2015-2020) where it only 15 papers published more than 5 years ago, it classified into 4 group object variable: application of a. year of publication. 15 4 9 7 4 9 2 <2015 2015 2016 2017 2018 2019 2020 b. publication grouping by subject area. 24 3 5 18 application of tqm relationship of tqm reviewing tqm (literature review) the impact of tqm c. publication grouping by continent area of author. 6 9 19 2 14 africa america asia australia europe d. publication grouping by country of author. figure 2. literature’s mapping. (continued on next page). int. j. prod. manag. eng. (2021) 9(1), 25-36creative commons attribution-noncommercial-noderivatives 4.0 international a systematic literature review of total quality management (tqm) implementation in the organization 31 http://creativecommons.org/licenses/by-nc-nd/4.0/ tqm, the impact of tqm, the review of tqm and the relationship of tqm. the papers were spreading into 5 continents (asia, africa, america, australia & europe) where it came from 40 countries around the world. the papers were spreading into 18 different industries sector from manufacturing, construction, oil & gas, hotel, education until to sme level. most of the researcher from those 50 papers were observed the linkage between the implementation of tqm with the increament a competitive advantage of the organization as well as it increased customer satisfaction. a clear of tqm advantage can be seen in the figure 3. a tqm relationship framework: from relationship framework on figure 3, we can see that tqm is one of the good concept to be use in the organization as a strategic approach which is suitable for many industries sector and still popular to be use in today situation while it still compatible with new management standard such as iso 9001 and also with today’s market situation. here we can see that tqm is widely used in many industries sector as well as used in so many countries around the world. tqm also still popular e. publication grouping by research object (sector). figure 2. literature’s mapping. (continued from preious page) figure 3. tqm relationship framework. int. j. prod. manag. eng. (2021) 9(1), 25-36 creative commons attribution-noncommercial-noderivatives 4.0 international permana et al. 32 http://creativecommons.org/licenses/by-nc-nd/4.0/ in todays era where it show on the publication of the research paper in last 5 years where it confirmed that researcher are still interesting to observe the tqm implementation in the organization and also confirmed that tqm is still being used and still compatible with today’s business environment with rapid and quick changing (agile). for the further research on tqm framework in the future, it can be improve continuously and adapt to the preparation of industry 4.0 implementation as well as society 5.0. these will improve the product design capability and process improvement capability where it will beneficial the organization (figure 4). 4. conclusion total quality management (tqm) is still widely used in many industries sector as well as used in so many countries around the world. one of the most reason why tqm is still suitable on this today’s situation is the fact that tqm is focus to increase customer satisfaction on improving the quality of product, quality of service and overall quality of the organization to deliver the best product or service solution to the customer. another reason is because tqm is implemented with total commitment from the management as well as total involvement from employess where it become a solid concept that are simple to understand and easy to implement. furthermore, its required the standard measurement method to determine the level or score of the tqm implementation in the organization since it must be quantifiable to do continuous improvement as it need to compete in today’s business situation. in addition, its strongly recommended to continue the further study on the several industries sector especially for the new start up industries such as e-commerce or digital start up to ensuring that tqm is still suitable on those new industries sector. acknowledgements: this article was completed thanks to the financial support from the university of mercu buana, jakarta-indonesia. it also completed with the purpose and motivation of the authors to have an innovate research thinking as well as the contribution to the future researcher. references alanazi, m.h. (2020). the mediating role of primary tqm factors and strategy in the relationship between supportive tqm factors and organisational results: an empirical assessment using the mbnqa model. cogent business and management, 7(1). https://doi.org/10.1080/23311975.2020.1771074 antunes, m.g., mucharreira, p.r., justino, m. do r.t., & quirós, j.t. (2018). total quality management implementation in portuguese higher education institutions. proceedings mdpi, 2(21), 1342. https://doi.org/10.3390/proceedings2211342 arifin, j. (2016). penguatan manajemen syariah melalui total quality managementbagi pelaku lembaga keuangan syariah di kota semarang. jurnal at-taqaddum, volume 8, nomor 2, november 2016, 8(2), 180. https://doi.org/10.21580/at.v8i2.1170 figure 4. future research framework. int. j. prod. manag. eng. (2021) 9(1), 25-36creative commons attribution-noncommercial-noderivatives 4.0 international a systematic literature review of total quality management (tqm) implementation in the organization 33 https://doi.org/10.1080/23311975.2020.1771074 https://doi.org/10.3390/proceedings2211342 https://doi.org/10.21580/at.v8i2.1170 http://creativecommons.org/licenses/by-nc-nd/4.0/ balasubramanian, m. (2016). total quality management [tqm] in the healthcare industry – challenges, barriers and implementation developing a framework for tqm implementation in a healthcare setup. science journal of public health, 4(4), 271. https://doi.org/10.11648/j.sjph.20160404.11 benzaquen, j., carlos, m., norero, g., armas, h., & pacheco, h. (2019). quality in private health companies in peru: the relation of qms & iso 9000 principles on tqm factor. international journal of healthcare management, 0(0), 1-9. https://doi.org/10.1080/20479700.2019.1644472 bigliardi, b., & galati, f. (2014). the implementation of tqm in r&d environments. journal of technology management and innovation, 9(2), 157–171. https://doi.org/10.4067/s0718-27242014000200012 bunglowala, a., & asthana, n. (2016). a total quality management approach in teaching and learning process. international journal of management (ijm), 7(5), 223-227. http://www.iaeme.com/masteradmin/uploadfolder/ijm_07_05_021/ijm_07_05_021.pdf busu, m. (2019). applications of tqm processes to increase the management performance of enterprises in the romanian renewable energy sector. processes mdpi. https://doi.org/10.3390/pr7100685 dahlgaard, j.j., kristensen, k., & kanji, g.k. (2002). fundamentals of total quality management: process analysis and improvement jens. original illustrations © taylor & francis 2002. https://doi.org/10.4324/9780203930021 dewi, h.p., lumbanraja, p., & matondang, r. (2015). implementation of total quality management and interpersonal communication in achieving student satisfaction through service quality at yayasan pendidikan islam, miftahussalam, medan. international journal of research and review, 2(6), 343-347. http://www.gkpublication.in/ijrr_vol.2_issue6_june2015/ijrr0066.pdf eltawy, n., & gallear, d. (2017). leanness and agility: a comparative theoretical view. industrial management and data systems, 117(1), 149-165. https://doi.org/10.1108/imds-01-2016-0032 fitriani, f. (2019). persiapan total quality management (tqm). adaara: jurnal manajemen pendidikan islam, 9(2), 908-919. https://doi.org/10.35673/ajmpi.v9i2.426 garcia-alcaraz, j.l., flor-montalvo, f.j., avelar-sosa, l., sánchez-ramírez, c., & jiménez-macías, e. (2019). human resource abilities and skills in tqm for sustainable enterprises. sustainability mdpi, 11(22), 6488. https://doi.org/10.3390/su11226488 george, s., & weimerskirch, a. (1998). total quality management: strategies and techniques proven at todays’ most successful companies (second ed.). john wiley & sons, inc. green, f.b. (2006). six-sigma and the revival of tqm. total quality management and business excellence, 17(10), 1281-1286. https://doi.org/10.1080/14783360600753711 gómez-lópez, r., serrano-bedia, a.m., & lópez-fernández, m.c. (2016). motivations for implementing tqm through the efqm model in spain: an empirical investigation. total quality management and business excellence, 27(11-12), 1224-1245. https://doi.org/10.1080/14783363.2015.1068688 haffar, m., al-karaghouli, w., & ghoneim, a. (2013). an analysis of the influence of organisational culture on tqm implementation in an era of global marketing: the case of syrian manufacturing organisations. international journal of productivity and quality management, 11(1), 96-115. https://doi.org/10.1504/ijpqm.2013.050570 hasan, k., islam, m.s., shams, a.t., & gupta, h. (2018). total quality management (tqm): implementation in primary education system of bangladesh. international journal of research in industrial engineering, 7(3), 370-380. https://doi.org/10.22105/riej.2018.128170.1041 houston, d. (2007). tqm and higher education: a critical systems perspective on fitness for purpose. quality in higher education, 13(1), 3-17. https://doi.org/10.1080/13538320701272672 kaname, o. (2003). handbook for tqm and qcc vol 1. in handbook (vol. 1). kantardjieva, m. (2015). the relationship between total quality management (tqm) and strategic management. journal of economics, business and management, 3(5), 537-541. https://doi.org/10.7763/joebm.2015.v3.242 kim, g.-s. (2016). effect of total quality management on customer satisfaction. international journal of engineering sciences & research technology, 5(6), 507-514. https://doi.org/10.5281/zenodo.55618 kiruthiga, k. (2016). major factors affecting the execution of total quality management in the construction industry in india. journal of chemical and pharmaceutical sciences, 9(2), e135-e140. kumar, s., & shanmuganathan, j. (2019). a structural relationship between tqm practices and organizational performance with reference to selected auto component manufacturing companies. international journal of management, 10(5). https://doi.org/10.34218/ijm.10.5.2019/009 kumar, u., kumar, v., de grosbois, d., & choisne, f. (2009). continuous improvement of performance measurement by tqm adopters. total quality management & business excellence, 20(6), 603-616. https://doi.org/10.1080/14783360902924242 kuo, c. (2016). effects of total quality management implementation and supply chain management capability on customer capital. the journal of global business management, 12(2), 47-60. lawrence, j.j., & mccollough, m.a. (2004). implementing total quality management in the classroom by means of student satisfaction guarantees. total quality management and business excellence, 15(2), 235-254. https://doi.org/10.1080/1478336032000149063 int. j. prod. manag. eng. (2021) 9(1), 25-36 creative commons attribution-noncommercial-noderivatives 4.0 international permana et al. 34 https://doi.org/10.11648/j.sjph.20160404.11 https://doi.org/10.1080/20479700.2019.1644472 https://doi.org/10.4067/s0718-27242014000200012 http://www.iaeme.com/masteradmin/uploadfolder/ijm_07_05_021/ijm_07_05_021.pdf https://doi.org/10.3390/pr7100685 https://doi.org/10.4324/9780203930021 http://www.gkpublication.in/ijrr_vol.2_issue6_june2015/ijrr0066.pdf https://doi.org/10.1108/imds-01-2016-0032 https://doi.org/10.35673/ajmpi.v9i2.426 https://doi.org/10.3390/su11226488 https://doi.org/10.1080/14783360600753711 https://doi.org/10.1080/14783363.2015.1068688 https://doi.org/10.1504/ijpqm.2013.050570 https://doi.org/10.22105/riej.2018.128170.1041 https://doi.org/10.1080/13538320701272672 https://doi.org/10.7763/joebm.2015.v3.242 https://doi.org/10.5281/zenodo.55618 https://doi.org/10.34218/ijm.10.5.2019/009 https://doi.org/10.1080/14783360902924242 https://doi.org/10.1080/1478336032000149063 http://creativecommons.org/licenses/by-nc-nd/4.0/ mensah, j.o., copuroglu, g., & fening, f.a. (2012). total quality management in ghana: critical success factors and model for implementation of a quality revolution. journal of african business, 13(2), 123-133. https://doi.org/10.1080/15228916.2012.693444 mercy, o., & taiye, t.b. (2015). strategic imperatives of total quality management and customer satisfaction in organizational sustainability. international journal of academic research in business and social sciences, 5(4), 1-22. https://doi.org/10.6007/ijarbss/v5-i4/1538 mitreva, e., cvetkovik, d., filiposki, o., taskov, n., & gjorshevski, h. (2016). the effects of total quality management practices on performance within a company for frozen food in the republic of macedonia. tem journal, 5(3), 339-346. https://doi.org/10.18421/tem53-14 morath, c., & doluschitz, r. (2009). total quality management in the food industry – current situation and potential in germany. applied studies in agribusiness and commerce, 3(3-4), 83-87. https://doi.org/10.19041/apstract/2009/3-4/18 musenze, i.a., & thomas, m.s. (2020). development and validation of a total quality management model for uganda’s local governments. cogent business and management, 7(1), 1-22. https://doi.org/10.1080/23311975.2020.1767996 neyestani, b., & juanzon, j.b.p. (2016). developing an appropriate performance measurement framework for total quality management (tqm) in construction and other industries. ira-international journal of technology & engineering (issn 2455-4480), 5(2), 32. https://doi.org/10.21013/jte.v5.n2.p2 ngambi, m.t., & nkemkiafu, a.g. (2015). the impact of total quality management on firm’s organizational performance marcel. american journal of management, 15(4), 57-76. nicolaou, n., & kentas, g. (2017). total quality management implementation failure reasons in healthcare sector. journal of health science 5 (2017) 110-113, 5(2), 110-113. https://doi.org/10.17265/2328-7136/2017.02.007 nugroho, t.w., & nurcahyo, r. (2018). analysis of total quality management (tqm) implementation in small medium industries. proceedings of the international conference on industrial engineering and operations management, 2018(jul), 607-618. oakland, j.s. (2003). total quality management text with cases. in butterworth-heinemann (third edit). butterworth-heinemann. phan, a.c., nguyen, h.t., nguyen, h.a., & matsui, y. (2019). effect of total quality management practices and jit production practices on flexibility performance: empirical evidence from international manufacturing plants. mdpi sustainability (switzerland), 11(11). https://doi.org/10.3390/su11113093 prajogo, d.i., & brown, a. (2004). the relationship between tqm practices and quality performance and the role of formal tqm programs: an australian empirical study. quality management journal, 11(4), 31-42. https://doi.org/10.1080/10686967.2004.11919131 ramlawati, & putra, a.h.p.k. (2018). total quality management as the key of the company to gain the competitiveness, performance achievement and consumer satisfaction. international review of management and marketing, 8(5), 60-69. rogers, r.e. (2013). implementation of total quality management a comprehensive training program. 1996 by the haworth press, inc. all rights reserved. sabet, e., adams, e., & yazdani, b. (2014). quality management in heavy duty manufacturing industry: tqm vs. six sigma. total quality management and business excellence, 27(1-2), 215-225. https://doi.org/10.1080/14783363.2014.972626 sader, s., husti, i., & daróczi, m. (2017). suggested indicators to measure the impact of industry 4.0 on total quality management. international scientific journal: industry 4.0, 2(6), 298-301. https://stumejournals.com/journals/i4/2017/6/298/pdf sadikoglu, e., & olcay, h. (2014). the effects of total quality management practices on performance and the reasons of and the barriers to tqm practices in turkey. laboratory management information systems: current requirements and future perspectives, 2014, 996-1027. https://doi.org/10.1155/2014/537605 sainis, g., haritos, g., kriemadis, t., & fowler, m. (2017). the quality journey for greek smes and their financial performance. production and manufacturing research, 5(1), 306-327. https://doi.org/10.1080/21693277.2017.1374891 santos, a.c. de s.g. dos, carvalho, l.m., souza, c.f. de, reis, a. da c., & freitag, a.e.b. (2019). total quality management: the case of an electricity distribution company. brazilian journal of operations & production management, 16(1), 53-65. https://doi.org/10.14488/bjopm.2019.v16.n1.a5 sari, & firdaus, a. (2018). the impact of total quality management implementation on small and medium manufacturing companies. esensi: jurnal bisnis dan manajemen, 8(1), 67-78. https://doi.org/10.15408/ess.v8i1.5852 sila, i., & walczak, s. (2017). universal versus contextual effects on tqm: a triangulation study using neural networks. production planning and control, 28(5), 367-386. https://doi.org/10.1080/09537287.2017.1296598 sivalai, t., & rojniruttikul, n. (2018). determinants of the state railway of thailand’s (srt) total quality management process: sem analysis. journal of international studies, 11(2). https://doi.org/10.14254/2071-8330.2018/11-2/9 small, e.p., ayyash, l., & hamouri, k. al. (2017). benchmarking performance of tqm principals in electrical subcontracting in dubai: a case study. procedia engineering, 196(june), 622-629. https://doi.org/10.1016/j.proeng.2017.08.050 sousa-mendes, g.h. de, gomes-salgado, e., & moro-ferrari, b.e. (2016). prioritization of tqm practices in brazilian medical device smes using analytical hierarchy process (ahp) glauco. dyna (colombia), 83(197), 195-203. https://doi.org/10.15446/dyna.v83n197.52205 steiber, a., & alänge, s. (2013). do tqm principles need to change? learning from a comparison to google inc. total quality management and business excellence, 24(1-2), 48-61. https://doi.org/10.1080/14783363.2012.733256 int. j. prod. manag. eng. (2021) 9(1), 25-36creative commons attribution-noncommercial-noderivatives 4.0 international a systematic literature review of total quality management (tqm) implementation in the organization 35 https://doi.org/10.1080/15228916.2012.693444 https://doi.org/10.6007/ijarbss/v5-i4/1538 https://doi.org/10.18421/tem53-14 https://doi.org/10.19041/apstract/2009/3-4/18 https://doi.org/10.1080/23311975.2020.1767996 https://doi.org/10.21013/jte.v5.n2.p2 https://doi.org/10.17265/2328-7136/2017.02.007 https://doi.org/10.3390/su11113093 https://doi.org/10.1080/10686967.2004.11919131 https://doi.org/10.1080/14783363.2014.972626 https://stumejournals.com/journals/i4/2017/6/298/pdf https://doi.org/10.1155/2014/537605 https://doi.org/10.1080/21693277.2017.1374891 https://doi.org/10.14488/bjopm.2019.v16.n1.a5 https://doi.org/10.15408/ess.v8i1.5852 https://doi.org/10.1080/09537287.2017.1296598 https://doi.org/10.14254/2071-8330.2018/11-2/9 https://doi.org/10.1016/j.proeng.2017.08.050 https://doi.org/10.15446/dyna.v83n197.52205 https://doi.org/10.1080/14783363.2012.733256 http://creativecommons.org/licenses/by-nc-nd/4.0/ suarez-barraza, m.f., & ablanedo-rosas, j.h. (2014). total quality management principles: implementation experience from mexican organisations. total quality management and business excellence, 25(5-6), 546-560. https://doi.org/10.1080/14783363.2013.86 7606 sukardi, r.a. (2016). pengaruh total quality management (tqm) terhadap kepuasan pelanggan pada matahari department store di plaza mulia samarinda. ejournal administrasi bisnis, 4(3), 758-772. sukdeo, n., pretorius, j.h., & vermeulen, a. (2017). the role of total quality management (tqm) practices on improving organisational performance in manufacturing and service organisations. proceedings of the international conference on industrial engineering and operations management, 2017(oct), 1133-1152. sutrisno, t.f.c.w. (2019). relationship between total quality management element, operational performance and organizational performance in food production smes. jurnal aplikasi manajemen, 17(2), 285-294. https://doi.org/10.21776/ub.jam.2019.017.02.11 sweis, r., ismaeil, a., obeidat, b., & kanaan, r.k. (2019). reviewing the literature on total quality management and organizational performance. journal of business & management (coes&rj-jbm), 7(3), 192-215. https://doi.org/10.25255/jbm.2019.7.3.192.215 talib, f., & rahman, z. (2015). identification and prioritization of barriers to total quality management implementation in service industry: an analytic hierarchy process approach. tqm journal, 27(5), 591-615. https://doi.org/10.1108/tqm-11-2013-0122 tervonen, p., pahkala, n., & haapasalo, h. (2009). development of tqm in steel manufacturers’ production. ibima business review, 1-3, 52-59. tesfaye, g., & kitaw, d. (2017). a tqm and jit integrated continuous improvement model for organizational success: an innovative framework. journal of optimization in industrial engineering, 22, 15-23. https://doi.org/10.22094/joie.2017.265 vukomanovic, m., radujkovic, m., & nahod, m.m. (2014). efqm excellence model as the tqm model of the construction industry of southeastern europe. journal of civil engineering and management, 20(1), 70-81. https://doi.org/10.3846/13923730.2013.843582 yang, c.o., & tsai, m.c. (2014). improving operations performance through tqm in the post-financial crisis era: an exploratory case study of a multinational im firm in the greater china region. total quality management and business excellence, 25(5-6), 561-581. https://doi.org/10.1080/14783363.2013.839167 yeng, s.k., jusoh, m.s., & ishak, n.a. (2018). the impact of total quality management (tqm) on competitive advantage: a conceptual mixed method study in the malaysia luxury hotel industries. academy of strategic management journal, 17(2), 1-9. zairi, m. (1991). total quality management for engineers. in ccc (vol. 1). woodhead publishing limited. https://doi.org/10.1533/9781845698911.1 žitkienė, r., & deksnys, m. (2018). organizational agility conceptual model. montenegrin journal of economics, 14(2), 115-129. https://doi.org/10.14254/1800-5845/2018.14-2.7 int. j. prod. manag. eng. (2021) 9(1), 25-36 creative commons attribution-noncommercial-noderivatives 4.0 international permana et al. 36 https://doi.org/10.1080/14783363.2013.867606 https://doi.org/10.1080/14783363.2013.867606 https://doi.org/10.21776/ub.jam.2019.017.02.11 https://doi.org/10.25255/jbm.2019.7.3.192.215 https://doi.org/10.1108/tqm-11-2013-0122 https://doi.org/10.22094/joie.2017.265 https://doi.org/10.3846/13923730.2013.843582 https://doi.org/10.1080/14783363.2013.839167 https://doi.org/10.1533/9781845698911.1 https://doi.org/10.14254/1800-5845/2018.14-2.7 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2018.10369 received 2017-06-08 accepted: 2018-07-23 categorisation of the main disruptive events in the sensitive products transportation supply chains quentin schoena1,b, raquel sanchisc1, raul polerc2, matthieu laurasa2, franck fontanilia3 and sébastien truptila4 a imt mines albi, campus jarlard, 81013 albi cedex 9, france. a1 quentin.schoen@mines-albi.fr, a2 matthieu.lauras@mines-albi.fr, a3 franck.fontanili@mines-albi.fr, a4 sebastien.truptil@mines-albi.fr. betablissement français du sang occitanie, avenue de grande bretagne, 31300 toulouse, france. quentin.schoen@efs.sante.fr cresearch centre on production management and engineering (cigip). universitat politècnica de valència (upv). calle alarcón, 03801 alcoy (alicante), spain. c1 rsanchis@cigip.upv.es, c1 rpoler@cigip.upv.es abstract: the upcoming logistic environment is about to modify deeply the way we supply products. in fact, some new trends are going to require more and more agility between a large number of stakeholders in open and dynamic networks. this should be possible to achieve thanks to new data collection and treatment abilities. considering this moving technological and logistic environment, it appears necessary to define and categorize more specifically the main disruptive events that can affect a supply chain. in fact, amount of data are collected on the field and must be helpful to make relevant decisions in case of disruption. in order to understand automatically what these data mean, it is necessary to detect and classify the disruptive events in order to find the best adaptation. this paper focuses on the sensitive products’ supply chains, that are facing with agility high requirements, based on their ability to detect disruptive events. we take as an example the blood supply chain. key words: disruptive event, supply chain, sensitive products, resilience, agility. 1. introduction in order to organize the vehicles rounds to pick-up and/or deliver products, companies are using scheduling systems, trying to deal with the vehicle routing problem (vrp) problem (toth and vigo, 2002). despite the optimized schedule, a lot of events may affect the on-going process like traffic jam, bad weather conditions, unexpected demand, vehicle failure, etc. these undesirable situations may have negative impacts on the supply chains and transportation performance. for some years we assist to a supply chain complexification on one hand and to very quick changes on the technological abilities on the other hand. the logistic environment evolves quickly and some new trends are about to modify the way goods are transported. the product customization and the temperature controlled supply chain imply a strict and effective traceability during the transportation steps in order to ensure the delivery of the right product at the right place and moment in the best conditions. meantime, logistic as a service schemes and responsibilities splitting increase the number of stakeholders involved in the transportation of each box between the sender and the recipient. finally, the supply chain customization makes the whole network more complex to satisfy the customers’ requirements who require anything / anytime / anywhere pick-up and delivery services. to cite this article: schoen, q., sanchis, r., poler, r., lauras, m., fontanili, f. and truptil, s.. (2018). categorisation of the main disruptive events in the sensitive products transportation supply chains. international journal of production management and engineering, 6(1), 79-89. https://doi.org/10.4995/ijpme.2018.10369 int. j. prod. manag. eng. (2018) 6(2), 79-89creative commons attribution-noncommercial-noderivatives 4.0 international 79 http://creativecommons.org/licenses/by-nc-nd/4.0/ thus, in order not to be drawn under the possibilities and demands, it appears necessary to improve agility on the whole supply chain in order to take advantage of the upcoming hyperconnected network and react in case of unexpected events. as an example, we examine in this study the case of sensitive products transportation. we define a transport as a sensitive one (i) because it is risky for the product itself (fragility, dangerous for the environment, high risk of theft, etc.) or/and (ii) because if a certain quality degree is not reached in the service and/or product (delay in the delivery, product injured). this will have high-impact negative consequences downstream the supply chain. the first category involves products that are expensive (electronics, precious metals, etc.), dangerous for the environment (radioactive, chemicals, etc.), rare (piece of art), fragile (animals, objects sensitive to shocks), and/or vital (organ, drugs). the second one involves any object, even water, blankets or dry food needed during humanitarian crisis for instance. in such a situation, common products are involved in sensitive transportation steps because they are required as a matter of emergency. these two categories are obviously not separated. indeed, as we explain it hereafter, some transportation combine both aspects when the type of goods transported are related to human health, as there is, besides the economic loss, the potential loss of human lives. for this reason, it is important to shape an agile supply chain with the main aim to build a resilient one, as agility is one of the key characteristics of the resilience capacity. when working effectively and efficiently, supply chains allow goods to be manufactured and transported in the right quantities, to the right places at the right time and with the right conditions in a cost effective manner (christopher et al., 2004). resilient supply chains allow goods to be delivered effectively and efficiently under unstable conditions that is the frequent way of supply chains actual operation. a resilient supply chain must be adaptable to face up to undesirable situations efficiently and to guarantee the survival (ruiz-martin et al., 2018). resilient processes should be agile to change quickly (christopher, 2005). christopher and peck’s (2004) conceptualization of a resilient supply chain includes the agility as a basic characteristic to recover after being disrupted and to provide a more rapid response to changed conditions. moreover, it is important to highlight that the feature of agility is even more crucial in the case of sensitive products, in order to react to an undesirable situation smartly to minimize the negative impacts. the objective of this paper is to categorize the main disruptive events that can affect the normal operation of a sensitive product supply chain focusing on the transportation step. this is a starting point to focus on agile aspects and the disruptive events detection to build a resilient supply chain. the agility detection step, put forward by barthe-delanoë et al., (2013) is a key element in order to automatically identify that something is going wrong. the identification of the main disruptive events that negatively affect a sensitive product supply chain has been performed based on the categorisation framework of disruptions developed by sanchis and poler (2014) and on the exhaustive study and analysis of the blood transportation supply chain in france. this has been performed through the historical data provided by the etablissement français du sang (french blood establishment: efs) of a whole region with 25 sites, representative of the other one in france. figure 1 shows the relationship between the different elements of the scope of this research (grey ones) whose main aim is focused on the categorisation of the main disruptive events in the transport of the sensitive products supply chains. resilient supply chains flexibility robustness redundancy agility … detection adaptation reactivity figure 1. conceptualization of the research scope to build a resilient supply chain. the paper is organized as follows. section 2 describes the literature review related to the upcoming logistic environment and how it will increase the need of resilience and agility. a former categorisation framework of disruptions is studied. section 3 focus on our proposal consisting of an extended version of the categorisation framework of disruptions and its use in the sensitive products supply chain during transportation steps. finally, section 4 uses the efs use case to discuss this proposition. int. j. prod. manag. eng. (2018) 6(2), 79-89 creative commons attribution-noncommercial-noderivatives 4.0 international schoen, q., sanchis, r., poler, r., lauras, m., fontanili, f. and truptil, s. 80 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. literature review 2.1. the upcoming logistic environment the logistic environment evolves quickly regarding the technologies and trends. this study considers some upcoming technologies that would allow improvement in our ability to collect and treat data automatically and in real time. this should be useful to take up the challenges of being resilient in the future logistic environment. first, thanks to new mobile devices able to collect and transmit data on the internet (iot: internet of things), the data collection in real time should not be a problem in a near future. these new technologies, linked with the cloud computing is creating a hyperconnected network of devices affordable by anyone (harris et al., 2015). big data aspects, described by wamba et al. (2015) as a combination of data volume, velocity, variety, veracity and value were a limitation to extract added value from these data-collection. considering how important and relevant must be considered the ability to use these data to support the decision making process in real time, we can expect in a near future that this kind of technologies, accessible through software as a service model will spread in the whole supply chain. harris et al., (2015) explain that the small and medium sized enterprises (smes) should also take advantage of these technologies. on a network aspect, we are usually considering predetermined and fixed network. the multimodality, defined by the united nations (1980), may represent a risk for the products to be transported because it is no longer possible to consider the vehicle as a black box between the sender and the recipient. currently, the transshipments may endanger the products (storage conditions, forgotten container, theft, etc.) whereas they may increase the effectiveness combining the advantages of each transportation mean used (harris et al., 2015). in light of the new data collection and treatment abilities, it should be easier to control these steps and less hazardous to organize them. moreover, as it is the case for softwares from several years, logistics activities are split between different companies like 3-4-5pl (party logistics) firms subcontracting a increasingly larger perimeter. this “service” trend, visible and used by the consumers (uber, airbnb, deliveroo, etc.) should spread in logistics activities with companies responsible of just one part of the whole supply chain. the last mile delivery is one of these potential perimeters. the distribution customization based on “anytime, anywhere delivery model” (dhl, 2016) complexifies the network between the sender and the recipient, requiring more agility and ability to pick up and deliver the products the closest to the sender and recipients. finally, automation inside hubs and automated vehicles themselves (van meldert et al., 2016) should provide more transportation possibilities and contribute to build an open and agile network. from the product point of view , we are currently tracking in real time the vehicles when containers are loaded inside. in this evolving network, more stakeholders involved in the supply chain and more transshipment may be settled. in this situation, with numerous stakeholders in charge of each container, instead of tracking the vehicles, it must be valuable to track directly each box. this idea is a part of the physical internet philosophy, making a parallel between the box on the supply chain network and the packets on the internet network (montreuil, 2011). this point of view appears relevant in the expected network and even more in a product customization trend. in fact, this “batch size one production” (dhl, 2016) prevents the possibility of replacing a container damaged by another one in the same storage area, because each product is different. a supply chain issue on a container may require to manufacture again the products inside. thus, the sensitivity of each of them increases as the traceability needs. moreover, the climate-controlled supply chain demand increases and (bogataj et al., 2005) the laws that control it are very strict. this sensitive supply chain also involves anomaly detection in real time to be resilient. figure 2 illustrates these evolutions, putting forward the increasingly more complex logistic environment. in this hyperconnected and open network where the containers are tracked independently in real time, resilience is a key element.   figure 2. expected supply chain environment. int. j. prod. manag. eng. (2018) 6(2), 79-89creative commons attribution-noncommercial-noderivatives 4.0 international categorisation of the main disruptive events in the sensitive products transportation supply chains 81 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2.2. resilience and agility resilient supply chains allow to detect the potential and/or the real disruptive event occurrence and increase agility. resilience was firstly coined by holling (1973), as a system that persists in a state of equilibrium (stability) and how dynamic systems behave when they are stressed and move from this equilibrium. gallopin (2006) explains that the capacity of resilience is the enabler for enterprises and supply chains to decrease the level of vulnerability to expected and unexpected risks, to determine how agile it is in reorganizing itself despite its changing environment, and assess how effective it may be in recovering in the least possible time and at the least possible expense. morales allende et al., (2017) explain that there is not entire consensus about the definition of the term resilience. there is a high amount of definitions in the literature. table 1 shows some of these literature review’s definitions of resilience but, in this case, applied to supply chains to understand the importance of being resilient to face potential disruptive events but also to identify the different elements that are related to this essential supply chains’ ability. sanchis and poler (2013) state that the ability of resilience in an enterprise or its supply chain is function of (i) the vulnerability; (ii) the adaptative capacity (adaptation) and (iii) the recovery ability. moreover the authors propose a framework with the main research areas related to the study of resilience that will serve as the foundation for further research. this framework relates the adaptation to aspects such as flexibility, agility, robustness, redundancy, … among others. one of the main remarkable features of the supply chain to be able to adapt to the new situation provoked by a disruptive event is the agility with which the supply chain is able to manage this new undesirable situation. it is conceptualised as the ability to better deal with unexpected events, to overcome unforeseen situations of business environment as to take benefits and opportunities of changes (swafford et al., 2008). therefore, agility is a prerequisite for building resilient supply chains. moreover, barthe-delanoë et al., (2013) explains that agility is a function of detection, adaptation and reactivity as follows: agility = (detection + adaptation) × reactivity (1) starting with the first element, detection is very important as the supply chain is not aware about the table 1. supply chains resilience definitions authors definition ponomarov and holcomb (2009) the adaptive capability of the supply chain to prepare for unexpected events, respond to disruptions, and recover from them by maintaining continuity of operations at the desired level of connectedness and control over structure and function. barroso et al., /2011) ability to react to the negative effects caused by disturbances that occur at a given moment in order to maintain the supply chain’s objectives. berle et al., (2011) ability of the supply chain to handle a disruption without significant impact on the ability to serve the supply chain mission ponis and koronis, (2012) the ability to proactively plan and design the supply chain network for anticipating unexpected disruptive (negative) events, respond adaptively to disruptions while maintaining control over structure and function and transcending to a post-event robust state of operations, if possible, more favourable than the one prior to the event, thus gaining competitive advantage wieland and marcus-wallenburg, (2013) the ability of a supply chain to deal with change either proactively or reactively. brandon-jones et al., (2014) the ability of a supply chain to return to normal operating performance, within an acceptable period of time, after being disturbed. int. j. prod. manag. eng. (2018) 6(2), 79-89 creative commons attribution-noncommercial-noderivatives 4.0 international schoen, q., sanchis, r., poler, r., lauras, m., fontanili, f. and truptil, s. 82 http://creativecommons.org/licenses/by-nc-nd/4.0/ disruptive event happening and most important, it is crucial to detect it as soon as possible, preferably in “real time”, in order to be able to take the most appropriate decisions and implement the most adequate actions to reduce and minimize the undesirable consequences of the disruptive event occurrence. adaptation is the following element and tries to modify the current conditions of the supply chain to make them compliant with the new supply chain circumstances. the last element is the reactivity which is in charge of altering supply chain behaviour when a disruptive event occurrence has been detected and the supply chain is adapting to this new context. agility allows on one hand to detect if the ongoing processes meet the requirements of the current situation, on the other hand to adapt the ongoing processes if necessary (barthe-delanoë et al., 2013). therefore, to build resilient supply chains, it is necessary to be as agile as possible. this also depends on the detection element to be able to discover that the current situation is not the expected one. for this reason, having deep knowledge about the potential disruptive events is not enough as it is also desirable to have the appropriate means to monitor and detect what causes the lack of supply chain resilience, i.e. to categorize the most relevant and probable disruptive events. 2.3. categorisation framework of disruptions the identified disruptive events have been classified based on the already defined categorisation framework of disruptions developed by sanchis and poler (2014). this framework was applied to typical manufacturing supply chains and the authors did not consider its application to other types as the sensitive products transportation supply chains. in this paper, we will confirm if the categorisation framework of disruptions presents enough generality to be applied to any singular supply chain, even though to any logistic environment, or on the contrary it needs to be updated. when developing the categorisation framework of disruptions, sanchis and poler (2014) pointed out that there is a high degree of confusion both in terms of disruption and in its constituent elements. for example a forest fire during summer nearby a road is a disruptive event whose source might be a cigarette. the consequences are the roads nearby cut off and the forest destruction. focusing here on the forest, it is easy to identify the source, the disruptive event and the consequences. however, considering the road, the cut off is the triggering event, the source is the fire and the consequences are deviations and traffic jam. next, considering a truck which contains sensitive products, the disruptive event is the traffic jam, the source is the road cut off and the consequences are still to be defined, depending on the products, their sensitivity (related to their characteristics or the context/demand), etc. however, we could consider that the source of traffic jam is the forest fire, setting apart the road cut off. this short example put forward the difficulty to clearly define the system considered and the disruption elements among all the events included in a butterfly effect.   figure 3. butterfly effect showing the difficulty to define clearly the nature of each event. svensson (2000) and kleindorfer and saad (2005) define disruption as an unexpected event that interrupts the normal flow of products and materials in a supply chain. this definition does not consider the ability to foresee this event and to reduce the risk of occurrence. barrroso et al. (2008) takes it into account and define disruption as a foreseeable or unforeseeable event, which affects directly the usual operation and stability of an enterprise or its supply chain. in this research work we add to this last definition the fact that this disruptive event is potentially damaging for the considered supply chain and its components if some parameters are forgotten in the analysis step. for instance, during its round, a carrier receives the instruction to deviate from this initial plan in order to deliver a customer who have ordered a product unexpectedly. inside the vehicle, if the containers/products state and their constraints are not checked before making the decision of deviation, the adaptation solution found may be more damaging than other solutions. in fact, the delay this deviation implies for the following deliveries may not be allowed or could injure the products. we consider int. j. prod. manag. eng. (2018) 6(2), 79-89creative commons attribution-noncommercial-noderivatives 4.0 international categorisation of the main disruptive events in the sensitive products transportation supply chains 83 http://creativecommons.org/licenses/by-nc-nd/4.0/ here that all the parameters require to make an adapted decision are detected by sensors on the field or indirectly. the categorisation framework of disruptions (sanchis and poler, 2014) considers that a disruption is composed by 3 elements as it is shown in figure 4: source: the trigger that causes and originates the disruption. disruptive event: incident that causes an expected or unexpected disturbance that alters the normal operation of the supply chain. consequence: impact of the disruptive event in form of negative effects on the supply chain. level/origin source disruptive event consequences perturbance impact disruption figure 4. elements of the categorisation framework (source: based on sanchis and poler, 2014). the authors divide the source element into the level in which the disruption have its origin: (i) within an enterprise of the supply chain, (ii) outside an enterprise but internal to the supply chain and (iii) external to the supply chain; and into the origin that causes the disruptive event, considering accidental, customer, energetic, equipment, financial, information and communications technologies (ict), infrastructure, man-made, natural, political, product, regulatory, supplier and terrorism. the consequences element is also divided into the following ones: (i) business interruption; (ii) damage to reputation/brand; (iii) delays and failure of due dates; (iv) failure to attract or retain top talent; (v) failure to meet customer needs; (vi) high inventories; (vii) impossibility to pay personnel, suppliers, taxes; (viii) increase of final products price; (ix) increase of production costs; (x) injury to end customers; (xi) injury to workers; (xii) loss of intellectual property/ data; (xiii) loss of networked communication; (xiv) physical damage; (xv) sales decrease; (xvi) understaffing; (xvii) unfulfilled orders. moreover, the authors differentiate the initial impact, the quickest negative signals of the disruptive event occurrence and the long-term consequences as the lasting ones. table 2 shows an example of the original categorisation framework of disruptions. as it could be observed, the examples are more related to a typical manufacturing supply chain. however, in this paper, we would evaluate if the categorisation framework of disruptions is as generic as necessary to be able to be applied to any type of supply chain. 3. proposal: extended categorisation framework and disruptive events 3.1. taking the detection into account based on the french blood establishment (efs) experience and the historical data about the main disruptive events that affect the blood transportation supply chain, the identification of the core disruptive events has been performed. moreover, for each of the disruptive events identified, the source where the disruptive event is originated and the consequences, as the negative outcome of the disruptive event occurrence have been analysed in detail accordingly table 2. small fragment of the categorisation framework of disruptions of sanchis and poler (2014).   level primary origin disruptive event initial impact long-term consequences in-, inter-, extraaccidental fire, gas leak, exp losions… injury to workers i, iii, v, xi, xii, xiii, xiv, xvii in-, intercustomer unanticip ated or very volatile demand high inventories / delay s and failure of due dates v, ix, xv, xvii in-, inter-, extraenergetic energy /water interrup tion (electricity , gas…) business interrup tion ii, iii, viii, ix, xvi in-, inter-, extraequip ments breakdown of machinery delay s and failure of due dates i, ii in-, inter-, extrafinancial economic slowdown sales decrease i, iv, vii in-, inter-, extraict lack of technology infrastructure to sup p ort business needs failure to meet customer needs xiii in-, inter-, extrainfrastructure transp ortation infraestructure failure (e.g.rail disrup tions) delay s and failure of due dates ii in-, inter-, extram an-made crime/theft/fraud/emp loy ee dishonesty loss of intellectual p rop erty /data ii, iii, v, xvi in-, inter-, extranatural natural disasters (e.g. earthquake, flooding, tsunami, tornados…) iii, v, xi, xiii, xiv, xvi in-, inter-, extrapolitical political instability or other socio-p olitical crises i, iii, v, x, xi, xiii, xiv, xvi, xvii in-, inter-, extraproduct nocive substances in p roducts damage to rep utation/brand / injury to end customers ii, v, x, xv in-, inter-, extraregulatory regulatory and legislative changes injury to end customers v, viii, ix, xv in-, intersup p lier natural resource scarcity /unavailability of raw materials delay s and failure of due dates i, ii, v, viii, xvii in-, inter-, extraterrorism international terror attacks business interrup tion ii, iii, v, xi, xii, xiii, xiv, xv, xvi, xvii … … … … … s ources consequences business interrup tion int. j. prod. manag. eng. (2018) 6(2), 79-89 creative commons attribution-noncommercial-noderivatives 4.0 international schoen, q., sanchis, r., poler, r., lauras, m., fontanili, f. and truptil, s. 84 http://creativecommons.org/licenses/by-nc-nd/4.0/ to the categorisation framework of disruptions presented. however, to apply this framework to the real and current situation of the sensitive products’ supply chains and considering the expected one, it has been recognized the necessity to add another element that put forward the detection that a disruptive event is happening or potentially will happen (deviation from the expected situation). for this reason, the existing framework (figure 4) has been extended (figure 5) including the detection element, that can be defined as the signal that allows the identification and perception of a disruptive event occurring or potentially occurring. level/origin source disruptive event detection perturbance signal disruption consequences impact figure 5. elements of the extended version of the categorisation framework of disruptions. in fact, in sensitive supply chains, if the products are transported through hundreds of kilometres, the detection step of a disruptive event is not as easy as it can be in a manufacturing factory where almost everything is under control or at least, delimited. during transportation processes, some parameters are not under control (traffic, weather…) and may affect others already difficult to monitor because of detection issues or data broadcast (truck compartment load and state, container state, vehicle location/speed/state, client demand). this surveillance aspect needs to be continuously reviewed and monitored in order to detect “in real time” or after few minutes the disruptive events. sensitive products, like the blood ones, may be damaged after few minutes and in case of urgent delivery, the receiver may not be able to wait for few more minutes. even if it may be easy for the driver to detect it through embedded sensors, in order to react correctly and in accordance to all the other containers transported and vehicles deployed, the information sharing and processing among the whole agents involved is essential. in fact, it is necessary to analyse correctly the whole situation to preview the consequences and make an adapted decision based on all the available means, considering the context (traffic, weather, distances, time, etc.) and current needs. therefore, on the detection step depends our ability to detect the disruptive event as soon as possible after it appears with enough relevant and reliable data. gathering them, it allows to draw a relevant picture of the situation in real time and to anticipate the consequences. the objective of detecting these events and being able to react with agility is to reduce as much as possible the negative impact of the consequences on the supply chain and the products. based on this, figure 5 shows the extended version of the categorisation framework of disruptions to be applied in the current sensitive products supply chain and broadly in the expected logistic environment. 3.2. extended categorisation framework of disruptions applied to sensitive products supply chain the identification of the main elements that shape the extended version of the categorisation framework of disruptions has been mainly performed through the analysis of the historical data provided by the efs, due to, our best knowledge, the lack of information in the literature. this supply chain, described more specifically in section 4 deals with very sensitive products through thousands of kilometres everyday and allow us to describe in table 3 the extended categorisation framework of disruptions to sensitive supply chain in general. the source element has been simplified as it has been only categorized the origin in which the disruptive event is happening. it is already preconceived that for the analysed disruptive events (table 3), the level in which the disruptive event occurs is external to the sensitive product transportation supply chains as it is related to the traffic, natural conditions, man-made disruptions, among others. the main disruptive events identified are eight. the number is not very high but the consequences if these disruptive events occur, could be, in some cases, disastrous. in fact, considering the both aspects that lead to define a supply chain as sensitive (cf. introduction) either the products themselves may be injured or the receiver stuck in a damaging situation. int. j. prod. manag. eng. (2018) 6(2), 79-89creative commons attribution-noncommercial-noderivatives 4.0 international categorisation of the main disruptive events in the sensitive products transportation supply chains 85 http://creativecommons.org/licenses/by-nc-nd/4.0/ finally, the detection element has been included as it is difficult to notice a disruptive event happening, mainly in real time. for this reason, it is so crucial to monitor continuously this element to be able to respond in an agile way. one of the challenging issue here is not just to detect and transmit these data to the monitoring system but to treat them in real time in order to understand what is happening globally on the field and find adapted solutions. as it appears in table 3, the categorisation is adapted to the upcoming environment described above, focusing on the container and not just on the vehicles. knowing the high number of stakeholders potentially involved in container transportation steps between a sender and a receiver, the detection must be as technically standardized as possible regardless of the disruption or the sensors used. we discuss in the last section the relevancy of this categorisation, considering the sensitive blood supply chain. 4. blood supply chain discussion we develop here aspects of the blood supply chain in order to study the extent to which the categorisation presented may be implemented. the blood transportation supply chain is considered as a sensitive one because of its own structure and the products sensitivity the efs deals with. in fact, in order to provide to the hospital enough blood products, the efs has to collect around 10.000 blood bags everyday through the country. each year, 2 millions of blood bags (2/3) are collected in “one-day collection sites” settled in public spaces, universities and firms. two major constraints imply to organize collection sites everywhere in france in thousands of different sites. firstly, donors have to wait at least 2 months between two blood donations, and several weeks between two platelets of plasma donation. as a consequence, the table 3. extended version of the categorisation framework of disruptions applied to sensitive products supply chains during transportation steps. source (origin) disruptive event detection consequences transport traffic jam current vehicle or container position is different from the expected position • disorganization (delay/cancellation of product delivery) natural bad weather conditions • increasing the risk of accident • product damage risks growth (delay, constraints respect, etc.) human theft either a position sensor on the container or a door sensor and detection by the driver or the recipient • malicious/dangerous use of the product forgotten or lost container either a position sensor on the container, the transport management system, the sender or the recipient • product damage risks growth (delay, constraints respect, etc.) • undelivered package vehicle truck compartment or container physical parameter failure physical parameter (temperature, humidity, etc.) out of limit during x minutes • disorganization (delay/cancellation of product delivery) vehicle breakdown current vehicle or container position is different from the expected position • product damage risks growth (delay, constraints respect, etc.) demand unexpected product order order received by phone call or directly on the transport management system • disorganization (delay/cancellation of product delivery) because of the verification of re-routing possibilities if a valid product is in the truck unexpected and urgent product order int. j. prod. manag. eng. (2018) 6(2), 79-89 creative commons attribution-noncommercial-noderivatives 4.0 international schoen, q., sanchis, r., poler, r., lauras, m., fontanili, f. and truptil, s. 86 http://creativecommons.org/licenses/by-nc-nd/4.0/ efs must not collect blood at the same place too often because potential donors will not be able to give their blood again. secondly, in order to store the highest diversity of blood types, the efs has to collect blood everywhere in the country in order to be close to the broadest diversity of people. however, the blood bags have to be processed in order to separate their components (plasma, platelets, and red blood cells) and tested to check the innocuousness of each donation. these steps require laboratory professionals and expensive equipment. thus, 12 efs centres process every day the 10000 donations and 4 qualification laboratories test everyday all the samples associated with the blood bags (more than 50000 samples). then, the blood products are stored in 130 distribution centres, the nearest to the hospitals where they are used to care receivers. as it can be observed, this supply chain has an x structure with a lot of suppliers sending raw materials to few processing centres that send end products to a lot of distribution centres. this implies transportation steps over around 20 million kilometres every year, dealing with around 2 million containers. the main difficulty is to reconcile this aspect with the sensitivity of each product in terms of storage temperature and lifecycle as described in table 4. table 4. labile blood products storage conditions. products lifetime temperature whole blood less than 48 h 18-24°c blood samples less than 48 h 2-10°c platelets 5 days 20-24°c shaken red blood cells 42 days 2-10°c plasma 1 year < –25°c finally, because it is not possible to store each type of product everywhere and some of them are required as a matter of emergency, a few dozens of urgent and unexpected transports are settled everyday on demand. as it can be viewed, the blood transportation supply chain presents high complexity due to the product characteristics and the usage of this singular product. for this reason, the categorisation framework of disruptions has been applied to this unusual case as an initial attempt to illustrate the main characteristics of the blood transportation supply chains and as a starting point to address all the efforts to detect as soon as possible the disruptive events categorised and adapt efficiently to the new situation to shape a resilient supply chain. the disruptive events presented in table 3 are coherent with the blood supply chain. unexpected demand coming from hospital may appear anytime while the trucks full of blood products are proceeding their round and it may be useful to deviate one of them. moreover, because of temperature sensitivity, being able to detect a temperature issue in a compartment is essential in order to find an adapted way of recovery. this framework should contribute to develop methods and tools to assess, analyse and propose actions to improve the resilience capacity in the sensitive products transportation and, in turn, the blood transportation supply chain. 5. conclusions the application of the extended version of the categorisation framework of disruptions establishes the starting point to study and analyse the proper actions to be implemented with regards to the current blood supply chain focusing on the transportation steps and the upcoming logistic environment. the special features of the blood transportation make this categorisation more critical if possible, as, besides the economic losses; we are talking about potential human lives losses. it is very important to know which disruptive events are the most serious (either per frequency or per criticality) to focus all the efforts on these ones to minimize the negative impact of its consequences. moreover, it is also very important to define the detection elements as they allow the discovery that the real situation is not the expected one. this element will provide reliable and real-time information that will permit to take decisions quicker and more efficient. the categorisation framework of disruptions provides a useful first direction to address the research towards the most appropriate actions to respond in an agile way and shape resilient supply chains. further research will be focused on defining the appropriate and specific detection means to be able to receive the relevant information when necessary and to define the appropriate actions to be implemented in each of the disruptive events identified based on int. j. prod. manag. eng. (2018) 6(2), 79-89creative commons attribution-noncommercial-noderivatives 4.0 international categorisation of the main disruptive events in the sensitive products transportation supply chains 87 http://creativecommons.org/licenses/by-nc-nd/4.0/ the information received from an agile perspective to achieve a resilient sensitive products supply chain. a machine learning perspective could be used in order to deduce for each case the best adaptation solution to implement. finally, these detection events may come from any device, internal (our own sensors) or external (weather, traffic jam, container/vehicle position, etc.) to the company. thus it is necessary to be able to collect and understand them. an adapted disruptive events metamodel must be necessary and this represents an interesting research topic to work on. acknowledgements this research work has been supported by the european virtual laboratory for enterprise interoperability, we would like to thank here. the french blood establishment (efs) is equally thanked for the data and examples used to validate and discuss our proposal. references barroso, a.p., machado, v.h., machado, v.c. (2008). a supply chain disturbances classification. in proceedings of the ieee international conference on industrial engineering and engineering management, singapore. https://doi.org/10.1109/ieem.2008.4738196 barroso, h.p., machado, v.h., machado, v.c. (2011). supply chain resilience using the mapping approach. supply chain management, pengzhong li (ed.), intech, available from: http://www.intechopen.com/articles/show/title/supply-chain-resilience-using-the-mappingapproach. barthe-delanoë, a.m., lauras, m., truptil, s., bénaben, f., pingaud, h. (2013). a platform for event-driven agility of processes: a delivery context use-case. in collaborative systems for reindustrialization (pp. 681-690). springer berlin heidelberg. https://doi.org/10.1007/978-3-642-40543-3_71 bogataj, m., bogataj, l., vodopivec, r. (2005). stability of perishable goods in cold logistic chains. international journal of production economics, 93, 345-356. https://doi.org/10.1016/j.ijpe.2004.06.032 brandon-jones, e., squire, b., autry, c., petersen, k.j. (2014). a contingent resource-based perspective of supply chain resilience and robustness. journal of supply chain management, 50(3), 55-73. https://doi.org/10.1111/jscm.12050 christopher, m. (2005). managing risk in the supply chain. in christopher, m. (ed.), logistics & supply chain management, 3rd ed. (pp. 231-58.). prentice-hall, harlow. christopher, m., peck, h. (2004). building the resilient supply chain.  the international journal of logistics management,  15(2), 1-14. https://doi.org/10.1108/09574090410700275 dhl, http://www.dhl.com/content/dam/downloads/g0/about_us/logistics_insights/dhl_logistics_trend_radar_2016.pdf harris, i., wang, y., wang, h. (2015) ict in multimodal transport and technological trends: unleashing potential for the future. international journal of production economics, 159, 88-103. https://doi.org/10.1016/j.ijpe.2014.09.005 holling, c.s. (1973). resilience and stability of ecological systems. annual review of ecology and systematics, 4, 1-23. https://doi.org/10.1146/annurev.es.04.110173.000245 kleindorfer, p., saad, g. (2005). managing disruption risks in supply chains. production and operations management, 14(1), 53-68. https://doi.org/10.1111/j.1937-5956.2005.tb00009.x montreuil, b. (2011). toward a physical internet: meeting the global logistics sustainability grand challenge. logistics research, 3(2-3), 71-87. https://doi.org/10.1007/s12159-011-0045-x morales allende, m., ruiz-martin, c., lopez-paredes, a., perez ríos, j.m. (2017). aligning organizational pathologies and organizational resilience indicators. international journal of production management and engineering, 5(2), 107-116. https://doi.org/10.4995/ijpme.2017.7423 ponis, s.t., koronis, e. (2012). supply chain resilience: definition of concept and its formative elements.  journal of applied business research, 28(5), 921. https://doi.org/10.19030/jabr.v28i5.7234 ponomarov, s.y., holcomb, m.c. (2009). understanding the concept of supply chain resilience.  the international journal of logistics management, 20(1), 124-143. https://doi.org/10.1108/09574090910954873 ruiz-martin, c., lópez-paredes, a., wainer, g. (2018). what we know and do not know about organizational resilience. international journal of production management and engineering, 6(1), 11-28. https://doi.org/10.1108/09574090910954873 sanchis, r., poler, r. (2014). enterprise resilience assessment: a categorisation framework of disruptions. dirección y organización, 54, 45-53. http://hdl.handle.net/10251/58456 sanchis, r., poler, r. (2013). definition of a framework to support strategic decisions to improve enterprise resilience. in ifac proceedings volumes (ifac-papersonline), 700–705. https://doi.org/10.3182/20130619-3-ru-3018.00600 svensson, g. (2000). conceptual framework for the analysis of vulnerability in supply chains. international journal of physical distribution & logistics management, 30(9), 731-749. https://doi.org/10.1108/09600030010351444 int. j. prod. manag. eng. (2018) 6(2), 79-89 creative commons attribution-noncommercial-noderivatives 4.0 international schoen, q., sanchis, r., poler, r., lauras, m., fontanili, f. and truptil, s. 88 https://doi.org/10.1109/ieem.2008.4738196 http://www.intechopen.com/articles/show/title/supply-chain-resilience-using-the-mapping-approach http://www.intechopen.com/articles/show/title/supply-chain-resilience-using-the-mapping-approach https://doi.org/10.1007/978-3-642-40543-3_71 https://doi.org/10.1016/j.ijpe.2004.06.032 https://doi.org/10.1111/jscm.12050 https://doi.org/10.1108/09574090410700275 http://www.dhl.com/content/dam/downloads/g0/about_us/logistics_insights/dhl_logistics_trend_radar_2016.pdf https://doi.org/10.1016/j.ijpe.2014.09.005 https://doi.org/10.1146/annurev.es.04.110173.000245 https://doi.org/10.1111/j.1937-5956.2005.tb00009.x https://doi.org/10.1007/s12159-011-0045-x https://doi.org/10.4995/ijpme.2017.7423 https://doi.org/10.19030/jabr.v28i5.7234 https://doi.org/10.1108/09574090910954873 https://doi.org/10.1108/09574090910954873 http://hdl.handle.net/10251/58456 https://doi.org/10.3182/20130619-3-ru-3018.00600 https://doi.org/10.1108/09600030010351444 http://creativecommons.org/licenses/by-nc-nd/4.0/ swafford, p.m., ghosh, s., murthy, n. (2008). achieving supply chain agility through it integration and flexibility. international journal of production economics, 116(2), 288-297. https://doi.org/10.1016/j.ijpe.2008.09.002 toth, p., vigo, d. (eds.). (2002). the vehicle routing problem. society for industrial and applied mathematics. https://doi.org/10.1137/1.9780898718515 un, (1980). united nations. convention on international multimodal transport of goods. van meldert, b., de boeck, l. (2016). introducing autonomous vehicles in logistics: a review from a broad perspective. faculty of economics and business, leuven, belgium. wamba, s. f., akter, s., edwards, a., chopin, g., gnanzou, d. (2015). how ‘big data’can make big impact: findings from a systematic review and a longitudinal case study. international journal of production economics, 165, 234-246. https://doi.org/10.1016/j.ijpe.2014.12.031 wieland, a., marcus-wallenburg, c. (2013). the influence of relational competencies on supply chain resilience: a relational view. international journal of physical distribution & logistics management, 43(4), 300-320. https://doi.org/10.1108/ijpdlm-08-2012-0243 int. j. prod. manag. eng. (2018) 6(2), 79-89creative commons attribution-noncommercial-noderivatives 4.0 international categorisation of the main disruptive events in the sensitive products transportation supply chains 89 https://doi.org/10.1016/j.ijpe.2008.09.002 https://doi.org/10.1137/1.9780898718515 https://doi.org/10.1016/j.ijpe.2014.12.031 https://doi.org/10.1108/ijpdlm-08-2012-0243 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.12961 received: 2020-01-09 accepted: 2020-05-27 total quality management, employee commitment and competitive advantage in nigerian tertiary institutions. a study of the university of lagos abimbola, b.o.a1, oyatoye, e.o.a2, oyenuga, o.g.a3 adepartment of business administration,faculty of management sciences, university of lagos, nigeria. a1 oabimbola@unilag.edu.ng, a2 eoyatoye@unilag.edu.ng, a3 goyenuga@unilag.edu.ng abstract: the quality of education offered by the tertiary institutions in nigeria in recent times has critically degenerated. unfortunately, this has led to the unprecedented rise in the poor quality of nigerian graduates which contributes to the very alarming high rate of unemployment in the country. this article therefore examines the relationship between total quality management (tqm), employee commitment and competitive advantage among tertiary institutions in lagos state with a particular focus on the university of lagos, using a survey research design with a random selection of 350 respondents from a population of 2047 staff strength across the 11 faculties in the university. the data used for this article were derived from a systematic review of the relevant literature and a structured questionnaire. the data obtained from the questionnaire was subjected to descriptive and inferential statistics. the result showed that the adoption of total quality management practices and employee commitment significantly affects competitive advantage. the findings suggest that proper adoption of tqm practices coupled with employee commitment will guarantee competitive advantage outcomes measured in terms of increase in revenue, customer satisfaction and employee satisfaction. tqm and employee commitment through its principles can successfully contribute to enhanced organizational performance and dynamic competitiveness. key words: tqm, employee commitment, competitive advantage, organizational performance, tertiary institutions. 1. introduction over the years, it has become very essential for organizations all over the world to continuously monitor its business environment due to its dynamic nature and then strive to adapt to the changes. to cope with these global changes, tertiary institutions seek for different methods to outshine the local institutions and compete with the international institutions and one of such ways is integrating quality into their educational system. to this respect, attention has shifted to the concept of total quality management (tqm). tqm is a philosophy that talks about the continuous effort of all the employees in an organization to ensure thorough quality products and services and to guarantee customer satisfaction. the success of every organization greatly lies on the commitment of its employees. for an organization to rise to the peak, the employees need to be wholly committed to the achievement of the objectives. employee commitment is the devotion that an employee has to its organization. ogini (2020) defined employee commitment as the psychological attachment employees shows towards their workplace. human resources are the greatest asset in an organization and can be the source of competitive advantage in an organization. employee commitment is a key determinant to an organization’s performance because high levels of employee commitment to organizational goal will achieve a higher to cite this article: abimbola, b.o., oyatoye, e.o., oyenuga, o.g. (2020). total quality management, employee commitment and competitive advantage in nigerian tertiary institutions. a study of the university of lagos. international journal of production management and engineering, 8(2), 87-98. https://doi.org/10.4995/ijpme.2020.12961 int. j. prod. manag. eng. (2020) 8(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international 87 http://creativecommons.org/licenses/by-nc-nd/4.0/ organizational performance. the objective of every organization is to develop a sustainable competitive advantage that will promote continuous growth and expansion. more so, the increase in globalization has equally increased the level of competition between organizations and this has led to the formulation of competitive strategies by organizations to distinguish itself from others. competitive advantage is the superiority that an organization has over other similar organizations. quality can be a major source of competitive advantage (economou & chatzikonstantinou, 2009) but the achievement of this objective greatly lies upon the human resources of the organization. the employees are majorly responsible for implementing total quality management practices which can guarantee competitive advantage. hence, this is the reason why the adoption of tqm practices with the help of committed employees in its implementation seems an effective approach for an organization to achieve the objective of prevailing over its competitors. 1.1. statement of problem from earlier reviewed literature, the quality of higher education in the nigerian tertiary institutions in the seventies could be compared with the type of education offered by top world higher institutions. however, in recent times, the quality of education offered by nigerian tertiary institutions has critically degenerated (moja, 2000). hence, there is an unprecedented rise in the poor quality of nigerian graduates which results in the very alarming high rate of unemployment in the country. to make things worse, most nigerian students in pursuit of a degree would not give a second thought to choosing home based tertiary institutions when provided with the option of going to tertiary institutions overseas. this is due to the decline in the quality of education provided by the home based institutions. students, parents and sponsors in recent times have also become more conscious of their right to demand good value for the money and time spent in getting quality education thus, the need for tertiary institutions to provide quality education. in addition, organizations such as the world bank, united nations and unesco emphasize on the important role of education in the world’s economic development, the development of “a knowledge based society” (ganguly, 2015). quality education contributes to overall development in the economy and tertiary institutions all over the world strive to be the best facilitator of quality education. however, with the intensity of global competition and as tertiary institutions fightfor competitive advantage, the question of what strategy a tertiary institution can use to outperform its competitors both locally and internationally is asked. quality has always been one of the key strategic choices available and the adoption of tqm practices seems one of the ways out to achieve sustainable competitive advantage. however, it is worthy of note that the implementation of tqm practices in an organization is in the hands of its employees. therefore, an organization that can boast of committed employees has a greater tendency of higher productivity which can lead to competitive advantage. 1.2. the objectives of the study 1. examine the effect of the adoption of total quality management practices on employee commitment 2. examine the significant effect of employee commitment on competitive advantage 3. investigate how the adoption of total quality management practices affect competitive advantage 4. determine the effect of the adoption of total quality management practices and employee commitment on competitive advantage 1.3. research questions and hypotheses of the study 1. how does the adoption of total quality management practices significantly affect employee commitment? 2. what is the significant effect of employee commitment on competitive advantage? 3. how does the adoption of total quality management practices significantly affect competitive advantage? 4. how does the adoption of total quality management practices and employee commitment significantly affect competitive advantage? 1.4. research hypotheses h0: adoption of total quality management practices does not significantly affect employee commitment. h0: there is no significant effect of employee commitment on competitive advantage. int. j. prod. manag. eng. (2020) 8(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international abimbola et al. 88 http://creativecommons.org/licenses/by-nc-nd/4.0/ h0: there is no significant effect of the adoption of total quality management practices on competitive advantage. h0: adoption of total quality management practices and employee commitment does not have any significant effect on competitive advantage. 2. literature review 2.1. european foundation for quality management (efqm) model different models have been developed such as the european foundation for quality management (efqm model) to steer organisations towards total quality management (tqm). the total quality management (tqm) concept is based on the efqm model. the model is based on the assumption that organizations should channel all its activities to sustain continuous improvement which is the concept of tqm. the new efqm model consists of seven criteria (as shown in figure 1) namely: purpose, vision and strategy,organisational culture and leadership,engaging stakeholders, creating sustainable value, driving performance and transformation, stakeholder perception, strategic and operational performance (efqm, 2020). 2.2. total quality management tqm has been broadly defined by so many researchers. it was simply defined by farooq et al. (2007) as an art of organizing the whole to achieve excellence. addae-korankye (2013) in his own view added that “tqm is a quest for excellence that ensures the right attitudes and controls are put in place to prevent defects and to achieve customer satisfaction through intense effectiveness and efficiency”. according to ngambi & nkemkiafu (2015), tqm is a “concept based on continuous improvement in the performance of processes in an organization and in the quality of the products and services that are the outputs of those processes”. from these definitions, tqm can be deduced as a theory that ensures customer satisfaction by providing quality through the continual efforts and involvement of everyone in the organization. there are principles that revolve around tqm as reviewed in different literature. based on this, some of these principles will be discussed below to give a general background on the nature of total quality management (figure 2). customer focus emphasizes that organizations should provide services that satisfies their customers. hence, who are the customers in higher education institutions? harvard university defined its customers as “as one to whom we provide information or service”. customers here are of two types, the internal and the external. the external customers are the students, parents, sponsors, employers and society at large while the workers in the organization (ali & shastri, 2010). although in tqm, the emphasis is on the external customers, however the internal customers should equally be made happy (najafabadi et al., 2008). ray et al. (2016) affirms this by saying that the requirements of the internal customer should firstly be satisfied. this will motivate them to provide fulfilling services to the external customers as satisfied internal customers equals to satisfied external customers. continuous improvement focuses on improving all the factors relating to the production of goods and services on an ongoing basis (stevenson, 2012). to meet up with the demands of the customers and to survive in an ever growing market, an organization must continuously improve its processes and methodologies to achieve a greater quality of its products. ray et al (2016) notes that continuous improvement is necessary in order to be above the challenges that might be brought up by the competition and to always satisfy the customer. team approach stresses the use of teams to solve problems, creating synergy and promoting the spirit of cooperation among employees to achieve the goal of an organization. the japanese understood this philosophy well and excelled as a result of making use of it. according to adediran & adediran (2008), an effective and efficient creation of goods and services in an organization depends extensively on a well-organized and arranged team. this goes on to say that without effective teamwork, achieving quality in an organization cannot be possible. top management commitment is supreme in the adoption and implementation of total quality management (alamutu et al., 2012) since they are entrusted with the role of coaching the other employees on the importance of tqm. according to mehra et al. (2001), top management provides leadership needs and gives their approval for training demands and most organizational changes. employee involvement entails empowering employees to make decisions through responsibilities int. j. prod. manag. eng. (2020) 8(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international total quality management, employee commitment and competitive advantage in nigerian tertiary institutions. a study of the university of lagos 89 http://creativecommons.org/licenses/by-nc-nd/4.0/ that will help in solving problems that will help organizations achieve their goals. in order to achieve proper implementation of tqm, every employee in the organization must be actively involved. employees are motivated when they are given the authority and responsibility for improvements. this allows them to make decisions and proffer necessary solutions to problems (stevenson, 2012). training prepares employees with the needed skills and techniques which will enable them improve quality in the production process. to achieve a successful tqm implementation, constant training programs should be put in place to continuously improve quality. with proper and effective training, employees will have adequate knowledge of the industry and structure of the organization. this will in turn increase their motivation to work, overall performance and loyalty to the organization. in addition, they will be able to provide quality services which will increase customer satisfaction and reduce complaints (sadikoglu & olcay, 2014). competitive benchmarking entails finding out organizations that are the best in their fields and carefully observing and learning their methods in figure 1. the structure of the new efqm model. source: adapted according to efqm (2020). retrieved from www.efqm.org. figure 2. total quality management principles. source: adapted from talib & rahman (2012). int. j. prod. manag. eng. (2020) 8(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international abimbola et al. 90 http://www.efqm.org http://creativecommons.org/licenses/by-nc-nd/4.0/ order to improve its mode of operations (stevenson, 2012). here, the organization measures its performance against its competitors by concentrating on some selected industry measures. according to camp (1993), benchmarking is a form of selfappraisal and self-advancement that involves an organization comparing its performance with its competitors for the purpose of identifying its strengths and weaknesses. 2.3. employee commitment employee commitment according to ongori (2007) is the degree of affection or devotion an employee feels towards the organization. to zheng (2010), it simply means the attitude of an employee to his organization. employee commitment means when employees devote their energy into executing their duties, which stems from a sense of obligation to the organization. employee commitment has been believed to predict the performance of workers which in turn determine the success of an organization (zheng, 2010). research has found out that several factors induces employee commitment. some of these factors can be compensation, job enrichment, employee empowerment (ongori, 2007), continuous training and development, job security to mention a few. in some studies, employee commitment is used synonymously with organizational commitment. committed employees are more likely to remain in an organization, hence employee commitment is of great importance to an organization because it will reduce absenteeism and turnover as well as improve organizational performance as a whole (dixit & bhati, 2012). 2.4. concept of competitive advantage competitive advantage is the edge that an organization has over its competitors when it develops or acquires attributes that makes it to outperform them (wang, 2014). an organization can boast of competitive advantage when it is able to distinguish itself from its competitors through its unique capabilities. an organization has to develop strategies that will give it an edge over its competitors in order to have a competitive advantage. in developing a competitive advantage, an organization must be able to identify its strengths and weaknesses by understanding the operating forces in the business environment as identified by michael porter. these forces are the threat of new entry, bargaining power of suppliers, bargaining power of buyers, threats of substitute products and competition in the industry (porter, 2008). according to porter, entry of new firms into the market with similar products poses a threat to the existing firms. the ability of other firms in the industry to replicate or substitute the products or services of a firm is also a threat. furthermore, the more suppliers a firm has in the industry, the more power it has over them. however, if the suppliers are few, the firm is at a disadvantage because they tend to determine the prices. in addition, the more the customers of a firm, the less power they have to determine prices, hence beneficial for the firm. the other approach to competitive advantage is the resource-based view. these resources are the physical and non-physical assets which an organization possesses. resources are very crucial and dependent on the smooth operations of an organization. however, what is imperative is how these resources are being used effectively and efficiently. if the resources of an organization are managed properly, it can lead to competitive advantage. competitive advantage according to al-rfou (2012) is the ability of an organization to outperform its competitors by effectively producing goods or services more than they do. for an organization to attain competitive advantage, one or a mix of the three approaches can be used: differentiation, cost leadership, or focus (porter, 1980). daft (1991) as cited in addaekorankye (2013) opines that it is very possible that organizations that have adopted total quality management use the above mentioned strategies. when the quality of products and services are improved, manufacturing costs are reduced. this is cost leadership and it leads to higher profits and returns. in addition, the firm will be able to command premium prices as a result of the uniqueness of their product (differentiation). 2.5. tqm, employee commitment and competitive advantage the level of competition among firms has been increased due to globalization and transnationalism of firms and one out of the many strategies for achieving competitive advantage in organizations around the worlds is providing a higher quality service around the world. customers demand for quality products and this has inspired many organizations to produce quality products to gain ground in the market place (kruger, 2001). customers no longer consider price as a motivator in buying products or services. presently quality int. j. prod. manag. eng. (2020) 8(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international total quality management, employee commitment and competitive advantage in nigerian tertiary institutions. a study of the university of lagos 91 http://creativecommons.org/licenses/by-nc-nd/4.0/ has become one of the yardsticks used to measure organizations and products compete in terms of their quality in the market place. a study conducted by lamptey (2009) showed that so many ghanaian organisations are wearing out and below standard due to their retrogression in quality management. this is because majority of these organizations most especially the service ones do not treat their customers right forgetting that in this era of globalization, customers have the power to make or mar an organization. in addition, getting quality employees is very imperative but the huge cost involved in the training and development has made majority of these organisations neglect it. this however can lead to the provision of inferior services, therefore, to attain competitive advantage in organisations, quality should take precedence (addae-korankye, 2013). in addition, having quality and committed employees in an organization is beneficial to the organizational performance. this is because committed employees are always excited to identify with the organization, believe in its goals and puts its interest at heart (george & jones, 1996). total quality management approaches are being embraced by organizations to improve quality, increase productivity, increase customer satisfaction and improve organizational performance. this has made many organizations to adopt and implement tqm strategies to counter the current global competition. total quality management is one of the strategies used in maximizing the competitive nature of firms by focusing on employee involvement, and continuous improvement in the quality of products and services and customer satisfaction (gaspersz, 2005). there have been a large number of studies that have acknowledged the fact that total quality management has a positive impact on an organization’s financial and overall performance. hence, it is important to appreciate the vital role tqm plays in giving an organization a competitive edge (kruger, 2001). 3. methodology 3.1. research design this study was conducted using the survey research design. the academic and non-academic staff across the faculties in the university of lagos constituted the population for this research. the university is composed of 11 faculties and they are art, basic medical sciences, clinical sciences, dental sciences, education, engineering, environmental sciences, law, management sciences, sciences, social sciences. the number of employees in each faculty was obtained from the human resource department of the university and as at the time of carrying this study, there was one thousand, four hundred and ninety-two (1492) academic and five hundred and fifty-five (555) non-academic staff across the faculties in the university. 3.2. sampling technique and instrument a sample size of 335 using the taro yemane (1967) formula was drawn out of the population. the study made use of multistage sampling technique. stratified technique was used to ensure that every faculty in the university is represented in the sample chosen in proportion to their population. hence for each of the faculty, simple random sampling was applied on proportionate basis to determine the required sample size in each faculty and the sample size was calculated using the stratified method thus: the population of employees in each faculty was obtained and respondents were selected from each faculty based on the population of each faculty in relation to the determined sample size. the questions for the study were adapted by making use of the variables total quality management practices, employee commitment and competitive advantage. the instrument was pilot tested with 40 respondents and a cronbach’s alpha coefficient of 0.879 was gotten. three hundred and thirty five questionnaires were administered but only 290 was returned which was approximately 87%. 3.3. data collection method in collecting the data for this study being quantitative, a close ended questionnaire was used for the study. this choice is influenced by the fact that it enables the researcher to cover a large sample with dependable and reliable result (kothari, 2004), guarantees higher response rate and because it is self-administered. self-administered questionnaires are normally completed by the respondents (saunders et al., 2011). other reasons why closed ended questionnaires was chosen is due to the constraints of time, rigors of analysis and high cost of administration posed by other instruments. int. j. prod. manag. eng. (2020) 8(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international abimbola et al. 92 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. results and discussion of findings 4.1. socio-demographic characteristics of the respondents socio-demographic characteristics frequency (n=290) percent gender male 165 56.9 female 125 43.1 total 290 100 age (years) less than 20 years 0 0 21-30 years 98 33.8 31-40 years 101 34.8 41-60 years 84 29.0 60 years and above 7 2.4 total 290 100 marital status single 105 36.2 married 145 50.0 divorced/separated 33 11.4 widow/widower 7 2.4 total 290 100 highest educational qualification nce, nd or equivalent 27 9.3 hnd, b.sc. or equivalent 96 33.1 m.sc./mba 129 44.5 phd 38 13.1 total 290 100 faculty art 36 12.4 basic medical sciences 11 3.8 clinical sciences 26 9.0 dental sciences 10 3.4 education 33 11.4 engineering 36 12.4 environmental sciences 22 7.6 law 13 4.5 management sciences 30 10.3 sciences 46 15.9 social sciences 27 9.3 total 290 100 staff designation academic 158 54.5 non-academic 132 45.5 total 290 100 employee cadre senior staff 220 75.9 junior level staff 70 24.1 total 290 100 length of service below 5 years 70 24.1 5-10 years 104 35.9 11-15 years 64 22.1 16 years & above 52 17.9 total 290 100 source: survey, 2018. 4.2. test of research hypotheses the hypotheses stated in this study were tested against the statistical table value to form the basis of accepting or rejecting the null hypotheses using simple and multiple regression analysis at 0.05 level of significance so as to determine whether the outcome of the study will lead to the rejection of the null hypothesis. the simple regression was used for hypotheses 1 to 3 while the multiple regression was used for hypothesis 4. 4.2.1. hypothesis 1 h0: adoption of total quality management practices does not significantly affect employee commitment. model summarya model r r square adjusted r square std. error of the estimate durbinwatson 1 0.594b 0.353 0.348 4.76960 1.692 a dependent variable: employee_commitment. b predictors: (constant), total_quality_management_practices. anovaa sum of squares df mean square f sig. regression residual total 659.055 7843.290 8502.345 1 288 289 659.055 27.234 24.200 0.000b a dependent variable: employee_commitment. b predictors: (constant), total_quality_management_practices. coefficientsa model unstandardized coefficients standardized coefficients t sigb std. error beta (constant) 33.434 2.288 0.278 14.614 0.000 tqmp 0.327 0.006 4.919 0.000 a dependent variable: employee_committment. tqmp: total_quality_management_practices. in order to test the above stated null hypothesis, simple regression coefficient was used. the r2 value of 7.8% explains the proportion of the total variation in employee commitment that is attributed to variation in total quality management. the adjusted r2 value of 7.4% shows the actual variation in the university’s employee commitment as a result of changes in its total quality management. from the anova table, since the p-value is 0.000 which is less than 0.05, this implies that the test is statistically significant. hence, the null hypothesis is rejected and it can therefore be concluded that the adoption int. j. prod. manag. eng. (2020) 8(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international total quality management, employee commitment and competitive advantage in nigerian tertiary institutions. a study of the university of lagos 93 http://creativecommons.org/licenses/by-nc-nd/4.0/ of total quality management practices significantly affects employee commitment. the coefficient table also shows that where total quality management practices do not exist, employee commitment is 33.434. also, for every unit increase in total quality management practices, employee commitment will increase marginally by 0.327. 4.2.2. hypothesis 2 h0: there is no significant effect of employee commitment on competitive advantage. model summarya model r r square adjusted r square std. error of the estimate durbinwatson 1 0.503b 0.253 0.250 5.11591 1.600 a dependent variable: competitive advantage. b predictors: (constant), employee commitment. anovaa model sum of squares df mean square f sig. regression 2551.918 1 2551.918 97.504 0.000b residual 7537.686 288 26.173 total 10089.603 289 a dependent variable: competitive advantage. b predictors: (constant), employee commitment. coefficientsa model unstandardized coefficients standardized coefficients t sig.b std. error beta (constant) 12.832 2.492 5.149 0.000 employee commitment 0.548 0.055 0.503 9.874 0.000 a. dependent variable: competitive advantage. the coefficient of determination r2 value of 25.3% explains the proportion of the total variation in competitive advantage that is attributed to variation in employee commitment. the adjusted r2 value of 25% shows the actual variation in competitive advantage as a result of changes in employee commitment. from the anova table, since the p-value is 0.000 which is less than 0.05, this implies that the test is statistically significant. hence, the null hypothesis is rejected and it can therefore be concluded that there is significant effect of employee commitment on competitive advantage. the coefficient table also shows that where employee commitment does not exist, employee commitment is 12.832. also, for every unit increase in employee commitment, competitive advantage will increase marginally by 0.548. 4.2.3. hypothesis 3 h0: there is no significant effect of the adoption of total quality management practices on competitive advantage. model summarya model r r square adjusted r square std. error of the estimate durbinwatson 1 0.444b 0.197 0.194 5.30436 1.546 a dependent variable: competitive_advantage. b predictors: (constant), total_quality_management_practices. anovaa model sum of squares df mean square f sig. regression 1986.376 1 1986.376 70.599 0.000b residual 8103.227 288 28.136 total 10089.603 289 a dependent variable: competitive_advantage. b predictors: (constant), total_quality_management_practices. coefficientsa model unstandardized coefficients standardized coefficients t sig.b std. error beta (constant) 17.897 2.325 7.696 0.000 total_quality_ management_ practices 0.567 0.068 0.444 8.402 0.000 a dependent variable: competitive_advantage. the coefficient of determination r2 value of 19.7% explains the proportion of the total variation in competitive advantage that is attributed to variation in total quality management. the adjusted r2 value of 19.4% shows the actual variation in competitive advantage as a result of changes in total quality management. from the anova table, since the p-value is 0.000 which is less than 0.05, this implies that the test is statistically significant. hence, the null hypothesis is rejected and it can therefore be concluded that there is significant effect of the adoption of total quality management practices on competitive advantage. the coefficient table also shows that where total quality management practices does not exist, competitive advantage is 17.897. also, for every unit increase in total quality management practices, competitive advantage will increase marginally by 0.567. int. j. prod. manag. eng. (2020) 8(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international abimbola et al. 94 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.2.4. hypothesis 4 h0: adoption of total quality management practices and employee commitment does not have any significant effect on competitive advantage. model summarya model r r square adjusted r square std. error of the estimate durbinwatson 1 0.594b 0.353 0.348 4.76960 1.692 a dependent variable: competitive_advantage. b predictors: (constant), employee_commitment, total_quality_ management_practices. anovaa model sum of squares df mean square f sig. regression 3560.617 2 1780.308 78.258 0.000b residual 6528.987 287 22.749 total 10089.603 289 a dependent variable: competitive_advantage. b predictors: (constant), employee_commitment, total_quality_ management_practices. coefficientsa model us. coefficients s. coeff. t sig collinearity statistics b std. error beta tolerance vif (constant) 2.918 2.759 1.058 0.291 tqmp 0.421 0.063 0.329 6.659 0.000 0.922 1.084 ec 0.448 0.054 0.411 8.319 0.000 0.922 1.084 dependent variable: employee_committment. us: unstandarized; s: standarized. tqmp: total_quality_management_practices. ec: employee_commitment. the r2 value of 35.3% explains the proportion of the total variation in competitive advantage that is attributed to variation in total quality management and employee commitment. the adjusted r2 value of 34.8% shows the actual variation in competitive advantage as a result of interactions of total quality management and employee commitment. from the anova table, since the p-value is 0.000 which is less than 0.05, this implies that the test is statistically significant. hence, the null hypothesis is rejected and it can therefore be concluded that the adoption of total quality management practices and employee commitment does have significant effect on competitive advantage. the coefficient table also shows that where total quality management practices and employee commitment does not exist, competitive advantage is 2.918. also, for every unit increase in total quality management practices, competitive advantage will increase marginally by 0.421. for every unit increase in total quality management practices, competitive advantage will increase marginally by 0.448. 4.3. sobel test the aim of the sobel (1982) test in this study was to thoroughly confirm the significance and effect of the mediating variable on the independent and dependent variable. the sobel (1982) test for this analysis was calculated using the sobel calculator by preacher and leonardelli (2018) and the result showed the p-value to be 0.00000915 which is less than 0.05. this implies that the test is significant. 4.4. discussion of findings the result of the analysis showed that the adoption of total quality management practices in an institution significantly affects employee commitment, as the result shows the acceptance of the alternative hypothesis as against the null given the p-value of 0.00 which is less than the significant value of 0.05. this finding is in consistent with the findings of several authors, one of which is allen and brady (1997) who suggested that the commencement of total quality management programs in organization has a positive influence on the employee’s commitment. karia and asaari (2006) also revealed that total quality management practices such as employee empowerment, continuous improvement and so on significantly affect organizational commitment. the result of the analysis also revealed that there is a significant effect of employee commitment on the competitive advantage of an institution, as the p-value of 0.00 is less than the significant value of 0.05 thereby resulting in the rejection of the null hypothesis and the acceptance of the alternative hypothesis. this finding is in line with the findings of mathur (2015) in his study on achieving competitive advantage through employees in which he outlined that employees are a valuable source of sustainable competitive advantage and that the success of an organization depends on the employee commitment towards the organization. furthermore, the analysis showed that there is significant effect of the adoption of total quality management practices on competitive advantage, as the p-value of 0.00 is less than the significant value of 0.05 thereby resulting in the rejection of the null hypothesis. this finding is in consonance with the findings of addae-korankye (2013) who found out that the proper implementation of total quality management can be a source of competitive advantage. this also goes in line with int. j. prod. manag. eng. (2020) 8(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international total quality management, employee commitment and competitive advantage in nigerian tertiary institutions. a study of the university of lagos 95 http://creativecommons.org/licenses/by-nc-nd/4.0/ the study of ganapavarapu and prathigadapa (2015) where it was revealed that total quality management activities cause a significant impact on competitive advantage of organizations in global businesses. finally, the result of the analysis also showed that total quality management adoption and employee commitment significantly affects the competitive advantage as the p-value of 0.00 is higher than the significant value of 0.05 thereby resulting in the acceptance of the null hypothesis and the rejection of the alternative hypothesis. 5. conclusion and recommendations the following conclusions have been drawn based on the findings of this study and the test of hypotheses. a. the adoption of total quality management practices in an institution has a tremendous impact on employee commitment. this means that institutions that embrace tqm practices will have employees that are highly committed to the organization. b. also, it is concluded that employee commitment is an important determinant on the competitive advantage of an institution. this goes on to say that committed employees will be more than willing to increase productivity which will in turn lead to competitive advantage. c. furthermore, it was established that the adoption of total quality management practices is another important predictor of the competitive advantage of an institution. this means that proper adoption of tqm practices will enable organizations operate at a more competitive level by increasing quality, minimizing cost and satisfying customers. d. finally, it was concluded that the adoption of total quality management practices and the employee commitment has an influence on the competitive advantage of an institution. the combination of proper adoption and implementation of tqm practices and employee commitment in an organization can strengthen an organization’s competitiveness. 5.1. recommendations the following recommendations have been proffered for better decision making and consideration for the purpose of improving better performance from the conclusions of the study. a. this study has established that employee commitment is an important determinant on the competitive advantage of an institution; therefore, the management should introduce total quality programs with dimensions such as employee empowerment, training and continuous improvement into the system. this will enable employees to have adequate knowledge of their duties and responsibilities which will in turn increase their motivation to work, and to be actively involved towards achieving organizational goals leading to overall performance and competitive advantage. b. furthermore, from the conclusion that the adoption of total quality management practices in an institution tremendously impact employee commitment, tqm dimensions such as top management commitment and employee involvement should be put in place. employees will be motivated to work by knowing that the organization is concerned about them. c. the management should create an inclusive communication environment and educate the employees on working as a team to make them aware that the success of an organization depends on them working as a whole and not in isolation. d. also, the government, the national university commission (nuc) as well as the governing councils in each institution should put in place policies that will allow institutions focus more on tqm measures and most importantly, provide funding for tqm based quality improvement programs. e. the governing councils in each institution should also endeavor to choose the right leaders in top management positions as it is only the right leaders that can generate as well as trigger commitment from employees. f. in addition, the top management leaders should have the understanding that the pursuit for quality efforts through the adoption of total quality management practices do result in a great level of employee commitment which correlates to competitive advantage. g. furthermore, there should be a department in charge of tqm and if possible sub-units in every faculty in order to cultivate a tqm culture and to make all the employees quality conscious and customer focused. int. j. prod. manag. eng. (2020) 8(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international abimbola et al. 96 http://creativecommons.org/licenses/by-nc-nd/4.0/ h. employees should regularly interact with their customers (students) giving them advice and always willing to assist them. i. also, avenues to receive feedback and complaints from customers should be put in place. this will enhance the performance of employees and make them more competitive. j. benchmarking measures should also be put in place in order to compare performance with the best in the field with the aim of improving mode of operations. references addae-korankye, a. (2013). total quality management (tqm): a source of competitive advantage. a comparative study of manufacturing and service firms in ghana. international journal of asian social science, 3(6), 1293-1305. adediran, o., adediran, o. (2008). total quality management: a test of the effect of tqm on performance and stakeholder satisfaction (master’s thesis). school of management, blekinge institute of technology. alamutu, s.a, hotepo, o.m, oyeobu, a.j., nwatulegwu, b.i. (2012). an evaluation of total quality management practices on business performance of the nigerian telecommunications sector: a case study of mtn nigeria limited. proc. iccem, 155-166 allen, m., brady, r. (1997). total quality management, organizational commitment, perceived organizational support, and intraorganizational communication, management communication quarterly, 10(3), 316-341. https://doi.org/10.1177/0893318997010003003 ali, m., shastri, r.k (2010). implementation of total quality management in higher education. asian journal of business management, 2(1), 9-16. al-rfou, a.n. (2012). achieving competitive advantage through enterprise resource planning system (erp). empirical evidence from jordan. international journal of asian social sciences, 2(6), 805-857. blackburn, r., rosen, b. (1993). total quality and human resources management: lessons learned from baldrige award-winning companies. academy of management perspectives, 7(3), 49-66. https://doi.org/10.5465/ame.1993.9411302347 camp, r.c. (1993). a bible for benchmarking, by xerox, financial executive, 9(4),23-7. daft, r. (1991). management. chicago: dryden press. dale, b.g., boaden, r.j., lascelles d.m. (1994). ‘total quality management an overview, managing quality (edited by dale, b. g.), new york: prentice hall. dixit, v., bhati, m. (2012). a study about employee commitment and its impact on sustained productivity in indian auto-component industry. european journal of business and social sciences, 1(6), 34–51. economou, v.p., chatzikonstantinou, p.g. (2009). gaining company’s sustained competitive advantage, is really a necessary precondition for improved organizational performance? the case of tqm. european research studies, 12(3), 83-100. https://doi.org/10.35808/ ersj/233 efqm (2020). the efqm model. retrieved from http://www.efqm.org/ [accessed 25 may 2020]. farooq, m.s., akhtar, m.s, ullah, s.z., memon, r.a. (2007). application of total quality management in education. journal of quality and technology management: iii, (i1), 87-97. ganapavarapu l.k., prathigadapa, s. (2015). study on total quality management for competitive advantage in international business. arabian journal business management review, 5(3). ganguly, a. (2015). exploring total quality management (tqm) approaches in higher education institutions in a globalized environment-case analysis of uk and sweden. british journal of education, 3(7), 83-106. gasperz, v. (2005). total quality control. catakan keempat, jakarta: pt. gramediapustoka utama. george, j.m., jones, g.r. (1996). understanding and managing organizational behavior. california: addison-wesley publishing company, inc. karia, n., asaari, m. (2006). the effects of total quality management practices on employees’ work-related attitudes, the tqm magazine, 18(1), 30-43. https://doi.org/10.1108/09544780610637677 kothari, c.i. (2004). research methodology, methods & techniques (2nd ed.). new delhi: new age international publishers limited. kruger, v. (2001). main schools of tqm: the big five. the tqm magazine, 13(3), 146-155. https://doi.org/10.1108/09544780110366042 lamptey, s.o. (2009). quality management systems of unilever ghana limited (master’s thesis). st. clement university, turks and caicos island. mathur, p. (2015). achieving competitive advantage through employees. international journal of arts, humanities and management sciences, 1(9), 66-71. mehra, s., hoffman, j.m., sirias, d. (2001). tqm as a management strategy for the next millennia. international journal of operations & production management, 21(5/6), 855-876. int. j. prod. manag. eng. (2020) 8(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international total quality management, employee commitment and competitive advantage in nigerian tertiary institutions. a study of the university of lagos 97 https://doi.org/10.1177/0893318997010003003 https://doi.org/10.5465/ame.1993.9411302347 https://doi.org/10.35808/ersj/233 https://doi.org/10.35808/ersj/233 http://www.efqm.org https://doi.org/10.1108/09544780610637677 https://doi.org/10.1108/09544780110366042 http://creativecommons.org/licenses/by-nc-nd/4.0/ moja, t. (2000). nigeria education sector analysis: an analytical synthesis of performance and main issues. world bank: new york mullins, l. (2009). management and organizational behaviour, 5th edition. portsmouth: pitman publishing. najafabadi, h.n, sadeghi s., habibzadeh, p. (2008). total quality management in higher education case study: quality in practice at university college of borås (master’s thesis). institution for engineering school, university college of borås. ngambi, m.t., nkemkiafu, a.g. (2015). the impact of total quality management on firm’s organizational performance. american journal of management, 15(4), 69. ogini, j.a. (2020). job promotion and employee commitment of public hospitals in rivers state, nigeria. international journal of economics and business management, 6(1), 10. ongori h. (2007). a review of the literature on employee turnover. african journal of business management, 49-54. porter, m. (1980). competitive strategy; free press: techniques for analyzing industries and competitors. new york. porter, m.e. (2008). the five competitive forces that shape strategy. harvard business review, 86(1), 57-71. preacher, k., leonardelli, g. (2018). calculation for the sobel test: an interactive calculation tool for mediation tests. retrieved july 17, 2018, from http://quantpsy.org/sobel/sobel.htm ray, s.s, nanda, s., ota, r. (2016). total quality management: a need for a sustainable competitive advantage. international journal of commerce, business and management (ijcbm), 5(3). sadikoglu, e., olcay, h. (2014). the effects of total quality management practices on performance and the reasons of and the barriers to tqm practices in turkey. hindawi publishing corporation advances in decision sciences, 2014, 537605. https://doi.org/10.1155/2014/537605 saunders, m.n., saunders, m., lewis, p., thornhill, a. (2011). research methods for business students, (5th ed.). india: pearson education. sobel, m.e. (1982). asymptotic confidence intervals for indirect effects in structural equation models. in s. leinhardt (ed.) sociological methodology, 1982 (pp. 290–312). american sociological association: washington, dc. https://doi.org/10.2307/270723 stevenson, w.j. (2012). operations management, (11th ed). new york: mcgraw-hill inc. talib, f., rahman, z. (2012) total quality management practices in manufacturing and service industries: a comparative study. int. j. advanced operations management, 4(3)155-176. https://doi.org/10.1504/ijaom.2012.047634 wang, h. (2014). theories for competitive advantage. in h. hasan (eds.), being practical with theory: a window into business research (pp. 33-43). australia: wollongong. wright, p., mcmaham, g., mcwilliams, a., (1994). human resources and sustained competitive advantage: a resource-based perspective. int. j. of hrm, 5(2), 301-326. https://doi.org/10.1080/09585199400000020 yamane, t. (1967). statistics: an introductory analysis (2nded). new york: harper & row. zheng, w., sharan, k., wei, j. (2010). new development of organizational commitment: a critical review (1960-2009). african journal of business management, 4(1), 12-20. int. j. prod. manag. eng. (2020) 8(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international abimbola et al. 98 https://doi.org/10.1155/2014/537605 https://doi.org/10.2307/270723 https://doi.org/10.1504/ijaom.2012.047634 https://doi.org/10.1080/09585199400000020 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.12944 received: 2020-01-08 accepted: 2020-05-25 production planning in 3d printing factories de antón, j.*, senovilla, j., gonzález, j.m., acebes, f., pajares, j. insisoc research group, university of valladolid, school of engineering industries, paseo del cauce 59, 47011 valladolid, spain. juan.anton@uva.es abstract: production planning in 3d printing factories brings new challenges among which the scheduling of parts to be produced stands out. a main issue is to increase the efficiency of the plant and 3d printers productivity. planning, scheduling, and nesting in 3d printing are recurrent problems in the search for new techniques to promote the development of this technology. in this work, we address the problem for the suppliers that have to schedule their daily production. this problem is part of the lonja3d model, a managed 3d printing market where the parts ordered by the customers are reorganized into new batches so that suppliers can optimize their production capacity. in this paper, we propose a method derived from the design of combinatorial auctions to solve the nesting problem in 3d printing. first, we propose the use of a heuristic to create potential manufacturing batches. then, we compute the expected return for each batch. the selected batch should generate the highest income. several experiments have been tested to validate the process. this method is a first approach to the planning problem in 3d printing and further research is proposed to improve the procedure. key words: additive manufacturing, production planning, packing problem, optimization, nesting. 1. introduction additive manufacturing (am), also known as 3d printing, is a manufacturing technique that allows to produce a diversity of parts from a 3d model. the process involves the successive addition of material layers until the part is completed. in the past few years, 3d printing has evolved considerably in both techniques and materials, compared to the production processes and logistics i.e., it is evolving and is currently undergoing its phase of industrialization. these technologies are called to play a central role in the next generation of production systems. the am process shows benefits unachievable by traditional manufacturing techniques. masscustomized production, prototyping, sustainable production, and minimized lead time and cost are some of the benefits that are attracting attention in the field of manufacturing (mehrpouya et al., 2019). as materials and techniques are constantly evolving, there is a need to develop a helpful framework to take advantage of the possibilities of am. in production terms, this means that printing time and cost must be reduced (gogate & pande, 2008). here is where the scheduling and packing problem takes on importance. given that 3d printing a part consists of a single operation, and that multiple parts can be printed at the same time, planning and scheduling for am brings a unique set of emerging opportunities and challenges while attempting to optimize the process (dvorak et al., 2018). in this context, there is a growing interest to facilitate the production scheduling of am systems. while tackling the optimization of the 3d printing process, we can either consider the manufacturing process or the whole production cycle. the first one refers to the process of setting up the machine for the production of one or more parts, the printing of those parts and the post-processing operations. on the other to cite this article: de anton, j., senovilla, j., gonzález, j.m., acebes, f., pajares, j. (2020). production planning in 3d printing factories. international journal of production management and engineering, 8(2), 75-86. https://doi.org/10.4995/ijpme.2020.12944 int. j. prod. manag. eng. (2020) 8(2), 75-86creative commons attribution-noncommercial-noderivatives 4.0 international 75 http://creativecommons.org/licenses/by-nc-nd/4.0/ hand, production planning includes all activities aimed at satisfying the order of a customer, i.e. receiving the part specifications, scheduling the part production and delivering the final product to the client. there are different approaches to what we have referred to as production planning. in consequence, studies carried out have focused on different parts of the process. in this work we review the scheduling problem for a factory that has to schedule the daily production of several 3d printers, trying to maximize the profit obtained for each one. nowadays, factories rely on their know-how and intuition to set the daily production planning. in the best case, they resort to prioritization techniques backed by qualitative parameters. both methods prove inefficient in time-saving and resource optimization. this situation demonstrates the need for an automated system that supports planning tasks. we propose the use of a heuristic to solve the planning of parts to be produced by a single machine in a daily shift. the paper aims to provide a practical solution to 3d factories to increase their productivity by tackling the production planning. the deliverable is a python program in which the input data of the parts is introduced and a platform layout is returned. to check their performance, the proposed heuristic will be assessed and compared to other existing techniques. the main motivation of this paper is to provide 3d companies with mechanisms that help them increase their productivity. having tools to optimize production capacity per unit of time leads to an increase in expected revenues per unit of time. the article is organized as follows: section 2 reviews the literature about planning, scheduling and nesting problems in am. section 3 describes the problem addressed and the main objectives. section 4 describes the method used to solve the packing problem and the winner determination problem, which is programmed using the software package python. in section 5, some practical examples are presented and the outcomes are analyzed. section 6 summarizes the main conclusions and future research is proposed. 2. literature review production planning in am starts when a customer sends the design specifications of a part that needs to be manufactured (dvorak et al., 2018). at this point, the manufacturer has to manage the printing of the part and the delivery to the customer, usually subjected to a due date. the key activity of this process turns out to be scheduling the parts that need to be printed. the scheduling process involves two fundamental questions: how to group parts for each print and how to place each part in the printing space of each print (i.e. nesting the parts) (wang et al., 2019). specialists in the 3d printing field have addressed several issues related to the planning, nesting and scheduling problem. some have acquired a comprehensive point of view of the production process, including shipping and delivery in their models. others have focused on the scheduling and nesting problem, studying the benefits of maximizing the use of the printing surface and the simultaneous manufacture of parts. in this section, we review the most relevant works that have faced the problems previously described. 2.1. production, scheduling and nesting problems the planning problem for a case in which orders from different distributed customers were satisfied by due dates was studied by chergui et al. (2018). a mathematical formulation of the problem was presented and a heuristic was also proposed to solve it. the heuristic solution proposed was programmed in python and tests carried out highlighted the importance of planning/scheduling for an optimized production with am. the nesting and scheduling problem in am was addressed by dvorak et al. (2018). they investigated a problem in which a set of parts with unique configurations and deadlines must be printed by a set of machines while minimizing time and satisfying both deadlines and constraints. the method proposed consisted of two main steps: first, they modeled the optimization problem of nesting parts into builds, then they scheduled those builds into machines. a similar approach was made by wang et al. (2019) who also resorted to an analogous two-step procedure to solve the problem and presented an improved visionbased placement method. the problem of maximizing the use of the bed (manufacturing surface) appears as one of the main issues in am production. kucukkoc et al. int. j. prod. manag. eng. (2020) 8(2), 75-86 creative commons attribution-noncommercial-noderivatives 4.0 international de antón et al. 76 http://creativecommons.org/licenses/by-nc-nd/4.0/ (2016) presented a mathematical model to optimize the use of resources. however, the model was not programmed or tested to determine its performance. the problem of production planning in am machines was described by li et al. (2017). a milp model (mixed-integer linear programming) was considered and it was solved using cplex. in addition, two heuristics (‘best fit’ and ‘adapted best fit’) were developed for minimizing the average cost of production per volume of material. as a result, it can be affirmed that am planning reduces costs considerably and the algorithms provided promising performance values within reasonable computing times. according to zhou et al. (2018) 3d printing has become a relevant service of cloud manufacturing, due to the vast personalization it offers to users. cloud manufacturing provides a suitable environment for integrating the resources of 3d printing technologies. it is a platform that allows the customers to carry out their corresponding orders, satisfying the individualized requirements of material and dimensions, among others (zhou et al., 2016). the platform groups the pieces and checks that the suppliers can do these orders. lonja3d project introduced by lópez-paredes et al. (2018) develops a model similar to that one presented by zhou et al. (2018), which is a cloud manufacturing platform. lonja3d model is based on a platform that allows grouping the orders from various customers that meet the same material and requirements in a single order the suppliers will receive. each supplier will have to decide which parts to do the printing. by manufacturing pieces from different clients simultaneously, the supplier manages to reduce its manufacturing costs, as shown by piili et al. (2015), thus offering more competitive prices. besides, the customer is favored by receiving the product at a lower price. li et al. (2017) and kucukkoc, (2019) tackled the chance of integrating a comprehensive nesting procedure valid to be integrated into the lonja3d model. the allocation of parts was solved considering their exact dimensions on the horizontal and vertical axes, rather than accounting the production area as a whole. our proposed heuristic figures out this problem, by assigning the ordered pieces to the available manufacturing area, keeping in mind to get the highest possible productivity. some authors have studied the benefits of packing parts in am. although we must note the production differences among the various am technologies, several studies report an increase in production efficiency. zhao et al. (2018) presented a case study where the packing algorithm showed could save over 50% of the manufacturing time compared to unpacked situations. also, piili et al. (2015) found that manufacturing parts simultaneously allows to reduce costs by 81-92% compared to the manufacture of parts separately, which proves that the optimal use of the manufacturing platform is the main variable that involves the supplier in terms of multi-item production. as a case of study to demonstrate the convenience of packing parts in the printing process, li et al. (2017) compared their novel heuristics for part scheduling (‘best fit’ and ‘adapted best fit’) with a regular assignment procedure (without a systematic approach). results obtained with both heuristics clearly outperformed the regular procedure. as a consequence, it is important to work out an appropriate distribution of parts in manufacturing batches for reducing costs and obtaining a higher economic benefit. 2.2. nesting problem approach while attempting to optimize the production planning of parts from different clients, li et al. (2017) focused on the process of assigning parts to jobs and jobs to machines (scheduling). nevertheless, they noticed a lack of efficiency in the nesting of the pieces in the manufacturing surface. one proposed further research line was to develop a nesting method for placing parts in the manufacturing bed that could be integrated into the scheduling process. in this work, we propose a model expected to fulfill those requirements. a part to be printed is characterized by raw material, width, length, height, volume, and filling percentage (i.e. a parameter used in am to determine the solidness of a piece). a 3d printing manufacturer usually will have different machines to produce with different raw materials (technologies). his goal is to maximize the return with each machine, which is possible when machines are working using their whole manufacturing volume without interruption 24 hours a day, 7 days a week. the orientation of parts is one of the most critical decisions to be made when it comes to am. cost, quality and time will be affected by the part’s orientation. notwithstanding, it has much to do with int. j. prod. manag. eng. (2020) 8(2), 75-86creative commons attribution-noncommercial-noderivatives 4.0 international production planning in 3d printing factories 77 http://creativecommons.org/licenses/by-nc-nd/4.0/ the technology employed; while in fdm orientation is not a key factor, in laser-based technologies it must be approached with caution (singhal et al., 2005). several studies have been made considering the orientation of a single part. zhang et al. (2016) reviewed the single-part orientation problem and proposed a new model aimed at achieving the orientation of a part that allows to reduce manufacturing time and cost. besides, an online tool was developed for users to find the right orientation according to their quality requirements. this importance of a part’s orientation lays on the anisotropic properties inherent to the layer-bylayer manufacturing process (shaffer et al., 2014). nevertheless, rotation around the vertical axis (z-axis) will be allowed since the critical direction is the z-axis itself, while orientation in the xy plane does not show a relevant influence on the final quality (sung-hoon et al., 2002). vertical rotation in the nesting process has also been considered in other works, such as in wang et al. (2019) and in wodziak et al. (1994). this assumption allows us to increase the search space so that the possibilities of finding a better solution are higher. the optimization problem for placing parts in the manufacturing surface has been introduced in several works with different approaches. as we have noticed for other characteristics, it has a great dependence on the 3d printing technology employed. wodziak., (1994) prioritized the number of parts produced rather than the percentage of occupation utilized in the stereolithography (sla) process. also for sla, canellidis et al. (2013) used three different criteria for placing the parts on the surface while the optimization objective was to maximize the area of the platform covered by the projections of the parts. moving on to powder-bed laser processes (e.g. selective laser sintering, selective laser melting, direct metal laser sintering), chergui et al. (2018) presented a model that considered delivery times in which the utilization of the manufacturing area was subordinated to the on-time service of the parts produced. in am the production cost (and in consequence the expected income) of a good directly depends on its mass (volume and filling percentage). to maximize the productivity of each 3d printer we should solve the puzzle that ensures for each manufacturing batch that the largest proportion of the manufacturing area will be occupied with the parts which have the highest filling percentage. in case the amount and number of different items to be produced is large enough, with differences in mass and geometries, there is a combinatorial explosion of possible manufacturing batches. to optimize the production planning, we will consider the profit increases with the total mass of each batch. 3. problem description 3d printing factories need to be prepared to produce lots of various parts from many customers. sometimes, the clients will order parts produced using different materials and manufacturing techniques. it opens the chance to break the lots from the clients and reorganize and combine them in case the items can be manufactured simultaneously. we have introduced that scheduling in am is divided into two steps: grouping parts for a print and placing parts in the printing space. in this work, we will delve into the second step, which is the most difficult to handle in computing terms. thus, we focus on nesting parts in the manufacturing surface as a part of the scheduling process in 3d printing. the nesting problem is also referred to as platform layout optimization. we propose a scenario in which a group of clients has made different orders of parts and the chosen supplier has to manufacture those parts. this is part of the supplier’s decision making on those parts to print within the lonja3d model. once a large set of parts has been grouped and assigned to a printer following the matching between parts’ requirements and printers’ parameters, the supplier must make the decision on which parts to print in each batch. since the capability of a printer is subjected to its printing surface, we assume that not all parts can be made at the same time. this means the supplier has to choose for the printer the parts to produce first and those that will be left for later batches. the problem we address is the placement of parts in the bed surface once the previous grouping step has been made. a heuristic procedure is developed, in which the primary optimization criterion is to maximize the income obtained for a batch and the manufacturing surface occupation is used to help derive the final solution. starting from a large set of parts to be allocated in a single printer, the problem is to solve the nesting of the parts on the platform. the input data will be int. j. prod. manag. eng. (2020) 8(2), 75-86 creative commons attribution-noncommercial-noderivatives 4.0 international de antón et al. 78 http://creativecommons.org/licenses/by-nc-nd/4.0/ the width, length, height, and filling percentage of each part from the set. also, the width, length, and height of the 3d printer will be introduced as data. the program will use those inputs to figure out an optimized layout with a subset of parts from the initial set. to solve this problem, the heuristic first seeks subsets with a large occupation percentage, and then select from those the subset showing the highest quantity of mass. consequently, that subset and its corresponding layout would be chosen by a manufacturer that tries to schedule the daily production of that printer in an optimized way. the remaining parts will be checked on the next production planning. 3.1. assumptions we have made some assumptions in order to limit the problem to our research objective. we are relaxing several requirements to simplify the situation so that we can offer a viable solution in the shortest possible time. a) to allocate each part in the manufacturing surface we will consider the minimum rectangle which guarantees it is inside. it is the projection of the geometry of the part on the xy plane plus the minimum tolerance needed to guarantee the quality is not compromised. b) the location of parts is worked out in 2d, in the base formed by the x and y axes. c) it is assumed that all parts on the list can be manufactured by the supplier (size, precision, and cost requirements are met). the grouping step of parts whose requirements match the features of their assigned printers has already been solved. d) parts can rotate 90 degrees around the vertical axis. e) dates of delivery are not considered as well as setup times, production times, or post-processing. we do not explore the “matching problem” among parts and printers. it is because the goal of this work is to develop a placing procedure for a set of parts in a manufacturing surface and check the workability of our heuristic . in any case, in section 2 we have pointed out some interesting works in which matching mechanisms were presented. thus, we will start from a list of parts to be printed and their assigned printer. in brief, we solve the problem of nesting parts from a large set in a single printer. 4. batch manufacturing optimization method to maximize the productivity of 3d printing machines, we propose a method inspired by combinatorial auctions (ca). as in a ca mechanism, we will define a method with two steps. the first one is to create manufacturing batches that occupy the maximum percentage of the manufacturing bed (the packing problem in ca). in the second one, the winner determination problem (wdp) in ca, we will evaluate for each batch the total mass (expected profit). while the most complex problem in combinatorial auctions is usually the wdp, in our case it is in the packing problem (pp) where the complexity is maximum. parts have previously been assigned to the printer with which their matching is higher. thus, the problem to solve is the “batch composition” to determine the layout in which parts must be done to obtain an optimized solution. this problem has been solved in other works as a two-step process, in which a grouping of parts was made prior to determine their placement on the surface. wang et al. (2019) starts from the whole list of parts to print and create a first division trying to put together parts with similar heights; then, parts of each group are divided into jobs for their successive printing. in our case, there is no previous grouping of parts and the batch composition is completely solved in the pp stage. this means that by obtaining an optimized platform layout we are solving two questions: which subset of parts must be selected and how those parts must be placed in the platform for the printing. optimization must be expressed in some kind of measure. this issue is here solved in the second step of the process. the winning choice will be selected according to our optimization criteria. this is what justifies the existence of a second stage in the procedure. should we want to select another parameter as the decision variable, we would only have to modify the winning criterion. the method will be implemented using python. the first stage tackles the pp and solves the allocation of parts on the bed. the second deals with the winner determination problem. 4.1. packing problem in the first stage, the pp is raised. it is started from a list of parts that will be reordered randomly in each int. j. prod. manag. eng. (2020) 8(2), 75-86creative commons attribution-noncommercial-noderivatives 4.0 international production planning in 3d printing factories 79 http://creativecommons.org/licenses/by-nc-nd/4.0/ of the simulations conducted. as we have assumed that parts will be simplified by their horizontal projections, the problem consists of finding the packing pattern that results in the largest occupied area. this problem is np-hard because as the number of parts increases, the number of possible combinations increases exponentially. being n the number of parts, the size of the solutions space is given by: 2n · n! (1) this is a consequence of the fact that there are n! sequences of rectangles (i.e. permutations) and each one can be placed in two ways since a 90° vertical rotation is allowed. thus, the search space is larger than the search space in the traveling salesman problem (equation 2). if 25 parts are given, then 1031 orthogonal packing patterns exist (jakobs, 1996). 225 · 25! > 1031 (2) we describe now the problem and the variables used for its characterization. the inputs of the parts are the name (pi), filling percentage (ri), length (li), width (wi), height (hi). the build platform area is given by name (aj), length (lj), width (wj) and height (hj), all of them in mm. we introduce the variable xij as the boolean variable that takes the value 1 if the part is assigned to the batch. also, we define two variables that indicate the position of the part on the xy plane by the top-left corner on the x-axis (cx) and the topleft corner on the y-axis (cy). first, we introduce two new terms that will be used through the algorithm: the list of available parts (lap), which contains the parts that remain to be assigned, and the list of available areas (laa), where a list of the unused areas can be found. figure 1. placement of the part pi on the area aj. the procedure begins with an empty surface. the first part will be allocated in the upper-left corner. then, two new subareas will be created from the remaining available surface: the first one by cutting from top to bottom, and the second one covering the remaining area. to start the allocation procedure, we choose the first part of the lap and try to assign it to the first area of the laa. to do this, we follow several steps (see figure 2): 1. we compare the width of the part (wi) and the width of the area (wj). a. wi < wj, then lengths are compared i. li < lj, then the part is assigned and xij takes the value 1. also, we create two new areas from the remaining available surface (see figures 3 and 4). the first one, (aj+1), is defined by: aj+1 = (li, wj – wi) (3) the length of the new area coincides with the length of the part assigned. the width is calculated as the difference between the width of the original area (aj) and the width of the part. similarly, we define(aj+2) by: aj+2 = (lj li, wj) (4) these two new areas are included in the laa, while the area aj is removed from the list. to define the position of these areas, they will be assigned their corresponding pair of coordinates from those variables defined at the beginning (cx and cy). ii. li > lj, we rotate the part 90° so that the new width wi’ coincides with the former length and the new length li’ coincides with the former width. once the change has been made, we repeat the checking procedure: b. wi’ < wj, then the length is checked ii. li’ < lj, then the part is assigned. two new subareas are created as explained in (i) and the laa is updated. c. wi’ > wj, we move on to the next area of the laa and restart from the first step. 2. wi > wj , we rotate the part 90° as we have explained in (ii) and repeat the checking procedure. int. j. prod. manag. eng. (2020) 8(2), 75-86 creative commons attribution-noncommercial-noderivatives 4.0 international de antón et al. 80 http://creativecommons.org/licenses/by-nc-nd/4.0/ a second piece follows the same procedure, i.e. if it does not fit in the first subarea, the second created subarea is checked for compatibility. this procedure is iterated until there are not available parts or surfaces that can be assigned. ¿wi < wj? no part pi fro m lap yes ¿li < lj? no rotate 90º xyplane allocate part piyes ¿wi’ < wj?no yes ¿li’ < lj? yes no try next area from l aa figure 2. flow chart for the allocation of a part. figure 3. dimensions of the new area aj+1. figure 4. dimensions of the new area aj+2. 4.2. winner determination problem (wdp) as well as other authors such as li et al. (2017) try to minimize the average production cost per volume of material, and others try to minimize the tardiness in the delivery of produced parts, see dvorak et al. (2018) and chergui et al. (2018), we acquire an alternative perspective of the planning problem: to maximize the income we get from each printer. this focus has its support on the fact that the goal of a company is to increase the profit, and resources and management systems are all subordinated to that goal (gupta & boyd, 2008). also, this approach runs in concordance with our scenario: do the daily planning for a 3d factory. once the manufacturing surface has been optimized (pp), the supplier has to choose among all the possible batches the one that offers the highest income. despite having solved the pp in 2d, parts are also characterized by their height (hi) and filling percentage (ri). the profit generated by each part has a close relationship with those two parameters and a decision criterion must be set to select the winner batch among those showing the highest occupancy percentage. one time the pp step finishes, the program provides two outputs that help determine the winner choice. the first one provides the given occupied manufacturing area. the second gives the total mass used for each batch. the last one allows us to solve the winner determination problem. parts have been simplified by their horizontal projections in 2d, and to determine their volume we will consider them as cubes. then, the volume of a part (vi) is given by: vi = li · wi · hi (5) in order to calculate the amount of material needed to manufacture a part (mi), we simply multiply the volume of the cube by the filling percentage (ri) as follows: mi = vi · ri (6) the total stuff of the batch is obtained by adding up the single masses of the parts assigned to it, and it is expressed as a volume (in mm3). this will allow to choose the batch that has the largest amount of stuff intending to obtain the highest income, being this a useful tool for the provider to maximize the return of the machine. int. j. prod. manag. eng. (2020) 8(2), 75-86creative commons attribution-noncommercial-noderivatives 4.0 international production planning in 3d printing factories 81 http://creativecommons.org/licenses/by-nc-nd/4.0/ costs in am can be simplified as a function of two components: a mass-dependent one and a constant that represents the pre-processing and postprocessing costs. c(m) = k + c · m (7) at the same time, material costs constitute a major proportion of the costs involved in producing a part with am techniques (thomas & gilbert, 2015). though this makes more sense for big parts rather than small ones (where the material cost is less relevant as compared to other am costs), we will assume that in our model it occurs indifferently. on an analogous reasoning, the income derived from the production of a part with am is mainly represented by the mass of the part multiplied by the price per kg. being i the income expected: i(m) = p · m (8) this is how we justify our winning choice as the batch with the highest mass. 5. experimentation: results a case of study has been used to validate the method. it is considered a 3d printer whose parameters are shown in table 1. in addition, 10 parts have been listed (see table 2). these meet the requirements already described. 120 simulations are done to derive a solution close to the optimal solution. the outputs of the simulations are displayed in tables 3 and 4. the results showed that in the simulation 77 the batch composed by parts p1, p9, p4, p7, p5, p2, and p8 got an occupation of the manufacturing area of 92.69% (37 075 mm2 of the 40 000 mm2 available). the mass used was 1 523 500.00 mm3. in simulation 26, it is obtained a batch with a higher percentage of occupation. the batch consists of the pieces p2, p7, p1, p3, p5, p6, p8 (see figure 5) with an occupancy rate of 97.63% (39 050 mm2 out of the 40 000 mm2 available). it is approximately 5% higher than the previous case. however, the amount of raw material is 1 502 500.00 mm3. that is a 21 000 mm3 difference from the previous case. as a consequence, the first option is selected as the winner since it should generate the greatest benefit for the manufacturer. figure 5. a) a batch that occupies the 92.69%. b) a batch which occupies the 97.63%. these results show that our heuristic achieves good performance in absolute terms. however, we will carry out some experiments to have a more accurate idea of how precise our approach is. we have addressed the packing problem and the winner determination problem separately, and now we introduce the findings reached. as for the pp, the two main concerns are the order of the parts in the list (lap) and the performance of our procedure compared to other authors’ studies. it seems reasonable to think that the order of parts to enter the algorithm could have an effect on the solution obtained. thus, we conduct an experiment forcing the order of the parts by size: first, we try allocating parts ordered from largest to smallest area; second, we try with parts ordered from smallest to largest area. results obtained reported that our heuristic performs better in the first case but compared to a random order of the parts it still shows a lower performance (see table 5). in conclusion, we can say that integrating local optimal search techniques can help save computing time and guide the solution. to check the performance of our heuristic, we have conducted experiments where we compared the solutions reached with those obtained by other authors. two scenarios are analyzed: a case with few parts and a case with a large number of parts to allocate. we run several experiments and, as an example, we will discuss the results obtained for an experiment with a list of 40 parts of 10 types and one with 106 parts of 20 types. because we are studying the pp, we analyze the results of surface occupation. for the case with low parts we use an experiment conducted by toro & granada-echeverri, (2007) since they include the geometric data of the parts. int. j. prod. manag. eng. (2020) 8(2), 75-86 creative commons attribution-noncommercial-noderivatives 4.0 international de antón et al. 82 http://creativecommons.org/licenses/by-nc-nd/4.0/ to solve the problem, they use the chu-beasley genetic algorithm and run it for 28 seconds. the best solution found reaches an occupancy percentage of 95%. running our heuristic for the same time we achieve 100% of occupation. the experiment with a list of 106 parts is introduced by cui, (2007) and we compare our results with those obtained by toro et al. (2008). the search space for this problem is up to 2106 · 106! = 9.2997 · 10201 solutions, so here the complexity is high. they table 1. features of the manufacturing area. name length(mm) width (mm) height (mm) area (mm2) volume(mm3) a1 200 200 200 40 000 8 000 000 table 2. attributes of parts. name length (mm) width (mm) height (mm) filling area (mm2) volume(mm3) p1 100 100 100 0.5 1 000 000 500 000 p2 100 100 100 0.5 1 000 000 500 000 p3 50 100 100 0.2 500 000 100 000 p4 50 100 100 0.2 500 000 100 000 p5 50 100 100 0.2 500 000 100 000 p6 50 100 100 0.2 500 000 100 000 p7 45 45 100 0.5 202 500 101 250 p8 45 45 100 0.5 202 500 101 250 p9 55 55 100 0.4 302 500 121 000 p10 80 80 100 0.3 640 000 192 000 table 3. batch associated with the largest amount of stuff. name length width height filling area (mm2) volume(mm3) stuff (mm3) p1 100 100 100 0.5 10 000 1 000 000 500 000 p9 55 55 100 0.4 3025 302 500 121 000 p4 50 100 100 0.2 5000 500 000 100 000 p7 45 45 100 0.5 2025 202 500 101 250 p5 50 100 100 0.2 5000 500 000 100 000 p2 100 100 100 0.5 10 000 1 000 000 500 000 p8 45 45 100 0.5 2025 202 500 101 250 total 37 075 3 707 500 1 523 500 table 4. batch with the highest percentage of the area covered. name length width height filling area (mm2) volume(mm3) stuff (mm3) p2 100 100 100 0.5 10 000 1 000 000 500 000 p7 45 45 100 0.5 2025 202 500 101 250 p1 100 100 100 0.5 10 000 1 000 000 500 000 p3 50 100 100 0.2 5000 500 000 100 000 p5 50 100 100 0.2 5000 500 000 100 000 p6 50 100 100 0.2 5000 500 000 100 000 p8 45 45 100 0.5 2025 202 500 101 250 total 39 050 3 905 000 1 502 500 table 5. comparison of the performance for the initial order of parts in the lap. order parts allocated % of area covered largest smallest p4, p5, p6, p7, p8, p12, p13, p14 & p15 88,19 smallest largest p4, p5, p6, p9, p10, p11, p13 & p14 44,44 random p1, p2, p3, p4, p6, p7, p8, p9, p10, p11, p13 & p14 100 int. j. prod. manag. eng. (2020) 8(2), 75-86creative commons attribution-noncommercial-noderivatives 4.0 international production planning in 3d printing factories 83 http://creativecommons.org/licenses/by-nc-nd/4.0/ propose a complex algorithm that merges different heuristic techniques (e.g. neighborhood search, simulated annealing) to find the best solution, which shows occupancy of 99.31%. we make 100 000 simulations with our heuristic and the best solution found reaches 94.03%. the conclusion reached for those experiments is that our heuristic performs well for a case with a relatively small number of parts, outperforming in some cases the results reported in the existing literature. nevertheless, as the complexity of the problem increases with the increase in the number of parts, the results are not as good as others analyzed, and computing time increases considerably. attending to the winner determination problem, we are pursuing the batch that has the highest amount of stuff. the two decisive parameters in this context are height and filling percentage. we conduct experiments to study their significance degree. in both studies we analyze a case in which we start from a list of 15 parts to allocate; we realize 10 000 simulations and choose the 8 best solutions. to study the importance of the filling percentage we consider that all parts have the same height. we compute the algorithm and analyze the results (see table 6). we see that the batch with the highest amount of stuff (i.e. the winner) is not the one with a higher percentage of area covered neither the one with the highest number of parts allocated. this can be explained because here the total mass is primarily determined by the filling percentage. in conclusion, we can say that parts with a high filling percentage have more value. table 6. experiment with parts of the same height. solution % area covered nº parts stuff (mm3) winner 97.22 11 20 800 000 max area 100 12 20 300 000 max nº parts 98.61 13 20 200 000 alternatively, we study the case keeping the filling percentage constant (and varying the heights) to analyze the significance of the height. as it happened in the previous study, the batch with the highest amount of stuff has less parts allocated and less area occupied than other solutions (see table 7). this allows us to prove that height has an important influence on the wdp and that parts with higher height can bring greater benefits to the detriment of others with greater area. table 7. experiment with parts of the same filling percentage. solution % area covered nº parts stuff (mm3) winner 95.14 11 79 500 000 max area 100 12 75 000 000 max nº parts 98.61 12 76 750 000 6. conclusions and future research production optimization can be understood in different ways. the approach taken in this paper is that suppliers should choose the batch that allows them to get the highest income. in a dynamic market, where customers are continuously making new orders, this allows to choose those parts that bring the maximum return at the end of the day. at the same time, the main problem is to solve the nesting of parts on the manufacturing surface trying to maximize the occupied area. a two-step method based on combinatorial auction has been presented. first, we solve the pp to create potential manufacturing batches. then we calculate the expected return for each one (the wdp). after conducting several experiments in which we compared our method with other techniques previously developed, we have reached interesting conclusions about its performance. we found that our program offers a better result when parts in the list are sorted from largest to smallest area. also, when the number of parts to allocate stays around 40, we obtain a quality solution in 5-10 minutes of computing time. however, for a list of about 100 parts the solutions show a lower quality and the computing time fires above 15 minutes. another finding proved was that the filling percentage and the height are relevant attributes, forcing to consider situations with a lower occupation of the bed but a higher use of material. once the conclusions have been discussed, we focus on the research lines that will enable the improvement of our procedure. we know that our method does better when parts are sorted from large to small. notwithstanding, parts in the list are randomly sorted. this opens up the possibility of integrating new search techniques to explore the neighborhood of the best solutions obtained in each simulation. with regards to the height, we have highlighted its importance in our wdp. nevertheless, differences in height among parts will probably translate into a int. j. prod. manag. eng. (2020) 8(2), 75-86 creative commons attribution-noncommercial-noderivatives 4.0 international de antón et al. 84 http://creativecommons.org/licenses/by-nc-nd/4.0/ waste of time in multi-parts manufacturing, since a short part cannot be removed from the plate until the tallest part is finished. a good solution could be a penalty factor for parts with noticeable differences in height. the time dimension is not included in our method. a main research line can be the integration of algorithms to calculate the time needed to manufacture a batch. at the same time, it will be interesting to expand this study (designed for a single machine) to a multimachine production scheduling. by integrating both problems (pp and wdp) in a comprehensive approach, we would need to introduce a bounding criterion to select a subset of the batches with a higher mass. this might be interesting according to our aim of choosing the batch with the highest mass. however, the searching process proves to be more flexible when setting the area covered as the first filtering parameter. besides, in the outlook of improving this procedure, we feel it is preferable to figure out both problems separately so that they can be addressed independently to improve their performance. acknowledgements this research has been partially financed by the project: “lonja de impresión 3d para la industria 4.0 y la empresa digital (lonja3d)” funded by the regional government of castile and leon and the european regional development fund (erdf, feder) with grant va049p17. references canellidis, v., giannatsis, j., dedoussis, v. (2013). efficient parts nesting schemes for improving stereolithography utilization. cad computer aided design, 45(5), 875-886. https://doi.org/10.1016/j.cad.2012.12.002 chergui, a., hadj-hamou, k., vignat, f. (2018). production scheduling and nesting in additive manufacturing. computers and industrial engineering, 126(may), 292-301. https://doi.org/10.1016/j.cie.2018.09.048 cui, y. (2007). an exact algorithm for generating homogenous t-shape cutting patterns. computers operations research, 34(4), 1107-1120. https://doi.org/https://doi.org/10.1016/j.cor.2005.05.025 dvorak, f., micali, m., mathieu, m. (2018). planning and scheduling in additive manufacturing. inteligencia artificial, 21(62), 40-52. https://doi.org/10.4114/intartif.vol21iss62pp40-52 gogate, a.s., pande, s.s. (2008). intelligent layout planning for rapid prototyping. international journal of production research, 46(20), 5607-5631. https://doi.org/10.1080/00207540701277002 gupta, m.c., boyd, l.h. (2008). theory of constraints: a theory for operations management. international journal of operations and production management, 28(10), 991-1012. https://doi.org/10.1108/01443570810903122 jakobs, s. (1996). on genetic algorithms for the packing of polygons. european journal of operational research, 88(1), 165-181. https://doi.org/10.1016/0377-2217(94)00166-9 kucukkoc, i. (2019). milp models to minimise makespan in additive manufacturing machine scheduling problems. computers and operations research, 105, 58-67. https://doi.org/10.1016/j.cor.2019.01.006 kucukkoc, i., li, q., zhang, d.z. (2016). increasing the utilisation of additive manufacturing and 3d printing machines considering order delivery times. in 19th international working seminar on production economics (pp. 195-201). innsbruck, austria. li, q., kucukkoc, i., zhang, d.z. (2017). production planning in additive manufacturing and 3d printing. computers and operations research, 83, 1339-1351. https://doi.org/10.1016/j.cor.2017.01.013 lópez-paredes, a., pajares, j., martín, n., del olmo, r., castillo, s. (2018). application of combinatorial auctions to create a 3dprinting market. advancing in engineering network, castro and gimenez eds. lecture notes in management and industrial engineering (in press), 12-13. mehrpouya, m., dehghanghadikolaei, a., fotovvati, b., vosooghnia, a., emamian, s.s., gisario, a. (2019). the potential of additive manufacturing in the smart factory industrial 4.0: a review. applied sciences, 9(18), 3865. https://doi.org/10.3390/app9183865 piili, h., happonen, a., väistö, t., venkataramanan, v., partanen, j., salminen, a. (2015). cost estimation of laser additive manufacturing of stainless steel. physics procedia, 78(august), 388-396. https://doi.org/10.1016/j.phpro.2015.11.053 shaffer, s., yang, k., vargas, j., di prima, m.a., voit, w. (2014). on reducing anisotropy in 3d printed polymers via ionizing radiation. polymer, 55(23), 5969-5979. https://doi.org/10.1016/j.polymer.2014.07.054 singhal, s.k., pandey, a.p., pandey, p.m., nagpal, a.k. (2005). optimum part deposition orientation in stereolithography. computer-aided design and applications, 2(1-4), 319-328. https://doi.org/10.1080/16864360.2005.10738380 int. j. prod. manag. eng. (2020) 8(2), 75-86creative commons attribution-noncommercial-noderivatives 4.0 international production planning in 3d printing factories 85 https://doi.org/10.1016/j.cad.2012.12.002 https://doi.org/10.1016/j.cie.2018.09.048 https://doi.org/https://doi.org/10.1016/j.cor.2005.05.025 https://doi.org/10.4114/intartif.vol21iss62pp40-52 https://doi.org/10.1080/00207540701277002 https://doi.org/10.1108/01443570810903122 https://doi.org/10.1016/0377-2217(94)00166-9 https://doi.org/10.1016/j.cor.2019.01.006 https://doi.org/10.1016/j.cor.2017.01.013 https://doi.org/10.3390/app9183865 https://doi.org/10.1016/j.phpro.2015.11.053 https://doi.org/10.1016/j.polymer.2014.07.054 https://doi.org/10.1080/16864360.2005.10738380 http://creativecommons.org/licenses/by-nc-nd/4.0/ sung‐hoon, a. (2002). anisotropic material properties of fused deposition modeling abs. rapid prototyping journal, 8(4), 248-257. https://doi.org/10.1108/13552540210441166 thomas, d.s., gilbert, s.w. (2015). costs and cost effectiveness of additive manufacturing: a literature review and discussion. additive manufacturing: costs, cost effectiveness and industry economics, 1-96. https://doi.org/10.6028/nist.sp.1176 toro, e., garces, a., ruiz, h. (2008). two dimensional packing problem using a hybrid constructive algorithm of variable neighborhood search and simulated annealing. revista facultad de ingeniería universidad de antioquia, 119-131. toro, e., granada-echeverri, m. (2007). problema de empaquetamiento rectangular bidimensional tipo guillotina resuelto por algoritmos genéticos. scientia et technica, 1(35). wang, y., zheng, p., xu, x., yang, h., zou, j. (2019). production planning for cloud-based additive manufacturing—a computer vision-based approach. robotics and computer-integrated manufacturing, 58(march), 145-157. https://doi.org/10.1016/j.rcim.2019.03.003 wodziak, j.r., fadel, g.m., kirschman, c. (1994). a genetic algorithm for optimizing multiple part placement to reduce build time. proceedings of the fifth international conference on rapid prototyping., (may), 201,210. https://doi.org/10.1.1.462.5661 zhang, y., gupta, r.k., bernard, a. (2016). two-dimensional placement optimization for multi-parts production in additive manufacturing. robotics and computer-integrated manufacturing, 38, 102-117. https://doi.org/10.1016/j.rcim.2015.11.003 zhao, z., zhang, l., cui, j. (2018). a 3d printing task packing algorithm based on rectangle packing in cloud manufacturing. lecture notes in electrical engineering, 460, 21-31. https://doi.org/10.1007/978-981-10-6499-9_3 zhou, l., zhang, l., laili, y., zhao, c., xiao, y. (2018). multi-task scheduling of distributed 3d printing services in cloud manufacturing. international journal of advanced manufacturing technology, 96(9-12), 3003-3017. https://doi.org/10.1007/s00170-017-1543-z zhou, l., zhang, l., xu, y. (2016). research on the relationships of customized service attributes in cloud manufacturing. asme 2016 11th international manufacturing science and engineering conference, msec 2016, 2, 1-8. https://doi.org/10.1115/msec2016-8530 int. j. prod. manag. eng. (2020) 8(2), 75-86 creative commons attribution-noncommercial-noderivatives 4.0 international de antón et al. 86 https://doi.org/10.1108/13552540210441166 https://doi.org/10.6028/nist.sp.1176 https://doi.org/10.1016/j.rcim.2019.03.003 https://doi.org/10.1016/j.rcim.2015.11.003 https://doi.org/10.1007/978-981-10-6499-9_3 https://doi.org/10.1007/s00170-017-1543-z https://doi.org/10.1115/msec2016-8530 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.11953 received: 2019-06-07 accepted: 2020-01-21 assembly line balancing by using axiomatic design principles: an application from cooler manufacturing industry yılmaz, ö. f.a*, demirel, ö. f.b, zaim, s.c, sevim, s.d akaradeniz technical university, department of industrial engineering, 61080, trabzon, turkey. bibn haldun university, school of business, 34494, i̇stanbul, turkey. cistanbul şehir university, department of industrial engineering, 34662, i̇stanbul, turkey. ci̇stanbul technical university, department of industrial engineering, 34367, i̇stanbul, turkey. a omerfarukyilmaz@ktu.edu.tr, b ofahrettin.demirel@ihu.edu.tr, c selimzaim@sehir.edu.tr, d sevims@itu.edu.tr abstract: the philosophy of production without waste is the fundamental belief behind lean manufacturing that should be adopted by enterprises. one of the waste elimination methods is assembly line balancing for lean manufacturing, i.e. yamazumi. the assembly line balancing is to assign tasks to the workstations by minimizing the number of workstations to the required values. there should be no workstation with the excessively high or low workload, and all workstations must ideally work with balanced workloads. accordingly, in this study, the axiomatic design method is applied for assembly line balancing in order to achieve maximum output with the installed capacity. in order to achieve this aim, all improvement opportunities are defined and utilized as an output of the study. computational results indicate that the proposed method is effective to reduce operators’ idle time by 12%, imbalance workload between workstations by 38%, and the total number of workers by 12%. as a result of these improvements, the production volume is increased by 23%. key words: lean manufacturing; axiomatic design; assembly line balancing. 1. introduction customer demands are the most important drivers for manufacturers. they want to have high-quality, low-cost products just-in-time and in full. hence, manufacturing companies should be more robust and responsive to continuous, variable and unpredictable demands from customers. in the fiercely competitive market, manufacturers need to use flexible, adaptive, active, and responsive strategies that meet customer demands at low cost and in a short time. to reach these targets, the lean manufacturing tools can be employed within an effective method. through the application of lean techniques, manufacturers can enable to eliminate wastes in their operations. lean manufacturing includes many tools to eliminate wastes such as kanban, smed, kaizen, value stream mapping, 5s, and visual management. assembly line balancing is one of the implementations of lean manufacturing, also known as yamazumi that helps to reduce overproduction, inventory, unnecessary motion, material handling activities, scrap, and operators’ idle time. the objective of assembly line balancing problems is to allocate operations to workstations in such a manner to optimize a criterion specified by considering the constraints regarding the accomplishment of the work. objectives addressed at line balancing problems are classified as two main types, technical and economic criteria. while to cite this article: yılmaz, ö. f., demirel, ö. f., zaim, s., sevim, s. (2020). assembly line balancing by using axiomatic design principles: an application from cooler manufacturing industry. international journal of production management and engineering, 8(1), 31-43. https://doi.org/10.4995/ijpme.2020.11953 int. j. prod. manag. eng. (2020) 8(1), 31-43creative commons attribution-noncommercial-noderivatives 4.0 international 31 mailto:omerfarukyilmaz@ktu.edu.tr mailto:omerfarukyilmaz@ktu.edu.tr mailto:selimzaim@sehir.edu.tr mailto:sevims@itu.edu.tr http://creativecommons.org/licenses/by-nc-nd/4.0/ the minimization of the number of workstations for cycle time is the one that is most widely used among technical criteria, the minimization of the cycle time for the number of workstations and minimization of the total idle time are the other technical criteria principally used with regard to the methods developed for assembly line balancing problems. from the operational cost perspective, minimizing the total cost of stations and labors are the most widely used economic criteria (ghosh and gagnon, 1989; yilmaz et al., 2016). in the current study, assembly line balancing problem (albp) is solved by using the axiomatic design approach to reduce waste and improve the system of a cooler manufacturing plant. the ad method was introduced about three decades ago, since then it has been used systematically both for academicians and practitioners in the industry (suh, 1990). the usefulness of the basic principles in terms of analysis, comparison, and selection is very effective in using ad in many areas such as production, organization, and system improvement. there is a set of important factors affecting the spread of ad. first, design methodologies developed for production systems cannot meet the needs due to continuous change in requirements. lean and agile manufacturing can easily adapt to rapidly changing conditions. the changes directly affect the production system; however, technological change at the factory level can change the production techniques, the features of the product and the way the workforce is employed (alcorta, 1999). ad is considered to be a robust approach to the changes in manufacturing systems. second, the manufacturing enterprises are inherently complex and the external factors affect the sustainability of the companies. the high ability of ad in the systematic propagation of functional requirements to the many aspects of a system design makes it an appropriate procedure in the context of system design (reynal et al., 1996). besides, thanks to ad, different levels of a system (in particular manufacturing system) can be interrelated in a holistic manner. third, the ongoing information revolution can affect the design process. today, the design is not guided by a decision-maker but it is systematically directed by a stepwise roadmap, which will be so crucial to apply the axioms (lipson et al., 2000). fourth, the separation of whats and hows in the ad results leads to flexibility providing superiority to ad versus other design methodologies. therefore, ad is an appropriate method to overcome the difficulties preventing to make decisions on the manufacturing system design (durmusoglu and satoglu, 2011). in this paper, we proposed a synthesized solution for an assembly line of the plant by using ad. the main objective of this study is to show the applicability of ad methodology to balance the assembly line. in this manner, a real case study is presented and a lean-based ad methodology is employed to balance the line by eliminating the waste. the goals with the implementation of the ad methodology are reducing the number of stations, total idle times, workload imbalance, and increasing the production rate. the scientific contributions of this paper are presented from two different sides as follows. from theoretical perspective, the main contribution of this paper is to implement a leanbased ad methodology to balance the assembly line. as for the managerial point of view, this study provides several managerial insights related to ad and its application on the assembly line by eliminating the wastes from the production system. the rest of this paper is organized as follows. a comprehensive literature review on ad is provided in section 2. the principles of ad are presented in section 3. the implementation of ad to the assembly line balancing problem is given in section 4. a real case study is given in section 5. concluding remarks and future research directions are presented in section 6. 2. literature review assembly lines play a significant role in the efficiency of production systems. setting up or reorganizing a line is a costly investment. therefore, it is important to organize the line effectively right from the beginning to the end. the basic problem during the design of the assembly line is to balance the workload with regard to workstations on production lines. unbalanced lines lead to inefficiency in production, increases in costs and further losses in technology and workloads (cakir, 2006). performance criteria used in the assembly line balancing problems is usually the minimization of the number of stations or the cycle time (ağpak et al, 2002; cevikcan and durmusoglu, 2011; yilmaz et al, 2016). int. j. prod. manag. eng. (2020) 8(1), 31-43 creative commons attribution-noncommercial-noderivatives 4.0 international yılmaz et al. 32 http://creativecommons.org/licenses/by-nc-nd/4.0/ many reviews of the literature are available for assembly lines design to reach several objectives by several methods (graves and lamar, 1983; mcmullen and frazier, 1998; guschinskaya et al., 2008; dolgui and ihnetsenka, 2009). ad is also one of the tools used in lean manufacturing and offers designers a very useful structure in order to achieve the final design object (nordlund and tate, 1996; matt, 2012). many companies use ad methodology to develop new products, processes, and approaches. thanks to the axiomatic approach, functional requirements can be determined by separating the design problem in a hierarchical manner. after ad theory and application principles were first developed by suh (1990), many studies have been carried out on the application of ad methodology over the last 30 years. in this section, the studies including ad applications are reviewed within four different streams: (i) cellular manufacturing systems, (ii) assembly lines, (iii) lean manufacturing system, and (iv) other types of production systems. the studies regarding the ad implementation for cellular manufacturing systems are reviewed below. black and schroer (1988) adapted the ad methodology to provide flexibility in the cellular manufacturing system. cochran et al. (2000) applied ad principles together with lean manufacturing principles and separated complex production systems into smaller and manageable systems. chen et al. (2001) used the independence axiom of ad with a decision support system to enhance the performance of a cellular manufacturing system. kulak et al. (2005) presented a road map to the design of cellular manufacturing system using ad principles. durmusoglu and satoglu (2011) a holistic methodology and road-map are presented to design a hybrid manufacturing system by implementing ad principles. ertay and satoglu (2012) used information axiom for a new product introduction to hybrid manufacturing systems. han et al. (2013) used ad methodology for virtual cellular manufacturing system design by focusing on the system cost and efficiency. the ad principles are also applied in assembly line design studies reviewed below. houshmand and jamshidnezhad (2004) used the ad methodology, which was developed based on lean manufacturing principles, to redesign a car body assembly line. matt (2012) used ad principles to control complexity dynamics in a mixed-model assembly system. matt (2013) proposed a design approach based on ad principles to control the sensitivity of assembly systems to fluctuations. hager et al. (2017) proposed a methodology for manufacturing system design, in particular assembly lines. celek et al. (2019) proposed an assembly system design methodology by implementing ad principles for aircraft fuselage structures assembly. rauch et al. (2019) defined the guidelines to implement the ad approach for designing flexible manufacturing and assembly systems. by doing so, the functional requirements were determined based on customer needs. to design a lean manufacturing system, ad is employed in several studies reviewed below. suh et al. (1995) provided an ad-based model for an ideal production system in line with lean principles. houshmand and jamshidnezhad (2002) combined capabilities and value stream mapping tool based on the ad method and developed a design model. houshmand and jamshidnezhad (2006) developed an ad methodology for lean manufacturing system design using process variables (pvs). nakao et al. (2007) proposed ad methodology for shortening lead time in the manufacturing process of tailor-made products. matt (2008) developed methodological guidance for the effective use of the axiomatic design in lean manufacturing systems. vinodh and aravindraj (2012) used axiomatic modeling for the design process of a lean manufacturing system. ad principles are also employed for other types of production systems reviewed below. gunasekera and ali (1995) employed the ad method conceptual stage of a metal forming process consisting of three-stage. suh (1997) proposed a new approach to define, classify, and design the system by considering ad methodology. babic (1999) developed a decision support system based on ad principles for a flexible manufacturing system. holzner et al. (2015) proposed ad for the systematic design of small and medium-sized enterprises (smes). khandekar and chakraborty (2016) proposed fuzzy axiomatic design principles to determine appropriate non-traditional machining processes by considering their importance. chakraborty et al. (2017) determined the design criteria so as to evaluate the re-manufacturability of products. int. j. prod. manag. eng. (2020) 8(1), 31-43creative commons attribution-noncommercial-noderivatives 4.0 international assembly line balancing by using axiomatic design principles: an application from cooler manufacturing industry 33 http://creativecommons.org/licenses/by-nc-nd/4.0/ the major finding from the review is that, to the best of our knowledge, a road map including all functional requirements of assembly line balancing has not been studied so far. hence, in this study, a methodology is developed using ad principles in order to fill this gap in the assembly line balancing literature. 3. principles of axiomatic design the main purpose of using ad is to ensure that design activities are carried out on a scientific basis with a theoretical foundation (suh, 2001). to this end, a systematic search process is carried out to achieve the best design solution among all alternative solutions. the basis of ad is to separate two important questions: what (objectives) to do and how (means) to do. in ad terminology, the objectives are represented by functional requirements (fr’s), while the solutions are represented by design parameters (dp’s). in the design process, the best set of design parameters that will meet the functional requirements are determined. the existence of design axiom is one of the most important concepts in ad. in this manner, the independence axiom corresponds to the first one, while the information axiom corresponds to the second axiom. they were introduced by suh (1990) as follows. axiom 1. the independence axiom (ia): ensure the independence for functional requirements axiom 2. the information axiom: information content should be minimized the relationship between the frs and dps can be expressed as {fr}=|a|{dp} where, {fr} corresponds to the functional requirement vector, {dp} is related to the design parameter vector, and |a| is the design matrix characterizing the design. each entry aij of |a| corresponds the ith fr to jth dp. the structure of |a| matrix describes the design type, which is considered by decision maker. so as to ensure the independence axiom, |a| matrix should be in form of uncoupled or decoupled design. |a| matrix is divided into three categories as explained follow. uncoupled design (most preferred): the |a| matrix is a diagonal matrix indicating the independence of fr-dp pairs. thus, each fr can be satisfied by simply focusing the corresponding dp. decoupled design (second choice): the corresponding |a| matrix is triangular. hence, the frs can be satisfied systematically fr1 to frn by just dealing with the first n dps. this design is most commonly encountered in real life. coupled design (undesirable): the |a| matrix does not define a special structure. thus, even a small change in any dp may have an impact all frs, simultaneously. coupled design should be avoided as much as possible while systems are designing. 4. assembly line balancing model through ad principles several main sources of waste are recognized in the assembly line and some practical solutions are suggested to alleviate these sources by using ad principles. the main waste in assembly lines is caused by the time differences between the operations that lead to idle time for the operators and high work-in-process. as a result, the utilization of the production line decreases. since the main goal is to meet customer demands on time, these irregularities lead to loose customers who are not satisfied. in order to prevent losing customers and profits, some preventive actions must be considered at the beginning of the system design. design is not just a random creative issue of an experienced expert but it is the product of systematic reasoning whose bases can be captured and generalized. when a company tries to become lean and more efficient, it will start to introduce lean concepts. making use of the ad approach, we analyze the assembly line system and propose a step by step plan toward lean manufacturing. according to the assembly line structure, it is necessary to redesign some of the activities which affect the line performance such as assigning tasks to operators, material handling between the work stations, eliminating non-value added operations, etc. the axiomatic design theory provides a framework to simplify the whole problem. it is also a hierarchical structure that eliminates all kinds of waste which is a prerequisite for other functional requirements. the frs represent the goals of the design or what we want to achieve, so they need to be improved by dps enhancing performance. int. j. prod. manag. eng. (2020) 8(1), 31-43 creative commons attribution-noncommercial-noderivatives 4.0 international yılmaz et al. 34 http://creativecommons.org/licenses/by-nc-nd/4.0/ at the first stage customer needs and attributes are recognized and formulated as frs and dp constraints. constraints establish the bounds on the acceptable design solutions and differ from frs in that they don’t have to be independent. in the following steps, the frs and dps are determined from highest to lowest level of the hierarchy by the zigzagging procedure which takes place between domains and specifies the relevant subproblems in the next level of the hierarchy. step 1: choose frs in the functional domain and mapping of frs in the physical domain. the highest level functional requirement is chosen to be meeting the required production volume on the assembly line. in order to meet customer demand on time, we should redesign the assembly line by eliminating non-value added tasks and make the line balanced. therefore the relevant design parameter of the function requirement is to balance the line to meet the required production volume. fr: meet the required production volume on the assembly line dp: balance the line to meet the required production volume the next step is to decompose the functional requirement that makes the design problem simple and easy to handle. step 2. decompose fr in the functional domainzigzagging between the domains. the following functional requirements are defined to achieve the highest level of fr. first of all to balance the line, operators and machines should work within the cycle time. morever, operators should know which task they will do to balance their operations, so their work definitions should have been defined before. finally, to meet the customer demand on time, we should always be informed. in these requirements, the sequence of frs is also important. all operators must perform the assigned tasks within predetermined cycle time. to do so, the machines can be operated within the production cycle time. another crucial decision here is that each operator must be assigned to the tasks in accordance with skill or skill levels. when these requirements met the information flow must be provided in a continuous manner. otherwise, the improvements cannot be permanent. fr1: make operators work within the production cycle time fr2: operate machines within the production cycle time fr3: ensure that operators are allocated to tasks according to their skills fr4: provide the continous information flow step 3. find the corresponding dpx’s by mapping frx’s in the physical domain. to satisfy the four frs defined above, we define the design parameters in the physical domain corresponding to the functional domain. to operate operators and machines within the production cycle time, firstly we should know that their working times. the first parameter is related to operators’ working times and it should be checked to control whether the requirement is met or not. the next parameter is also related to the second requirement. because it is not possible to assign any worker to a task without knowing the skill or skill levels, it must also be checked. the system must be designed to provide continuous information flow, which is related to the fourth requirement. dp1: check operators’ working times dp2: check machines’ working times dp3: check operators’ skill/skill levels (knowledge) dp4: design a system which strengthens the continous information network step 4. determine the design matrix. after frs and dps are defined, the corresponding design matrix is constituted. it is important that the design matrix (dm) must satisfy the independence axiom (ia) of ad principles. the design equation and the dm corresponding to the fr-dp sets are as follows. fr fr fr fr x x x x dp dp dp dp 1 2 3 4 0 0 0 0 0 0 0 0 0 0 0 0 1 2 3 4 = r t ssssssssss r t ssssssssss r t ssssssssss v x wwwwwwwwww v x wwwwwwwwww v x wwwwwwwwww int. j. prod. manag. eng. (2020) 8(1), 31-43creative commons attribution-noncommercial-noderivatives 4.0 international assembly line balancing by using axiomatic design principles: an application from cooler manufacturing industry 35 http://creativecommons.org/licenses/by-nc-nd/4.0/ the design matrix at this level is an uncoupled design and satisfies ia of ad. in the dm above, a symbol x represents a strong relationship between the corresponding fr-dp pair. step 5. decomposition of frxs and dpxs frxs and dpxs are very comprehensive parameters and cannot be applied directly. therefore, decomposition is inevitable to acquire a practical hierachy. first of all, fr1 is decomposed to the following sublevels. the processing times first need to be determined, after that non-value added tasks should be defined to be eliminated. the cycle time must not be exceeded while assigning the operators to them. fr11: define processing times fr12: define non-value added task fr13: allocate task to the operators without exceeding the cycle time the first requirement necessitates time study ensuring to be conducted by the first parameter which is required by all frs. ecrs analysis is at the core of the second parameter which must be carried out for second and third frs. lastly, the third parameter is strongly related to the third requirement. the corresponding dps may be stated as: dp11: time study dp12: improve operations with ecrs (eliminate, combine, rearrange, simplify) analysis dp13: balance the assembly line the design matrix for the above set of frs and dps are fr fr fr x x x x x x dp dp dp 11 12 13 0 0 0 11 12 13 => > >h h h this is a decoupled design satisfying ia of ad. allocation of task to operators may become practicable just after the tasks are defined and processing times are determined subject to cycle time. the functional requirement fr2 (operate machines within the production cycle time) may be decomposed with dp2 (check machines’ working times) as follows: fr21: define machine operation times fr22: define non-value adding operations effecting machine operation times fr23: allocate operations to machines without exceeding the cycle time the time study for machines must be carried out for all requirements. machine operations need to be improved by implementing ecrs analysis and it is used both second and third frs. balancing machine times is just used by the third fr. the corresponding dps are as follows: dp21: time study for machines dp22: improve machine operations with ecrs analysis dp23: balance machine times the design equation and the dm corresponding to the fr-dp sets are as follows. fr fr fr x x x x x x dp dp dp 1 2 3 0 0 0 1 2 3 2 2 2 2 2 2 => > >h h h once again, this is decoupled design satisfying ia of ad. before balancing machine times, machine operation times are defined by using time study and non-value adding operations are determined and improved by ecrs analysis. the functional requirement fr22 (define nonvalue adding operations effecting machine operation times) may be decomposed with dp22 (make improvements with ecrs analysis) as follows: fr221: reduce material handling fr222: reduce walking time fr223: reduce failures in machines fr224: minimize movements int. j. prod. manag. eng. (2020) 8(1), 31-43 creative commons attribution-noncommercial-noderivatives 4.0 international yılmaz et al. 36 http://creativecommons.org/licenses/by-nc-nd/4.0/ fr225: prevent waiting fr226: prevent overproduction fr227: reduce mistaken operations in these relations, except first dp, each dp must be conducted for the corresponding fr. the first dp is also used by second fr reducing walking times. when frs are examined, it is observed that each of them is independent of others. for this reason, constructing independent dps is more plausible to carry out requirements. the corresponding dps may be stated as: dp221: put equipments near the machines dp222: place machines close to each other dp223: apply jidoka dp224: modify layout dp225: feed the line on time dp226: establish pull system dp227: apply poka-yoke the design matrix for the above set of frs and dps are fr fr fr fr fr fr fr x x x x x x x x dp dp dp dp dp dp dp 221 222 223 224 225 226 227 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 221 222 223 224 225 226 227 = r t sssssssssssssssssss r t sssssssssssssssssss r t sssssssssssssssssss v x wwwwwwwwwwwwwwwwwww v x wwwwwwwwwwwwwwwwwww v x wwwwwwwwwwwwwwwwwww moreover, the functional requirement fr3 (ensure that operators know which operations they will do) are decomposed with dp3 (check operators’ knowledge). fr31: define operations fr32: measure operators’ current knowledge about the operations fr33: define required trainings and apply them fr34: visualization of results applying standard operation procedures must be carried out for all requirements. for instance, the results cannot be visualized without procedures. the second parameter is required by the last three frs. constructing a training matrix can be used to plan training and visualize the results. the skill matrix can just be employed for visualization. design parameters of these frs are as follows: dp31: apply standard operation procedures dp32: make examination during the operation process dp33: training matrix dp34: skill matrix the design matrix for the above set of frs and dps are fr fr fr fr x x x x dp dp dp dp x x x x x x 31 32 33 34 0 0 0 0 0 0 1 2 3 4 3 3 3 3 = r t ssssssssss r t ssssssssss r t ssssssssss v x wwwwwwwwww v x wwwwwwwwww v x wwwwwwwwww this a decoupled design satisfying ia of ad. to constitute training matrix, operations should be defined and the operators’ ability on these operations should be measured. then, to see the effects of trainings on operators, skill matrix is formed. training is one of the most important functions to perform the line balancing. by applying training plan, workers should be equipped with different abilities to compensate their operations and also reduce the number of operators. therefore, the functional requirement fr33 (define required trainings) are decomposed with dp33 (training matrix) as follows: fr331: improve operators skill level to do one more operation dp331: multi-functional operator training programme finally, the functional requirement fr4 (provide the continous information flow) is decomposed as follows: fr41: ensure continuity of information flow between departments int. j. prod. manag. eng. (2020) 8(1), 31-43creative commons attribution-noncommercial-noderivatives 4.0 international assembly line balancing by using axiomatic design principles: an application from cooler manufacturing industry 37 http://creativecommons.org/licenses/by-nc-nd/4.0/ dp41: application of report system and visualization without applying a report system and visualization, continuous information flow cannot be ensured. the frs of fr41 level are as follows: fr411: define required data to improve the continous information flow fr412: hold up to date information the corresponding dps can be stated as dp411: install new data system or improve current system dp412: update the system continuously current system improvement is necessary for both improving continuous information flow and up to date information. on the other hand, updating the system is just used by holding up to date information. the design matrix for the above set of frs and dps are fr fr x x x dp dp 411 412 0 411 412 =; ; ;e e e this is decoupled design satisfying ındependence axiom of ad again. the functional requirement of fr412 and the design parameter related with this fr are as follows: fr4121: control and track data system dp4121: assign responsible operator assigning a responsible operator is necessary to control and track the data system. this decomposition generally shown in figure 1 is then applied to our case performed in a cooler manufacturing company. 5. an application from cooler manufacturing industry the developed balancing design is applied to one of the assembly lines in sfa cool company in turkey. we chose the assembly line of the product ‘slim’ including 26 different task and 32 operators to perform these tasks. the workstations, the assigned tasks and the assigned operators are shown in table 1. figure 2 shows the precedence relationship diagram. table 1. assembly line before ad. before after work station assigned task assigned operator work station assigned task assigned operator 1 body assembly (1) 1-2 1 body assembly (1) 1-2 2 evaporator assembly (2) 3 2 evaporator (2) and compressor assembly (3) 3 3 compressor assembly (3) 4 3 welding preperation (4) and welding (5) 4 4 welding preperation (4) 5 4 condanser fan assembly (6) 5 5 welding (5) 6 5 evaporator fan cable assembly (7) 6 6 condanser fan assembly (6) 7 6 condanser assembly (8) 7 7 evaporator fan cable assembly (7) 8 7 thermostat assembly (9) 8 8 condanser assembly (8) 9 8 compressor fan assembly (10) 9 9 thermostat assembly (9) 10 9 helyum test (11) 10 10 compressor fan assembly (10) 11 10 eva. styr. ass. (12) and int. metals ass (13) 11 11 helyum test (11) 12 11 glass and led ass. (14) and glass clean. (15) 12-13 12 evaporator styraphor assembly (12) 13 12 sticker (16) 14 13 internal metals assembly (13) 14 13 shelf assembly (17) 15 14 glass and led assembly (14) 15-16 14 top cover assembly (18) 16-17 15 glass cleaning (15) 17 15 vacuum (19) 18 16 sticker (16) 18 16 gas pomp (20) 19 17 shelf assembly (17) 19 17 electrical gas leakage test (21) 20 18 top cover assembly (18) 20-21 18 performance test (22) 21 19 vacuum (19) 22 19 external metal assembly (23) 22-23 20 gas pomp (20) 23 20 cleaning (24) 24 21 electrical gas leakage test (21) 24 21 quality control (25) 25 22 performance test (22) 25 22 packaging (26) 26-27-28 23 external metal assembly (23) 26-27 24 cleaning (24) 28 25 quality control (25) 29 26 packaging (26) 30-31-32 int. j. prod. manag. eng. (2020) 8(1), 31-43 creative commons attribution-noncommercial-noderivatives 4.0 international yılmaz et al. 38 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 1. the decomposition of assembly line balancing system. int. j. prod. manag. eng. (2020) 8(1), 31-43creative commons attribution-noncommercial-noderivatives 4.0 international assembly line balancing by using axiomatic design principles: an application from cooler manufacturing industry 39 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5.1. current situation of the assembly line the company works for 540 minutes per shift. 155 units of slims should be produced per shift. therefore the cycle time of slim product is (540×60)/155=167 seconds/product. there are 32 operators working to produce slims whose operating times are shown in figure 3 before line balancing. as can be seen in figure 3, there is no balance in the line in terms of workload. some are above and some are below the cycle time. the aim is to balance the line, balance the workload and also reduce the idle times of workers and then providing effectively working balanced line. line balancing aims to reduce non-value added operations in the system. when these operations are eliminated, the idle times are shown in figure 4. moreover, average efficieny of the line is 55% which is calculated by summing all efficiencies of 32 operators after dividing by the number of operators. then the balancing ratio is calculated as 42% which is aimed to increase at least 80%. this ratio is important to have a good balanced line calculated by dividing the average of all 32 operator’s operating time by the maximum operating time of all operators. 5.2. improvements by axiomatic design principles firstly, ad method is applied defined in the previous section and then mention about the line improvements which are the results of ad method. here, some important fr&dp pairs are considered. the following fr&dp pair is related to fr3 (ensure that operators are allocated to tasks according to their skills) and dp3 (check operators’ knowledge) pairs. when the standard operation procedures are defined, operators know which operation should be done and they perform their operations following the standard operation procedures. in order to ensure this functional requirement, standard operation procedures are prepared for products and according to these procedures, training plan for operators is determined. after the training is completed, skill matrices are prepared for operators showing their capacity about the operation. in order to increase and ensure the continuity of information flow, key performance indicators are defined and charts are prepared to track them. fr41: ensure continuity of information flow between departments dp41: application of report system and visualization after that, boards are prepared, which contain all these indicators for every department to ensure visualization. hence everybody can see whenever they want to learn values of indicators. if there is a problem, when the department operative visits this board he/she can see the problem and communicate with the related department or person and solve the problem easily. figure 2. precedence relationship diagram. figure 3. operating times of the operators before line balancing. figure 4. operating times of the operators after balancing. int. j. prod. manag. eng. (2020) 8(1), 31-43 creative commons attribution-noncommercial-noderivatives 4.0 international yılmaz et al. 40 http://creativecommons.org/licenses/by-nc-nd/4.0/ fr4121: control and track data system dp4121: assign responsible operator to prevent that, area responsibles are determined who are responsible for entering data. for example, at the end of every hour these operators write production amount to the production table. if they have a problem or stoppage etc. during the production, they enter these data to the boards and give information to the department supervisor. some examples of tracking boards applied in the company are quality problems board, performance indicators (performance, productivity, quality) tracking board, area audit board, hourly production figures board, the department’s request and the last one is the production plan board. after this point, we study on balancing the workload, reduce the idle times and increase efficiency of the line. according to the following principles, some improvements are achieved and changes: distribute the tasks to operators. the operators should be occupied with tasks to reduce their idle times. the general instruction is that operator’s working time should not exceed the available time. try different combinations of tasks allocation, aiming at balancing the working time per operator till the best scenario is found. 5.3. results of the improvements when the assembly line is balanced, operators’ efficiency, productivity, saturity etc. changed. by considering the differences between the operating times of the operators, their working times are balanced as well. the operation times of the operators are shown in figure 5 for the new situation. the idle time of the operators is also reduced by an application of assembly line balancing. the average idle time was 155 seconds and the longest idle time was 222 seconds before balancing. after balancing, the idle times are reduced for each operator. for instance, the longest idle time is 147 seconds and average 70 seconds (see figure 6). the other performance indicators of the new assembly line system are summarized in table 2. the number of operators reduced to 28 from 32. before balancing, 118 units are produced per shift and the demand cannot satisy. after balancing, 155 units are produced per shift. to do so, the customers’ demands are met for this model. besides, the determined balancing ratio is reached and 32% saving is obtained. productivity is the most effective and important indicator for balancing. that shows how much labor required assembling this product. before balancing, 0.52 manhours are required to produce 1 unit. after balancing, this value changed to 0.40 manhours. that is, 0.12 manhours are gained and productivity is increased to 0.40 manhours. table 2. performance indicators of the assembly line. before balancing after balancing savings number of operators 32 28 4 output (unit) 118 155 37 balancing ratio (%) 42 80 38 productivity (manhour/unit) 0.52 0.40 0.12 6. conclusions in today’s globalizing world and international competition, enterprises have become aware that the key to industrial success is effective manufacturing systems, and geared their attention to how such systems may be set up with low costs. in the new system, the way to reduce manufacturing costs figure 5. idle time of the operators before line balancing. figure 6. idle times of the operators after balancing. int. j. prod. manag. eng. (2020) 8(1), 31-43creative commons attribution-noncommercial-noderivatives 4.0 international assembly line balancing by using axiomatic design principles: an application from cooler manufacturing industry 41 http://creativecommons.org/licenses/by-nc-nd/4.0/ involves producing standardized products in large volumes which is only possible by assembly lines. assembly line balancing (alb) is one of the most important problems for assembly lines to increase productivity. therefore, in this study, the alb problem is investigated from a different perspective in which the design of the assembly line has a strong impact on balancing. the main novelty of this study is that the ad method is employed as a line balancing tool by using lean principles. by doing so, line balancing improvements, reducing material handling, increasing communication, and better tracking of data are achieved. implementation of ad leads to determine the wastes that need to be eliminated to boost the system performance. the functional requirements are determined following the lean principles to balance the line effectively. these requirements answer the question of “what the system needs”. after the requirements are specified, the design parameters are determined to answer the question of “how the requirement can be met”. the proposed method is applied to the cooler manufacturing plant where the slim product is considered. in order to balance the assembly line, firstly the current situation is analyzed. after that potential improvement points are identified in the direction of axiomatic design results. at the end of all these improvements, the line is balanced to a ratio of 80%. the expected ratio is 75%, but better results are achieved. also, the production volume increased from 118 units to 155 units per shift. the operators’ idle time reduces and it is used as effective production times. the main contribution of this study is to apply the axiomatic design to balance the assembly line with lean principles. this study can be extended in several directions: (i) the vagueness inherent of processes can be considered through fuzzy, stochastic or robust modeling, (ii) ad method can be extended to a methodology by considering cellular manufacturing features, and (iii) other performance criteria can be taken into consideration. references ağpak, k , gökçen, h , saray, n , özel, s . (2013). stokastik görev zamanlı tek modelli u tipi montaj hattı dengeleme problemleri i̇çin bir sezgisel. gazi üniversitesi mühendislik mimarlık fakültesi dergisi , 17 (4). retrieved from https://dergipark.org.tr/en/pub/gazimmfd/ issue/6654/89311 alcorta, l. (1999). flexible automation and location of production in developing countries.  the european journal of development research, 11(1), 147-175. https://doi.org/10.1080/09578819908426731 babic, b. (1999). axiomatic design of flexible manufacturing systems. international journal of production research, 37(5), 1159-1173. https://doi.org/10.1080/002075499191454 black, j. t., schroer, b. j. (1988). decouplers in integrated cellular manufacturing systems. journal of engineering for industry, 110(1), 77-85. https://doi.org/10.1115/1.3187846 cakir, b. (2006). a simulation annealing algoirthm for stochastic process time based assembly line balancing, m.s. thesis, gazi university. celek, o. e., yurdakul, m., ic, t. (2019). axiomatic design of a reconfigurable assembly system for aircraft fuselages (no. 2019-01-1359). sae technical paper. https://doi.org/10.4271/2019-01-1359 cevikcan, e., durmusoglu, m. b. (2011). minimising utility work and utility worker transfers for a mixed-model assembly line. international journal of production research, 49(24), 7293-7314. https://doi.org/10.1080/00207543.2010.537385 chen, s. j. g., chen, l. c., lin, l. (2001). knowledge-based support for simulation analysis of manufacturing cells.  computers in industry, 44(1), 33-49. https://doi.org/10.1016/s0166-3615(00)00071-3 chakraborty, k., mondal, s., mukherjee, k. (2017). analysis of product design characteristics for remanufacturing using fuzzy ahp and axiomatic design. journal of engineering design, 28(5), 338-368. https://doi.org/10.1080/09544828.2017.1316014 cochran, d. s., eversheim, w., kubin, g., sesterhenn, m. l. (2000). the application of axiomatic design and lean management principles in the scope of production system segmentation.  international journal of production research,38(6), 1377-1396. https://doi.org/10.1080/002075400188906 dolgui, a., ihnatsenka, i. (2009). branch and bound algorithm for a transfer line design problem: stations with sequentially activated multispindle heads.european journal of operational research, 197(3), 1119-1132. https://doi.org/10.1016/j.ejor.2008.03.028 durmusoglu, m. b., satoglu, s. i. (2011). axiomatic design of hybrid manufacturing systems in erratic demand conditions. international journal of production research, 49(17), 5231-5261. https://doi.org/10.1080/00207543.2010.510487 ertay, t., satoğlu, s. i. (2012). system parameter selection with information axiom for the new product introduction to the hybrid manufacturing systems under dual-resource constraint. international journal of production research, 50(7), 1825-1839. https://doi.org/10.1080/00207543.2011.560205 int. j. prod. manag. eng. (2020) 8(1), 31-43 creative commons attribution-noncommercial-noderivatives 4.0 international yılmaz et al. 42 https://dergipark.org.tr/en/pub/gazimmfd/issue/6654/89311 https://dergipark.org.tr/en/pub/gazimmfd/issue/6654/89311 https://doi.org/10.1080/09578819908426731 https://doi.org/10.1080/002075499191454 https://doi.org/10.4271/2019-01-1359 https://doi.org/10.1080/00207543.2010.537385 https://doi.org/10.1016/s0166-3615(00)00071-3 https://doi.org/10.1080/09544828.2017.1316014 https://doi.org/10.1080/002075400188906 https://doi.org/10.1016/j.ejor.2008.03.028 https://doi.org/10.1080/00207543.2010.510487 https://doi.org/10.1080/00207543.2011.560205 http://creativecommons.org/licenses/by-nc-nd/4.0/ ghosh, s., gagnon, r. j. (1989). a comprehensive literature review and analysis of the design, balancing and scheduling of assembly systems. the international journal of production research, 27(4), 637-670. https://doi.org/10.1080/00207548908942574 graves, s. c., lamar, b. w. (1983). an integer programming procedure for assembly system design problems. operations research, 31(3), 522-545. https://doi.org/10.1287/opre.31.3.522 gunasekera, j. s., ali, a. f. (1995). a three-step approach to designing a metal-forming process. jom, 47(6), 22-25. https://doi.org/10.1007/bf03221198 guschinskaya, o., dolgui, a., guschinsky, n., levin, g. (2008). a heuristic multi-start decomposition approach for optimal design of serial machining lines. european journal of operational research, 189(3), 902-913. https://doi.org/10.1016/j.ejor.2006.03.072 hager, t., wafik, h., faouzi, m. (2017). manufacturing system design based on axiomatic design: case of assembly line. journal of industrial engineering and management, 10(1), 111-139. https://doi.org/10.3926/jiem.728 han, w. m., zhao, j. l., chen, y. (2013). a virtual cellular manufacturing system design model based on axiomatic design theory. in applied mechanics and materials (vol. 271, pp. 1478-1484). trans tech publications. https://doi.org/10.4028/www.scientific.net/ amm.271-272.1478 holzner, p., rauch, e., spena, p. r., matt, d. t. (2015). systematic design of sme manufacturing and assembly systems based on axiomatic design.procedia cirp, 34, 81-86. https://doi.org/10.1016/j.procir.2015.07.010 houshmand, m., jamshidnezhad, b. (2002). conceptual design of lean production systems through an axiomatic approach. in proceedings of icad2002 second international conference on axiomatic design. houshmand, m., jamshidnezhad, b. (2004). a lean manufacturing roadmap for an automotive body assembly line within axiomatic design framework. international journal of engineering transactions, 17(1), 51-72. houshmand, m., jamshidnezhad, b. (2006). an extended model of design process of lean production systems by means of process variables. robotics and computer-integrated manufacturing, 22(1), 1-16. https://doi.org/10.1016/j.rcim.2005.01.004 khandekar, a. v., chakraborty, s. (2016). application of fuzzy axiomatic design principles for selection of non-traditional machining processes. the international journal of advanced manufacturing technology, 83(1-4), 529-543. kulak, o., durmusoglu, m. b., tufekci, s. (2005). a complete cellular manufacturing system design methodology based on axiomatic design principles. computers & industrial engineering, 48(4), 765-787. https://doi.org/10.1016/j.cie.2004.12.006 lipson, h., suh, n. p. (2000). towards a universal knowledge database for design automation. in proceeding of icad2000, first international conference on axiomatic design, pg (vol. 250258, pp. 21-23). matt, d. t. (2008). template based production system design.  journal of manufacturing technology management,  19(7), 783-797. https://doi.org/10.1108/17410380810898741 matt, d. t. (2012). application of axiomatic design principles to control complexity dynamics in a mixed-model assembly system: a case analysis.international journal of production research, 50(7), 1850-1861. https://doi.org/10.1080/00207543.2011.565086 matt, d. t. (2013). design of a scalable assembly system for product variety: a case study.  assembly automation,  33(2), 117-126. https://doi.org/10.1108/01445151311306627 mcmullen, p. r., frazier, g. v. (1998). using simulated annealing to solve a multiobjective assembly line balancing problem with parallel workstations. international journal of production research, 36(10), 2717-2741. https://doi.org/10.1080/002075498192454 nakao, m., kobayashi, n., hamada, k., totsuka, t., yamada, s. (2007). decoupling executions in navigating manufacturing processes for shortening lead time and its implementation to an unmanned machine shop. cirp annals-manufacturing technology, 56(1), 171-174. https://doi.org/10.1016/j.cirp.2007.05.041 nordlund, m., tate, d., suh, n. p. (1996). growth of axiomatic design through industrial practice. in 3rd cirp workshop on design and the implementation of intelligent manufacturing systems, tokyo, japan (vol. 6, pp. 77-84). rauch, e., spena, p. r., matt, d. t. (2019). axiomatic design guidelines for the design of flexible and agile manufacturing and assembly systems for smes. international journal on interactive design and manufacturing (ijidem), 13(1), 1-22. https://doi.org/10.1007/s12008-018-0460-1 reynal, v. a., cochran, d. s. (1996). understanding lean manufacturing according to axiomatic design principles. suh, n. p. (1990). the principles of design (vol. 990). new york: oxford university press. suh, n. p. (1995). designing-in of quality through axiomatic design.  ieee transactions on reliability,  44(2), 256-264. https://doi.org/10.1109/24.387380 suh, n. p. (1997). design of systems. cirp annals-manufacturing technology,46(1), 75-80. https://doi.org/10.1016/s0007-8506(07)60779-3 suh, n. p. (2001). axiomatic design: advances and applications (the oxford series on advanced manufacturing). vinodh, s., aravindraj, s. (2012). axiomatic modeling of lean manufacturing system. journal of engineering, design and technology, 10(2), 199-216. https://doi.org/10.1108/17260531211241185 yilmaz, o. f., cevikcan, e., durmusoglu, m. b. (2016). scheduling batches in multi hybrid cell manufacturing system considering worker resources: a case study from pipeline industry.  advances in production engineering & management,  11(3). https://doi.org/10.14743/apem2016.3.220 int. j. prod. manag. eng. (2020) 8(1), 31-43creative commons attribution-noncommercial-noderivatives 4.0 international assembly line balancing by using axiomatic design principles: an application from cooler manufacturing industry 43 https://doi.org/10.1080/00207548908942574 https://doi.org/10.1287/opre.31.3.522 https://doi.org/10.1007/bf03221198 https://doi.org/10.1016/j.ejor.2006.03.072 https://doi.org/10.3926/jiem.728 https://doi.org/10.4028/www.scientific.net/amm.271-272.1478 https://doi.org/10.4028/www.scientific.net/amm.271-272.1478 https://doi.org/10.1016/j.procir.2015.07.010 https://doi.org/10.1016/j.rcim.2005.01.004 https://doi.org/10.1016/j.cie.2004.12.006 https://doi.org/10.1108/17410380810898741 https://doi.org/10.1080/00207543.2011.565086 https://doi.org/10.1108/01445151311306627 https://doi.org/10.1080/002075498192454 https://doi.org/10.1016/j.cirp.2007.05.041 https://doi.org/10.1007/s12008-018-0460-1 https://doi.org/10.1109/24.387380 https://doi.org/10.1016/s0007-8506(07)60779-3 https://doi.org/10.1108/17260531211241185 https://doi.org/10.14743/apem2016.3.220 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.12360 received: 2019-09-19 accepted: 2020-01-21 feasibility study of a facility to produce injection molded parts for automotive industry yudianto, a. a1*, tan, h.a2, qu, z.a3, xue, q.a4, naveen, a.c.a5, mushtaq, m.a6, gopi, k.s.a7 adepartment of mechanical and aerospace engineering, politecnico di torino corso duca degli abruzzi 24, 10129, torino, italy. a1 s250659@studenti.polito.it abstract: this study aims at the preliminary assessment in designing a complete stand-alone industrial facility to produce injection molded parts for the automotive industry. a draft design solution to allow the company to evaluate the capital investment was performed giving an estimated solution in project profitability. proposed successive design steps were developed. it includes the definition of input data and information, quantity determination, plat layout diagrams, machine selection, selection of material handling equipment, plant layout design including space requirements of production centers, aísles, support functions. moreover, the outdoor facility masterplan design is also proposed. finally, investment calculation via cash flow analysis is calculated. key words: feasibility study; facility design; plant layout design. 1. introduction facility layout design has been considered as one of the essential aspects of a company to have a significant reduction in the operational costs of the firm (kovàcs, 2019). the proper planning of the overall layout production plant needs to be deeply conducted as the first step of a company to run the business. however, everything starts with an assessment of the planned business area in terms of the practicality of a proposed plant of the business sector. a preliminary feasibility assessment has been widely conducted as the first, foremost step before starting a business or developing projects (hwang et al., 2019, halil et al., 2016, ma et al., 2019, eksangsri and jaiwang, 2014, chingua et al., 2019). the main objective of a feasibility assessment is to ‘de-risk’ potential full trial funding (morgan, 2018). considering some assessments in the feasibility study performed by mentioned researchers, this research activity can mainly provide reassurance and potentially avoid waste and to obtain an answer to the project in question of the company, which in this case is especially applied in the automotive sector. the complete work is to develop a feasibility study aimed at the preliminary design of a complete stand-alone industrial facility to manufacture aluminum alloy and injection molded parts for the automotive industry. however, this study is one part of the complete feasibility analysis of the plant. this paper focuses only on the production of injection molded parts. in this case, the industry is planning to manufacture bumpers, car dashboards and wheel arches. the mentioned facility comprises a fullyequipped production plant layout, raw material storage, work in process areas, finished product warehouse, auxiliary materials warehouse, and to cite this article: yudianto, a., tan, h., qu z., xue q., naveen a.c., mushtaq, m., gopi k.s. (2020). feasibility study of a facility to produce injection molded parts for automotive industry. international journal of production management and engineering, 8(1), 45-57. https://doi.org/10.4995/ijpme.2020.12360 int. j. prod. manag. eng. (2020) 8(1), 45-57creative commons attribution-noncommercial-noderivatives 4.0 international 45 https://orcid.org/0000-0003-3944-8948 http://creativecommons.org/licenses/by-nc-nd/4.0/ auxiliary function spaces. a complete masterplan design for both indoor and outdoor master plant is proposed. finally, an economic assessment is performed to estimate project profitability. 2. analysis and assessment methodology approach 2.1. description of plant location the manufacturing plant is allocated in a completely new building or greenfield solution. this solution uses a piece of land located in borgaro, torino, italy. the location is planned inside the area indicated in figure 1. figure 1. study area (source: google maps). 2.2. product description, routings, and production requests figure 2 shows the sample of the components to manufacture. the bumpers are made up of an external body assembled with an internal component named cross beam. it allows the bumper fastening to the vehicle body through specific hooks. this bumper is for small vehicles. the cross beam and external body are assembled by a gluing machine that carries out the operation in a semi-automatic way. additionally, the car dashboard assembly consists of some parts which are body, insert, air duct, instrument panel frame, glove compartment interior and exterior. the dashboard body and insert are welded utilizing a vibration welding machine while the other components are manually assembled. in the end, dashboards are subjected to quality inspection and then are sent to the finished product warehouse. instead, the wheel arch is a part of the vehicle body to provide a connection with the outer body portion. it surrounds the vehicle at the top and connects with the vehicle interior. in total, there are four wheel arches in a vehicle front right, front left, rear right, and rear left. the production routing of the bumpers starts with raw material receiving, granule loading in the press hoppers, injection molding, part unloading, cross beam gluing, final inspection, packaging and transport to the warehouse, storage in warehouse and shipping. the routing for car dashboards starts with raw material receiving, granule loading, body injection molding, insert injection molding, component injection molding, final inspection, packaging and transport to the warehouse, storage in the warehouse, and shipping. then, the routing for wheel arches is raw material receiving, granule loading in the press hopper, front wheel arch injection molding, rear wheel arch injection molding, part unloading, final inspection, packaging and transport to the warehouse, storage in the warehouse and finally shipping process. (a) (b) (c) (d) (e) (f) (g) figure 2. sample figure of (a) bumper, (b) wheel arch, (c) air duct, (d) insert, (e) instrument panel frame, (f) glove compartment exterior and (g)glove compartment interior (cagliano and chiabert, 2018) table 1. assigned unit loads. product/component box size [mm] units/box finished bumper 1900×800×500 2 finished dashboard 1800×800×700 4 bumper body 1900×800×500 4 bumper cross beam 1400×800×600 8 dashboard body 1400×800×600 6 dashboard insert 1400×800×600 10 air duct 1000×800×800 56 instr panel frame 1000×800×800 100 glove comp – ext 1000×800×800 80 glove comp – int 1000×800×800 80 front wheel arch 1900×800×800 2 rear wheel arch 1800×800×800 2 int. j. prod. manag. eng. (2020) 8(1), 45-57 creative commons attribution-noncommercial-noderivatives 4.0 international yudianto et al. 46 http://creativecommons.org/licenses/by-nc-nd/4.0/ on an annual basis, the production request of each product is 132,000 units for bumpers, 120,000 units for car dashboards and 116,000 units for wheel arches. the operation runs 270 days a year, six days/week, three shifts for monday tuesday – wednesday – thursday-friday, two shifts for saturday. therefore, the total is 17 shifts/week. the working hour is 7.5 hour per shift. however, office staff works 8 hours per day one shift from monday to friday. the finished products are requested to be packed in a box with the detail mentioned in table 1. it is also suggested that no more than two boxes can be stored on a single pallet. the unit load type for all components is a cardboard box on a wooden pallet. 2.3. raw materials it is requested for the raw material storage to have a capacity to meet the requirements of a minimum of five days of production for each of the three products manufactured in the facility. for all the products, the material consists of granules of thermoplastic material available in the market in octagonal-based cartoon boxes (octa bins) placed on pallets. each octabin contains of 1000 kg of granules and has 1000×1200×1500 (h) mm of size. the adhesive or glue is available in 25 kg drums. each pallet carries six drums, and the total pallet sizes are 1000×1200×700 (h) mm. the raw material requirement of each component can be seen in table 2. the abbreviation “pol.” stands for polypropylene. raw materials arrive on trucks to the receiving docks and are stocked in the storage area. the delivery of raw materials is scheduled for a maximum of three days per week. table 2. raw material requirement. part raw material quantity/ unit [kg] bumper body talc-filled pol 30% 3.6 bumper crs beam talc-filled pol 30% 0.77 bumper +crs beam adhesive 0.25 dashboard body glass-filled pol 12% 3.19 dashboard insert abs 2.06 air duct glass-filled pol 12% 0.87 instr panel frame glass-filled pol 12% 0.66 glove comp ext glass-filled pol 12% 0.54 glove comp int glass-filled pol 12% 0.53 front wheel arch talc-filled pol 15% 5.6 rear wheel arch talc-filled pol 15% 5.2 therefore, it is then possible to calculate the total raw material needed per shift and the total number of pallets to store the raw material. here, the production process is assumed to have 1% of scrap in every operation, and the calculation also considers the production request described in the previous section. the estimation of the total raw material per shift can be observed in table 3. table 3. raw material estimation per shift and the total number of pallets needed. material sum [kg/shift] storage[pallet] talc-filled pol 30% 769 12 glass-filled pol 12% 932 14 abs 333 5 talc-filled pol 15% 827 13 adhesive 44 5 2.4. machine layout the layout choice is determined by the number of products, production volume, the homogeneity of the product, the routings, etc. the proper choice of the machine and layout positioning is crucial to minimize the sum of manufacturing and material handling costs (schaller, 2008). the approach that is used in this study is that, first of all, it is necessary to create a multicolumn process chart to show the production routing of each product under production. then an activity relationship diagram needs to be created. finally, considering the size of the machine and the available space, the machine layout can be determined. figure 3. machine layout position. the multicolumn process chart indicates which workstation or machine has the most relationship with which other workstation. figure 3 shows the solution of the machine layout proposed. 2.5. production scheduling scheduling activities could be one of the determinants in deciding the further facility design of the plant. many scheduling techniques, optimization int. j. prod. manag. eng. (2020) 8(1), 45-57creative commons attribution-noncommercial-noderivatives 4.0 international feasibility study of a facility to produce injection molded parts for automotive industry 47 http://creativecommons.org/licenses/by-nc-nd/4.0/ approaches, and evaluation to have more effective production scheduling and improvement have been proposed by many researchers (hazaras et al., 2014, al-aomar, 2006, sun et al., 2015). however, a simple continuous production scheduling is utilized as a preliminary decision of the study. considering the production request per week, the routing, the requirement of the storage capacity of a minimum of five days, the following production schedule is possible to be applied in the production process. even if the maintenance schedule can be adjusted to have the desired throughput (yang et al., 2007), a simple assumption of two hours setting and maintenance schedule is set. 2.6. machine and workstation the production equipment to manufacture bumpers, car dashboards, and wheel arches, together with average hourly production rate and the approximate machine cost need to be determined. the number of machines for each product can be calculated by using equation 1. in this case, takt time is the time required to produce one unit of product. considering 92% of efficiency and average scrap rate equal to 1%, it is possible to calculate the takt time. then, by using the time standard, the number of machines required to produce each product is possible to calculate. finally, the result of the number of machines required can be calculated. the calculation considers the same machine might be used to produce different products. therefore, the computation includes the total time during the process. #machine takt time takt time 1 1time standard time standard = = (1) based on the specification of each machine dimension, then, it is necessary to calculate the total spaces required to determine the best layout that has been mentioned in the previous section. however, it is also necessary to consider a free space to allow the workers to perform maintenance and repair action around the machine. the offset area of 1000 mm is required to be added to space. moreover, some additional spaces to have a comfortable working environment and spacious workstation are needed. in this case, the total area is multiplied by 150% to have a spacious area. 2.7. number of employees the employees consist of direct and indirect employees. direct employees are those who work directly in the production line, and indirect employees are the one who works as officers or administration staff. for the direct employees, the number of operators for each machine needed is considering the size and the kind of job they are performing. one 3000 t injection molding press machine requires two operators. considering the number of machines and total shifts per day in the schedule, it finally requires 12 operators. furthermore, for 2500 t machine by considering again production schedule and because the dashboard body and rear wheel arch require two operators and the other two needs one for each, the arrangement is possible to be made one operator for one machine and two for the others. therefore, it needs three operators for each shift. in total, nine operators are required. the gluing machine and welding machine need two operators for each, and the other remaining machines need one operator per machine. therefore, 39 additional operators are required. finally, a total of 60 direct employees is required. it is, however, required to have indirect people who help the direct employee. literature suggests having an additional 20% of direct employee that works as helpers and support functions, this includes the person who works as an administrative staff in the office (stephens and meyers, 2013). these employees are considered as indirect employees. additionally, one cell supervisor per shift and warehouse workers per shift are also required. taking into account, there are three shifts per day, and there are seven cells and the number of warehouses, the additional employee needed are 33 people. in total, there are 45 indirect employees required. 2.8. plant support functions receiving area functions are locating trailers at receiving docks, unloading material, opening, inspecting, counting, reporting material received, moving to the raw material warehouse (stephens and meyers, 2013). receiving docks and areas must be sized to execute daily receiving functions. the approximation of size determination is calculated by the following formula. # of receiving docks = t×tt (2) receiving area size=(s×q)×tm (3) int. j. prod. manag. eng. (2020) 8(1), 45-57 creative commons attribution-noncommercial-noderivatives 4.0 international yudianto et al. 48 http://creativecommons.org/licenses/by-nc-nd/4.0/ q is the quantity of incoming raw material unit loads per unit of time. in this case, we have 20 pallets/hour. t is the number of trucks per unit of time (arrival rate) that is 1 truck/hour. s is the size of the unit load of floor area which is 1000×1200 mm. tt is the time required to unload the truck (unloading service time) which equals 45 min/truck or 0.75 hours/truck. tm is the time required to receive and move unit loads to the raw material warehouse (moving service time) which is 2.66 hours by assuming having 4 min/pallet. the number of receiving docks is 1 and the required size is approximately 63.84 m2. finally, take into account the additional space for aisle and space for handling and maneuvering and also the plant condition, the final area required is about 198 m2. shipping area functions are packaging finished goods for shipping, weighing, loading trailers, documentation: bills of lading (stephens and meyers, 2013). it is also necessary to design space for packaging, staging, aisles, offices, trailer parking and roadways. one primary constraint in determining the required area is the trailer size. the size of the trailer is commonly called feu (fourth-feet equivalent unit) which has the dimension of 8 feet wide, 40 feet long, and 7 feet high. if it is converted into cubic meter the total volume becomes 63.431 m3. at this point, it is necessary to calculate the number of trucks per day. the following equation is then used. # of trucks per day volume trailer volume product = (4) however, the total volume product remains to calculate. taking in to account all data that have been mentioned before, the number of boxes per day and the required size to store is depicted in table 4. therefore, the number of trucks required to ship the product is 9 trucks per day. table 4. volume of product. production #box/day vol prod/day [m3] bumper 258 196.08 car dashboard 117 117.936 front wheel arches 113 136.8 rear wheel arches 113 129.6 total volume 580.416 apart from the area needed to store the unit loads, the area for staging and area for maneuvering for material handling is required. considering the location and position of the land and the position of the product flow in the plant, the approximated area of the shipping activities is about 535 m2. figure 4. receiving area. figure 5. shipping area. maintenance is a service to the company’s equipment. more commonly a central maintenance area is designed to include equipment, machine overhead areas, maintenance supply, and spare parts storage areas (stephens and meyers, 2013). it is assumed to have approximately 150 m2 of the total area of the maintenance room. a tool room is made up of machines and an assembly area similar to production. the tool room size is the total of all the equipment space requirements times 200%. the assumed total area needed for the tool room is 30 m2. the utility room includes battery charging spaces, heating, electrical panels, air compressor, and air conditioning. these areas are kept separate from the regular traffic. electrical panels are fenced off, heaters are kept clean, and air compressors are located in a particular construction because they are noisy. the assumed area needed is approximately 70 m2. 2.9. employee facilities several employee facilities need to be taken into account in this section. the first employee facility is parking lots. office parking may be different from factory parking because it can incorporate the visitor parking spaces in this area. additionally, it might int. j. prod. manag. eng. (2020) 8(1), 45-57creative commons attribution-noncommercial-noderivatives 4.0 international feasibility study of a facility to produce injection molded parts for automotive industry 49 http://creativecommons.org/licenses/by-nc-nd/4.0/ also have different parking locations for managers or guests. the wider the parking spaces are, the less door damage there is in the parking lot. it is suggested to have a minimum of 18.58 m2 for each car (stephens and meyers, 2013). considering there are 105 employees in total, the total available area for parking is about 1,950 m2. additional parking lots for managers and guests could be added separately. the second facility is the employee main entrance. this is where the employees enter the plant will influence the placement of parking, the locker room, time card racks, restrooms, and cafeterias. the flow of people into a factory is from their cars into the plant via the employee entrance to their lockers and the cafeteria to wait for the start of their shifts. the employee entrance is where security, time cards, bulletin boards, and sometimes the personnel departments are located. depending on the management’s attitude and corporate requirements, employee entrance can vary from a simple doorway with a time card rack and time clock to a series of offices and gates through which to pass. the size of the employee entrance must consider individual requirements. the door could measure 1.8 m with an aisle or walkway leading into the plant. locker rooms give employees space to change from their street clothes to their work clothes and a place to keep their personal effects while working. their coats, lunches, street shoes, and so forth will be kept in lockers. the size of a locker room can be initially sized by multiplying the number of employees by one m2 per employee. break facilities comprise of cafeterias with serving lines, dining rooms, off-site diners for any typical plant. a table (or tables) allowing a minimum of 0.6 m width by 0.3 m length of table space per person. chairs or seats with back support for each person likely to be eating at one time. for the individual table space for a person 0.18 m2. however, this is adjustable as the need. the next one is medical facilities. these facilities should give first aid and accident treatment; under special conditions, full medical care in general practice may be arranged for the workers and their families at the industrial health center. contents of medical kit are the size and layout of the workplace, the number, and distribution of employees throughout the workplace, the nature of any hazards and the severity of the risk, location of the workplace and known occurrence of accidents or illnesses. an office layout is the next consideration of these facilities. it needs a total of about 678 m2 by considering the 20-25 m2 surrounding area, aisles, stairs, and various services. the last employee facilities are restrooms and toilets. the number of toiles must be distributed in several places as the need. the local building code may dictate how many toilets are necessary. the number of washbasins is equal to the number of toilets. the size of a restroom is 1.3 m2 per toilet, washbasin, and entryway, and 0.83 m2 for urinals. 2.10. unit loads the main unit loads, in this case, are the unit loads for raw material, for semi-finished products and finished products. the main unit load is the pallet. in order to calculate the number of required pallets in each function department, the requirement of at least five days of minimum storage that has been mentioned in the previous section needs to be taken into account. secondly, the requirement of no more than 2 stacks per box for finished products, total production quantity of components for one week for both semi-finished and finished products. finally, the total pallets that are required for the production process are as mentioned in table 5. table 5. total pallets. item qty raw materials 59 finished bumper 264 finished dashboard 123 bumper body 60 bumper cross beam 35 dashboard body 42 dashboard insert 50 air duct 8 instrument panel frame 9 glove comp interior part 11 glove comp exterior part 11 front wheel arch 71 rear wheel arch 106 total 633 2.11. material handling equipment equipment to move and transport the material in the plant needs to be chosen as the need. many suggested approaches, advanced techniques, and methods for structuring and deciding the choices for outsourcing material handling could be performed (klingenberg et al., 2010). however, int. j. prod. manag. eng. (2020) 8(1), 45-57 creative commons attribution-noncommercial-noderivatives 4.0 international yudianto et al. 50 http://creativecommons.org/licenses/by-nc-nd/4.0/ the approach utilized in this study is to estimate the main equipment only. however, the main handling equipment is the forklift. to obtain the number of the forklifts needed, the different paths that will be done by the different products have to consider. the weight of the pallets is known, then considering the capacity and dimensions of the forklift, it is possible to obtain the number of the unit load/day that can be carried by one forklift. the average distance considers the journey and return. therefore, we can get the number of forklifts needed, considering that this is a quite rough computation and useful only as a first approximation. by considering the allowed maximum speed of the forklift is 8 km/h, and the availability of each forklift per day is 8 hours, the following calculation of the number of forklifts can be calculated. n h n tf i i i i= | (5) where ni is the number of unit loads that have to be moved per day of each item. ni is the number of the unit load carried by one forklift in one trip. ti is time to perform one trip by the forklift moving item i. h is the availability of each forklift per day. the distance travel is calculated based on the designed overall layout of the plant. as a result, the total forklift needed is approximately 8 units. 2.12. warehouse space utilization and the performance optimization of warehouses are one of several aspects that need to consider in designing the storage system, and it depends on the kind of the warehouses (derhami et al., 2019). in this case, the first storage is the raw material warehouse. after the material has received in the raw material receiving area, it must be moved and stored in the raw material warehouse. the raw material storage must have the capacity to meet the requirements of 5 days of production for each of the three products manufactured in the facility. therefore, calculating the number of pallets in warehouse data in table 6 were obtained. table 6. number of pallets in warehouse. ta lc -fi lle d po ly pr op yl en e 30 % g la ss -fi lle d po ly pr op yl en e 12 % a b s ta lc -fi lle d po ly pr op yl en e 15 % si ng le co m po ne nt ad he si ve 12 15 9 17 6 total = 159 pallets these total number of pallets are utilized to determine the area of the warehouse. then, these data are then used to design the number of the desired column, horizontal and vertical dimension and the aisle dimension. considering the amount of quantity and the moving frequency, it is decided to use the drivethrough rack type. the design and placement of the raw material warehouse need to be designed in such a way by considering also the aisles. the total approximate area of the raw material warehouse, therefore, is 263 m2. furthermore, it is necessary to calculate the space for storing intermediate products. as we know from the production schedule all the components are produced in different shifts and need to be stored until another component of the same product has produced before assembling takes place. therefore, to store the intermediate products, the warehouse with the capacity to store the maximum number of intermediate components produced in a week has been designed nearby assembly line. components will be moved and stored in unit loads, considering that no more than 2 boxes can be stored on a single pallet, placed one over the other. wheel arches can directly be stored in the finished product warehouse as it does not have any assembling process. the maximum number of unit loads/boxes required to store into the intermediate warehouse is in table 7. table 7. max number unit loads. b um pe r b od y b um pe r c ro ss be am d as hb oa rd b od y d as hb oa rd in se rt a ir d uc t in st p an el fr am e g lo ve c om p in t g lo ve c om p ex t 64 69 84 0 0 6 11 14 total = 246 the total number of bumper body pallets to be stored in the warehouse of all the products are 64. therefore, the number of necessary columns is 64/ (5×7) = 1.83 ~ 2 columns by considering 35 shelves in a column, 5 vertically and 7 horizontally. the total number of the bumper crossbeam, dashboard body, dashboard inserts (having the same size) pallets to be stored in the warehouse of all the products are 153. therefore, the number of necessary columns is 153/ (4×7) = 5.5 ~ 6 columns by considering 28 shelves in a column, 4 vertically and 7 horizontally. the total number of the air duct, instrument panel frame, glove compartment –exterior part, glove compartment– interior part (having the int. j. prod. manag. eng. (2020) 8(1), 45-57creative commons attribution-noncommercial-noderivatives 4.0 international feasibility study of a facility to produce injection molded parts for automotive industry 51 http://creativecommons.org/licenses/by-nc-nd/4.0/ same size) pallets to be stored in the warehouse of all the products are 31. therefore, the number of necessary columns is 31/ (3×7) = 1.5 ~ 2 columns by considering 21 shelves in a column, 3 vertically and 7 horizontally. the type of storage chosen is the drive-through rack. the approximate total area needed for intermediate warehouse storage is 337 m2. the last warehouse is the warehouse for finished products. since the final products have different sizes of the box, to estimate the size, one shelf has 1.9×0.8×0.8 meters in size, so that it is possible to accommodate any product in it. the finished product warehouse must have the capacity to hold an inventory equivalent to 3 working days for each product and will be outfitted with appropriate material handling and storage equipment. finished products are arranged within boxes and they are shipped by trucks. table 8. boxes for production. production /shift 3 days no. of boxes bumper 172 1548 774 (2 bumpers/ box) dashboard 156 1404 351 (4 dashboard/ box) f whl arches 75 675 337 (2/box) r whl arches 75 675 337 (2/box) total = 1799. the warehouse consists of racks arranged in columns and each rack has 4 shelves (i.e. 4 height layers) by having 6 rows per column. the total number of boxes to be stored in the warehouse of all the products is 1799. considering there are 5 stacks in each rack the proposed design of the warehouse is shown in figure 8. the total area required is 1092 m2 approximately. figure 6. raw material warehouse. figure 7. intermediate warehouse. figure 8. finished product warehouse. 2.13. economic investment after all, the economic calculation of investment unit cost needs to be performed. table 9 shows the rough estimation of the economic investment including the main building, warehouse, material handling equipment purchase, unit load, and the machine. it is estimated that the total investment unit cost is approximately 32,105,545 €. furthermore, table 10 shows the operation cost for production. the total cost is about 16,614,502 €. it is also necessary to calculate the net present value (npv) of the industry of this solution. this value represents the sum of the present values for all of these cash-outs and cash-ins. for the calculation of npv, it is compulsory to set the selling price. from the data in the market, the bumper prices vary in a int. j. prod. manag. eng. (2020) 8(1), 45-57 creative commons attribution-noncommercial-noderivatives 4.0 international yudianto et al. 52 http://creativecommons.org/licenses/by-nc-nd/4.0/ range from 40€ to 100€, the dashboard prices vary in a range from 150€ to 250€, the wheel arch prices vary from 45€ to 85€. here, the calculation sets three different selling prices to simulate the results of the npv. table 11-13 shows the results of all three cases of selling prices. the lowest selling price scenario would give the company a profit in year 7. instead, the second pricing scenario will give the company profit in the year of 3. then, for the last pricing scenario, the company will gain profit in the second year. however, this study needs to be further analyzed to determine the best-selling price further in the business plan of the company. 3. discussions industrial facility design is highly dependent on the characteristic of the masterplan layout design (pòvoa, 2002). putting all together, it is possible to design the indoor master plan and outdoor master plan layout. however, some other considerations need to be taken into accounts, such as the building structure the positioning of each department, aisles, pedestrian path, forklift direction, industrial doors, emergency exit doors, and maneuvering area for material handling equipment. therefore, the calculated area might be slightly different from the proposed layout due to the mentioned consideration. the indoor layout masterplan, as it is depicted in figure 10, shows the complete proposed layout for the plant. as it has been mentioned in the introduction, the hatched area is the location of another production plant which is not included in this study. starting from the receiving area at the left side of the plant, the raw material is docked, checked and then moved to the raw material warehouse. the required machine to produce the component is then located next to the raw material warehouse for the sake of the material moving efficiency. the location of the machine is decided based on the most appropriate layout that has been discussed in the section machine layout. practically, due to the different sizes of the required machine and the required routings, some changes are needed concerning the building characteristic, especially the presence of pillars. here, the material moving efficiency is essential to be considered. however, it is possible to have an improvement in the further assessment of the efficiency in the material handling activity with the mathematical formulation (fu et al., 1997), which is not performed in this stage of the study. then, the intermediate warehouse is located on the right side of the main production area, followed by the finished product warehouse. however, the type of storing technique for the warehouse either raw material warehouse, intermediate warehouse and final product warehouse is extremely crucial. it will affect so much the required area and the final cost of investment. finally, the shipping area could be placed on the very right side of the plant. therefore, the flow of the production cycle is from the left side of the plant to the right side of the plant with respect to the proposed area. in addition, the placement of the plant support function and employee facilities will follow. the utility room, battery charging area, tool and maintenance room, toilets, must be connected with aisles effectively. the main entrance must be placed in the accessible area and in this case, is in the lower part of the proposed layout, nearby the cafeteria, hall, medial room and offices and also locker room. this layout allows the direction and the separation of the direct and indirect employees to go to their workplaces. the emergency exit must be placed in several places by considering the level of emergency of the plant. there are several approaches that can be implemented in the determination of a facility sitting for fire and explosion scenarios (jung et al., 2011). however, simple fixed distance measurement is implemented in this study by considering the distance of the potentially explosive machine to the location of the emergency exit. instead, figure 9 shows the proposed outdoor facilities of the plant. with respect to the location being studied, all the entrance and exit for the car and trucks are decided in towards santa cristina streets considering the size of the road. it is suggested to have different gates between the main gate and the gate for trucks. the main gate is the gate for employees and guests. after the security point, the employees park their car in front of the plant. the main entrance is exactly next to the parking area. the parking area needs to consider also the necessary space and the special places for disable. the separated parking area is also available for guests and some important people next to the main entrance. service station consisting of the gas source, the water source is placed separately from the main building. the flow of the incoming and outgoing material and products is separated from the main gate. the truck gate 1 indicates the entrance of the trailer for incoming raw material. it directs to the receiving area where the maneuvering space is also available at this point. the direction employs one way or one direction flow so that the trucks enter from truck gate 1 and exit int. j. prod. manag. eng. (2020) 8(1), 45-57creative commons attribution-noncommercial-noderivatives 4.0 international feasibility study of a facility to produce injection molded parts for automotive industry 53 http://creativecommons.org/licenses/by-nc-nd/4.0/ by truck gate 2. on the right side of the plant, there is also a maneuvering area where the shipping area is located. the area is wide enough for the trailer to park the truck. the proposed area also considers the possibility for the plant to expand the company. the open area, in this case, can be used as the expansion area for the company in the future. the outer part of the company area is separated by fences and the distance of the outer fences needs to respect the regulation of the minimum distance from the main road. 4. conclusion several points can be derived from the analysis and the proposed assessment method of this study. the design process of the study follows several steps. first of all, the definition of the input data, definition, quantity of required equipment needs to be determined. then, it is essential to have an idea about the dimensionless plant layout diagrams to have first knowledge of the plan. the machine selection based on independent research, literature research or experiment needs to be conducted. the design of the workstation also needs to consider the principal of workstation ergonomics. from the data obtained in the first step of the study, the decision of the material handling equipment needs to be considered. after that, it is possible to get o the plant layout conventional design, through sizing of production centers, aisles, support functions’ space requirements, building frame size definition, and others. after finishing the plant for indoor layout, an outdoor facility plan design is the next step to be conducted. finally, the investment evaluation via cash flow analysis needs to be performed. the main contribution of this paper relies on the clear process of initial assessment in the case of automotive sector with some examples of data which was described and proposed. however, the proposed data and calculation is the estimation of all. in practice and further study, several points could be eliminated or even be added. this study is one part of the feasibility design of the requested work. the next activity is to determine another part of the study which is the production of other automotive parts that are indicated and allocated for the hatched area of the indoor masterplan. 5. acknowledgments gratitude and appreciation are expressed to prof anna corinna cagliano and politecnico di torino for complete support and advisor for this project. table 9. investment unit cost. unit single price quantity total cost[€] steel structure €/m2 950 16200 15390000 building systems €/m2 360 16200 5832000 industrial door €each 9300 4 37200 emergency exit door €each 2800 6 16800 protection grids €/m 100 8 800 receiving/shipping dock €each 3500 2 7000 warehouse drive-in racks €unit space 70 63.84 4468.8 gravity flow racks €unit space 320 547.2 175104 material handling equipment forklift (lifting height 3-6m) €each 31200 8 249600 drum lift and tip trolley €each 1234 3 3702 hand pallet trucks €each 800 4 3200 unit load wooden pallet €/unit 689 30 20670 machine 3000t injection molding press €/unit 2100000 2 4200000 2500t injection molding press €/unit 1800000 2 3600000 1000t injection molding press €/unit 850000 2 1700000 gluing machine €/unit 300000 1 300000 welding machine €/unit 415000 1 415000 assembly bench €/unit 30000 2 60000 test bench €/unit 15000 6 90000 int. j. prod. manag. eng. (2020) 8(1), 45-57 creative commons attribution-noncommercial-noderivatives 4.0 international yudianto et al. 54 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 10. operation cost raw material total total cost [€] talc-filled polypropylene 30% 34.605 t 1330 €/t 46024 glass-filled polypropylene 12% 41.940 t 1380 €/t 57877 abs 14.985 t 2020 €/t 30269 talc-filled polypropylene 15% 37.215 t 1330 €/t 49496 single-component adhesive 1980 kg 5 €/kg 9900 # direct labor €/h h/year total cost 60 38 6075 13,851,000 manufacturing o/h head total cost utilities, plant management 3 €/h 97200 h 291,600 facilities management 4.3 €/m2 54936 m2 236,225 selling, general and administrative total cost maintenance 642,111 selling and administrative 1,400,000 table 11. case 1 selling price and npv. production revenue €/unit production total cost (€) bumper 40 132,000 5,280,000 car dashboard 150 120,000 18,000,000 wheel arch 45 116,000 5,220,000 23,280,000 year 0 year 1 year 2 year 3 year 4 year 5 year 6 year 7 cf -32,105,545 6,665,498 6,665,498 6,665,498 6,665,498 6,665,498 6,665,498 6,665,498 npv -32,105,545 -25,876,108 -20,054,204 -14,613,172 -9,528,095 -4,775,687 -334,184 2,998,565 table 12. case vselling price and npv. production revenue €/unit production total cost (€) bumper 75 132,000 9,900,000 car dashboard 200 120,000 24,000,000 wheel arch 65 116,000 7,540,000 33,900,000 year 0 year 1 year 2 year 3 year 4 year 5 year 6 year 7 cf -32,105,545 17,285,498 17,285,498 17,285,498 17,285,498 17,285,498 17,285,498 17,285,498 npv -32,105,545 -15,950,874 -853,051 13,257,065 26,444,088 38,768,410 50,286,467 58,929,216 table 13. case 3 selling price and npv. production revenue €/unit production total cost (€) bumper 100 132,000 13,200,000 car dashboard 250 120,000 30,000,000 wheel arch 85 116,000 9,860,000 43,200,000 year 0 year 1 year 2 year 3 year 4 year 5 year 6 year 7 cf -3,2105,545 26,585,498 26,585,498 26,585,498 26,585,498 26,585,498 26,585,498 26,585,498 npv -32,105,545 -7,259,285 15,961,518 37,663,204 57,945,153 76,900,246 94,615,286 107,908,035 int. j. prod. manag. eng. (2020) 8(1), 45-57creative commons attribution-noncommercial-noderivatives 4.0 international feasibility study of a facility to produce injection molded parts for automotive industry 55 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 9. outdoor layout master plan. figure 10. indoor layout master plan. int. j. prod. manag. eng. (2020) 8(1), 45-57 creative commons attribution-noncommercial-noderivatives 4.0 international yudianto et al. 56 http://creativecommons.org/licenses/by-nc-nd/4.0/ references al-aomar, r. (2006). capacity-constrained production scheduling of multiple vehicle programs in automotive pilot plant. international journal of production research, 44(13), 2573-2604. https://doi.org/10.1080/00207540500521212 cagliano, a. c., chiabert, p. (2018). plant and manufacturing system lecture notes. politecnico di torino, italy. chingua, s. nyemba, w. r., boora, k., mbohwa, c. (2019). feasibility study of the materials handling and development of a sustainable conveying system in plastics recycling and manufacture. procedia manufacturing, 33, 383-390. https://doi.org/10.1016/j.promfg.2019.04.047 derhami, s., smith, j. s., gue, k. r. (2019). space-efficient layouts for block stacking warehouse. iise transaction, 51(9), 957-971. https://doi.org/10.1080/24725854.2018.1539280 eksangsri, t., jaiwang, t. (2014). feasibility study on reuse of washed water in electronic industry: case study for flexible printed circuit board manufacturing in thailand. procedia environmental sciences, 20, 206-214. https://doi.org/10.1016/j.proenv.2014.03.027 fu, m., kaku, b. k. (1997). minimizing work-in-process and material handling in the facilities layout problem. iie transactions, 29, 29-36. https://doi.org/10.1080/07408179708966309 halil, f. m., nasir, n. m., hassan, a. a., shukur, a. s. (2016). feasibility study and economic assessment in green building projects. procedia-social and behavioral sciences, 222, 56-64. https://doi.org/10.1016/j.sbspro.2016.05.176 hazaras, m. j., swartz, c. l. e., marlin, t. e. (2013). industrial application of a continuous-time scheduling framework for process analysis and improvement. i&ec research industrial & engineering chemistry research, 53, 259-273. https://doi.org/10.1021/ie4006904 hwang, d. k., cho, k., moon, j. (2019). feasibility study on energy audit and data driven analysis procedure for building energy efficiency: bench-marking in korean hospital buildings. journal energy 12(15), 3006. https://doi.org/10.3390/en12153006 jung, s., ng, d., ovalle, c. d., roman, r, v., mannan, m. s. (2011). new approach to optimizing the facility sitting and layout for fire and explosion scenarios. i&ec research industrial & engineering chemistry research, 50, 3928-3937. https://doi.org/10.1021/ie101367g kingenberg, w., boksma, j. d. (2010). a conceptual framework for outsourcing of material handling activities in automotive: differentiation and implementation. international journal of production research, 48(16), 4877-4899. https://doi.org/10.1080/00207540903067177 kovàcs, g. (2019). layout design for efficiency improvement and cost reduction. bull. pol. ac.: tech., 67(3), 547-555. https://doi.org/10.24425%2fbpasts.2019.129653 ma, t., yang, h., lu. l., qi, r. (2017). feasibility study of developing a zero-carbon-emission green deck in hong kong. energy procedia 105, 1155-1159. https://doi.org/10.1016/j.egypro.2017.03.487 morgan, b., hejdenberg, j., krapels, s. h., amstrong, d. (2018). do feasibility studies contribute to, or avoid, waste in research? plos one 13(4), e0195951. https://doi.org/10.1371/journal.pone.0195951 pòvoa, a. p. b., (2002). optimal design and layout of industrial facilities: a simultaneous approach. ind. eng. chem. res, 41, 3601-3609. https://doi.org/10.1021/ie010660s schaller, j. (2008). incorporating cellular manufacturing into supply chain design. international journal of production research, 46(17), 4925-4945. https://doi.org/10.1080/00207540701348761 stephens, m. p., meyers, f. e., (2013). manufacturing facilities design and material handling fifth edition. purdue university press, west lafayette, indiana. sun, l., luan, f., pian, j. (2015). an effective approach for scheduling of refining process with uncertain iterations in steel-making and continuous casting process. ifac-papersonline, 48(3), 1966-1972. https://doi.org/10.1016/j.ifacol.2015.06.376 yang, z., djurdjanovic, d., ni, j. (2007). maintenance scheduling for a manufacturing system of machines with adjustable throughput. iie transactions, 39, 1111-1125. https://doi.org/10.1080/07408170701315339 int. j. prod. manag. eng. (2020) 8(1), 45-57creative commons attribution-noncommercial-noderivatives 4.0 international feasibility study of a facility to produce injection molded parts for automotive industry 57 https://doi.org/10.1080/00207540500521212 https://doi.org/10.1016/j.promfg.2019.04.047 https://doi.org/10.1080/24725854.2018.1539280 https://doi.org/10.1016/j.proenv.2014.03.027 https://doi.org/10.1080/07408179708966309 https://doi.org/10.1016/j.sbspro.2016.05.176 https://doi.org/10.1021/ie4006904 https://doi.org/10.3390/en12153006 https://doi.org/10.1021/ie101367g https://doi.org/10.1080/00207540903067177 https://doi.org/10.24425%2fbpasts.2019.129653 https://doi.org/10.1016/j.egypro.2017.03.487 https://doi.org/10.1371/journal.pone.0195951 https://doi.org/10.1021/ie010660s https://doi.org/10.1080/00207540701348761 https://doi.org/10.1016/j.ifacol.2015.06.376 https://doi.org/10.1080/07408170701315339 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2019.10745 received 2018-10-13 accepted: 2019-01-20 using system simulation to search for the optimal multi-ordering policy for perishable goods huang, y.c.a1, chang, x.y.a2, and ding, y.a.a3 adepartment of industrial management, national pingtung university of science and technology. no.1, xuefu rd., neipu township, pingtung county 912, taiwan (r.o.c.) a1 ychuang@mail.npust.edu.tw, a2 pink4141114@gmail.com, a3 yian0213@gmail.com abstract: this paper explores the possibility that perishable goods can be ordered several times in a single period after considering the cost of marginal contribution, marginal loss, shortage, and purchasing under stochastic demand. in order to determine the optimal ordering quantity to improve the traditional newsvendor and maximize the total expected profits, and then sensitivity analysis is taken to realize the influence of the parameters on total expected profits and decision variables respectively. in addition, this paper designed a multi-order computerized system with monte carlo method to solve the optimal solution under stochastic demand. based on numerical examples, this paper verified the feasibility and efficiency of the proposed model. finally, several specific conclusions are drawn for practical applications and future studies. key words: perishable goods, single-period, multi-ordering, newsvendor model, monte carlo method. 1. introduction there are many goods which are shorter period than the durable commodities in reality. as time goes by, the value of the goods will rapidly decline. this type of goods is very common in our life such as newspapers, magazines, fresh food, and milk, and so on. before the start of the sales cycle, decision maker often needs to determine how many the goods to be ordered for the entire cycle, and no more ordering before the expiry date. this type of goods will be discussed in newsvendor model. in addition, these products are called as perishable goods or seasonal goods according to their characteristics. there were many kinds of research on newsvendor problems in academic community; they discussed the inventory method, demand situations, single or twice orders and so on. in the past literature, several scholars have discussed the second order in a single period. if the first ordering quantity is sold out, there has time to the end of the period, then determine the second order should be taken or not, and proved that in some cases the expected profit of order twice is higher than order once. however, past literature did not discuss the single-period and multi-order situations, although the expiry period of perishable goods is very short, but if only ordered once before the sales cycle, and do not consider the situation that all the perishable goods was sold out before the to cite this article: huang, y.c., chang, x.y., ding, y.a. (2019). using system simulation to search for the optimal multi-ordering policy for perishable goods. international journal of production management and engineering, 7(1), 49-62. https://doi.org/10.4995/ijpme.2019.10745 int. j. prod. manag. eng. (2019) 7(1), 49-62creative commons attribution-noncommercial-noderivatives 4.0 international 49 http://creativecommons.org/licenses/by-nc-nd/4.0/ expiry period, then it may be not an optimal ordering strategy, eventually. this paper is to improve the traditional newsvendor model, and to explore whether the perishable goods should be ordered more than one time in a single period, to achieve the goal of maximizing the total expected profit. the aim of this paper is to determine whether a perishable commodity should be ordered more than once in order to maximize the total expected profit. the purposes of the paper are as follows: 1. to establish a stochastic model under singleperiod and multi-ordering. 2. proposed an optimal ordering strategy for singleperiod and multi-ordering. 3. proved the total expected profit of multi-ordering is better than the single order under stochastic demand. 2. literature review perishable goods were ordered in the case of uncertain demand to meet the needs of the sales cycle. therefore, the order should be carefully determined when ordering. there were many kinds of research on newsvendor problems which discussed the inventory method, demand situations, single, and twice orders. this paper will discuss the optimal ordering strategy for perishable commodities under single period. 2.1. order once in a single period dian (1990) derived an algorithm to determine a sequence of supply quantities which minimizes total costs of overand undersupply in the most adverse demand conditions. fujiwara et al. (1997) considered the problem of ordering and issuing policies arising in controlling finite-life-time freshmeat-carcass inventories in supermarkets. they developed a mathematical model describing actual operations and then simplify the sub-product run out period so that optimal ordering and issuing policies were easily established. the newsvendor problem is also called single-period problem (spp). khouja (1999) built taxonomy of the spp literature and delineated the contribution of the different spp extensions. khouja (2000) extended the spp to the case in which demand was pricedependent and multiple discounts with prices under the control of the newsvendor were used to sell excess inventory. they developed two algorithms for determining the optimal number of discounts under fixed discounting cost for a given ordering quantity and realization of demand. chun (2003) assumed that the customer’s demand was represented as a negative binomial distribution, and determined the optimal product price based on the demand rate, buyers’ preferences, and length of the sales period. for the case where the seller can divide the sales period into several short periods, finally proposed a multi-period pricing model. dye and ouyang (2005) extended padmanabhan and vrat’s model (1995) by proposing a timeproportional backlogging rate to make the theory more applicable in practice. alfares and elmorra (2005) extended the analysis of the distribution-free newsvendor problem to the case when shortage cost was taken into consideration. a model was presented for determining both an optimal ordering quantity and a lower bound on the profit under the worst possible distribution of the demand. chen and chen (2009) presented a newsvendor model with a simple reservation arrangement by introducing the willingness rate, represented as the function of the discount rate, into the models. and mathematical models were developed, and the solution procedure was derived for determining the optimal discount rate and the optimal ordering quantity. in addition, some scholars put forward that the idea of demand forecast updated, which focus only on the trade-off between exact requirements and additional costs, and often assuming that the supplier’s capabilities were unrestricted, but in real life is not the case. zheng et al. (2016) investigated an extension of the newsvendor model with demand forecast updating under supply constraints. in studying the manufacturer-related effects, two supply modes are investigated: supply mode a, which has a limited ordering time scale, and supply mode b, which has a decreasing maximum ordering quantity. a comparison of the different supply scenarios demonstrated the negative effects of increased purchasing cost and ordering time and quantity restrictions when demand forecast updating implemented. 2.2. order twice in a single period gallego and moon (1993) extended the analysis to the recourse case, where there was a second purchasing int. j. prod. manag. eng. (2019) 7(1), 49-62 creative commons attribution-noncommercial-noderivatives 4.0 international huang, y.c., chang, x.y. and ding, y.a. 50 http://creativecommons.org/licenses/by-nc-nd/4.0/ opportunity; to the fixed ordering cost case, where a fixed cost was charged for placing an order; to the case of random yields; and to the multi-item case, where multiple items compete for a scarce resource. azoury and miller (1984) used the concept of flexibility it was anticipated that the quantity ordered under the non-bayesian policy would be greater than or equal to that under a bayesian policy. this result was established for the n-period non depletive inventory model. lau and lau (1998) considered the very common situation in which a single-period newsvendor type product may be ordered twice during a period. they extended the basic model to consider a non-negligible set-up cost for the second order; it served as an illustration of how one might want to extend their basic two-order model to handle a large number of different combinations of additional factors such as the second-order’s delivery delay time and price differential. chung and james (2001) extended the classic newsvendor problem by introducing reactive production. production occurs in two stages, an anticipatory stage and a reactive stage. their model reduces to a single-period model with piecewiselinear convex costs. they obtain an analogue of the well-known critical fractile formula of the classic newsvendor model. pando et al. (2013) presented of the newsvendor problem where an emergency lot can be ordered to provide for a certain fraction of shortage. this fraction was described by a general backorder rate function which is non-increasing with respect to the unsatisfied demand. an exponential distribution for the demand during the selling season was assumed. an expression was obtained in a closed form for the optimal lot size and the maximum expected profit. 2.3. literature review in this paper, we explored the single-period and multi-order strategy for perishable goods. the relevant literature was summarized and shown in table 1. 3. construction of the mathematical model this paper proposed the concept of single-period and multi-order strategy for perishable goods, then developed the total expected profits model to determine the optimal ordering quantity and quantity of order. furthermore, we will prove that the multiorder is superior to the single-order for perishable goods. we will introduce the simulation method and program flow chart in section 3.7. 3.1. the assumptions of this paper 1. the model assumes no lead time. each ordering must pay the same ordering cost. if the goods sold out in this period, then the subsequent ordering quantity can be delivered before the start of next period. 2. the demand is a random variable. the marginal contribution, marginal loss, shortage cost, salvage value, and delivery costs are all known and fixed. 3. the sales quantity of each period can be known by the pos system, and the distribution of demand can be reasonably estimated by historical data and goodness-of-fit test. 4. do not consider the quantity discount and restrictions of storage space. 3.2. definitions of symbols i: the period, i = 1,2,3…n n: the number of time intervals in expired period j: the jth ordering xi: the demand quantity of i th time interval (xi is a random variable) yj: the total demand from j th ordering to the end of sales cycle (yj is a random variable). y xj i i k n j = = | coj: ordering cost of j th ordering cp: purchase cost per unit of perishable goods price: price per unit of perishable goods s: salvage value per unit of perishable goods cs: shortage cost per unit of perishable goods mp: marginal contribution, mp=price–cp, where price >cp ml: marginal loss, ml=cp–s , where cp > s qj: the ordering quantity of j th ordering (qj is a decision variable) int. j. prod. manag. eng. (2019) 7(1), 49-62creative commons attribution-noncommercial-noderivatives 4.0 international using system simulation to search for the optimal multi-ordering policy for perishable goods 51 http://creativecommons.org/licenses/by-nc-nd/4.0/ fj ( yj): probability density function of yj fj ( yj): cumulative distribution function of yj kj : the ordering time point of j th ordering mπj (qj): marginal profit under ordering quantity qj and jth ordering mrj (qj): marginal revenue under ordering quantity qj and j th ordering mcj (qj): marginal cost under ordering quantity qj and jth ordering tπj (qj): total expected profit under ordering quantity qj and j th ordering tπ1 (q1): total expected profit under ordering quantity qj and 1 st ordering tπm (q1,q2,…,qj): sum of total expected profit under ordering quantity (q1,q2,…,qj) after symbols definition, the concept of multiple orders in single period for perishable goods can be shown in figure 1. the expiry period can be divided into n, and x1,x2,x3,…,xn respectively represents the demand quantity at period 1, 2, 3…n. only the 1st ordering time point is sure, the other ordering time points k1,k2,… are uncertain. if the demand of the entire cycle can be satisfied by the first ordering quantity, then k2 will not happen. if the initial ordering quantity cannot satisfy the demand of the entire cycle, and reordering has a positive profit, then 2nd ordering will be taken and the time point is k2. the others are reasoned by analogy. figure 1. the schematic of single-period and multi-order structure for perishable goods. based on the symbol definition and figure 1, the demand of yj is y xj i i k n j = = | 3.3. ordering strategy this section describes the mathematical model of the ordering strategy. 3.3.1. ordering strategy assuming that the demand of ith period is xi, the first ordering quantity is q1, and the total demand of whole period is y1, so y xi i n 1 1 = = | the total expected profit tπ1 (q1) under single order strategy is shown in equation (1): t q y mp q y ml f y dy q mp y q c f y dy co q sq 1 1 1 1 10 1 1 1 1 1 1 1 1 1 1 ∞ 1 1 $ $ $ $ $ $ r = + ^ ^ ^ ^ ^h h h h h 6 6 @ @ # # (1) table 1. the comparison between literature and this paper. project author shortage cost salvage value order cost total expected profit maximization twice orders multiorders system simulation azoury and miller (1984)   dian (1990)   gallego and moon (1993)    fujiwara et al. (1997)   lau, h. and h. lau (1998)  khouja (2000)    chung and james (2001)   dye and ouyang (2005)  pando et al. (2013)      this paper        int. j. prod. manag. eng. (2019) 7(1), 49-62 creative commons attribution-noncommercial-noderivatives 4.0 international huang, y.c., chang, x.y. and ding, y.a. 52 http://creativecommons.org/licenses/by-nc-nd/4.0/ tπ1 (q1) can be taken a first order derivative with respect to q1 and set the result be equal to zero to obtain the optimal ordering quantity q1 that maximizes the total expected profit, as shown in equation (2): q t q ml f y dy mp c f y dy f q mp ml c mp c 0 ∂ ∂ q s q s s 1 1 1 0 1 1 1 1 1 1 1 1 ∞ 1 1 ( $ $ r = + + = = + + + ^ ^ ^ ^ ^h h h h h# # (2) q f mp ml c mp c s s 1 1 1` = + + +c m (3) after finding out the optimal ordering quantity and if tπ1 (q1) < 0, it means the expected profit is negative, then the decision maker will not make an order to purchase the perishable goods; conversely, if tπ1 (q1) ≥ 0, it means the expected profit is positive, then the decision maker will make an order to purchase the perishable goods with the optimal ordering quantity (q1). the second order derivative of the total expected profit tπ1 (q1) with respect to q1 to verify whether the tπ1 (q1) is a concave function of q1: q q t q q ml f y dy mp c f y dy ml f q mp c f q ml mp c f q 0 ∂ ∂ ∂ ∂ ∂ ∂ < q sq s s 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 ∞1 1 $ $ $ $ $ r = + + = + = + +^ e f ^ ^ ^ ^ ^ ^ ^ ^ h h o h h h h h h hp# # (4) from equation (4) know that q t q 0∂ ∂ <2 2 1 1r ^ h , so the tπ1 (q1) is the concave function of q1, therefore, q f mp ml c mp c s s 1 1 1= + + +c m is an optimal ordering quantity, and can make tπ1 (q1) have a maximum value. 3.3.2. the construction of multi-order in single period problem the multi-order means that the decision maker may deliver one or more orders during the expiry period, each ordering quantity can be denoted by q1,q2,…,qj, respectively. the total expected profit is expressed by tπm (q1,q2,…,qj), so we have , , ,t q q q t q m j j j j j 1 2 1 gr r= = ^ ^h h| the total expected profit of jth order, tπj (qj), can be expressed as shown in equation(5): t q y mp q y ml f y dy q mp y q c f y dy co j j j j j q j j j j j j sq j j j j 0 ∞ j j $ $ $ $ $ $ r = + ^ ^ ^ ^ ^h h h h h 7 7 a a # # (5) where ~ , , , , , . y f y and y x q f mp ml c mp c j j1 2 j j j j ii k n j j s s1 j ` g = = + + + = = ^ c h m | (6) |j max j t q 0 >j jr= ^ h# 3.3.3. compare single-order and multi-order multi-order in single period will occur when the first ordering quantity was sold out and second order before the end of the sales cycle is still profitable. it can be inferred that the total expected profit of multiorder will be greater than the single-order, it means tπm (q1,q2,…,qj)≥ tπ1 (q1). the proof was shown in proposition 1. proposition 1. tπm (q1,q2,…,qj)≥ tπ1 (q1) proof: , , ,t q q q t q m j j j j j 1 2 1 a gr r= = ^ ^h h| and , , , ,t q j j0 1 2 >j j 6 gr =^ h , , ,t q q q t q t q t q > m j j j j j 1 2 1 1 2 1 1 gr r r r = + = ^ ^ ^ ^ h h h h | so tπm (q1,q2,…,qj)≥ tπ1 (q1) q.e.d. (7) 3.3.4. without considering the shortage cost when we do not consider the shortage cost, the total expected profit of perishable goods in 1st ordering tπ1 (q1) was shown in equation (8): t q y mp q y ml f y dy q mp f y dy co q q 1 1 1 1 10 1 1 1 1 1 1 1 1 ∞ 1 1 $ $ $ $ $ r = + -^ ^ ^ ^h h h h6 6 @ @# # (8) int. j. prod. manag. eng. (2019) 7(1), 49-62creative commons attribution-noncommercial-noderivatives 4.0 international using system simulation to search for the optimal multi-ordering policy for perishable goods 53 http://creativecommons.org/licenses/by-nc-nd/4.0/ based on first order condition (so called foc), we have equation (9) and (10) as follows: q t q ml f y dy mp c f y dy f q mp ml mp 0 ∂ ∂ q sq 1 1 1 0 1 1 1 1 1 1 1 1 ∞ 1 1 ( $ $ r = + + = = + ^ ^ ^ ^ ^h h h h h# # (9) q f mp ml mp 1 1 1` = + a k (10) when we consider the shortage cost, the optimal ordering quantity q1 is f mp ml c mp c s s 1 1 + + +c m ; whereas, when we do not consider the shortage cost, the optimal ordering quantity q1 is f mp ml mp 1 1 + a k . if cs=0, then mp ml c mp c mp ml mp s s + + + = + , so f mp ml c mp c f mp ml mp s s 1 1 1 1 + + + = + -c am k . if cs > 0, then mp ml c mp c mp ml mp> s s + + + + , so .f mp ml c mp c f mp ml mp≥ s s 1 1 1 1 + + + + -c am k it means when the shortage cost exists, the optimal ordering quantity will increase. 3.4. goodness-of-fit test this paper collected sales data, and based on the historical data at different periods to take the goodnessof-fit test to estimate the demand distribution and its population parameters. the kolmogorov-smirnov test (k-s test) is a goodness-of-fit test. the test is a nonparametric statistical method to test the sampling data whether follows a specific theoretical distribution, such as uniform distribution, normal distribution, exponential distribution and so on. the testing steps are as follows: step 1: building a hypothesis suppose that the actual distribution function of random variable x is f(x), and the specific theoretical distribution function is given as f0(x). the hypothesis of this test is: 1. null hypothesis h0: x ~ f0(x) 2. alternative hypothesis h1: ~x0 (h1 is the supplementary set of h0 ) step 2: calculating the testing statistic let x1,x2,…,xn be a set of random sample taken from the population distribution f0(x), and let f(x) be the actual distribution function, the testing statistic ,d max f x f x x 0 6= -^ ^h h , the testing statistic d is the maximum absolute difference between the actual distribution function f(x) and the specific theoretical distribution function f0(x). step 3: rejection region if d>dα, then reject h0, where dα is a critical value of d. after goodness-of-fit test to estimate the demand distribution of and then construct the mathematical model to search for the optimal ordering strategy. 3.5. the additive property of distributions assuming that the demand distribution for each period can be estimated from past sales data through by goodness-of-fit test, and then we need to discuss whether the distribution has the property of additive. let, y x j ii k n j = =| where ~ ,x f x and x i i i i^ h ╨ ,x i j ≠j 6 and ( ) e y e x e x j ii k n ii k n ii k n j j j n = = = = = = ^ ah k| | | (11) v y v x v x j ii k n ii k n ii k n 2 j j j v = = = = = = ^ ^ ah h k| | | (12) if xi follows normal distribution, it can be denoted as ~ ,x ni i i 2n v_ i , and y xj ii k n j = =| ,then ~ ,y n j ii k ii k nn 2 j j n v= =a k|| (13) the common distributions are summarized in table 2 to justify their additive property. 3.5.1. the discussion on ordering quantity under the premise of additive property or xii k n j= | follows the central limit theorem, and if mp, ml and cs are known, then qj ≥ qj+1, it means q1 ≥ q2 ≥… ≥qj. the proof is shown in proposition 2. int. j. prod. manag. eng. (2019) 7(1), 49-62 creative commons attribution-noncommercial-noderivatives 4.0 international huang, y.c., chang, x.y. and ding, y.a. 54 http://creativecommons.org/licenses/by-nc-nd/4.0/ proposition 2. : if mp, ml and cs are known and fixed, then q1 ≥ q2 ≥…≥ qj . proof: given y xj ii k n j = =| and xi ≥0 so y y y≥ ≥ ≥ ≥ ≥ ≥ j y y y1 2 j1 2(g gn n n and v y v y v y≥ ≥ ≥ j1 2 g^ ^ ^h h h it has , and, as shown in figure 2: therefore qj≥ qj+1. by the same way, we can prove that q1 ≥ q2 ≥…≥ qj. q.e.d. figure 2. schematic of f c f c≥j j1 11+-^ ^h h . 3.5.2. the discussion on total expected profit in each period if xi is additive, and mp, ml and cs are known, and co1=co2=…=coj, then tπ1(q1) ≥ tπ2(q2) ≥ …≥ tπj(qj) ≥ 0. the proof is shown in proposition 3. proposition 3 : if , mp, ml and cs are known and co1=co2=…=coj, then tπ1(q1) ≥ tπ2(q2) ≥ …≥ tπj(qj) ≥ 0. proof: f yj ja ^ h is an increasing function of yj, and if y1>y2, then fj(y1)≥fj(y2). if tπj(qj) ≥ 0, then mπj(qj) ≥ 0, where mrj(qj) mcj(qj) so mrj(qj) ≥ mcj(qj), and mrj(qj) =mp·p(yj≥qj); mcj(qj)= ml·p(yj < qj) mp p y q ml p y q≥ ≥ <j j j j` $ $^ ^h h likewise mp p y q ml p y q≥ ≥ <j j j j1 1 1 1$ $+ + + +^ ^h h and because of yj≥yj+1 and qj≥qj+1, so t q mp p y h ml p y h≥ <j j j jh q 1 j $ $r = -=^ ^ ^h h h7 a| table 2. additive justification of common distributions. distribution f(x) xi i 1= | additive normal , ~ ,f x e x n 2 1 x i i i2 22 2 $ n v= v n^ _ ^ h i h ~ ,x n i i n i i n i i n 1 1 2 1 n v = = = d n| | |  exponential , ~ , ~f x e x exp x iid x i im m= m-^ ^h h (a special case of gamma distribution) ~ ,x gamma n 1i i n 1 a b m = = = b l|  uniform , ~ , , ~f x b a x u x iid 1 0 1 i i= -^ ^h h ,x u n0≠i i n 1= ^ h|  gamma , ~ ,f x x e x gamma x i i 1 $ $ a b a b c = a a b -^ ^ ^h h h ~ ,x gammaii n i i n 1 1 a b = = d n| |  poisson !f x x e x$m= m^ h ~x poissoni i n i i n 1 1 m = = d n| |  bernoulli f x p qx x1$= -^ h (a special case of binomial distribution) ~ ,x b n pi i n 1= ^ h|  binomial f x x n p qx n x$= -^ `h j ~ ,x b n pi i n i i n 1 1= = d n| |  geometric f x q px 1 $= -^ h (a special case of negative binomial distribution) ~ ,x nb r pi i r 1= ^ h|  int. j. prod. manag. eng. (2019) 7(1), 49-62creative commons attribution-noncommercial-noderivatives 4.0 international using system simulation to search for the optimal multi-ordering policy for perishable goods 55 http://creativecommons.org/licenses/by-nc-nd/4.0/ t q mp p y h ml p y h≥ < j j j jh q 1 1 1 11 j 1 $ $ r = + + + += + ^ ^ ^ h h h7 a| also qj≥ qj+1, and f y f y<j j j j1 % % +_ _i i , so f y f y1 1>j j j j1% % +_ _ _ _ii ii7 a . in other words, p y y p y y> > >j j j j1 % % +_ _i i . so t q t q≥j j j j1 1r r + +^ ^h h . q.e.d. 3.6. sensitivity analysis the sensitivity analysis is taken to realize the influences of the system parameters on total expected profit are shown as follows. 1. the influence of marginal contribution (mp) on total expected profit (tπj(qj)) has a same changing direction. mp t q y f y dy q f y dy 0 ∂ ∂ > j j j j j j q j j j jq0 ∞j j $ $ r = + ^ ^ ^ h h h# # (14) 2 the influence of marginal loss (ml) on total expected profit (tπj (qj)) has an opposite changing direction. ml t q q y f y dy 0∂ ∂ < j j j j j j j q 0 j $ r = ^ ^ ^ h h h# (15) 3 the influence of shortage cost (cs) on total expected profit (tπj (qj)) has an opposite changing direction. c t q y q f y dy 0∂ ∂ < s j j j j j j jq ∞ j $ r = ^ ^ ^ h h h# (16) 4 the influence of delivery cost (coj) on total expected profit (tπj (qj)) has an opposite changing direction. co t q 1 0∂ ∂ < j j jr = ^ h (17) 3.7. system simulation the monte carlo simulation will be applied and introduced as follows. 3.7.1. monte carlo simulation monte carlo simulation is a simulation; it can generate random numbers that follow a specific probability distribution. based on the random numbers and given mathematical model to find out the optimal solution that maximizes the total expected profit or minimizes the total expected cost. in this study, the monte carlo method was applied to simulate the demand of each period. after collecting the past sales data and building the demand distribution of each period by goodness-of-fit test, using a random number generator to create a random number between 0 and 1, and let it denote fx(x). applying the inverse function of fx(x) to find out the value of random variable that follows a specific distribution. the steps of monte carlo simulation are as follows: step 1: collecting historical sales data. step 2: using goodness-of-fit test to estimate the population’s parameters and demand distribution of each period. step 3: using random number generator to create a random number (u) between 0 and 1, and u~uniform(0,1). step 4: finding the cumulative distribution function of the demand distribution (f(x)). step 5: let u= f(x) step 6: x= f-1(u). step 7: repeat step 4 to 6 until the required random numbers are satisfied. 3.7.2. the relationship between system simulation and uniform distribution a random variable u is generated, and u~u(0,1), then let fx(x)=u, therefore x= fx -1(u). if x~fx , where fx is a cumulative distribution function (c.d.f ) of x. in other words, a random number u can be obtained from the random number generator, where u has a uniform distribution between 0 and 1, and then given x~fx and let fx(x)=u can be used to obtain a mapping value of random variable (x). the proof is shown in property4. property 4. : given x~fx and . .x c r v ! (continuous random variable), let fx(x)=u, where u~u(0,1), then x= fx -1(u). int. j. prod. manag. eng. (2019) 7(1), 49-62 creative commons attribution-noncommercial-noderivatives 4.0 international huang, y.c., chang, x.y. and ding, y.a. 56 http://creativecommons.org/licenses/by-nc-nd/4.0/ proof: let u~u(0,1), then f u p u u dt u1≤ u u 0 $= = =^ ^h h # , ,u 0 1! 6 @ if fx(x)=u, then f u p u u p f x u p f f x f u p x f u f f u u ≤ ≤ ≤ ≤ u x x x x x x x 1 1 1 1 $ = = = = = = ^ ^ ^ ^ ^ ^ ^ h h h h h h h 6 7 7 7 a a @ a and . q.e.d. p x x p f u x p f f u f x p u f x f f x f x ≤ ≤ ≤ ≤ x x x x x u x x 1 1 = = = = = ^ ^ ^ ^ ^ ^ ^ h h h h h h h 7 7 6 6 7 @ @ a a a since fx is a non-decreasing function of x, it means if a > b, then fx(a) ≥ fx(b), and if . .x c r v ! , then fx(a) > fx(b). 3.8. the flow chart of the proposed system simulation this paper uses the visual basic software to develop a multi-ordering computerized system; the system flow chart is shown in figure 3. 4. example analysis this chapter will base on the statistical analysis described as above to search for the optimal ordering strategy under single-period and multi-ordering situations. at first, describes the problem and then put the data into simulation system to find out the optimal ordering quantity and total expected profit, then analysis and discuss the simulation results. finally, sensitivity analysis is carried out to verify the feasibility and correctness of the proposed model. 4.1. example description suppose there is a convenience store sells monthly magazine, and its price is $ 120 at cost $ 60. if it is not sold after the end of the sales cycle, it will only be worth $ 1 sold to the recycling dealer. considering the shortage cost is equal to the marginal contribution of the magazine, and each ordering and delivery cost is 50, and assuming that no lead time, when the expected profit of each ordering is 0 will also carry out an order to satisfy the customer’s need. the sales period of the magazine is 30 days and divided into 3 periods, so each period is 10 days. we collected sales data over the past years and took the goodnessof-fit test to estimate the demand distribution of each period. we found that the demand distribution of each period is a normal distribution, which is xi~n(µi,σi 2). the influences of the mean and variance of three periods on the number of orders, the optimal ordering quantity and the total expected profit is discussed. therefore, the mean and variance are classified as large, medium and small, respectively. the large, medium and small of mean were 10, 20 and 30; the large, medium and small of standard deviation were ,3 1 6 1 9 1, and i i in n n . therefore, there are 729 ((3×3)3) combinations. the random demand (xi) for period i is calculated by the monte carlo simulation method. each experiment is repeated 1000 times. the model proposed in this paper does not limit the ordering quantity, as long as tπj (qj)≥0, the order will be delivered. in this example, there are three possible ordering time points: the first time point is at the beginning of the sales period to meet the demand of entire period; the second time point is at the beginning of period 2 when the magazine was sold out in period 1, and reorder to meet the needs of period 2 and 3; the third time point is at the beginning of period 3 when the magazine was sold out in period 2, and then reorder to meet the need of time3. the purpose of this paper is to decide the optimal multi-ordering policy under stochastic demand to maximize the total expected profit. 4.2. general situation putting the values of mp, ml and cs into equation (6) to find out the optimum ordering quantity qj, and calculate the total expected profit tπj (qj) by equation (5). if tπj (qj)≥0, then takes an order and the ordering quantity is qj; if tπj (qj)<0, then do not take an order. 4.2.1. analysis of single data we now randomly select the combination no. 8 which ordering twice in a sales period (it has three periods) to explain. the mean demand of the period 1 is 30 and its standard deviation is 10, the mean demand of period 2 is 30 and its standard deviation int. j. prod. manag. eng. (2019) 7(1), 49-62creative commons attribution-noncommercial-noderivatives 4.0 international using system simulation to search for the optimal multi-ordering policy for perishable goods 57 http://creativecommons.org/licenses/by-nc-nd/4.0/ is 10 and the mean demand of time interval 3 is 10 and its standard deviation is 1.67. the data is shown in table 3. according to table 3, it can be found that there happened 145 times of twice ordering in 1000 experiments. if order occurs twice, the second ordering quantity will be less than the first ordering quantity (11 < 76). therefore, the proposition 2 was verified. when the second order occurs in combination no. 8, the final total expected profit will be greater than which is only order once (4493.97 > 3116.16), so the proposition 1 is verified. 4.2.2. analysis of order twice data according to the simulation results where each combination was performed 1000 times. we show partial results of order twice in table 4. according to table 4, it can be found that if order occurs twice, the second ordering quantity will be less than the first ordering quantity (q2 < q1). therefore, the proposition 2 is verified. in addition the final total expected profit of order twice will be greater than which is only order once (tπm (q1,q2) > tπ1 (q1)), so the proposition 1 is verified. in general, if q1 is less than or close to (µ1+3σ1), then it has the opportunity to order twice and the time of second order is at the end of period 1; when q1 is less than or close to µ1+(µ2+3σ2) or close to (µ1+3σ1)+µ2, it has the opportunity to order twice and the time of second order is at the end of the period 2. it was known from the examples that ordering twice is likely to occur in (µ1 ≥ µ2 ≥ µ3) or (µ2 ≥ µ1 and µ2 ≥ µ3) condition. figure 3. system flow chart. int. j. prod. manag. eng. (2019) 7(1), 49-62 creative commons attribution-noncommercial-noderivatives 4.0 international huang, y.c., chang, x.y. and ding, y.a. 58 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.3. shortage cost under the other parameters are fixed, we will discuss the magnitude of shortage cost that influences the optimal ordering strategy and total expected profit, simultaneously. there are three kinds of situations need to consider: (1) thinking of the shortage cost is as the opportunity cost, it means that cs=mp; (2) thinking of the shortage cost is as the opportunity cost plus customer run off cost, it means that cs > mp; (3) thinking of the shortage costs is as fictitious loss, it means that cs=0. applying the simulation system developed in this paper, the results are obtained and shown in figure 4. according to figure 4, it can be found that the ordering quantity will be increased when shortage cost rises. those results just verify the inference in section 3.3.4. 4.4. order three times’ conditions based on section 3.3.4, we knew that the ordering quantity of considering the shortage cost is greater than the one of do not consider. under do not consider the shortage cost, it can be found that ordering more than one time would occur in some particular combinations, and those results also proved that multi-ordering policy for perishable goods in expiry period (which can be divided into several periods) is worthy. the combinations of order three times are shown in table 4. according to we found that the situation of order three times is likely to occur only once in 1000 table 4 times random simulations under some specific combinations. usually it occurs at the mean and variation of period 1 are large, and the mean of period 3 is small. from table 4 we can find the ordering quantity is decreasing each time, it means q1 > q2 > q3. therefore, the proposition 2 was verified. in addition, the total expected profit is shown in figure 5. reorder conditions are based on tπj (qj)≥0. therefore, that can be known the total expected profits will increase when the order number is rising, so the proposition 1 was verified. when we do not consider shortage cost (it means cs=0) and then execute 1000 times simulations for each combination. it can be found that when table 3. total expected profit and ordering quantity of combination no. 8. number time 1 time 2 time 3 j=1 j=2 µ1 σ1 µ2 σ2 µ3 σ3 total times of ordering tπ1(q1) total average q1 total times of ordering tπm(q1 ,q2) total average q1 +q2 8 30 10 30 10 10 1.7 855 3,116.16 76 145 4,493.97 87 cs=0 cs=mp cs>mp no.246 order once total order quantity average 80 86 87 no.246 order twice total order quantity average 110 117 119 no.333 order once total order quantity average 60 65 66 no.333 order twice total order quantity average 70 75 77 80 86 87 110 117 119 60 65 66 70 75 77 55 85 115total ordering quantity figure 4. the effects of various shortage costs on ordering quantity. int. j. prod. manag. eng. (2019) 7(1), 49-62creative commons attribution-noncommercial-noderivatives 4.0 international using system simulation to search for the optimal multi-ordering policy for perishable goods 59 http://creativecommons.org/licenses/by-nc-nd/4.0/ (µ1 ≥ µ2 ≥ µ3) and (µ1, σ1, σ2) are very large, furthermore, q1 is less than or close to (µ1+3σ1) and q2 is less than or close to (µ2+3σ2), order three times situations will be happened. 4.5. sensitivity analysis the influences of the system parameters on total expected profit are shown as follows. according to table 5, it can be observed that tπj (qj) will increase when mp is rising. it showed that mp and tπj (qj) has a positive correlation. therefore, equation (14) was verified. it can be observed that tπj (qj) will decrease when ml is rising. it showed that ml and tπj (qj) has a negative correlation. therefore, equation (15) was verified. it can be observed that tπj (qj) will decrease when cs is rising. it showed that cs and tπj (qj) has a negative correlation. therefore, equation (16) was verified. it can be observed that tπj (qj) will decrease when coj is rising. it showed that coj and tπj (qj) has a negative correlation. therefore, equation (17) was verified. 0,0 1.500,0 3.000,0 4.500,0 6.000,0 7.500,0 7 8 34 35 36 61 62 order once in thousand times total expected profit averages 3.257,5 3.250,2 2.806,1 2.804 2.830,7 2.326,4 2.332,1 order twice in thousand times total expected profit averages 4.541,9 4.631,7 3.965,9 4.016,4 4.050,3 3.413,8 3.495,7 order thrice in thousand times total expected profit averages 6.693 7.050 5.493 5.850 5.731 4.650 4.650 total expected profit figure 5. total expected profit under three times ordering. table 4. total expected profit and ordering quantity for order twice and three times. no. i=1 i=2 i=3 j=1 j=2 j=3 µ1 σ1 µ2 σ2 µ3 σ3 total times of ordering tπ1(q1) total average q1 average total times of ordering tπm(q1,q2) total average q1+q2 average total times of ordering tπm(q1,q2,q3) total average q1+q2+q3 average 7 30 10 30 10 10 3.3 870 3,092.94 76 130 4,466.06 76+11.3 49 30 10 20 2.2 20 6.7 981 3,375.84 75 19 4,891.47 75+23 89 30 5 30 10 10 1.7 899 3,396.94 75 101 4,588.55 75+11 99 30 5 30 5 10 1.1 961 3,692.56 73 39 4,683.49 73+10 116 30 5 20 6.7 10 1.7 940 2,999.72 64 60 4,008.30 64+11 134 30 5 20 2.2 10 1.7 993 3,194.83 63 7 3,896.00 63+11 269 20 6.7 30 3.3 10 1.7 951 3,061.81 63 49 4,046.29 63+11 656 10 1.1 30 10 10 1.7 919 2,304.68 55 81 3,481.16 55+11 7 30 10 30 10 10 3.3 734 3,257.5 70 265 4,541.9 70+10.1 1 6,693 70+40+10 8 30 10 30 10 10 1.7 743 3,250.2 70 256 4,631.7 70+10 1 7,050 70+40+10 34 30 10 20 6.7 10 3.3 796 2,806.1 60 203 3,965.9 60+10.3 1 5,493 60+30+10 35 30 10 20 6.7 10 1.7 797 2,804 60 202 4,016.4 60+10.1 1 5,850 60+30+10 36 30 10 20 6.7 10 1.1 791 2,830.7 60 208 4,050.3 60+10.1 1 5,731 60+30+10 61 30 10 10 3.3 10 3.3 806 2,326.4 50 193 3,413.8 50+11.3 1 4,650 50+20+10 62 30 10 10 3.3 10 1.7 811 2,332.1 50 188 3,495.7 50+11.4 1 4,650 50+20+10 int. j. prod. manag. eng. (2019) 7(1), 49-62 creative commons attribution-noncommercial-noderivatives 4.0 international huang, y.c., chang, x.y. and ding, y.a. 60 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5. conclusions this paper establishes a single-period and multiordering mathematical model to revise the traditional newsvendor model and based on numerical examples to verify its feasibility and profitability. the purpose of this paper is to modify the traditional newsvendor model from single-order to multiorder to maximize the total expected profit. with consideration of marginal contribution, marginal loss, and shortage cost, the total expected profit for multiple orders will be better than for single order, and the amount of each order placed under multiple orders and its corresponding expected profit will gradually decrease. based on numerical examples, the perishable goods will be ordered three times only in few cases. the most order times is once and twice, and as long as order times is more than once, the total expected profit will increase when the times of ordering is increasing. in this paper, monte carlo method is used to simulate stochastic demand in each period, and we also designed a computerized system to search for the optimal multi-ordering strategy to maximize the total expected profit. finally, numerical examples are proposed to demonstrate the effectiveness and feasibility of the proposed model. finally, this paper only studies a single perishable goods, it can be studied for multi-perishable goods in the future. the model can be added in different limiting factors such as space or budget. references azoury, k.s., miller, b.l. (1984). a comparison of the optimal ordering levels of bayesian and non-bayesian inventory models, management science, 30(8), 993-1003. https://doi.org/10.1287/mnsc.30.8.993 alfares, h.k., elmorra, h.h. (2005). the distribution-free newsboy problem: extensions to the shortage penalty case, international journal of production economics, 93-94(8), 465-477. https://doi.org/10.1016/j.ijpe.2004.06.043 chung, c.s., flynn, j. (2001). a newsboy problem with reactive production, computers & operations research, 28(8), 751-765. https://doi.org/10.1016/s0305-0548(00)00006-x chun, y.h. (2003). optimal pricing and ordering policies for perishable commodities, european journal of operational research, 144(1), 68-82. https://doi.org/10.1016/s0377-2217(01)00351-4 chen, l.h., chen, y.c. (2009). a newsboy problem with a simple reservation arrangement. computers & industrial engineering, 56(1), 157160. https://doi.org/10.1016/j.cie.2008.04.010 dian, j.r. (1990). a high-low search algorithm for a newsboy problem with delayed information feedback. operations research, 38(5), 838-846. https://doi.org/10.1287/opre.38.5.838 dye, c.y., ouyang, l.y. (2005). an eoq model for perishable items under stock-dependent selling rate and time-dependent partial backlogging. european journal of operational research, 163(3), 776-783. https://doi.org/10.1016/j.ejor.2003.09.027 fujiwara, o., soewandi, h., sedarage, d. (1997). an optimal ordering and issuing policy for a two-stage inventory system for perishable products. european journal of operational research, 99(2), 412-424. https://doi.org/10.1016/s0377-2217(95)00365-7 table 5. the influence of mp, ml, cs, coj on q1 and tπ1(q1). number 183 (µ1=30, σ1=3.33, µ2=30, σ2=3.33, µ3=30, σ3=3.33) 547 (µ1=10, σ1=3.33, µ2=10, σ2=3.33, µ3=10, σ3=3.33) q1 tπ1(q1) q1 tπ1(q1) mp 60 93 4,978.4 33 1,363.4 70 93 5,868.8 33 1,655.4 80 93 6763.8 33 1943.4 ml 49 93 5,002.9 33 1,417.1 59 93 4,978.4 33 1,363.4 69 92 4,942.1 32 1,339.2 cs 0 90 5,076.1 30 1,474.5 60 93 4,978.4 33 1,363.4 80 93 4,953.7 33 1,351 coj 0 93 5,016.9 33 1,423 50 93 4,978.4 33 1,363.4 100 93 4915.5 33 1332.1 int. j. prod. manag. eng. (2019) 7(1), 49-62creative commons attribution-noncommercial-noderivatives 4.0 international using system simulation to search for the optimal multi-ordering policy for perishable goods 61 https://doi.org/10.1287/mnsc.30.8.993 https://doi.org/10.1016/j.ijpe.2004.06.043 https://doi.org/10.1016/s0305-0548(00)00006-x https://doi.org/10.1016/s0377-2217(01)00351-4 https://doi.org/10.1016/j.cie.2008.04.010 https://doi.org/10.1287/opre.38.5.838 https://doi.org/10.1016/j.ejor.2003.09.027 https://doi.org/10.1016/s0377-2217(95)00365-7 http://creativecommons.org/licenses/by-nc-nd/4.0/ gallego, g., moon, i. (1993). the distribution free newsboy problem: review and extensions. journal of the operational research society, 44(8), 825-834. https://doi.org/10.1057/jors.1993.141 khouja, m. (1999). the single-period (news-vendor) problem: literature review and suggestions for future research. omega, 27(5), 537553. https://doi.org/10.1016/s0305-0483(99)00017-1 khouja, m.j. (2000). optimal ordering, discounting, and pricing in the single-period problem. international journal of production economics, 65(2), 201-216. https://doi.org/10.1016/s0925-5273(99)00027-4 lau, h., lau, h. (1998). decision models for single-period products with two ordering opportunities. international journal of production economics, 55(1), 57-70. https://doi.org/10.1016/s0925-5273(98)00040-1 padmanabhan, g., vrat, p. (1995). eoq models for perishable items under stock dependent selling rate. european journal of operational research, 86(2), 281-292. https://doi.org/10.1016/0377-2217(94)00103-j pando, v., luis, a.s.j., juan, g.l., sicilia, j. (2013). a newsboy problem with an emergency order under a general backorder rate function. omega, 41(6), 1020-1028. https://doi.org/10.1016/j.omega.2013.01.003 zheng, m., wu, k., shu, y. (2016). newsvendor problems with demand forecast updating and supply constraints. computers & operations research, 67, 193-206. https://doi.org/10.1016/j.cor.2015.10.007 int. j. prod. manag. eng. (2019) 7(1), 49-62 creative commons attribution-noncommercial-noderivatives 4.0 international huang, y.c., chang, x.y. and ding, y.a. 62 https://doi.org/10.1057/jors.1993.141 https://doi.org/10.1016/s0305-0483(99)00017-1 https://doi.org/10.1016/s0925-5273(99)00027-4 https://doi.org/10.1016/s0925-5273(98)00040-1 https://doi.org/10.1016/0377-2217(94)00103-j https://doi.org/10.1016/j.omega.2013.01.003 https://doi.org/10.1016/j.cor.2015.10.007 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.14561 received: 2020-11-03 accepted: 2021-07-24 lean six sigma implementation, a systematic literature review tampubolon, s.a1* , purba, h.h.a2 a industrial engineering department, mercu buana university, jln. meruya selatan no 1, kembangan, jakarta 11650, indonesia. a1 nomlas26@yahoo.com, a2 humiras.hardi@mercubuana.ac.id abstract: organizations must be able to meet customer needs in today’s complex market situation and business environment, the needs and essentials for their satisfaction such as high product quality, competitive costs and faster delivery. organization need to apply a comprehensive concept and method on managing this requirement. this systematic review intends to identify how lean six sigma implementation in many industries. lean six sigma (lss) is a method that has been widely used in research in various fields and continues to grow, to get the most common solution it is necessary to review the method. this research is to observe concept and method still relevant to be use and effectively improved the business performance and customer satisfaction. for the identity of the lss papers, a total of 50 research papers were reviewed which met the criteria, research object, country of research and year of publication and result of research. the result show that lss is still being used and successfuly help the organization to improve their competitiveness, improve quality, reduce costs, increase customer satisfaction, increase productivity, and increase employee morale. key words: lean manufacturing, dmaic, lean-six sigma, six-sigma, vsm. 1. introduction all sectors open to global competition have equal opportunities from various organizations around the world. in an effort to respond to the pressure created by global competition from all sectors, opportunities for organizations from different parts of the world have equal opportunities in global competition so that companies have to adopt competitive and innovative methods, which in most cases, tend to emphasize quality and focus on customers. apart from focusing on improving quality, products, services and processes, most of these approaches also focus on customer satisfaction. the use of quality methods is increasing. lean manufacturing technique is a concept adopted to eliminate waste and processes that do not add any value to customer satisfaction. it also aims to increase the efficiency and effectiveness of the company. meanwhile, the six-sigma method is needed to reduce process variability. motorola was one of the companies that was successful in adopting the six-sigma method in the 1980s, in an effort to increase the level of quality by reducing variability in manufacturing operations in a continuous and consistent manner (olanrewaju et al., 2019). six-gima dedicates what customers want from a product of the highest quality. on the other hand, lean manufacturing in particular is focused on reducing waste and non-adding value for customer satisfaction. now lean six sigma (lss) has become a leading business improvement methodology that has been successfully implemented since being implemented in all types of businesses. the goal of lss is to drive to cite this article: tampubolon, s., purba, h.h. (2021). lean six sigma implementation, a systematic literature review. international journal of production management and engineering, 9(2), 125-139. https://doi.org/10.4995/ijpme.2021.14561 int. j. prod. manag. eng. (2021) 9(2), 125-139creative commons attribution-noncommercial-noderivatives 4.0 international 125 https://orcid.org/0000-0002-2885-5946 http://creativecommons.org/licenses/by-nc-nd/4.0/ business improvement with the key features of lean and six sigma and incorporate these features into an integrated approach towards improving business performance. six-sigma focuses on eliminating critical to quality (ctq) issues affecting business organizations, while companies focus on systematically creating value and reducing waste (thomas et al., 2016). 1.1. the principles of lean when done correctly, lean can create huge impact of business in efficiency, cycle time, lower costs and improved competitiveness. eliminate waste –the non-value-added components in any process is the goal of lean–. and remember, lean isn’t restricted to manufacturing. inventory management, and even client interaction, lean can improve how a team works together. according to researchers (sibalija & vidosav, 2014) that the basic principles of making lean are: (a) value flow, where the value flow for each product must be identified, implying both value added and non-value added activities. value flow is the sequence of events that flow from concept to delivery, adding value to customer expectations. lean overall purpose is to eliminate activities that do not create value. there are seven known wastes in manufacturing, one of which is waste which can be categorized into nonadded value. on average, less than 1% of activity is value added, and typically, resources are focused on increasing 1% and ignoring 99% of opportunities, (b) value, where the value for a particular product must be determined precisely from the point of view of the customer who uses it (c) pull: the customer will actually draw value from the producer. in the production chain, when the customer last pulls the product, each production process triggers to produce the product simultaneously, (d) flow: flow value is determined by customer requirements and created without having to be interrupted. flow can be understood as a continuous movement of product, which supports the flow of one part and work cell over a production line. in determining the steps for value creation it is necessary to pay attention to the strict order so that the product flows smoothly towards the customer. (e) perfection: perfection of production must be continuously attempt step by step. since the determined value of the value flow activity is identified, wasted steps are eliminated, and flow and pull are introduced, the continuous improvement process is continued until the state of perfection is that no waste is made into a perfect value. muda is the japanese term label for value creation through the elimination of wasteful activity, and is a basic lean principle called “toyota’s seven wastes” as identified by ohno (1988) in describing the toyota production system (tps) in the general category: over-processing; overproduction; transport; unnecessary / excessive movements; waiting; defects; unnecessary inventory. 1.2. the concept of lean make it earlier with no longer time and valuable with no waste generation is the lean works on operation format. simply performing better productivity in eliminating waste from the manufacturing process is the concept of lean manufacturing. maximize value to the customer while using a few resources as possible is the aims of lean. so any activity that does not add value from a customer’s point of view is waste (sabale & thorat, 2019). 1.3. the eight wastes of lean production listed below are eight wastes, or processes and resources, that add no value to customers: (1) overproduction of the product, (2) waiting, perhaps people waiting or equipment that is not needed, (3) investing more time into a product than what the customer needs, such as a design that requires high-tech machinery or a lot of costs for features unnecessary, (4) excess inventory, (5) unnecessary transportation, (6) unnecessary movement of people, equipment or machines, (7) unused waste of talent and ingenuity, (8) disabilities, which require more a lot of work and costs for corrections. 1.4. the relationship between six-sigma and lean principles according to researchers tohidi (2012), process improvement efforts seek to correct problems by eliminating the causes of variations in the process while leaving the basic processes intact. process improvement refers to the strategy of finding solutions to eliminate the root causes of performance problems in existing processes in our company. in six sigma terms, the process improvement team finds the critical x (cause) that creates the unwanted y (defect) generated by the process. 5s is a set of techniques. they are used to improve workplace practices that facilitate visual control and lean implementation. 5s is the basis for continuous improvement, zero defects, cost reduction, and a int. j. prod. manag. eng. (2021) 9(2), 125-139 creative commons attribution-noncommercial-noderivatives 4.0 international tampubolon and purba 126 http://creativecommons.org/licenses/by-nc-nd/4.0/ safe work area and is a systematic way to improve workplaces, processes and products through the involvement of production line employees. dmaic (define, measure, analyze, improve, and control) is a structured problem solving methodology that is widely used in business. these phases lead the team logically from defining the problem through implementation solutions related to the underlying causes, and establishing best practices to ensure solutions stay in place. the relationship between the six sigma model and lean principles is shown in figure 1. figure 1. relationship between six sigma models and lean principles. 1. the first step in lean starts from identifying problems that are directly obtained and understood from managerial descent to the production line, while in six-sigma the define step can be sourced from the production line, customer complaints or top managerial wish plans. 2. the second step is to carry out value stream mapping, which is to identify the time required for each work process with a focus on added value to customers, this is the same as measuring the dependent variable (y) at the measurement stage and measuring the independent variable (x) at the analysis stage. 3. flow and pull steps, is an improvement step that must be done. flow value is determined by customer requirements. in the production chain, when the customer last pulls the product, each production process triggers to produce the product simultaneously. 4. perfect step, is a step of stability of the improved process that we monitor at all times during the control phase. benefits of lean-six-sigma (lss) the following are the benefits of lean-six-sigma (lss): (olanrewaju et al., 2019). 1. the process speed and output are uniform, so cycle times can be reduced. 2. efforts to continuously increase productivity. 3. reduction of defective products that are in the process of production. 4. efforts to save space and reduce costs. 1.5. obstacles and difficulties implementing lean six-sigma based on the research of albliwi et al, (2013) lack of top management commitment and involvement, lack of communication, lack of training and education, limited resources and others are some common factors for failure in term of lss implementation. pereira et al., (2017) only 10% or less of the companies trying to implement lean manufacturing practices are successful in understanding and implementing the philosophy in their processes. lean production systems are quite difficult to implement. there are many difficulties for it on the side of the operational and supporting framework and, even today, they cause a number of problems, confusion and controversy. even in companies starting a lean implementation process, the difficulty is just as great. implementation of lean manufacturing is not an easy task, and there have been several issues and issues reported about the difficulties the company faced during implementation. the understanding and willingness of top management to make a very large investment so that this method can work is also an important obstacle. extensive knowledge of lean and six-sigma based on the implementation of statistical science is a must for the success of lean six-sigma. in addition to implementing lss in the service industry, it is very difficult to do this because the basic reference of lean and six-sigma is the manufacturing industry. int. j. prod. manag. eng. (2021) 9(2), 125-139creative commons attribution-noncommercial-noderivatives 4.0 international lean six sigma implementation, a systematic literature review 127 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. methods of research by reviewing the previously published literature this magazine addresses the most commonly resolved topics with the lean six-sigma (lss) method approach in many sector. this research was initiated by obtaining scientific journals through academic journals in the field of lss which were published in leading journal database. the research articles that are the researcher’s concentration are related to lss and are carried out in many sector. the articles are reviewed starting from articles published since 2009 ~ 2020, according to predetermined criteria, then searches can be carried out. the journals we reviewed come from many countries such as india, portuguese, italy, sweden, brazilia, serbia, usa, uk, malaysia, jordan, norwegian, indonesia, south afrika, mexico, poland, egypt, philipines, purdue, australia, ireland, thailand, croatia, irag, lebanon, russia, slovak & bangladesh. by using keywords, paper submissions can be determined. figure 2. study of framework. 3. result & discussion let us explain the findings from several journals that we have reviewed in terms of the object of research. a systematic literature review is conducted to identify relevant opportunities for the introduction and development of a successful lean six sigma approach in many kinds of industries. there are 50 papers that have been completed in-depth review. consideration of the object of research and the results of each paper are evaluated. table 1 shows the complete list of reviewed papers. there are several papers that have successfully implemented lss in several organizational sectors. manufacturing: ghaziabad, from india (2012) applied lss in manufacturing company by a). capabilities for wider widths hard alloys rolling was developed in hot rolling mill and b) in the hot mill process to eliminate the downtime due to strip / coil slippage during the 5xxx hard alloy winding, measurements of all parameters in six sigma are carried out. from this activity it can be concluded that, the practice of lss helps the company to remain in the market with a relatively small investment for a new and more powerful hot mill. in portuguese, pereira et al., (2019) there are many areas of organizations in the mold industry, which can be said to have had great success when they made use of optimization tools. the event that contributed the most to the end can be done with pareto analysis. increased study of the mold industry is obtained from the study of cnc machines with the approach of applying lean six sigma tools. in sweden, schön et al., (2010) did a research in three company. the surveys that are distributed to these companies are the result of research conducted. in many aspects of job satisfaction experienced by workers after following the lss towards positive changes, these conclusions can be concluded through the survey results found. in malaysia, jirasukprasert et al., (2014) conducted an investigation into a rubber company to find the factors causing defects. after optimizing two production process variables, which in turn can help the organization learn to reduce defects, a reduction of about 50 percent in glove leakage defects can be achieved. an approach through six sigma and a streamlined problem solving methodology that can effectively improve the glove-making process by reducing the number of defects is the conclusion that can be given. indrawati & ridwansyah, (2015) from indonesia, measure six-sigma level in an iron ores industry. iron ore drying production is not optimal due to its low production efficiency, general and influential waste products are an indication that the processing is improper and produces defects. in this research, to increase the capability of the manufacturing int. j. prod. manag. eng. (2021) 9(2), 125-139 creative commons attribution-noncommercial-noderivatives 4.0 international tampubolon and purba 128 http://creativecommons.org/licenses/by-nc-nd/4.0/ process, it is carried out using the lean six-sigma method, the proposed improvement program is develop to overcome the problems through lean six sigma consistency. lss has succeeded in improving organizational performance and its competitive advantage, the conclusions of several researchers who have conducted a comprehensive review and evaluation of the implementation of lss in the organization. researchers taylor et al., (2011) from the usa made observations about dominant lean versus dominant six sigma in an article published in the international journal of lean six sigma. finally, they recommend that lean is dominant and consists of two subordinate methods six sigma. and statistical process control can make significant progress to the organization. healthcare: barnabé et al. (2016) did the observation in healthcare institutions in italy. various results can be expected. first, increased efficiency as a result of the anticipated improvements in professional performance. the key lss factors are employee empowerment, communication, institutional leadership, and personnel training, as the conclusion of the project. arcidiacono & pieroni, (2018) another researcher in the same object of observation in term of healthcare in italy. by improving the quality of experience felt by patients while reducing healthcare costs, this is the conclusion of the application of lean six sigma 4.0. oil & gas: nascimento et al. (2019) did observation oil and gas company in brazilia to explore the synergy between six sigma principles and lean production (lp). for problem solving through continuous and gradual improvement, until they conclude that lss is developed taking into account the integration of six sigma and lean production (lp) will provide a systemic and holistic approach. multi object: mvsit et al. (2016) from india, did measurement to evaluate enabler ratings is the basic framework of ahp. albliwi, et al, (2013) find out there are 34 critical failure factors for the spread of lss. garment: rodrigues nogueira et al. (2017) from india, did measurement defect in textile industry by using combining the variability reduction tools and techniques of six sigma is a function of lean sigma. winning customer loyalty and improving bottom-line results, dmaic is a proposed framework integrating lean tools in the six sigma methodology. following lean six sigma to find out the main defect causes, as well as the causal variables, and then suggest a logical solution to minimize the defects is the conclusion of this project. another researcher nedra et al. (2019) also from india,. to improve process performance for clothing in small-to-medium enterprises and smes. perform a combination technique between applying pdca, and controlling, improving each dmaic step that is carried out continuously. the use of the pdca-dmaic technique is better when applied with a certified company, than with an uncertified one, it is much more effective. automobile: in usa, ellis (2016). in the process of making a car, an exploration of the application of lean six sigma is carried out to find the root causes of the inefficiency of the production process. the adoption of lean six sigma to increase orders at car manufacturing facilities has successful in increased efficiency for the company. pharmaceutical: al-shourah et al. (2018) calculated quality impact programs in pharmaceutical companies in jordan by simple regression, according to the result of calculation they decided, in order to improve production performance there is an influence of the manufacturing system which we are familiar with the alpha (α) value which is statistically significant at the significance level (α≤0.05). aerospace: in uk, thomas et al. (2016) conducted research at aerospace companies to create an integrated approach between six sigma and lean elements. after calculating all the factors, they concluded a successful financial savings of more than £ 2 million was proposed. telecommunications: in south africa (shuttleworth, 2015) conducted research in south african telecommunication companies, regarding when implementing lss management support is very important. in the final section this research concludes that lss ultimately contributes to improving yields and can improve service delivery. shipping: another researcher from uk (garzareyes et al., 2016) observing the framework of the ship loading process in the iron pellet industry and commercial time as a form of lss implementation to improve key performance and operating main indicators. after calculating all parameters and extract problem by using value stream mapping (vsm), result of this project can be save about $ 300,000 usd every year. int. j. prod. manag. eng. (2021) 9(2), 125-139creative commons attribution-noncommercial-noderivatives 4.0 international lean six sigma implementation, a systematic literature review 129 http://creativecommons.org/licenses/by-nc-nd/4.0/ education: li et al. (2019) from purdue, doing observations in college, how to use six sigma can improve service processes. they realize that the strength of lss in the service process in higher education has a very big impact, so that implementation must continue to be pursued. hospital: in usa, furterer (2012) conducted research on applying lean six sigma problemsolving methodologies and tools to improve linen loss in acute care hospitals. improve the key operational metrics the team is able to perform. he concluded that lss was a great application of problem solving methodology and tools. it: kundu & murali manohar (2012) conduct observations in the it support services sector and identify csfs from a lean implementation perspective. they concluded that the it support services sector had not been systematically examined and investigated for the application of lean principles to the csf section. financial: in sidney (chelliah & skinner, 2016) conduct research to maximize financial benefits with the implementation of lean six sigma. finally, they recommend strategies for using key metrics to drive leadership-driven decisions about improvement. public sector: last, research the public sector, researcher from malaysia, kowang et al. (2019) did a research success factors for lss implementation in private and public sectors. in this section, knowledge of lss and a culture of continuous improvement that contributes to the sustainability of lss implementation can be realized. table 1. existing literature review of the lean six sigma. no paper identity research object country result 1 (ghaziabad, 2012) develop ability to eliminate break time and hard alloys with wider width rolling in the hot rolling mill process india resulting in better order compliance and massive reductions in inventory, and cycle times were dramatically reduced from 47 to 20 days, 2 (pereira et al., 2019) the optimization of internal process portuguese the lead time is 12 weeks, which corresponds to approximately 164 h of cycle time (oct), 23 h related to change-over (c/o), totaling in average of 60% availability time. 3 (barnabé et al., 2016) evaluation of health institutions for performance improvement being analyzed italy significant savings of over € 28,000 annually which are then reallocated in a different and more efficient manner than over 65 days of hospitalization. 4 (schön et al., 2010) six sigma influences job satisfaction. sweden job satisfaction according to four categories of personal change; an overall assessment of the influence of six sigma on the company; influence on the organization of the careerist impression. 5 (nascimento et al., 2019) six-sigma principles and exploring synergies of lean production (lp) brazilia consider the integration of pdca (kaizen), dmaic methodology (from six sigma) and lp principles, a conceptual framework of lss can be proposed. 6 (sibalija & vidosav, 2014) the approaches for integration of six sigma and lean serbia in general, to keep sustaining of the business by implementing lss synergy. 7 (furterer, 2012) improve the linen processes usa by undertaking the implementation of automated linen and scrub dispensers and operational improvements, the team was able to improve key linen operating metrics, saving $ 77,480 for the first year and a soil-to-clean linen ratio of 16%. 8 (mvsit et al., 2016) ranking of these enablers and identification of appropriate enablers. india lean six-sigma is structured as a hierarchy and pairwise comparison matrix is formed and arranged, to evaluate enabler ratings is the basic framework of ahp. (table 1, continued on next page) int. j. prod. manag. eng. (2021) 9(2), 125-139 creative commons attribution-noncommercial-noderivatives 4.0 international tampubolon and purba 130 http://creativecommons.org/licenses/by-nc-nd/4.0/ no paper identity research object country result 9 (albliwi, et al, 2013) understand more deeply the critical failure factors for the application of lss in various sectors uk in the organization there are 34 critical failure factors for the spread of lss. poor selection and priority of lss projects, lack of training lack of top management attitudes. 10 (kowang et al., 2019) success factors for lss implementation malaysia important factors in the implementation of lss, top management support and communication agreed by previous researchers as success factors for the innovative culture of lss. 11 (fullerton et al., 2014) operations personnel with their internal decision making can be supported by lean map usa of the 244, there are 119 factories that have some form of calculating lean implementation. 12 (taylor et al., 2011) lean dominant versus six sigma dominant usa there is an assumption that lean six sigma is not standardized. the dominant lean six sigma article ensures that lean is the dominant philosophy and six sigma is the subordinate tool. 13 (al-shourah et al., 2018) in pharmaceutical companies to improve production performance jordan “there is a statistically significant impact of lean management and six sigma, quality programs, just in time, manufacturing systems, at the level of significance (α≤0.05) in improving the performance of production in pharmaceutical companies in the amman stock exchange”. 14 (assarlind & aaboen, 2014) adopting lean six sigma to identify strengths (in the form of converters and inhibitors) norwegian the need for recognition in the adoption process at peak tech. the order of the organization level from level 0 to level 1, and from level 1 to level 2. how they contribute to the enterprise adoption process and not such converters and blockers is more interesting. 15 (jirasukprasert et al., 2014) reducing defective products process of rubber gloves in the manufacturing industry malaysia from million product, found production defect (dpmo) from 195,095 ppm to 83,750 ppm, thus the practice of six-sigma can increase the sigma level from 2.4 to 2.9. 16 (motiani & kulkarni, 2019) business process of outsourcing and knowledge process in industry. india organizational factors such as cost drivers and strategic alignment were observed variations. to ensure smooth implementation top management commitment is a must in all cases. 17 (siregar et al., 2019) the focus on several journals published on a particular topic is subject to review indonesia almost all types of manufacturing industries such as shipping, automotive, paper, pharmaceuticals, iron, etc. 21 journals reviewed, there are 19 journals for case study of the application of lss. 18 (pervez et al., 2020) a conceptual framework for the implementation of six sigma practices and lean green malaysia the framework which incorporates the practices of lean green six sigma’s effect on the sustainability performance of smes. 19 (hill et al., 2018) maintenance repair and overhaul (mro) in an aerospace facility. uk down from an average calls of 29 per month to just 13 per month for late calls. 20 shuttleworth, 2015) a south african telecommunications company implementing lss south african top management support determines the success of the lss project directly or indirectly. 21 (valles et al., 2009) reduce the level of defects at a semiconductor company méxico before the improvement, it was found that the sigm level was 3.35 sigma and an increase of 0.37 sigma was obtained which represented the elimination of 1.88% units which did not match or equal to 18,788 ppm. (table 1, continued on next page) int. j. prod. manag. eng. (2021) 9(2), 125-139creative commons attribution-noncommercial-noderivatives 4.0 international lean six sigma implementation, a systematic literature review 131 http://creativecommons.org/licenses/by-nc-nd/4.0/ no paper identity research object country result 22 (indrawati & ridwansyah, 2015) improve the manufacturing process capability, in iron ores industry. indonesia the drying iron ore production process is not optimal because the production efficiency is only at the level of 52%. process capability is at the 2.96 sigma level, because the non necessary non value added (nnva) activity is 14.20% and non value added (nva) is 33.67%. 23 (maleszka & linke, 2016) the impact of lean six sigma tools in some polish production companies poland the list of tools most often used by companies are: pareto diagram and ishikawa diagram 5s and kaizen, kaizen, 5s. 24 (gijo et al., 2011) reducing defects in automotive companies in the fine grinding process. india in the fine grinding process defect can be reduce from 16.6 to 1.19%. 25 (arcidiacono & pieroni, 2018) increase the qoe felt by patients while reducing health care costs. italy we can assume that most of the corrective activities should be carried out in this process with respect to the mht value = 348 minutes. 26 (garza-reyes et al., 2016) framework of the ship loading process in the iron pellet industry and commercial time as a form of lss implementation to improve key performance and operating main indicators uk the loading of ships can be increased processing capability and commercial time by more than 30 percent. result of this project can be save about $ 300,000 usd every year. 27 (ellis, 2016) to improve the order processing process in an automobile manufacturing facility, determine the root causes of process inefficiencies and make recommendations usa related to the order processing process can be done by postponing the order or updating the spec pro sooner will eliminate the most influential defects. 28 (elbermawy et al., 2014) improving a pharmaceutical industry in supply chain processes. egypt order preparation time from 206 to 121 time/ mins. 29 (sanidad & dalimot, 2019) effects of implementing projects using lean six sigma, in various companies in the philippines philippines “determining financial needs” and “dividend policy” areas with 97% and 94%, respectively. 30 (li et al., 2019) improving the service process in university with the application of six sigma purdue the strength of six sigma in the service process in university can be seen through its application. 31 (ahmed et al., 2018) lean six sigma effects in hospital malaysia in malaysian health care organizations it was found that there was no significant relationship between top management commitment and quality performance. 32 (rodrigues nogueira et al., 2017) the final product in the textile industry, defects that occur can be reduced india after applying lean six sigma the percentage of defects can be reduced from 8.25 to 2.63. and for the industry sigma level value, it shifts from level 2.9 to level 3.1 33 (kundu & murali manohar, 2012) csfs is it support services sector as a guideline india certain important regulating and facilitating factors are the keys to a successful lean application. 34 (chelliah & skinner, 2016) maximize financial gain australia based on management practices that involve process management and strategic metrics, it is the project improvement strategy that should be chosen. (table 1, continued on next page) (table 1, continued from previous page) int. j. prod. manag. eng. (2021) 9(2), 125-139 creative commons attribution-noncommercial-noderivatives 4.0 international tampubolon and purba 132 http://creativecommons.org/licenses/by-nc-nd/4.0/ no paper identity research object country result 35 (nedra et al., 2019) for smes, small and medium enterprises in the clothing business, process improvements are carried out india the lead time decreased from 39.47 days to 30.23 days. increasing the sigma level from the 1.45 level to the 3.85 level and increasing the cp processability from 0.5 to 1.3 and 36 (brown et al., 2019) surgery admission improvement rate (dosa) ireland the preoperative test duplication has decreased from 83 to <2% surgery admission improvement rate has increased from 10 to 75%. 37 (trakulsunti et al., 2020) dispensing errors in the inpatient teaching hospital pharmacy can be reduced thailand of the 20,000 days of hospitalization per month error expenditure has decreased from 6 to 2 incidents, representing a 66.66% decrease. 38 (dewi et al., 2012) minimize waste in prime line international ltd indonesia waste in the production process is waiting with an incidence percentage of 95.81% and a sigma level of 0.00, a defect with a percentage of incidence of 2.64% and a level sigma 2.84, and also overproduction by percentage incidence of 0.76% and the sigma level of 3.55. 39 (bratić, 2011) at the croatian graphics company, the six sigma approach is carried out and presents the results of empirical research croatian in a croatian graphic company, tools analysis, variance (48.15%), regression analysis (49.38%) and process mapping (56.79%) were carried out as part of the six-sigma technique analysis. 40 (ketan & nassir, 2016) improvement of aluminum hot extrusion process capability irag profit increased from id 127,000 to id 223,000 per 1000 kg, sigma level was increased from level 1.4 to level 2.4, processing yield (y) was increased from 46% to 81%, and dpmo reduction from 536,804 ppm to 185,795.09 ppm. 41 (thomas et al., 2016) six sigma framework in an aerospace company and strategic lean uk reducing value-added time by 5% and, reducing non-value-added time by 44.5%, building on 20.5% reduction in time, and increasing on-timein-full (otif) delivery to customers by 26.5%. 42 (sabry, 2014) implementation in some of lebanese hospitals performance indicators lebanon factors 1, 2, 3, 4, and 6 namely (executive commitment), (adopting a philosophy), (benchmarking), (training) and (closer supplier relationship) are the five critical success factors for the differences of the 17 quality factors of the sixsigma program quality with significant at the 5% level. 43 (barnabé et al., 2016) conduct analysis and evaluation for performance improvement in health institutions italy significant savings of over € 28,000 annually which are then reallocated in a different and more efficient manner than over 65 days of hospitalization. 44 (kremcheeva & kremcheev, 2019) six sigma method implementation in the educational process russia the learning process before students gain competence, it requires input and output of skills and knowledge processes that are transformed and also expanded. 45 (ha et al., 2016) quality improvement a mass immunizations project at the u.s. naval academy usa the project implementation process proved to be controlled with a capability index of 1.18 and a performance index of 1.10, resulting in a damage rate of 0.04% and an average immunization waiting time decreased by 79% and staff decreased by 10%. 46 (sachin & dileeplal, 2017) improve the production process in a foundry industry india casting rejection decreased from a rate of 15.61% to 7.40%. (table 1, continued on next page) (table 1, continued from previous page) int. j. prod. manag. eng. (2021) 9(2), 125-139creative commons attribution-noncommercial-noderivatives 4.0 international lean six sigma implementation, a systematic literature review 133 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. mapping of the journal according to the papers mapping on figure 2~5, there are 50 papers reviewed in total, all of papers are published in the last 11 years (2009 – 2020). the papers are spreading into 12 different industries sector from manufacturing, healthcare, oil & gas, multi object , garment, automobile, public sector, pharmaceutical, aerospace, telecommunications, shipping, education, hospital, it, and graphic. where the author as well as the object research location was spreading into 27 countries globally. most of the researcher from those 50 papers were observed the linkage between the implementation of lss with identify appropriate papers to review and the significant improvement of the organization as well as it increased customer satisfaction. of course the author also includes several literature reviews as input on the obstacles and success of lss implementation. the following map journals are divided into 3 categories, namely the number of publications per year, the country of origin of the researcher and the third is the industry or material of the studied object. the year of publication of the journal studied is as follows (figure 3). from the graphs presented, most journals in 2016 & 2019 were 20%, followed by journals in 2014 was 12% and 2012, 2017 and 2018 as much as 8 %, the rest can be seen in the graphs. the country of origin of the researcher can be seen in the pareto chart (figure 4), where the majority of the researchers are coming from india, as the research in manufacture, garment and it. the object of the study can be seen from figure 5, where most of the research object comes from manufacture. this survey literature uses random sampling, with the keyword “lean six sigma” on data base provider in the international journal. no paper identity research object country result 47 (girmanová et al., 2017) to avoid increased internal costs associated with poor product quality of metallurgical products slovak the defective product indicator value per one million opportunities decreased drastically from 81,038 ppm or equal to 2.9 sigma, to 39,636 ppm dpmo or equal to 3.3 sigma. 48 (prabu et al., 2013) quality improvement in foundry process. india increase sigma level quality from 3.90 to 3.97. 49 (al-qatawneh et al., 2019) to apply six sigma in the field of health care logistics so it is necessary to propose a framework jordan this work expresses the idea of prioritizing all stored items based on performance, cost and criticality in hospital management. 50 (rahman et al., 2017) reduce defect rate at garment industry bangladesh we can achieve the desired 2% breakdown rate by the end of the project deadline. (table 1, continued from previous page) figure 3. years publication. int. j. prod. manag. eng. (2021) 9(2), 125-139 creative commons attribution-noncommercial-noderivatives 4.0 international tampubolon and purba 134 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5. lean six sigma and agenda for future research many countries in the world have implemented lss extensively in many industrial sectors. in the current era, lss is also still popular, which can be seen in the publication of research papers in the last 10 years which confirms that researchers are still interested in observing the implementation of lss in organizations. with today’s business environment and rapid change it is emphasized that lss is still in use and still is appropriate. for the future research framework on lss, in preparation for the implementation of industry 4.0, it must be continuously improved and adjusted. this will enhance process improvement capabilities and product design capabilities that will benefit the figure 4. country of researcher. figure 5. research object. int. j. prod. manag. eng. (2021) 9(2), 125-139creative commons attribution-noncommercial-noderivatives 4.0 international lean six sigma implementation, a systematic literature review 135 http://creativecommons.org/licenses/by-nc-nd/4.0/ company. for future research, six sigma collects data to attain its goal. collected data should be analyzed to create an optimum and proper decision. however, industry 4.0 technologies change together an enormous amount of information. therefore, traditional data analysis techniques do not seem to be sufficient because they require more prolonged and value. it is possible to profit from advanced techniques, which are suitable for large data, like big data analytics and process mining, additionally to traditional techniques to create effective decisions, shown in figure 6. also, in the future research collaboration of 2 components, there are industry 4.0 and lean six sigma, can provide a guide that makes easier, faster, more reliable, and satisfied decisions with data for improving quality in processes. with the use of industry 4.0 technology in the implementation of lean six-sigma, data input, data processing and analysis to conclusions can be made very quickly and accurately, which in turn can increase efficiency and increase productivity very significantly. 6. conclusion according to the mapping results. in accordance with the results of the mapping that has been made, the most published years can be published, in 2016 and 2019. in the research country section, india is the country that produces the most publications. the most research object comes from the manufacturing sector. it is realized that the results of the mapping above do not describe the real situation as a whole, but only a picture when collecting data according to keywords. the main review of this article is a combination of six sigma tools and techniques and lean manufacturing principles that change the process level and product quality produced in the industry. the combination of the two concepts we call lean six sigma (lss) is a powerful tool that can influence processes and customer / stake holder satisfaction in an organization. the adoption of lean six sigma is still starting in the last decade or two in developing countries. in context, extensive records have reduced researchers’ the increasing number of current research on leansix-sigma is a testament to the great attention and research interest in this area. in developed countries lean-six-sigma has been practiced most of the time, but still focuses on interests in manufacturing processes and management activities. all types of industries can apply lean-six-sigma for better customer / stakeholder satisfaction and productivity through continuous improvement in business activities. (olanrewaju et al., 2019). for example base on garza-reyes et al., (2016) research concluded improvement both the capability of its ship loading process and commercial time by more than 30 percent, resulting in operational savings in the range of $300,000 usd per year. in the manufacturing section, the increase in the impact of improvements is felt to provide great benefits to the organization, this can be seen in table 1, point to numbers 1, 2, 4, 8, 11, 12, 14, 15, 16, 21, 22, 23, 24, 38, 40, 46 and 47. then from the other side, we can see that lss publications from year to year continue to increase, this shows that these tools have increased use in various sectors. in many countries in the world lean six sigma (lss) is still widely used and applied in many industrial sectors. lss focuses on providing the best product or service solutions and increasing customer satisfaction are the reasons why lss is still suitable for the current situation, it is necessary to figure 6. future research framework. int. j. prod. manag. eng. (2021) 9(2), 125-139 creative commons attribution-noncommercial-noderivatives 4.0 international tampubolon and purba 136 http://creativecommons.org/licenses/by-nc-nd/4.0/ improve service quality, organizational quality and product quality, as a whole. of course, there are obstacles in implementing lss, such as lack of top management commitment and involvement, lack of communication, lack of training and education, limited resources and others implementation. references ahmed, s., abd manaf, n.h., & islam, r. (2018). effect of lean six sigma on quality performance in malaysian hospitals. international journal of health care quality assurance, 31(8), 973–987. https://doi.org/10.1108/ijhcqa-07-2017-0138 albliwi, s., antony, j., abdul halim lim, s. and van der wiele, t. (2014). critical failure factors of lean six sigma: a systematic literature review, international journal of quality & reliability management, 31(9), 1012-1030. https://doi.org/10.1108/ijqrm-09-2013-0147 al-qatawneh, l., abdallah, a.a.a., & zalloum, s.s.z. (2019). six sigma application in healthcare logistics: a framework and a case study. journal of healthcare engineering, 2019. https://doi.org/10.1155/2019/9691568 al-shourah, a.a., al-tarawneh, r.t., & ali, f. (2018). the integration of lean management and six sigma strategies to improve the performance of production in industrial pharmaceutical. 13(8), 207–216. https://doi.org/10.5539/ijbm.v13n8p207 arcidiacono, g., & pieroni, a. (2018). the revolution lean six sigma 4.0. international journal on advanced science, engineering and information technology, 8(1), 141–149. https://doi.org/10.18517/ijaseit.8.1.4593 assarlind, m., & aaboen, l. (2014). chalmers publication library. barnabé, f., giorgino, m.c., guercini, j., & bianciardi, c. (2016). performance enhancement and continuous improvement in healthcare: how lean six sigma “hits the target.” international journal of business and social science, 7(5), 15. bratić, d. (2011). six sigma: a key driver for process improvement. communications of the ibima, 2011, 1–15. https://doi. org/10.5171/2011.823656 brown, r., grehan, p., brennan, m., carter, d., brady, a., moore, e., teeling, s.p., ward, m., & eaton, d. (2019). using lean six sigma to improve rates of day of surgery admission in a national thoracic surgery department. international journal for quality in health care: journal of the international society for quality in health care, 31(1), 14–21. https://doi.org/10.1093/intqhc/mzz083 chelliah, j., & skinner, a. (2016). organizational transformation: strategic application of lean six sigma for high performance. 12(1), 85–92. elbermawy, m.f., al manhawy, a.a., & ibrahim, h.e.a. (2014). implementation of lean six sigma for improving supply chain processes in a pharmaceutical industry. international journal of scientific & engineering research, 5(8), 519–529. ellis, s.f. (2016). the application of lean six sigma to improve a business process: a study of the order processing process at an automobile manufacturing facility. university of south carolina scholar commons, 139. fullerton, r.r., kennedy, f.a., & widener, s.k. (2014). lean manufacturing and firm performance: the incremental contribution of lean management accounting practices. journal of operations management, 32(7–8), 414–428. https://doi.org/10.1016/j.jom.2014.09.002 furterer, s. (2012). applying lean six sigma to reduce linen loss in an acute care hospital. international journal of engineering, science and technology, 3(7), 39–55. https://doi.org/10.4314/ijest.v3i7.4s garza-reyes, j.a., al-balushi, m., antony, j., & kumar, v. (2016). a lean six sigma framework for the reduction of ship loading commercial time in the iron ore pelletising industry. production planning and control, 27(13), 1092–1111. https://doi.org/10.1080/09537287.2 016.1185188 ghaziabad, m.t. (2012). improvement process for rolling mill through the dmaic six sigma approach. 6(3), 221–231. gijo, e.v., scaria, j., & antony, j. (2011). application of six sigma methodology to reduce defects of a grinding process. quality and reliability engineering international, 27(8), 1221–1234. https://doi.org/10.1002/qre.1212 girmanová, l., šolc, m., kliment, j., divoková, a., & mikloš, v. (2017). application of six sigma using dmaic methodology in the process of product quality control in metallurgical operation. acta technologica agriculturae, 20(4), 104–109. https://doi.org/10.1515/ata2017-0020 ha, c., mccoy, d.a., taylor, c.b., kirk, k.d., fry, r.s., & modi, j.r. (2016). using lean six sigma methodology to improve a mass immunizations process at the united states naval academy. military medicine, 181(6), 582–588. https://doi.org/10.7205/milmed-d-15-00247 hill, j., thomas, a.j., thomas, a.j., & the, s.e. (2018). the implementation of a lean six sigma framework to enhance operational performance in an mro facility the implementation of a lean six sigma framework to. production & manufacturing research, 6(01), 26–48. https://doi.org/10.1080/21693277.2017.1417179 indrawati, s., & ridwansyah, m. (2015). manufacturing continuous improvement using lean six sigma: an iron ores industry case application. procedia manufacturing, 4(iess), 528–534. https://doi.org/10.1016/j.promfg.2015.11.072 int. j. prod. manag. eng. (2021) 9(2), 125-139creative commons attribution-noncommercial-noderivatives 4.0 international lean six sigma implementation, a systematic literature review 137 https://doi.org/10.1108/ijhcqa-07-2017-0138 https://doi.org/10.1108/ijqrm-09-2013-0147 https://doi.org/10.1155/2019/9691568 https://doi.org/10.5539/ijbm.v13n8p207 https://doi.org/10.18517/ijaseit.8.1.4593 https://doi.org/10.1016/j.jom.2014.09.002 https://doi.org/10.4314/ijest.v3i7.4s https://doi.org/10.1080/09537287.2016.1185188 https://doi.org/10.1080/09537287.2016.1185188 https://doi.org/10.1002/qre.1212 https://doi.org/10.1515/ata-2017-0020 https://doi.org/10.1515/ata-2017-0020 https://doi.org/10.7205/milmed-d-15-00247 https://doi.org/10.1080/21693277.2017.1417179 https://doi.org/10.1016/j.promfg.2015.11.072 http://creativecommons.org/licenses/by-nc-nd/4.0/ jirasukprasert, p., garza-reyes, j. a., kumar, v., & lim, m. k. (2014). a six sigma and dmaic application for the reduction of defects in a rubber gloves manufacturing process. international journal of lean six sigma, 5(1), 2-21. https://doi.org/10.1108/ijlss-03-2013-0020 ketan, h., & nassir, m. (2016). aluminium hot extrusion process capability improvement using six sigma. advances in production engineering and management, 11(1), 59–69. https://doi.org/10.14743/apem2016.1.210 kowang, t.o., yew, l.k., hee, o.c., fei, g.c., & long, c.s. (2019). lean six sigma implementation: does success means sustainability? international journal of academic research in business and social sciences, 9(6), 907–914. https://doi.org/10.6007/ijarbss/v9i6/6051 kremcheeva, d.a., & kremcheev, e.a. (2019). implementation of the six sigma method in the educational process. journal of physics: conference series, 1384(1), 012022. https://doi.org/10.1088/1742-6596/1384/1/012022 kundu, g., & murali manohar, b. (2012). critical success factors for implementing lean practices in it support services. international journal for quality research, 6(4), 301–312. li, n., laux, c.m., & antony, j. (2019). how to use lean six sigma methodology to improve service process in higher education: a case study. international journal of lean six sigma, 10(4), 883–908. https://doi.org/10.1108/ijlss-11-2018-0133 maleszka, a., & linke, m. (2016). improvement of management process by using lean six sigma tools in some big organisation of food industry. polish journal of natural sciences, 31(1), 101–112. motiani, n., & kulkarni, a. (2019). sustainability and impact of lean six sigma practices: learnings from some kpo / bpo organizations. 11, 442–452. https://doi.org/10.35940/ijitee.k1076.09811s19 mvsit, s.k., haleem, a., & millia, j. (2016). lean six sigma implementation: an analytic hierarchy process approach lean six sigma implementation: an analytic hierarchy process approach. december 2014. nascimento, d.l. de m., goncalvez quelhas, o.l., gusmão caiado, r.g., tortorella, g.l., garza-reyes, j.a., & rocha-lona, l. (2019). a lean six sigma framework for continuous and incremental improvement in the oil and gas sector. international journal of lean six sigma, 11(3), 577–595. https://doi.org/10.1108/ijlss-02-2019-0011 nedra, a., néjib, s., yassine, c., & morched, c. (2019). a new lean six sigma hybrid method based on the combination of pdca and the dmaic to improve process performance: application to clothing sme. industria textila, 70(05), 447–456. https://doi.org/10.35530/ it.070.05.1595 olanrewaju, f., chima uzorh, a., & nnanna, i. (2019). lean six sigma methodology and its application in the manufacturing industry – a review. american journal of mechanical and industrial engineering, 4(3), 40. https://doi.org/10.11648/j.ajmie.20190403.11 pereira, a.m.h., silva, m.r., domingues, m.a.g., & sá, j.c. (2019). lean six sigma approach to improve the production process in the mould industry: a case study. quality innovation prosperity, 23(3), 103–121. https://doi.org/10.12776/qip.v23i3.1334 pereira, c.m., anholon, r., & batocchio, a. (2017). obstacles and difficulties implementing the lean philosophy in brazilian enterprises. brazilian journal of operations & production management, 14(2), 218. https://doi.org/10.14488/bjopm.2017.v14.n2.a10 pervez, m., abu, n., & owee, t. (2020). the development of a sustainability framework via lean green six sigma practices in smes based upon rbv theory. 12(5), 135–156. prabu, k., makesh, j., naveen raj, k., devadasan, s.r., & murugesh, r. (2013). six sigma implementation through dmaic: a case study. international journal of process management and benchmarking, 3(3), 386–400. https://doi.org/10.1504/ijpmb.2013.058162 rahman, a., shaju, s.u.c., sarkar, s.k., hashem, m.z., hasan, s.m.k., mandal, r., & islam, u. (2017). a case study of six sigma definemeasure-analyze-improve-control (dmaic) methodology in garment sector. independent journal of management & production, 8(4), 1309. https://doi.org/10.14807/ijmp.v8i4.650 rodrigues nogueira, r., lemos cotrim, s., & camila lapasini leal, g. (2017). lean six sigma implementation on a yarn textile industry. revista gestão da produção operações e sistemas, 12(2), 67–94. https://doi.org/10.15675/gepros.v12i2.1634 sabale, s., & thorat, s. (2019). applications of lean six sigma. 1999–2004. sabry, a. (2014). factors critical to the success of six-sigma quality program and their influence on performance indicators in some of lebanese hospitals. arab economic and business journal, 9(2), 93–114. https://doi.org/10.1016/j.aebj.2014.07.001 sachin, & dileeplal. (2017). six sigma methodology for improving manufacturing process in a foundry industry. international journal of advanced engineering research and science, 4(5), 237172. sanidad, b.g., & dalimot, m.m. (2019). the effects of lean six sigma projects in the companies. operational and financial efficiencies. 21(6), 82–88. schön, k., bergquist, b., & klefsjö, b. (2010). the consequences of six sigma on job satisfaction: a study at three companies in sweden. international journal of lean six sigma, 1(2), 99–118. https://doi.org/10.1108/20401461011049494 sibalija, t., & vidosav, m. (2014). integrating lean with / within six sigma, international journal ’total quality management & excellence’. 38(4). siregar, k., ariani, f., ginting, e., & p, t.d.m. (2019). lean six sigma for manufacturing industry: a review lean six sigma for manufacturing industry: a review. in iop conference series: materials science and engineering, 505(1), 012056. https://doi.org/10.1088/1757899x/505/1/012056 int. j. prod. manag. eng. (2021) 9(2), 125-139 creative commons attribution-noncommercial-noderivatives 4.0 international tampubolon and purba 138 https://doi.org/10.1108/ijlss-03-2013-0020 https://doi.org/10.14743/apem2016.1.210 https://doi.org/10.6007/ijarbss/v9-i6/6051 https://doi.org/10.6007/ijarbss/v9-i6/6051 https://doi.org/10.1088/1742-6596/1384/1/012022 https://doi.org/10.1108/ijlss-11-2018-0133 https://doi.org/10.35940/ijitee.k1076.09811s19 https://doi.org/10.1108/ijlss-02-2019-0011 https://doi.org/10.35530/it.070.05.1595 https://doi.org/10.35530/it.070.05.1595 https://doi.org/10.11648/j.ajmie.20190403.11 https://doi.org/10.12776/qip.v23i3.1334 https://doi.org/10.14488/bjopm.2017.v14.n2.a10 https://doi.org/10.1504/ijpmb.2013.058162 https://doi.org/10.14807/ijmp.v8i4.650 https://doi.org/10.15675/gepros.v12i2.1634 https://doi.org/10.1016/j.aebj.2014.07.001 https://doi.org/10.1108/20401461011049494 https://doi.org/10.1088/1757-899x/505/1/012056 https://doi.org/10.1088/1757-899x/505/1/012056 http://creativecommons.org/licenses/by-nc-nd/4.0/ shuttleworth, c.c. (2015). management support for the application of lean six sigma methodology to improve customer satisfaction in a south african telecommunications company. problems and perspectives in management, 13(4), 205-214. taylor, j., sinn, j., ulmer, j., & badar, m. (2015). proposed progression of lean six sigma. the journal of technology studies, 41(1), 2-8. thomas, a.j., francis, m., fisher, r., & byard, p. (2016). implementing lean six sigma to overcome the production challenges in an aerospace company. production planning and control, 27(7–8), 591–603. https://doi.org/10.1080/09537287.2016.1165300 tohidi, h. (2012). six sigma methodology and its relationship with lean manufacturing system. advances in environmental biology, 6(2), 895–906. trakulsunti, y., antony, j., dempsey, m., & brennan, a. (2020). reducing medication errors using lean six sigma methodology in a thai hospital: an action research study. international journal of quality and reliability management, 38(1), 339-362. https://doi.org/10.1108/ijqrm-10-2019-0334 valles, a., sanchez, j., noriega, s., & gómez nuñez, b. (2009). implementation of six sigma in a manufacturing process: a case study. international journal of industrial engineering: theory applications and practice, 16(3), 171–181. dewi w.r., setyanto, n.s., tantrika, c.f.m. (2012). implementasi metode lean six sigma sebagai upaya minimasi waste pada pt. prime line internasional. jurnal rekayasa dan manajemen sistem industri, 1(1), 47–56. int. j. prod. manag. eng. (2021) 9(2), 125-139creative commons attribution-noncommercial-noderivatives 4.0 international lean six sigma implementation, a systematic literature review 139 https://doi.org/10.1108/ijqrm-10-2019-0334 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.12935 received: 2020-01-06 accepted: 2020-06-07 a k-chart based implementation framework to attain lean & agile manufacturing zaheer, s.a1, amjad, m.s. a2, rafique, m.z.a3, khan, m.a.a4 a department of mechanical engineering, the university of lahore, pakistan. a1 shafaat1990@yahoo.com, a2 saadamjad95@gmail.com, a3 muhammadzeeshanrafique@gmail.com, a4 mohammad.aamir@me.uol.edu.pk abstract: lean manufacturing has always ensured production optimization by eliminating wastes, and its implementation has helped in improving the operational performance of the organization since it eliminates the bottlenecks from the processes, thus making them efficient. in lean scenarios, the focus is on “waste” elimination, but in agile manufacturing, the focus is on the ability of comprehension of changing market dynamics and the resilience. one of the major factors in the combined implementation of lean and agile approaches is inadequate planning, monitoring and lack of awareness regarding changing market trends, and this can be countered by utilizing the effective tool of k-chart. through a systematic literature review, the authors establish the requirement of effective planning and monitoring in the implementation of integrated lean and agile approach, concluding that k-chart is a handy tool to adopt for their effective implementation. the result provides a new vision of lean implementation through k-chart, whereas it provides clarity to practitioners by presenting a k-chart based implementation framework for achieving favourable results. being a literature review the research work can be validated through a case study approach in future through a comparative analysis between various implementation techniques and k-chart. key words: k-chart, lean manufacturing, agile manufacturing, operational excellence. 1. introduction lean manufacturing is vital in the elimination of wastes and improvement of the operational efficiency, in addition to optimized production (khodeir & othman, 2016; moyano-fuentes et al., 2012). every industry is using lean for improvements in its process since long (koskela, 1992). lean manufacturing minimizes the hurdles that occur in processing and improves the production rate (bhamu & singh sangwan, 2014; rafique et al., 2017). lean principles have been successfully applied on the systems in which lead times and operational efficiency were improved (matt & rauch, 2013). on the contrary, many industries adopt lean system but usually falter in its implementation, which is not a straightforward task. this complexity of lean has produced awareness in manufacturers and researchers of the whole world to apply such tools that are beneficial for lean in the system (swank, 2003), to acquire the best results like improve lead times, customers demand and controlling cost and quality (goldsby, griffis, & roath, 2006; rafique, ab rahman, saibani, arsad, & saadat, 2016). in order to do so, the lean tool of value stream mapping is beneficial since it categorizes the process to value added and non-value added times, laying groundwork for improvements (christian & zimmers, 1999). lean can be considered as a philosophy which helps to make the system effective and efficient for industries. the cost of labours, customers satisfaction, reduce lead time, reduce wastes can be minimized with the help of lean and agile (fagerholm et al., 2015). leanness eliminates the wastes and the process becomes more to cite this article: zaheer, s., amjad, m.s., rafique, m.z., khan, m.a. (2020). a k-chart based implementation framework to attain lean & agile manufacturing. international journal of production management and engineering, 8(2), 123-135. https://doi.org/10.4995/ijpme.2020.12935 int. j. prod. manag. eng. (2020) 8(2), 123-135creative commons attribution-noncommercial-noderivatives 4.0 international 123 https://orcid.org/0000-0003-3265-2081 http://creativecommons.org/licenses/by-nc-nd/4.0/ efficient. lean manufacturing is used to control the production according to its environment and company’s demand (maqbool et al., 2019; naslund, 2008). responding to, and taking advantage of changes through strategic utilization of managerial and manufacturing methods and tools are the pivotal concepts of agile manufacturing (sharifi & zhang, 2001). diversification and product individuality are the backbones of agile manufacturing, whilst being able to swiftly responding to market change (waters, 2007). the success of agility lies in making through a turbulent and a period of uncertainty, and striving for excellence by prospering in a competitive environment, thus setting the foundations for organizational success (yauch, 2011). it was deduced by mccullen and towill (2001) that lean manufacturing can be considered a sub category of agile manufacturing. in order to ensure smooth implementation of lean and agile manufacturing, it is imperative to have an efficient tool for its implementation that facilitates the change. cause & effect, fishbone diagram, pareto chart are a few important techniques, but do not cater to the requirement of simultaneous planning and monitoring. in this scenario, the objective of this research is to introduce a novel technique for planning, monitoring and implementation of lean, known as k-chart (abdullah et al., 2006). the research introduces a new concept of k-chart in manufacturing context that will provide a new direction to researchers. in addition to that, it will provide clarity to practitioners regarding the use of k-chart in implementation of lean and agile techniques for improving the production process. the paper is structured in the following manner; research methodology will be is discussed in section 2, section 3 comprises of the literature review, results are in section 4. in the 5th section, the conclusions are drawn along with discussion on the limitation of research and future implications. 2. research methodology in order to achieve a robust research methodology, it is imperative to designate a clear research context. therefore, the research onion strategy was developed by m. saunders, lewis, and thornhill (2009) and m. n. saunders (2011).nesensohn (2014) further explains that division of research onion into four distinct layers in order to achieve favorable results. a research onion comprises of the following layers: research philosophies (ontology, epistemology) research logics (inductive & deductive) research purposes (explanatory, exploratory & descriptive) research approaches (qualitative & quantitative) research strategies (survey, case study, phenomenology, ethnography) research techniques (data collection, procedures) research philosophy being the outermost layer of research onion specifies the starting point of the research work (nesensohn, 2014; saunders, 2011). it is subdivided into ontology, epistemology and axiology (elnadi, 2015; srichuachom, 2015). it can be seen that quantitative research is attributed to positivism philosophy, whereas qualitative research is associated with the concept of interpretivism. after selection of positivism philosophy, the next task was the study of data collection techniques for the subject matter since it is of paramount importance in every research. from numerous data collection procedures, the authors selected the literature review philosophy. the literature review suggests reviewing the data available in literature (elnadi, 2015; robson & mccartan, 2016). the literature reviews inclusive of previous researches and findings, which are gathered from previous publications, journal papers, articles, conference papers, books and other previous reviews. according to robson and mccartan (2016) the leading reason to conduct literature review are: to comprehend the research topic and to acquire research knowledge classifying previous works and examining them to attain results realizing gaps in previous research evading mistakes and limitations of previous research circumventing inadequate work after selection of literature review as research methodology, the authors developed a comprehensive search and selection criteria for lean manufacturing, agile manufacturing & k-chart. this has culminated in development of a roadmap that helped in arriving at the most pertinent research works in line with the subject matter. the designed search and selection criteria has helped in selection of the most relevant journals and research works, which have been cited in the succeeding section. the authors were careful in int. j. prod. manag. eng. (2020) 8(2), 123-135 creative commons attribution-noncommercial-noderivatives 4.0 international zaheer et al. 124 http://creativecommons.org/licenses/by-nc-nd/4.0/ selecting the research works based on the relevance, coherence and relationship to the manuscript aims. the literature was collected from high impact factor and eminent journals in the field of manufacturing, production and their management. in addition to that, credible conference papers were also used in the literature review. it was aimed that the selection criteria returns those articles that carry the most impact in their relevant field, and are in line with the research aims. using the sciencedirect database, the research works for the period 1997-2020 were selected, with primary focus on works from 20132020 the initial search resulted in 134 papers on lean manufacturing, and 138 articles on agile and lean-agile manufacturing. 19 papers related to lean manufacturing were selected after careful consideration; and for agile manufacturing domain 18 papers were selected the papers related to supply chain domain were excluded, and the manufacturing domain was considered only. there was no such criterion of selection of research method from the selected articles, the authors included case studies, surveys, conceptual models, mathematical models, etc. so that the core message for both approaches in conveyed. 3. literature review 3.1. lean manufacturing the very first instance of introducing the manufacturing sector to lean was in 1990, when james p. womack wrote the book the machine that changed the world. the transition from mass production to lean became a necessity since it presented an efficient and economical solution to the manufacturing industry (sabbagh et al., 2016; womack et al., 1990). the era of mass production came to a halt in 1980s and was duly replaced by agile or flexible production, which marked the beginnings of revolution in industrial management (duguay et al., 1997). this led to the conclusion that lean manufacturing enables industries to walk towards a path of business excellence (mamat et al., 2015; mejabi, 2003). koskela (1992) indicated that the major categorization of lean philosophy can be done by segregating the processes in production systems to conversion activities and flow activities, commencement of research w o rk idea generatio n data a vailable? n o n o r esearch possible yes conductin g comprehensive literature review defin e s election criteria defin e s earch criteria hig h impact journals, conference papers, bo oks s election on basis of r elevance & coherence m a nual s earch citation b ased s earch inclusio n or exclusio n exclsu io n out of process inclsuio n final s election & exit s election of r esearch context s election of r esearch philosophy s election of d ata collection technique literature review s election figure 1. search & selection criteria. table 1. search & selection criteria. study material implementation of lean manufacturing, implementation of agile manufacturing, importance of k-chart, modern planning & monitoring techniques period covered 1997–2020 data used from sites emerald, science direct, springer, scopus, taylor and francis online keywords used to search from database lean manufacturing, agile manufacturing, planning & monitoring, k-chart, implementation framework int. j. prod. manag. eng. (2020) 8(2), 123-135creative commons attribution-noncommercial-noderivatives 4.0 international a k-chart based implementation framework to attain lean & agile manufacturing 125 http://creativecommons.org/licenses/by-nc-nd/4.0/ the former being value adding and the latter being time and consuming, while adding no value to the product. lean was further defined in following steps by (womack et al., 1990) that included: explicitly define the end customer need which in in turn define the value eliminate wastes and map each process for each product make the processes flow by continuously making sure that there is no waste, waiting and no downtime introducing a pull system design, production and delivery should be a function of customer need following the philosophy of “continuous improvement lean manufacturing is not just a fancy term thrown everywhere, it’s a philosophy (bhasin & burcher, 2006) that enables industries to do better by employing the principles of continuous improvement and changing the organizational culture. moreover, lean manufacturing helps in waste reduction (bhamu & singh sangwan, 2014; sahwan et al., 2012) thus causing an increase in quality and quickly responding to customer demand the universality of lean is such that its concepts and tools were developed in the services sector (swank, 2003) and the implications can be clearly seen in the better business and operational performance of the organization (nawanir et al., 2013; nordin et al., 2012). the facet of just-in-time manufacturing has a significant impact on the operational performance (rahman et al., 2010; rehman et al., 2018) implementation of lean manufacturing principles considerably reduces the production time (rahani & al-ashraf, 2012), regardless of the size of the organization (matt & rauch, 2013). it has been observed that for achieving long term strategic goals and objectives, it is necessary to implement lean manufacturing techniques to reduce waste and providing detailed guidelines to the management for process improvement (sundar et al., 2014). by basing their work on implementation of lean strategies, (anvari et al., 2011) concluded that it improved productivity and increased competitiveness. however, it was seen that the attitude of workers and complete understanding of the concepts are of primary importance in effective implementation of lean system (nordin et al., 2010). a coalescence of just-in-time & total quality management is necessary for increased employee involvement, increased productivity levels and higher quality standards (sriparavastu & gupta, 1997). by using kanban model, industries can reduce operational costs and can make their workstations more flexible. (rahman et al., 2013). an organization that desires to reduce the lead time, escalate the product quality, increase the flexibility and lower the costs must adopt lean production techniques (martínez sánchez & pérez pérez, 2001). an inverse relation has been observed between the inventory and the extent of leanness, irrespective of the organizational volume – thus reducing the logistics charges over the course of time (chun wu, 2003). lm can affect organizational performance, not only at the operations level but also at the business level. by demonstrating the existence of direct and indirect effects of lm practices on bp (nawanir et al., 2013) provide clear evidence that lm implementation is important to enhance companies’ performance. 3.2. agile manufacturing in lean environment, the focus is on “waste” elimination, but in agile environment, the focus is on the ability of understanding the market changes and responding to them. an important difference is that lean supply is associated with level scheduling, whereas agile supply means reserving capacity to cope with volatile demand.(martin & towill, 2000). supply chains, supply chain management change, and evolve mainly under the pressure of the competition. generally, the existing activities of supply chain management aim at the cost reduction by using instruments for their leanness (lean management concept) or at higher service level by higher flexibility (agile management concept) (konecka, 2010). agility has four underlying principles: delivering value to the customers; being ready for change; valuing human knowledge and skills; and forming virtual partnerships (gunasekaran, 1999) the main characteristics of agile manufacturing summarized by yusuf, sarhadi, and gunasekaran (1999) are given as follows: high end customization of products without compromising on quality manufacturing products and offering services with value added content increased responsiveness to shifting paradigms, uncertain conditions and environmental issues amalgamation of diverse technologies agile manufacturing was termed to be of supreme importance in the survival and prosperity of organizations with the volatile business climate int. j. prod. manag. eng. (2020) 8(2), 123-135 creative commons attribution-noncommercial-noderivatives 4.0 international zaheer et al. 126 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 2. literature review on lean manufacturing. author name & year methodology findings(authors’ points of view) womack and jones (1996) case study the following five steps are effective to ensure leanness: defining the customer value defining value stream subjecting the value stream to a flow introducing a pull system continuous determination towards achieving excellence: martínez sánchez and pérez pérez (2001) survey lean indicators can be classified into six groups: elimination of non-value added activities continuous improvement just-in-time production & delivery teamwork integration of suppliers flexible information systems abdulmalek and rajgopal (2007) case study a simulation model was run it was concluded that application of lean manufacturing principles significantly reduces production lead time by reducing wip inventory, using hybrid production system and introduction total productive maintenance. anand and kodali (2010) literature review the ten levels of implementation framework are given as follows: evaluate prepare organization for lm implementation defining value identification of value stream creating a process flow process improvement using spc, pokayoke etc. stabilizing improvements advancing by letting the customer pull establishing the use of philosophies such as tpm, tm, six sigma etc. pursue perfection losonci et al. (2011) case study & survey it was identified that the critical intrinsic factors (commitment, belief) and external factors (lean work method, communication) affect the success of the lean implementation from workers’ point of view. the stereotypical gender values can intensify the impact of factors related to the particular process type rahani & al-ashraf (2012) case study reduction of waiting time, economic impact of time improvement and lower rejection rates. matt and rauch (2013) case study lean production principles can be conveniently applied to smes resulting in productivity improvement. metternich et al. (2013) case study the number of operators and number of machine tools can be adjusted to suit the production demand, thus accounting for customer volatility. nawanir et al. (2013) case study & survey the author suggest that business performance and lean practices have a direct relationship, resulting in increased productivity. bhamu and singh sangwan (2014) literature review external support is required to enhance adoption of lm in smes. one of the critical implementation factors of lm is simultaneous adoption of leanness in supply chain sundar et al. (2014) literature review a roadmap was proposed for implementation of lm techniques such as every product every interval & continuous improvement. rohani and zahraee (2015) case study the authors applied the approach of value stream mapping in a color factory to achieve lead time and value added time reduction. salonitis and tsinopoulos (2016) survey the authors conducted a survey for greek manufacturing sector and concluded that organizational culture is of paramount importance when it comes to the successful implementation of lean manufacturing. botti et al. (2017) case study the authors developed a mathematical relationship between lean manufacturing and workplace ergonomics, which was then applied to an italian manufacturing firm. the results indicate that hybrid assembly lines are beneficial for maximum productivity. sartal et al. (2017) hypotheses testing the authors developed various hypotheses regarding the integration of lean manufacturing with green practices and it approaches to conclude that the developed conceptual model suggests that it approaches are incumbent for achieving manufacturing excellence. marodin et al. (2018) survey the authors use a survey approach to develop a relationship between the lean product development and lean manufacturing, and conclude that the aforesaid interact in a positive manner. ghobadian et al. (2018) literature review the authors discuss the case of various industries and propose that lean manufacturing concepts can be extended to sustainability and posit that additive manufacturing is of vital importance in sustainability. abu et al. (2019) literature review & survey the authors discuss the barriers in the implementation of lean manufacturing in various malaysian industries ad suggest that 5s, employee training and quality control are essential for harnessing the lean benefits. yadav et al. (2020) case study/survey the authors developed a framework for lean manufacturing implementation in developing economies using the hybrid-fuzzy mathematical modeling tools. int. j. prod. manag. eng. (2020) 8(2), 123-135creative commons attribution-noncommercial-noderivatives 4.0 international a k-chart based implementation framework to attain lean & agile manufacturing 127 http://creativecommons.org/licenses/by-nc-nd/4.0/ (sharifi & zhang, 2001) achieved by astute utilization of the resources. being a strategic process, agile manufacturing breeds winners – ranging from suppliers to end user customer by using the integration of core components (jin-hai et al., 2003). the core components, as described yusuf et al. (1999) are given as follows: core competence management value driven enterprise capability for reconfiguration virtual driven enterprise agile philosophy improves the capability of the supply chain to respond towards customer’s requirements by the virtue of flexibility. thus, the product/service quality is increased, lead times are shortened despite variation in volume, customer satisfaction is increased and the products are delivered on time. the aforementioned qualities lead to increase in productivity and profitability (carvalho et al., 2012; tao & zhang, 2017). table 3. literature review on agile manufacturing. author name & year methodology findings(authors’ points of view) gunasekaran (1998) conceptual model the author suggests that lean manufacturing techniques in conjunction with digital technologies can help in achieving agility. yusuf et al. (2004) conceptual model/ case study the authors suggest that lean and agile practices can be integrated in a harmonious manner to achieve improved business performance. baker (2006) survey/case study design of distribution centers is a critical part of communication and visibility improvement in agile supply chains. iskanius et al. (2006) survey/case study to achieve agility, the flexibility should be focused on operations as opposed to human capital. process integration in supply chain requires switching the mindset. ismail and sharifi (2006) literature review supply chain design interacts with market, supply chain, business environment, technology to support the dynamic characteristics of agile supply chain. vonderembse et al. (2006) case study & conceptual framework the authors considered product lifecycle in which it was concluded that earlier phase requires agile practices, whereas maturity and decline phases require lean principles. baramichai et al. (2007) case study & conceptual framework the authors developed a conceptual framework which helps in improving the business performance through agile means, focusing on supply chain reconfiguration. gunasekaran et al. (2008) framework/ case study for achieving agility in manufacturing, integration with it is necessary for smart working. inman et al. (2011) survey / conceptual framework the authors suggested that lean practice of jit has positive relationship with agile principles in which the operational and marketing performances are improved. costantino et al. (2012) case study the authors discussed the role of decision making in manufacturing systems and suggest that agile practices are necessary to achieve efficient performance. constantinescu et al. (2014) literature review the authors identified the drivers and antecedents of agile manufacturing, and conclude that pragmatic usability applicability and mass customization are of extreme importance in agile manufacturing environment. pawlowski and pawlowski (2015) survey the authors surveyed the polish manufacturing industry to conclude that mass customization, organizational shrewdness and resource flexibility are of vital importance in agile manufacturing. leite and braz (2016) survey & case study the authors suggest that agile manufacturing practices contribute positively to financial performance of the industry, however it remains a relatively unknown approach. sindhwani and malhotra (2017) structural modeling the authors have discussed the enablers of agile manufacturing and suggest that top management commitment, organizational support and it integration are extremely important for agile manufacturing implementation. gunasekaran et al. (2018) case study the authors shared the results of four uk based companies and suggested that agile manufacturing needs to be integrated with bigdata business analytics in order to achieve manufacturing excellence. ghobakhloo and azar (2018) survey the authors conducted a survey with iranian manufacturing companies, developed hypotheses and used structural equation modelling to suggest that lean manufacturing is a precursor to implementation of agile manufacturing, where the former improves financial performance. gunasekaran et al. (2019) literature review the authors suggest that the five major agile competencies include transparent customization supply chains of agile nature, automation, employee empowerment and technology integration. khalfallah and lakhal (2020) survey the authors took responses from 205 tunisian manufacturing companies and suggest that lean approaches supplement the agile practices except for jit delivery, whereas agile manufacturing practices add towards improved operational performance. int. j. prod. manag. eng. (2020) 8(2), 123-135 creative commons attribution-noncommercial-noderivatives 4.0 international zaheer et al. 128 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. results & discussion k-chart was introduced by abdullah et al. (2006) as a simplistic research planning and monitoring technique. it shows the scope, methodologies, key findings/results and timelines by giving them in shape of ladder or tree; thus a detailed micro level layout can be developed through k-chart (abdullah et al.). it should not be confused with the mathematical modelling technique used for statistical process control (gani & limam, 2013; kumar et al., 2006). k chart plays an important role in planning and monitoring through tree diagram. without using k chart, errors such as like delays, incorrect monitoring and inadequate utilization of resources can be encountered. in addition to this, k chart assists in cost reduction, interlink between the whole processes of the system. the main role of k chart depends on layers and each layer describes itself (abdullah et al.). the uses of k-chart found through literature are in table 4. the results in the table 4 indicate that k-chart is an excellent tool for planning and monitoring, but has only been employed in research works and a singular instance of supply chain management. therefore, the authors have used this technique for lean implementation. rafique et al. (2017) have introduced an excellent implementation framework for lean but it does not cater to the requirements of controlling and monitoring. the tool, pareto charts, ishikawa/fish bone/cause and effect, flow chart and gantt chart (jasiulewiczkaczmarek, 2013) are common tools, which can’t provide the clear structure of the system, hence, a new tool is introduced named k chart, which is a useful tool for any research and industry to become sustainable. it provides the clear picture of research scope, explanation of expected and ongoing results and issues, stepwise methodology for any research planning. it is clear that, k chart is a useful tool, which distinguish the different layers and its corresponding outcomes on micro level. a k-chart consists of issues, methodologies, results and time line. a k-chart basically organizes the issues from the broad ones to the specific ones within the area under study. the broader issues are placed at the higher branches of the tree diagram and dissected into various specific issues (sub-issues) underneath it. the issues are then designated into general, complementary and focused issues. 4.1. using lean tool of value stream mapping klean manufacturing has always been at the forefront of improving production performance of an organization; doing so by introducing various tools and techniques to curb the losses. out of the lean tools of vsm, 5s, kanban, andon, tpm etc. vsm (value stream mapping) is of significant importance in which the processes are mapped to evaluate the performance and identify the bottlenecks in the production (rother & shook, 2003). it is proposed that current state map of the process is drawn and checked for areas for improvement (singh et al., 2011). the tool of vsm carries such universality that it does not only cater to the production process, but the services as well (keyte & locher, 2004). the lead times, cycle times and changeover times are calculated, in which the value added activities and non-value added activities are differentiated table 4. literature review on k-chart. author name & year research focused area findings abdullah et al. (2006) research planning & monitoring the authors introduced the techniuque of k-chart for efficient planning and monitoring of research, by using a layered approach in which processes are divided into sub issues and methodologies, enabling the user to achieve time based monitoring. alfaris et al. (2019) enterprise resource planning the authors have performed a literature review in which they have described the importance of k-chart in research planning and monitoring, exhibiting its importance by applying it to their case study based research where the use of k-chart has provided clarity regarding the planning and execution of research. yahaya (2016) power systems the author has used the k-chart technique in planning and monitoring the research work in which the research was divided into various subgroups and subsystems for efficient working and delivery. yen (2013) electronics the author has used k-chart for planning and stringent monitoring of the research work in order to create an electronics based alert system. ter ji-xi (2013) electronics in order to perform a project on object monitoring for people suffering with dementia, the authors employed k-chart technique to achieve favourable results. immawan et al. (2015) supply chain management the authors have developed a framework to assess the supply chain performance of indonesian industries by employing k-chart technique for planning and monitoring of research. int. j. prod. manag. eng. (2020) 8(2), 123-135creative commons attribution-noncommercial-noderivatives 4.0 international a k-chart based implementation framework to attain lean & agile manufacturing 129 http://creativecommons.org/licenses/by-nc-nd/4.0/ (serrano lasa et al., 2008; seth & gupta, 2005). it is advised to simulate the values for validation through power software tool (lian & van landeghem, 2007) so that the issues in data collection are identified and addressed. by answering the eight questions suggested by rother and shook (2003), a future state map is drawn in which the lean wastes of downtime (defects, overproduction, waiting, non-utilized resources, transport, inventory, motion and extra processing) are eliminated and cycle times are reduced (seth & gupta, 2005). the future state map gives a direction regarding the changes in production process or production layout (hines & rich, 1997; rahani & al-ashraf, 2012). the tool of value stream mapping can be applied to any industry regardless of its size or production quantum (chen et al., 2010; grewal, 2008). 4.2. using agile tool of technological unification in case of agile manufacturing, the authors have suggested the use of technological unification in which the processes would be automated, and highly stressful laborious processes would be replaced by technologically intensive processes. agile systems seem provide rapid and cost-effective response to new (unplanned) product model introductions and dynamic capacity allocation to meet unpredictable demand (elkins et al., 2004). in order to achieve the aforesaid, the driving force behind the concept of agile manufacturing pivots on the following points, as described by yusuf et al. (1999): eliminating human error by using automation broadening the customer segment and ensuring the highest quality standards ever-competitive manufacturing climate proactive manufacturing – thinking one step ahead from customers coalescence of manufacturing and management best practices it is a well-known fact that holistic lean manufacturing implementation requires 3-5 years, and requires top management commitment, employee training and embracing the lean thinking approach. similarly, the literature review suggested that in case of agile manufacturing, the successful implementation is incumbent upon the top management commitment to achieve higher operational performance (khalfallah & lakhal, 2020). therefore, a conscious effort should be made towards integrated implementation of both approaches for optimized manufacturing and operational performances. in the current literature, there exists a paucity of approaches that involve implementation of said approaches on system and sub-system levels, divided into various layers (abu et al., 2019; ghobakhloo & azar, 2018). in this regard, k-chart serves as an effective tool due to its strong monitoring capability (abdullah et al., 2006). it involves, problem identification, objective setting, definition of deliverables and setting various milestones. the drivers of lean and agile manufacturing are the independent variables which correspond to the results such as improved production performance, financial gains, operational efficiency, reduced lead times, etc. in comparison to the conventional methods of fishbone diagram, ishikawa, pareto chart, etc. the k-chart takes a holistic view of the implementation of the said approaches. in the case of lean and agile manufacturing implementation, the sub-issues included operational issues, strategic issues and methodology for implementation. consequently, the following layer of the sub issues encompassed manufacturing, organizational nature and production philosophies. this is followed by the methodology layer in which technical and experimental approaches are discussed; in which the former explains the antecedents whereas the latter discusses the dependent variables and the tools used. after successful experimentation, the production optimization process is commenced using lean and agile tools of value stream mapping and technological unification, validated through rigorous simulations. concomitant to that, the real time improvements include lead time reduction, nonvalue-added time reduction, reduction in excessive operations and workforce, and improved valueadded time. therefore, it can be stated that the implementation of k-chart based approach helps immensely in smooth implementation of desired manufacturing approaches and carries the ability to meticulously plan and monitor the progress. the use of k-chart can be extended to project management, program management, research tracking, etc. 5. conclusion through a systematic literature review, the authors have discussed the implementation of k-chart in combined implementation of lean and agile int. j. prod. manag. eng. (2020) 8(2), 123-135 creative commons attribution-noncommercial-noderivatives 4.0 international zaheer et al. 130 http://creativecommons.org/licenses/by-nc-nd/4.0/ manufacturing, using the tools of value stream mapping and technological unification respectively. a k-chart has been drawn which is divided into various layers and sub-layers, eventually culminating in the production optimization. as per the k-chart given in figure 2, the system level is defined which caters to manufacturing, which is divided into three sub issues of operational, methodology based and production optimization to attain sustainable lean & agile manufacturing through value stream mapping operational mechanical (manufacturing) strategicmethodology manufacturing production philosophies organization lean operation production optimization management & support lean deployment responsiveness competency flexibility quickness value stream mapping improved innovation leagile (reduce lead) current & future state map computer simulation simulation results implementation model real system validation cycle time reduction non value added time reduction excessive operations reduction excessive labor reduction improvement in value added time lead time reduction sub-issue 1 (type based) sub-issue 2 (type based) sub-elements methodology layer 1 (technical approach) methodology layer 2 (experimental approach) methodology layer 3 (data type) result layer 1 (model type) result layer 2 (model validation) result layer 3 (expected outcomes) system figure 2. k-chart based lean & agile implementation. int. j. prod. manag. eng. (2020) 8(2), 123-135creative commons attribution-noncommercial-noderivatives 4.0 international a k-chart based implementation framework to attain lean & agile manufacturing 131 http://creativecommons.org/licenses/by-nc-nd/4.0/ strategic, which have been further divided into manufacturing, organizational and production paradigms. the sub-elements succeed the subissues, which have been further divided into three methodology layers. the results layers follow which are divided into three tiers as well, helping in arriving at the most pertinent results. with the increased level of planning and monitoring, the research introduces a new concept of k-chart in manufacturing context that will provide a new direction to researchers. in addition to that, it will provide clarity to practitioners regarding the use of k-chart in implementation of lean and agile techniques for improving the production process. however, the proposed idea requires validation through a case study that is the obvious limitation of this research work. building on this limitation, the future research works can be case study based and a comparative analysis between various implementation techniques and k-chart to provide a clear picture of its efficacy. references abdullah, m.k., mohd suradi, n., jamaluddin, n., mokhtar, a.s., abu talib, a., zainuddin, m.f. (2006). k-chart: a tool for research planning and monitoring. j. of quality management and analysis, 2(1), 123-130. abdullah, m.k., suradi, n.r.m., jamaluddin, n., mokhtar, s., talib, a.r.a., zainuddin, m.f. k-chart: a tool for research planning and monitoring, 7. abdulmalek, f.a., rajgopal, j. (2007). analyzing the benefits of lean manufacturing and value stream mapping via simulation: a process sector case study. international journal of production economics, 107(1), 223-236. https://doi.org/10.1016/j.ijpe.2006.09.009 abu, f., gholami, h., saman, m.z.m., zakuan, n., streimikiene, d. (2019). the implementation of lean manufacturing in the furniture industry: a review and analysis on the motives, barriers, challenges, and the applications. journal of cleaner production, 234, 660-680. https://doi.org/10.1016/j.jclepro.2019.06.279 alfaris, m., edikuncoro, g., savitri, a., yogiari, d., sulistio, j. (2019). a literature review of sustain enterprise resource planning. paper presented at the iop conference series: materials science and engineering. https://doi.org/10.1088/1757-899x/598/1/012128 anand, g., kodali, r. (2010). development of a framework for implementation of lean manufacturing systems. international journal of management practice, 4(1), 95. https://doi.org/10.1504/ijmp.2010.029705 anvari, a., ismail, y., hojjati, s.m.h. (2011). a study on total quality management and lean manufacturing: through lean thinking approach. world applied sciences journal, 12(9), 1585-1596. baker, p. (2006). designing distribution centres for agile supply chains. international journal of logistics research and applications, 9(3), 207-221. https://doi.org/10.1080/13675560600859136 baramichai, m., zimmers, e.w., marangos, c.a. (2007). agile supply chain transformation matrix: an integrated tool for creating an agile enterprise. supply chain management: an international journal, 12(5), 334-348. https://doi.org/10.1108/13598540710776917 bhamu, j., singh sangwan, k. (2014). lean manufacturing: literature review and research issues. international journal of operations & production management, 34(7), 876-940. https://doi.org/10.1108/ijopm-08-2012-0315 bhasin, s., burcher, p. (2006). lean viewed as a philosophy. journal of manufacturing technology management, 17(1), 56-72. https://doi.org/10.1108/17410380610639506 botti, l., mora, c., regattieri, a. (2017). integrating ergonomics and lean manufacturing principles in a hybrid assembly line. computers & industrial engineering, 111, 481-491. https://doi.org/10.1016/j.cie.2017.05.011 carvalho, h., azevedo, s.g., cruz-machado, v. (2012). agile and resilient approaches to supply chain management: influence on performance and competitiveness. logistics research, 4(1-2), 49-62. https://doi.org/10.1007/s12159-012-0064-2 chen, j.c., li, y., shady, b.d. (2010). from value stream mapping toward a lean/sigma continuous improvement process: an industrial case study. international journal of production research, 48(4), 1069-1086. https://doi.org/10.1080/00207540802484911 christian, p.h., zimmers jr, e.w. (1999). age of agile manufacturing puts quality to the test. quality progress, 32(5), 45. chun wu, y. (2003). lean manufacturing: a perspective of lean suppliers. international journal of operations & production management, 23(11), 1349-1376. https://doi.org/10.1108/01443570310501880 constantinescu, c., matarazzo, d., dienes, d., francalanza, e., bayer, m. (2014). modeling of system knowledge for efficient agile manufacturing: tool evaluation, selection and implementation scenario in smes. procedia cirp, 25, 246-252. https://doi.org/10.1016/j.procir.2014.10.035 costantino, n., dotoli, m., falagario, m., fanti, m.p., mangini, a.m. (2012). a model for supply management of agile manufacturing supply chains. international journal of production economics, 135(1), 451-457. https://doi.org/10.1016/j.ijpe.2011.08.021 duguay, c.r., landry, s., pasin, f. (1997). from mass production to flexible/agile production. international journal of operations & production management, 17(12), 1183-1195. https://doi.org/10.1108/01443579710182936 int. j. prod. manag. eng. (2020) 8(2), 123-135 creative commons attribution-noncommercial-noderivatives 4.0 international zaheer et al. 132 https://doi.org/10.1016/j.jclepro.2019.06.279 https://doi.org/10.1088/1757-899x/598/1/012128 https://doi.org/10.1504/ijmp.2010.029705 https://doi.org/10.1080/13675560600859136 https://doi.org/10.1108/13598540710776917 https://doi.org/10.1108/ijopm-08-2012-0315 https://doi.org/10.1108/17410380610639506 https://doi.org/10.1016/j.cie.2017.05.011 https://doi.org/10.1007/s12159-012-0064-2 https://doi.org/10.1080/00207540802484911 https://doi.org/10.1108/01443570310501880 https://doi.org/10.1016/j.procir.2014.10.035 https://doi.org/10.1016/j.ijpe.2011.08.021 https://doi.org/10.1108/01443579710182936 http://creativecommons.org/licenses/by-nc-nd/4.0/ elkins, d.a., huang, n., alden, j.m. (2004). agile manufacturing systems in the automotive industry. international journal of production economics, 91(3), 201-214. https://doi.org/10.1016/j.ijpe.2003.07.006 elnadi, m. (2015). an innovative framework for implementing lean principles in product-service system. fagerholm, f., ikonen, m., kettunen, p., münch, j., roto, v., abrahamsson, p. (2015). performance alignment work: how software developers experience the continuous adaptation of team performance in lean and agile environments. information and software technology, 64, 132-147. https://doi.org/10.1016/j.infsof.2015.01.010 gani, w., limam, m. (2013). on the use of the k-chart for phase ii monitoring of simple linear profiles. journal of quality and reliability engineering, 2013. https://doi.org/10.1155/2013/705450 ghobadian, a., talavera, i., bhattacharya, a., kumar, v., garza-reyes, j.a., o’regan, n. (2018). examining legitimatisation of additive manufacturing in the interplay between innovation, lean manufacturing and sustainability. international journal of production economics. 219, 457-468 ghobakhloo, m., azar, a. (2018). business excellence via advanced manufacturing technology and lean-agile manufacturing. journal of manufacturing technology management, 29(1), 2-24. https://doi.org/10.1108/jmtm-03-2017-0049 goldsby, t.j., griffis, s.e., roath, a.s. (2006). modeling lean, agile, and leagile supply chain strategies. journal of business logistics, 27(1), 57-80. https://doi.org/10.1002/j.2158-1592.2006.tb00241.x grewal, c. (2008). an initiative to implement lean manufacturing using value stream mapping in a small company. international journal of manufacturing technology and management, 15(3-4), 404-417. https://doi.org/10.1504/ijmtm.2008.020176 gunasekaran, a. (1998). agile manufacturing: enablers and an implementation framework. international journal of production research, 36(5), 1223-1247. https://doi.org/10.1080/002075498193291 gunasekaran, a. (1999). agile manufacturing: a framework for research and development. international journal of production economics, 62(1-2), 87-105. https://doi.org/10.1016/s0925-5273(98)00222-9 gunasekaran, a., lai, k., edwincheng, t. (2008). responsive supply chain: a competitive strategy in a networked economy. omega, 36(4), 549-564. https://doi.org/10.1016/j.omega.2006.12.002 gunasekaran, a., yusuf, y.y., adeleye, e.o., papadopoulos, t. (2018). agile manufacturing practices: the role of big data and business analytics with multiple case studies. international journal of production research, 56(1-2), 385-397. https://doi.org/10.1080/00207543.2017.1395488 gunasekaran, a., yusuf, y.y., adeleye, e.o., papadopoulos, t., kovvuri, d., geyi, d.a.g. (2019). agile manufacturing: an evolutionary review of practices. international journal of production research, 57(15-16), 5154-5174. https://doi.org/10.1080/00207543.2018.1530478 hines, p., rich, n. (1997). the seven value stream mapping tools. international journal of operations & production management, 17(1), 4664. https://doi.org/10.1108/01443579710157989 immawan, t., arkeman, y., maulana, a. (2015). sustainable supply chain management for make to stock-make to order production typology case study: batik industry in solo indonesia. supply chain management (sscm), 7(11). inman, r.a., sale, r.s., green, k.w., whitten, d. (2011). agile manufacturing: relation to jit, operational performance and firm performance. journal of operations management, 29(4), 343-355. https://doi.org/10.1016/j.jom.2010.06.001 iskanius, p., haapasalo, h., page, t. (2006). requirements for change in a traditional industry to be competitive: transformation towards an agile supply chain. international journal of agile systems and management, 1(3), 258. https://doi.org/10.1504/ijasm.2006.010942 ismail, h.s., sharifi, h. (2006). a balanced approach to building agile supply chains. international journal of physical distribution & logistics management, 36(6), 431-444. https://doi.org/10.1108/09600030610677384 jasiulewicz-kaczmarek, m. (2013). sustainability: orientation in maintenance management: case study. in p. golinska (ed.), ecoproduction and logistics (pp. 135-154). berlin, heidelberg: springer berlin heidelberg. https://doi.org/10.1007/978-3-642-23553-5_9 jin-hai, l., anderson, a.r., harrison, r.t. (2003). the evolution of agile manufacturing. business process management journal, 9(2), 170189. https://doi.org/10.1108/14637150310468380 keyte, b., locher, d.a. (2004). the complete lean enterprise: value stream mapping for administrative and office processes: productivity press. https://doi.org/10.1201/b16650 khalfallah, m., lakhal, l. (2020). the impact of lean manufacturing practices on operational and financial performance: the mediating role of agile manufacturing. international journal of quality & reliability management. https://doi.org/10.1108/ijqrm-07-2019-0244 khodeir, l.m., othman, r. (2016). examining the interaction between lean and sustainability principles in the management process of aec industry. ain shams engineering journal. konecka, s. (2010). lean and agile supply chain management concept in the aspect of risk management. logforum, 6(4), 23-31. koskela, l. (1992). application of the new production philosophy to construction (72). stanford: stanford university. retrieved from http://www.leanconstruction.org.uk/media/docs/koskela-tr72.pdf kumar, s., choudhary, a., kumar, m., shankar, r., tiwari, m. (2006). kernel distance-based robust support vector methods and its application in developing a robust k-chart. international journal of production research, 44(1), 77-96. https://doi.org/10.1080/00207540500216037 int. j. prod. manag. eng. (2020) 8(2), 123-135creative commons attribution-noncommercial-noderivatives 4.0 international a k-chart based implementation framework to attain lean & agile manufacturing 133 https://doi.org/10.1016/j.ijpe.2003.07.006 https://doi.org/10.1016/j.infsof.2015.01.010 https://doi.org/10.1155/2013/705450 https://doi.org/10.1108/jmtm-03-2017-0049 https://doi.org/10.1002/j.2158-1592.2006.tb00241.x https://doi.org/10.1504/ijmtm.2008.020176 https://doi.org/10.1080/002075498193291 https://doi.org/10.1016/s0925-5273(98)00222-9 https://doi.org/10.1016/j.omega.2006.12.002 https://doi.org/10.1080/00207543.2017.1395488 https://doi.org/10.1080/00207543.2018.1530478 https://doi.org/10.1108/01443579710157989 https://doi.org/10.1016/j.jom.2010.06.001 https://doi.org/10.1504/ijasm.2006.010942 https://doi.org/10.1108/09600030610677384 https://doi.org/10.1007/978-3-642-23553-5_9 https://doi.org/10.1108/14637150310468380 https://doi.org/10.1201/b16650 https://doi.org/10.1108/ijqrm-07-2019-0244 http://www.leanconstruction.org.uk/media/docs/koskela-tr72.pdf https://doi.org/10.1080/00207540500216037 http://creativecommons.org/licenses/by-nc-nd/4.0/ leite, m., braz, v. (2016). agile manufacturing practices for new product development: industrial case studies. journal of manufacturing technology management, 27(4), 560-576. https://doi.org/10.1108/jmtm-09-2015-0073 lian, y.h., van landeghem, h. (2007). analysing the effects of lean manufacturing using a value stream mapping-based simulation generator. international journal of production research, 45(13), 3037-3058. https://doi.org/10.1080/00207540600791590 losonci, d., demeter, k., jenei, i. (2011). factors influencing employee perceptions in lean transformations. international journal of production economics, 131(1), 30-43. https://doi.org/10.1016/j.ijpe.2010.12.022 mamat, r.c., md deros, b., ab rahman, m.n., khalil omar, m., abdullah, s. (2015). soft lean practices for successful lean production system implementation in malaysia automotive smes: a proposed framework. jurnal teknologi, 77(27). https://doi.org/10.11113/jt.v77.6910 maqbool, y., rafique, m.z., hussain, a., ali, h., javed, s., amjad, m.s.,... atif, m. (2019). an implementation framework to attain 6r-based sustainable lean implementation—a case study. ieee access, 7, 117561-117579. https://doi.org/10.1109/access.2019.2936056 marodin, g., frank, a.g., tortorella, g.l., netland, t. (2018). lean product development and lean manufacturing: testing moderation effects. international journal of production economics, 203, 301-310. https://doi.org/10.1016/j.ijpe.2018.07.009 martin, c., towill, d.r. (2000). supply chain migration from lean and functional to agile and customised. supply chain management: an international journal, 5(4), 206-213. https://doi.org/10.1108/13598540010347334 martínez sánchez, a., pérez pérez, m. (2001). lean indicators and manufacturing strategies. international journal of operations & production management, 21(11), 1433-1452. https://doi.org/10.1108/01443570110407436 matt, d.t., rauch, e. (2013). implementation of lean production in small sized enterprises. procedia cirp, 12, 420-425. https://doi.org/10.1016/j.procir.2013.09.072 mccullen, p., towill, d. (2001). achieving lean supply through agile manufacturing. integrated manufacturing systems, 12(7), 524-533. https://doi.org/10.1108/eum0000000006232 mejabi, o.o. (2003). framework for a lean manufacturing planning system. international journal of manufacturing technology and management, 5(5/6), 563. https://doi.org/10.1504/ijmtm.2003.003710 metternich, j., bechtloff, s., seifermann, s. (2013). efficiency and economic evaluation of cellular manufacturing to enable lean machining. procedia cirp, 7, 592-597. https://doi.org/10.1016/j.procir.2013.06.038 moyano-fuentes, j., sacristán-díaz, m., josé martínez-jurado, p. (2012). cooperation in the supply chain and lean production adoption: evidence from the spanish automotive industry. international journal of operations & production management, 32(9), 1075-1096. https://doi.org/10.1108/01443571211265701 näslund, d. (2008). lean, six sigma and lean sigma: fads or real process improvement methods?. business process management journal, 14(3), 269-287. https://doi.org/10.1108/14637150810876634 nawanir, g., kong teong, l., norezam othman, s. (2013). impact of lean practices on operations performance and business performance: some evidence from indonesian manufacturing companies. journal of manufacturing technology management, 24(7), 1019-1050. https://doi.org/10.1108/jmtm-03-2012-0027 nesensohn, c. (2014). an innovative framework for assessing lean construction maturity. liverpool john moores university. nordin, n., deros, b.m., wahab, d.a. (2010). a survey on lean manufacturing implementation in malaysian automotive industry. international journal of innovation, management and technology, 1(4), 374. nordin, n., deros, b.m., wahab, d.a., rahman, m.n.a. (2012). a framework for organisational change management in lean manufacturing implementation. int. j. services and operations management, 12(1), 101-117. https://doi.org/10.1504/ijsom.2012.046676 pawlowski, k., pawlowski, e. (2015). modern manufacturing practices and agile enterprise. anticipated scope of implementation and empirical results from polish enterprises. procedia manufacturing, 3, 464-471. https://doi.org/10.1016/j.promfg.2015.07.209 rafique, m.z., ab rahman, m.n., saibani, n., arsad, n. (2017). a systematic review of lean implementation approaches: a proposed technology combined lean implementation framework. total quality management & business excellence, 30(3-4), 386-421. https://doi.org/10.1080/14783363.2017.1308818 rafique, m.z., ab rahman, m.n., saibani, n., arsad, n., saadat, w. (2016). rfid impacts on barriers affecting lean manufacturing. industrial management & data systems, 116(8), 1585-1616. https://doi.org/10.1108/imds-10-2015-0427 rahani, a.r., al-ashraf, m. (2012). production flow analysis through value stream mapping: a lean manufacturing process case study. procedia engineering, 41, 1727-1734. https://doi.org/10.1016/j.proeng.2012.07.375 rahman, n.a.a., sharif, s.m., esa, m.m. (2013). lean manufacturing case study with kanban system implementation. procedia economics and finance, 7, 174-180. https://doi.org/10.1016/s2212-5671(13)00232-3 rahman, s., laosirihongthong, t., sohal, a.s. (2010). impact of lean strategy on operational performance: a study of thai manufacturing companies. journal of manufacturing technology management, 21(7), 839-852. https://doi.org/10.1108/17410381011077946 rehman, a.u., alkhatani, m., umer, u. (2018). multi criteria approach to measure leanness of a manufacturing organization. ieee access, 6, 20987-20994. https://doi.org/10.1109/access.2018.2825344 robson, c., mccartan, k. (2016). real world research: john wiley & sons. int. j. prod. manag. eng. (2020) 8(2), 123-135 creative commons attribution-noncommercial-noderivatives 4.0 international zaheer et al. 134 https://doi.org/10.1108/jmtm-09-2015-0073 https://doi.org/10.1080/00207540600791590 https://doi.org/10.1016/j.ijpe.2010.12.022 https://doi.org/10.11113/jt.v77.6910 https://doi.org/10.1109/access.2019.2936056 https://doi.org/10.1016/j.ijpe.2018.07.009 https://doi.org/10.1108/13598540010347334 https://doi.org/10.1108/01443570110407436 https://doi.org/10.1016/j.procir.2013.09.072 https://doi.org/10.1108/eum0000000006232 https://doi.org/10.1504/ijmtm.2003.003710 https://doi.org/10.1016/j.procir.2013.06.038 https://doi.org/10.1108/01443571211265701 https://doi.org/10.1108/14637150810876634 https://doi.org/10.1108/jmtm-03-2012-0027 https://doi.org/10.1504/ijsom.2012.046676 https://doi.org/10.1016/j.promfg.2015.07.209 https://doi.org/10.1080/14783363.2017.1308818 https://doi.org/10.1108/imds-10-2015-0427 https://doi.org/10.1016/j.proeng.2012.07.375 https://doi.org/10.1016/s2212-5671(13)00232-3 https://doi.org/10.1108/17410381011077946 https://doi.org/10.1109/access.2018.2825344 http://creativecommons.org/licenses/by-nc-nd/4.0/ rohani, j.m., zahraee, s.m. (2015). production line analysis via value stream mapping: a lean manufacturing process of color industry. procedia manufacturing, 2, 6-10. https://doi.org/10.1016/j.promfg.2015.07.002 rother, m., shook, j. (2003). learning to see: value stream mapping to add value and eliminate muda: lean enterprise institute. sabbagh, o., ab rahman, m.n., ismail, w.r., wan hussain, w.m.h. (2016). methodology implications in automotive product– service systems: a systematic literature review. total quality management & business excellence, 28(13-14), 1632-1668. https://doi.org/10.1080/14783363.2016.1150169 sahwan, m.a., rahman, m.n.a., deros, b.m. (2012). barriers to implement lean manufacturing in malaysian automotive industry. jurnal teknologi, 59, 107-110. https://doi.org/10.11113/jt.v59.2571 salonitis, k., tsinopoulos, c. (2016). drivers and barriers of lean implementation in the greek manufacturing sector. procedia cirp, 57, 189-194. https://doi.org/10.1016/j.procir.2016.11.033 sartal, a., llach, j., vázquez, x.h., de castro, r. (2017). how much does lean manufacturing need environmental and information technologies? journal of manufacturing systems, 45, 260-272. https://doi.org/10.1016/j.jmsy.2017.10.005 saunders, m., lewis, p., thornhill, a. (2009). understanding research philosophies and approaches. research methods for business students, 4, 106-135. saunders, m.n. (2011). research methods for business students, 5/e: pearson education india. serrano lasa, i., ochoa laburu, c., de castro vila, r. (2008). an evaluation of the value stream mapping tool. business process management journal, 14(1), 39-52. https://doi.org/10.1108/14637150810849391 seth, d., gupta, v. (2005). application of value stream mapping for lean operations and cycle time reduction: an indian case study. production planning & control, 16(1), 44-59. https://doi.org/10.1080/09537280512331325281 sharifi, h., zhang, z. (2001). agile manufacturing in practice application of a methodology. international journal of operations & production management, 21(5/6), 772-794. https://doi.org/10.1108/01443570110390462 sindhwani, r., malhotra, v. (2017). a framework to enhance agile manufacturing system. benchmarking: an international journal. https://doi.org/10.1108/bij-09-2015-0092 singh, b., garg, s.k., sharma, s.k. (2011). value stream mapping: literature review and implications for indian industry. the international journal of advanced manufacturing technology, 53(5-8), 799-809. https://doi.org/10.1007/s00170-010-2860-7 srichuachom, u. (2015). the impact of lean approaches to support quality developments in thailand: an investigation of a claim of universality of lean thinking. university of southampton. sriparavastu, l., gupta, t. (1997). an empirical study of just-in-time and total quality management principles implementation in manufacturing firms in the usa. international journal of operations & production management, 17(12), 1215-1232. https://doi.org/10.1108/01443579710182954 sundar, r., balaji, a.n., kumar, r.m.s. (2014). a review on lean manufacturing implementation techniques. procedia engineering, 97, 1875-1885. https://doi.org/10.1016/j.proeng.2014.12.341 swank, c.k. (2003). the lean service machine. harvard business review, 9. tao, f., zhang, m. (2017). digital twin shop-floor: a new shop-floor paradigm towards smart manufacturing. ieee access, 5, 20418-20427. ter ji-xi, j. (2013). object locator for people with dementia (demicator). universiti teknikal malaysia melaka. vonderembse, m.a., uppal, m., huang, s.h., dismukes, j.p. (2006). designing supply chains: towards theory development. international journal of production economics, 100(2), 223-238. https://doi.org/10.1016/j.ijpe.2004.11.014 waters, c.d.j. (2007). supply chain risk management: vulnerability and resilience in logistics. london; philadelphia: kogan page. womack, j.p., jones, d.t. (1996). beyond toyota: how to root out waste and pursue perfection. 13. womack, j.p., jones, d.t., roos, d. (1990). the machine that changed the world. new york, ny: rawson associates. yadav, g., luthra, s., huisingh, d., mangla, s.k., narkhede, b.e., liu, y. (2020). development of a lean manufacturing framework to enhance its adoption within manufacturing companies in developing economies. journal of cleaner production, 245, 118726. https://doi.org/10.1016/j.jclepro.2019.118726 yahaya, s.m.b. (2016). a new investigation method of power quality behavior and improve by using low single phase inverter. universiti teknikal malaysia melaka. yauch, c.a. (2011). measuring agility as a performance outcome. journal of manufacturing technology management, 22(3), 384-404. https://doi.org/10.1108/17410381111112738 yen, s.b.k. (2013). fmna system (forget me not alert system). universiti teknikal malaysia melaka. yusuf, y.y., gunasekaran, a., adeleye, e.o., sivayoganathan, k. (2004). agile supply chain capabilities: determinants of competitive objectives. european journal of operational research, 159(2), 379-392. https://doi.org/10.1016/j.ejor.2003.08.022 yusuf, y.y., sarhadi, m., gunasekaran, a. (1999). agile manufacturing: the drivers, concepts and attributes. 11. int. j. prod. manag. eng. (2020) 8(2), 123-135creative commons attribution-noncommercial-noderivatives 4.0 international a k-chart based implementation framework to attain lean & agile manufacturing 135 https://doi.org/10.1016/j.promfg.2015.07.002 https://doi.org/10.1080/14783363.2016.1150169 https://doi.org/10.11113/jt.v59.2571 https://doi.org/10.1016/j.procir.2016.11.033 https://doi.org/10.1016/j.jmsy.2017.10.005 https://doi.org/10.1108/14637150810849391 https://doi.org/10.1080/09537280512331325281 https://doi.org/10.1108/01443570110390462 https://doi.org/10.1108/bij-09-2015-0092 https://doi.org/10.1007/s00170-010-2860-7 https://doi.org/10.1108/01443579710182954 https://doi.org/10.1016/j.proeng.2014.12.341 https://doi.org/10.1016/j.ijpe.2004.11.014 https://doi.org/10.1016/j.jclepro.2019.118726 https://doi.org/10.1108/17410381111112738 https://doi.org/10.1016/j.ejor.2003.08.022 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2020.12368 received: 2019-09-22 accepted: 2020-06-07 a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks cavdur, f. a1*, kaymaz, e. a2 a department of industrial engineering, bursa uludag university, nilufer, bursa, (turkey). a1 fatihcavdur@uludag.edu.tr, a2 eliifkaymaz@gmail.com abstract: in assembly lines, rework stations are generally used for reprocessing defective items. on the other hand, using rework stations for this purpose only might cause inefficient usage of the resources in this station especially in an assembly line with a low defective rate. in this study, a mixed-integer programming model for cycle time minimization is proposed by considering the use of rework stations for performing parallel tasks. by linearizing the non-linear constraint about parallel tasks using a variate transformation, the model is transformed to a linear-mixed-integer form. in addition to different defective rates, different rework station positions are also considered using the proposed model. the performance of the model is analyzed on several test problems from the related literature. key words: assembly line balancing, cycle time minimization, rework station, parallel tasks, mixed-integer programming. 1. introduction in the last station of an assembly line, it is quite frequent that some quality control procedures are also carried out in addition to the other tasks performed in the station. if a particular product does not pass the quality control check (i.e., a defective product), it is sent to the rework station for performing the necessary corrective operations to transform it into a non-defective product. in assembly lines, rework stations are generally used for performing rework operations of defective products. on the other hand, using rework stations for this purpose only might cause inefficient usage of the resources in this station (i.e., low utilizations of operators, equipment etc.) especially in an assembly line with a low defective rate. it might be possible to increase the efficiency of the resources in a rework station by designing the station for performing standard tasks also in addition to the rework operations. in such a setting, one of the alternative utilizations of a rework station might be that it can be used for performing parallel tasks. in other words, some tasks might be parallelized so that they are assigned to both the rework station and a standard station where the task is performed in only one of these stations in each cycle sequentially (i.e., the task of the first product is performed in the standard station whereas the rework station is used to perform the task of the second product and so on). in this study, we propose an integer programming model considering the utilization of the rework station for parallel tasks in such a setting. the organization of the study is as follows. in section 2, a brief literature review on assembly line balancing and different parallelism concepts in assembly line balancing is presented. section 3 defines the problem considered in the study. we present the details of the proposed mixed-integer programming model in section 4 and a numerical example for illustration in section 5. the performance of the model is tested using some sample problems from the literature as summarized in section 6. section 7 includes the final remarks. to cite this article: cavdur, f., kaymaz, e. (2020). a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. international journal of production management and engineering, 8(2), 109-121. https://doi.org/10.4995/ijpme.2020.12368 int. j. prod. manag. eng. (2020) 8(2), 109-121creative commons attribution-noncommercial-noderivatives 4.0 international 111 https://orcid.org/0000-0001-8054-5606 https://orcid.org/0000-0001-9111-6209 mailto:fatihcavdur@uludag.edu.tr mailto:eliifkaymaz@gmail.com http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. literature review the concept of balancing assembly lines was introduced by bryton in 1954 (bryton, 1954) and the assembly line balancing (alb) problem was defined by salveson in 1955 (salveson, 1955). it is possible to find different studies in the literature about the classification of alb problems (ghosh and gagnon, 1989; sivasankaran and shahabudeen, 2014; boysen et al., 2007). the study of ghosh and gagnon (1989) classifies alb problems as single-model and multi/mixed-model alb problems. a single type of item is produced in a single-model assembly line whereas more than different types of items are produced in a multi/mixed-model assembly line. we can make another classification of alb problem as the problems with deterministic and stochastic task times. an assembly line balancing problem with the following assumptions is defined as the simple assembly line balancing problem (salbp) (baybars, 1986): all input parameters are known with certainty. a task cannot be split among two or more stations. the tasks cannot be processed in arbitrary sequences due to technological precedence requirements. all tasks must be processed. all stations are capable of processing any tasks. task times are independent of the station at which they are performed and of the preceding or following tasks. any task can be processed at any station. assembly line is considered to be serial with no feeder or parallel sub-assembly lines. assembly system is assumed to be designed for a unique model of a single product. cycle time is given and fixed (i.e., salbp-1). number of stations is given and fixed (i.e., salbp-2). some assumptions about the salbp are rather restrictive compared to real-life assembly line systems resulting with the increasing number of studies on generalized assembly line balancing problems (galbp) including various constraints and problem features such as parallel stations and parallel tasks. comprehensive studies regarding the salbp and galbp are presented by becker and scholl (2006) and scholl and becker (2006). in addition to the categorization of the salp as salbp-1 and salbp-2 defined with the aforementioned assumptions, it can be further generalized with respect to the objective function of the problem and divided into four categories as salbp-1, salbp-2, salbp-e and salbp-f where the recently added categories (i.e., salbp-e and salbp-f) considers both the cycle time and the number of stations at the same time for assembly line balancing, differently from salbp-1 where the cycle time is given and fixed and salbp-2 where the number of stations is given and fixed (school and becker, 2006; wei and chao, 2011). although assembly lines are usually classified as straight and u-shaped lines with respect to their designs, some other versions such as parallel lines and two-sided lines are also considered (gokcen et al., 2006; ozcan and toklu, 2010; kara et al., 2011). a straight alb problem is considered in this study. various approaches are used for solving alb problems. some of these are exact methods such as branch-and-bound algorithm and dynamic programming. a comprehensive review of these is presented in the study of boysen et al. (2007). other approximate methods include various heuristics developed for some specific alb problems and metaheuristics used for many different types of problems (battaia and dolgui, 2013). some examples of metaheuristic approaches frequently used in the literature are genetic algorithms (anderson and ferris, 1994), ant colony optimization (sabuncuoglu et al., 2009), simulated annealing (cercioglu et al., 2009) and tabu search (suwannarongsri and puangdownreong, 2008). it is noted that one of the factors affecting the cycle time of an assembly line is about the parallelism concept that can be classified into different categories such as paralleling assembly lines (suer 1998; gokcen et al., 2006), workstations (bard, 1989; askin and zhou, 1997, tiacci et al., 2006, simaria and vilarinho., 2001), tasks (pinto et al., 1975; kaplan, 2004; kazemi et al., 2011) and works (bartholdi, 1993; kim et al., 2000; lee et al., 2001) considering the previous studies in the literature. note that paralleling tasks as considered in this study is defined as the assignment of a task to more than one workstation. we also note the difference between our study and some other studies in which it is possible to divide tasks into smaller units and assign any of these units (not the task itself) to int. j. prod. manag. eng. (2020) 8(2), 109-121 creative commons attribution-noncommercial-noderivatives 4.0 international cavdur & kaymaz 112 http://creativecommons.org/licenses/by-nc-nd/4.0/ different stations whereas a task, as a whole only, can be assigned to more than one station in our study due to the assumption that tasks cannot be divided into smaller units. in other words, a task can be assigned to more than one station (a standard workstation and the rework station, in our case) where consecutive tasks are performed in the alternative workstations sequentially (i.e., 1st task in the standard station, 2nd task in the rework station, 3rd task in the standard station and so on) to balance the workloads of both the corresponding standard workstation and the rework station. we also note that paralleling workstations gains more interest from the researchers compared to paralleling tasks considering the previous studies in the literature focusing on both concepts. although some of the previous studies consider parallel tasks (pinto et al., 1975; kaplan, 2004; kazemi et al., 2011), according to the best of our knowledge, none of them conceptualized and formulated the utilization of the rework station for parallel task assignment constituting the main contribution of our study. balancing an assembly line considering task times only might cause extreme workload on some operators which is one of the main causes of occupational accidents (baykasoglu and akyol, 2014; mutlu and ozgormus, 2012). we note an increasing interest in the consideration of ergonomic factors in assembly line balancing is in recent years. guner and hasgul (2012), for instance, propose an integer programming model considering ergonomic factors. in another study, efe et al. (2014) consider assembly line worker assignment and balancing problem (alwabp) focusing on the workload differences due to the ages and genders of operators. in the study of kara et al., (2014), a model is proposed for integrating some ergonomic factors into assembly line balancing. we finalize this section by noting some remedial actions from the related literature as these studies include some similarities to ours in that some schema (i.e., policy) is proposed to improve some metrics (i.e., cycle time minimization) of an assembly line. some examples of such remedial actions are stopping the line (silverman and carter, 1986), offline repair (gokcen and baykoc, 1999; kottas and lau, 1976), hybrid lines (lau and shtub, 1987) and multiple manning (shtub, 1984). among these remedial actions, the most commonly used are stopping the line and offline repair (altekin et al., 2016). the first one can be defined as stopping the assembly line to complete the missing tasks if the tasks assigned to a station exceed the cycle time of the total operation duration whereas the offline repair remedial action is used for unfinished tasks at the end of the cycle. as the reader might note although the idea of the offline repair remedial action has some conceptual similarity to the one proposed in our study, the methodologies are totally different as the offline repair remedial action is used for unfinished tasks at the end of the cycle to improve the line balance whereas we consider assigning parallel tasks to the rework station for the same purpose which constitutes the main contribution of our work since it is not considered in none of the aforementioned studies as well as the other related papers accessible to us for review. 3. problem description the utilizations the resources in the rework station might vary according to the defective rate of the assembly line. a low defective rate causes inefficient uses of the resources in the rework station. in such cases, it might be advantageous to use the rework station not only for rework operations but also some of the other standard tasks. by using the rework station for this purpose, some of the tasks performed on other standard workstations can also be assigned to the rework station as parallel tasks to be performed on both the corresponding standard workstations as well as the rework station. in other words, a specialized version of parallel task assignment is considered in this study where a parallel task is assigned to both the rework station and a standard station where the task is performed in only one of these stations in each cycle sequentially (i.e., the task of the first product is performed in the standard station whereas the rework station is used to perform the task of the second product and so on). utilization of the rework station for parallel tasks might contribute to the minimization of the cycle time of the assembly line. on the other hand, the defective rate of the assembly line must be considered while assigning standard tasks to the rework station since a low (high) defective rate results in more (less) assignments. in this study, we consider three different defective rates. in addition to the defective rate of the assembly line, the position in which the rework station is located is also considered in this study. since precedence relations of the tasks change the order in which they are performed, the number of tasks that can be assigned to the rework station can vary depending on the position of the rework station. it might be possible to improve the cycle time by changing the position int. j. prod. manag. eng. (2020) 8(2), 109-121creative commons attribution-noncommercial-noderivatives 4.0 international a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks 113 http://creativecommons.org/licenses/by-nc-nd/4.0/ of the rework station. within the scope of the study, three different alternative designs are created where the rework station is located in the last three station positions as illustrated in figure 1, figure 2 and figure 3, respectively where the flows to the rework station are colored as red for defective products and blue for the standard products (i.e., parallel tasks). in figure 1, the rework station is in last station position (i.e., the (n+1)th station position) in an assembly line including n standard workstations. in a similar setting in figure 2, the rework station is in the nth station position and it serves as a workstation where both the corrective operations for the defective products coming from the last station of the assembly line as well as the parallel tasks ws n ws n + 1 (rw station) defective products ws n – 1 standard products (parallel tasks) corrected and standard products standard products ws n – 2 ws 1 end of line figure 1. positioning the rework station as the (n + 1)th station. ws n + 1 ws n (rw station) defective products ws n – 1 standard products (parallel tasks) standard products corrected products ws n – 2 ws 1 end of line figure 2. positioning rework station as the nth station. ws n ws n – 1 (rw station) defective products ws n – 2 standard products (parallel tasks) standard products corrected products ws n + 1 end of line ws 1 figure 3. positioning rework station as the (n – 1)th station. int. j. prod. manag. eng. (2020) 8(2), 109-121 creative commons attribution-noncommercial-noderivatives 4.0 international cavdur & kaymaz 114 http://creativecommons.org/licenses/by-nc-nd/4.0/ for the products coming from the previous station (i.e., (n – 1)th station) are performed. in figure 3, the rework station is located in the (n – 1)th station position where it serves similarly at its new position. in addition to the aforementioned positions, the rework station can be moved to the previous station positions; however, since the rework station is primarily used as a station where defective products from the last station are corrected, moving away from the last station position increases the transportation times and distances of the defective products requiring corrections. on the other hand, instead of being located at the last station position, the number of potential tasks that can be assigned to the rework station can be increased by placing the rework station in positions near the last station (such as the (n – 1)th station, the (n – 2)th station positions) depending on the precedence relations constraints. it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4. methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: i tasks, i = 1, …, m j workstations j = 1, … , n parameters: m total number of tasks n total number of workstations ti processing time for task i pik precedence relation matrix element, equals 1 if task i is predecessor of task k and 0 otherwise r rework station position α penalty for parallel task assignment β defective rate coefficient, 1 ≤ β ≤ 1.75 λ scaling factor for the objective function terms γ penalty used to limit the number of jobs that can be assigned to the rework station variables: c cycle time xij equals 1 if task is assigned to station j; 0 otherwise yi equals 1 if task i is assigned to the rework station; 0 otherwise zij equals 1 if task i is assigned to both the rework station and station j; 0 otherwise objective function: a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (1) constraints: a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (2) a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (3) a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (4) a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (5) a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (6) a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (7) a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (8) a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (9) a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks. 6 | int. j. prod. manag. eng. (yyyy) vv(nn), ppp-ppp creative commons attribution-noncommercial 3.0 spain it is noted from the foregoing discussion that when the position of the rework station moves towards to the first station position, we might have more flexibility in assigning tasks to the rework station depending on the precedence relations, but the transportation times/distances of the corrective operations increase. on the contrary, if the rework station position moves towards to the last station position, the flexibility might be lost in task assignments, but transportation times/distances of the corrective operations decrease in return. 4 methodology this section details the integer programming model developed to minimize the cycle time for the problem described in the previous pages. indexes: 𝑖𝑖𝑖𝑖 tasks, 𝑖𝑖𝑖𝑖 = 1, … , 𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗 workstations 𝑗𝑗𝑗𝑗 = 1, … , 𝑛𝑛𝑛𝑛 parameters: 𝑚𝑚𝑚𝑚 total number of tasks 𝑛𝑛𝑛𝑛 total number of workstations 𝑡𝑡𝑡𝑡+ processing time for task 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝+precedence relation matrix element, equals 1 if task 𝑖𝑖𝑖𝑖 is predecessor of task 𝑘𝑘𝑘𝑘 and 0 otherwise 𝑟𝑟𝑟𝑟 rework station position 𝛼𝛼𝛼𝛼 penalty for parallel task assignment 𝛽𝛽𝛽𝛽 defective rate coefficient, 1 ≤ 𝛽𝛽𝛽𝛽 ≤ 1.75 𝜆𝜆𝜆𝜆 scaling factor for the objective function terms 𝛾𝛾𝛾𝛾 penalty used to limit the number of jobs that can be assigned to the rework station variables: 𝑐𝑐𝑐𝑐 cycle time 𝑥𝑥𝑥𝑥+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to station 𝑗𝑗𝑗𝑗; 0 otherwise 𝑦𝑦𝑦𝑦+ equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to the rework station; 0 otherwise 𝑧𝑧𝑧𝑧+: equals 1 if task 𝑖𝑖𝑖𝑖 is assigned to both the rework station and station 𝑗𝑗𝑗𝑗; 0 otherwise objective function: min 𝑧𝑧𝑧𝑧 = 𝜆𝜆𝜆𝜆𝑐𝑐𝑐𝑐 + (1 − 𝜆𝜆𝜆𝜆) de 𝛼𝛼𝛼𝛼𝑦𝑦𝑦𝑦+ f +gh i (1) constraints: 1 ≤ e 𝑥𝑥𝑥𝑥+: jkh :gh ≤ 2, ∀𝑖𝑖𝑖𝑖 (2) 𝑥𝑥𝑥𝑥-: ≤ 1 − 𝑥𝑥𝑥𝑥+n, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗, 𝑘𝑘𝑘𝑘, 𝑞𝑞𝑞𝑞: 𝑝𝑝𝑝𝑝+= 1; ∀𝑞𝑞𝑞𝑞 ≥ 𝑗𝑗𝑗𝑗 + 1 (3) e s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh + e 𝑡𝑡𝑡𝑡+ (1 − 𝑦𝑦𝑦𝑦+)𝑥𝑥𝑥𝑥+: f +gh ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 (4) 𝛽𝛽𝛽𝛽u de s 𝑡𝑡𝑡𝑡+ 2 t 𝑥𝑥𝑥𝑥+:𝑦𝑦𝑦𝑦+ f +gh i ≤ 𝑐𝑐𝑐𝑐, ∀𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (5) 𝑥𝑥𝑥𝑥+: = 𝑦𝑦𝑦𝑦+, ∀𝑖𝑖𝑖𝑖; 𝑗𝑗𝑗𝑗 = 𝑟𝑟𝑟𝑟 (6) 𝑦𝑦𝑦𝑦+ = e 𝑥𝑥𝑥𝑥+: jkh :gh − 1, ∀𝑖𝑖𝑖𝑖 (7) 𝑥𝑥𝑥𝑥+: ∈ {0,1}, ∀𝑖𝑖𝑖𝑖, 𝑗𝑗𝑗𝑗 (8) 𝑦𝑦𝑦𝑦+ ∈ {0,1}, ∀𝑖𝑖𝑖𝑖 (9) 𝑐𝑐𝑐𝑐 ≥ 0 (10) (10) the objective function in equation (1) includes the weighted sum of the cycle time and the total number of parallel tasks. in this expression, α is the penalty for each parallel task and λ ∈ [0,1] is int. j. prod. manag. eng. (2020) 8(2), 109-121creative commons attribution-noncommercial-noderivatives 4.0 international a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks 115 http://creativecommons.org/licenses/by-nc-nd/4.0/ defined as the scaling factor between the objective function components. in order to emphasize the effects of defective rate and rework station position, this study has been carried out with the assumption that there is no negative effects of parallel task assignment (i.e., for α = 0 and λ = 1 ) and the effects of these parameters similarly can be analyzed in future studies. the constraint given by equation (2) ensures that tasks are assigned to at most two stations and at least one station. in other words, if a task is assigned to the rework station, (i.e., if it is a parallel task), it is then assigned to another standard station (i.e., the task is assigned to two stations). if it is not a parallel task, it can only be assigned to one station. the constraint given in equation (3) indicates the precedence relations between tasks. it is ensured with equation (4) that the total processing times of the tasks assigned to a station do not exceed the cycle time. since parallel tasks are assigned to two stations at the same time, half of the total processing times of the tasks is taken into consideration when calculating the cycle time. similarly, the tasks that are assigned to the rework station are limited depending on the defective rate of the assembly line using equation (5) where defective rate coefficient β ∈ [1,1.75] in the equation reserves some time for rework operations (i.e., corrective operations) where the extreme values (1 and 2) represent the cases in which no rework operations are performed at all and the rework station only performs rework operations. the other parameter γ is defined as the penalty used to limit the number of tasks that can be assigned to the rework station for various reasons, such as ergonomic factors due the rework station position. it might be desired that the number of jobs assigned to the rework station might be limited in the case that the rework station moves towards to the first station position by defining γ = n / r as the ratio of the total number of stations to the rework station position. it is noted that with β γ factor, the number of tasks assigned to the rework station can be limited beyond the defective rate since γ = n / r parameter takes larger values as the rework station moves towards to the first station position. in the context of this study, all computations are performed with the assumption of no effects of assigning parallel tasks or moving the rework station (i.e., γ = 1). a detailed sensitivity analysis can be performed in future studies to examine the effects of these parameters. equation (6) and equation (7) show the relationships between the corresponding variables by ensuring that if a task is assigned to the rework station, it must be a parallel task assigned to another standard workstation. variable definitions are given by equation (8), equation (9), and equation (10). note that the constraints in equation (4) and equation (5) are nonlinear constraints due to the term of xijyi, for all i and j. we however note that it can be easily linearized since both variables are binary and a linear model can be obtained by introducing the new variables defined as zij=xijyi, for all i and j, as shown in equation (11), equation (12) and equation (13). zij ≥ xij+yi–1, ∀ i,j (11) zij ≤ xij, ∀ i,j (12) zij ≤ yi, ∀ i,j (13) 5. numerical example we illustrate the proposed model using the jackson test problem with 11 tasks and three and four stations from the literature. the precedence relations the jackson sample together with the processing times of the tasks is shown in figure 4. we solve the problem for three different defective rates and rework station positions as detailed in the previous section. note that the rework station is added as an additional station to the original problem. in other words, in the threestation version of the jackson sample, the rework station added as the fourth station, and thus, three setting considered are the ones where the rework station is located in the second, third and fourth station positions. similarly, the model is solved for three different defective rates (i.e., β=1.25, β=1.5, and β=1.75) in addition to the case of zero defective production (i.e., β=1.25 ) in order to show the effect of the defective rate. the results are given in table 1. in addition, the results obtained for different rework station positions (for r = 4, r = 3 and, r = 2 respectively) are shown in figure 5, figure 6 and figure 7 for a defective rate coefficient of β=1.25 where we represent the standard tasks in white, parallel tasks in light gray and rework operations in dark gray. the effect of changing the position of the rework station in the figures is clearly observed. when the rework station is in the last station position, the number of potential tasks that can be assigned to the rework station is more limited depending on precedence relations and only a parallel task (with task number 11) can be assigned as shown in figure 5. accordingly, it is seen that the total time of the tasks int. j. prod. manag. eng. (2020) 8(2), 109-121 creative commons attribution-noncommercial-noderivatives 4.0 international cavdur & kaymaz 116 http://creativecommons.org/licenses/by-nc-nd/4.0/ performed in the rework station is considerably less than the cycle time. on the other hand, by changing the position of the rework station by locating it in the third and second positions, it becomes possible to assign more tasks to the rework station, which makes it possible to achieve more improvements in the cycle time as seen in figure 6 and figure 7, respectively. 1(6) 2(2) 3(5) 4(7) 5(1) 6(2) 7(3) 8(6) 9(5) 10(5) 11(4) figure 4. precedence diagram and process times for jackson data set. the results in table 1 show how the optimal solution changes depending on the position of the rework station and the defective rate of the assembly line. it is noted, for a 25% of defective rate (i.e., ) for instance, that when the rework station is in the second and third workstation positions, the cycle time is 12.5 time units whereas it is 15 time units when the rework station is in the fourth station position at the end of the assembly line. similarly, increases in defective rate, also increases the cycle time due to more rework operations performed in the station. in addition to the jackson sample, the test problems of mitchells, sawyer and kilbrid are also solved for different number of stations to show the performance of the model and presented in the next section. 6. computational results in this section, the performance of the model is tested using various alb problem test samples (i.e., β=1.00 the test problems of jackson, mitchells, heskiaoff, sawyer, kilbrid) from the literature. the results are summarized in table 2. the data about the test problems used in the study can be accessed at http://assembly-line-balancing.mansci.de 1 4 7 11 2 6 9 3 8 10 5 11 0 2 4 6 8 10 12 14 16 1 2 3 4 ti m e station figure 5. assignments for β=1.25 and r=4. 1 2 3 7 3 6 4 9 4 8 5 11 5 10 10 0 2 4 6 8 10 12 14 1 2 3 4 ti m e station figure 6. assignments for β=1.25 and r=3. table 1. results for the jackson sample. ns ct st rw station position β=1.00 β=1.25 β=1.50 β=1.75 ct st ct st ct st ct st 3+1 16.00 0.23 2 12.00 0.27 12.50 0.03 13.00 0.30 13.00 0.10 3 12.50 0.22 12.50 0.06 13.00 0.27 13.00 0.18 4 15.00 0.17 15.00 0.03 15.00 0.20 15.00 0.03 4+1 12.00 0.11 3 9.50 0.25 10.00 0.13 10.50 0.28 10.50 0.10 4 9.50 0.26 10.00 0.10 10.50 0.28 10.50 0.08 5 11.00 0.11 11.00 0.03 11.00 0.11 11.00 0.02 ns (number of stations), ct (cycle time-time unit), st (solution time-second). int. j. prod. manag. eng. (2020) 8(2), 109-121creative commons attribution-noncommercial-noderivatives 4.0 international a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks 117 http://assembly-line-balancing.mansci.de http://creativecommons.org/licenses/by-nc-nd/4.0/ test problems are solved using gurobi optimizer in mathematical programming language (mpl) on a personal computer (intel (r) core i7-7500 cpu 2.70 ghz 2.90 ghz). as in the previous section, all computations are performed for α = 0, λ = 1 and γ = 1 in order to emphasize the effects of defective rate and rework station position. in future studies, the effects of these parameters can be similarly investigated. the results are summarized in table 2 where two different number-of-stations combinations from the study of ugurdag et al., (1997) are taken into consideration. 1 3 4 8 2 4 5 10 3 5 7 11 6 8 9 0 2 4 6 8 10 12 14 1 2 3 4 ti m e station figure 7. assignments for β=1.25 and r=2. table 2. computational results. problem ns ct st rw station position β=1.00 β=1.25 β=1.50 β=1.75 ct st ct st ct st ct st jackson (11 task) 3+1 16.00 0.23 2 12.00 0.27 12.50 0.03 13.00 0.30 13.00 0.10 3 12.50 0.22 12.50 0.06 13.00 0.27 13.00 0.18 4 15.00 0.17 15.00 0.03 15.00 0.20 15.00 0.03 4+1 12.00 0.11 3 9.50 0.25 10.00 0.13 10.50 0.28 10.50 0.10 4 9.50 0.26 10.00 0.10 10.50 0.28 10.50 0.08 5 11.00 0.11 11.00 0.03 11.00 0.11 11.00 0.02 mitchell (21 task) 3+1 35.00 ≈ 0.00 2 31.00 0.05 31.00 0.11 31.00 0.03 31.00 0.09 3 31.00 0.06 31.00 0.10 31.00 0.05 31.00 0.10 4 33.50 0.01 33.50 0.04 33.50 0.02 33.50 0.05 5+1 21.00 0.02 4 18.50 0.17 18.50 0.37 19.00 0.17 19.50 0.32 5 18.00 0.13 18.50 0.18 19.00 0.22 19.50 0.26 6 20.50 0.05 20.50 0.09 20.50 0.03 20.50 0.07 heskiaoff (28 task) 4+1 256.00 0.01 3 205.00 0.42 213.50 2.57 219.50 0.98 224.00 1.28 4 205.00 0.21 213.50 1.82 219.50 1.78 224.00 0.26 5 247.00 0.03 247.00 0.07 247.00 0.36 247.00 0.05 5+1 205.00 0.06 4 171.00 0.28 176.88 3.10 181.00 1,080.00 184.00 2.79 5 171.00 0.34 176.88 3.73 181.00 5.47 184.00 1.65 6 198.00 0.05 198.00 0.08 198.00 0.23 198.00 0.12 sawyer (30 task) 5+1 65.00 0.08 4 55.00 2.27 56.00 4.71 57.50 2.98 58.50 4.05 5 57.00 1.47 57.50 2.97 57.50 2.65 58.63 3.57 6 61.00 0.09 61.00 0.10 61.00 0.16 61.00 0.10 8+1 41.00 0.08 7 37.00 8.59 37.50 7.19 38.00 4.67 38.00 7.67 8 36.50 2.84 37.00 6.50 38.00 8.88 38.00 8.52 9 39.00 2.01 39.00 2.12 39.00 3.10 39.00 3.66 kilbrid (45 task) 3+1 184.00 0.33 2 141.00 0.58 145.50 0.85 150.75 0.60 154.88 1.17 3 138.00 0.55 145.50 1.30 150.75 0.94 154.88 3.77 4 182.00 0.25 182.00 0.08 182.00 0.05 182.00 0.09 6+1 92.00 0.20 5 79.00 1.48 81.50 4.31 83.00 3.80 84.00 5.04 6 79.00 2.03 81.50 2.38 83.00 4.16 84.00 2.58 7 91.00 0.15 91.00 0.55 91.00 0.41 91.00 0.41 ns (number of stations), ct (cycle time-time unit), st (solution time-second). int. j. prod. manag. eng. (2020) 8(2), 109-121 creative commons attribution-noncommercial-noderivatives 4.0 international cavdur & kaymaz 118 http://creativecommons.org/licenses/by-nc-nd/4.0/ the computational results in this section are presented in a similar form to the results of the numerical example given for the jackson sample in the previous section. in addition to modified versions of the problems with the added rework stations, the original test problems (the ones without the rework stations) are also solved. similar to the numerical example computations given in the previous section, the problems are solved for three different defective rates (i.e., β=1.25, β=1.50 and β=1.75) in addition to the case of zero defective production (i.e., β=1.00). similarly, three rework station positions are considered by positioning it as the last three stations of the assembly line. all problems are solved to optimality with the corresponding solution times varying between fractions of a second (for various problems such as the all versions of the jackson sample with 11 tasks and mitchells sample with 21 tasks) and 1,080 seconds (for the heskiaoff sample with 28 tasks and 6 stations where the rework station is located in the 4th station position). it is also noted however that the maximum solution time of 1,080 seconds seems to be an outlier since the optimal solutions for all other cases (even for larger samples of sawyer with 30 tasks and kilbrid with 45 tasks) are obtained in significantly smaller durations (i.e., less than 10 seconds). we can observe the effects of different defective rates and rework station positions in table 2. 7. conclusions in this study, we propose a mixed-integer programming model for minimizing the cycle time of an assembly line considering the use of the rework stations for parallel tasks in addition to the rework operations. using the proposed model, the tasks are assigned to the rework station and a standard workstation in parallel to utilize the resources in the rework station. by linearizing the nonlinear constraints of the proposed model, it is transformed into a linear-mixed-integer program. we test the model using some test problems from the literature where we analyze the effects of different defective rates and the rework station position. on the other hand, it is noted that the applicability of the proposed model might be limited in some real-life environments due to the difficulties about parallel task implementation in the assembly lines without a particular level of automation. in such situations, it becomes even more important to consider human factors especially for the workers responsible for performing parallel tasks and rework operations. nevertheless, it might be easier to deal with such ergonomics difficulties in the near future with the technological developments yielding highly automated smart production systems. since the proposed model is obtained by variable transformation for linearization, model size is significantly increased compared to the original nonlinear model. as a result of this increase, applying the model on larger-size problems is expected to result in longer solution times. developing some heuristic and meta-heuristic approaches to deal larger-size problems is important in future studies. in this study, in order to emphasize the effects of defective rate and rework station position, all computations all computations are performed with the corresponding parameter combination setting (i.e., α = 0, λ = 1 and γ = 1) ignoring the potential negative effects of parallel task assignment to be considered in future studies. in addition, more comprehensive experiments can be considered in the future to analyze the effects of the defective rate and rework station position. finally, the validation of the solutions obtained by the proposed approach using simulation might also be useful to analyze the bottleneck effects of parallel task assignment especially for the problems involving uncertainty. references altekin, f.t., bayindir, z.p., gumuskaya, v. (2016). remedial actions for disassembly lines with stochastic task times. computers & industrial engineering, 99, 78-96. https://doi.org/10.1016/j.cie.2016.06.027 anderson, e.j., ferris, m.c. (1994). genetic algorithms for combinatorial optimization: the assemble line balancing problem. orsa journal on computing, 6(2), 161-173. https://doi.org/10.1287/ijoc.6.2.161 askin, r.g., zhou, m. (1997). a parallel station heuristic for the mixed-model production line balancing problem. international journal of production research, 35(11), 3095-3106. https://doi.org/10.1080/002075497194309 bard, j.f. (1989). assembly line balancing with parallel workstations and dead time. the international journal of production research, 27(6), 1005-1018. https://doi.org/10.1080/00207548908942604 int. j. prod. manag. eng. (2020) 8(2), 109-121creative commons attribution-noncommercial-noderivatives 4.0 international a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks 119 https://doi.org/10.1016/j.cie.2016.06.027 https://doi.org/10.1287/ijoc.6.2.161 https://doi.org/10.1080/002075497194309 https://doi.org/10.1080/00207548908942604 http://creativecommons.org/licenses/by-nc-nd/4.0/ bartholdi, j.j. (1993). balancing two-sided assembly lines: a case study. international journal of production research, 31(10), 2447-2461. https://doi.org/10.1080/00207549308956868 battaia, o., dolgui, a. (2013). a taxonomy of line balancing problems and their solution approaches. international journal of production economics, 142(2), 259-277. https://doi.org/10.1016/j.ijpe.2012.10.020 baybars, i. (1986). a survey of exact algorithms for the simple assembly line balancing problem. management science, 32(8), 909-932. https://doi.org/10.1287/mnsc.32.8.909 baykasoglu, a., demirkol akyol, s. (2014). ergonomic assembly line balancing. journal of the faculty of engineering and architecture of gazi university, 29(4), 785-792. https://doi.org/10.17341/gummfd.00296 becker, c., scholl, a. (2006). a survey on problems and methods in generalized assembly line balancing. european journal of operational research, 168(3), 694-715. https://doi.org/10.1016/j.ejor.2004.07.023 boysen, n., fliedner, m., scholl, a. (2007). a classification of assembly line balancing problems. european journal of operational research, 183(2), 674-693. https://doi.org/10.1016/j.ejor.2006.10.010 bryton, b. (1954). balancing of a continuous production line. master’s thesis, northwestern university, evanston. cercioglu, h., ozcan, u., gokcen, h., toklu, b. (2009). a simulated annealing approach for parallel assembly line balancing problem. journal of the faculty of engineering and architecture of gazi university, 24(2), 331-341. efe, b., kremer, g.e.o., kurt, m. (2018). age and gender-based workload constraint for assembly line worker assignment and balancing problem in a textile firm. international journal of industrial engineering, 25(1), 1-17. ghosh, s., gagnon, r.j. (1989). a comprehensive literature review and analysis of the design, balancing and scheduling of assembly systems. the international journal of production research, 27(4), 637-670. https://doi.org/10.1080/00207548908942574 gokcen, h., baykoc, ö.f. (1999). a new line remedial policy for the paced lines with stochastic task times. international journal of production economics, 58(2), 191-197. https://doi.org/10.1016/s0925-5273(98)00123-6 gokcen, h., agpak, k., benzer, r. (2006). balancing of parallel assembly lines. international journal of production economics, 103(2), 600609. https://doi.org/10.1016/j.ijpe.2005.12.001 guner, b., hasgul, s. (2012). u-type assembly line balancing with ergonomic factors for balance stability. journal of the faculty of engineering and architecture of gazi university, 27(2), 407-415. kaplan, o. (2004). assembly line balancing with task paralleling. master’s thesis, metu, ankara. kara, y., ozguven, c., yalcın, n., atasagun, y. (2011). balancing straight and u-shaped assembly lines with resource dependent task times. international journal of production research, 49(21), 6387-6405. https://doi.org/10.1080/00207543.2010.535039 kara, y., atasagun, y., gokcen, h., hezer, s., demirel, n. (2014). an integrated model to incorporate ergonomics and resource restrictions into assembly line balancing. international journal of computer integrated manufacturing, 27(11), 997-1007. https://doi.org/10.1080/0951192x.2013.874575 kazemi, s.m., ghodsi, r., rabbani, m., tavakkoli-moghaddam, r. (2011). a novel two-stage genetic algorithm for a mixed-model u-line balancing problem with duplicated tasks. the international journal of advanced manufacturing technology, 55(9-12), 1111-1122. https://doi.org/10.1007/s00170-010-3120-6 kim, y.k., kim, y., kim, y.j. (2000). two-sided assembly line balancing: a genetic algorithm approach. production planning & control, 11(1), 44-53. https://doi.org/10.1080/095372800232478 kottas, j.f., lau, h.s. (1976). a total operating cost model for paced lines with stochastic task times. aiie transactions, 8(2), 234-240. https://doi.org/10.1080/05695557608975072 lau, h.s., shtub, a. (1987). an exploratory study on stopping a paced line when incompletions occur. iie transactions, 19(4), 463-467. https://doi.org/10.1080/07408178708975421 lee, t.o., kim, y., kim, y.k. (2001). two-sided assembly line balancing to maximize work relatedness and slackness. computers & industrial engineering, 40(3), 273-292. https://doi.org/10.1016/s0360-8352(01)00029-8 mutlu, o., ozgormus, e. (2012). a fuzzy assembly line balancing problem with physical workload constraints. international journal of production research, 50(18), 5281-5291. https://doi.org/10.1080/00207543.2012.709647 ozcan, u., toklu, b. (2010). balancing two-sided assembly lines with sequence-dependent setup times. international journal of production research, 48(18), 5363-5383. https://doi.org/10.1080/00207540903140750 pinto, p., dannenbring, d.g., khumawala, b.m. (1975). a branch and bound algorithm for assembly line balancing with paralleling. the international journal of production research, 13(2), 183-196. https://doi.org/10.1080/00207547508942985 sabuncuoglu, i., erel, e., alp, a. (2009). ant colony optimization for the single model u-type assembly line balancing problem. international journal of production economics, 120(2), 287-300. https://doi.org/10.1016/j.ijpe.2008.11.017 salveson, m.e. (1955). the assembly line balancing problem. the journal of industrial engineering, 18-25. scholl, a., becker, c. (2006). state-of-the-art exact and heuristic solution procedures for simple assembly line balancing. european journal of operational research, 168(3), 666-693. https://doi.org/10.1016/j.ejor.2004.07.022 int. j. prod. manag. eng. (2020) 8(2), 109-121 creative commons attribution-noncommercial-noderivatives 4.0 international cavdur & kaymaz 120 https://doi.org/10.1080/00207549308956868 https://doi.org/10.1016/j.ijpe.2012.10.020 https://doi.org/10.1287/mnsc.32.8.909 https://doi.org/10.17341/gummfd.00296 https://doi.org/10.1016/j.ejor.2004.07.023 https://doi.org/10.1016/j.ejor.2006.10.010 https://doi.org/10.1080/00207548908942574 https://doi.org/10.1016/s0925-5273(98)00123-6 https://doi.org/10.1016/j.ijpe.2005.12.001 https://doi.org/10.1080/00207543.2010.535039 https://doi.org/10.1080/0951192x.2013.874575 https://doi.org/10.1007/s00170-010-3120-6 https://doi.org/10.1080/095372800232478 https://doi.org/10.1080/05695557608975072 https://doi.org/10.1080/07408178708975421 https://doi.org/10.1016/s0360-8352(01)00029-8 https://doi.org/10.1080/00207543.2012.709647 https://doi.org/10.1080/00207540903140750 https://doi.org/10.1080/00207547508942985 https://doi.org/10.1016/j.ijpe.2008.11.017 https://doi.org/10.1016/j.ejor.2004.07.022 http://creativecommons.org/licenses/by-nc-nd/4.0/ shtub, a. (1984). the effect of incompletion cost on line balancing with multiple manning of work stations. the international journal of production research, 22(2), 235-245. https://doi.org/10.1080/00207548408942450 silverman, f.n., carter, j.c. (1986). a cost-based methodology for stochastic line balancing with intermittent line stoppages. management science, 32(4), 455-463. https://doi.org/10.1287/mnsc.32.4.455 simaria, a.s., vilarinho, p.m. (2001). the simple assembly line balancing problem with parallel workstations-a simulated annealing approach. int j ind eng-theory, 8(3), 230-240. sivasankaran, p., shahabudeen, p. (2014). literature review of assembly line balancing problems. the international journal of advanced manufacturing technology, 73(9-12), 1665-1694. https://doi.org/10.1007/s00170-014-5944-y suer, g.a. (1998). designing parallel assembly lines. computers & industrial engineering, 35(3-4), 467-470. suwannarongsri, s., puangdownreong, d. (2008). optimal assembly line balancing using tabu search with partial random permutation technique. international journal of management science and engineering management, 3(1), 3-18. https://doi.org/10.1080/175096 53.2008.10671032 tiacci, l., saetta, s., martini, a. (2006). balancing mixed-model assembly lines with parallel workstations through a genetic algorithm approach. international journal of industrial engineering, 13(4), 402. ugurdag, h.f., rachamadugu, r., papachristou, c.a. (1997). designing paced assembly lines with fixed number of stations. european journal of operational research, 102(3), 488-501. https://doi.org/10.1016/s0377-2217(96)00248-2 wei, n.c., chao, i.m. (2011). a solution procedure for type e simple assembly line balancing problem. computers & industrial engineering, 61(3), 824-830. https://doi.org/10.1016/j.cie.2011.05.015 int. j. prod. manag. eng. (2020) 8(2), 109-121creative commons attribution-noncommercial-noderivatives 4.0 international a mixed-integer programming model for cycle time minimization in assembly line balancing: using rework stations for performing parallel tasks 121 https://doi.org/10.1080/00207548408942450 https://doi.org/10.1287/mnsc.32.4.455 https://doi.org/10.1007/s00170-014-5944-y https://doi.org/10.1080/17509653.2008.10671032 https://doi.org/10.1080/17509653.2008.10671032 https://doi.org/10.1016/s0377-2217(96)00248-2 https://doi.org/10.1016/j.cie.2011.05.015 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.16140 received: 2021-08-29 accepted: 2022-04-07 enhance quality improvement through lean six sigma in division side board clavinova piano’s hernadewitaa*, indra setiawanb, hendrac ab master of industrial engineering, universitas mercu buana, jl. meruya selatan no.1, jakarta 11650, indonesia. c mechanical engineering, universitas sultan ageng tirtatayasa, banten, indonesia. ab* hernadewita@mercubuana.ac.id, c hendra@untirta.ac.id abstract: a lean production system is one of the main factors that every company must have to improve quality, especially reducing waste that occurs in the production line. the musical instrument industry such as piano manufacturing includes industries that reduce waste by improving the manufacturing system. the company’s strategy to eliminate this waste is to create an effective and efficient work system. the piano production process still has several problems, such as the presence of waste. this is an obstacle in improving the quality of production by reducing the waste that occurs. for this reason, this research focuses on improving the quality of production by reducing waste that occurs, finding the cause of the problem and taking corrective action. the method used in this research is the integration of lean and six sigma and the dmaic framework. the process is waste flow and production problems are identified in define phase and residual measurement is measured in measurement phase. fishbone diagrams and the application of fmea are used to analyze the factors that cause problems that occur and prioritize improvements to solve these problems. the implementation of value stream mapping (vsm) is applied to the enhance of quality improvement phase to reduce waste. the results showed that product quality increased from an average sigma level of 3.53 to 3.79 where overproduction decreased by 41% and side board production leadtime decreased by 373 second. key words: quality improvement, lean production, dmaic, value stream mapping. 1. introduction nowadays business development is going very fast, especially in the manufacturing industry. the rapid development of the manufacturing industry will have an impact on intense competition in the domestic and international markets (hernadewita et al., 2019). the competitiveness among manufacturing industries encourages each company to increase their own productivity in various ways (gupta et al., 2016; soundararajan & reddy, 2020). in addition, the best way to win the competition in the global market is to improve production quality by minimizing waste (henny et al., 2019). the productivity and quality improvement must be proportional to the increase in product value to customers by increasing in the quality of production process (costa et al., 2020). the musical instrument industry is one of the electronic manufacturing industries that produces the upright piano’s and clavinova piano’s. companies are being develop various improvement strategies to increase productivity and quality (santos et al., 2020). the production process of the clavinova piano is devided into several sub-processes, one of which is the side board manufacturing process. in the production process, there have been still occured some problems that need to be fixed. as on preliminary observed, there were found wastes to cite this article: hernadewita, setiawan, i., hendra. (2022). enhance quality improvement through lean six sigma in division side board clavinova piano’s. international journal of production management and engineering, 10(2), 173-181. https://doi.org/10.4995/ijpme.2022.16140 int. j. prod. manag. eng. (2022) 10(2), 173-181creative commons attribution-noncommercial-noderivatives 4.0 international 173 mailto:hadeita@yahoo.com http://creativecommons.org/licenses/by-nc-nd/4.0/ such as ineffective processes, improper layout, overproduction and poor production quality. to improve the quality of its production, the side board division needs to create a streamlined production process so that it will create a more effective and efficient production line. a balance between effective and efficient can be made by reducing waste (karim & arif-uz-zaman, 2014). according to liker & meier (2006) waste is any activity that has non-value added. to eliminate waste, companies can use the concept of integration of lean and six sigma (narottam et al., 2019; megawati et al., 2020). these two concepts are one of the main business process strategies that can be used by various companies to improve manufacturing performance (pugna et al., 2016; raval et al., 2019). this concept can also identify complex problems (nandakumar et al., 2020). according to mahato et al. (2017) propose that the integration of lean and six sigma can provide effective results for optimization of production costs so that efficient production costs are obtained. research on lean six sigma has also been developed in large-scale manufacturing industries (swarnakar et al., 2020). the application of lean six sigma can not only be applied in large industries but also in small and medium industries by improving the production process (alexander et al., 2019). lean is a sustainable strategy to eliminate waste (kumar et al., 2019). the loss of waste will increase the company’s productivity (prayugo & zhong, 2021). a tool that can be used to reduce waste is value stream mapping (vsm) (lacerda et al., 2016). vsm can assist companies in identifying non-value added activities (kosasih et al., 2019). meanwhile, six sigma is an improvement to reduce variance in products so that they can provide value to customers (belu et al., 2018; setiawan & setiawan, 2020). the popular six sigma method used is dmaic (bhargava & gaur, 2021). dmaic is used as a systematic approach to eliminate waste and the best way to improve quality in the production process (garg et al., 2020). this study aims to help companies to eliminate waste on the production floor and propose corrective actions to improve the production quality of the side board division. based on previous research conducted in the musical instrument industry (rochman & agustin, 2017), it was found that the six sigmadmaic was able to reduce the level of defects in the soundboard and side piano production process. santosa & sugarindra (2018) were found that the lean manufacturing method can reduce waste in the upright piano sanding process. both studies focused on the upright piano. thus, there is a gap to conduct recent studies with different subjects. 2. methodology this study uses the lean six sigma integration method. the stages that design improvements use the systematic steps define, measure, analyze, improve, control (dmaic). the define phase begins by mapping the production process, defining waste and determining the critical to quality (ctq). the measurement phase starts with measuring the amount of production, measuring the distance and time of transportation, creating a control chart and measuring the sigma value. the analysis phase identifies the dominant causative factor. the improve phase designs the improvement plan and implements the improvements. the last phase of process control is by setting key performance indicators (kpis). the study framework for this research can be seen in figure 1. 3. result and discussion 3.1. define phase define is the first step in the dmaic stage, the steps taken are making a production process flow figure 1. study frame work. int. j. prod. manag. eng. (2022) 10(2), 173-181 creative commons attribution-noncommercial-noderivatives 4.0 international hernadewita et al. 174 http://creativecommons.org/licenses/by-nc-nd/4.0/ and defining the waste that occurs in the side board production line. sipoc diagram making uses side board production process activity data obtained through observation. sipoc diagram of the side board production process describes the side board production flow from suppliers to customers which can be seen in figure 2. there are 7 wastes of lean manufacturing in industrial sector such as transportation, overproduction, defects, motion, inventory, waiting and overprocessing. the results of this research on the piano music instrument industry show that there are 3 wastes of lean manufacturing which dominantly affect the quality of product, namely overproduction, transportation and defects. the results obtained in this research are: 1. overproduction. overproduction occurs because the number of products produced is more than the quantity ordered by the customer. overproduction is a type of waste that can support the occurrence of other wastes. 2. transportation. transportation is included in non-value added activities. the disadvantage of this waste is the addition of material handling, transportation equipment, moving distances, additional space for the movement of goods and storage. transportation can also cause product defects due to handling. 3. defect. the defect occurs because the product produced does not meet the specifications set by the company. defects that occur in the side board process are high enough so that customer satisfaction related to quality will be reduced, so it is necessary to improve the process so that the resulting product is close to zero defect. the type of defect in this research are scratch, not flat, exfoliate, bubble, gloss and other. by using the figure 2. sipoc diagram of the side board production process. figure 3. pareto diagram of accumulated defects. int. j. prod. manag. eng. (2022) 10(2), 173-181creative commons attribution-noncommercial-noderivatives 4.0 international enhance quality improvement through lean six sigma in division side board clavinova piano’s 175 http://creativecommons.org/licenses/by-nc-nd/4.0/ pareto diagram in figure 3, the biggest defect that occurs in the side board production line occurs in a scratch defect with the percentage 34,5. 3.2. measure phase this stage measures the waste that occurs on the production line, namely calculating the output of the amount of production, measuring the time and distance of transportation. then map the entire side board production flow using value stream mapping. this mapping is used to find out which activities are included in the non-value added. the following value stream mapping can be seen in figure 4. based on the value stream mapping map as shown in figure 4, it can be seen that the non-value added (nva) activity is 385 second, the value added (va) activity is 781 second and the non-value added necessary (nnva) activity is 112 second. product control limits are used to determine product variance. the number of samples used for observation was 27 times. p chart before improvement of case in the electrican music pianos can be seen in figure 5. figure 5 shows that there are several points outside the of product control limit. this shows that the production process is not completely under control. therefore it is necessary to make improvements to get a good production process. figure 4. value stream mapping of the side board process. figure 5. p-chart of defect before improvement. int. j. prod. manag. eng. (2022) 10(2), 173-181 creative commons attribution-noncommercial-noderivatives 4.0 international hernadewita et al. 176 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.3. analyze phase analyze phase is identifies the main factors causing each waste. fishbone diagram are used as the tool for analyse and identify the cause of waste problem. the factors that cause each waste at fishbone diagram can be seen in figure 6, figure 7 and figure 8. after the root cause has been identified, the next step is to calculate the risk priority number (rpn) value using the failure mode and effect analysis (fmea) method. this calculation is used to determine the priority ranking for improvement. this rpn assessment is carried out by expert judgment. table 1 is a calculation of the rpn value of each waste. figure 6. fishbone diagram of overproduction. figure 7. fishbone diagram of transportation. figure 8. fishbone diagram of defect. int. j. prod. manag. eng. (2022) 10(2), 173-181creative commons attribution-noncommercial-noderivatives 4.0 international enhance quality improvement through lean six sigma in division side board clavinova piano’s 177 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.4. improve phase based on the results of the analysis in the previous stage, the improvement of each waste can be done by the why, what, where, when, who, how (5w+1h) method. analysis of improvements with the 5w+1h method can be seen in table 2. the results of improvement of each waste using 5w+h are: 1. overproduction. improvements to overproduction, namely providing operators with daily production cards and posting them to general information boards. added target control led to the production line with the importance of avoiding overproduction. make short-term production forecasts. 2. transportation. transportation improvement is to do kaizen re-layout in the material storage area. change the make to stock method to make to order. eliminates material movement from supplier to wip area. the effect obtained reduces the side board production cycle time by 261 seconds. 3. defect. improvement of defects scratch, namely doing weekly validation on jigs and trolley, forming a total productive maintenance (tpm) team and doing tpm weekly. changed the method of material storage by providing felt barriers between materials 3.5. control phase this phase performs product control improvements with the p chart. after improvement of product control, the production process become stable. all samples are within product control limits. p chart pf product control limits after repair can be seen in figure 9. in control phase also calculates the dpmo value and the sigma level. calculation of the sigma value was carried out by taking a sample of 6 months and after the improvement took a sample of 2 months. during the observation, the sigma value show that table 1. calculation of the rpn value for each waste. waste potential failure mode sev potential failure effects occ potential cause of failure det rpn rank overproduction sop is not perfect 5 excess production ouput 6 lack of production control 7 210 1 transportation the layout is not good 6 long time transportation 6 process location is very far 7 252 1 defect nonstandard process capacity 6 between materials rubbing against each other 7 no protective appearance 6 252 1 trolley and jig is bad 6 the material hit the sharp edge 6 chipped felt 7 252 1 figure 9. p chart of the defect after improvement. int. j. prod. manag. eng. (2022) 10(2), 173-181 creative commons attribution-noncommercial-noderivatives 4.0 international hernadewita et al. 178 http://creativecommons.org/licenses/by-nc-nd/4.0/ it has increased. figure 10 show the results of sigma values comparison before and after improvement of production process. the average difference value every month is 3.5 except at january and february. after the improvement and implementation plan are carried out, the production process control is carried out by applying kpi as a measure of success. the kpi determination at this control stage is expected to have a significant influence and role on the target action plan planned by the side board production line. the target of this action plan is derived from the company’s vision and mission as well as the company’s strategy towards a zero-defect target. table 2. matrix of 5w+1h analysis. waste causes why what where when who how why should it be fixed? what needs to be fixed? where are the repairs done? when is it implemented? who is carrying out? how to implement? overproduction lack on production control maintain the production is not excessive. sop terkait schedule dan quantity produksi line side board at the beginning of production and model change operator side board carry out briefings at the beginning of each shift to increase production targets. provide operators with daily production plan cards. installing led control production system in the production line. transportation remote location keep the production leadtime shorter layout pabrik line side board december 2020 team of kaizen wood working changed the system: from make to stock to job order. thus, factory re-layout by eliminating material transfer from suppliers to the wip line side board area defect-scratch felt appearnace keep no scratch defect products penggantian felt yang tidak layak pakai line side board saat felt terkelupas team of jig engineeing establish a tpm team. conduct weekly validation. preventive on the jig, so that when an abnormal jig is found, repairs are immediately carried out. material storage methodmissed placed keep no scratch defect products memberikan sekat pembatas antar material line side board december 2020 tim of kaizen wood working changing the method of stacking material, which was originally without a partition, is now given a barrier using white felt clamp stiffness keep no defective decoction product jig jig engineering at the jig abnormal team of jig engineeing establish a tpm team. conduct weekly validation. preventive on the jig, so that when an abnormal jig is found, repairs are immediately carried out. defect-peeled off operator -lack of understanding keep no flaky defective products pembaruan sop line side board december 2022 staff engineering add the sop points related to the correct mentoring process flow, starting from the direction of the mentoring process and standard equipment requirements to be used. int. j. prod. manag. eng. (2022) 10(2), 173-181creative commons attribution-noncommercial-noderivatives 4.0 international enhance quality improvement through lean six sigma in division side board clavinova piano’s 179 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. conclusion the main problem that occurs in the side board production line is the low quality of the production process which causes waste on the production floor. the low quality of the production process has a loss impact on the company. to overcome this problem, several corrective actions must be taken, including: 1. improved overproduction by giving operators with daily production cards, installing led controls on production lines and forecasting short-term production. the effect is that overproduction has decreased by 41%. 2. improved kaizen re-layout on the side board production line can reduce the overall cycle time by 261 seconds 3. perform tpm on jigs and trolleys. the effect is that the quality of the product has increased the average sigma level from 3.56 to 3.79 its suggested in conducting empirical studies based on causality statistics to determine the variables that have a significant effect on company productivity and quality. references alexander, p., antony, j., & rodgers, b. (2019). lean six sigma for smalland medium-sized manufacturing enterprises: a systematic review. international journal of quality and reliability management, 36(3), 378–397. https://doi.org/10.1108/ijqrm-03-2018-0074 belu, n., rizea, a.d., niţu, e.l., gavriluţǎ, a.c., & gavriluţǎ, a.c. (2018). an application of six sigma to ppm reduction in the relationship with the external customer. iop conference series: materials science and engineering, 400(6). https://doi.org/10.1088/1757-899x/400/6/062006 bhargava, m., & gaur, s. (2021). process improvement using six-sigma (dmaic process) in bearing manufacturing industry: a case study. iop conference series: materials science and engineering, 1017(1). https://doi.org/10.1088/1757-899x/1017/1/012034 costa, j.p., lopes, i.s., & brito, j.p. (2020). six sigma application for quality improvement of the pin insertion process. procedia manufacturing, 38(2019), 1592–1599. https://doi.org/10.1016/j.promfg.2020.01.126 garg, a., raina, k., & sharma, r. (2020). reducing soldering defects in mobile phone manufacturing company: a dmaic approach. iop conference series: materials science and engineering, 748(1), 1–10. https://doi.org/10.1088/1757-899x/748/1/012027 gupta, a., sharma, p., malik, s.c., agarwal, n., & jha, p.c. (2016). stage of the amplifier production process: a dmaic. international journal of reliability, quality and safety engineering, 23(6), 1–13. https://doi.org/10.1142/s021853931640012x henny, h., andriana, i., latifah, a.n., & haryanto, h. (2019). the application lean six sigma method approach to minimize waste. iop conference series: materials science and engineering, 662(2). https://doi.org/10.1088/1757-899x/662/2/022089 hernadewita, h., rochmad, i., hendra, h., hermiyetti, h., & yuliani, e.n. s. (2019). an analysis of implementation of taguchi method to improve production of pulp on hydrapulper milling. international journal of production management and engineering, 7(2), 125. https://doi.org/10.4995/ijpme.2019.10163 karim, a., & arif-uz-zaman, k. (2014). a methodology for effective implementation of lean strategies and its performance evaluation in manufacturing organizations. business process management journal, 19(1), 169–196. https://doi.org/10.1108/14637151311294912 kosasih, w., sriwana, i.k., sari, e.c., & doaly, c.o. (2019). applying value stream mapping tools and kanban system for waste identification and reduction (case study: a basic chemical company). iop conference series: materials science and engineering, 528(1). https://doi.org/10.1088/1757-899x/528/1/012050 figure 10. sigma value comparison before and after improvement. int. j. prod. manag. eng. (2022) 10(2), 173-181 creative commons attribution-noncommercial-noderivatives 4.0 international hernadewita et al. 180 https://doi.org/10.1108/ijqrm-03-2018-0074 https://doi.org/10.1088/1757-899x/400/6/062006 https://doi.org/10.1088/1757-899x/1017/1/012034 https://doi.org/10.1016/j.promfg.2020.01.126 https://doi.org/10.1088/1757-899x/748/1/012027 https://doi.org/10.1142/s021853931640012x https://doi.org/10.1088/1757-899x/662/2/022089 https://doi.org/10.4995/ijpme.2019.10163 https://doi.org/10.1108/14637151311294912 https://doi.org/10.1088/1757-899x/528/1/012050 http://creativecommons.org/licenses/by-nc-nd/4.0/ kumar, n., jarial, s.k., & narwal, m.s. (2019). lean six sigma in brazil: a literature review. international journal of lean six sigma, 10(1), 435–472. https://doi.org/10.1108/ijlss-09-2017-0103 lacerda, a.p., xambre, a.r., & alvelos, h.m. (2016). applying value stream mapping to eliminate waste: a case study of an original equipment manufacturer for the automotive industry. international journal of production research, 54(6), 1708–1720. https://doi.org/10.1080/00207543.2015.1055349 liker, j.k., & meier, d. (2006). the toyota way fieldbook (1st ed.). mcgraw-hill education. mahato, s., rai dixit, a., & agrawal, r. (2017). application of lean six sigma for cost-optimised solution of a field quality problem: a case study. proceedings of the institution of mechanical engineers, part b: journal of engineering manufacture, 231(4), 713–729. https://doi.org/10.1177/0954405417694060 megawati, e., wicaksono, p.a., & nurkertamanda, d. (2020). reducing defect in furniture industry using a lean six sigma approach. aip conference proceedings, 2217(april). https://doi.org/10.1063/5.0004282 nandakumar, n., saleeshya, p.g., & harikumar, p. (2020). bottleneck identification and process improvement by lean six sigma dmaic methodology. materials today: proceedings, 24, 1217–1224. https://doi.org/10.1016/j. matpr.2020.04.436 narottam, y., mathiyazhagan, k., & kumar, k. (2019). literature review: continuous improvement through lean six sigma. international journal of productivity and quality management, 28(1), 3–27. https://doi.org/10.1504/ ijpqm.2019.102423 prayugo, j., & zhong, l. (2021). green productivity: waste reduction with green value stream mapping. a case study of leather production. international journal of production management and engineering, 9(1), 47–55. https://doi.org/10.4995/ijpme.2021.12254 pugna, a., negrea, r., & miclea, s. (2016). using six sigma methodology to improve the assembly process in an automotive company. procedia social and behavioral sciences, 221, 308–316. https://doi.org/10.1016/j.sbspro.2016.05.120 raval, s.j., kant, r., & shankar, r. (2019). benchmarking the lean six sigma performance measures: a balanced score card approach. benchmarking, 26(6), 1921–1947. https://doi.org/10.1108/bij-06-2018-0160 rochman, y.a., & agustin, a. (2017). minimization of defective products in the department of press bridge & rib through six sigma dmaic phases. iop conference series: materials science and engineering, 215(1). https://doi.org/10.1088/1757-899x/215/1/012035 santos, g., carlos, j., ricardo, s., pulido, j., jimenez, g., santos, g., pulido, j., & hernández, h. (2020). improvement of productivity and quality in the value chain improvement of productivity and quality in the value chain through lean manufacturing – a case study. procedia manufacturing, 41, 882–889. https://doi.org/10.1016/j.promfg.2019.10.011 santosa, w.a., & sugarindra, m. (2018). implementation of lean manufacturing to reduce waste in production line with value stream mapping approach and kaizen in division sanding upright piano, case study in: pt.x. matec web of conferences, 154(1), 8–11. https://doi.org/10.1051/matecconf/201815401095 setiawan, i., & setiawan. (2020). defect reduction of roof panel part in the export delivery process using the dmaic method: a case study. jurnal sistem dan manajemen industri, 4(2), 108–116. https://doi.org/10.30656/jsmi.v4i2.2775 soundararajan, k., & reddy, k.j. (2020). productivity and quality improvement through dmaic in sme. international journal of productivity and quality management, 31(2), 271–294. https://doi.org/https://doi.org/10.1504/ ijpqm.2020.110027 swarnakar, v., singh, a.r., & tiwari, a.k. (2020). effect of lean six sigma on firm performance: a case of indian automotive component manufacturing organization. materials today: proceedings, 07(august). https://doi.org/10.1016/j. matpr.2020.07.115 int. j. prod. manag. eng. (2022) 10(2), 173-181creative commons attribution-noncommercial-noderivatives 4.0 international enhance quality improvement through lean six sigma in division side board clavinova piano’s 181 https://doi.org/10.1108/ijlss-09-2017-0103 https://doi.org/10.1080/00207543.2015.1055349 https://doi.org/10.1177/0954405417694060 https://doi.org/10.1063/5.0004282 https://doi.org/10.1016/j.matpr.2020.04.436 https://doi.org/10.1016/j.matpr.2020.04.436 https://doi.org/10.1504/ijpqm.2019.102423 https://doi.org/10.1504/ijpqm.2019.102423 https://doi.org/10.4995/ijpme.2021.12254 https://doi.org/10.1016/j.sbspro.2016.05.120 https://doi.org/10.1108/bij-06-2018-0160 https://doi.org/10.1088/1757-899x/215/1/012035 https://doi.org/10.1016/j.promfg.2019.10.011 https://doi.org/10.1051/matecconf/201815401095 https://doi.org/10.30656/jsmi.v4i2.2775 https://doi.org/https://doi.org/10.1504/ijpqm.2020.110027 https://doi.org/https://doi.org/10.1504/ijpqm.2020.110027 https://doi.org/10.1016/j.matpr.2020.07.115 https://doi.org/10.1016/j.matpr.2020.07.115 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2013.1565 editorial this is the first number of the international journal of production management and engineering (ijpme). why this new journal? from the research centre on production management and engineering at the universitat politècnica de valència, it is understood that the numerous results being obtained in the research and transfer projects, including the generation of new knowledge and practical applications in companies and industrial sectors, are not reaching all the potentially interested people. although there are many journals that deal with enterprise issues, we want to address the resolution or solution proposal of the real problems that companies are facing. we also wish that the ijpme becomes an agile channel between researchers and readers. problems, cases and proposals will be addressed from an engineering point of view. industrial engineering, system engineering, process engineering, etc., are visions that the ijpme will enhance in the future. the ijpme’s objectives promote the progress and understanding of phenomena related with all aspects of production engineering and management. in particular, the ijpme covers proposals of strategic, tactic or operative challenges in industrial or service organizations, such as individual enterprises, and supply or distribution network approaches. moreover, cross-disciplinary working areas, such a business process modelling, operation research, performance measurement systems, enterprise engineering or information technology, etc., as decision support systems are welcome. strategy and design: physical systems design, supply or distribution network design, business models & modelling, business process, behavioural issues, collaborative networks, risk management, software services and enterprise architectures, kpis, inventory policies, etc., based on new models, methodologies, techniques, tools, etc. operations and control: demand forecasting, production planning & scheduling, order promising, project management, multi-level lot sizing, routing, etc., using exact heuristic or meta-heuristic methods based on fuzzy logic, mathematical models, intelligence software agents, methodologies, algorithms, etc. the journal is arranged in two sections: the editorial and papers section. the editorial section publishes a visionary short papers (2-3 pages max.) in order to guide future journal contributors by identifying the first number of ijpme: an exciting challenge vicéns-salort, e., gómez-gasquet, p., boza, a., franco pereyra, r.d., mula, j., & cuenca, ll. editorial team of ijpme. ijpme@upv.es centro de investigación de gestión e ingeniería de la producción (cigip). universitat politècnica de valència, cno. de vera, s/n, 46022, valencia, spain. evicens@cigip.upv.es pgomez@cigip.upv.es aboza@cigip.upv.es dfranco@cigip.upv.es fmula@cigip.upv.es llcuenca@cigip.upv.es 1int. j. prod. manag. eng. (2013) 1(1), 1-2creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2013.1565 mailto:ijpme@upv.es mailto:llcuenca}@cigip.upv.es emerging research trends inside different areas; the papers section. it publishes theoretical and empirical peer-reviewed articles. during this initial period in which the journal should shape its personality, for the purpose of becoming a reference in its area, a policy has been made to focus on quality, with two annual numbers, and on the selection of no more than eight papers per number. the publication dates will be the second week of january and the first week of july. all the ijpme numbers will offer the widest possible view of the forefront of engineering proposals. the journal is available in an open environment under creative commons licence (http:// creativecommons.org/licenses/by-nc/3.0/es/deed.en) in accordance with current trends and the framework proposed by the european union. this ensures an agile, widespread dissemination of the works. the ojs (open journal service) platform and the editorial through the polipapers portal-facilitated journals achieve these objectives. in order to meet the quality criteria required by databases, the ijpme has welcomed the statement of ethics and the good practice polipapers portal, based on the committee on publications ethics (http://publicationethics.org/). in this first issue, some of the most relevant authors in the field of engineering have been invited. the editorial team is pleased to present the following papers: pablo aparicio ruiz, maria rodríguez palero, luis onieva giménez survival function of a power transformer and a switch by means of nonparametric estimators. luiz c. r. carpinetti, rafael h. p. lima institutions for collaboration in industrial clusters: proposal of a performance and change management model. bernard grabot, yue ming, raymond houé mrp-based negotiation in collaborative supply chains. paul w. p. j. grefen, remco m. dijkman hybrid control of supply chains: a structured exploration from a systems perspective. andrew c.l. lyons lessons from empirical studies in product and service variety management. colin raßfeld, dominikrößle, roland jochem integrated and modular design of an optimized process architecture. acknowledgements: we would like to conclude this first editorial by thanking the universitat politècnica de valència for the facilities provided to us through the ojs platform and the editorial staff. we also thank the initiative of the research centre on production management and engineering, and the people who have devoted their efforts to identify and promote the importance of engineering, whose dissemination task today we join with enthusiasm. 2 int. j. prod. manag. eng. (2013) 1(1), 1-2 creative commons attribution-noncommercial 3.0 spain vicéns-salort, e., gómez-gasquet, p., boza, a., franco pereyra, r.d., mula, j., & cuenca, ll. http://creativecommons.org/licenses/by-nc/3.0/es/deed.en http://creativecommons.org/licenses/by-nc/3.0/es/deed.en http://publicationethics.org/ http://dx.doi.org/10.4995/ijpme.2013.1517 http://dx.doi.org/10.4995/ijpme.2013.1517 http://dx.doi.org/10.4995/ijpme.2013.1517 http://dx.doi.org/10.4995/ijpme.2013.1517 http://dx.doi.org/10.4995/ijpme.2013.1502 http://dx.doi.org/10.4995/ijpme.2013.1502 http://dx.doi.org/10.4995/ijpme.2013.1502 http://dx.doi.org/10.4995/ijpme.2013.1502 http://dx.doi.org/10.4995/ijpme.2013.1516 http://dx.doi.org/10.4995/ijpme.2013.1516 http://dx.doi.org/10.4995/ijpme.2013.1516 http://dx.doi.org/10.4995/ijpme.2013.1544 http://dx.doi.org/10.4995/ijpme.2013.1544 http://dx.doi.org/10.4995/ijpme.2013.1544 http://dx.doi.org/10.4995/ijpme.2013.1557 http://dx.doi.org/10.4995/ijpme.2013.1557 http://dx.doi.org/10.4995/ijpme.2013.1557 http://dx.doi.org/10.4995/ijpme.2013.1558 http://dx.doi.org/10.4995/ijpme.2013.1558 http://dx.doi.org/10.4995/ijpme.2013.1558 pme i j international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.16736 received: 2021-11-25 accepted: 2022-01-24 comparative study of whale optimization algorithm and flower pollination algorithm to solve workers assignment problem huthaifa al-khazraji a control and system engineering department, university of technology-iraq, baghdad, iraq. 60141@uotechnology.edu.iq abstract: many important problems in engineering management can be formulated as resource assignment problem (rap). the workers assignment problem (wap) is considered as a sub-class of rap which aims to find an optimal assignment of workers to a number of tasks in order to optimize certain objectives. wap is an np-hard combinatorial optimization problem. due to its importance, several algorithms have been developed to solve it. in this paper, it is considered that a manager is required to provide a training course to his workers in order to improve their level of skill or experience to have a sustainable competitive advantage in the industry. the training cost of each worker to perform a particular job is different. the wap is to find the best assignment of workers to training courses such that the total training cost is minimized. two metaheuristic optimizations named whale optimization algorithm (woa) and flower pollination algorithm (fpa) are utilized to final the optimal solution that reduces the total cost. matlab software is used to perform the simulation of the two proposed methods into wap. the computational results for a set of randomly generated problems of various sizes show that the fpa is able to find good quality solutions. key words: servitization, resource assignment problem, workers assignment problem, metaheuristic optimization, whale optimization algorithm, flower pollination algorithm. 1. introduction with the increase in competition in the global market, industrial companies are forced to improve their manufacturing processes via cutting costs and increasing process efficiency (ostadi et al., 2021). in this direction, it’s become necessary for decisionmakers to find the best strategies that utilize their resources in order to have the best performance (lin and chiu, 2018). the problem of finding the best utilization of resources in the industrial companies is named a resource assignment problem (rap). among many varieties of resources, human resources play a significant role in the success of the industrial organization if they are well allocated to different services or systems, with an aim to maximize or minimize certain objectives related to performance and productivity (bouajaja and dridi, 2017). therefore, worker assignment problem (wap) is defined as a process of assigning workers among various tasks for maximization of the profit (or efficiency) or minimization of the cost (or time). a wap is the most widely used in the context of industrial and engineering management such as in production planning and maintenance management (krokhmal and pardalos, 2009). for some industrial processes, the load needs to be distributed among workers in such a way that the time is minimized or/and the efficiency is increased. moreover, it can be noticed that there are particular workers who can perform some of the jobs with less to cite this article: al-khazraji, h. (2022). comparative study of whale optimization algorithm and flower pollination algorithm to solve workers assignment problem. international journal of production management and engineering, 10(1), 91-98. https://doi.org/10.4995/ijpme.2022.16736 int. j. prod. manag. eng. (2022) 10(1), 91-98creative commons attribution-noncommercial-noderivatives 4.0 international 91 https://orcid.org/0000-0002-6290-3382 http://creativecommons.org/licenses/by-nc-nd/4.0/ time or more efficient way than others due to their experience or skills (mahmoud, 2009). caron et al. (1999) considered the case of wap with additional constrain that an unassigned worker cannot be given a certain job unless the unassigned worker has the qualification to perform that particular job. however, if the manager focuses only on the workers which have skills or experiences to perform the tasks without having the right blanching of distrusting the loads among all workers, a worker fatigue problem might be happed in the long term (yadav et al., 2020; demiral, 2017). besides, intangible resources represented by the skill and experience of the workers are valuable and scarce. moreover, the experiences that the workers gained in the past are not enough to have a sustainable competitive advantage in the present (ruiz et al., 2020). therefore, a manager has to provide a training course continuously to his workers in order to improve their level of skill or experience to have a sustainable competitive advantage in the industry. the problem of wap that is considered in this paper can be stated as follows: if the manager needs to assign n of jobs to n of workers where the training cost of each worker to perform that particular job is different. the problem is to find the best assignment of workers to training courses such that the total training cost is minimized. the problem is np-hard problems with enormous search spaces (ammar et al., 2013). pentico (2007) reviewed the mathematical model for most of the variations of the assignment problem (ap) which is the general form of wap. bouajaja and dridi (2017) presented a comprehensive review study on the numerous approaches that applied to solve varieties classes of wap in different application areas. the general ap (gap) can be formulated as integer linear programming. different methods such as exact, heuristic and metaheuristic are developed to solve the problem. kuhn (1955) developed the well-known hungarian method to solve the general ap. in ross and soland (1975), a branch and bound (b&b) technique is proposed to solve the general ap. xuezhi and xuehua (1996) described how ap can be solved using dynamic programming. exact methods such as the hungarian method, b&b technique and dynamic programming are only effective in certain problems with a small size of decision variables. therefore, larger-sized problems are often solved by using heuristic and metaheuristic to obtain highquality solutions with reasonable computational time (bouajaja and dridi, 2017). in terms of heuristic methods, cattrysse et al. (1994) proposed a column generation heuristic method where the problem was formulated as a set partitioning problem. on other hand, among many metaheuristic methods, ant colony optimization (aco) was the most approach that is utilized to solve the problem. for example, wang et al. (2009) presented a detailed procedure to apply aco to wap where the objective was to maximize efficiency. demiral (2017) implemented aco for a set of randomly generated wap. three objectives of wap (minimization cost, maximization sales and maximization profit) were investigated in the study. statistical analysis based on mean, standard deviation and variance was performed to help decisionmakers to select the best objective based on their perspective. in the same direction, suliman (2019) examined the performance of aco in comparison with the traditional hungarian method for solving the wap with the size of 3×3 in terms of running time, number of iteration and quality of solutions. besides aco, chu and beasley (1997) presented a genetic algorithm (ga) for solving the ap. jia and gong (2008) solved the multi-objective wap using multiobjective particle swarm optimization (mopso). in this paper, two metaheuristic optimizations named whale optimization algorithm (woa) and flower pollination algorithm (fpa) are utilized final the optimal solution that reduces the total cost. matlab software is used to perform the simulation of the two proposed methods into wap. 2. training course training can be defined as a process of developing programs to ensure that employees are provided the right skills that are needed to achieve better positive in the market (halawi and haydar, 2018). the importance of providing training courses to the workforce of the industrial companies appears to be a recognition strategy toward gaining a competitive advantage in the global market competition (sharma, 2014). workforces are required continuous training courses to have sustainability professional qualifications to cope with the recent advanced technology within industry 4.0. according to walsh and volini (2017), 80% of human resource managers reported that workforce training is one of the biggest problems to improve the effectiveness and competitiveness of industrial companies. in this direction, industrial companies are compelled to provide workforce training strategies continuously int. j. prod. manag. eng. (2022) 10(1), 91-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-khazraji 92 http://creativecommons.org/licenses/by-nc-nd/4.0/ to increase productivity. the present paper consideres that a manager is required to assign n of jobs to n of workers where the training cost of each worker to perform a particular job is different. the problem is to find the best assignment of workers to training courses such that the total training cost is minimized. 3. mathematical model consider there are (n) of workers (w) are required to be assigned to (n) of training courses (s) as shown in figure 1. each worker wi (i=1,2,…,n) can be assign to any course sj (j=1,2,…,n) with different cost (ci,j). the problem is to find the best assignment of workers to courses such that the total cost of training is minimized. the number of the workers was assumed equal to the number of courses in this study. the wap is formulated as (krokhmal and pardalos, 2009): ∑∑ = = n i n j ijijxcmin 1 1 (1) s.t ∑ = = n i ijx 1 1 (j=1,2,…,n) (2) ∑ = = n j ijx 1 1 (i=1,2,…,n) (3) where xij has two values, either 1 if the worker i is assigned to job j, otherwise is zero. the constraint in equation (2) satisfies that each training course is assign to a worker and the constraint in equation (3) satisfies that each worker is assign to a training course. figure 1. workers assignment problem. 4. solution approach bio-inspired algorithms are considered powerful in solving np-hard combinatorial optimization problems (yang, 2009). therefore, two algorithms which are inspired from the biological systems in nature are proposed to final the optimal solution that reduces the total cost. these two algorithms are whale optimization algorithm (woa) and flower pollination algorithm (fpa). the next subsections explain these two algorithms. 4.1. whale optimization algorithm whale optimization algorithm (woa) is a population-based swarm optimization algorithm. it was developed by mirjalili and lewis in 2016. woa mimics the bubble-net hunting behavior of humpback whales. the mathematical model of this algorithm consists of two processes named exploitation and exploration. in the exploitation process, the position of the humpback whale is updated based on the location of the prey using a bubble-net attacking strategy. in this strategy, the movement towards the prey is performed by two mechanisms (satapathy et al., 2018). the first one is the shrinking encircling mechanism. this behavior is represented by the following equations (mirjalili and lewis, 2016): q=2r1 (4) a=2ar2–a (5) d=q·p*(t)–p(t) (6) p(t+1)=p*(t)–a·d (7) where a coefficient value linearly decreased from 2 to 0 for each iteration q coefficient value calculated as given in equation (4) a coefficient value calculated as given in equation (5) d coefficient value calculated as given in equation (6) r1,r2 random value between [0,1] t current iteration int. j. prod. manag. eng. (2022) 10(1), 91-98creative commons attribution-noncommercial-noderivatives 4.0 international comparative study of whale optimization algorithm and flower pollination algorithm to solve workers assignment problem 93 http://creativecommons.org/licenses/by-nc-nd/4.0/ p* position of the prey p position of the whale the second mechanism is named spiral updating position where in this strategy the whale moves towards the prey in a helix-shaped movement. this behavior is represented by the following equations (mirjalili and lewis, 2016): d'=│ p*(t)–p(t) │ (8) p(t+1)=d;ebl cos(2πl)+p*(t) (9) where d' coefficient value calculated as given in equation (8) b constant used to define the shape of the logarithmic spiral l random value between [0,1] to model the changes between these two strategies, it was assumed that there is a probability of 50% to select between the shrinking encircling strategy and the spiral one to update the position of the current whale in the simulation of the algorithm. the mathematical model of this scenario can be formulated by selecting a random value (rand), then if the value of the random value >50 the movement of the individual will be performed based on equation (7), otherwise wiii be performed based on equation (9) (mirjalili and lewis, 2016). in the exploration process, the position of the humpback whale is updated randomly. the basic idea of this strategy is to ensure the search space explored globally. this strategy is represented by the following equations (mirjalili and lewis, 2016): d=│q·prand(t)–p(t) │ (10) p(t+1)=prand (p)+a·d (11) where prand random position chosen from the current population the pseudo code of fpa is illustrated in figure 2. 1. input ✓ objective function (fitness function), population size (n), coefficient value a, number of iteration t 2. initialization ✓ initialize population ✓ evaluate objective function ✓ find p* 3. loop: ✓ for t = 1:t ✓ for i = 1: n ✓ update q as in eq. (4), a as in eq. (5) and select random value θ ✓ if θ > 0.5 ➢ if |a| < 1:update the position of the current whale based on eq. (7) ➢ if |a) > 1:update the position of the current whale based on eq. (9) ✓ else ➢ update the position of the current whale based on eq. (11) ✓ perform greedy selection and update p" ✓ if there is no convergence of the current solution & if t >t go to loop 4. print the optimal solution figure 2. the pseudo code of woa. 4.2. flower pollination algorithm flower pollination algorithm (fpa) is a swarmbased meta-heuristic optimization developed by yang in 2012. the fpa mimics the pollination phenomena in the flower. the main idea of the pollination in flower is to transfer the pollen from the male into the female. this process can be classified based on the way that pollen is transferred into biotic and abiotic. in the biotic, the pollinator can be animal or insect, whereas in the abiotic, the pollinator is the wind and diffusion in water (abdel-basset and shawky, 2019). the optimization procedure of the fpa starts with randomly initialized a population of n of flowers within the search space as given in equation (12): yi = yl+rand*(yu–yl) (12) where i counter (i= 1,2,3,…) rand initial solution yi random value between [0,1] yl lower bound of the search space yu upper bound of the search space in the fpa, there are two ways to search of the optimum value. these are the global pollination and local pollination (yang, 2012). in the global pollination stage, the movement of each individual in the population is directed by the one that has the int. j. prod. manag. eng. (2022) 10(1), 91-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-khazraji 94 http://creativecommons.org/licenses/by-nc-nd/4.0/ best cost function found yet. this can be represented mathematically as: yinew = yiold + σ*(yiold –yg) (13) where yinew new solution yiold old solution yg the current best solution σ the step size the step size σ can be set fixed or follow a random steps such as lèvy flight. in terms of local pollination, the algorithm selects two solutions randomly, and then the new solution is generated based on the following (yang, 2012): yinew = yiold + ε*(yj –yk) (14) where yj a solution chosen randomly yk a solution chosen randomly ε random value between [0,1] the pseudo code of fpa is illustrated in figure 3 1. input ✓ objective function (fitness function), population size (n), switch probability (p), number of iteration (t) 2. initialization ✓ initialize population n flowers based on eq. (12) ✓ evaluate objective function and assign yz 3. loop: ✓ for t = 1: t ✓ for i = 1: n ✓ if rand < p ➢ generate a step size (σ) ➢ generate a new solution based on eq. (13) (global pollination) ✓ else ➢ choose two solutions randomly among all solutions ➢ generate a new solution based on eq. (14) (local pollination) ✓ perform greedy selection and update yg ✓ if there is no convergence of the current solution & if t > t go to loop 4. print the optimal solution figure 3. the pseudo code of fpa 5. simulation study for evaluating the performance of the two algorithms (woa and fpa) to solve wap, a set of different size randomly generated problems have been used. three sizes (n=5,10,15) of wap are considered to perform the evaluation as given in the appendix i. matlab software is used to perform the simulation. matlab becomes a powerful tool in wide applications in engineering, economics and management. it can handle different computational algorithms with a reasonable time. on other words, different algorithms could be tested and evaluated with less time. as a consequence of using simulation, more knowledge and insight can be gained to enhance the solution of wap. all simulations were conducted on a computer with intel(r) core(tm) cpu i7-4500 and 8 gb ram. the woa and fpa parameters are presented in table 1. for justify the comparison between the woa and fpa, the size of the population and the number of iteration are set equally. the coefficient value (a) is set to 2 as recommended by the (mirjalili and lewis, 2016). in the same way, the value of the switch probability (p) is set to 0.5 as recommended by the yang (2012). both algorithms were run 10 times and the statistical data such as the average (avg.), the maximum (max), the minimum (min) and the standard deviation (std.) were recorded for each algorithm. table 2 presents the statistical data of the experiments for the three problem size of wap using woa and fpa. table 1. woa and fpa algorithms parameters. parameters woa fpa number of population (n) 50 50 number of iteration (t) 100 100 coefficient value (a) 2 switch probability (p) 0.5 table 2. statistics for solving wap using woa and fpa size method avg. min max std. 5 woa 30 30 30 0 fpa 30 30 30 0 10 woa 64.6 63 67 1.175 fpa 62.2 61 63 0.78 15 woa 104.1 103 105 0.738 fpa 101.8 101 102 0.422 it can be noticed form table 2 that in general for small size problem, both algorithms are obtained a good solution results. however, table 2 show that if the size of the problem increased (i.e. n=10 and n=15), the fpa is recommended to solve the wap. fpa shows better performance in terms of obtaining a less average value, less minimum value, less maximum value and less standard deviation. int. j. prod. manag. eng. (2022) 10(1), 91-98creative commons attribution-noncommercial-noderivatives 4.0 international comparative study of whale optimization algorithm and flower pollination algorithm to solve workers assignment problem 95 http://creativecommons.org/licenses/by-nc-nd/4.0/ 6. conclusions the worker assignment problem (wap) is an important problem faced by manufacturing companies. wap is an np-hard combinatorial optimization problem. human resources play a significant role in the overall performance of manufacturing companies. workforce training is one of the biggest problems in the industrial companies to improve effectiveness and competitiveness. the paper considers the problem of assigning workers to training courses in order to improve the level of skill or experience of the worker to have a sustainable competitive advantage in the industry. the training cost to perform a particular job of each worker is different. the wap is to find the best assignment of workers to training courses such that the total training cost is minimized. two metaheuristic optimizations named whale optimization algorithm (woa) and flower pollination algorithm (fpa) are utilized to final the optimal solution that reduces the total cost. the simulations results reveal that for a small size problem, both algorithms are obtained a good solution result. however, for a large size problem (i.e. n=10 and n=15), the fpa is recommended to solve the wap. fpa shows better performance in terms of obtaining a less average value, less minimum value, less maximum value and less standard deviation. references abdel-basset, m., & shawky, l. a. (2019). flower pollination algorithm: a comprehensive review. artificial intelligence review, 52(4), 2533-2557. https://doi.org/10.1007/s10462-018-9624-4 ammar, a., pierreval, h., & elkosentini, s. (2013). workers assignment problems in manufacturing systems: a literature analysis. in proceedings of 2013 international conference on industrial engineering and systems management (iesm) (pp. 1-7). ieee. bouajaja, s., & dridi, n. (2017). a survey on human resource allocation problem and its applications. operational research, 17(2), 339-369. https://doi.org/10.1007/s12351-016-0247-8 caron, g., hansen, p., & jaumard, b. (1999). the assignment problem with seniority and job priority constraints. operations research, 47(3), 449-453. https://doi.org/10.1287/opre.47.3.449 cattrysse, d. g., salomon, m., & van wassenhove, l. n. (1994). a set partitioning heuristic for the generalized assignment problem. european journal of operational research, 72(1), 167-174. https://doi.org/10.1016/0377-2217(94)90338-7 chu, p. c., & beasley, j. e. (1997). a genetic algorithm for the generalised assignment problem. computers & operations research, 24(1), 17-23. https://doi.org/10.1016/s0305-0548(96)00032-9 demiral, m. f. (2017). ant colony optimization for a variety of classic assignment problems. in international turkish world engineering and science congress, antalya. halawi, a., & haydar, n. (2018). effects of training on employee performance: a case study of bonjus and khatib & alami companies. international humanities studies, 5(2). jia, z., & gong, l. (2008). multi-criteria human resource allocation for optimization problems using multi-objective particle swarm optimization algorithm. in 2008 international conference on computer science and software engineering, 1, 1187-1190. ieee. https://doi.org/10.1109/csse.2008.1506 koleva, n., & andreev, o. (2018, june). aspects of training in the field of operations management with respect to industry 4.0. in 2018 international conference on high technology for sustainable development (hitech) (pp. 1-3). ieee. https://doi.org/10.1109/hitech.2018.8566581 krokhmal, p. a., & pardalos, p. m. (2009). random assignment problems. european journal of operational research, 194(1), 1-17. https://doi.org/10.1016/j.ejor.2007.11.062 kuhn, h. w. (1955). the hungarian method for the assignment problem. naval research logistics quarterly, 2(1-2), 83-97. https://doi.org/10.1002/nav.3800020109 lin, j. t., & chiu, c. c. (2018). a hybrid particle swarm optimization with local search for stochastic resource allocation problem. journal of intelligent manufacturing, 29(3), 481-495. https://doi.org/10.1007/s10845-015-1124-7 mahmoud, k. i. (2009). split assignment with transportation model for job-shop loading (case study). journal of engineering, 15(2). mirjalili, s., & lewis, a. (2016). the whale optimization algorithm. advances in engineering software, 95, 51-67. https://doi.org/10.1016/j.advengsoft.2016.01.008 pentico, d. w. (2007). assignment problems: a golden anniversary survey. european journal of operational research, 176(2), 774-793. https://doi.org/10.1016/j.ejor.2005.09.014 int. j. prod. manag. eng. (2022) 10(1), 91-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-khazraji 96 https://doi.org/10.1007/s10462-018-9624-4 https://doi.org/10.1007/s12351-016-0247-8 https://doi.org/10.1287/opre.47.3.449 https://doi.org/10.1016/0377-2217(94)90338-7 https://doi.org/10.1016/s0305-0548(96)00032-9 https://doi.org/10.1109/csse.2008.1506 https://doi.org/10.1109/hitech.2018.8566581 https://doi.org/10.1016/j.ejor.2007.11.062 https://doi.org/10.1002/nav.3800020109 https://doi.org/10.1007/s10845-015-1124-7 https://doi.org/10.1016/j.advengsoft.2016.01.008 https://doi.org/10.1016/j.ejor.2005.09.014 http://creativecommons.org/licenses/by-nc-nd/4.0/ ross, g. t., & soland, r. m. (1975). a branch and bound algorithm for the generalized assignment problem. mathematical programming, 8(1), 91-103. https://doi.org/10.1007/bf01580430 ruiz, m., igartua, j. i., mindeguia, m., & orobengoa, m. (2020). understanding and representation of organizational training programs and their evaluation. international journal of production management and engineering, 8(2), 99109. https://doi.org/10.4995/ijpme.2020.12271 satapathy, p., mishra, s. p., sahu, b. k., debnath, m. k., & mohanty, p. k. (2018, april). design and implementation of whale optimization algorithm based pidf controller for agc problem in unified system. in international conference on soft computing systems (pp. 837-846). springer, singapore. https://doi.org/10.1007/978-981-13-1936-5_85 sharma, h. (2014). importance and performance of managerial training in indian companies-an empirical study. the journal of management development, 33(2), 75-89. https://doi.org/10.1108/jmd-11-2013-0144 suliman, a. s. a. (2019). using ant colony algorithm to find the optimal assignment. al-anbar university journal of economic and administration sciences, 11(25). ostadi, b., taghizadeh yazdi, m., & mohammadi balani, a. (2021). process capability studies in an automated flexible assembly process: a case study in an automotive industry. iranian journal of management studies, 14(1), 1-37. walsh, b. & volini, e. (2017). rewriting the rules for the digital age. deloitte university press. new york. wang, z., li, s., wang, y., & li, s. (2009, august). the research of task assignment based on ant colony algorithm. in 2009 international conference on mechatronics and automation (pp. 2334-2339). ieee. xuezhi, q., & xuehua, w. (1996). dynamic programming model of a sort of optimal assignment problem [j]. mathematics in practice and theory, 3. yadav, n., banerjee, k., & bali, v. (2020). a survey on fatigue detection of workers using machine learning. international journal of e-health and medical communications (ijehmc), 11(3), 1-8. https://doi.org/10.4018/ijehmc.2020070101 yang, x. s. (2009). firefly algorithms for multimodal optimization. in international symposium on stochastic algorithms (pp. 169-178). springer, berlin, heidelberg. https://doi.org/10.1007/978-3-642-04944-6_14 yang, x. s. (2012). flower pollination algorithm for global optimization. in international conference on unconventional computing and natural computation (pp. 240-249). springer, berlin, heidelberg. int. j. prod. manag. eng. (2022) 10(1), 91-98creative commons attribution-noncommercial-noderivatives 4.0 international comparative study of whale optimization algorithm and flower pollination algorithm to solve workers assignment problem 97 https://doi.org/10.1007/bf01580430 https://doi.org/10.4995/ijpme.2020.12271 https://doi.org/10.1007/978-981-13-1936-5_85 https://doi.org/10.1108/jmd-11-2013-0144 https://doi.org/10.4018/ijehmc.2020070101 https://doi.org/10.1007/978-3-642-04944-6_14 http://creativecommons.org/licenses/by-nc-nd/4.0/ appendix i this appendix presents the three randomly generated wap that were used in this study. size example (ci,j) optimal assignment (xi,j) minimum cost 5 ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣ ⎡ 67978 108776 710968 89688 768710 ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣ ⎡ 10000 00001 00010 00100 01000 30 10 ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣ ⎡ 101096789788 1167988108710 79787710976 101069789788 119798988610 811107987678 910796108779 77787610978 61098789788 119798868710 ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣ ⎡ 0001000000 0100000000 0000000001 0010000000 0000000010 0000000100 0000100000 0000010000 1000000000 0000001000 60 15 ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣ ⎡ 1161099711109799788 10117971167988108710 877111099781098769 77679810877878910 8781197778786979 788108910796108778 6118981187988108710 118771079787710976 89768119798988810 7119871010910769788 11897861098789788 791079810877879610 878116798797871011 78978101096789788 97789101069789788 ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣ ⎡ 010900000000000 000000100000000 000000000000010 001000000000000 000000000001000 000000000100000 100000000000000 000000000000001 000100000000000 000000000010000 000001000000000 000000000000010 000010000000000 000000001000000 000000010000000 90 int. j. prod. manag. eng. (2022) 10(1), 91-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-khazraji 98 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.16666 received: 2021-11-15 accepted: 2021-12-19 mapping the scientific structure of organization and management of enterprises using complex networks olivares-gil, a. a1 , arnaiz-rodríguez, a. b,c , ramírez-sanz, j.m. a2 , garrido-labrador, j.l. a3 , ahedo, v. d1 , garcía-osorio, c. a4 , santos, j.i. d2 , & galán, j.m. d3* a universidad de burgos, departamento de ingeniería informática, escuela politécnica superior, ed. a1, avda. cantabria s/n. 09006, burgos, spain. a1 aolivares@ubu.es, a2 jmrsanz@ubu.es, a3 jlgarrido@ubu.es, a4 cgosorio@ubu.es b ellis (european lab. for learning and intelligent systems) unit alicante, universidad de alicante, edificio nuevos institutos ctra. san vicente s/n. 03690, san vicente del raspeig, alicante, españa. adrian@ellisalicante.org c universidad de alicante, departamento de ciencia de la computación e inteligencia artificial, 03080, san vicente del raspeig, alicante, spain. d universidad de burgos, departamento de ingeniería de organización, escuela politécnica superior, ed. a1, avda. cantabria s/n. 09006, burgos, spain. d1 vahedo@ubu.es, d2 jisantos@ubu.es, d3 jmgalan@ubu.es abstract: understanding the scientific and social structure of a discipline is a fundamental aspect for scientific evaluation processes, identifying trends and niches, and balancing the trade-off between exploitation and exploration in research. in the present contribution, the production of doctoral theses is used as a proxy to analyze the scientific structure of the knowledge area of business organization in spain. to that end, a complex networks approach is selected, and two different networks are built: (i) the social network of co-participation in thesis examining committees and thesis supervision, and (ii) a bipartite network of theses and thesis descriptors. the former has a modular structure that is partially explained by thematic specialization in different subdisciplines. the latter serves to assess the interdisciplinary structure of the discipline, as it enables the characterization of affinity levels between fields, research poles and thematic clusters. our results have implications for the scientific evaluation and formal definition of related fields. key words: complex networks, community detection, doctoral theses, pattern recognition, interdisciplinarity, organization and management of enterprises. 1. introduction science mapping is an essential tool for understanding the structure of science and determining both the scientific strategy and the evaluation criteria of scientific production (sedighi, 2016). this process is often conducted through the formal analysis of networks such as journal citation networks or cocitation networks, among others (newman, 2003). one of the ways in which the mapping process can be carried out is through the study of the doctoral theses produced within a given field. such an approach, that uses theses as interaction entities, presents some interesting particularities. the most notable of all of them is probably that the effort to cite this article: olivares-gil, a., arnaiz-rodríguez, a., ramírez-sanz, j.m., garrido-labrador, j.l., ahedo, v., garcía-osorio, c., santos, j.i.,& galán, j.m. (2022). mapping the scientific structure of organization and management of enterprises using complex networks. international journal of production management and engineering, 10(1), 65-75. https://doi.org/10.4995/ijpme.2022.16666 http://polipapers.upv.es/index.php/ijpme int. j. prod. manag. eng. (2022) 10(1), 65-75creative commons attribution-noncommercial-noderivatives 4.0 international 65 https://orcid.org/0000-0002-3378-197x https://orcid.org/0000-0001-5567-801x https://orcid.org/0000-0003-0189-8046 https://orcid.org/0000-0002-3441-4223 https://orcid.org/0000-0002-9812-388x https://orcid.org/0000-0002-1206-1084 https://orcid.org/0000-0002-6653-043x https://orcid.org/0000-0003-3360-7602 mailto:aolivares@ubu.es mailto:jmrsanz@ubu.es mailto:jlgarrido@ubu.es mailto:cgosorio@ubu.es mailto:adrian@ellisalicante.org mailto:vahedo@ubu.es mailto:jisantos@ubu.es mailto:jmgalan@ubu.es http://creativecommons.org/licenses/by-nc-nd/4.0/ and commitment required to undertake a doctoral dissertation —often greater than in other scientific enterprises, where collaborations may be more punctual and opportunistic— has the potential to indicate more robust trends and research lines. doctoral theses are a crucial source of information to understand the scientific structure of disciplines and identify the social dimension of science, its main actors and protagonists, and how they are related to each other (repiso, torres, & delgado, 2011). in the case of spain, the influence of supervisors on the composition of the thesis evaluation committee is a well-known phenomenon (villarroya, barrios, borrego, & frías, 2008). thereupon, the co-occurrence of members and supervisors combined with the thesis descriptors can provide relevant information on both the academic and social structure of the scientific field(s) under consideration. scientific areas are the central element in the evaluation and development of the scientific career. in spain, the knowledge area of business organization —organización de empresas— is one of the most varied in terms of subjects and one of the most numerous in the number of academics. previous theses-based analyses of the scientific structure of this area have shown: (i) an unequal distribution of participation in thesis evaluation committees (compatible with a truncated powerlaw); (ii) a modular structure; and (iii) a positive assortativity among network members belonging to the same scientific association (garrido-labrador et al., 2022). the present work continues this line of research and further uses complex networks analysis tools (newman, 2003) to extend and deepen previous findings and intuitions on the teseo database. more specifically, we combine the co-participation networks in thesis examination committees in organization and management of enterprises with the unesco descriptors (unesco, 1988) that define the topics and areas of each doctoral thesis. notably, by expanding the information on the theses at the subdiscipline level, we find a more detailed scientific map of the field and can relate it to the levels and areas of scientific specialization. eventually, we complete our analysis by identifying the interdisciplinary relations of organization and management of enterprises with the rest of existing domains. 2. methods and data sources 2.1. methodological framework: network science the methodological framework used in this work to formalize the problem is network science (also called complex networks or network analysis). this approach has received much attention within the academia, and its development and applications have grown significantly in recent years (latora, nicosia, & russo, 2017). modeling a system as a network is a very general abstraction, as networks allow describing the interactions between elements of systems of a very diverse nature —from social and biological to technological and information phenomena (newman, 2003). the advantages of modeling a system as a network lie not only in the intuitiveness of the description but also in the fact that, once a system has been described in network terms, it can be analyzed using the powerful mathematical apparatus of graph theory. such apparatus makes it possible to extract, analyze and summarize information on the functioning of the system in a powerful and tractable way, both in static and dynamic terms. in a network approach, the elements that constitute a given system are modeled as nodes, and the interactions between them as links. according to the characteristics of these links and/or nodes, networks are divided into different types: weighted/ unweighted, unimodal/bimodal, bipartite, etc. once the network is built, the resulting topology allows to formally infer many relevant properties and patterns of the system as a whole (mata, 2020), the relative importance of the nodes (rodrigues, 2019), and/ or whether it presents structure in the mesoscale (fortunato & hric, 2016), among others. in specific applications, it is also possible to determine how the interaction structure —network structure— conditions the processes under study and vice versa. application examples of network science are bountiful and diverse, ranging from epidemic models (pastorsatorras, castellano, van mieghem, & vespignani, 2015), interaction between species and the evolution of social groups in ecology (bascompte, 2007), to the management of communications in nuclear emergency plans (ruiz-martin, ramirez-ferrero, gonzalez-alvarez, & lópez-paredes, 2015), the identification of efficacious combination therapies in drug development (cheng, kovács, & barabási, 2019) and many other applications in many different int. j. prod. manag. eng. (2022) 10(1), 65-75 creative commons attribution-noncommercial-noderivatives 4.0 international olivares-gil et al. 66 http://creativecommons.org/licenses/by-nc-nd/4.0/ contexts (havlin et al., 2012; newman, 2018; schweitzer et al., 2009). remarkably, since the erdős number became famous (grossman, 1997), the use of networks to analyze scientific interconnectedness and productivity has developed an important tradition. in fact, scientometrics and bibliometrics have made intensive use of network analysis to identify academic patterns. typically, the networks built to that end are based on article citations, being co-citation networks and bibliographic coupling common approximations (newman, 2018). the reason behind using article citations is that they are a good proxy for scholarly activity as, in general, when one article cites another, it indicates that the cited article is relevant in some way to the citing article. the first analyses in this line date back to the 1960s with the pioneering work of price (1965). in these studies, the articles constituted the nodes of the citation network, and directed links were used to indicate which articles cited or were cited by others. as regards co-citation networks, their links represent the number of other articles that simultaneously cite both, being hence undirected and weighted. eventually, in bibliographic coupling studies, the weight of the links represents the number of articles cited by both papers. thanks to these complementary approaches, it is possible not only to map different scientific areas, but also to shed light onto the relative influence of different scientific ideas, their evolution, the similarity or difference between papers, etc. furthermore, it is worth noting that the analysis of scientific networks is not restricted to article networks. as a matter of fact, co-authorship and social relations within the academia have also been explored using network approaches. notably, some of these works have served to better understand the social dimension of science and the formation processes behind the patterns found. see, for instance, the high clustering and small distance between researchers, compatible with the small-world property (watts, 1999), and/ or the heavy-tailed collaborative distributions identified in scientific networks (newman, 2001a, 2001b, 2001c). even models have been developed to characterize the evolution of co-authorship networks (barabási et al., 2002). 2.2. data the data used in this work was collected using the teseo database compiled by the spanish ministry of education, culture and sports (https://www.educacion.gob.es/teseo). this repository contains a unified database with all the doctoral theses from spanish universities since 1976. the database provides information on the title of the thesis, the university, the author, the date, the supervisors, the examining committee, and the thesis classification according to the unesco nomenclature for the fields of science and technology. this terminology was an international effort that began in 1966 and was successfully completed in 1988 to create a global standard system for classifying science and technology. although, initially, its objective was to classify research articles and doctoral theses, today, the classification standard is used for broader purposes — classification of research projects, academic positions, research lines, etc. (martínez-frías & hochberg, 2007). basically, the nomenclature is organized into three hierarchical levels of aggregation. the first level (2-digit code) is the level corresponding to the scientific field (e.g., chemistry, physics, medical science); the second level (4-digit code) establishes the level of scientific discipline (e.g., within chemistry: analytical chemistry, biochemistry, inorganic chemistry, etc.); and, finally, the third level (6-digit code) determines the level of the subdiscipline, thus corresponding to individual specializations in science and technology. in our contribution, we have filtered the complete database to the theses that contain the unesco code 5311xx —organization and management of enterprises. unlike previous studies, in this work, we have selected not only those theses that include the four-digit code but also those that include the six-digit code, i.e., the other nine subdisciplines that comprise the discipline: sales management, industry studies, manpower management, financial management, operations research, marketing, optimum production levels, organization of production and advertising. as for the names of the supervisors and members of the thesis examination committees, they have been pre-processed to reduce the lack of consistency presented by teseo in some fields of the database (castelló i cogollos, bueno cañigral, & valderrama zurián, 2019), and/or to identify possible academics who have been registered under different names. the filtered database has been formalized into a tripartite network (figure 1), that is latter transformed into two different bimodal and bipartite networks — int. j. prod. manag. eng. (2022) 10(1), 65-75creative commons attribution-noncommercial-noderivatives 4.0 international mapping the scientific structure of organization and management of enterprises using complex networks 67 https://www.educacion.gob.es/teseo http://creativecommons.org/licenses/by-nc-nd/4.0/ recall that bimodal means that there are two different groups of nodes, and bipartite implies that edges run only between nodes of unlike group. in the first bipartite network, the node groups are the theses and the scholars, while in the second network, the groups are the theses and their descriptors. as regards the theses-scholars network, there is a link between a scholar and a thesis if the academic has been a supervisor or a member of the examination committee of that thesis. note that the set of the unesco codes of each thesis has been kept as an attribute of the corresponding thesis node. the resulting network is constituted by 7911 scholar nodes and 3572 theses that were defended from 14th october 1991 to 27th february 2020. figure 1. the data structure of the analysis of this work corresponds to a tripartite network with three types of nodes, i) academics who have been part of an evaluation committee (tec) or supervised a thesis (s); ii) theses defended in the scientific field; and iii) unesco scientific field descriptors of each thesis. as for the second bipartite and bimodal network, there is a link between each thesis and each of its descriptors —i.e., the unesco codes used to describe it in teseo. recall that since the set of descriptors of each thesis is determined by its author, some inconsistencies might be found in relation to the categorization of topics. 3. analysis and results 3.1. communities of scholars and their specialization the first step of our analyses consisted in performing a simple weighting projection of the scholarstheses bipartite network onto the scholars’ space. one-mode projections are commonly used when dealing with bimodal/bipartite networks, as the set of mathematical tools available for their analysis is much more developed. however, since one-mode projections inevitably lead to the loss of information, it is important to choose a projection procedure with a suitable type of weighting, i.e., one that allows preserving as much information as possible. among the different projection possibilities, simple weighting is probably the most frequent of all of them. it consists in assigning a weight to the links that is equal to the number of times the common association is repeated (zhou, ren, medo, & zhang, 2007). in our case, the result of the one-mode projection onto the scholars was an undirected monomodal network in which the scholars constitute the nodes, and there exists a link between two of them if they have coincided in the same thesis (either as supervisors or as members of the examination committee). interestingly, the scholar monomodal network thus obtained reveals a giant component that contains more than 90% of the nodes —i.e., 90% of its nodes are connected to each other and belong to the same component. in addition, recall that each of the nodes in this network is endowed with two attributes: the first is the number of theses in which the researcher has participated; the second is her profile of specialization in organization and management of enterprises. to calculate this latter attribute, we propagated the unesco thesis descriptors to each scholar at the 6-digit (subdiscipline) level. more specifically, we considered the number of thesis descriptors so that they summed up to one for each thesis. for example, in a thesis with two descriptors, organization of production and industry studies, each subdiscipline counted 0.5. on its part, if the thesis had only the organization of production descriptor, it counted 1.0. once all the theses related to each researcher had been recorded in accordance with this procedure, we calculated the relative frequency of each unesco code for each person, hence obtaining the different research specialization profiles. to interpret this network, we filtered it to include only the scholars who had been in at least 10 doctoral theses —i.e., the scholars with a degree equal to or greater than 10. this threshold reduced the network from 7911 nodes and 41433 links to 305 nodes and 3088 links respectively, thus serving to eliminate noise, avoid considering spurious structure and helping to int. j. prod. manag. eng. (2022) 10(1), 65-75 creative commons attribution-noncommercial-noderivatives 4.0 international olivares-gil et al. 68 http://creativecommons.org/licenses/by-nc-nd/4.0/ identify the core patterns. the resulting network is typically referred to as backbone. afterwards, we explored the backbone with the louvain algorithm for community detection (blondel, guillaume, lambiotte, & lefebvre, 2008). such algorithm allows determining if a network presents a modular structure, i.e., if it has nodes densely connected to each other but weakly connected to the rest of the network. to that end, the louvain algorithm uses a modularity maximization heuristic, i.e., it seeks to maximize the difference between the number of actual links between each pair of nodes, and the expected number of links if they had been established at random while preserving the degree of each node. in the modularity formula (1) m is the number of links in the network; ki is the degree of node i; and ci is an integer representing the community of node i; δ(cicj) denotes the kronecker delta, which equals 1 if both nodes belong to the same community and 0 otherwise. q = 1 2m ∑ ij (aij − ki kj 2m )δ (cicj )  (1) remarkably, in the backbone of our scholars’ network, the algorithm found a modularity value of 0.584 and identified nine different communities (see figure 2), hence revealing the social structure of the network of researchers in the field of organization and management of enterprises in spain. in this regard, an interesting research question is whether such network can be explained according to the scientific sub-specialization of each community and/ or in accordance with the social relations between its members. to try to answer it, we conducted various complementary analyses. first, we analyzed the general specialization profile of each community from the profiles of the researchers who belong to it. the community profile was calculated in two ways: (i) through a consensus distribution obtained by averaging the profile of all the researchers in the community, and (ii) through a weighted average using the number of theses —i.e., the degree of the scholar— as weight; we found that the results are robust to both approximations. the subdiscipline distributions obtained for each community are represented in the lower part of figure 2. to determine whether the clusters are specialized or generalist, we calculated the normalized entropy of each subdiscipline distribution according to equation (2), where xi represents each subdiscipline and n the total number of subdisciplines —recall that the higher the entropy, the more generalist the specialization profile of the community and vice versa. n o r m s = − ∑ni=1 p(xi)log2p(xi) log2n   (2) the results show two highly specialized communities: community 4 (purple) in marketing, and community 6 (pink) in marketing and advertising; interestingly, these two communities are weakly connected between them. community 0 (blue) is also specialized in marketing, but with a certain level of financial management; it acts as a bridge between the purple and pink communities. the rest of the communities show an intermediate degree of specialization: the green community is focused on financial management; the communities 1 (orange) and 3 (red) present a relevant relative focus on manpower management; the 7 (gray) and 8 (yellow) combine manpower management with marketing and financial management in the former case and organization of production in the latter; community 5 (brown), which is strongly associated with the association for the development of management engineering (asociación para el desarrollo de la ingeniería de organización, adingor) —see garrido-labrador et al. (2022)— also presents an intermediate level of specialization in which organization of production, operations research, and industry studies are overrepresented. it is noteworthy that the most specialized communities are more peripheral in the network, as might be expected. however, some peripheral communities are rather generalist, hence challenging the explanation of the patterns observed exclusively on the grounds of scientific specialization, at least for the scale of analysis selected. to better understand this latter issue, we decided to address the problem as follows. we built an additional network in which the nodes represented each of the nine communities, and the weight of the links represented the number of shared links (edges between them) in the original projected network (figure 3). once such a network of communities was built, we used as proxy for the social component of the problem the weight of the links. under this approach, if two nodes have a high weight between them, they are assumed to be closely socially related and vice versa. as regards the second component of the problem —i.e., the similarity/dissimilarity in the thematic int. j. prod. manag. eng. (2022) 10(1), 65-75creative commons attribution-noncommercial-noderivatives 4.0 international mapping the scientific structure of organization and management of enterprises using complex networks 69 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 2. at the top of the figure, the backbone of the network projected onto the scholars is shown. the colors indicate the different communities obtained using the louvain algorithm. at the bottom of the figure, the specialization profile of each community is represented by its distribution and normalized entropy. int. j. prod. manag. eng. (2022) 10(1), 65-75 creative commons attribution-noncommercial-noderivatives 4.0 international olivares-gil et al. 70 http://creativecommons.org/licenses/by-nc-nd/4.0/ specialization of the different communities— it was formalized by means of the cosine similarity (3) of their specialization profiles (obtained from the descriptors of the different theses that each community is linked to in the bipartite network). cosine similarity is the cosine of the angle between two n-dimensional vectors, and it is calculated as the dot product of the two vectors divided by the product of their magnitudes —equation (3). cosθ = a ∙ b a b = ∑ni=1 ai bi ∑ni=1 a 2 i ∑ n i=1 b 2 i (3) in order to determine if the social proximity between communities (weights of the links in the community network) and their thematic proximity (cosine similarity) were correlated, we used the spearman coefficient. in this regard, to ascertain whether the correlation value thus obtained was significant or not, a reference correlation value (baseline) was required. to establish it, we assumed a null model according to which the network of scholars would have formed exclusively on the basis of the social relations between its members —i.e., regardless of their thematic specialization. thereupon, we maintained the empirically found social network by keeping the thesis evaluation committees untouched and randomized only the theses assigned to each scholar. comparison of the empirical results and the simulation results of the null model suggest that the formation of the different communities is partially driven by thematic specialization. figure 4 shows the p-value of the normalized entropy in each of the communities found. in all cases, the null hypothesis figure 3. collapsed community network of academics. in this network the nodes are the different communities previously identified in figure 2 and the weight of the links represent the number of shared links between them. the size of the node is proportional to the number of scholars in each community. figure 4. the figure shows the expected distribution of normalized entropy under the null hypothesis that thesis committees are randomly assigned. the empirically observed value is shown in yellow. in all cases, the null hypothesis is rejected at 0.05 significance level. int. j. prod. manag. eng. (2022) 10(1), 65-75creative commons attribution-noncommercial-noderivatives 4.0 international mapping the scientific structure of organization and management of enterprises using complex networks 71 http://creativecommons.org/licenses/by-nc-nd/4.0/ is rejected, i.e., the empirical communities are significantly more specialized than what would be expected according to the null model. subsequently, we studied the nature of the relationships between the 9 communities. our results show a positive spearman correlation of 0.428 between the social proximity between communities (number of shared links) and their thematic similarity (measured by the cosine similarity between the thematic specialization vectors of each community). although these results indicate a more intense interaction with the more thematically similar communities, the empirical correlation value is within what would be expected according to the distribution of the null model —i.e., that obtained by randomizing the theses assigned to each scholar through the permutation test (see figure 5). thus, with these results, it is not possible to conclude that the relationships found between communities are driven by thematic similarity rather than by the structure of the co-participation network. 3.2. organization and management of enterprises ego-network in the previous section, we identified the social structure of the field of organization and management of enterprises in spain and shed light onto the relationship of such structure with its own subdisciplines. here, we complete our study by assessing its interdisciplinary nature, i.e., how the discipline of organization and management of enterprises relates to other fields. to that end, we analyzed our second bipartite network, the one of theses and descriptors. please note that, as previously stated, this network was built by filtering the whole teseo database to include only those theses that have at least one unesco descriptor related to the discipline of organization and management of enterprises at the 4and/or 6-digit level. in this case, the bimodal network was transformed into a unimodal one by projecting it onto the descriptors using hyperbolic weighting (newman, 2001b). the reason behind the choice of hyperbolic weighting is as follows. in simple weighting, each node from the mode that we do not project onto —the theses in this case— contributes the same —a unit— to the weight of the respective link. however, in the problem at hand, it seems reasonable to think that two descriptors will be more closely related if they typically appear together but in the absence of any other descriptors —or accompanied by just a few of them— than if they appear together but in conjunction with plenty of other descriptors. this phenomenon —known as saturation effect— is accounted and compensated for in hyperbolic weighting. specifically, in hyperbolic weighting, the marginal contribution of each node decreases with the number of nodes to which it is connected to in the initial bipartite network. thereupon, in our particular case, it gives a weight to the link that is inversely proportional to the number of unesco codes present in the thesis. once the monomodal projection onto the descriptors was obtained, for the sake of interpretation it was analyzed at the 4-digit level. accordingly, in the projection process the descriptors of the different subdisciplines were assimilated to the discipline to which they belong —i.e., to the immediately higher 4-digit level. self-loops between fields were not considered in the analysis. the outcome obtained is an ego-network with 176 different nodes and organization and management of enterprises as the central node. in figure 5. the top part of the figure represents the relationship between the interaction proximity between the communities, measured by the number of shared theses (as evaluation committee or supervision) among their members, versus the scientific proximity measured by the cosine similarity of the thematic distributions of each community. the spearman coefficient found is 0.428. although the value is positive and relatively high, the significance analysis under the null hypothesis (bottom of the figure) shows that the value is compatible with the null hypothesis and cannot be rejected. int. j. prod. manag. eng. (2022) 10(1), 65-75 creative commons attribution-noncommercial-noderivatives 4.0 international olivares-gil et al. 72 http://creativecommons.org/licenses/by-nc-nd/4.0/ figura 6, however, the network shown has only 44 nodes, as it is the core obtained after removing the nodes that in the projection had a weighted degree of less than 10. figure 6 provides an accurate picture of the relationships between the discipline of organization and management of enterprises and the rest of the scientific disciplines in spain. in its immediate circle of proximity, there is an important association with other disciplines in the field of economic sciences (field 53), as could be expected. such connection is particularly intense with economic sciences and sectorial economics. more surprising is the relationship of the discipline with others in the field of psychology, such as social psychology and occupational and personal psychology, which form a cluster with other domains that are not so intensely related, such as different disciplines of sociology and psychology. apart from this, there is another cluster of theses that connects organization and management of enterprises with more instrumental disciplines, such as mathematics, statistics and operations research, being computer science the closest to the field of organization and management of enterprises. this latter cluster is also closer to applied and technological disciplines, especially in the industrial field. 4. conclusions and implications the conclusions and implications of the present study can be divided in accordance with the two perspectives adopted in our analyses. figure 6. ego-network of the organization and management of enterprises descriptor and its relationship with other descriptors in the teseo database. the circles represent the weighted degree ranges after the projection. colors represent the field of each discipline according to unesco classification. int. j. prod. manag. eng. (2022) 10(1), 65-75creative commons attribution-noncommercial-noderivatives 4.0 international mapping the scientific structure of organization and management of enterprises using complex networks 73 http://creativecommons.org/licenses/by-nc-nd/4.0/ the analysis of the backbone of the projection of the theses-scholars network onto the scholars demonstrates that specialization plays an important role in the discipline and shapes and determines the evaluation relationships. this structure is markedly modular, so it may be necessary to take it into account in general assessment processes so as to capture the nuances and differences of evaluation specific to each subdiscipline. secondly, our graph reveals the social patterns and research topics addressed in spain, allowing the identification of possible niches and research opportunities still to be developed within the discipline. the ego-network obtained after projecting the theses-descriptors network onto the descriptors also provides relevant insights. it confirms the eclectic nature of business organization, as it shows how it interacts with a wide range of disciplines. the map obtained shows that these relationships have various degrees of intensity, defining different circles of interaction. furthermore, the proposed approach also evinces the role of the discipline as a crossroads between a formal, technological pole and another pole focused on human relations and its scientific disciplines. besides, this latter analysis provides a formal tool to clarify the concept of related field (área afín), relevant in the spanish academic system. the accreditation committees and the areas of knowledge assigned to each of them are based on this idea. in addition, commissions for the selection of university faculty are usually composed of researchers from the same area as the required position and, in case of difficulty, by members of related areas. when there are problems in the assignment of teaching duties in universities caused by a deficit of human resources in a given field, sometimes it is considered that academics can teach specific courses from related areas. nevertheless, the definitions of the similarity and affinity between scientific fields and disciplines are not precise; they can evolve and are not exempt from possible subjectivities. such is so that the lists of related areas are generally approved by each university’s governing board and may differ from one to another. although the results obtained in our descriptors network do not show the relationship of the areas of knowledge, but rather the relationship between the disciplines according to the unesco international standard system, a mapping between the areas, the disciplines and the subdisciplines could serve to identify similarities and to establish the associations of affinity both appropriately and dynamically. acknowledgments the authors acknowledge financial support from the spanish ministry of science, innovation and universities (red2018-102518-t), the spanish state research agency (pid2020118906gb-i00 and pid2020-119894gb-i00 via aei/10.13039/501100011033), the junta de castilla y león – consejería de educación (bu055p20), fundación la caixa (2020/00062/001) and from nvidia corporation and its donation of the titan xp gpus that facilitated this research. this work was partially supported by the european social fund, as the authors josé miguel ramírez-sanz, josé luis garrido-labrador and alicia olivaresgil are the recipient of a predoctoral grant from the department of education of junta de castilla y león (va) (orden edu/875/2021). in addition, this work was also partially supported by the generalitat valenciana via its conselleria de innovación, universidades, ciencia y sociedad digital, as adrián arnaiz is recipicient of a predoctoral grant. the authors would also like to thank dr. manzanedo, dra. saiz-bárcena, dr. solé parellada, dr. izquierdo and dr. del olmo for their insightful help to improve the manuscript. references barabási, a., jeong, h., néda, z., ravasz, e., schubert, a., & vicsek, t. (2002). evolution of the social network of scientific collaborations. physica a: statistical mechanics and its applications, 311(3–4), 590–614. https://doi.org/10.1016/s0378-4371(02)00736-7 bascompte, j. (2007). networks in ecology. basic and applied ecology, 8(6), 485–490. https://doi.org/10.1016/j. baae.2007.06.003 blondel, v. d., guillaume, j.-l., lambiotte, r., & lefebvre, e. (2008). fast unfolding of communities in large networks. journal of statistical mechanics: theory and experiment, 2008(10), p10008. https://doi.org/10.1088/17425468/2008/10/p10008 int. j. prod. manag. eng. (2022) 10(1), 65-75 creative commons attribution-noncommercial-noderivatives 4.0 international olivares-gil et al. 74 https://doi.org/10.1016/s0378-4371(02)00736-7 https://doi.org/10.1016/j.baae.2007.06.003 https://doi.org/10.1016/j.baae.2007.06.003 https://doi.org/10.1088/1742-5468/2008/10/p10008 https://doi.org/10.1088/1742-5468/2008/10/p10008 http://creativecommons.org/licenses/by-nc-nd/4.0/ castelló i cogollos, l. c., bueno cañigral, f. j., & valderrama zurián, j. c. (2019). análisis de redes sociales y bibliométrico de las tesis españolas sobre drogodependencias en la base de datos teseo. adicciones, 31(4), 309–323. https://doi.org/10.20882/adicciones.1150 cheng, f., kovács, i. a., & barabási, a.-l. (2019). network-based prediction of drug combinations. nature communications, 10(1), 1197. https://doi.org/10.1038/s41467-019-09186-x fortunato, s., & hric, d. (2016). community detection in networks: a user guide. physics reports, 659, 1–44. https://doi.org/10.1016/j.physrep.2016.09.002 garrido-labrador, j. l., ramírez-sanz, j. m., ahedo, v., arnaiz-rodríguez, a., garcía-osorio, c., santos, j. i., & galán, j. m. (2022). network analysis of co-participation in thesis examination committees in an academic field in spain. dirección y organización. grossman, w. j. (1997). paul erdos: the master of collaboration. algorithms and combinatorics, 14, 467–475. havlin, s., kenett, d. y., ben-jacob, e., bunde, a., cohen, r., hermann, h., … solomon, s. (2012). challenges in network science: applications to infrastructures, climate, social systems and economics. the european physical journal special topics, 214(1), 273–293. https://doi.org/10.1140/epjst/e2012-01695-x latora, v., nicosia, v., & russo, g. (2017). complex networks. principles, methods and applications. cambridge, uk: cambridge university press. https://doi.org/10.1017/9781316216002 martínez-frías, j., & hochberg, d. (2007). classifying science and technology: two problems with the unesco system. interdisciplinary science reviews, 32(4), 315–319. https://doi.org/10.1179/030801807x183605 mata, a. s. da. (2020). complex networks: a mini-review. brazilian journal of physics, 50(5), 658–672. https://doi.org/10.1007/s13538-020-00772-9 newman, m. e. j. (2001a). scientific collaboration networks. i. network construction and fundamental results. physical review e statistical physics, plasmas, fluids, and related interdisciplinary topics, 64(1), 8. https://doi.org/10.1103/physreve.64.016131 newman, m. e. j. (2001b). scientific collaboration networks. ii. shortest paths, weighted networks, and centrality. physical review e statistical physics, plasmas, fluids, and related interdisciplinary topics, 64(1), 7. https://doi.org/10.1103/physreve.64.016132 newman, m. e. j. (2001c). the structure of scientific collaboration networks. proceedings of the national academy of sciences, 98(2), 404–409. https://doi.org/10.1073/pnas.98.2.404 newman, m. e. j. (2003). the structure and function of complex networks. siam review, 45(2), 167–256. https://doi.org/10.1137/s003614450342480 newman, m. e. j. (2018). networks. oxford, uk: oxford university press. https://doi.org/10.1093/ oso/9780198805090.001.0001 pastor-satorras, r., castellano, c., van mieghem, p., & vespignani, a. (2015). epidemic processes in complex networks. reviews of modern physics, 87(3), 925–979. https://doi.org/10.1103/revmodphys.87.925 price, d. j. s. (1965). networks of scientific papers. science, 149(3683), 510–515. https://doi.org/10.1126/ science.149.3683.510 repiso, r., torres, d., & delgado, e. (2011). análisis bibliométrico y de redes sociales en tesis doctorales españolas sobre televisión (1976/2007). (spanish). comunicar, 18(37), 151–159. https://doi.org/10.3916/c37-2011-03-07 rodrigues, f. a. (2019). network centrality: an introduction. in e. e. n. macau (ed.), a mathematical modeling approach from nonlinear dynamics to complex systems (pp. 177–196). springer. https://doi.org/10.1007/978-3-319-78512-7_10 ruiz-martin, c., ramirez-ferrero, m., gonzalez-alvarez, j. l., & lópez-paredes, a. (2015). modeling of a nuclear emergency plan: communication management. human and ecological risk assessment: an international journal, 21(5), 1152–1168. https://doi.org/10.1080/10807039.2014.955383 schweitzer, f., fagiolo, g., sornette, d., vega-redondo, f., vespignani, a., & white, d. r. (2009). economic networks: the new challenges. science, 325(5939), 422–425. https://doi.org/10.1126/science.1173644 sedighi, m. (2016). application of word co-occurrence analysis method in mapping of the scientific fields (case study: the field of informetrics). library review, 65(1–2), 52–64. https://doi.org/10.1108/lr-07-2015-0075 unesco, n. (1988). proposed international standard nomenclature for fields of science & technology. paris: united nations educational, scientific and cultural organization. villarroya, a., barrios, m., borrego, a., & frías, a. (2008). phd theses in spain: a gender study covering the years 19902004. scientometrics, 77(3), 469–483. https://doi.org/10.1007/s11192-007-1965-8 watts, d. j. (1999). small world. princeton, nj: princeton university press. https://doi.org/10.1515/9780691188331 zhou, t., ren, j., medo, m., & zhang, y. c. (2007). bipartite network projection and personal recommendation. physical review e, 76(4), 046115. https://doi.org/10.1103/physreve.76.046115 int. j. prod. manag. eng. (2022) 10(1), 65-75creative commons attribution-noncommercial-noderivatives 4.0 international mapping the scientific structure of organization and management of enterprises using complex networks 75 https://doi.org/10.20882/adicciones.1150 https://doi.org/10.1038/s41467-019-09186-x https://doi.org/10.1016/j.physrep.2016.09.002 https://doi.org/10.1140/epjst/e2012-01695-x https://doi.org/10.1017/9781316216002 https://doi.org/10.1179/030801807x183605 https://doi.org/10.1007/s13538-020-00772-9 https://doi.org/10.1103/physreve.64.016131 https://doi.org/10.1103/physreve.64.016132 https://doi.org/10.1073/pnas.98.2.404 https://doi.org/10.1137/s003614450342480 https://doi.org/10.1093/oso/9780198805090.001.0001 https://doi.org/10.1093/oso/9780198805090.001.0001 https://doi.org/10.1103/revmodphys.87.925 https://doi.org/10.1126/science.149.3683.510 https://doi.org/10.1126/science.149.3683.510 https://doi.org/10.3916/c37-2011-03-07 https://doi.org/10.1007/978-3-319-78512-7_10 https://doi.org/10.1080/10807039.2014.955383 https://doi.org/10.1126/science.1173644 https://doi.org/10.1108/lr-07-2015-0075 https://doi.org/10.1007/s11192-007-1965-8 https://doi.org/10.1515/9780691188331 https://doi.org/10.1103/physreve.76.046115 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.14985 received: 2021-01-20 accepted: 2021-07-20 integrating tier-1 module suppliers in car sequencing problem jung, e. hbpo gmbh. china. nazimemre@hotmail.com abstract: the objective of this study is to develop a car assembly sequence that is mutually agreed between car manufacturers and tier-1 module suppliers such that overall modular supply chain efficiency is improved. in the literature so far, only constraints of car manufacturers have been considered in the car sequencing problem. however, an assembly sequence from car manufacturer imposes a module assembly sequence on tier-1 module suppliers since their assembly activities are synchronous and in sequence with assembly line of that car manufacturer. an imposed assembly sequence defines a certain demand rate for tier-1 module suppliers and has significant impacts on operational cost of these suppliers which ultimately affects the overall modular supply chain efficiency. in this paper, a heuristic approach has been introduced to generate a supplier cognizant car sequence which does not only provide better operational conditions for tier-1 module suppliers, but also satisfies constraints of the car manufacturer. key words: car sequencing, module assembly, synchronous assembly. 1. introduction car manufacturers have been seeking ways for more flexible and efficient processes to cope with the evolving environment of the automotive industry. in this regard, modules are perceived as an engineering tool for companies to manage complex products by dividing them into subassemblies. modular assembly concept offers car manufacturers the ability of efficient mass customization by enabling the postponement of final assembly of a product until customer orders have been received (fredriksson and gadde, 2005). one of the distinctive characteristics of modularity is synchronous production. synchronous production is defined by doran (2002) as an integrated supply chain approach which ensures delivery of products that are defect-free and match the exact requirements of the customer reflecting vehicle rather than model. because of this production model, there is high pressure on the module suppliers since the whole vehicle assembly process at car manufacturer depends on the timely delivery of their modules in the right sequence (larsson, 2002). assembly sequence of the car manufacturer imposes a module assembly sequence on tier-1 module suppliers since their assembly activities are synchronous and in sequence with assembly line of that car manufacturer. an imposed assembly sequence defines a certain demand rate for tier-1 module suppliers and has significant impacts on operational cost of these suppliers which ultimately affects the overall modular supply chain efficiency. in this study, we try to develop a car assembly sequence that is mutually agreed between car manufacturers and tier-1 module suppliers such that overall modular supply chain efficiency is improved. to cite this article: jung, e. (2021). integrating tier-1 module suppliers in car sequencing problem. international journal of production management and engineering, 9(2), 113-123. https://doi.org/10.4995/ijpme.2021.14985 int. j. prod. manag. eng. (2021) 9(2), 113-123creative commons attribution-noncommercial-noderivatives 4.0 international 113 https://orcid.org/0000-0003-0221-8874 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. literature review car sequencing in mixed model assembly line depends on the controlling goals or purposes, such as to minimize the variation in rate of consuming the parts of the sequence (monden, 1998). gottlieb et al. (2003) explained that common car sequencing problems in the literature involve scheduling cars along an assembly line where options are installed at different assembly stations. these assembly stations are designed to handle a certain percentage of the entire assembly work while cars are passing along the assembly line. this approach intends to minimize work overload at any assembly station. installing different options at a station results in various assembly times faced by that station. if car sequence lets several consecutive cars with same options which require longer assembly time to be assembled at a certain assembly station, then work overload is possible at that station. when workers get too much workload, there is a high possibility of making assembly mistakes which leads to increase in cost of quality or causing a line stoppage in the worst case. on the other hand, the workers in the successive stations may be idle while waiting for those options to be installed at that earlier assembly station and this leads again to higher production cost. therefore, cars requiring this option must be spaced such that the capacity of the station is never exceeded. parrello et al. (1986) and solnon et al. (2008) introduced sequencing rules ho:no for each labour-intensive option o, which restrict the occurrence of this option to at most ho in any subsequence of no successive models. the goal is to find a sequence, which does not violate any of the given sequencing rules or – if such a sequence is not existent – minimizes rule violations. boysen et al. (2009) define this sequencing rule as of type ho:no, which means that out of no successive models, only ho may contain the option o in order to avoid work overload. this is also known as the option rule in the literature. drexl and kimms (2001) provide an intuitive example: “assume that 60% of the cars manufactured on the line require the option ‘sunroof’. moreover, assume that five cars pass the station where the sunroofs are installed during the time for the installation of a single copy. then, three operators (installation teams) are necessary for the installation of sunroofs. hence, the capacity constraint of the final assembly line for the option ‘sunroof’ is three out of five in a sequence, or ‘3:5’ for short.” since 70% of the car value is built on the assembly line on the average, car sequencing problems in the literature considered final assembly line constraints that ensure load balancing and component supply to find an assembly sequence (gagne et al., 2006). in the literature, workload balancing or minimizing work overload and levelling component usage are two basic objectives of sequencing. the csp is strongly np-hard (estellon and gardi, 2006). to solve a csp problem with one hundred or so vehicles and few options, the use of constraint programming or integer programming has limits and several heuristics have been proposed such as ant colony optimization, greedy algorithms or local search (estellon et al., 2008). many researchers studied car sequencing problem with component levelling objectives, mainly as a key element of jit philosophy, but none has considered constraints of module assembly in that problem as a synchronous production and in-sequence delivery concept. module suppliers are critical partners in the automotive supply chain due to characteristics of synchronous production and delivery. due to the same characteristics, module suppliers are directly affected by production scheduling process of car manufacturer, especially car sequencing. assembling modules that are synchronized with car assembly line and delivering them in sequence leave no room for module suppliers to implement their own production schedules but to follow the one from car manufacturer. therefore, car sequence of the car manufacturer is directly affecting production output and productivity of module suppliers. car sequence generated by any available algorithm imposes an assembly sequence on tier-1 module suppliers, which ultimately defines the requested demand rate from these suppliers. in this study, a car assembly sequence that is mutually agreed between car manufacturer and module suppliers is developed. the proposed approach considers not only the objectives of car manufacturer, as all academic studies have done so far, but also the operational constraints of tier-1 module suppliers since assembly activities of these suppliers are synchronous with assembly line of car manufacturers. 3. impact of car sequencing on tier-1 module suppliers figure 1 illustrates a common module call-off and sequential delivery process of tier-1 module int. j. prod. manag. eng. (2021) 9(2), 113-123 creative commons attribution-noncommercial-noderivatives 4.0 international jung 114 http://creativecommons.org/licenses/by-nc-nd/4.0/ suppliers to car manufacturers following the vehicle assembly sequence. a, b and c represent car models to be produced where their subscripts define options of these vehicles. respective modules are assembled and delivered to the car assembly line by tier1 module suppliers matching with the car assembly sequence at the car manufacturer accordingly. in figure 1, there are five cars to be assembled in a sequence which is planned earlier by the car manufacturer (i.e. a1, c1, a2, b1, a1). notations of car models a, b and c present their options meaning that car model a has two available options (i.e. a1 and a2), whereas car model b and car model c have only single option (i.e. b1 and c1). these cars consist of three modules (i.e. module x, module y, and module z), which are to be assembled at their respective assembly stations (i.e. station x, station y, and station z). these modules are provided by different suppliers. module x is supplied solely by supplier 1 for all car models. module y is supplied by two different suppliers (i.e. supplier 2 and supplier 3). supplier 2 assembles and delivers module y for car model a, whereas supplier 3 assembles and delivers module y for car models b and c. in case of module z, there are three module suppliers, and each supplier supplies only one car model (i.e. supplier 4 for car model a, supplier 5 for car model b, and supplier 6 for car model c). in case of supplier 1, the module supplier is responsible for building all necessary modules for all car models produced by the car manufacturer and adjusting the sequence of the modules according to the car assembly sequence defined by the car manufacturer. this case is relatively simple to manage for both the car manufacturer and the module supplier. it is even feasible to match module assembly sequence in the module supplier exactly with the car assembly sequence at the car manufacturer. it is also possible that modules for each car model are supplied by different module suppliers, such as module z case. in this case, each module supplier receives assembly work order for the respective car model as well as the overall car assembly sequence for reference since each supplier is responsible for sequential delivery. after assembly of respective module by each supplier, these modules should be placed in the correct sequence that matches with figure 1. tier-1 module call-off and sequential delivery concept in automotive industry. int. j. prod. manag. eng. (2021) 9(2), 113-123creative commons attribution-noncommercial-noderivatives 4.0 international integrating tier-1 module suppliers in car sequencing problem 115 http://creativecommons.org/licenses/by-nc-nd/4.0/ the car assembly sequence of the car manufacturer. at this stage, owing to complexity of the delivery process, information exchange between all parties becomes extremely critical, and the involvement of the car manufacturer is unavoidable to define, and lead tasks related to delivery process and responsibilities between module suppliers. assembly work at a tier-1 module supplier starts only when an assembly order from car manufacturer is received. after assembly process is completed, delivery of assembled modules is done respecting the car assembly sequence and within the given time frame defined by the car manufacturer. due to the nature of synchronous production, module suppliers have different operational challenges than just-in-time (jit) suppliers. a jit component only becomes critical when stock levels of supplier are insufficient to meet forecast order volumes (hellingrath, 2008). in contrast, a tier-1 module supplier must be ready for production at any time as required by nature of synchronous production concept. it is obliged to assemble and deliver modules as soon as an assembly order from the car manufacturer is received. therefore, its module is a critical component, and the capacity of tier-1 module supplier becomes critical as well. assembly line of module supplier is designed bearing in mind average production output of the car manufacturer. workforce assigned for module assembly job is also defined and dispatched to module assembly line according to this average output level. if a car manufacturer follows a uniform demand rate d that is equal to this average output, workload at the module supplier would be quite balanced following car manufacturer’s assembly line speed. however, a fluctuating demand rate from the car manufacturer would result in a changing workload at the assembly line of the module supplier. at a lower than average demand rate, assembly line workers at the module supplier would be facing idle time. on the contrary, when the demand rate is higher than average demand rate, the module assembly line workers would be overloaded at their respective assembly stations. in order to cope with such workload conditions at module suppliers, car manufacturer and module suppliers usually agree on a certain flexibility in addition to an average demand rate d. this flexibility is necessary for avoiding any possible vehicle assembly line stoppages at the car manufacturer. it is termed as “flexibility corridor” in the literature. the flexibility corridor is defined by a negotiable percentage of the contract volume, where the supplier is obliged to cover all requests (niemann et al., 2019). if actual volume goes beyond this flexibility corridor (i.e. maximum volume level), then additional measures must be taken such as additional line investment and/or hiring additional workforce. therefore, an improvement in the overall modular supply chain seems to be feasible by reducing this gap as much as possible. however, module supplier alone is not able to reduce this gap without cooperation of the car manufacturer. module supplier that faces a big demand drop would have to deal with operational issues such as unstable inventory, excess assembly line capacity, idle workforce, and fixed operational costs. on the other hand, if total output of the car manufacturer is stable, other module supplier providing modules to other car models at the same assembly line of the car manufacturer would face an opposite trend, an increase in demand. at the end, car manufacturer may offer the former supplier a huge compensation amount for its loss due to missing volume of the project, meanwhile it had to settle capacity issues with the latter supplier. otherwise car manufacturer might review its strategy to outsource modules to avoid such schedule related economic impact on module suppliers. this economic impact seems to be avoided only if modules for all car models are provided by one module supplier. since, regardless of the vehicle mix and individual models, module supplier is obliged to respond to the whole vehicle schedule and sequence, it would be exposed to total production volume of the car manufacturer as demanded volume instead of only one specific car model. however, allowing only one module supplier for all the cars on the assembly line leads to a monopoly at the end. in that case, purchasing power of car manufacturer over the module suppliers will be damaged, which might lead to other commercial issues between car manufacturer and the module suppliers. a heuristic approach has been introduced to show the impact of car sequencing on tier-1 module suppliers and to generate a supplier cognizant car sequence trying to eliminate this impact. we start by generating a car sequence as a first step and review impacts of this car sequence in terms of demand imposed on tier-1 module suppliers by the car manufacturer. afterwards, possibility of improvement is studied by involving tier-1 module suppliers in the car sequencing problem. figure 2 shows the methodology used in this study. int. j. prod. manag. eng. (2021) 9(2), 113-123 creative commons attribution-noncommercial-noderivatives 4.0 international jung 116 http://creativecommons.org/licenses/by-nc-nd/4.0/ create additional option rules for each tier-1 module supplier check impact of the new car sequence s 2 on tier-1 module suppliers generate a new car sequence (s 2) satisfying option rules of both the car manufacturer and tier-1 module suppliers implement new car sequence s 2 at the car manufacturer are demand rates on tier-1 module suppliers improved? y n check impact of the generated car sequence s 1 on tier-1 module suppliers generate an initial car sequence (s 1) satisfying option rules of the car manufacturer figure 2. methodology to develop supplier cognizant car sequencing. 4. car sequencing problem car sequencing problem is a well-studied problem in literature as well as in the automotive industry. in this study, we have used the car sequencing algorithm written in java programming language by localsolver (benoist et al., 2011), a french software editor company, specializing in the field of optimization and decision support. localsolver utilizes a hybrid approach of very large-scale neighbourhood search (vlns) and very fast local search (vfls), which is the best-known approach for solving car sequencing problems (regin and puget, 1997; estellon et al., 2006; estellon et al., 2008). it is claimed to have a hybrid local search heuristic based on very fast explorations of small neighbourhoods (benoist et al., 2011). the algorithm applies one transformation at each iteration to the current sequence which modifies it only very locally. pseudo code of the algorithm is shown in figure 3. once initial sequence is built by a greedy algorithm, five basic transformation strategies are used in the algorithm as listed below: swap, forward insertion, backward insertion, reflection, and random shuffle (estellon et al., 2008). in this section, sample problem carseq_300_8_20_25 from csplib (gagne et al., 2006) is used in order to analyze the feasibility and efficiency of the proposed supplier cognizant car sequencing concept. all numerical experiments were performed on a standard computer equipped with the operating system windows 10 64-bits and the chip intel pentium g3420 (3.20 ghz, ram 4 gb). the problem is defined as shown in figure 4. 300 cars must be manufactured in this problem. the number of options is 8, and the class size is 20. first line of the problem states the number of cars (i.e. total demand) that are to be produced; number of options available for these cars and number of classes (i.e. number of car models) in this demand algorithm localsolver begin; compute initial sequence; while number of violations > 0 and execution time limit is not reached do choose transformation (swap, forward/backward insertion, reflection or shuffle) and positions where applying it; if transformation does not result in a higher number of violations compared to current sequence then update current sequence by performing it; end if; end do; return current sequence; end; figure 3. pseudo code of the car sequencing algorithm. int. j. prod. manag. eng. (2021) 9(2), 113-123creative commons attribution-noncommercial-noderivatives 4.0 international integrating tier-1 module suppliers in car sequencing problem 117 http://creativecommons.org/licenses/by-nc-nd/4.0/ package. second and third line of the problem define the option rules. for each option (columns), the maximum number of cars allowed with that option in a block (second line) and the block size (third line) are shown. the rest of the problem presents car model in the first column (i.e., model index number); number of cars to be produced for this model in the second column; and for each option (remaining columns), whether this model requires this option or not (1 or 0). a feasible solution (i.e. assembly sequence for 300 cars) with zero violation obtained by localsolver after 2,207,246 iterations in 44 seconds is shown in figure 5. the number at the first row shows the number of option rule violations achieved by the proposed car sequence. index of car classes is shown in the car sequence from the second row onwards. the table must be read from the left to the right and from the top to the bottom. therefore, feasible car sequence starts with the first car at the first column in the second row and is followed by its successor car on its right in the table. let us assume that a module supply chain is involved in this production and two module suppliers are delivering module x (such as seat module) to the car manufacturer but for different car models. naturally, variants of module x for different car models are different from each other. however, product structure and characteristics in each module x are similar, therefore it is feasible that module x for each model can be assembled on the same assembly line at both module suppliers. let us define two tier-1 module suppliers delivering module x for the car models presented in the csplib problem carseq_300_8_20_25 as supplier a and supplier b. let us also assume that supplier a assembles and delivers modules for car models (or car classes) from number 0 to number 8, and supplier b assembles and delivers modules for car models from number 9 to number 19. since a feasible car sequence is generated, following this car sequence, a demand would be imposed on these tier-1 module suppliers. supplier a and supplier b synchronize their module assembly lines with the assembly line of the car manufacturer following this car sequence. let us consider that production rate at the car manufacturer (p) is 15 cars per hour. it means that completing assembly of 300 cars is a matter of 20 hours. supplier a is assigned to assemble and deliver modules for 152 cars out of 300 cars to be produced. remaining demand of 148 cars is to be supplied by supplier b. in an ideal case, following their demand of 152/300 cars and 148/300 cars, supplier a and supplier b would follow an hourly demand rate of 7.6 modules per hour and 7.4 modules per hour respectively. therefore, we can consider that average hourly demand rate of 7 300 8 20 2 1 2 1 3 1 1 2 3 3 4 3 5 3 3 4 0 13 1 0 0 0 0 0 0 0 1 19 0 1 0 0 0 0 0 0 2 12 0 0 1 0 0 0 0 0 3 19 0 0 0 1 0 0 0 0 4 11 0 0 0 0 1 0 0 0 5 26 0 0 0 0 0 1 0 0 6 16 0 0 0 0 0 0 1 0 7 18 0 0 0 0 0 0 0 1 8 18 0 1 0 1 1 1 0 0 9 10 0 0 1 1 0 1 0 0 10 14 1 0 1 0 0 0 0 0 11 10 0 0 1 0 0 0 1 0 12 14 1 0 0 0 1 1 1 1 13 14 0 1 1 0 0 0 0 0 14 13 0 0 1 0 1 1 1 0 15 14 0 0 0 0 1 1 0 0 16 15 0 1 0 1 0 0 1 1 17 18 1 1 0 1 0 0 1 0 18 11 1 0 0 0 0 0 1 0 19 15 1 1 1 1 1 0 0 0 figure 4. csplib problem carseq_300_8_20_25. 0 8 11 4 8 10 11 8 4 7 16 15 4 17 5 4 17 5 7 17 5 0 16 5 0 17 5 10 16 5 0 16 5 0 16 5 7 19 14 7 3 14 13 3 12 7 19 4 18 8 10 6 8 10 18 8 0 6 19 5 6 19 5 6 19 5 6 1 9 11 1 3 12 13 3 14 1 3 12 1 7 14 1 3 12 13 3 14 7 19 12 2 7 17 15 7 16 5 4 16 15 2 17 5 0 17 5 2 17 5 10 16 15 0 16 15 0 17 5 0 17 15 10 17 5 2 16 5 7 17 5 10 17 5 4 13 12 3 1 14 3 13 6 9 1 18 9 1 18 9 1 6 9 1 6 9 1 18 9 1 11 5 19 6 5 19 11 4 8 18 10 8 6 10 8 18 10 8 11 7 8 6 10 8 11 7 8 6 2 19 12 7 19 5 6 19 15 2 16 5 0 17 15 2 16 15 2 17 15 7 17 15 10 16 15 0 16 5 10 16 15 2 17 15 7 17 5 7 19 12 2 1 14 3 13 12 4 13 14 3 1 14 3 1 14 3 13 12 4 19 6 2 8 18 0 8 11 7 8 11 10 8 6 0 8 11 7 19 12 2 1 14 3 13 18 9 1 18 9 13 18 3 13 12 3 1 14 3 13 12 3 13 14 3 1 6 9 13 4 12 19 figure 5. feasible solution for csplib problem carseq_300_8_20_25. int. j. prod. manag. eng. (2021) 9(2), 113-123 creative commons attribution-noncommercial-noderivatives 4.0 international jung 118 http://creativecommons.org/licenses/by-nc-nd/4.0/ or 8 modules are two acceptable workload scenarios for these module suppliers. however, car sequence generated in figure 5 imposes actual demand rates on both suppliers which are quite different than the ideal case. table 1 shows this actual demand that both suppliers face for 20 hours (based on 15 cars per hour production rate of the car manufacturer). it can be observed that during 20 hours of production, there are 10 production windows (i.e. one-hour production time) when supplier a and supplier b are facing either higher or lower hourly demand rates than their average hourly demand rate. specifically, at 15th and 19th hours, it can be observed that the gap between actual demand rates and average demand rates for both suppliers are big. supplier a is facing idle time due to lower demand rate than its average demand rate (47% lower at 15th hour and 34% lower at 19th hour). on the contrary, supplier b is overloaded with higher demand rate than its average demand rate (49% higher at 15th hour and 35% higher at 19th hour). the remaining 8 production windows, when average demand rate is not followed, also impose a demand fluctuation of approximately 20% on both suppliers. table 1. hourly demand rate of module suppliers. (unit: number of modules). supplier a supplier b 1st hour 9 6 2nd hour 9 6 3rd hour 8 7 4th hour 9 6 5th hour 8 7 6th hour 8 7 7th hour 8 7 8th hour 7 8 9th hour 7 8 10th hour 7 8 11th hour 8 7 12th hour 8 7 13th hour 9 6 14th hour 6 9 15th hour 4 11 16th hour 8 7 17th hour 9 6 18th hour 9 6 19th hour 5 10 20th hour 6 9 5. supplier cognizant car sequencing the purpose of supplier cognizant car scheduling is to avoid any idle time or work overload at the module suppliers if possible. it means that the hourly demand rate faced with each module supplier needs to be distributed as uniform as possible over the production time. in this section, possibility of improving the gap between actual demand rate and average demand rate for supplier a and supplier b have been studied by involving them in the car sequencing problem. we integrate tier-1 module suppliers in the car sequencing problem by utilizing option rules. we introduce modules of supplier a and supplier b as additional options for each car model (i.e. number of options in csplib carseq_300_8_20_25 increases from 5 to 7). only one module (either delivered by supplier a or supplier b) can be assigned to a car model. afterwards, we must define option rules for these modules. the aim of the supplier cognizant car sequencing is to provide tier-1 module suppliers a uniform demand rate as smooth as possible. therefore, we need to consider the size of production rate of the car manufacturer as block size for module options. at this point, we would like to calculate average demand rate for supplier a and supplier b. it is worth mentioning that demand rate is important for these suppliers as they can calculate their necessary takt time and design their module assembly lines accordingly. for any tier-1 module supplier s (s = 1,..,s), whose assembly line is synchronized with the assembly line of the car manufacturer, average demand rate of the supplier (ds) can be calculated as; d s = p · module demand to be supplied by suppliers total number of cars to be produced (1) where. s s =1 d s = p following equation (1), we can calculate average demand rates for supplier a and supplier b as below. d a = 15 · 152 300 = 7.6 modules per hour d b = 15 · 148 300 = 7.4 modules per hour now, we can define the option rules for these modules from supplier a and supplier b. option block sizes will be equal to car manufacturer’s production rate p (15 modules per hour or 15 cars in one hour) since it was the base for calculating average demand rates of module suppliers as well. within this block size of 15 cars, we expect to see 7.6 cars equipped with modules from supplier a and 7.4 cars equipped with int. j. prod. manag. eng. (2021) 9(2), 113-123creative commons attribution-noncommercial-noderivatives 4.0 international integrating tier-1 module suppliers in car sequencing problem 119 http://creativecommons.org/licenses/by-nc-nd/4.0/ modules from supplier b. therefore, we need to allow 8 cars in a row of 15 cars for modules from supplier a and 8 cars in a row of 15 cars for modules from supplier b. the methodology of integrating module suppliers in car sequencing problem can be summarized as below: step 1. consider each module supplier as an artificial car option. add one new car option for each module supplier. step 2. for the newly added car option, set block size equal to the production rate of the car manufacturer (i.e. cars per hour). step 3. calculate average demand rate for each module supplier using equation (1) and set maximum number of allowed cars in newly added car option block equal to the closest higher integer value ( ds ). step 4. assign module suppliers to respective car models. modified csplib problem carseq_300_8_20_25 is shown in figure 6. a feasible car sequence is obtained by localsolver without any violations after 49,944,691 iterations in 1,386 seconds as shown in figure 7. figure 6. modified problem carseq_300_8_20_25. this new car sequence does not only provide better operational conditions for tier-1 module suppliers, but it also satisfies constraints of the car manufacturer. therefore, in addition to the benefits of supplier cognizant car sequencing to tier-1 module suppliers and therefore to module supply chain, it does not have any negative impact on the car manufacturer in terms of car sequencing. figures 8 and 9 present the hourly demand rates for supplier a and supplier b respectively during 20 hours of production for both initial car sequence case and modified car sequence (i.e. supplier cognizant car sequence) case. table 2. hourly demand rate of module suppliers (supplier cognizant car sequence). (unit: number of modules). supplier a supplier b 1st hour 8 7 2nd hour 8 7 3rd hour 8 7 4th hour 7 8 5th hour 8 7 6th hour 7 8 7th hour 8 7 8th hour 8 7 9th hour 7 8 10th hour 7 8 11th hour 8 7 12th hour 7 8 13th hour 8 7 14th hour 7 8 15th hour 8 7 16th hour 7 8 17th hour 8 7 18th hour 7 8 19th hour 8 7 20th hour 8 7 0 8 18 2 8 11 7 19 5 6 19 15 6 13 9 6 1 9 18 1 9 6 1 9 6 1 9 18 1 9 6 13 3 12 13 4 11 8 10 6 8 7 11 8 10 6 19 7 12 13 3 14 7 19 5 4 17 5 4 16 15 4 17 15 7 16 5 0 16 5 0 17 5 10 17 5 0 17 15 7 17 5 4 16 5 10 17 4 15 16 2 5 16 7 15 17 7 5 16 0 5 16 10 5 17 7 5 16 2 15 17 0 5 19 4 12 13 3 14 1 3 12 13 3 14 1 3 14 13 3 12 13 3 14 1 3 12 13 3 14 1 3 12 13 3 14 1 2 12 19 7 14 1 3 12 2 19 12 7 19 5 6 13 9 6 1 9 11 4 8 18 10 8 18 4 8 11 10 8 6 10 8 11 7 8 11 10 8 18 0 8 18 10 8 18 7 19 14 0 1 14 3 13 12 3 1 12 3 13 12 3 13 14 3 1 14 3 1 14 0 19 5 0 17 15 0 16 15 2 17 5 4 16 5 7 17 0 5 16 10 15 17 4 7 16 5 2 16 5 10 16 7 5 17 2 15 17 0 15 17 2 5 17 7 15 6 19 5 11 1 9 18 1 9 18 0 8 11 2 1 12 2 19 6 5 19 18 15 19 11 7 8 6 10 8 6 10 8 6 2 19 figure 7. feasible solution for modified problem carseq_300_8_20_25. int. j. prod. manag. eng. (2021) 9(2), 113-123 creative commons attribution-noncommercial-noderivatives 4.0 international jung 120 http://creativecommons.org/licenses/by-nc-nd/4.0/ 6. conclusion in this research, we endeavoured to show the impact of car assembly sequence, which is imposed by car manufacturer on tier-1 module suppliers. as this impact affects the cost performance of the module supply chain, we argued that car sequencing, which is done by the car manufacturer, should consider modules as one of constraints while deciding car sequence to be assembled to lessen this impact. as there is no such study in the literature so far, we suggested a new approach to integrate module assembly in the car sequencing problem. we proposed a supplier cognizant car assembly sequencing concept by adding module suppliers as options and defining their option rules following the average demand rates they are required to respond to. the main reason behind suggesting a supplier cognizant car assembly sequence is to avoid idle time or work overload of assembly operators at module suppliers figure 8. change in hourly demand rate of supplier a. figure 9. change in hourly demand rate of supplier b. int. j. prod. manag. eng. (2021) 9(2), 113-123creative commons attribution-noncommercial-noderivatives 4.0 international integrating tier-1 module suppliers in car sequencing problem 121 http://creativecommons.org/licenses/by-nc-nd/4.0/ that is caused by an imposed car assembly sequence. both idle time and work overload happening at module suppliers lead to wastage of resources and have cost impact on the module suppliers as well as overall module supply chain. the overall module supply chain profitability would improve if the module supply chain cost could be reduced by implementing a supplier cognizant car sequencing, which targets to satisfy both operational constraints of car manufacturer as well as module suppliers. at the end, we showed that a car assembly sequence, which still satisfies the workload requirements of the car manufacturer and respects workload of module assembly line of tier-1 module suppliers, is possible. option rules were used as hard constraints for car sequencing problem. the option rules were defined considering requirements of car manufacturer and operational constraints of module suppliers. further studies can be conducted by using soft constraints. this approach would allow having different weights for operational constraints of car manufacturer and tier-1 module suppliers. at the end, some constraints imposed by car manufacturer may have priority over operational constraints of tier-1 module suppliers, therefore they should be respected even if they might end up resulting in certain cost due to unsatisfied operational constraints of those module suppliers. moreover, this study can be enhanced by exploring different volume scenarios between module suppliers, which may reduce the impact imposed by car manufacturer on the module suppliers. the implications of this study for manager of car manufacturers are clear. the improvement possibility of the tier-1 module supply chain is evident and worth exploring. supplier cognizant car sequencing does not only eliminate waste of resources but also ensures reliability of module supply chain by eliminating high demand fluctuations. depending on several factors such as which modules to consider and to what extent supplier cognizant car sequencing can be realized, performance of the concept will differ. references benoist, t., gardi, f., megel, r., nouioua, k. 2011. localsolver 1.x: a black-box local-search solver for 0-1 programming. 4or a quarterly journal of operations research, 9(299). https://doi.org/10.1007/s10288-011-0165-9 boysen, n., fliedner, m., scholl, a. 2009. sequencing mixed-model assembly lines: survey, classification and model critique. european journal of operational research, 192, 349-373. https://doi.org/10.1016/j.ejor.2007.09.013 doran, d. 2002. manufacturing for synchronous supply: a case study of ikeda hoover ltd. integrated manufacturing systems, 13(1), 18-24. https://doi.org/10.1108/09576060210411477 drexl, a., kimms, a. 2001. sequencing jit mixed-model assembly lines under station-load and part-usage constraints. management science, 47,(3), 480–491. https://doi.org/10.1287/mnsc.47.3.480.9777 estellon, b., gardi, f. 2006. car sequencing is np-hard: a short proof. journal of the operational research society, 64, 1503-1504. https://doi.org/10.1057/jors.2011.165 estellon, b., gardi, f., nouioua, k. 2006. large neighborhood improvements for solving car sequencing problems. rairo operations research, 40(4), 355-379. https://doi.org/10.1051/ro:2007003 estellon, b., gardi, f., nouioua, k. 2008. two local search approaches for solving real-life car sequencing problems. european journal of operational research, 191(3), 928–944. https://doi.org/10.1016/j.ejor.2007.04.043 fredriksson, p., gadde, l.e. 2005. flexibility & rigidity in customization and build-to-order production. science direct industrial marketing management, 34, 695-705. https://doi.org/10.1016/j.indmarman.2005.05.010 gagne, c., gravel, m., price, w. 2006. solving real car sequencing problems with ant colony optimization. european journal of operational research, 174(3), 1427–1448. https://doi.org/10.1016/j.ejor.2005.02.063 gottlieb, j., puchta, m., solnon., c. 2003. a study of greedy, local search and ant colony optimization approaches for car sequencing problems. in applications of evolutionary computing, lecture notes in computer science, 2611. springer, berlin, heidelberg. https://doi.org/10.1007/3-540-36605-9_23 hellingrath, b. 2008. key principles of flexible production and logistics networks. build to order: the road to the 5-day car, springer-verlag, london, 177-180. https://doi.org/10.1007/978-1-84800-225-8_10 larsson, a. 2002. the development and regional significance of the automotive industry: supplier parks in western europe. international journal of urban and regional research, 26(4), 767-784. https://doi.org/10.1111/1468-2427.00417 monden, y. 1998. toyota production systems: an integrated approach to just-in-time, 3rd edition. industrial engineering & management press, norcoss. https://doi.org/10.1007/978-1-4615-9714-8 int. j. prod. manag. eng. (2021) 9(2), 113-123 creative commons attribution-noncommercial-noderivatives 4.0 international jung 122 https://doi.org/10.1007/s10288-011-0165-9 https://doi.org/10.1016/j.ejor.2007.09.013 https://doi.org/10.1108/09576060210411477 https://doi.org/10.1287/mnsc.47.3.480.9777 https://doi.org/10.1057/jors.2011.165 https://doi.org/10.1051/ro:2007003 https://doi.org/10.1016/j.ejor.2007.04.043 https://doi.org/10.1016/j.indmarman.2005.05.010 https://doi.org/10.1016/j.ejor.2005.02.063 https://doi.org/10.1007/3-540-36605-9_23 https://doi.org/10.1007/978-1-84800-225-8_10 https://doi.org/10.1111/1468-2427.00417 https://doi.org/10.1007/978-1-4615-9714-8 http://creativecommons.org/licenses/by-nc-nd/4.0/ niemann, j., seisenberger, s., schlegel, a., putz, m. 2019. development of a method to increase flexibility and changeability of supply contracts in the automotive industry. 52nd cirp conference on manufacturing systems, ljubljana, slovenia, june 12-14. https://doi.org/10.1016/j.procir.2019.03.045 parrello, b.d., kabat, w.c., wos, l. 1986. job-shop scheduling using automated reasoning: a case study of the car-sequencing problem. journal of automated reasoning, 2(1), 1–42. https://doi.org/10.1007/bf00246021 regin, j.c., puget, j.f. 1997. a filtering algorithm for global sequencing constraints. in: smolka g. (eds) principles and practice of constraint programming-cp97. lecture notes in computer science, 1330. springer, heidelberg. https://doi.org/10.1007/bfb0017428 solnon, c., cung, v.d., nguyen a., artigues, c. 2008. the car sequencing problem: overview of state-of-the-art methods and industrial casestudy of the roadeef’2005 challenge problem. european journal of operational research, 191, 912-927. https://doi.org/10.1016/j. ejor.2007.04.033 int. j. prod. manag. eng. (2021) 9(2), 113-123creative commons attribution-noncommercial-noderivatives 4.0 international integrating tier-1 module suppliers in car sequencing problem 123 https://doi.org/10.1016/j.procir.2019.03.045 https://doi.org/10.1007/bf00246021 https://doi.org/10.1007/bfb0017428 https://doi.org/10.1016/j.ejor.2007.04.033 https://doi.org/10.1016/j.ejor.2007.04.033 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.15300 received: 2021-03-22 accepted: 2021-05-04 phyron: cognitive computing for the creation of an innovative intelligence experience center ruiz, m.a , rodriguez, j.j.b, erlaiz, g.c, olibares, i.d a isea s.coop. (mondragon corporation), spain. b indaba consultores s.l., spain. b sareteknika servicios globales postventa s. coop., spain. b lanalden s.a., spain. a mruiz@isea.eus, b jjrodriguez@indaba.es, c gerlaiz@sareteknika.com, d iolibares@sareteknika.es abstract: this research presents the results of a project called “phyron: cognitive computing for the creation of an innovative intelligence experience center”, funded by the basque government (economic development, sustainability and environment department). the project started in april 2019 and it will end in december 2021. its main objective was to arrange an industrial research about cognitive computing. the main aim was the application of these systems for the development of an intelligent experience center (iexc) to facilitate: i) enrichment of processes, products and services, in general client experiences, ii) automatic generation of technical predictions related to the product and the client behaviour through the exploitation of acquired knowledge, and iii) rationalization and automation of the processes that are involved in the after sale services both at technical and management level. the technological outcome presented in this paper is built using cognitive engines to enable learning from the client experience, and predictive models to anticipate client necessities. key words: cognitive computing, digitalization, digital transformation, predictive models, algorithms, big data. 1. introduction digitalization is the central pillar of the current fourth industrial revolution that we are living now (hagberg et al., 2016). digitalization is progressively penetrating in sectors such as industry, banking or retail. these sectors are related to “industrial internet or cyber-physical systems”. in the past, added value in industry was based on what traditionally was understood as product and production, however, nowadays added value is related to data in the company. this fact has a direct impact on how competitiveness is understood (only thorough elements such as product design, quality or productive efficiency). moreover, the field based on data is feed by new topics connected to information and communication technologies, such as big data and cloud computing. due to digitalization industry and services are converging. this enables the entry of new agents with no productive capacities in the industry, and definitively the concept of manufacture industry is now being analysed. 2. objective in this paper the results of a project called phyron are presented, in which the principal objective is to cite this article: ruiz, m., rodriguez, j.j., erlaiz, g., olibares, i. (2021). phyron: cognitive computing for the creation of an innovative intelligence experience center. international journal of production management and engineering, 9(2), 103-112. https://doi.org/10.4995/ijpme.2021.15300 int. j. prod. manag. eng. (2021) 9(2), 103-112creative commons attribution-noncommercial-noderivatives 4.0 international 103 https://orcid.org/0000-0001-6900-4382 http://creativecommons.org/licenses/by-nc-nd/4.0/ to arrange an industrial research about cognitive computing. the main aim is the application of these systems for the development of an intelligent experience center (iexc). the principal goal of this development is to facilitate technological improvements aligned with the business processes they support. in this case, technological advances for the improvement and efficacy of the call center service are: i) enrichment of processes, products and services, in general client experiences, ii) automatic generation of technical predictions related to the product and the client behaviour through the exploitation of acquired knowledge, and iii) rationalization and automation of the processes that are involved in the after sale services both at technical level and management level. the principal technological steps for the achievement of these goals are the definition of technological infrastructures for huge amount of data treatment, extraction, capture process and aggregation, and definition of predictive models. technology explained in section 5 (results). in this way, new knowledge about cognitive computing and its application in the industrial field is generated. an innovative product/service that combines infrastructure, software, and necessary tools for the development and implantation of an innovative intelligent experience center (iexc) is the final result of the project. phyron is based on the development of an iexc for after sale services and repairing of home appliances. the consortium behind this project is composed by four companies: sareteknika servicios globales postventa s. coop., as an expert company in the after sale global services in the field of home appliances (spain, portugal, andorra). indaba consultores s.l., as an expert company in the design and implantation of information systems and knowledge management based on computing technologies. lanalden s.a., experts in service through contact centre technology for process improvement, and technological innovation. isea s.coop., r&d institute of mondragon corporation. agent of the basque network for science, technology and innovation. they are experts in the development of technological research projects and launching of new activities. sareteknika and lanalden modelled the solution for making it accessible for service and industrial companies. this new product/service is built with cognitive engines that used predictive models to anticipate customer needs. lanalden was the company in which the validation and pilot test was arranged. finally, isea leads the work packages related to knowledge dissemination and new business model definition. 3. literature review: artificial intelligence (ai) when the term ai appeared in the 1950s, there was access to a limited amount of data, and therefore ai did not evolve in the way and with the speed that was envisioned. nowadays, however, there is a big amount of data which facilitates the development of ai (forrester, 2017). ai technologies need a large amount of information to execute algorithms (engelmore, 1987), and consequently obtain a result which is close to the reality. these algorithms based their complex calculus on data. consequently, data is turned into the key factor for the success of the process. the objective of ai is not to replace the human being; it is to facilitate the analysis of the volume of data necessary for making intelligent decisions (steels, 1993). in order to carry out these skills, a set of mathematical and statistical techniques is required to develop algorithms for the improvement of tasks based on experience and learning (machine learning). therefore, the access to a greater volume of data facilitates the learning of ai through commands to improve decision making. ai is therefore the responsible for providing the necessary intelligence for the extraction of relevant knowledge and information (bringsjord and schimanski, 2003; agrawal et al., 2017). as a previous step to any interpretation of data from ai perspective, it is essential to ensure the high quality of data (bond and gasser, 2014). for this purpose, a set of infrastructures and technologies is required to provide solutions and process huge data sets (structured, unstructured and semi-structured data). data to be processed needs to be based on relational data rules (easily processable) in order int. j. prod. manag. eng. (2021) 9(2), 103-112 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 104 http://creativecommons.org/licenses/by-nc-nd/4.0/ to be considered structured data. unstructured data concept, however, refers to data that has no value until it is ordered and classified (comments on social networks, images, audios, sensor data, etc). 3.1. application of ai in the following years disruptive business models will emerge and this will force companies to understand that digital transformation more than a trend is considered an essential point to remain competitive (pwc, 2017). in the following list the advantages of ai are presented: resource and time optimization through process automation and repetitive tasks. cost reduction for the long period. increase in productivity and efficiency. decision making improvement due to efficacy and the increasing speed. creation of new activity lines or business opportunities. improvement of client satisfaction through the obtaining of different perspectives to predict their preferences and offer a better and personalized experience. application of human abilities through computational services. machine learning (zhang, 2020) and ai technologies are helping marketers to correlate and synthesize variables from different sources, identity patterns of behaviour, and inter consumer interest or purchase intent. this will result into the definition of individualized one-to-one marketing strategies that aim to achieve the highest level of personalization. therefore, companies must define the best doing tasks. in this way, definition of responsibilities, roles and processes for efficiency maximization will be more efficient (strong, 2016). the collaboration between humans and machines is questioned by some experts; this relationship is understood as the transformation of tasks previously developed by humans into tasks developed by robots. experts predict that up to 40% of jobs in the us and up to 30% in the uk can be lost due to ai. in addition, this trend will also lead to creation of new jobs in the fields related to soft skills. the reorganization of existing skills and the large scale hiring of new workers will be also a direct consequence of ai (pwc, 2017). the new scenario resulted from ai technologies will move on from the traditional focus on lowvalue activities towards high value activities. the widespread automation of low-skilled jobs will change the business model of companies; they will be focused on workforce with higher productivity and fewer staff. during this transition, the main pillar will be the development of human capital and the application of new and more agile hiring strategies that reflect the change in the necessary skills paradigm for the competitiveness. moreover, the automation of processes through the use of machines will enable the necessity of human skills, such as intuition, critical thinking, or creativity. however, there are also some barriers related to the implementation of ia technologies: fear of change long periods of time for integration adaptation to the use of advanced technology lack of specialized talent in ai technologies lack of knowledge and maturity in the field of solutions and providers legal problems not standardized impact 3.1.1. machine learning (ml), deep learning (dl) and client experience in a more specialized level under machine learning (zhang, 2020) is deep learning. it facilitates learning processes through computer models acting like neural networks in human brain. its added value is the ability to learn from data through algorithms (goldberg and holland, 1988). in this way, rules of behaviour under algorithms replicate human rules of behaviour. the application of deep learning (rusk, 2016; lecun et al., 2015) enables the interpretation of the reality trough image recognition or natural language analysis. these functionalities are known as cognitive services. these services are focused on the replication of sensitive abilities of human beings, which makes them essential in the field of customer experience. int. j. prod. manag. eng. (2021) 9(2), 103-112creative commons attribution-noncommercial-noderivatives 4.0 international phyron: cognitive computing for the creation of an innovative intelligence experience center 105 http://creativecommons.org/licenses/by-nc-nd/4.0/ the application possibilities of this cognitive technology are very extensive (hollnagel and woods, 1983). the association for the development of customer experience identified a 64% of implementation of customer voice programs (voc) in companies. the objective of these programs is to facilitate the reception of client messages, identifying strengths and weaknesses of the service. the information about the client experience is extracted from the interactions in different channels, such as rrss, emails, recordings of contact center calls, security videos, etc. cognitive services would make possible the increase of the efficiency of these programs for the spread of analysed unstructured data and understanding through language recognition and image recognition algorithms. as a result, it could be stated that nowadays ai market is a trending sector with support of worldwide technological experts. in this way, new customer demand is more efficiently responded. the companies leading this development offer a wide variety of ai solutions (chatbots, virtual assistants, machine learning algorithms…). they also offer the possibility to integrate other products such as crm to increase the power of the solutions in a joint and linked strategy. the anticipation of certain future scenarios can be very enriching for decision-making. predictive models enable the anticipation of events for trust based relationship with clients, to personalize interactions with them and to anticipate to their needs. predictive analytics are frequently used for both private and public sectors. the analysis of customer lifecycle results in different solutions for direct applications. in the case of sale process, additional product recommendation models (cross-selling) or predictive demand models are widely used solutions among international and spanish companies. the differential value of this technique is the use of segmentation models. they facilitate the perception of the existing logic behind two different clients with similar behaviours. 3.1.2. intelligence experience center during the next years, consumers will regularly interact with cognitive computing-based services. cognitive systems and artificial intelligence in general, facilitates the management of huge volumes of data in companies. cognitive systems are changing the way in which people and systems interact (high, 2012). this has a visible impact in the communication process settled down in contact centres. converge and technological advance, such as big data, cloud computing or the internet of things (iot) are the levers for the change in this context, and they are integrated in the system to respond to the new customer needs. they are driving growth and innovation in ai. these technologies will enable management of both information and strategic decision making through ai, to elevate customer service to a more personalized level (pwc, 2018). the new emerging customer is a more demanding and informed consumer who places new demands on companies and expects a greater degree of customization from companies. different studies prove that the biggest challenge for spanish companies is the personalization of products and services to respond to client real needs. 4. methods 4.1. general process big data projects generally require 2 processes: i) design of the solution, and ii) implementation of the solution (wu et al., 2013; mousannif et al., 2014; katal et al., 2013; jin et al., 2015). these processes are divided in 5 phases: phase 01 data sources. this first phase requires the identification of data sources, type of source (relationship database, spreadsheets), file format (.txt, .doc, .xls…), or data character (data structure). in a big data environment, it will be important to determine all the data sources that are required for the execution of the project. phase 02 data ingestion. after the identification of data sources and their presentation, the next step is to select the techniques for extracting data, it can be extracted from a traditional etl (extract, transform and load or “extract, transform and load” process). this technique could be applied on traditional structured data, crawling techniques to extract data from web pages, design interoperability services, application of different apis or application programming interface, subroutines, functions or specific procedures. phase 03 data storing. in this phase, the required data storage solution is designed and implemented int. j. prod. manag. eng. (2021) 9(2), 103-112 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 106 http://creativecommons.org/licenses/by-nc-nd/4.0/ (relational database, datawarehouse or a non-sql database, sized to the project requirements, in terms of volume, etc.). phase 04 data processing. in this phase, the data processing or intelligence application techniques required in the project will be designed and implemented. initially, data mining techniques will be applied. data processing is arranged using tools such as weka, and observing the results obtained. after the study, the algorithms that offer the best results can be implemented. the need to apply natural language processing (nlp, natural language processing) and semantic web techniques was also analyzed. phase 05 data analysis. finally, in this phase, graphical and usable interfaces were developed to show the outcome of the project to the users. 4.2. methodology considering this general and standard methodology, phyron is developed during 33 months (from 2019 to 2021), and the main tasks of the project are three: table 1. principal tasks for the execution of the methodlogy (own source). task name duration 01 technological research, case of uses definition, and project requirements 2019 02 predictive models design and development 2020 03 solution implantation, validation and result exploitation 2021 in the following figure 1 the detail of the methodology is presented and in the next paragraphs each phase will be explained: 1. technical requirement definition: 1.1 generation of new knowledge in computer science 1.2 definition of the cases of use 1.3 collection of security and privacy requirements 1.4 definition and collection of security and privacy requirements 1.5 collection and definition of the requirements for the design of the infrastructure 1.6 collection and definition of requirements for the predictive models this task was focused on the generation of new knowledge in computer science. the main objective was to have a general perspective about machine learning techniques and their application to industrial environments. in addition, the selection and description of the cases of use was done. this is the pillar for the requirements of data sources, security, infrastructure, and predictive models. security refers to the anonymization of data. in order to define the infrastructure is necessary to describe hardware and software sets. extraction, storage and process requirements were defined. regarding predictive models, the most important objective was to facilitate decision making through the identification of entry parameters. moreover, the requirements for graphical and attractive interfaces were defined. 2. data treatment: 2.1 standardization and homogenization 2.2 data sources exploitation and analysis 2.3 characteristics extraction 2.4 design of the process for extraction, storage and aggregation 2.5 implementation of the process 2.6 validation of the process the first step is related to the quality control applied to data, data source standardization was compulsory. data source analysis was done based on the definition of the cases of use. in this way, available variables were selected for their use in the predictive models. this analysis showed that more variables than expected were needed. cleaning and processing treatments were applied to variables to facilitate their use. finally, the process for extraction, capture and processing was developed through the use of the engine for anonymization. for the validation, selected variables were exported step by step. once the validation was done all the variables were exported. figure 1. principal phases. int. j. prod. manag. eng. (2021) 9(2), 103-112creative commons attribution-noncommercial-noderivatives 4.0 international phyron: cognitive computing for the creation of an innovative intelligence experience center 107 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3. data privacy: 3.1 design of the process for data anonymization 3.2 implementation of anonymization process 3.3 implementation of infrastructure security data managed in this process belongs to particular users (individuals). consequently, it was necessary to implement the anonymization process to guarantee its safe use in different units of storage. anonymization engine made the variables invisible acting on its identity attribute. security of the infrastructure was implemented in parallel with validation. 4. infrastructure development: 4.1 installation of the software 4.2 validation of the infrastructure installation was dependant on the requirements defined in the first stage, and it was composed of different software sets to satisfy different functions such as extraction and data capture, storage, resource management, and predictive model generation through machine learning algorithms. apart from the installation, configuration of each of the software sets facilitated validation of the infrastructure. 5. predictive models and display interface: 5.1 predictive models design 5.2 design of the display interface 5.3 development and implementation of predictive models 5.4 implementation of the display interface 5.5 adaptation of predictive models for their execution in real time 5.6 validation of the predictive models and display interface firstly, existing correlations between data were identified to simplify the complexity of the data set, as well as to identify how these variables could be enriched. when data was analysed, inputs and outputs for the final result were classified. the final result was defined according to the requirements in the first stage. in that sense, when defining predictive models it was essential on the one hand the scalability of the variables, and on the other hand, model parallelization. these characteristics enabled the future integration of new parameters and future adaptation of variables to new necessities. for the display interface design user centered design philosophy was followed and mockups were used. moreover, information architecture and interaction were executed to optimize user experience. predictive models were validated independently. secondly, display interfaces were validated. finally, usability proofs were arranged to identify the possible mistakes related to design and development. 6. implementation of iexc phyron and experience modelling: 6.1 handbook and protocols for taking action 6.2 pilot test 6.3 identification of failures this phase was based on the definition of contents to facilitate the use of developed tools. a training program was defined for workers from lanalden. a real case study was arranged to validate the tool, for a high quality failure identification phase. 5. results the main objective of this project was to analyse new profiles and necessities of final users of sareteknika. the basis to extract this information was the available information in sareteknika about usual casuistry and actors involved in their after-sale service processes. this process was necessary to guarantee the total functionality of the system to solve real cases in after sale service. the main results could be defined as following: i) specifications of the architecture and extraction system, ii) anonymization process, and iii) infrastructure construction, and predictive models. a. specifications architecture and extraction system during the first year, the extraction and processing system was defined and implemented. a selection was initially made for all the information collected, and considering relevant data aligned with the established treatment objectives. the quality of the hidden knowledge to be discovered depended on both machine learning algorithm or the technique used and the quality of data to be treated. it is necessary to have a coherent raw material reception process to obtain consistent conclusions. int. j. prod. manag. eng. (2021) 9(2), 103-112 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 108 http://creativecommons.org/licenses/by-nc-nd/4.0/ then, transformation and cleaning of data for the modelling tools was developed. techniques to ensure the quality of data were used: cleaning: this phase was focused on correcting data anomalies that can result in inaccurate conclusions. these anomalies can result from missing data or from values that do not fulfil expected behaviour. it is important to identify the origin of these problems and their distribution before applying different corrective measures (statistical techniques to solve them). transformation: when the predictive power of the attributes is not very clear, it is necessary to build new attributes by applying some transformation to the original attributes (grouping, separate dates to integers, converting categorical values into numbers). dimensionality reduction: select the appropriate set of attributes for the specific task to be performed. it is carried out based on statistical techniques combined with empirical methods. it will facilitate optimization of analytical models. in the following figure 2 the principal entities related to the notifications of failure could be seen: figure 2. principal entities related to the notifications of failure (own source). b. anonymisation process initially, existing identifiers in automatic identification systems (ais) were eliminated from the phyron data source system. it was the first step to avoid individual identification through phyron data treatment. due to the objective pursued in phyron, most of the sensitive attributes about clients and existing in the original data source are not necessary: bank, iban, bank card, contract amounts and repairs. only two attributes have been maintained: “charged”, and “payment_date”. regarding the quasi-identifiers that have been used (postal code, town, province), k-anomymization techniques have been applied to achieve an acceptable level of anonymization considering affected users. c. infrastructure development phyron infrastructure is composed of four main modules: i) technological infrastructures for huge amount of data treatment, ii) extraction, capture, process and aggregation, iii) advanced analytics and predictive models, and iv) methodology and protocol for taking action. component 1: technological infrastructures for huge amount of data treatment the technological infrastructure functioned as the basis to collect and process data to create predictive models for the personalized prediction of each client situation. the adequate dimensioning of the technological infrastructure is a key factor to arrange the necessary data processing. an oversizing would increase the associated costs due to both the material cost and its management. the oversizing has no improvement in the processing time for predictions. failure to meet storage and processing demand will decelerate the project development. the software needed to manage and process data and generate models to be run in real time. it is a key factor for the sizing of the technological infrastructure. figure 3. software and data flow in phyron (own source). phyron technological infrastructure is composed by software sets which belong to four different families: i) data sources, ii) data feed, iii) data int. j. prod. manag. eng. (2021) 9(2), 103-112creative commons attribution-noncommercial-noderivatives 4.0 international phyron: cognitive computing for the creation of an innovative intelligence experience center 109 http://creativecommons.org/licenses/by-nc-nd/4.0/ collection and resource management, and iv) data analysis and model generation. the implemented technological infrastructure will collect data for the predictive models (without prior treatment). data flow shown in the following figure 3 will use spark streaming to process data in real time. figure 4. software and data flow for streaming process (own source). one of the fundamental aspects to implement a technological infrastructure is its ability to scale. the storage and processing needs will change depending on both the case of use and the amount of data generated. the original design and implementation is done for previously treated data set and processing method. in addition, depending on possible future needs, it is desirable that the software offers integration with other sets of software packages for additional functions. in this sense, open source software offers the necessary independence for flexibility when integrating other software packages or developing connectors to facilitate integration. component 2: extraction, capture, process and aggregation phyron is based on data extraction technologies, both structured and unstructured data. the main challenge is this sense was to investigate an adjusted solution for each set of structured data. this concept of “data” refers to available and previously stored data in relational storage systems. data generated in real time. moreover, different unstructured data is necessary as input for the predictive models. the solutions proposed for each type of unstructured data will contemplate the origin, frequency and, possible anomalies in the system. this results in the necessity to use different techniques depending on the data. a clear example of the pre-processing of unstructured data is the data collected through free text fields specified by the technicians. these fields facilitate the freely definition of the causes for failure or service provided, resulting in a text that is easily understood by human being, but unintelligible for information systems. it could not be used without prior preprocessing and extraction of characteristics. apart from the ml algorithms selected, the quality of the hidden knowledge is also influenced by the quality of the data treated. when inconsistences are found in the flow, the conclusions extracted from the system are inconsistent. component 3: advanced analytics and predictive models phyron aimed to develop different techniques that enable the development of predictive models that can be used “online”, although their generation will be “offline”. the novelty arranges from the holistic approach of integrating the design and development of the techniques. finally, developed tools were validated with real use cases. during the last year of the project, the design of the predictive models will be developed, in parallel to the visualization interfaces creation. for the design phase it is necessary to contemplate both the result of the project, and the way to present the result. component 4: methodology and protocol for taking action the fourth component of the iexc phyron was based on the elaboration of methodologies and documentation to collect the principal guidelines to facilitate an efficient customer experience, and optimizing the use of phyron. in the same way, protocols for taking action in a personalized way and services associated to each of the contact phases with the client were defined. training plans for the last year were designed. d. predictive models development this process was developed in parallel with the design of the display of the results. it is necessary int. j. prod. manag. eng. (2021) 9(2), 103-112 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 110 http://creativecommons.org/licenses/by-nc-nd/4.0/ an alignment between “what to obtain” and “how to present it”. in this way, added value in the process is increased. this process is developed in an iterative form, an initial design is implemented and its outcomes are used for the redesign and optimization of the solution. during this phase initial requirements were defined and it was especially relevant to follow them to guarantee an optimum integration degree (steyerberg et al., 2001; kuhn, 2008; biecek, 2018). variables that could be considered predictors were identified in a personalised way for each of the cases of planned use. this analysis was based on: i) familiarization with data, ii) identification of data quality, iii) identification of initial ideas, and iv) detection of data sets to form hypotheses about hidden information. after this process, the final result was the dataset to begin with the predictive model. once the infrastructure was assembled and data treatment process was designed and implemented, the predictive models were designed both for their execution based on historical data and their execution in real time. firstly design and development of models oriented to the existing possibility of repairing and the identification of the necessary materials for the repairing were done. conclusions from the subsequent model and its tests may lead to the convenience of reformulating its design, modifying the pre-processing of certain variables, or even incorporating new variables discarded in a first moment. results obtained from the models bring an improvement in the classification with respect to the distribution of variables. 6. conclusions the final result of this project is a product and a new service related to cognitive computing: i) cognitive platform based on the application of machine learning algorithms for the development of predictive models for consumer behaviour analysis, and ii) new service based on cognitive services related to intelligent experience centre. regarding process optimization it is necessary to mention that time and resources will be extremely improved due to phyron solution. costs in the long period, productivity and operative efficiency, decision making processes, new activity lines, client satisfaction, and application of human abilities through automated computational services will be improved. the result of the project is the design of a cognitive system for an intelligent experience center. this platform will be offered to the service sector using a commercial brand to convey this technological solution. acknowledgements we would like to thank the basque government for their support in the development of this project. special thanks to the economic development, sustainability and environment department. references agrawal, a., gans, j., goldfarb, a. (2017). what to expect from artificial intelligence. mit sloan management review. biecek, p. (2018). dalex: explainers for complex predictive models in r. the journal of machine learning research, 19(1), 3245-3249. bond, a.h., gasser, l. (eds.). (2014). readings in distributed artificial intelligence. morgan kaufmann. bringsjord, s., schimanski, b. (2003). what is artificial intelligence? psychometric ai as an answer. in ijcai, pp. 887-893. engelmore, r.s. (1987). artificial intelligence and knowledge based systems: origins, methods and opportunities for nde. in review of progress in quantitative nondestructive evaluation, pp. 1-20. springer, boston, ma. https://doi.org/10.1007/978-1-4613-1893-4_1 goldberg, d.e., holland, j.h. (1988). genetic algorithms and machine learning. machine learning 3, 95-99. https://doi.org/10.1023/a:1022602019183 hagberg, j., sundstrom, m., egels-zandén, n. (2016). the digitalization of retailing: an exploratory framework. international journal of retail & distribution management. 44(7), 694-712. https://doi.org/10.1108/ijrdm-09-2015-0140 high, r. (2012). the era of cognitive systems: an inside look at ibm watson and how it works. ibm corporation, redbooks, 1-16. int. j. prod. manag. eng. (2021) 9(2), 103-112creative commons attribution-noncommercial-noderivatives 4.0 international phyron: cognitive computing for the creation of an innovative intelligence experience center 111 https://doi.org/10.1007/978-1-4613-1893-4_1 https://doi.org/10.1023/a:1022602019183 https://doi.org/10.1108/ijrdm-09-2015-0140 http://creativecommons.org/licenses/by-nc-nd/4.0/ hollnagel, e., woods, d.d. (1983). cognitive systems engineering: new wine in new bottles. international journal of man-machine studies, 18(6), 583-600. https://doi.org/10.1016/s0020-7373(83)80034-0 jin, x., wah, b.w., cheng, x., wang, y. (2015). significance and challenges of big data research. big data research, 2(2), 59-64. https://doi.org/10.1016/j.bdr.2015.01.006 katal, a., wazid, m., goudar, r.h. (2013). big data: issues, challenges, tools and good practices. in 2013 sixth international conference on contemporary computing (ic3), 404-409. ieee. https://doi.org/10.1109/ic3.2013.6612229 kuhn, m. (2008). building predictive models in r using the caret package. journal of statistical software, 28(5), 1-26. https://doi.org/10.18637/jss.v028.i05 forrester (2017). “predictions 2018: the honeymoon for ai is over”. lecun, y., bengio, y., hinton, g. (2015). deep learning. nature, 521(7553), 436-444. https://doi.org/10.1038/nature14539 mousannif, h., sabah, h., douiji, y., sayad, y.o. (2014). from big data to big projects: a step-by-step roadmap. in 2014 international conference on future internet of things and cloud, 373-378. ieee. https://doi.org/10.1109/ficloud.2014.66 pwc. (2017). uk economic outlook. recuperado de http:// www.pwc.co.uk/economics pwc. (2018). “bots, machine learning, servicios cognitivos. realidad y perspectivas de la inteligencia artificial en españa, 2018”. rusk, n. (2016). deep learning. nature methods, 13(35), 35-35. https://doi.org/10.1038/nmeth.3707 steels, l. (1993). the artificial life roots of artificial intelligence. artificial life, 1(1,2), 75-110. https://doi.org/10.1162/artl.1993.1.1_2.75 steyerberg, e.w., harrell jr, f.e., borsboom, g.j., eijkemans, m.j.c., vergouwe, y., habbema, j.d.f. (2001). internal validation of predictive models: efficiency of some procedures for logistic regression analysis. journal of clinical epidemiology, 54(8), 774-781. https://doi.org/10.1016/s0895-4356(01)00341-9 strong, a.i. (2016). applications of artificial intelligence & associated technologies. science [etebms-2016], 5(6). wu, x., zhu, x., wu, g.q., ding, w. (2013). data mining with big data. ieee transactions on knowledge and data engineering, 26(1), 97-107. https://doi.org/10.1109/tkde.2013.109 zhang, x.d. (2020). machine learning. in a matrix algebra approach to artificial intelligence, 223-440. springer, singapore. https://doi.org/10.1007/978-981-15-2770-8_6 int. j. prod. manag. eng. (2021) 9(2), 103-112 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz et al. 112 https://doi.org/10.1016/s0020-7373(83)80034-0 https://doi.org/10.1016/j.bdr.2015.01.006 https://doi.org/10.1109/ic3.2013.6612229 https://doi.org/10.18637/jss.v028.i05 https://doi.org/10.1038/nature14539 https://doi.org/10.1109/ficloud.2014.66 http:// www.pwc.co.uk/economics https://doi.org/10.1038/nmeth.3707 https://doi.org/10.1162/artl.1993.1.1_2.75 https://doi.org/10.1016/s0895-4356(01)00341-9 https://doi.org/10.1109/tkde.2013.109 https://doi.org/10.1007/978-981-15-2770-8_6 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.16130 received: 2021-08-27 accepted: 2021-10-02 cyber-physical production system assessment within the manufacturing industries in the amazon moisés andrade coelho a1*, franciel andrade de oliveira a2, lindara hage dessimoni a3, & nicole sales libório a4 a hard sciences and technology institute (icet), federal university of amazonas (ufam), nossa senhora do rosário, 3863, itacoatiara/amazonas/brazil. a1 moises.acoelho@gmail.com, a2 frankdade14@gmail.com, a3 lindarahage@hotmail.com, a4 nicoleufam10@gmail.com abstract: cyber-physical production systems (cpps) represent a relevant aspect related to industry 4.0 and the advances promoted by the digitization and use of artificial intelligence in the production environment in the search for the development of smart factories. this study aims to assess the maturity level of cyber-physical production system (cpps) within manufacturing industries in the amazon. the research uses a quali-quantitative approach to analyze the problem by conducting exploratory case studies (indepth case) and the research framework used aimed to evaluate and measure the cpps within three manufacturing industries in the amazon (n = 3) to measure their maturity. findings reveal a positive relationship between the type of production system adopted by the company, the level of automation, and the maturity of the cpps. the proposed methodology can assist other companies in the development of the technological strategy, supporting the digital transformation process in order to obtain competitive advantage. the study contributes by addressing the topic of cyber-physical production systems from the point of view of operations management and strategy. key words: industry 4.0, cyber-physical production system, operations management, strategic process, amazon. 1. introduction industry 4.0 increases the digitization of manufacturing with the cyber-physical system (cps), in which connected networks of humans and robots interact and work together with shared and analyzed information, supported by big data and cloud computing along value chains whole industrial plants (yang, 2017). cps bring more functionality, autonomy, multi-area integration, application in all sectors, such as industry, medical systems, service/ user relationship, among others (baheti & gill, 2011) being characterized by robustness, security, and protection (sha et al., 2008) and where selforganized manufacturing, context-/situation-aware control and symbiotic human-robot collaboration can play an important role in transforming current factories into factories of the future with greater stability and security (wang et al, 2015). the application of cps in the production and manufacturing environment gave rise to the term cyber-physical production systems (cppss) which have great potential to make production systems intelligent, resilient and self-adaptive (wu et al., 2019a). in cppss systems, the model generated for cyberspace forecasting incorporates data from sensor networks for each critical asset to reflect changes. they are online networks of social machines to cite this article: coelho, moises andrade; de oliveira, franciel andade; dessimoni, lindara hage; libório, nicole sales. (2022). cyber-physical production system assessment within the manufacturing industries in the amazon. international journal of production management and engineering, 10(1), 51-64. https://doi.org/10.4995/ijpme.2022.16130 int. j. prod. manag. eng. (2022) 10(1), 51-64creative commons attribution-noncommercial-noderivatives 4.0 international 51 https://orcid.org/0000-0001-6102-766x https://orcid.org/0000-0002-9710-7415 https://orcid.org/0000-0002-4891-9855 https://orcid.org/0000-0003-2595-4049 mailto:frankdade14@gmail.com mailto:lindarahage@hotmail.com mailto:nicoleufam10@gmail.com http://creativecommons.org/licenses/by-nc-nd/4.0/ that interconnecting to information technology with mechanical and electronic components that communicate using rfid technology (radio frequency identification), for example (lee et al., 2017; delloite, 2015). since 2011, models have emerged to assess industry 4.0, such as the acatech maturity model (schuh et al., 2020), impuls – vdma (lichtblau et al., 2015), uni-warnick (agca et al., 2017) and pricewaterhousecoopers – pwc (geisbauer et al., 2016). recently studies have investigated cppss in the most different contexts, such as digital twin (zhang et al., 2019; zang et al., 2020; liu et al., 2020; ait-alla et al., 2019), internet of things (berlak et al., 2020; stock et al., 2020), cloud computing (mourtizis & vlachou, 2016; mourad et al., 2020; borangiu et al., 2020), voice assistant (afanasev et al., 2019) and shop floor (romero-silva & hernández-lópez, 2020; govender et al., 2019; torres et al., 2019; rocha et al., 2019). there are researchs/reports that assess industry 4.0 readiness, however the characteristics that make up cyber-physical production systems from the point of view of operations management remain unexplored. therefore, the following research questions were developed for investigations: rq1: does the company implement practices associated with cyber-physical production systems? rq2: does the research framework provide adequate assessment of cpps practices? rq3: what is the level of implementation observed in the company concerning the literature? in this context, this study aims to assess the maturity level of cyber-physical production systems (cpps) within manufacturing industries in the amazon. the second section deals with the literature background on the concepts and applications of cpps existing over the last few years in the literature. section three discusses the study’s methodology and its research framework; section four deals with the presentation of the case study companies; section five presents the results and discussion carried out throughout the study. finally, the last section presents the conclusion with the final position in the study. 2. literature background 2.1. the concept of cppss the implementation of industry 4.0 in manufacturing environments occurs with the development and complete industrial implementation of cyberphysical production – cps (francalanza et al., 2017). cpss represent an emerging area of research that has attracted the attention of many researchers due to expectations that they will play an important role in the design and development of future engineering systems (sanislav & miclea, 2012), being introduced by helen gill of national science foundation in the united states in 2006 (gill, 2008). cps can be considered an extremely important step in the development of future manufacturing (monostori et al., 2016) to make manufacturing more competitive by integrating advanced computing and cps to adapt and take advantage of the current big data environment (lee et al., 2013). over the last few years the concept of cps has expanded as a result of the advancement and complexity of instruments that are related to industry 4.0 improving for cyber-physical production system (cpps). cppss consist of autonomous and cooperative elements, as well as subsystems; these elements are connected and can communicate independent situations, at all levels of production, from the field device and process level to the factory and production planning levels (francalanza et al., 2018), that is, denotes a mechatronic system (physical world) coupled with software and digital information entities (cyber part), both allowing the concept of the intelligent factory for the industry 4.0 paradigm (cruz salazar et al., 2019). cppss offer new possibilities for production planning and control (pps) due to their characteristics of decentralized organization, real-time capacity, and intelligent data processing (berger et al., 2019), integrating physical and computational resources due to sensors and power increasingly available processing tools. this allows the use of data, to create additional value, such as monitoring or optimization of conditions (bunte et al., 2019) and involves multidisciplinary engineering activities that depend on the exchange of effective and efficient knowledge int. j. prod. manag. eng. (2022) 10(1), 51-64 creative commons attribution-noncommercial-noderivatives 4.0 international coelho et al. 52 http://creativecommons.org/licenses/by-nc-nd/4.0/ for better collaboration between engineers from different disciplines (meixner et al., 2019). lee et al. (2018) present a cpps architectural framework, where the core elements (sub-system) are (1) big data analytics, (2) detection and coordination, and (3) kpi simulation. among the components are big data storage, quality prediction model builder, model repository, real-time data listener, quality and productivity detector, coordinator, cyber model builder, simulation engine, and reference kpi builder. in conclusion, the manufacturing processes at cpps allow for more self-regulation and self-organization in semi-autonomous production teams, create greater complexity and dynamics in the work process, require more collaboration, communication, and problem-solving skills from operators, enable new forms of human-machine-human collaboration, and can significantly increase the productivity and quality of manufactured products (mühlfelder, 2019). 2.2. cppss applications numerous authors have expressed interest in the applications of cppss in the most diverse contexts. four categories of studies related to cpps stand out. the first related to industrial application and manufacturing, highlighting the studies of implementation of cpps in a real factory to predict the quality of metal casting and control of the operation from the design of a cpps architecture framework (lee et al., 2018), or even, implementation of the cyber-physical production system in intelligent manufacture estimating how companies want to implement disruptive technologies (drennanstevenson, 2019), 5c architecture as a cyberphysical development guide for industrial application (lee et al., 2017), chawla et al. (2020) presented a generalized synergic framework between different production facilities and in different geographical locations to obtain an efficient and energy-saving cpps for the production of different types of jobs in the context of industry 4.0, implementation of cpps using low-cost devices in a simulated factory to control the industrial process and integrate shop floor communications using the amqp advanced message queuing protocol (llamuca et al., 2019), ferreira et al. (2020) contributed data on the applicability of cps components in the current smes (small and medium enterprises) environment, with the intention of improving the performance of manufacturing processes (within the concept of cpps), lins & oliveira (2020) presented the standardization of the retrofitting process to transform old equipment into a cpps, and pinzone et al. (2020) dealt with the implications of operative and sustainability functionalities on the health, learning and operational performance of human workers within the concept of cpps. the second category presents studies involving new cpps application models, highlighting the studies of a new event-based approach (berger et al., 2019), such as the 3s-oriented design concept for resilient and integrated cyber-physical systems based on three standards: stability, security, and systematicity (hu et al., 2016), design patterns of multi-agent systems (mas), indicating that agent-based patterns greatly benefit the design of cpps (cruz salazar et al., 2018), a new event-based approach (berger et al., 2019), prist et al. (2019) who proposed an intermediate layer in the architecture that allows each device, production line and machine to be connected independently, despite the protocol adopted, bunte et al. (2019) analyzed the existing reference architecture on their cognitive skills (related to cpps), architectures that define the structure and interaction of components software development in cpps (mayrhofer et al., 2019), a software module for a monitoring system focused on the evolution of the icps industrial cyber-physical system (iglesias sagardui & arellano, 2019), development of a computer-based model to simulate physical aspects of the material flow using a physical mechanism in cpps (glatt & aurich, 2019), eckhart et al. (2019) proposed a methodology called security development lifecycle for cyber-physical production system (sdl-cpps) that aims to promote security designed for cpps, stock et al. (2019), approach to create a cyber-physical data access layer, based on in the self-description capability of cpps components, wu et al. (2019b) proposed a meta model to formalize the integrative link between cpps and enterprise information systems (eis), patalas-maliszewska & schlueter (2019) explored the possibility of integrating the knowledge management system (kms) and system integrator (s) in cyberphysical production systems (cpps), and simulation of various degrees of autonomy in a cpps using a hybrid lab approach (gronau, 2019). the third category involves digital twin studies and their application to smart shop-floor at scale via digital twin (dt), highlighting the opportunities to use dt for cpps to support work scheduling during operation (zhang et al., 2019), the concept of intelligent digital twin that can be used to perform autonomous cpps (ashtari talkhestani et al., 2019), int. j. prod. manag. eng. (2022) 10(1), 51-64creative commons attribution-noncommercial-noderivatives 4.0 international cyber-physical production system assessment within the manufacturing industries in the amazon 53 http://creativecommons.org/licenses/by-nc-nd/4.0/ cyber-physical system (cps) and digital twin to build the interconnection and interoperability of a physical shop floor and correspondent cybershop floor (ding et al., 2019), transforming traditional manufacturing into a cpps through the e-core approach (loucopoulos et al., 2019), and park et al. (2020) proposed a digital twin based cpps architectural framework that overcomes performance obstacles. in conclusion, the fourth category involves support for the decision-making process. highlights are the studies of alves & putnik (2019) presented research on the influence of the duration of decision making at cpps on the performance of the manufacturing system, for different programming paradigms, an innovation system for effective planning of production and maintenance integrated into the cpps complex using multi-criteria decision-making (schreiber et al., 2019). 3. methodology a mixed method was adopted to analyze the problem. this approach enables a better understanding of the research problems that each of the approaches (quantitative and qualitative) would allow separately. the combined methods make it possible to expand the understanding of research problems (miguel, 2012; creswell, 2009), as occurred in this study. the purpose of this research was to conduct exploratory case studies (mccutcheon & meredith, 1993) in which the conclusions obtained from the analysis of the data will be based on empirical evidence. the choice of single cases is justified by the need for greater depth in the research framework (voss et al., 2002). the case study is a methodological procedure in which it examines a phenomenon as a whole, using multiple data collection methods to collect information from one or a few entities, such as people, groups, or organizations. it examines contemporary events where the behavior of the research subjects cannot be manipulated, having a generalized character to the theoretical prepositions. it can also be used to analyze longitudinal change processes. thus, it aims to expand and generalize theories and not populations and universes (benbasat et al., 1987; eisenhardt, 1989; yin, 1994). the research framework used in the study aimed to evaluate and measure the cyber-physical production system (cpps) within three manufacturing industries in the amazon to identify the maturity level of these organizations. the constructs were determined from pricewaterhousecoopers (2014), european parliament (2016), thiede (2018), thiede et al. (2016), germany trade & invest (2014), delloite (2015), lee et al. (2017) and lee et al. (2018). this research framework considers 15 constructs, consisting of a form containing 42 questions to provide an adequate assessment of the evaluated productive system. the constructs considered are (1) organization of the machines in a network; (2) integration of machines and the production process; (3) sensors and control elements; (4) data exchange and control in real-time; (5) kpi simulation, (6) dashboard, (7) data treatment and storage, (8) system interoperability; (9) level of autonomy; (10) vertical integration; (11) connection; (12) conversion; (13) cyber; (14) process cognition and optimization; and (15) configuration artificial intelligence and machine learning. table 1 presents the relation between constructs and source. table 1. relation between constructs and source. constructs source organization of the machines in a network delloite (2015)integration of machines and the production process sensors and control elements data exchange and control in real-time pwc (2014) kpi simulation lee et al. (2018) dashborad data treatment and storage thiede (2018) & thiede et al. (2016) system interoperability ep (2016) level of autonomy vertical integration gtai (2014) connection lee et al. (2017) conversion cyber process cognition and optimization configuration artificial intelligence and machine learning source: authors. each question received a score according to the evidence that was presented by the company int. j. prod. manag. eng. (2022) 10(1), 51-64 creative commons attribution-noncommercial-noderivatives 4.0 international coelho et al. 54 http://creativecommons.org/licenses/by-nc-nd/4.0/ and observed in loco on a likert scale (1 to 5) to adequately measure the evaluated production system. the scores are: 1 practice not even thought out, rarely occurs, does not apply to the reality of the company; 2 some awareness, but disorderly and occasional responses, informal systems, basic level of implementation of cpps; 3 consciousness and appropriate formal systems but could be further improved, intermediate level of implementation of cpps; 4 effective and highly developed systems, advanced level of cpps implementation, including provisions for improvement and development; 5 highly effective and developed systems, highly advanced level of cpps implementation, including self-configuration and organization of systems. the delimitation of the universe of this research was three manufacturing industries located in the amazon. a form was used to identify organizational characteristics, based on guérin et al. (2011) and aspects related to cppss, based on pricewaterhousecoopers (2014), european parliament (2016), thiede (2018), thiede et al. (2016), germany trade & invest (2014), delloite (2015), lee et al. (2017) and lee et al. (2018). as research techniques (marconi & lakatos, 2002) were used: (1) indirect documentation (documentary and bibliographic research); (2) intensive direct observation (in-loco observation and open structured interview); and (3) extensive direct observation (application of form). the study took place in four moments: (1) open structured interview (vergara, 2009) based on a script, based on guérin et al. (2001), where fundamental information was identified to complement the forms applied in the company, later; (2) documentary research took place intending to collect preliminary data in written documents (reports, internal reports, and website) and structured observation (vergara, 2009); there was (3) application of the form with those responsible for the organization (after observation and interviews); and concluding, (4) quantitative and qualitative data were analyzed and tabulated. the quantitative data obtained from the script responses were tabulated in a summary table, grouped according to the content, and stratified according to the structure of the evaluation and measurement form of cyber-physical production systems. for qualitative data, discourse analysis (bardin, 1977) was used based on the following steps: (1) pre-analysis (systematization and establishment of interpretation indicators), (2) data exploration (coding, classification, and categorization), and (3) treatment of results, inference, and interpretation. a summary of the methodological procedures used is presented in table 2. table 2. summary of the methodological procedures. stage method source comments approach to the problem mixed methods miguel (2012), creswell (2009) interpretation of the opinion of the interviewees use of productive quantitative data type of research exploratory case study (single case) mccutcheon & meredith (1993), voss et al. (2002) chemical company thermal power plant plastic injection molding company located in the amazon procedure indirect documentation intensive direct observation extensive direct observation marconi & lakatos (2002), vergara (2009), guérin et al. (2001) reports, internal reports, and website data gathering multiple case study in loco observation = 15 days open structured interview = 20 hours of interview form application = 05 days marconi & lakatos (2002), vergara (2009), guérin et al. (2001) quality department operations management department industrial engineering department observation in the productive and administrative area interviews with company managers and employees application form with company managers and employees analysis of data analysis of content bardin (1977) description, understanding, and explanation of research framework (evaluation and measurement of a cyberphysical production system (cpps) from the following steps: (1) pre-analysis, (2) data exploration, and (3) treatment of results, inference, and interpretation. source: authors. int. j. prod. manag. eng. (2022) 10(1), 51-64creative commons attribution-noncommercial-noderivatives 4.0 international cyber-physical production system assessment within the manufacturing industries in the amazon 55 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. companies 4.1. case 1. chemical company the first case represents a company in the chemical sector, being one of the best known and respected manufacturers of hygiene and cleaning products in the world, operating in the categories of hygiene and cleaning, personal care, insecticides, and more recently, domestic storage. currently, the plant located at the industrial pole of manaus (pim) has approximately 320 employees. 4.2. case 2. thermal power plant the second case represents a thermal power plant, operating since 2002. the company has an installed capacity of 9mw, with a continuous production process, and with a high level of automation. currently, the plant located in amazon has approximately 60 employees. 4.3. case 3. plastic injection molding company the third case represents a plastic injection molding company, being an important national player in the segment of plastics for the electronics, home appliance, automobile, and electrical products industries. currently, the plant located at the pim has approximately 400 employees. 5. results and discussion rq1: does the company implement practices associated with cyber-physical production systems? case 1. chemical company (table 3). table 3. case 1 – chemical company. constructs evidences organization of the machines in a network the company’s machines are not organized in networks and also do not use rfid technology or other technology. the machines do not share information between them, such as stock levels, problems or failures, changes in orders, or levels of demand. also, remote access to them is not possible. integration of machines and the production process there are no sensors or control elements in the company that allow the machines/equipment to be connected to the manufacturing plant, fleets, work networks, and human beings. sensors and control elements over the years the company has developed a series of improvements in machines and equipment as a result of the adoption of lean thinking, which included the implementation of tpm. the machines have several productive control sensors or fail-safe devices (poka-yoke), however, they do not communicate with each other. data exchange and control in real-time information is collected by operators directly from each machine at each workstation through-out the production process. then these data are inserted in spreadsheets and also in the erp system (enterprise resource planning). there is no real-time control of the production process, however, the data collected from each workstation, manually, is used to monitor pro-duction. kpi simulation company has kpis in its production process, however they are not simulation results. dashboard the information generated by the production process is located in visible places on the plant data treatment and storage there is no standardized communication proto-col between machines/equipment and data sys-tems. the information is not categorized and there is also no responsibility for analyzing the data generated by the systems. system interoperability there is no system interoperability, does not exist the connection and communication between human and smart devices available (real and virtual). level of autonomy the level of automation adopted by the company is low considering industry 4.0. just a production process has a high level of automation. vertical integration vertical integration (suppliers, company, and consumers) takes place from the network of machines/equipment using erp. connection there is no standardized communication protocol between machines/equipment and data systems. however, the company presents an initial stage of connection between activities, erp and machines in its production process conversion there is no big data storage and other methods for conversion meaningful information cyber there is no application and generation of predictive/cyber models. process cognition and optimization there is no application of decision-making and reasoning methods to recommend operations aimed at maintaining optimal production. configuration artificial intelligence and machine learning the system does not provide features that are self-configuring or involve machine learning and artificial intelligence. source: authors. (table 3 continues in the next column) (table 3 continues from the previous column) int. j. prod. manag. eng. (2022) 10(1), 51-64 creative commons attribution-noncommercial-noderivatives 4.0 international coelho et al. 56 http://creativecommons.org/licenses/by-nc-nd/4.0/ case 2. thermal power plant (table 4). table 4. case 2 – thermal power plant. constructs company has kpis in its production process, however they are not simulation results. organization of the machines in a network the machines are organized in a network using plcs for production control. integration of machines and the production process the process can be controlled remotely or automatically. its main process variables to be controlled: pressure and temperature. sensors are coupled to the equipment to provide system data to then be analyzed and stored, then they are shown on the screens of the computers supervised by the operators. sensors and control elements the information comes from sensors that capture the process variables of the industrial plant. data exchange and control in real-time the data is still transmitted informally, only with the help of intranet via radio, e-mail, and documentary. there is control of the entire production process in realtime kpi simulation company has kpis in its production process, however they are not simulation results. dashboard information generated by the production process is available in the plant’s control room data treatment and storage the company uses a supervisory system designed to capture and store information about the production process in a database. system interoperability there is system interoperability through the connection and communication between human and machines (real and virtual) level of autonomy approximately 85% of the production process is automated. vertical integration there is no vertical integration with the raw material supplier connection the company adopts well-defined communication protocols due to the characteristic of the production system. communication between machines is observed at different stages of the production process. conversion there is no big data storage and other methods for conversion meaningful information cyber there is no application and generation of predictive/cyber models. process cognition and optimization there is no application of decisionmaking and reasoning methods to recommend operations aimed at maintaining optimal production. configuration artificial intelligence and machine learning the system does not provide features that are self-configuring or involve machine learning and artificial intelligence. source: authors. case 3. plastic injection molding company (table 5). table 5. case 3 – plastic injection molding company. constructs evidences organization of the machines in a network the machines are connected to a central network, where the production data of all assets are stored, however, they do not communicate with each other, sharing only the storage database. integration of machines and the production process the company has a performance, production, and problem control based on remote sensing, which sends the collected data immediately to the production control telemetry system. sensors and control elements there is no direct communication between the machines, the information generated by the productivity sensors and controllers present in the assets are diagrammed and sent to a control platform data exchange and control in real-time there is the collection of information directly from machines and workstations. this information is available on a panel in the production area. there is real-time control of a significant part of the production process. kpi simulation company has kpis in its production process, however they are not simulation results. dashboard the information generated by the production process is visible on a dashboard in the production area and in the erp system. data treatment and storage the company has a communication of its data which is transmitted digitally. there is a great dependence on a direct communication between the pcp, engineering, logistics, and production teams, which can make it many times the exchange of information that is crucial for decision-making on the process is inefficient. (table 4 continues in the next column) (table 5 continues in the next page) (table 4 continues from the previous column) int. j. prod. manag. eng. (2022) 10(1), 51-64creative commons attribution-noncommercial-noderivatives 4.0 international cyber-physical production system assessment within the manufacturing industries in the amazon 57 http://creativecommons.org/licenses/by-nc-nd/4.0/ constructs evidences system interoperability there is system interoperability through the connection and communication between human and machines (real and virtual). especially in production process. level of autonomy the level of automation adopted by the company is low considering industry 4.0. just a production process has a high level of automation. vertical integration vertical integration (suppliers, company, and consumers) takes place from the network of machines/ equipment using erp. connection it is observed the adoption of communication protocols in the production process, through erp. however, there is a greater need for connection between the production process and other areas of the company. conversion there is no big data storage and other methods for conversion meaningful information cyber there is no application and generation of predictive/cyber models. process cognition and optimization there is no application of decisionmaking and reasoning methods to recommend operations aimed at maintaining optimal production. configuration artificial intelligence and machine learning the system does not provide features that are self-configuring or involve machine learning and artificial intelligence. source: authors. rq2: does the research framework provide adequate assessment of cpps practices? the instrument deals with 15 constructs that correspond to the main aspects that involve the concept of cyber-physical production systems. the instrument allows a comprehensive analysis of a company, seeking to identify the current stage of adaptation to the precepts related to cyber-physical production systems (pricewaterhousecoopers, 2014; european parliament, 2016; thiede, 2018; thiede et al., 2016; germany trade & invest, 2014; delloite, 2015; lee et al., 2017; lee et al., 2018), in addition to identifying what level the company is about the 5c architecture presented in lee et al. (2017). also, the complementary research techniques used in the in-depth case provided a diversity of sources of information and provided the scientific reliability necessary for this case study. the realization of an open structured interview, structured observation, content analysis, and document research enabled an in-depth case that contributes to the literature by deepening the observations in the real context about the application of concepts involving cyber-physical production systems in a company. in conclusion, the research framework made it possible to assess the maturity level of cyber-physical production systems within three manufacturing industries in the amazon, making it an original contribution to operations management. when associated with other research techniques (open structured interview, structured observation, content analysis, and documentary research) it provided a more comprehensive and deeper understanding of the phenomenon within the companies studied, as pointed out in voss et al. (2002), miguel (2012), creswell (2009) and mccutcheon & meredith (1993). the internal validity of the case study is confirmed by the systematic comparison of the literature concerning the research framework, whereas the reliability of the study is justified by the preparation of the database that was organized, integrated, and synthesized of the information obtained from different sources of evidence, resulting from the various research techniques employed (villarreal, 2017; villarreal & calvo, 2015; villarreal & landetta, 2010). rq3: what is the level of implementation observed in the company concerning the literature? case 1. chemical company considering the information collected about the cyber-physical production system theme at the chemical company, the radar graph was created (figure 1) presenting the measurement to the adoption of practices related to the theme. the constructs with the best performance were (1) integration of machines and the production process (2.0), (2) data exchange and real-time control (2.0), and (3) organization and networked machines in line with delloite (2015) and pricewaterhousecoopers (2014). the connection level averaged 1.5 and the conversion level averaged 1.25. cyber, cognition, and configuration scored 1.0. finally, the company’s overall average was 1.29. the chemical company has a level of implementation of the concepts of cyberphysical production systems in an initial stage to the connection level (lee et al., 2017) due to the process of integration of the machines and the use of control sensors still to be found in a stage that there is no single communication between these elements in the (table 5 continues from the previous page) int. j. prod. manag. eng. (2022) 10(1), 51-64 creative commons attribution-noncommercial-noderivatives 4.0 international coelho et al. 58 http://creativecommons.org/licenses/by-nc-nd/4.0/ production environment, in addition to the lack of a single communication protocol. the standardization of the communication equipment of the manufacture is still lacking. 0 1 2 3 4 5 q.1 q.2 q.3 q.4 q.5 q.6 q.7 q.8 q.9 q.10 q.11 q.12 q.13 q.14 q.15 q.16 q.17 q.18 q.19 q.20 q.21q.22q.23q.24 q.25 q.26 q.27 q.28 q.29 q.30 q.31 q.32 q.33 q.34 q.35 q.36 q.37 q.38 q.39 q.40 q.41 q.42 overa ll performa nce organization of the machines in a network integration of machines and the production process sensors and control elements data exchange and control in real-time system interoperability level of autonomy vertical integration connection conversion cyber process cognition and optimization configuration artificial intelligence and machine learning kpi simulation dashboard data treatment and storage figure 1. performance – chemical company. source: authors. case 2. thermal power plant figure 2 shows the radar graph generated from the information collected in the thermal power plant about the cyber-physical production system. the best performing constructs were (1) organization of the networked machines with an average of 3.0, followed by (2) sensors and control elements (3.0), (3) integration of the machines and the production process (2.0), and (4) data exchange and control in real-time (2.0) in line with delloite (2015) and pricewaterhousecoopers (2014). the connection level averaged 2.5 and the conversion level averaged 1.25. cyber, cognition, and configuration scored 1.0. finally, the company’s overall average was 1.52. the thermal power plant has a level of implementation of the concepts of cyber-physical production systems in an intermediate stage at the connection level, being possible to observe characteristics of the conversion level (lee et al., 2017). the very nature of the company’s activity reinforces the results because it is an organization that has a continuous production system, with little variability in the volume of production and no variability in terms of the diversity of final products. a certain standardization of manufacturing communication equipment is evident and raw data from the production process are converted into “meaningful” information. case 3. plastic injection molding company figure 3 shows the radar graph generated from the information collected in the plastic injection molding company regarding the cyber-physical production system. the best performing constructs were (1) organization of the machines in a network (3.0), followed by (2) sensors and control elements (3.0), (3) integration of the machines and the production process (2.5), and (4) data exchange and real-time control (2.0) delloite (2015) and pricewaterhousecoopers (2014). the connection level averaged 2.0 and the conversion level averaged 1.75. cyber, cognition, and configuration scored 1.0. in conclusion, the company’s overall average was 1.6. the plastic injection molding company has a level of implementation of the concepts of cyberphysical production systems in a similar stage to that observed in the thermal power plant (connection level, being possible to observe characteristics of the conversion level). the very nature of the company’s activity reinforces the results because the production process has a moderate-high level of automation. it was possible to observe a certain standardization of the manufacturing communication equipment with a higher level of conversion and visualization of product data becoming “meaningful” information. 0 1 2 3 4 5 q.1 q.2 q.3 q.4 q.5 q.6 q.7 q.8 q.9 q.10 q.11 q.12 q.13 q.14 q.15 q.16 q.17 q.18 q.19 q.20 q.21q.22q.23q.24 q.25 q.26 q.27 q.28 q.29 q.30 q.31 q.32 q.33 q.34 q.35 q.36 q.37 q.38 q.39 q.40 q.41 q.42 overa ll performa nce organization of the machines in a network integration of machines and the production process sensors and control elements data exchange and control in real-time system interoperability level of autonomy vertical integration connection conversion cyber process cognition and optimization configuration artificial intelligence and machine learning kpi simulation dashboard data treatment and storage figure 3. performance – plastic injection molding company. source: authors. in summary, the plastic injection molding company obtained superior performance in the evaluation, followed by the thermal power plant and chemical company indicating that the cyber-physical 0 1 2 3 4 5 q.1 q.2 q.3 q.4 q.5 q.6 q.7 q.8 q.9 q.10 q.11 q.12 q.13 q.14 q.15 q.16 q.17 q.18 q.19 q.20 q.21q.22q.23q.24 q.25 q.26 q.27 q.28 q.29 q.30 q.31 q.32 q.33 q.34 q.35 q.36 q.37 q.38 q.39 q.40 q.41 q.42 overa ll performa nce organization of the machines in a network integration of machines and the production process sensors and control elements data exchange and control in real-time system interoperability level of autonomy vertical integration connection conversion cyber process cognition and optimization configuration artificial intelligence and machine learning kpi simulation dashboard data treatment and storage figure 2. performance – thermal power plant. source: authors. int. j. prod. manag. eng. (2022) 10(1), 51-64creative commons attribution-noncommercial-noderivatives 4.0 international cyber-physical production system assessment within the manufacturing industries in the amazon 59 http://creativecommons.org/licenses/by-nc-nd/4.0/ production system has better performance when considering the type of production system adopted by the company and the level of automation in line with the cyber-physical production system (cpps) architecture framework presented in lee et al. (2018). companies evaluated even at an early stage of implementing concepts of cyber-physical production systems already have an advantage in the competitive environment (lee et al., 2013) about the competition and moving towards the concept of intelligent factories (cruz salazar et al., 2019; drennan-stevenson, 2019) in the amazon context, with data processing becoming intelligent, monitoring, optimization, and multidisciplinary engineering activities according to berger et al. (2019), bunte et al. (2019) and meixner et al. (2019). this study differs from the works presented in this research regarding the application of cpps (hu et al., 2016; lee et al., 2017; lee et al., 2018; drennan-stevenson, 2019; cruz salazar et al., 2018; ashtari talkhestani et al., 2019; among others presented), as its focus is on the evaluation of cyberphysical production systems from the point of view of operations management, while most of the works on the subject work from the point of view of the science of computing. however, the approach to work with a focus on the shop floor stands out, especially romero-silva & hernández-lópez (2020), govender et al. (2019), torres et al. (2019), and rocha et al. (2019). 6. conclusion this study aimed to assess the maturity level of cyberphysical production systems within manufacturing industries in the amazon. the research framework enabled a comprehensive assessment of companies by using different research techniques, making an original contribution to operations management, identifying opportunities for improvement, and assessing maturity in this context. the contributions of this research are relevant for academics and professionals. the theoretical contribution is in expanding the theme by conducting research focusing on operations management’s point of view on the subject. we seek to initiate a discussion of cyberphysical production systems from the perspective of planning, management, and evaluation characteristic of business management and expanding the body of knowledge related to the topic. all variables on the cpps evaluation and measurement form have been previously examined in the literature. the study contributes to the development of research in the amazon context and its application in other realities. among the managerial implications, the research contributed to the reflection on the part of the companies that participated in the study as to the current stage in which they are about the theme. the results can be used to optimize initiatives for productive and managerial excellence. the adopted form can assist other companies in the evaluation of their cyber-physical production systems and can be applied to evaluate an industrial sector as a whole. the proposed methodology can assist other companies in the development of the technological strategy, supporting the digital transformation process in order to obtain competitive advantage. our findings reveal a positive relationship between the type of production system adopted by the company, the level of automation, and the maturity of the cpps. the limitations of the research are associated with limited sample size, although the study was carried out in the form of an in-depth case, which enabled a greater level of deepening of the reality of the companies, and the impossibility of carrying out a longitudinal analysis. for future research, it is recommended to conduct similar research in a specific industrial sector, in addition to expanding the number of cases reported per study. references afanasev, m.y., fedosov, y.v., andreev, y.s., krylova, a.a., shorokhov, s.a., zimenko, k.v., & kolesnikov, m.v. (2019). a concept for integration of voice assistant and modular cyber-physical production system. in 2019 ieee international conference on industrial informatics (indin) pp. 27–32. institute of electrical and electronics engineers inc. https://doi.org/10.1109/indin41052.2019.8972015 ait-alla, a., kreutz, m., rippel, d., lütjen, m., & freitag, m. (2019). simulation-based analysis of the interaction of a physical and a digital twin in a cyber-physical production system. ifac-papersonline, 52(13), 1331–1336. https://doi.org/10.1016/j.ifacol.2019.11.383 agca, o., gibson, j., godsell, j., ignatius, j., wyn davies, c., & xu, o. (2017). an industry 4 readiness assessment tool. international institute for product and service innovation, 1–19. int. j. prod. manag. eng. (2022) 10(1), 51-64 creative commons attribution-noncommercial-noderivatives 4.0 international coelho et al. 60 https://doi.org/10.1109/indin41052.2019.8972015 https://doi.org/10.1016/j.ifacol.2019.11.383 http://creativecommons.org/licenses/by-nc-nd/4.0/ alves, c., & putnik, g.d. (2019). cyber-physical production system (cpps) decision making duration time impact on manufacturing system performance. fme transactions, 47(4), 675–682. https://doi.org/10.5937/fmet1904675a ansari, f., hold, p., & khobreh, m. (2020). a knowledge-based approach for representing jobholder profile toward optimal human– machine collaboration in cyber physical production systems. cirp journal of manufacturing science and technology, 28, 87–106. https://doi.org/10.1016/j.cirpj.2019.11.005 ashtari talkhestani, b., jung, t., lindemann, b., sahlab, n., jazdi, n., schloegl, w., & weyrich, m. (2019). an architecture of an intelligent digital twin in a cyber-physical production system. at-automatisierungstechnik, 67(9), 762–782. https://doi.org/10.1515/auto-2019-0039 baheti, r., & gill, h. (2011). cyber-physical systems. the impact of control technology, p. 161-166. bardin, l. (1977). l’analyse de contenu. presses universitaires de france. benbasat, i., goldstein, d.k., & mead, m. (1987). the case research strategy in studies of information systems. mis quarterly: management information systems, 11(3), 369–386. https://doi.org/10.4135/9781849209687.n5 berger, c., zipfel, a., braunreuther, s., & reinhart, g. (2019). approach for an event-driven production control for cyber-physical production systems. in procedia cirp, 79, 349–354. elsevier b.v. https://doi.org/10.1016/j.procir.2019.02.085 berlak, j., berg, j., stangl, m., & baumgart, l. (2020). real-time iot-based production planning and control of industrial robots in an automated cyber-physical production system under dynamic conditions: lessons learned from a make-to-order usage case. in lecture notes in computer science (including subseries lecture notes in artificial intelligence and lecture notes in bioinformatics), 12014 lncs, pp. 485–492. springer. https://doi.org/10.1007/978-3-030-45096-0_59 borangiu, t., morariu, o., răileanu, s., trentesaux, d., leitão, p., & barata, j. (2020). digital transformation of manufacturing. industry of the future with cyber-physical production systems. romanian journal of information science and technology, 23(1), 3–37. brauner, p., & ziefle, m. (2019). why consider the human-in-the-loop in automated cyber-physical production systems? two cases from cross-company cooperation. in 2019 ieee international conference on industrial informatics (indin), 2019-july, pp. 861–866. institute of electrical and electronics engineers inc. https://doi.org/10.1109/indin41052.2019.8972142 brauner, p., philipsen, r., valdez, a.c., & ziefle, m. (2019). human interaction under risk in cyber-physical production systems. in advances in intelligent systems and computing, 822, 421–430. springer verlag. https://doi.org/10.1007/978-3-319-96077-7_45 bunte, a., fischbach, a., strohschein, j., bartz-beielstein, t., faeskorn-woyke, h., & niggemann, o. (2019). evaluation of cognitive architectures for cyber-physical production systems. in 2019 ieee international conference on emerging technologies and factory automation, etfa, 2019-september, pp. 729–736. institute of electrical and electronics engineers inc. https://doi.org/10.1109/etfa.2019.8869038 chawla, v.k., angra, s., suri, s., & kalra, r.s. (2020). a synergic framework for cyber-physical production systems in the context of industry 4.0 and beyond. international journal of data and network science, 4(2), 237–244. https://doi.org/10.5267/j.ijdns.2019.12.002 creswell, j.w. (2009). research design: qualitative, quantitative, and mixed methods approaches. 3. ed. california, usa: sage. cruz salazar, l.a., ryashentseva, d., lüder, a., & vogel-heuser, b. (2019). cyber-physical production systems architecture based on multi-agent’s design pattern—comparison of selected approaches mapping four agent patterns. international journal of advanced manufacturing technology, 105(9), 4005–4034. https://doi.org/10.1007/s00170-019-03800-4 delloite. (2015). industry 4.0: challenges and solutions for the digital transformation and use of exponential technologies. suíça: delloite. ding, k., chan, f.t. s., zhang, x., zhou, g., & zhang, f. (2019). defining a digital twin-based cyber-physical production system for autonomous manufacturing in smart shop floors. international journal of production research, 57(20), 6315–6334. https://doi.org/10.1080/00207543.2019.1566661 drennan-stevenson, k. (2019). real-world implementation of cyber-physical production systems in smart manufacturing: cognitive automation, industrial processes assisted by data analytics, and sustainable value creation networks. journal of self-governance and management economics, 7(3), 14–20. https://doi.org/10.22381/jsme7320192 eckhart, m., ekelhart, a., lüder, a., biffl, s., & weippl, e. (2019). security development lifecycle for cyber-physical production systems. in 2019 iecon proceedings (industrial electronics conference), 2019-october, pp. 3004–3011. ieee computer society. https://doi.org/10.1109/iecon.2019.8927590 eisenhardt, k.m. (1989). building theories from case study research. academy of management review, 14(4), 532-550. https://doi.org/10.5465/amr.1989.4308385 european parliament. (2016). industry 4.0. bruxelas: ep. ferreira, j., lopes, f., doumeingts, g., sousa, j., mendonça, j.p., agostinho, c., & jardim-goncalves, r. (2020). empowering smes with cyber-physical production systems: from modelling a polishing process of cutlery production to cpps experimentation. in studies in computational intelligence, 864, 139–177. springer. https://doi.org/10.1007/978-3-030-38704-4_7 int. j. prod. manag. eng. (2022) 10(1), 51-64creative commons attribution-noncommercial-noderivatives 4.0 international cyber-physical production system assessment within the manufacturing industries in the amazon 61 https://doi.org/10.5937/fmet1904675a https://doi.org/10.1016/j.cirpj.2019.11.005 https://doi.org/10.1515/auto-2019-0039 https://doi.org/10.4135/9781849209687.n5 https://doi.org/10.1016/j.procir.2019.02.085 https://doi.org/10.1007/978-3-030-45096-0_59 https://doi.org/10.1109/indin41052.2019.8972142 https://doi.org/10.1007/978-3-319-96077-7_45 https://doi.org/10.1109/etfa.2019.8869038 https://doi.org/10.5267/j.ijdns.2019.12.002 https://doi.org/10.1007/s00170-019-03800-4 https://doi.org/10.1080/00207543.2019.1566661 https://doi.org/10.22381/jsme7320192 https://doi.org/10.1109/iecon.2019.8927590 https://doi.org/10.5465/amr.1989.4308385 https://doi.org/10.1007/978-3-030-38704-4_7 http://creativecommons.org/licenses/by-nc-nd/4.0/ francalanza, e., borg, j., & constantinescu, c. (2017). a knowledge-based tool for designing cyber physical production systems. computers in industry, 84, 39–58. https://doi.org/10.1016/j.compind.2016.08.001 francalanza, e., mercieca, m., & fenech, a. (2018). modular system design approach for cyber physical production systems. in procedia cirp, 72, 486–491. elsevier b.v. https://doi.org/10.1016/j.procir.2018.03.090 geisbauer, r., vedso, j., & schrauf, s. (2016). industry 4.0: building the digital enterprise. 2016 global industry 4.0 survey, 1–36. retrieved from www.pwc.com/industry40. gill, h. (2008). from vision to reality: cyber-physical systems. present. hcss natl. work. new res. dir. high confd. transp. cps automotive, aviat. rail. glatt, m., & aurich, j.c. (2019). physical modeling of material flows in cyber-physical production systems. in procedia manufacturing, 28, 10–17. elsevier b.v. https://doi.org/10.1016/j.promfg.2018.12.003 gronau, n. (2019). determining the appropriate degree of autonomy in cyber-physical production systems. cirp journal of manufacturing science and technology, 26, 70–80. https://doi.org/10.1016/j.cirpj.2019.05.001 germany trade & invest. (2014). industrie 4.0. germany, gtai, 2014. guérin, f., laville, a., daniellou, f., duraffourg, j., kerguelen, a. (2001). compreender o trabalho para transformá-lo: a prática da ergonomia. são paulo: blücher: fundação vanzolini. govender, e., telukdarie, a., & sishi, m.n. (2019). approach for implementing industry 4.0 framework in the steel industry. in 2019 ieee international conference on industrial engineering and engineering management, pp. 1314–1318. ieee computer society. https://doi.org/10.1109/ieem44572.2019.8978492 hu, f., lu, y., vasilakos, a.v., hao, q., ma, r., patil, y., … xiong, n.n. (2016). robust cyber-physical systems: concept, models, and implementation. future generation computer systems, 56, 449–475. https://doi.org/10.1016/j.future.2015.06.006 iglesias, a., sagardui, g., & arellano, c. (2019). industrial cyber-physical system evolution detection and alert generation. applied sciences (switzerland), 9(8). https://doi.org/10.3390/app9081586 lee, j., lapira, e., bagheri, b., & kao, h. an. (2013). recent advances and trends in predictive manufacturing systems in big data environment. manufacturing letters, 1(1), 38–41. https://doi.org/10.1016/j.mfglet.2013.09.005 lee, j., jin, c., & bagheri, b. (2017). cyber physical systems for predictive production systems. production engineering, 11(2), 155–165. https://doi.org/10.1007/s11740-017-0729-4 lee, j.h., do noh, s., kim, h.j., & kang, y.s. (2018). implementation of cyber-physical production systems for quality prediction and operation control in metal casting. sensors (switzerland), 18(5). https://doi.org/10.3390/s18051428 lichtblau, k., stich, v., bertenrath, r., blum, m., bleider, m., … schroter, m. (2015). industrie 4.0 readiness. impuls-stiftung, 0–76. liu, c., jiang, p., & jiang, w. (2020). web-based digital twin modeling and remote control of cyber-physical production systems. robotics and computer-integrated manufacturing, 64. https://doi.org/10.1016/j.rcim.2020.101956 lins, t., & oliveira, r.a. r. (2020). cyber-physical production systems retrofitting in context of industry 4.0. computers and industrial engineering, 139. https://doi.org/10.1016/j.cie.2019.106193 llamuca, e.s., garcia, c.a., naranjo, j.e., & garcia, m.v. (2019). cyber-physical production systems for industrial shop-floor integration based on amqp. in proceedings 2019 international conference on information systems and software technologies, ici2st 2019, pp. 48–54. institute of electrical and electronics engineers inc. https://doi.org/10.1109/ici2st.2019.00014 loucopoulos, p., kavakli, e., & chechina, n. (2019). requirements engineering for cyber physical production systems. in lecture notes in computer science (including subseries lecture notes in artificial intelligence and lecture notes in bioinformatics) 11483 lncs, pp. 276–291. springer verlag. https://doi.org/10.1007/978-3-030-21290-2_18 marconi, m.a., & lakatos, e.m. (2002). técnicas de pesquisa: planejamento e execução de pesquisas, amostragens e técnicas de pesquisa, elaboração, analise e interpretação de dados. 5. ed. são paulo: atlas. mayrhofer, m., mayr-dorn, c., zoitl, a., guiza, o., weichhart, g., & egyed, a. (2019). assessing adaptability of software architectures for cyber physical production systems. in lecture notes in computer science (including subseries lecture notes in artificial intelligence and lecture notes in bioinformatics), 11681 lncs, pp. 143–158. springer verlag. https://doi.org/10.1007/978-3-030-29983-5_10 mccutcheon, d.m., & meredith, j.r. (1993). conducting case study research in operations management. journal of operations management, 11(3), 239–256. https://doi.org/10.1016/0272-6963(93)90002-7 miguel, p.a.c. (org.). (2012). metodologia de pesquisa em engenharia de produção e gestão de operações. 2. ed. rio de janeiro: elsevier: abepro. meixner, k., rabiser, r., & biffl, s. (2019). towards modeling variability of products, processes and resources in cyber-physical production systems engineering. in acm international conference proceeding series (vol. b). association for computing machinery. https://doi.org/10.1145/3307630.3342411 monostori, l., kádár, b., bauernhansl, t., kondoh, s., kumara, s., reinhart, g., … ueda, k. (2016). cyber-physical systems in manufacturing. cirp annals, 65(2), 621–641. https://doi.org/10.1016/j.cirp.2016.06.005 int. j. prod. manag. eng. (2022) 10(1), 51-64 creative commons attribution-noncommercial-noderivatives 4.0 international coelho et al. 62 https://doi.org/10.1016/j.compind.2016.08.001 https://doi.org/10.1016/j.procir.2018.03.090 http://www.pwc.com/industry40 https://doi.org/10.1016/j.promfg.2018.12.003 https://doi.org/10.1016/j.cirpj.2019.05.001 https://doi.org/10.1109/ieem44572.2019.8978492 https://doi.org/10.1016/j.future.2015.06.006 https://doi.org/10.3390/app9081586 https://doi.org/10.1016/j.mfglet.2013.09.005 https://doi.org/10.1007/s11740-017-0729-4 https://doi.org/10.3390/s18051428 https://doi.org/10.1016/j.rcim.2020.101956 https://doi.org/10.1016/j.cie.2019.106193 https://doi.org/10.1109/ici2st.2019.00014 https://doi.org/10.1007/978-3-030-21290-2_18 https://doi.org/10.1007/978-3-030-29983-5_10 https://doi.org/10.1016/0272-6963(93)90002-7 https://doi.org/10.1145/3307630.3342411 https://doi.org/10.1016/j.cirp.2016.06.005 http://creativecommons.org/licenses/by-nc-nd/4.0/ mourad, m.h., nassehi, a., schaefer, d., & newman, s.t. (2020). assessment of interoperability in cloud manufacturing. robotics and computer-integrated manufacturing, 61. https://doi.org/10.1016/j.rcim.2019.101832 mourtzis, d., & vlachou, e. (2016). cloud-based cyber-physical systems and quality of services. tqm journal, 28(5), 704–733. https://doi.org/10.1108/tqm-10-2015-0133 mühlfelder, m. (2019). analysis and design of a cyber-physical production system (cpps) in sensor manufacturing. a case study. in advances in intelligent systems and computing, 822, pp. 391–400. springer verlag. https://doi.org/10.1007/978-3-319-96077-7_41 park, k.t., lee, j., kim, h.j., & noh, s.d. (2020). digital twin-based cyber physical production system architectural framework for personalized production. international journal of advanced manufacturing technology, 106(5–6), 1787–1810. https://doi.org/10.1007/s00170-019-04653-7 patalas-maliszewska, j., & schlueter, n. (2019). model of a knowledge management for system integrator(s) of cyberphysical production systems (cpps). in lecture notes in mechanical engineering (pp. 92–103). pleiades publishing. https://doi.org/10.1007/978-3-030-18715-6_8 prist, m., monteriu, a., freddi, a., pallotta, e., cicconi, p., giuggioloni, f., … longhi, s. (2019). cyber-physical manufacturing systems for industry 4.0: architectural approach and pilot case. in 2019 ieee international workshop on metrology for industry 4.0 and iot, metroind 4.0 and iot 2019 proceedings (pp. 219–224). institute of electrical and electronics engineers inc. https://doi.org/10.1109/metroi4.2019.8792880 pinzone, m., albè, f., orlandelli, d., barletta, i., berlin, c., johansson, b., & taisch, m. (2020). a framework for operative and social sustainability functionalities in human-centric cyber-physical production systems. computers and industrial engineering, 139. https://doi.org/10.1016/j.cie.2018.03.028 pricewaterhousecoopers. (2014). industry 4.0: opportunities and challenges of the industrial internet. germany: pwc. rocha, a. di., tripa, j., alemao, d., peres, r.s., & barata, j. (2019). agent-based plug and produce cyber-physical production system test case. in 2019 ieee international conference on industrial informatics (indin), 2019-july, pp. 1545–1551. institute of electrical and electronics engineers inc. https://doi.org/10.1109/indin41052.2019.8972169 romero-silva, r., & hernández-lópez, g. (2020). shop-floor scheduling as a competitive advantage: a study on the relevance of cyber-physical systems in different manufacturing contexts. international journal of production economics, 224. https://doi.org/10.1016/j.ijpe.2019.107555 sanislav, t., & miclea, l. (2012). cyber-physical systems concept, challenges and research areas. ceai, 14(02), 28-33. sha, l., gopalakrishnan, s., liu, x., & wang, q. (2008). cyber-physical systems: a new frontier. in proceedings 2008 ieee international conference on sensor networks, ubiquitous, and trustworthy computing, pp. 1–9. https://doi.org/10.1109/sutc.2008.85 schreiber, m., vernickel, k., richter, c., & reinhart, g. (2019). integrated production and maintenance planning in cyber-physical production systems. in procedia cirp, 79, 534–539. elsevier b.v. https://doi.org/10.1016/j.procir.2019.02.095 schuh, g., anderl, r., dumitrescu, r., krüger, a., & hompel, m. (2020). industrie 4.0 maturity index. managing the digital transformation of companies–update 2020. acatech study. stock, d., schel, d., & bauernhansl, t. (2019). cyber-physical production system self-description-based data access layer. in 2019 ieee international conference on emerging technologies and factory automation, etfa, 2019-september, pp. 168–175. institute of electrical and electronics engineers inc. https://doi.org/10.1109/etfa.2019.8869486 stock, d., schel, d., & bauernhansl, t. (2020). middleware-based cyber-physical production system modeling for operators. in procedia manufacturing, 42, 111–118. elsevier b.v. https://doi.org/10.1016/j.promfg.2020.02.031 thiede, s. (2018). environmental sustainability of cyber physical production systems. in procedia cirp, 69, 644–649. elsevier b.v. https://doi.org/10.1016/j.procir.2017.11.124 thiede, s., juraschek, m., & herrmann, c. (2016). implementing cyber-physical production systems in learning factories. in procedia cirp, 54, 7–12. elsevier b.v. https://doi.org/10.1016/j.procir.2016.04.098 torres, p., dionísio, r., malhão, s., neto, l., ferreira, r., gouveia, h., & castro, h. (2019). cyber-physical production systems supported by intelligent devices (smartboxes) for industrial processes digitalization. in 5th international forum on research and technologies for society and industry: innovation to shape the future, rtsi 2019 proceedings, pp. 73–78. institute of electrical and electronics engineers inc. https://doi.org/10.1109/rtsi.2019.8895553 vergara, s.c. (2009). métodos de coleta de dados no campo. são paulo: atlas. villarreal, o. (2017). is it desirable, necessary and possible to perform research using case studies? cuadernos de gestion, 17(1), 147-172. https://doi.org/10.5295/cdg.140516ov villarreal, o., & landeta, j. (2010). case study as a scientific research methodology in business economics and strategic management. an application to internationalisation. investig. eur. direccion econ. empres., 16(3), 31-52. https://doi.org/10.1016/s1135-2523(12)60033-1 int. j. prod. manag. eng. (2022) 10(1), 51-64creative commons attribution-noncommercial-noderivatives 4.0 international cyber-physical production system assessment within the manufacturing industries in the amazon 63 https://doi.org/10.1016/j.rcim.2019.101832 https://doi.org/10.1108/tqm-10-2015-0133 https://doi.org/10.1007/978-3-319-96077-7_41 https://doi.org/10.1007/s00170-019-04653-7 https://doi.org/10.1007/978-3-030-18715-6_8 https://doi.org/10.1109/metroi4.2019.8792880 https://doi.org/10.1016/j.cie.2018.03.028 https://doi.org/10.1109/indin41052.2019.8972169 https://doi.org/10.1016/j.ijpe.2019.107555 https://doi.org/10.1109/sutc.2008.85 https://doi.org/10.1016/j.procir.2019.02.095 https://doi.org/10.1109/etfa.2019.8869486 https://doi.org/10.1016/j.promfg.2020.02.031 https://doi.org/10.1016/j.procir.2017.11.124 https://doi.org/10.1016/j.procir.2016.04.098 https://doi.org/10.1109/rtsi.2019.8895553 https://doi.org/10.5295/cdg.140516ov https://doi.org/10.1016/s1135-2523(12)60033-1 http://creativecommons.org/licenses/by-nc-nd/4.0/ villarreal, o., & calvo, n. (2015). from the triple helix model to the global open innovation model: a case study based on international cooperation for innovation in dominican republic. journal of engineering and technology management, 35(1), 71-92. https://doi.org/10.1016/j.jengtecman.2014.10.002 voss, c., tsikriktsis, n., & frohlich, m. (2002). case research in operations management. international journal of operations and production management, 22(2), 195–219. https://doi.org/10.1108/01443570210414329 yang, lu. (2017). industry 4.0: a survey on technologies, application and open research. journal of industrial information integration, 6, 1-10. https://doi.org/10.1016/j.jii.2017.04.005 wang, l., törngren, m., & onori, m. (2015). current status and advancement of cyber-physical systems in manufacturing. journal of manufacturing systems, 37, 517–527. https://doi.org/10.1016/j.jmsy.2015.04.008 wu, x., goepp, v., & siadat, a. (2019a). cyber physical production systems: a review of design and implementation approaches. in 2019 ieee international conference on industrial engineering and engineering management, pp. 1588–1592. ieee computer society. https://doi.org/10.1109/ieem44572.2019.8978654 wu, x., goepp, v., & siadat, a. (2019b). the integrative link between cyber physical production systems and enterprise information systems. in proceedings of 2019 international conference on computers and industrial engineering, cie, 2019-october. computers and industrial engineering. yin, r.k.(1994). case study research, design and methods, 2nd ed. newbury park, sage publications. zhang, h., zhang, g., & yan, q. (2019). digital twin-driven cyber-physical production system towards smart shop-floor. journal of ambient intelligence and humanized computing, 10(11), 4439–4453. https://doi.org/10.1007/s12652-018-1125-4 zhang, h., yan, q., & wen, z. (2020). information modeling for cyber-physical production system based on digital twin and automationml. international journal of advanced manufacturing technology, 107(3–4), 1927–1945. https://doi.org/10.1007/s00170-020-05056-9 int. j. prod. manag. eng. (2022) 10(1), 51-64 creative commons attribution-noncommercial-noderivatives 4.0 international coelho et al. 64 https://doi.org/10.1016/j.jengtecman.2014.10.002 https://doi.org/10.1108/01443570210414329 https://doi.org/10.1016/j.jii.2017.04.005 https://doi.org/10.1016/j.jmsy.2015.04.008 https://doi.org/10.1109/ieem44572.2019.8978654 https://doi.org/10.1007/s12652-018-1125-4 https://doi.org/10.1007/s00170-020-05056-9 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j international journal of production management and engineering lessons from empirical studies in product and service variety management lyons, a. c. university of liverpool management school, liverpool l69 7zh, uk. a.c.lyons@liverpool.ac.uk abstract: for many years, a trend for businesses has been to increase market segmentation and extend product and service-variety offerings in order to provid more choice for customers and gain a competitive advantags. however, there have been relatively few variety-related, empirical studies that have been undertaken. in this research, two empirical studies are presented that address the impact of product and service variety on business and business function performance. in the first (service-variety) study, the focus concerns the relationship between service provision offered by uk-based, third-party logistics (3pl) providers and the operational and financial performance of those providers. here, the results of a large survey identify the most important services offered by 3pls and the most important aspects of 3pl operational performance. also, the research suggests that the range of service variety offered by 3pls does not directly influence the 3pls’ financial performance. the second (product-variety) study presents the findings from an analysis of data from 163 manufacturing plants where the impact of product variety on the performance of five business functions is examined. an increase in product variety was found to influence business functions differently depending on the combination of customisation and variety offered to customers. key words: variety management, sku proliferation, mass customisation. 1. introduction success for many businesses is often dependent on their ability to innovate, develop new ideas and introduce new products and services. in modern, highly-competitive business environments it is difficult for sales to be maintained or grown from a fixed portfolio of products or services. rather, sales growth is dependent on the ability of a business to stimulate an existing market or penetrate a different one by offering new choices. consequently, product development has become more rapid, manufacturing systems have become more flexible and stock-keeping unit (sku) proliferation and variety continue to increase (fisher & ittner, 1999; hu et al., 2011; meyr, 2004) . differentiation of products has gone beyond the simple and prosaic categories of age, size and gender to include regional and national tastes, personal attributes and personal lifestyle. the management of the complexity associated with wide product diversity has become is core to competitive advantage. however, despite the fact that high product variety may lead to an increase in sales, it does not necessarily guarantee an increase in business profits or competitiveness. moreover, product variety can have a positive effect on both sales and market share, but can also have negative consequences for business performance (yeh & chu, 1991). for example, higher product variety may increase manufacturing costs through an increase in the complexity of the production process. it can also cause higher complexity of the demand forecasting process and make obdurate the alignment of supply with demand in the supply chain (whang & lee, 1998; randall & ulrich, 2001). many businesses have started to recognise that ‘more’ is not necessarily better and that a trade-off exists between product variety and business function performance (thonemann & bradley, 2002). those increasing variety in their products and services should also, therefore, consider the impact of product variety on the performance and cost profile of their business functions. decisions relating to product variety can be viewed as focusing on how to innovate, engineer and produce products and services with the requisite level of customer choice. however, only by extending this focus to other business functions can the full implications of product variety be revealed (ramdas, 2003). the fundamental question concerns the level of variety offered. the solution necessarily concerns the need to assess the benefits in relation to the increased cost and resource burden. http://dx.doi.org/10.4995/ijpme.2013.1557 received: 2013-05-20 accepted: 2013-06-10 https://ojs.upv.es/index.php/ijpme 55int. j. prod. manag. eng. (2013) 1(1), 55-62creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2013.1557 2. empirical studies 2.1. study 1 – service variety management in recent years, many 3pl providers have extended their service range to include warehousing, freight forwarding, packaging and managing product returns. extending service variety has increased competition amongst 3pl providers. the aim of this study is to provide an evaluation of the relationship between the variety of services offered by ukbased, third-party logistics (3pl) providers and the performance of those providers. the evaluation is based on a recent survey (liu & lyons, 2011). to support the development of competitive strategies and for mitigating investment risks, logisticians need to understand the relationship that exists between 3pl performance and different logistics service offerings. previous research concerning the relationship between service variety and performance has made only a limited contribution to understanding the relationship that exists between 3pl performance and the range of service provision. in addition, there has been relatively little attention given to empirical studies of both providers and customers. this research has set out to address these gaps by empirically exploring the relationships between service variety and performance from both a provider and customer point of view. the level of the provider’s service capabilities should meet the customer’s requirements. therefore, the review of the service capabilities is based on both a provider and customer perspective. the key phases of the study methodology are shown in figure 1. in summary, the methodology concerned the identification of the performance and service capabilities of 3pl providers. this consisted of a review of previous studies and the development of a survey questionnaire. cluster analysis was used to distinguish 3pls in terms of service variety. a one-way analysis of variance (anova) was used to test whether there were significant differences in 3pl operational and financial performance. a simple regression analysis was used to evaluate the relationship between operational performance and the financial performance of 3pl providers. lyons, a. c. 56 int. j. prod. manag. eng. (2013) 1(1), 55-62 creative commons attribution-noncommercial 3.0 spain figure 1. study 1 methodology. table 1. financial and operational performance indicators. items f1. gross profit margin f2. sales growth o1. to deliver expedited shipments/speed of delivery o2. to offer short delivery lead-time o3. on time and accurate delivery o4. higher customer satisfaction ratings o5. to enhance customer success o6. lower customer complaints (percentage of total sales) o7. to deliver goods in an undamaged state o8. to accommodate special or non-routine requests o9. to handle unexpected events o10. to provide quicker response to customers o11. to operate with low overall operating cost as a percentage of sales o12. to improve the rate of utilisation of facilities/equipment/manpower in providing the services o13. aggressiveness in increasing the value-added content of services o14. aggressiveness in the reduction of order cycle time o15. to provide new and better services/ speed of introduction for new services a postal survey was used as the principal method of data collection for this study. financial performance was measured on a two-item scale: gross profit and sales margin. the fifteen indicators for operational performance were identified by referring to previous logistics research (ellinger et al., 2002; fawcett & smith, 1995) and from discussions with logistics academics and practitioners (see table 1). the 32 service variety indicators were identified by referring to previous logistics research (lai, 2004; murphy & poist, 2000; stefansson, 2006) and by conducting personal interviews with practitioners (see table 2). 621 3pls that provided transportation and warehouse-related services were contacted as part of the survey. a further 595 large manufacturing customer companies were also contacted. for 3pls, the effective population size was reduced to 513 as 93 respondents indicated that their companies only provided services for internal users, 11 service providers did not provide any transportation/ warehousing or value-added related services and 4 of the respondents did not provide services for manufacturing. the total usable number of responses was 112. therefore, the overall response rate was 21.8% (112/513). for customers, 168 usable questionnaires were obtained. therefore, the total response rate was 28.2 per cent (168/595). the reliability of a questionnaire is associated with the consistency of responses to questions. reliability is usually expressed on the basis of the cronbach’s alpha coefficient. levels of 0.70 or more are generally accepted as representing good reliability (hair et al., 2006). with the exception of financial performance for the 3pls (0.605), all of the reliability scores exceeded this minimum reliability standard of 0.70. 2.2. study 1 results for customers, outbound distribution was viewed as the most important service capability by respondents, followed by order fulfilment, rate negotiation, tracking and tracing and interfacing with erp systems (mean scores were derived from a sevenpoint scale where 1 represented very unimportant and 7 signified very important). table 3 highlights the ten services identified as being the most important for customers. on-time and accurate delivery was viewed as the most important aspect of operational performance by respondents for customers, followed by undamaged state delivery, and higher customer satisfaction. table 4 presents an importance ranking and highlights the top five most-important items. in order to classify the 3pls according to the variety of their service provision, a cluster analysis was undertaken using the 32 items. there are two approaches that are most-widely used for this procedure: the hierarchical method and the nonhierarchical method. in this research, a hierarchical lessons from empirical studies in product and service variety management 57int. j. prod. manag. eng. (2013) 1(1), 55-62creative commons attribution-noncommercial 3.0 spain table 2. service variety indicators. s1. inbound transportation s9. inventory management s17. simple processing s25. selection of software s2. outbound distribution s10. pick and pack s18. bar code scanning s26. interfacing with erp systems s3. merge in transit s11. order fulfilment s19. rfid s27. invoicing/billing function s4. rate negotiation s12. cross-docking s20. edi capability s28. freight bill auditing/ payment s5. carrier selection s13. product returns s21. electronic commerce s29. billing the final customer s6. freight forwarding s14. labelling/marking s22. tracking and tracing s30. insurance service s7. storage s15. packaging s23. logistics information systems s31. consulting services s8. storage of special requirements s16. relabeling/ repackaging s24. order management systems s32. management reports table 3. importance of 3pl service capabilities to customers. rank service capabilities mean 1 s2. outbound distribution 6.288 2 s11. order fulfilment 6.119 3 s4. rate negotiation 5.966 4 s22. tracking and tracing 5.774 5 s26. interfacing with erp systems 5.770 6 s23. logistics information systems 5.690 7 s7. storage 5.689 8 s32. management reports 5.669 9 s9. inventory management 5.593 10 s29. billing the final customer 5.590 cluster analysis, by way of ward’s (1963) partitioning technique and the squared euclidean distancemethod, was used, and provided the most suitable number of clusters. all responding firms were assigned initially to these clusters. a non-hierarchical technique, that is, k-means cluster analysis was subsequently used to re-assign the respondents into the most appropriate clusters through an iterative process. the 92 responding firms (in order to conduct a k-means cluster analysis, 20 of the 112 were excluded due to missing data) were assigned to three clusters: 36 in cluster 1, 21 in cluster 2 and 35 in cluster 3. a one-way anova was used to examine which of the service capabilities differed across the three clusters. 27 items were found to significantly differ. only five items (s1: inbound transportation, s2: outbound distribution, s6: freight forwarding, s29: billing the final customer and s30: insurance service) did not significantly differ across the three clusters. the first type (cluster 1, n=36) accounts for 39.1% of the sample. these types of 3pl achieved a mediumlevel of capability concerning transportationrelated services (s1, and s3~s6), warehousingrelated (s9~s13), value-added services (s14~s17), information technology (s21~s25), finance-related (s27~s30) and other services (s31 and s32). compared to cluster 3, they have a much higher capability in warehousing-related, valued-added and information technology-related services. the second type (cluster 2, n=21, 22.8%) possesses a high level of capability in most of the 32 logistics service items. this suggests that they are comprehensive 3pls. the final type of 3pl (cluster 3, n=35, 38.0%) possesses a medium-level of capability in carrying out the three aspects of transportation-related service (s1, s2 and s4), the one aspect of warehousing-related (s7), all of the finance-related services (s27~s30) and one aspect of other services (s32). this type of 3pl under-performed in most warehousing-related, value-added, and information technology aspects of provision. these firms are traditional transportation companies. to determine if 3pl clusters differ in terms of financial and operational performance, another oneway analysis of variance (anova) was undertaken. the anova results presented in table 5 indicate that statistically significant differences, that is p < 0.05, lyons, a. c. 58 int. j. prod. manag. eng. (2013) 1(1), 55-62 creative commons attribution-noncommercial 3.0 spain table 4. importance of 3pl operational performance to customers. rank operational performances mean 1 o3. on time and accurate delivery 6.62 2 o7. undamaged state delivery 6.57 3 o4. higher customer satisfaction 6.10 4 o6. lower customer complaints (percentage of total sales) 5.99 5 o1. to deliver expedited shiments/speed of delivery 5.85 table 5. anova analysis of performance differences across the three clusters. items 1 (n=36) 2 (n=21) 3 (n=35) f f1. gross profit margin 4.57 4.62 4.09 1.630 f2. sales growth 4.36 5.05 4.37 2.183 o1. to deliver expedited shipments/speed of delivery 5.26 5.52 5.26 0.696 o2. to offer short delivery lead-time 5.50 5.55 5.17 1.325 o3. on time and accurate delivery 5.64 5.90 5.54 1.043 o4. higher customer satisfaction ratings 5.61 6.05 5.49 3.465* o5. to enhance customer success 5.31 5.90 5.14 4.515* o6. lower customer complaints (percentage of total sales) 5.53 5.86 5.29 1.763 o7. to deliver goods in an undamaged state 5.77 6.05 5.50 2.376 o8. to accommodate special or non-routine requests 6.11 6.10 5.76 1.484 o9. to handle unexpected events 6.00 6.19 5.97 0.422 o10. to provide quicker response to customers 5.89 6.14 5.62 2.550 o11. to operate with low overall operating cost as a percentage of sales 4.67 4.81 4.26 1.654 o12. to improve the rate of utilization of facilities/ equipment/manpower in providing the services 4.86 5.33 4.50 5.081* o13. aggressiveness in increasing the value-added content of services 4.63 5.38 4.49 5.422* o14. aggressiveness in the reduction of order cycle time 4.57 5.30 4.41 5.421* o15. to provide new and better services/speed of introduction for new services 5.00 5.38 4.55 3.814* overall operational performance** 5.36 5.70 5.13 6.319* * represents significant level p < 0.05 ** means the average of all aspects of operational performance. *** pairwise differences shown are significant at the 0.05 level. existed among the three 3pl clusters in some of the operational performance items. results of a chi-square analysis revealed that total sales volume and number of employees significantly differed across the three clusters at the p < 0.05 significance level. 2.3. study 1 conclusions a positive and significant relationship was found between operational performance and the 3pl financial performance. this finding suggests that if 3pls can improve their operational performance, they will increase the financial performance of their businesses. it implies that customers will be more satisfied with using their services. the influences of service variety on 3pl operational performance were partially supported. it appears that 3pl clusters with a wide service variety offering generally have better operational performance. results showed the ratings differed significantly in six of the fifteen aspects of operational performance. aligning high levels of operational performance with quality (i.e., o4 and o5) and innovation (i.e., o13~o15) was found to be a necessary strategy for the uk’s 3pl providers. however, the impact of service variety on the 3pl providers’ financial performance was not supported. this implies that the range of service provision offered by 3pls cannot directly influence the 3pls’ financial performance. through a better operational performance, 3pl providers with a broader range of service provision that correspond to the key priorities of customers will gain superior financial performance. 2.4. study 2 – product variety management the key aim of this second study was to explore and compare the impact of product variety on business function performance. five business functions were examined: engineering, manufacturing, purchasing, logistics and marketing. a questionnaire composed of 37 questions concerning the impact of product variety on business function performance and 5 questions related to product variety and customisation were sent to 1,500 manufacturing companies. all target companies were large enterprises (les) with total sales values in excess of £2 million. les were selected as the intention was to target manufacturers that had resources to invest in increasing product variety. individuals from 163 companies with increasing variety responded to the impact of product variety questions from the survey questionnaire, which is an acceptable, overall response rate overall of 15%. respondents were asked to “indicate the impact of increased product variety on each item where the manufacturer had had a variety increase in its main product family during the past five years” using a 1-10 scale, on which 1 indicated the lowest increase and 10 the highest increase. 2.5. study 2 results performance within each function was captured from a number of individual items: four in engineering (α=0.866), sixteen in manufacturing (α=0.952), three in purchasing (α=0.883), nine in logistics (α=0.946) and five in marketing (α=0.891). the average impact of an increase in product variety on each of the different business functions was found to be as follows: marketing (m=4.86), engineering (m=4.65), manufacturing (m=4.05), purchasing (m=4.03), and logistics (m=3.87). items can be regarded as both the cost and non-cost related aspects of business function performance. costrelated items have strong correlations with each other (p<0.01). non-cost related performance can be either positive or negative: 1) positive: competitive advantage; customer satisfaction; market share; product flexibility; utilisation of standardised parts; postponement; outsourcing, 2) negative: demand forecast uncertainty; scheduling complexity; design complexity; manufacturing complexity; part variety; supervision effort; total quality control; manufacturing lead time; process variety; workin-process inventory; finished goods inventory; purchased component/part variety; purchased part inventory; delivery time; order process complexity. one of the most significant motivations for an increase in product variety is the ability to customise the product. thus, with regard to customisation, the first of these factors, a high level of customisation is expected to have a corresponding high level of product variety and a level of variety that is higher than a low level of customisation. the three dimensions (fundamental, intermediate and peripheral) of variety were tested in relation to each customisation type (pure standardisation, segmented standardisation, customised standardisation, tailored customisation, pure customisation) using a oneway analysis of variance (anova). table 6 depicts the results. the results show significant statistical differences at the .05 and .01 levels. typically, high customisation types were expected to display higher product variety than low customisation types with a general increase in variety across the ps to lessons from empirical studies in product and service variety management 59int. j. prod. manag. eng. (2013) 1(1), 55-62creative commons attribution-noncommercial 3.0 spain pc continuum. however, unexpectedly, tailored customisation (tc) displayed the highest level of product variety. this can be explained by the fact that empirically pc industries do not typically use their full variety-producing capabilities. to examine the different impacts of increased product variety on the performance of the different business functions, a one-way analysis of variance (anova) was undertaken. the anova results (table 7) indicate that statistically significant differences exist among the different customisation types. ps typically is impacted upon the most by an increase in product variety, followed by ss, cs, tc and pc. that is, the impact of increased product variety decreased across the ps to pc continuum. this is as expected and is attributable to an increase of the business function flexibility in the more-customised types. overall, 7 items (p < 0.01), 11 items (p < 0.05) and 5 items (p < 0.1) out of the 37 items showed significant differences according to customisation type. 2.6. study 2 findngs the unit cost of the product exhibited a significant difference across the continuum of customisation types for the engineering function. increased overhead, direct labour and material costs owing to increased product variety lead to a higher unit cost. however, customised standardisation (cs) and tailored customisation (tc) often use component sharing in the design of product families, which reduces overhead cost and the increase of the unit cost of a product can be reduced compared with pure standardisation (ps) and segmented standardisation (ss) even allowing for ps and ss making use of appropriate economies of scale. manufacturing, material and process technology investment cost displayed statistically significant differences across the customisation types, and in accordance with the expected trend across the continuum. ps incurs the highest escalation in manufacturing and material costs, followed by ss, cs, tc, and pc. the results highlight that a flexible manufacturing system and supporting business function design are essential factors to mitigate the trade-off between product variety and increased manufacturing cost. in the case of process technology investment cost, pc is affected less than cs, followed by tc, ss, and ps. the anova test demonstrates the highest increase in the use of standardised parts for ps, followed by cs, ss, tc, and pc. it is worthy of note that cs had the highest increase in the use of postponement, followed by ps, ss, tc, and pc. the result implies that the cs environments typically employ an assemble-to-order (ato) production logic and are heavily reliant on postponement strategies and modularisation. as expected, with respect to product flexibility, low customisation types such as ps and ss are affected more than high customisation types due to an increase in the use of standardised material. process and part variety, manufacturing complexity and lead time are most adversely affected for the ps and ss types with an increase in product variety. ps displays the highest increase in purchasing costs with product variety increases. ps suffers from a policy that typically requires the purchase of high volumes from selected suppliers and is consequently more adversely affected by increased parts and material variety than the more customised types. further down the continuum, tc demonstrates the greatest increase in purchasing costs. similarly, ps displays the highest increase in purchased components and materials, followed by cs, ss, tc and pc. market mediation costs, including inventory holding, mark-down, and lost sales, are primarily influenced by demand uncertainty. although uncertainty of demand increases and forecasting accuracy decreases generally from a make-to-stock (mts) to a designto-order (dto) strategy, the ps type may be affected more in the cost of inventory holding, mark-downs, and lost sales due the position of its decoupling point. pc typically has low market mediation cost because of the upstream decoupling point that allows inventory holding and stock-out costs to be affected less by an increase in variety. ps incurs a higher increase in transportation costs than the more customised types. the results indicate that the increased costs of the low customisation types may exceed the increased costs associated with less-than-truck-loads (ltl) in lyons, a. c. 60 int. j. prod. manag. eng. (2013) 1(1), 55-62 creative commons attribution-noncommercial 3.0 spain table 6. anova analysis of variety differences according to customisation type. fundamental variety intermediate variety peripheral variety mean ps 3.19 3.23 2.94 ss 3.09 3.47 3.29 cs 3.75 4.02 4.02 tc 4.14 4.24 4.05 pc 3.77 3.80 3.70 tot 3.67 3.83 3.69 f 4.400** 3.016* 3.885** sig 0.002 0.019 0.005 high customisation types. as a consequence, the rate of increasing ltl cost due to disaggregated shipping is likely to decrease as product variety increases. in addition, high customisation types often require the delivery of products directly to customers, imposing delivery cost on the end customer, which could feasibly reduce the overall cost associated with transportation. purchased parts, finished goods and work-in-process inventory such as semi-finished parts, exhibited the highest increases in cost with the low-level customisation types. 2.7. study 2 conclusions the impact of product variety on the performance of five business functions (engineering, manufacturing, purchasing, logistics and marketing) was examined through a study of 163 manufacturing plants. each plant was classified as one of five customisation types: pure standardization (ps), segmented standardisation (ss), customised standardisation (cs), tailored customisation (tc) or pure customisation (pc) which provided a continuum across which performance trends could be assessed. the relationships between business function performance, degree of customisation and the level of product variety offered were also researched. an increase in product variety was found to influence business functions differently depending on the existing combination of the degree of customisation and the level of product variety offered. overall, the marketing function was found to be impacted the most by an increase in product variety, followed by the engineering, manufacturing, purchasing and logistics function. the research also revealed that an increase in product variety in low customisation types increases customer satisfaction, market share and competitive advantage more than in high customisation types. however, product variety increases in low customisation types also impose higher costs than high customisation types. furthermore, product variety increases in low customisation types were found to lead to a higher take-up of variety control strategies (for example, the use of standardised parts, postponement, and product flexibility) than in high customisation types. also, the prevailing degree of customisation was found to be a more significant factor than the existing level of product variety for determining the impact of a variety increase on a number of key functional attributes including manufacturing cost, material cost, transportation cost, manufacturing complexity, manufacturing lead time and demand forecast uncertainty. 3. conclusions in this research, two empirical studies were presented in order to provide a consolidated piece of work that addressed the impact of product and service variety on business and business function performance. in the first (service-variety) study, the relationship between service provision offered by uk-based, thirdparty logistics (3pl) providers and the operational and financial performance of those providers was analysed. the study found that the range of service variety offered by 3pls does not directly influence the 3pls’ financial performance. this study makes a significant contribution to the prevailing knowledge of both logistics and variety management by providing an approach that links service variety with the operational and financial performance of 3pls. the study findings have implications for practice and research. a limitation of this study is that the results can only be generalised to large manufacturers and the study itself was carried out a single point in time. the second (product-variety) study presented the findings from an analysis of data from 163 manufacturing plants where the impact of product variety on the performance of five business functions was examined. an increase in product variety was found to influence business functions differently depending on the combination of customisation and variety offered to customers. marketing performance was found to be most dramatically affected by an increase in product variety, followed by engineering, manufacturing, purchasing and logistics performance. a key limitation of this study is that the research focused on the main customisation type of each manufacturing plant. however, mixed rather than single customisation types commonly occur. the implications, trade-offs and synergies associated with such multiple scenarios have not been considered. an appropriate topic for future variety-management research concerns the examination of how manufacturers can optimise the provision of multiple products with different decoupling points, different levels of variety and different degrees of customisation. also, future research could concern qualitative case studies to understand the development of service capabilities and operational performance for 3pls and how structural equation modelling (sem) could be used to understand if there are any cause and effect relationships between service dimensions and performance. lessons from empirical studies in product and service variety management 61int. j. prod. manag. eng. (2013) 1(1), 55-62creative commons attribution-noncommercial 3.0 spain acknowledgements: this author would like to thank chiung-lin liu and juneho um from the university of liverpool with their help in supporting this research. references ellinger, a.e., ellinger, a.d., & keller, s.b. (2002). logistics managers’ learning environments and firm performance. journal of business logistics, 23(1), 19-37. doi:j.2158-1592.2002.tb00014.x fawcett, s.e., & smith, s.r. (1995). logistics measurement and performance for united-states-mexican operations under nafta. transportation journal, 34(3), 25-34. fisher, m.l. & ittner, c.d. (1999). the impact of product variety on automobile assembly operations: empirical evidence and simulation analysis. management science, 45(6), 771-786. http://dx.doi.org/10.1287/mnsc.45.6.771 hair, j.f., black, b., babin, b., anderson, r.e., & tatham, r.l. (2006). multivariate data analysis. 6ed. upper saddle river, nj: pearson prentice hall. hu, s.j., ko, j., weyand, l., elmaraghy, h.a., lien, t.k., koren, y., bley, h., chryssolouris, g., nasr, n., & shpitalni, m. (2011). assembly system design and operations for product variety. cirp annals manufacturing technology, 60(2), 715-733. doi:10.1016/j.cirp.2011.05.004 lai, k.h. (2004). service capability and performance of logistics service providers. transportation research part e: logistics and transportation review, 40(5), 385-399. doi:10.1016/j.tre.2004.01.002 liu, c.l. & lyons, a.c. (2011). an analysis of third-party logistics performance and service provision. transportation research part e: logistics and transportation review, 47(4), 547-570. doi:10.1016/j.tre.2010.11.012 meyr, h. (2004). supply chain planning in the german automotive industry. or spectrum, 26(4), 447-470. doi:10.1007/s00291-004-0168-4 murphy, p.r., & poist, r.f. (2000). third-party logistics: some user versus provider perspectives. journal of business logistics, 21(1), 121-133. stefansson, g. (2006). collaborative logistics management and the role of third-party service providers. international journal of physical distribution & logistics management, 36, 76-92. doi:10.1108/09600030610656413 ramdas, k. (2003). managing product variety: an integrative review and research directions. production and operations management, 12(1), 79-101. doi: 10.1111/j.1937-5956.2003.tb00199.x randall, t. & ulrich, k. (2001). product variety, supply chain structure, and firm performance: analysis of the u. s. bicycle industry. management science, 47(12), 1588-1604. doi:10.1287/mnsc.47.12.1588.10237 thonemann, u.w. & bradley, j.r. (2002). the effect of product variety on supply-chain performance. european journal of operational research, 143(3), 548-569. doi:10.1016/s0377-2217(01)00343-5 ward, j.h. (1963). hierarchical grouping to optimize an objective function. journal of the american statistical association, 58(301), 236244. doi:10.1080/01621459.1963.10500845 whang, s. & lee, h. (1998). value of postponement in product variety management: research advances, boston: kluwer academic publishers. yeh, k.h. & chu, c.h. (1991). adaptive strategies for coping with product variety decisions. international journal of operations & production management, 11(8), 35-47. lyons, a. c. 62 int. j. prod. manag. eng. (2013) 1(1), 55-62 creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.1002/j.2158-1592.2002.tb00014.x http://dx.doi.org/10.1287/mnsc.45.6.771 http://dx.doi.org/10.1016/j.cirp.2011.05.004 http://dx.doi.org/10.1016/j.tre.2004.01.002 http://dx.doi.org/10.1016/j.tre.2010.11.012 http://dx.doi.org/10.1007/s00291-004-0168-4 http://dx.doi.org/10.1108/09600030610656413 http://dx.doi.org/10.1111/j.1937-5956.2003.tb00199.x http://dx.doi.org/10.1287/mnsc.47.12.1588.10237 http://dx.doi.org/10.1016/s0377-2217(01)00343-5 http://dx.doi.org/10.1080/01621459.1963.10500845 pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.1857 received 2013-11-09 accepted 2013-12-02 a non parametric estimation of service level in a discrete context. cardós, m.i, babiloni, e.ii, estellés, s.iii and guijarro, e.iv dpto. de organización de empresas, universidad politécnica de valencia, camino de vera s/n, 46022 valencia, spain. i mcardos@doe.upv.es ii mabagri@doe.upv.es iii soesmi@omp.upv.es iv esguitar@doe.upv.es abstract: an exact method for the estimation of the cycle service level has been proposed for periodic review stock policies in a discrete demand context for any known i.i.d. demand distribution. however, the implementation of this method in real environments has previously to manage some important and eventually cumbersome issues such as: (i) the identification of the appropriate demand distribution and its validation; (ii) the estimation of the parameters of the demand distribution; and (iii) the calculation of temporal aggregates of the demand distribution in order to estimate the expected service level. this paper shows some difficulties linked to these issues and proposes an alternative approach based on the observed demand frequencies, so that these issues are avoided and the accuracy of the service level estimation seems to be improved. key words: periodic review, service level, demand distribution. 1. introduction probably the most important and useful problem studied by inventory control is the selection of the stock policy and the estimation of its parameters. for example, in the periodic review (r, s) policy, the review period r is usually predetermined by factors like the transportation schedule, so in practice managers see this problem as the determination of the optimal base stock level s such that total costs are minimized or some target customer service level is fulfilled. estimating the parameters of the stock policy subject to a target service constraint is by far the most frequent approach in practice. this paper focuses on the cycle service level (csl) and periodic review policy but the approach proposed in this paper also applies even if an alternative service metric or continuous review policy is selected. finally, the demand distribution is assumed to be i.i.d. and discrete and sample demand data is available. (cardós et al. 2009) propose a comprehensive set of procedures for the exact calculation of csl with backlog and lost sales, not only for the periodic review policy but also for the continuous review one and also approximate expressions in every case. these expressions apply for any i.i.d. demand distribution being discrete and known. for the sake of simplicity, this paper focuses on the case in which backlog is allowed and whose exact estimation of the csl, according to (cardós, babiloni, palmer, & albarracín j.m. 2009) can be obtained by the expression ( ) (0) 1 (0) l r r f s f c sl f = (1) being f(·) the cumulative distribution function of demand, s the base stock, r the cycle and l the lead time. the application of that formula, or the appropriate one in different circumstances such as for example in a demand lost context or when a base stock policy is used, is just the last step of the procedure to compute the service level using sample data, as shown in figure 1. usually it is assumed in the literature that the demand distribution is known, but in practice it is not the case so that we have to cope with the first steps of the estimation procedure. 47int. j. prod. manag. eng. (2014) 2(1), 47-52creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2014.1857 mailto:mabagri@doe.upv.es figure 1. procedure steps to estimate the service level from sample data. the purpose of this paper is twofold: (i) to gain insight into the practical and technical difficulties of estimating the service level from demand sample data and its effect on the accuracy of the service level; and (ii) to propose an alternative non parametric approach so that these issues are avoided. the rest of this paper is organized as follows. section 2 presents the most important practical issues related with the estimation of the service level considering: (i) demand distribution selection and validation using sample data even when they are scarce, (ii) alternatives for the estimation of the demand distribution parameters, and (iii) calculation of the temporal aggregates of the demand. section 3 is devoted to introduce our proposed non parametric approach and provide some illustrative examples. finally, conclusions and further research are presented in section 4. 2. steps to estimate the service level 2.1. demand distribution selection and validation the demand distribution pattern has to be modelled from the available demand data. continuous distributions are very used for modelling the demand pattern (dunsmuir and snyder 1989), (schultz 1987), (yeh et al. 1997) and normal distribution is especially frequent even for discrete demand. although normal distributions may provide acceptable results even in the discrete demand case depending on its characteristics (average, variance, etc) it is more accurate modelling the discrete demand with a discrete distribution (janssen et al. 1996), (strijbosch et al. 2000), (vereecke and verstraeten 1994). poisson distribution is recommended by (silver et al. 1998) for slow moving class a items. compound poisson distribution is also used when the probability of zero demand is significant but (strijbosch, heuts, & van der schoot 2000) explain that this compound distribution is the result of modelling the number of received orders as poisson and also the size of the orders, but this is often an unrealistic starting point table 1. data of three class a items from an spare parts system. item ranked average variance r p pearson's test a1 1 5,9550 11,9221 5,9429 0,4995 pass a2 13 0,2415 4,6251 0,0522 0,0133 fail a3 47 0,0549 0,3069 0,1789 0,0120 fail because data demand is usually aggregated on a daily basis becoming into a compound bernouilli distribution. negative binomial distribution can be used as an alternative to the poisson distribution especially when the sample variance exceeds the sample mean. not surprisingly (syntetos and boylan 2006) point out that the negative binomial distribution is able to model demand patterns belonging to every demand category (smooth, erratic, intermittent and lumpy). first of all, from a practical point of view, the most suitable distributions are poisson and negative binomial. usually the demand pattern is selected as poisson when the variance differs from the average in no more than 10 per cent; if not so a negative binomial distribution is preferred. bernouilli compound distributions are rarely used because of the kind of difficulties explained below. in order to illustrate the main difficulties related with the distribution function selection, we consider the daily demands of three items of class a from the spare parts of an airline company during 911 consecutive days (see table 1). there are 941 items and selected ones are ranked 1, 13 and 47 respectively considering the number of units demanded during the period. obviously items a1, a2 and a3 belong to class a. in these three cases variance is much higher than the average, so following the usual rule a negative binomial distribution is selected and its parameters r and p are estimated. last column shows the results of the pearson’s chi-squared test. 48 int. j. prod. manag. eng. (2014) 2(1), 47-52 creative commons attribution-noncommercial 3.0 spain cardós, m., babiloni, e., estellés, s. and guijarro, e. http://en.wikipedia.org/wiki/variance http://en.wikipedia.org/wiki/mean demand histogram of item a1 seems to fit to a negative binomial distribution (see figure 1) confirmed by pearson’s chi-squared test. demand histograms of items a2 and a3 do not seem to fit a negative binomial distribution (see figures 2 and 3) and obviously pearson’s chi-squared test fails in both cases. unfortunately usually the validity of the demand distribution is not checked because: (i) the application of this test is quite cumbersome and impractical for large inventories; (ii) poisson and negative binomial are usually the only available options; and (iii) there is not a manageably distribution function able to fit demand patterns with so many peaks. it could be argued that these peaks could be anomalies in the demand but this is not the case for items a2 and a3. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 data 18 53 81 95 88 105 107 92 76 60 42 37 21 11 8 8 3 3 0 1 0 2 0 0 0 nb fitted 15 44 76 101 113 112 102 88 71 55 41 30 21 15 10 7 4 3 2 1 1 0 0 0 0 0 20 40 60 80 100 120 fr eq ue nc y item a1 figure 2. demand histogram for item a1 and negative binomial expected frequencies. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 data 888 5 4 0 1 0 3 1 1 0 0 0 2 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 4 0 0 nb fitted 727 37 19 13 10 8 7 6 5 4 4 3 3 3 3 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 0 5 10 15 20 25 30 35 40 45 50 fr eq ue nc y item a2 figure 3. demand histogram for item a2 and negative binomial expected frequencies. 49int. j. prod. manag. eng. (2014) 2(1), 47-52creative commons attribution-noncommercial 3.0 spain a non parametric estimation of service level in a discrete context. additionally, the validation of a demand distribution using a statistical test requires a number of non zero demand periods but it is not always possible as it happens with a3 item even being a class a item. 2.2. estimation of parameters second, the estimation of the parameters of a demand distribution may be obtained using maximum likelihood estimators which estimate the parameters in order to make the observed data the most probable. these estimators have a number of desirable properties, but in some cases the estimators are unsuitable or do not exist. for example, the estimation of the parameters of a bernouilli compound distribution requires the use of a computer to solve simultaneously two equations (2) being p the probability of zero demand, λ the poisson rate, n the size of the sample, xi the demand data and m the number of non zero demands. this situation also applies to the negative binomial distribution. another estimation approach is the method of moments which uses as many moments as parameters have to be estimated, replaces the moments by the sample moments and derives the expressions of the parameters. for example, for the negative binomial distribution with parameters r and p (3) being x̂ and 2ŝ the sample moments of first and second order. maximum likelihood estimators tend to offer better estimations than the moments method, but the estimators based on moments can be quickly and easily calculated. 2.3. obtaining temporal aggregates of the demand first of all, it should be noted that the probability of no demand during r consecutive periods is needed in expression (1) and can be calculated directly whatever the demand distribution would be as (4) once the demand distribution has been selected and its parameters have been estimated, the last step before applying expression (1) is to develop temporal aggregates of the demand. for example, the cumulative distribution in l consecutive periods fl(s) is the kind of temporal aggregate needed to compute 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 data 901 1 1 0 0 1 7 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 nb fitted 413 73 42 30 24 20 17 15 13 12 11 10 9 8 8 7 7 6 6 6 5 5 5 5 4 4 4 4 4 4 3 0 5 10 15 20 25 30 35 40 45 50 fr eq ue nc y item a3 figure 4. demand histogram for item a3 and negative binomial expected frequencies. 50 int. j. prod. manag. eng. (2014) 2(1), 47-52 creative commons attribution-noncommercial 3.0 spain cardós, m., babiloni, e., estellés, s. and guijarro, e. csl and it can be obtained using the properties of the sum of i.i.d. distribution functions. we only need to develop an expression for fl(s) in a convenient form to be used. this is quite straightforward for poisson and negative binomial distributions since both maintain their own distribution and (5) however it is not so easy for many other distributions such as bernouilli compound that becomes into a binomial compound distribution and its aggregates involve quite complex and long calculations. anyway, fl(s) can always be obtained based on f(.) using the convolution of two discrete distributions (6) 3. proposed non parametric approach the difficulties explained below appear when demand data do not fit a convenient distribution function such as poisson or negative binomial. these problems are difficult to manage when they occur, but there are also interactions among them making it harder. for example, if you improve the fitting of the distribution function using a complex compound distribution, then it leads to an impractical analytical expression for the demand during the lead time. we propose a different approach by defining the demand distribution as (7) being fri the sample relative frequency. this formulation avoids the need of identifying and validating the demand distribution and has no parameters to be estimated. the performance of this approach can be illustrated with an example where we know the demand distribution and we compare the performance of parametric and non parametric approaches. first, the demand is the sum of a poisson distribution with λ=0.01 and three times a bernouilli distribution with p=0.1 being p the probability of non zero demand. second, we compute the exact csl using expressions (1), (4) and (6). third, demand is simulated 30 times for 1,000 consecutive days and parametric and non parametric estimations of csl are obtained each time using the generated sample data. finally the average csl estimation for each base stock and procedure is obtained (see figure 5) resulting in better estimates from the non parametric one. 1 2 3 4 5 6 7 8 9 10 exact 0,566 0,567 0,969 0,977 0,977 1,000 1,000 1,000 1,000 1,000 non parametric 0,565 0,565 0,969 0,975 0,975 1,000 1,000 1,000 1,000 1,000 negative binomial 0,749 0,870 0,929 0,961 0,978 0,987 0,993 0,996 0,998 0,999 0,500 0,550 0,600 0,650 0,700 0,750 0,800 0,850 0,900 0,950 1,000 cs l csl vs. s figure 5. illustrative example comparing the parametric and non parametric procedures for different base stocks (s=1..10) 51int. j. prod. manag. eng. (2014) 2(1), 47-52creative commons attribution-noncommercial 3.0 spain a non parametric estimation of service level in a discrete context. 4. conclusions and practical implications although exact estimation procedures are available for computing csl and other service level metrics in discrete contexts, its estimation remains being a challenge due to the practical difficulties involved in identifying and validating the demand distribution, estimating the parameters of the distribution and developing temporal aggregates. when demand patterns can be modelled using poisson or negative binomial distributions, then the estimation of the service level is quite simple because of their distribution properties. other discrete distributions like bernouilli compound may be an interesting alternative when the probability of no demand is quite high, but the calculation of temporal aggregates become much more complex. anyway, the validation of the demand distribution is always very time-consuming so that usually it is not performed extensively. as a consequence, we propose a non parametric approach that outperforms the conventional parametric estimation: (i) it is not necessary to estimate the parameters of the distribution or to validate the distribution pattern; and (ii) the estimation of the service level seems to outperform the parametric one. given the impact and practical implications of these results in operational management, further research will focus on developing an extensive experiment in order to check the influence of sample size and other factors including the demand characteristics. acknowledgements: this research was part of a project supported by the universitat politècnica de valència, with reference number paid-06-11/2022. references cardós, m., babiloni, e., palmer, m.e., albarracín, j.m. (2009). effects on undershoots and lost sales on the cycle service level for periodic and continuous review policies. ieee transactions on information forensics and security, 819-824. doi:10.1109/iccie.2009.5223854 dunsmuir, w.t.m. snyder, r.n. (1989). control of inventories with intermittent demand. european journal of operational research, 40(1), 16-21. available from: isi:a1989u671400002. doi:10.1016/0377-2217(89)90267-1 janssen, f., heuts, r., kok, t. (1996). on the (r,s,q) inventory model when demand is modelled as a compound bernoulli process. european journal of operational research, 104(3), 423-436. doi:10.1016/s0377-2217(97)00009-x schultz, c.r. (1987). forecasting and inventory control for sporadic demand under periodic review. journal of the operational research society, 38(5), 453-458. available from: isi:a1987h521100009 silver, e.a., pyke, d.f., peterson, r. (1998). inventory management and production planning and scheduling, third edition. john wiley & sons (ed.). new york, usa. strijbosch, l.w.g., heuts, r.m.j., van der schoot, e.h.m. (2000). a combined forecast inventory control procedure for spare parts. journal of the operational research society, 51(10), 1184-1192 available from: isi:000089702100010. doi:10.1057/palgrave.jors.2601013 syntetos, a.a. boylan, j.e. (2006). on the stock control performance of intermittent demand estimators. international journal of production economics, 103(1), 36-47. available from: isi:000238785300005. doi:10.1016/j.ijpe.2005.04.004 vereecke, a. verstraeten, p. (1994). an inventory management model for an inventory consisting of lumpy items, slow movers and fast movers. international journal of production economics, 35(1-3), 379-389. available from: isi:a1994nz26500050. doi:10.1016/09255273(94)90106-6 yeh, q.j., chang, t.p., chang, h.c. (1997). an inventory control model with gamma distribution. microelectronics and reliability, 37(8), 11971201. available from: isi:a1997xf22800008. doi:10.1016/s0026-2714(96)00295-8 52 int. j. prod. manag. eng. (2014) 2(1), 47-52 creative commons attribution-noncommercial 3.0 spain cardós, m., babiloni, e., estellés, s. and guijarro, e. http://dx.doi.org/10.1109/iccie.2009.5223854 http://dx.doi.org/10.1016/0377-2217(89)90267-1 http://dx.doi.org/10.1016/s0377-2217(97)00009-x http://dx.doi.org/10.1057/palgrave.jors.2601013 http://dx.doi.org/10.1016/j.ijpe.2005.04.004 http://dx.doi.org/10.1016/0925-5273(94)90106-6 http://dx.doi.org/10.1016/0925-5273(94)90106-6 http://dx.doi.org/10.1016/s0026-2714(96)00295-8 pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering received: 2021-11-22 accepted: 2022-06-26 preparing for future e-waste from photovoltaic modules: a circular economy approach david hidalgo-carvajal a1*, ruth carrasco-gallego a2 a department of organization engineering, business administration and statistics, escuela técnica superior de ingenieros industriales (etsii), universidad politécnica de madrid (upm), c/ josé gutiérrez abascal 2, 28006 madrid, spain. a1* david.hidalgo.carvajal@upm.es, a2 ruth.carrasco@upm.es abstract: the increasing adoption rate of photovoltaic power generation shows that renewable energies have a bright future. yet, this could be overshadowed by the unintended consequence of increased generation of waste of electric and electronic equipment (weee) at the installations’ end-of-life (eol) stage. as countries find themselves dealing with the increasing weee issue, they may adopt different practices which, if wrongly implemented, could potentially backfire, creating additional issues especially among vulnerable social groups. this work proposes improving the weee management system by including the informal recyclers in the equation, benefitting social groups and material recovery through by delivering materials along different streams in the closed-loop supply chain. the proposed model intends to support the circular economy approach on waste management systems. key words: recycling, weee management model, informal recyclers, photovoltaic panels. 1. introduction through the past decades, world’s population has rapidly increased and, if current trends continue, is predicted to reach 9.7 billion by 2050, and peak at 11 billion by 2100 (united nations, 2019). with a global trend of ageing population, an aggravated food waste problem, surging demands in energy consumption, as well as an increasing demand of new products from customers’ side (lofthouse & prendeville, 2018), is evident that our current traditional linear manufacturing model calls for action and urgent change. the circular economy (ce) business model seems to provide a much needed answer to this change by implementing different approaches, such as the 10r strategies (reike et al., 2018). the concept of ce embodies an alternate method aiming to alter the linear consumption pattern (i.e., take-make-dispose) and make it more sustainable, creating “a production-distribution-consumption model that is regenerative and restorative by design” (hidalgo-carvajal et al., 2021). this concept has gained relevant traction recently among scholars and practitioners, evolving around different disciplines and approaches. however, as the concept is transversal and common to different disciplines, and each of them has its own version of the definition (van loon et al., 2021), this creates a discrepancy between the widespread consensus received by the objectives and means of the ce and the difficulty of defining what it is (korhonen et al., 2018). additionally, up to date, most of the advanced analyses from academics and practitioners has been mainly focused on two of the dimensions: environmental and economic. moreover, the social aspects of the ce have been largely ignored, sometimes only being considered as to cite this article: hidalgo-carvajal, d., carrasco-gallego, r. (2022). preparing for future e-waste from photovoltaic modules: a circular economy approach. international journal of production management and engineering, 10(2), 131-141. https://doi.org/10.4995/ijpme.2022.16712 https://doi.org/10.4995/ijpme.2022.16712 int. j. prod. manag. eng. (2022) 10(2), 131-141creative commons attribution-noncommercial-noderivatives 4.0 international 131 https://orcid.org/0000-0001-7771-0906 https://orcid.org/0000-0002-9542-3836 mailto:david.hidalgo.carvajal@upm.es mailto:ruth.carrasco@upm.es http://creativecommons.org/licenses/by-nc-nd/4.0/ side-effects, “peripherally and sporadically integrated into the circular economy concept” (geissdoerfer et al., 2017), even when these have been mentioned by organizations as part of their policies (bubicz et al., 2021). as can be seen, the benefits related to an appropriate implementation of the ce framework will help countries and governments “to meet the objectives of the 2030 agenda for sustainable development” (european commission, 2018) from all different perspectives. world’s current linear manufacturing model continues to overlook all dimensions of sustainability (chuang & lin, 2015) along the whole supply chain. among the gradually rising manufacturing of new products and the consequent resulting waste after their endof-life (eol), electric and electronic equipment (eee) take a key place. as highlighted by guzzo et al. (2021), eee “plays a central role in society as it facilitates day-to-day tasks, improves living conditions and work environments, and facilitates communication”. furthermore, obsolescence plays a fundamental part in the amount of waste (weee) generated per year, with approximately 53.6 million tons generated during last year (forti et al., 2020), consisting of both, the absolute obsolescence and the relative obsolescence (shittu et al., 2021). amongst the rapidly increasing manufacturing of eee, one stands out due to its potential: photovoltaic (pv) modules, with all its components. as an important fact, there were more solar installations in 2019 than fossil fuel and nuclear power additions combined, which occurred for the fifth year in a row (ren21, 2020), following a trend led by major greenhouse gas emitters like china, the united states, and india (harrington, 2017). this has been also supported by the decreasing cost of solar pv which has globally plummeted by 99 percent in over the last decades, reaching a point of cost competitiveness with conventional generation technologies (fossil fuels), even without government subsidies (lazard, 2019). yet, an unexpected outcome of having access to cheaper technology, will be an increase in the waste generation from pv installations at their endof-life stage in the near future. the international renewable energy agency (irena) predicts that between 1.7 and 8 million tons of pv waste will be generated in 2030 and between 60 and 78 million tons in 2050 (irena & ieapvps, 2016). the later implies that pv modules’ waste could exceed by 10% the total amount of electronic waste generated by other devices. this could be translated in an increase in the economic value associated with the recoverable raw materials through recycling, with a rough estimate of usd 450 million in 2030 and usd 15 billion in 2050 (irena & iea-pvps, 2016). with proper weee management many benefits can be obtained, ranging from avoiding wrong chemical disposal of elements and substances dangerous to human health and environment, to recovering tons of materials that can be reused or reintroduced in the manufacturing processes. according to the most recent data, merely 17.4% of the total weee is correctly recycled (forti et al., 2020), whereas the rest finds a secondary market in less-developed countries (wang et al., 2016), or is “misplaced” either on their way to the recycling facilities or during the sorting and recycling process (bigum et al., 2017; pekarkova et al., 2021). one of the strongly recommended suggestions to improve waste management (wm) is to include “the long tradition and experience available in the informal sector” (agamuthu, 2010) and complement it with the regulations and working conditions (hygiene, safety and fair payment) provided by a formal sector (asim et al., 2012). furthermore, incorporating the informal recycling sector (irs) into wm proves to serve as a facilitator for ce implementation, by including the often forgotten aspects of the social dimension in the equation (awasthi et al., 2019; mies & gold, 2021), and lastly, supporting accomplishing the sustainable development goals (sdgs) (valencia, 2019). as the informal systems continue to exist and thrive side-by-side with the formal system (oteng-ababio et al., 2013), it becomes necessary to propose a model that combines them both and generates value through the ce approach. thus, the goal of this study is to answer the following question: how do the informal recycling sector can be integrated in the waste management systems to support achieving the sdgs? the main objective of this document is to draw a picture on the potential benefits in waste reduction derived from including irs along weee management systems of pv modules. through the review of academic and grey literature, diverse impacts (good and bad) from implementing different practices in weee management were identified. also, areas where irs could be included along the reverse supply chain process for weee have been suggested. int. j. prod. manag. eng. (2022) 10(2), 131-141 creative commons attribution-noncommercial-noderivatives 4.0 international hidalgo-carvajal & carrasco-gallego 132 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. methods to complete this exploration, the authors followed a qualitative research methodology, suggested by multivocal literature reviews (yasin & hasnain, 2012), consisting of reviewing academic and grey literature, to identify a wide range of relevant documents to the discussion. in the academic area, two main databases were used given their larger subject and pertinent journal range: scopus and web of science. additionally, to complement the search for grey literature, google scholar was used. for this purpose, the research considered academic documents in the period of 1992-2021, using the following search strings in the literature review: (“weee” or “photovoltaic module” or “pv panel”) and (“recycl*” or “circular economy” or “closed loop supply chain”), retrieving a total of 1,915 unique academic articles, from which 36 were included in the final sample. the research design and selection process can be seen in figure 1. it was found that most academic documents were peer-reviewed articles (60%), followed by conference papers (23%), literature reviews (8%) and book chapters (7%). and the publications can be allocated to the following subjects: environmental sciences, engineering, energy, material sciences, business management, and other areas. figure 1. research design. 3. results although the hypothesis is that as the weee continues to growth, its management improves along with it; however, through the literature review, was possible to identify that this is not always the case. additionally, considering that waste from pv modules will increase in the future, is necessary to complement the current pv module recycling system, making sure that weee management practices are properly adopted. 3.1. pv modules recycling after completing the literature review on pv recycling processes, we identified that the module availability on today’s market is split in two main categories: silicon-based panel (c-si) and nonsilicon based materials such as cadmium telluride (cdte), and copper indium gallium (di) selenide (cigs)) (chowdhury et al., 2020). presently, most of the recycling efforts are focused on c-si pv type (chowdhury et al., 2020), counting with three specific recycling types: mechanical, thermal and chemicals (kang et al., 2012). a special case, which is still under discussion, are the solar thermal modules, which have a different material composition than pvs; however, for this study, only pv modules have been considered. additionally, as mentioned by besiou & van wassenhove (2016), given the long life cycle of the pv waste and the existence of precious/rare materials (silver, indium, gallium and germanium) in it, a new approach to properly manage this waste becomes necessary. moreover, with a growing interest around pv module’ recycling worldwide, steps need to be taken. right on this track, europe has carried most of the development on this area. companies and research institutions have worked together into more efficient ways to develop pv from recycled materials (solar world ag, first solar, pv cycle), innovating recycling processes (saperatec, fraunhofer ise, recyclia, screlec), increasing the life cycle of components (auo) and generating collaborative knowledge (iea-pvps). in addition, legislation has been developed at regional level in the european union (weee directive), which helps individual countries to adapt it into their own legislations (sousa et al., 2018). data cr iter ia identification d atabase selection: wos & scopus t ime inter val: 1992—2021 doc types: academic papers, written in english, peer reviewed sear ch: • step 1 — i dentification – using the search string – all terms combined ("pv panel" or "photovoltaic module" or "weee") and ("circular economy" or "recycl* ” or “closed loop supply chain” ) – 2,847 documents • step 2 — practical screening removal of duplicates from both databases – 1,915 documents • step 3 — quality screening – removal of papers non relevant to the research objective – 146 documents • step 4 — i ncluded – final selection – 39 documents int. j. prod. manag. eng. (2022) 10(2), 131-141creative commons attribution-noncommercial-noderivatives 4.0 international preparing for future e-waste from photovoltaic modules: a circular economy approach 133 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.2. weee practices on one hand, cases on which companies carried socially and environmentally responsible practices were clustered in “good case practices”. some companies and countries have created e-waste policies and proposed weee management strategies (delgado et al., 2006), while others have focused on creating social enterprises and services around weee (lixandru et al., 2017), and a few have invested in education programs for their citizens and research around the issue (awasthi et al., 2019). while most of the initiatives focus on dealing with the weee once it is generated by improving the collection system (cleanspot, relight, c-servees) and the recycling process (c-servees, gatec, inventorization, cfc3-recycling), a key element for successful weee management systems is reducing the generation of waste (matsushida electric, siemens/fujitsu), and generating enterprises focused on sustainable weee recycling (social market economy). good practices have been largely and consistently allocated in developed markets (li et al., 2013: richter & koppejan, 2016; sousa et al., 2018), as these have investing resources, and more detailed management plans to deal with weee (dieste et al., 2017). plenty of good practices have been developed in specific areas, such as education (morris & metternicht, 2016; nikoloudakis & rangoussi, 2019), policy making and implementation (yu et al., 2010; pariatamby & victor, 2013; daum et al., 2017), use of diverse set of technologies (stejskal, 2016; wang & wang, 2019; coughlan & fitzpatrick, 2020), improve sorting practices (picon et al., 2010; barletta et al., 2015), inventorying practices (yumoto & shiratori, 2009), logistics and network design (gamberini et al., 2010; mar-ortiz et al., 2011; kilic et al., 2015; islam & huda, 2018; nowakowski & mrówczyńska, 2018), and creating enterprises with social impact (papaoikonomou et al., 2009; gonzález et al., 2017). in table 1, the identified good practices, per region, can be found. although it is reassuring to see plenty of these initiatives spreading around the world, is important to highlight that only one of these initiatives proposes the inclusion of irs as part of the waste management systems (wms), specifically in the asian region. on the other hand, “bad practices” present unintended consequences, and refer to cases on which practices are preceded by poor management of weee residues, and are not aligned with the sdgs (goodship et al., 2019). evidence of bad experience can be located all around the world; however, it was found that most of these unsafe practices frequently occur on developing countries (ongondo et al., 2011), where a lack of regulation allows for different failures in the system (honda et al., 2016; mihai et al., 2019). poor storage, wrong transportation and sorting practices are some of the key issues that countries currently face on this matter (shinkuma & huong, 2009; de souza et al., 2016; singhal et al., 2019). a summary of the identified bad practices, per region, can be found in table 2. furthermore, two main direct consequences of the poor management remain latent: environmental impact and social impact. among the environmental impacts, key issues are derived from inadequate management of hazardous materials (zhang et al., 2017), open burning (cesaro et al., 2019), improvised metallurgical processing (li et al., 2015), and/or misplaced waste (bigum et al., 2017; pekarkova et al., 2021), creating conditions for further contamination. on the social impact side, the long-term impacts on health (mihai, 2020) and working conditions (lima et al., 2016) of the informal recyclers are the key issue. according to (marke et al., 2020), an easy way to solve both issues is through improving regulations on the topic. nevertheless, regulating informal recyclers is the most difficult part given the essential role they play on the delicate waste management ecosystem (williams, 2016). although the weee generation continues to increase yearly, the amount of bad practices seems to increase with it (althaf et al., 2019) as there exist either insufficient control or a lack of it over the waste flows (rochman et al., 2017). table 1. identified good practices per region. location practice europe asia latin america africa arab states region automated processing x x e-waste policy x x transportation of weee x weee management x x x x x ict tools x x social enterprises x x consulting services x inventorying x x education x x x x int. j. prod. manag. eng. (2022) 10(2), 131-141 creative commons attribution-noncommercial-noderivatives 4.0 international hidalgo-carvajal & carrasco-gallego 134 http://creativecommons.org/licenses/by-nc-nd/4.0/ additionally, on the other extreme of the supply chain, extraction of raw materials also identifies several bad practices, mostly related to the social impacts, largely on working conditions (labor exploitation, numerous accidents and deaths, air poorly breathable and stiffing heat), hygiene problems and diseases, including cancer (santillánsaldivar et al., 2021). therefore, the “bad practices” issue calls for attention from the eee manufacturers and all parties along the entire supply chain. 3.3. the irs in the weee management based on the examples of good and bad practices, the role and importance of the informal sector on the solid wm, especially for weee, becomes critical, considering also that these need to be positioned and aligned with ce principles (ranjbari et al., 2021). inclusion of the informal recyclers in the system provides significant economic benefits to those working under these systems, despite the related health and social issues (wilson et al., 2006). moreover, irs is heavily composed by social groups which are mostly poor and marginalized. despite this fact, the informal sector is usually considered by governments as a sector that needs to be either eradicated or formalized, as they are unregulated, unlicensed and, in addition, they do not pay taxes (andrianisa et al., 2016). under this assumption the authors propose including the irs in the pv circular supply chain cycle, taking advantage of their extended experience as waste precollectors (wpc) and waste segregation experts. considering the fact that irs is a common practice around the globe, frequently in areas of emerging markets (chi et al., 2011; umair et al., 2016; hai et al., 2017; ghisolfi et al., 2017; echegaray & hansstein, 2017; parajuly et al., 2018; asibey et al., 2020), the impact that including irs in the waste table 2. identified bad practices per region. location practice europe asia latin america africa unregulated system x x x x informal sector x x x open burning x x poor storage system x x x poor transportation x x poor sorting x x unsafe handling x nonexistent legislation x "misplaced" weee x stolen materials x figure 2. including the informal recycling sector in the circular economy supply chain model (own development). raw material manufacture distribution use and maintenance disposal weee classification pv waste separation component sorting raw material for industry i1 raw material for industry i2 raw material for industry in formal recycling industry maintenance for industry i1 maintenance for industry i2 maintenance for industry in int. j. prod. manag. eng. (2022) 10(2), 131-141creative commons attribution-noncommercial-noderivatives 4.0 international preparing for future e-waste from photovoltaic modules: a circular economy approach 135 http://creativecommons.org/licenses/by-nc-nd/4.0/ management system could have is vast and positive for these populations and economic sectors. the model presented in figure 2 proposes a wms for weee on which irs becomes a “visible” part of the system. the traditional linear supply chain is represented in blue, showing the extraction to waste model. when the waste is properly disposed, formal recycling industry (fri), in orange, receives it and transform it into materials that can be used for maintenance (as spare parts) or recycled to be sold as raw materials for the same industry or to another industry (i1 … in), as can be seen in grey. however, due to mismanagement, not all weee is properly collected by the fri, and it ends up in the hands of the irs upon being disposed in collection points or landfills. as shown in green, weee classification by irs generates a stream of material for reuse (own usage and second-hand market), sale to repair shops (within the same or another industry) and supply of materials to fri. additionally, by including both, the fri, and the irs in the weee management, as the recovered number of critical materials increases, it would reduce drastically the need for extraction of critical raw materials (crm), which are key for the industrial value chain of different sectors such as new technologies and medical sector, as well as for energy and environment sectors (for renewable energy technologies), in this specific case, the pv modules. as mentioned by santillán-saldivar et al. (2021), maximizing the recovering and recycling of materials would also mean reinserting these materials into the domestic economy, having a longlasting impact. the proposed model complements a gap in the research where only improvements to the fri have been discussed, by establishing a formal collaboration between the fri and the irs in the weee management. moreover, this model supports the implementation of circular economy in three fronts: first, recovering a greater amount of materials directly impacts the economic part by reducing the cost of production as the need for virgin raw materials is reduced. second, given the reduction of need for raw materials and the avoidance of inadequate disposal of weee in landfills, it directly reduces the environmental impact. finally, a social positive impact is reached by generating job opportunities and social inclusion of the irs in the wm. moreover, this model could be used not only for weee, but could also be considered as a tool to improve current urban wms. this model follows the work proposed by wilson et al. (2006) and valencia (2019), and includes the role of irs into weee management from a circular economy perspective, which would, potentially, benefit many different industries and social groups. 4. main contribution the performed literature review shows a need to tackle the weee issue from different fronts: first, a need for a review of the different weee flows; second, a demand for identification of current weee management practices; third, a need for linking together the informal and formal recycling sectors under the wms. despite the growing amount of different eee (i.e., mobile phones, computers, internet-enabled devices, renewable energy assemblies, among others), there seems to be a lack of preparedness for the future management of this waste once the eee become obsolete and is discarded at its eol. additionally, this work contributes to the literature by providing a model which complements the reduction, reuse and recycle (3r) practices currently in place. furthermore, the proposed model aims to integrate a usually excluded social group that could help solve the growing issue of the weee management by being integrated into the wms. this would have a positive influence in the following key areas: firstly, reducing the environmental impact of this waste; secondly, increasing the economic value of the recovered material and improving the domestic economy; thirdly, and more importantly, providing a beneficial social effect for this social group, increasing their own economic conditions, and directly benefitting their health and working conditions. 5. conclusions waste management systems are confronted with the reality of increasing e-waste. this has been partially driven by the decrease of prices, increase of living standards, and consumption of goods and services. moreover, some technologies have become more affordable, due to continued government support and cumulative innovation. among these, adoption int. j. prod. manag. eng. (2022) 10(2), 131-141 creative commons attribution-noncommercial-noderivatives 4.0 international hidalgo-carvajal & carrasco-gallego 136 http://creativecommons.org/licenses/by-nc-nd/4.0/ of solar technologies leads the trend as they added more installations in 2019 than fossil fuel and nuclear power combined. however, a future unintended effect is an increase in the generation of weee at the eol of those installations. to deal with this future effect, several initiatives have been conducted around the world to either advance in the materials and manufacturing processes or improve current recycling practices. reviewing the experiences with general weee management, it was possible to identify good and bad practices on the topic. unfortunately, bad practices are largely present in developing countries. moreover, these bad practices are related to poor and marginalized social groups, as they rely on waste generation to generate income. this informal sector is usually considered by governments as a sector that needs to be either eradicated or formalized, as they are unregulated, even though they play a key role on waste management. moreover, the irs continues to exist and thrive side-by-side with the fri, which, if both were bound to work together, would improve the waste management through an improved closedloop supply chain. to potentiate the impact that irs could have in the weee management, we propose an improved closed-loop supply chain model which incorporates irs and represents how they are able to supplement fri’s work, and directly contribute to circular economy principles and development of sdgs. as a future path of research, the development of specific indicators that combine the contribution from irs and fri is needed. this invites for a multidisciplinary involvement around the topic through a more systemic view, reviewing the impacts they have on each other. additionally, considering the increased development of solar thermal technologies, research on the proper waste management of these materials is needed. this study is relevant for academics and decision makers working on the topic, as it proposes a supply chain integrating outcast social groups, creating long lasting impacts at economic, social, and environmental levels. acknowledgements: the authors would like to thank the instituto brasileiro de desenvolvimento e sustentabilidade (iabs) for their interest and support on our research. additionally, we would like to thank the projects funding david hidalgo-carvajal’s research: the wedistrict project [founded by the european union’s horizon 2020 research and innovation programme under grant agreement n°857801], the “campus upm circulares” project within the upm research program [programa propio upm 2020. acción estratégica en ciencia y tecnología], and “the circular and regenerative campus” community from the eelisa european university alliance. references agamuthu, p. (2010). the role of informal sector for sustainable waste management. waste management & research, 28(8), 671–672. https://doi.org/10.1177/0734242x10377834 althaf, s., babbitt, c.w., & chen, r. (2019). forecasting electronic waste flows for effective circular economy planning. resources, conservation and recycling, 151, 104362. https://doi.org/10.1016/j.resconrec.2019.05.038 andrianisa, h.a., brou, y.o.k., & séhi bi, a. (2016). role and importance of informal collectors in the municipal waste pre-collection system in abidjan, côte d’ivoire. habitat international, 53(april), 265–273. https://doi.org/10.1016/j. habitatint.2015.11.036 asibey, m.o., lykke, a.m., & king, r.s. (2020). understanding the factors for increased informal electronic waste recycling in kumasi, ghana. international journal of environmental health research, 1–16. https://doi.org/10.1080/ 09603123.2020.1755016 asim, m., batool, s.a., & chaudhry, m.n. (2012). scavengers and their role in the recycling of waste in southwestern lahore. resources, conservation and recycling, 58(january), 152–162. https://doi.org/10.1016/j.resconrec.2011.10.013 awasthi, a.k., li, j., koh, l., & ogunseitan, o.a. (2019). circular economy and electronic waste. nature electronics, 2, 86–89. https://doi.org/10.1038/s41928-019-0225-2 barletta, i., johansson, b., cullbrand, k., bjorkman, m., & reimers, j. (2015). fostering sustainable electronic waste management through intelligent sorting equipment. 2015 ieee international conference on automation science and engineering (case), 459–461. https://doi.org/10.1109/coase.2015.7294122 besiou, m., & van wassenhove, l.n. (2016). closed-loop supply chains for photovoltaic panels. journal of industrial ecology, 20(4), 929–937. https://doi.org/10.1111/jiec.12297 int. j. prod. manag. eng. (2022) 10(2), 131-141creative commons attribution-noncommercial-noderivatives 4.0 international preparing for future e-waste from photovoltaic modules: a circular economy approach 137 https://doi.org/10.1177/0734242x10377834 https://doi.org/10.1016/j.resconrec.2019.05.038 https://doi.org/10.1016/j.habitatint.2015.11.036 https://doi.org/10.1016/j.habitatint.2015.11.036 https://doi.org/10.1080/09603123.2020.1755016 https://doi.org/10.1080/09603123.2020.1755016 https://doi.org/10.1016/j.resconrec.2011.10.013 https://doi.org/10.1038/s41928-019-0225-2 https://doi.org/10.1109/coase.2015.7294122 https://doi.org/10.1111/jiec.12297 http://creativecommons.org/licenses/by-nc-nd/4.0/ bigum, m., damgaard, a., scheutz, c., & christensen, t.h. (2017). environmental impacts and resource losses of incinerating misplaced household special wastes (weee, batteries, ink cartridges and cables). resources, conservation and recycling, 122(july), 251–260. https://doi.org/10.1016/j.resconrec.2017.02.013 bubicz, m.e., dias barbosa-póvoa, a.p.f., & carvalho, a. (2021). social sustainability management in the apparel supply chains. journal of cleaner production, 280, 124214. https://doi.org/10.1016/j.jclepro.2020.124214 cesaro, a., belgiorno, v., gorrasi, g., viscusi, g., vaccari, m., vinti, g., jandric, a., dias, m.i., hursthouse, a., & salhofer, s. (2019). a relative risk assessment of the open burning of weee. environmental science and pollution research, 26(11), 11042–11052. https://doi.org/10.1007/s11356-019-04282-3 chi, x., streicher-porte, m., wang, m.y.l., & reuter, m.a. (2011). informal electronic waste recycling: a sector review with special focus on china. waste management, 31(4), 731–742. https://doi.org/10.1016/j.wasman.2010.11.006 chowdhury, md. s., rahman, k.s., chowdhuruy, t., nuthammachot, n., techato, k., akhtaruzzaman, md., tiong, s.k., sopian, k., & amin, n. (2020). an overview of solar photovoltaic panels’ end-of-life material recycling. energy strategy reviews, 27(january), 100431. https://doi.org/10.1016/j.esr.2019.100431 chuang, s.-h., & lin, h.-n. (2015). co-creating e-service innovations: theory, practice, and impact on firm performance. international journal of information management, 35(3), 277–291. https://doi.org/10.1016/j.ijinfomgt.2015.01.002 coughlan, d., & fitzpatrick, c. (2020). trialling the preparation for reuse of consumer ict weee in ireland. journal of cleaner production, 256, 120512. https://doi.org/10.1016/j.jclepro.2020.120512 daum, k., stoler, j., & grant, r. (2017). toward a more sustainable trajectory for e-waste policy: a review of a decade of e-waste research in accra, ghana. international journal of environmental research and public health, 14(2), 135. https://doi.org/10.3390/ijerph14020135 de souza, r.g., clímaco, j.c.n., sant’anna, a.p., rocha, t.b., do valle, r. de a.b., & quelhas, o.l.g. (2016). sustainability assessment and prioritisation of e-waste management options in brazil. waste management, 57, 46–56. https://doi.org/10.1016/j.wasman.2016.01.034 delgado, l., savage, m., lindblom, j., & institute for prospective technological studies. (2006). implementation of waste electric and electronic equipment directive in the eu 25. publications office. dieste, m., viagi, a.f., panizzolo, r., & biazzo, s. (2017). a comparison of different models for collection of weee in europe. international journal of environmental science and development, 8(8), 591–596. https://doi.org/10.18178/ ijesd.2017.8.8.1021 echegaray, f., & hansstein, f.v. (2017). assessing the intention-behavior gap in electronic waste recycling: the case of brazil. journal of cleaner production, 142, 180–190. https://doi.org/10.1016/j.jclepro.2016.05.064 european commission. (2018). on a monitoring framework for the circular economy (com(2018) 29 final). european commission. https://eur-lex.europa.eu/legal-content/en/txt/?uri=com%3a2018%3a29%3afin forti, v., baldé, c.p., kuehr, r., & bel, g. (2020). the global e-waste monitor 2020: quantities, flows, and the circular economy potential. united nations university (unu)/united nations institute for training and research (unitar) – co-hosted scycle programme, international telecommunication union (itu) & international solid waste association (iswa). gamberini, r., gebennini, e., manzini, r., & ziveri, a. (2010). on the integration of planning and environmental impact assessment for a weee transportation network—a case study. resources, conservation and recycling, 54(11), 937– 951. https://doi.org/10.1016/j.resconrec.2010.02.001 geissdoerfer, m., savaget, p., bocken, n.m.p., & hultink, e.j. (2017). the circular economy – a new sustainability paradigm? journal of cleaner production, 143, 757–768. https://doi.org/10.1016/j.jclepro.2016.12.048 ghisolfi, v., diniz chaves, g. de l., ribeiro siman, r., & xavier, l.h. (2017). system dynamics applied to closed loop supply chains of desktops and laptops in brazil: a perspective for social inclusion of waste pickers. waste management, 60(february), 14–31. https://doi.org/10.1016/j.wasman.2016.12.018 gonzález, x.m., rodríguez, m., & pena-boquete, y. (2017). the social benefits of weee re-use schemes. a cost benefit analysis for pcs in spain. waste management, 64, 202–213. https://doi.org/10.1016/j.wasman.2017.03.009 goodship, v., stevels, a., & huisman, j. (eds.). (2019). waste electrical and electronic equipment (weee) handbook (second edition). woodhead publishing, an imprint of elsevier. guzzo, d., rodrigues, v.p., & mascarenhas, j. (2021). a systems representation of the circular economy: transition scenarios in the electrical and electronic equipment (eee) industry. technological forecasting & social change, 163(120414), 1–17. https://doi.org/10.1016/j.techfore.2020.120414 hai, h.t., hung, h.v., & quang, n.d. (2017). an overview of electronic waste recycling in vietnam. journal of material cycles and waste management, 19(1), 536–544. https://doi.org/10.1007/s10163-015-0448-x harrington, r. (2017). this incredible fact should get you psyched about solar power. business insider. https://www.businessinsider.com/solar-power-potential-how-much-energy-does-sun-have-2017-3 int. j. prod. manag. eng. (2022) 10(2), 131-141 creative commons attribution-noncommercial-noderivatives 4.0 international hidalgo-carvajal & carrasco-gallego 138 https://doi.org/10.1016/j.resconrec.2017.02.013 https://doi.org/10.1016/j.jclepro.2020.124214 https://doi.org/10.1007/s11356-019-04282-3 https://doi.org/10.1016/j.wasman.2010.11.006 https://doi.org/10.1016/j.esr.2019.100431 https://doi.org/10.1016/j.ijinfomgt.2015.01.002 https://doi.org/10.1016/j.jclepro.2020.120512 https://doi.org/10.3390/ijerph14020135 https://doi.org/10.1016/j.wasman.2016.01.034 https://doi.org/10.18178/ijesd.2017.8.8.1021 https://doi.org/10.18178/ijesd.2017.8.8.1021 https://doi.org/10.1016/j.jclepro.2016.05.064 https://eur-lex.europa.eu/legal-content/en/txt/?uri=com%3a2018%3a29%3afin https://doi.org/10.1016/j.resconrec.2010.02.001 https://doi.org/10.1016/j.jclepro.2016.12.048 https://doi.org/10.1016/j.wasman.2016.12.018 https://doi.org/10.1016/j.wasman.2017.03.009 https://doi.org/10.1016/j.techfore.2020.120414 https://doi.org/10.1007/s10163-015-0448-x https://www.businessinsider.com/solar-power-potential-how-much-energy-does-sun-have-2017-3 http://creativecommons.org/licenses/by-nc-nd/4.0/ hidalgo-carvajal, d., carrasco-gallego, r., & morales-alonso, g. (2021). from goods to services and from linear to circular: the role of servitization’s challenges and drivers in the shifting process. sustainability, 13(8), 4539. https://doi.org/10.3390/su13084539 honda, s., khetriwal, d.s., & kuehr, r. (2016). regional e-waste monitor east and southeast asia. irena, & iea-pvps. (2016). end-of-life management: solar photovoltaic panels. international renewable energy agency and international energy agency photovoltaic power systems. https://www.irena.org/publications/2016/jun/ end-of-life-management-solar-photovoltaic-panels islam, m.t., & huda, n. (2018). reverse logistics and closed-loop supply chain of waste electrical and electronic equipment (weee)/e-waste: a comprehensive literature review. resources, conservation and recycling, 137, 48–75. https://doi.org/10.1016/j.resconrec.2018.05.026 kang, s., yoo, s., lee, j., boo, b., & ryu, h. (2012). experimental investigations for recycling of silicon and glass from waste photovoltaic modules. renewable energy, 47(november), 152–159. https://doi.org/10.1016/j.renene.2012.04.030 kilic, h.s., cebeci, u., & ayhan, m.b. (2015). reverse logistics system design for the waste of electrical and electronic equipment (weee) in turkey. resources, conservation and recycling, 95, 120–132. https://doi.org/10.1016/j. resconrec.2014.12.010 korhonen, j., nuur, c., feldmann, a., & birkie, s.e. (2018). circular economy as an essentially contested concept. journal of cleaner production, 175(february), 544–552. https://doi.org/10.1016/j.jclepro.2017.12.111 lazard. (2019). lazard’s levelized cost of energy analysis – version 13.0. lazard. https://www.lazard.com/media/451086/ lazards-levelized-cost-of-energy-version-130-vf.pdf li, j., lopez n., b.n., liu, l., zhao, n., yu, k., & zheng, l. (2013). regional or global weee recycling. where to go? waste management, 33(4), 923–934. https://doi.org/10.1016/j.wasman.2012.11.011 li, j., zeng, x., chen, m., ogunseitan, o.a., & stevels, a. (2015). “control-alt-delete”: rebooting solutions for the e-waste problem. environmental science & technology, 49(12), 7095–7108. https://doi.org/10.1021/acs.est.5b00449 lima, a.b., rocha, t., pimentel, m., & feichas, l. (2016). the brazilian government efforts to support electronic recycling facilities to comply with environmental sound practices. 2016 electronics goes green 2016+ (egg), 1–8. https://doi.org/10.1109/egg.2016.7829837 lixandru, a., venkatesan, p., jönsson, c., poenaru, i., hall, b., yang, y., walton, a., güth, k., gauß, r., & gutfleish, o. (2017). identification and recovery of rare-earth permanent magnets from waste electrical and electronic equipment. waste management, 68(october), 482–489. https://doi.org/10.1016/j.wasman.2017.07.028 lofthouse, v., & prendeville, s. (2018). human-centred design of products and services for the circular economy – a review. the design journal, 21(4), 451–476. https://doi.org/10.1080/14606925.2018.1468169 marke, a., chan, c., taskin, g., & hacking, t. (2020). reducing e-waste in china’s mobile electronics industry: the application of the innovative circular business models. asian education and development studies, 9(4), 591–610. https://doi.org/10.1108/aeds-03-2019-0052 mar-ortiz, j., adenso-diaz, b., & gonzález-velarde, j.l. (2011). design of a recovery network for weee collection: the case of galicia, spain. journal of the operational research society, 62(8), 1471–1484. https://doi.org/10.1057/ jors.2010.114 mies, a., & gold, s. (2021). mapping the social dimension of the circular economy. journal of cleaner production, 321, 128960. https://doi.org/10.1016/j.jclepro.2021.128960 mihai, f.-c. (2020). electronic waste management in romania: pathways for sustainable practices. in handbook of electronic waste management (pp. 533–551). elsevier. https://doi.org/10.1016/b978-0-12-817030-4.00024-3 mihai, f.-c., gnoni, m.-g., meidiana, c., ezeah, c., & elia, v. (2019). waste electrical and electronic equipment (weee): flows, quantities, and management—a global scenario. in electronic waste management and treatment technology (pp. 1–34). elsevier. https://doi.org/10.1016/b978-0-12-816190-6.00001-7 morris, a., & metternicht, g. (2016). assessing effectiveness of weee management policy in australia. journal of environmental management, 181, 218–230. https://doi.org/10.1016/j.jenvman.2016.06.013 nikoloudakis, n., & rangoussi, m. (2019). on the role of (vocational) education in the adoption of green and recycling practices: a case study on weee. 020053. https://doi.org/10.1063/1.5116980 nowakowski, p., & mrówczyńska, b. (2018). towards sustainable weee collection and transportation methods in circular economy—comparative study for rural and urban settlements. resources, conservation and recycling, 135, 93–107. https://doi.org/10.1016/j.resconrec.2017.12.016 ongondo, f.o., williams, i.d., & cherrett, t.j. (2011). how are weee doing? a global review of the management of electrical and electronic wastes. waste management, 31(4), 714–730. https://doi.org/10.1016/j.wasman.2010.10.023 oteng-ababio, m., melara arguello, j.e., & gabbay, o. (2013). solid waste management in african cities: sorting the facts from the fads in accra, ghana. habitat international, 39(july), 96–104. https://doi.org/10.1016/j.habitatint.2012.10.010 int. j. prod. manag. eng. (2022) 10(2), 131-141creative commons attribution-noncommercial-noderivatives 4.0 international preparing for future e-waste from photovoltaic modules: a circular economy approach 139 https://doi.org/10.3390/su13084539 https://www.irena.org/publications/2016/jun/end-of-life-management-solar-photovoltaic-panels https://www.irena.org/publications/2016/jun/end-of-life-management-solar-photovoltaic-panels https://doi.org/10.1016/j.resconrec.2018.05.026 https://doi.org/10.1016/j.renene.2012.04.030 https://doi.org/10.1016/j.resconrec.2014.12.010 https://doi.org/10.1016/j.resconrec.2014.12.010 https://doi.org/10.1016/j.jclepro.2017.12.111 https://www.lazard.com/media/451086/lazards-levelized-cost-of-energy-version-130-vf.pdf https://www.lazard.com/media/451086/lazards-levelized-cost-of-energy-version-130-vf.pdf https://doi.org/10.1016/j.wasman.2012.11.011 https://doi.org/10.1021/acs.est.5b00449 https://doi.org/10.1109/egg.2016.7829837 https://doi.org/10.1016/j.wasman.2017.07.028 https://doi.org/10.1080/14606925.2018.1468169 https://doi.org/10.1108/aeds-03-2019-0052 https://doi.org/10.1057/jors.2010.114 https://doi.org/10.1057/jors.2010.114 https://doi.org/10.1016/j.jclepro.2021.128960 https://doi.org/10.1016/b978-0-12-817030-4.00024-3 https://doi.org/10.1016/b978-0-12-816190-6.00001-7 https://doi.org/10.1016/j.jenvman.2016.06.013 https://doi.org/10.1063/1.5116980 https://doi.org/10.1016/j.resconrec.2017.12.016 https://doi.org/10.1016/j.wasman.2010.10.023 https://doi.org/10.1016/j.habitatint.2012.10.010 http://creativecommons.org/licenses/by-nc-nd/4.0/ papaoikonomou, k., kipouros, s., kungolos, a., somakos, l., aravossis, k., antonopoulos, i., & karagiannidis, a. (2009). marginalised social groups in contemporary weee management within social enterprises investments: a study in greece. waste management, 29(5), 1754–1759. https://doi.org/10.1016/j.wasman.2008.11.012 parajuly, k., thapa, k.b., cimpan, c., & wenzel, h. (2018). electronic waste and informal recycling in kathmandu, nepal: challenges and opportunities. journal of material cycles and waste management, 20(1), 656–666. https://doi.org/10.1007/s10163-017-0610-8 pariatamby, a., & victor, d. (2013). policy trends of e-waste management in asia. journal of material cycles and waste management, 15(4), 411–419. https://doi.org/10.1007/s10163-013-0136-7 pekarkova, z., williams, i.d., emery, l., & bone, r. (2021). economic and climate impacts from the incorrect disposal of weee. resources, conservation and recycling, 168(105470), 1–10. https://doi.org/10.1016/j.resconrec.2021.105470 picon, a., ghita, o., iriondo, p.m., bereciartua, a., & whelan, p.f. (2010). automation of waste recycling using hyperspectral image analysis. 2010 ieee 15th conference on emerging technologies & factory automation (etfa 2010), 1–4. https://doi.org/10.1109/etfa.2010.5641201 ranjbari, m., saidani, m., shams esfandabadi, z., peng, w., lam, s.s., aghbashlo, m., quatraro, f., & tabatabaei, m. (2021). two decades of research on waste management in the circular economy: insights from bibliometric, text mining, and content analyses. journal of cleaner production, 314, 128009. https://doi.org/10.1016/j.jclepro.2021.128009 reike, d., vermeulen, w.j.v., & witjes, s. (2018). the circular economy: new or refurbished as ce 3.0? — exploring controversies in the conceptualization of the circular economy through a focus on history and resource value retention options. resources, conservation and recycling, 135, 246–264. https://doi.org/10.1016/j.resconrec.2017.08.027 ren21. (2020). renewables 2020 global status report. renewable energy policy network for the 21st century. richter, j.l., & koppejan, r. (2016). extended producer responsibility for lamps in nordic countries: best practices and challenges in closing material loops. journal of cleaner production, 123, 167–179. https://doi.org/10.1016/j. jclepro.2015.06.131 rochman, f.f., ashton, w.s., & wiharjo, m.g.m. (2017). e-waste, money and power: mapping electronic waste flows in yogyakarta, indonesia. environmental development, 24, 1–8. https://doi.org/10.1016/j.envdev.2017.02.002 santillán-saldivar, j., cimprich, a., shaikh, n., laratte, b., young, s.b., & sonnemann, g. (2021). how recycling mitigates supply risks of critical raw materials: extension of the geopolitical supply risk methodology applied to information and communication technologies in the european union. resources, conservation and recycling, 164, 105108. https://doi.org/10.1016/j.resconrec.2020.105108 shinkuma, t., & huong, n.t.m. (2009). the flow of e-waste material in the asian region and a reconsideration of international trade policies on e-waste. environmental impact assessment review, 29(1), 25–31. https://doi.org/10.1016/j.eiar.2008.04.004 shittu, o.s., williams, i.d., & shaw, p.j. (2021). global e-waste management: can weee make a difference? a review of e-waste trends, legislation, contemporary issues and future challenges. waste management, 120, 549–563. https://doi.org/10.1016/j.wasman.2020.10.016 singhal, d., tripathy, s., & jena, s.k. (2019). sustainability through remanufacturing of e-waste: examination of critical factors in the indian context. sustainable production and consumption, 20, 128–139. https://doi.org/10.1016/j. spc.2019.06.001 sousa, r., agante, e., cerejeira, j., & portela, m. (2018). eee fees and the weee system – a model of efficiency and income in european countries. waste management, 79, 770–780. https://doi.org/10.1016/j.wasman.2018.08.008 stejskal, b. (2016). animated 3d training software—inovative tool for the weee reuse and recycling industry. infrastruktura i ekologia terenów wiejskich, iv/2/2016, 1463–1471. https://doi.org/10.14597/infraeco.2016.4.2.108 umair, s., anderberg, s., & potting, j. (2016). informal electronic waste recycling in pakistan. the journal of solid waste technology and management, 42(3), 222–235. https://doi.org/10.5276/jswtm.2016.222 united nations. (2019). world population prospects 2019. department of economic and social affairs population division. valencia, m. (2019). informal recycling sector (irs), contribution to the achievement of the sdgs, and a circular economy. in responsible consumption and production. springer. https://doi.org/10.1007/978-3-319-71062-4_107-1 van loon, p., diener, d., & harris, s. (2021). circular products and business models and environmental impact reductions: current knowledge and knowledge gaps. journal of cleaner production, 288, 125627. https://doi.org/10.1016/j.jclepro.2020.125627 wang, x.v., & wang, l. (2019). digital twin-based weee recycling, recovery and remanufacturing in the background of industry 4.0. international journal of production research, 57(12), 3892–3902. https://doi.org/10.1080/00207543.20 18.1497819 wang, z., zhang, b., & guan, d. (2016). take responsibility for electronic-waste disposal. nature, 536(7614), 23–25. https://doi.org/10.1038/536023a int. j. prod. manag. eng. (2022) 10(2), 131-141 creative commons attribution-noncommercial-noderivatives 4.0 international hidalgo-carvajal & carrasco-gallego 140 https://doi.org/10.1016/j.wasman.2008.11.012 https://doi.org/10.1007/s10163-017-0610-8 https://doi.org/10.1007/s10163-013-0136-7 https://doi.org/10.1016/j.resconrec.2021.105470 https://doi.org/10.1109/etfa.2010.5641201 https://doi.org/10.1016/j.jclepro.2021.128009 https://doi.org/10.1016/j.resconrec.2017.08.027 https://doi.org/10.1016/j.jclepro.2015.06.131 https://doi.org/10.1016/j.jclepro.2015.06.131 https://doi.org/10.1016/j.envdev.2017.02.002 https://doi.org/10.1016/j.resconrec.2020.105108 https://doi.org/10.1016/j.eiar.2008.04.004 https://doi.org/10.1016/j.wasman.2020.10.016 https://doi.org/10.1016/j.spc.2019.06.001 https://doi.org/10.1016/j.spc.2019.06.001 https://doi.org/10.1016/j.wasman.2018.08.008 https://doi.org/10.14597/infraeco.2016.4.2.108 https://doi.org/10.5276/jswtm.2016.222 https://doi.org/10.1007/978-3-319-71062-4_107-1 https://doi.org/10.1016/j.jclepro.2020.125627 https://doi.org/10.1080/00207543.2018.1497819 https://doi.org/10.1080/00207543.2018.1497819 https://doi.org/10.1038/536023a http://creativecommons.org/licenses/by-nc-nd/4.0/ williams, i.d. (2016). global metal reuse, and formal and informal recycling from electronic and other high-tech wastes. in r.m. izatt (ed.), metal sustainability (pp. 23–51). john wiley & sons, ltd. https://doi.org/10.1002/9781119009115. ch2 wilson, d.c., velis, c., & cheeseman, c. (2006). role of informal sector recycling in waste management in developing countries. habitat international, 30, 797–808. https://doi.org/10.1016/j.habitatint.2005.09.005 yasin, a., & hasnain, m.i. (2012). on the quality of grey literature and its use in information synthesis during systematic literature reviews [master thesis report]. blekinge institute of technology. yu, j., williams, e., ju, m., & shao, c. (2010). managing e-waste in china: policies, pilot projects and alternative approaches. resources, conservation and recycling, 54(11), 991–999. https://doi.org/10.1016/j.resconrec.2010.02.006 yumoto, t., & shiratori, t. (2009). study on recycling of metals in weee-investigation for making an inventory of metal contents-. journal of mmij, 125(2), 75–80. https://doi.org/10.2473/journalofmmij.125.75 zhang, h., duan, h., zuo, j., song, m., zhang, y., yang, b., & niu, y. (2017). characterization of post-disaster environmental management for hazardous materials incidents: lessons learnt from the tianjin warehouse explosion, china. journal of environmental management, 199, 21–30. https://doi.org/10.1016/j.jenvman.2017.05.021 int. j. prod. manag. eng. (2022) 10(2), 131-141creative commons attribution-noncommercial-noderivatives 4.0 international preparing for future e-waste from photovoltaic modules: a circular economy approach 141 https://doi.org/10.1002/9781119009115.ch2 https://doi.org/10.1002/9781119009115.ch2 https://doi.org/10.1016/j.habitatint.2005.09.005 https://doi.org/10.1016/j.resconrec.2010.02.006 https://doi.org/10.2473/journalofmmij.125.75 https://doi.org/10.1016/j.jenvman.2017.05.021 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j international journal of production management and engineering mrp-based negotiation in collaborative supply chains grabot, b., ming y., & houé r. university of toulouse, inpt, lgp-enit, 47 avenue d’azereix, bp1629, f-65016 tarbes cedex, france bernard.grabot@enit.fr yue.ming@enit.fr raymond.houe@enit.fr abstract: although collaboration between partners is now considered as a mean to increase the performance of the supply chains, most of them are still managed in a directive way through cascades of classical mrp/mrp2 systems, in which the constraints of the suppliers, especially the smallest ones, are poorly taken into account. we suggest in this article to assess the interest of the integration into collaborative processes of some practices based on mrp, identified after a study in aeronautical supply chains. within these processes, classical parameters of mrp would be negotiated instead of being imposed by the large companies. key words: collaborative supply chain, aeronautical industry, mrp2, negotiation. 1. introduction recent studies on supply chain management consider that information sharing, joint-planning, cooperation and strategic partnerships over the entire supply chain are conditions for building more efficient and reactive supply chains (see for instance (sahin and powell robinson, 2005)). nevertheless, large industrial supply chains, like in the automotive or aeronautical sectors, still use classical production management methods, and especially mrp2 (manufacturing resource planning), as the backbone of the exchanges between suppliers and customers (van donselaar et al., 2000). when, as often, customers have more power on their suppliers than the opposite, the suppliers, especially if they are small companies, may be reluctant to express their local constraints to their customers. this difficulty is still increased when these constraints are poorly consistent with supply chain management and lean manufacturing principles, now intensively promoted by large companies. the risk is then that these constraints turn into hidden practices, which may decrease the performance of the supply chain. this article first aims at identifying such practices through interviews in companies of the aeronautic sector (section 2), then at “publishing” these constraints and including them in negotiation processes, even if the negotiation on the considered items may be considered as incompatible with nowadays habits (section 3). our purpose is to show that including some parameters of mrp into a negotiation process could lead in some cases to a “win-win” situation which could benefit to all the supply chain. 2. analysis of real practices in the aeronautic sector 2.1. the context of aeronautical supply chains the aeronautical industry has an oligopolistic market structure characterized by high technological, financial and market entry barriers (tyson, 1992). to lower these barriers, the aircraft firms implement a production organization characterized by a pyramidshaped hierarchic structure (convergent network) http://dx.doi.org/10.4995/ijpme.2013.1516 received: 2013-05-15 accepted: 2013-05-31 https://ojs.upv.es/index.php/ijpme 27int. j. prod. manag. eng. vol. 01 (2013): 27-37creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2013.1516 with an assembly area at the top where the parts and components coming from three main sub-sectors (engines, equipment and avionics, and airframe) are assembled for obtaining the final product (esposito and passaro, 2009). three main production levels may be distinguished: at the first level, the leader firm, which directly operates in the airframe sub-sector, organizes the flow of parts, components, systems and information and coordinates the program and assembly of the final product. the firms of the second production level manufacture complex parts and components of the aircraft both for the sub-systems manufacturers and for the assembler (fuselage, wings, motors, land gears etc.). generally, these large firms are leaders in their own program, and belong at the same time to the first and second level. the third production level includes small and medium enterprises (sme), subcontracting firms who generally work for the second production level. from the second level firms, they receive the information on the production process, the manufacturing specifications, the technical service etc. when the manufacturing process is over, they transfer the ordered components and relevant information to the second-level customers. the suppliers in this level are approved by the leader firm, which checks if they are able to meet the quality standards required and if the production process is realized according to the procedures imposed in the program. in the last few years, suppliers have become increasingly involved in the production of parts with high added value, including the management of their upstream suppliers, but have also been invited to share the risks of the programs with the aircraft designer (esposito and passaro, 2009). a major difference netween this sector and the automotive industry is the higher diversity of the parts, aircrafts being produced in small series, but containing much more different parts (up to 1 million parts for a large aircraft), with long cycle times. this results in three important constraints: long term planning. since the aircraft production cycle is much longer than for other products, it requires the involved partners (subcontractor, supplier) to have long terms procurement, manufacturing and transportation plans, including to monitor changes during a long period. high flexibility production. manufacturing an aircraft is not a mass production process; each component is possibly different depending on the type of aircraft. this requires a high flexibility and reactivity of each supply network member when performing his production processes. efficient relationship. as an industry dealing with many diverse and expensive parts, inventories are to be kept at the minimum level while ensuring a good availability of each component, shortage having tremendous consequences. thereby, the storage and delivery of raw materials, components, semi-finished products and subsystems should be optimized, which highly depends on the coordination between supply network members. within aeronautical supply chains, the presence of many smes leads to specific advantages, but also to some constraints. leaders of smes are usually very receptive to technological constraints but the planning process may be informal. managers and operatives are more likely to be directly involved with the customers, two ways and faceto-face communication being the norm in smes (ghobadian and gallear, 1996). as a consequence, research shows that smes are more responsive to market needs, more adaptable to change, and more innovative in their ability to meet the customers’ demand, but less oriented on information technology and management tools allowing long/middle term visibility (appiah-adu and singh, 1998; quayle, 2003). the traditional approaches and methodologies promoted by large companies (lean manufacturing, mrp...) are therefore considered by some authors as not suitable for smes, supposed to prefer logical reasoning approaches over systematic planning approaches, like aggregate production plans or production forecasts (thakkar et al., 2008). 2.2. supply chain management in aeronautical supply chains supply chains (more exactly “supply networks”, their structure being never linear) can be managed using two main techniques: centralized, or decentralized planning. centralized planning can be performed using an aps (advanced planning system, (stadtler and kilger, 2007)), able to provide an optimal planning for all the members of a supply chain. in practice, this results in a poor autonomy of each company, which would be dedicated to a given supply chain, which is not the case in the aeronautical industry. on the other hand, decentralized planning can be performed using a point-to-point relationship, each partner receiving demands from his customers, that he translates into a supply plan for his own suppliers. this can be done using the mrp2 method 28 int. j. prod. manag. eng. vol. 01 (2013): 27-37 grabot, b., ming y., & houé r. creative commons attribution-noncommercial 3.0 spain (manufacturing resource planning) (see for instance (adams and cox, 1985)), which is the base of the production management modules of all the erp (enterprise resource planning) systems. using mrp2, and for each partner of the supply chain, forecasts should be used as inputs for building a sales and operation plan (sop) with a 1-3 years horizon in the aeronautical sector (see figure 1). a master production schedule (mps) can then be deduced at lower term. on the base of the obtained sequenced requirements on the final products, the bills of materials are used for generating on one side a supply plan, and on the other side a production plan (material requirement planning step). the adequacy between the load generated by the production plan and the capacity of the company is checked (load planning), then the production is scheduled, with a typical horizon of 1-2 weeks. since adaptations of the initial production program may be necessary in order to react to unexpected events, the forecasts usually contain three zones: firm period, corresponding to confirmed orders (typically: 1 to 2 months); flexible period (typically: 3-6 months), in which the orders may vary within given ratios, e.g. +/-20%, and “free” period, only given for information. the total horizon of the forecasts is usually 2-3 years, in order to allow the suppliers to anticipate large variations of the demand. therefore, supply chain management is supposed to be implemented in this sector through a cascade of mrp systems, one in each company (see figure 1), the supply plan of each company being used to create the forecasts sent to the suppliers. it is therefore mandatory that each partner uses consistent methods, and is able to efficiently perform his local role for propagating information upstream the supply chain. as a consequence, it is usually considered that smes should switch from simple financial plans to forecast based planning (thakkar et al., 2008). in that purpose, the use of erp systems including mrp modules, is more and more considered as a condition for smes to join supply chains (lenny koh and simpson, 2005). we shall see in next section that the use of these systems may in some cases be difficult. 3. from local practices to negotiation processes 3.1. emergence of local practices in supply chains during three projects launched by large companies aiming at analyzing the product flow in their supply chains (ming, 2011), interviews have been performed in several tenths of both large and small companies. they showed that many local practices were added to the global process described in figure 1 in order to cope with local objectives or constraints of the supply chain members. the performed analysis has mainly concerned four major operational processes: the answer to a “request for quotation”, having for result the creation of a middle/long term relationship between customer and supplier; the middle term order management; the orders fulfillment (short term) and the supplier development, through audit and transfer of various tools and techniques, among which mrp and lean manufacturing (this point will not be detailed here). some findings of the study are summarized hereafter. in each case, we have tried to distinguish (even if it may be ambiguous in some cases), cooperative practices, i.e. practices aiming at helping the partner, from selfish practices, aiming at preserving local interests. customer’s side/cooperative practices: attempts for protecting the small suppliers from variations of the demand have been noticed from several large companies, who try for instance, to smooth the demand between two periods when sending a supply plan to their small suppliers. others try to increase the lot sizes of their commands in order to help their suppliers. some companies also noticed that sending precise due dates for all their orders was not always justified, since safety inventories were available for some parts. as a consequence, they decided to mention their level of inventory with each order, in order to 29int. j. prod. manag. eng. vol. 01 (2013): 27-37 mrp-based negotiation in collaborative supply chains creative commons attribution-noncommercial 3.0 spain   figure 1. supply chain management as a cascade of mrp systems (grabot and mayère, 2009) allow their suppliers to assess themselves the criticity of the respect of the due date. customer’s side/selfish practices: most of the large companies admit that they choose suppliers for which they represent an important sale ratio (in order to have power over them), but not too much (in order to avoid any responsibility if they decide to decrease their orders). partnerships are usually signed for several years, all the contracts including a decrease of the price through time (up to 8% per year). the service ratio (percentage of parts delivered on time on a given period, often the month) is a very important indicator. when their ratio is not good, many companies turn the status of their order to “urgent” (which has a specific significance in the aeronautical industry) in order to be sure to be delivered on time, with the result of the destabilization of small suppliers having to cope with up to 30% of urgent parts. the supply network defines a “pert-like” structure converging on the assembly of an aircraft, in which not all the orders (and suppliers) are on the critical path. therefore, meeting the due dates does not have the same importance for all the parts and suppliers. nevertheless, the logisticians, each of them managing a group of suppliers at the customer’s, do not always know what are the “real” urgencies; therefore, they have no other choice than setting the same pressure on all the suppliers for all the parts, which may destabilize them in overloaded situations. supplier’s side/cooperative practices: even if the production program of each aircraft is relatively well known and stable, many hazards may occur during the manufacturing cycle, with the result of urgencies, or even changes in the definition of the parts. all the suppliers are used to these constraints, and do their best for dealing with them, even if it may be costly. the scarcity of some raw materials (e.g. some aeronautical alloys and casted parts) has considerably increased the supply times in the recent years; some suppliers have now to order raw materials before the corresponding orders are confirmed... in that case, they take an important financial risk, to the benefit of the supply chain. supplier’s side/selfish practices: each supplier has “important” customers (usually those who generate a high cash flow) and less important ones. therefore, when their capacity is insufficient, they privilege their important customers. nevertheless, it is not rare that all the customers work for the same final assembler. in that case, the supply chain is in a way “competing against itself”, not always for its final benefit since the most important customer for a local supplier is not necessarily the critical one for the supply chain. some suppliers having a critical competence (for instance surface treatment, which is object of strict regulations and is therefore a scarce resource) may use this power for imposing to their customers a link between price and cycle time, which can hardly be taken into account in the classical mrp framework. in many cases, having to decrease their prices through time leads the suppliers to questionable solutions. many of them try for instance to increase their lot sizes, by grouping the demand on parts having similar characteristics. the result can sometimes be both early parts (creating inventories), and late ones. these examples show a complex and sometimes paradoxical situation, were real attempts to help the partner cohabit with selfish behaviors, that often remain hidden. the usual way to address this problem is to promote standardized “best practices” supposed to improve the global performance of the supply chain to the final benefit of each partner. nevertheless, some authors have shown that if the appropriate mechanisms are not in place, the supplier may not perceive the benefits associated with these investments and may reject the initiative to modify or improve their processes (krause et al. 1998). for others, difficulties in developing “state of the art” capabilities in management, technology or co-operation is inherent to relationships between large and small companies (chen and chen, 2002; blomqvist, 2002). in order to cope with this problem, a different research paradigm may be chosen. all behaviors have a justification for the companies adopting them, even if they can be considered as “positive” or “negative” according to usual criteria. whatever their consequence can be, and even if they are not consistent with present habits, we have therefore chosen to consider some of the typical practices that we have observed, and try to include them in an open negotiation process, instead of keeping them hidden or informal. the practices we have chosen to test are described in next section. 30 int. j. prod. manag. eng. vol. 01 (2013): 27-37 grabot, b., ming y., & houé r. creative commons attribution-noncommercial 3.0 spain 3.2. a negotiation process for integrating hidden practices into collaborative processes an explanation of the practices listed in previous section may be that many parameters usually defined in the contract between customer and supplier heavily depend on the context and on the situation of each partner. for instance, the price of the parts (including its evolution through time) is defined when the contract is concluded (even if it can be renegotiated in some cases). it is nevertheless clear that the production costs depend on a precise context: for instance, parts produced during extra-hours are more expensive than if produced during normal hours. on the other hand, since the price is an important criterion for choosing a partner, accepting that he may change it after the contract has been concluded is somehow inconsistent. similarly, the periods of the forecast (firm, flexible and free) should depend on the uncertainty of the final demand, and not on a local context. nevertheless, we have decided here to assess what would happen if the identified local (and sometimes “hidden”) practices would be included in formal negotiation processes. as an illustration, we have chosen to focus on four practices often mentioned during the interviews (directly or indirectly): the periods of the forecast. the difference between the length of the flexible period received by the customer and the one he sends to a supplier is a way to put pressure (shorter period) or to protect (longer period) the supplier. similarly, sending to a supplier a firm period shorter than the supply cycle of his raw materials obliges the supplier to take risks. we suggest to generalize these behaviors by allowing a negotiation on the length of the periods of forecasts. the possible load variations during the flexible periods of the forecasts. these variations are usually defined in the contracts, and can be up to 50% in the aeronautical sector. nevertheless, high variations may be acceptable by the supplier in some periods, while even low ones may be problematic in other periods, depending on the global load of the supplier. therefore, we suggest to adjust a possible variation through negotiation. the prices and cycle times. prices and cycle times are linked by the resources used, even if these two items are implicitly considered as independent. as a consequence, we suggest that the prices in a given situation are adjusted according to the real costs induced by each situation. the order priority and lot sizes. lot sizes are usually defined in the contracts, but their variation may allow to balance productivity and cycle time for a supplier. similarly, suppliers have to sequence orders coming from their various customers without an objective view of their real criticity for each customer. we suggest that these two points could be discussed for a better mutual interest. 3.2.1. negotiation on the periods of the forecasts the real issue in the choice of the periods of forecast is risk sharing: risk taken by the supplier when he orders raw materials or releases production orders on the base of the flexible period of the forecasts he receives, and risk taken by the customer when he accepts to send to a supplier a firm period longer than the one he himself receives from his own customer. however, the interest of protective behaviors from the customer depends on the actual situations of the supplier. therefore, we propose to put the periods of forecast into the middle term negotiation process, which would allow to make the length of the periods more flexible, being negotiated on the base of the real requirements and actual necessities of both customer and supplier. the business process diagram describing the conditions for releasing a negotiation on the periods of the forecasts is summarized in figure 2, with the customer’s activities in the top and the supplier’s ones in the bottom on the figure. using the mrp2 method, forecasts coming from customer’s customer are inputs of the customer’s sop, and then used to generate the mps (master production schedule) (point ① in figure 2). the mps provides more detailed production requirements to the mrp (material requirement planning) calculation (point ②). the supply plan (one of the outputs of mrp) is calculated using the bom (bill of materials), supply lead time, material inventory level, etc., according to the contractual horizons, including lengths of firm, flexible and free periods (point ③). he supply plan is the base of the forecasts received by the supplier (point ④). the supplier makes then his own mrp calculation (point ⑤), resulting in a supply plan (not mentioned in figure 2) and a load plan (point ⑥). since he has taken into account his cycle time and the cycle time of his suppliers, the supplier is able to see whether this load planning is consistent or not, or in other terms whether he takes 31int. j. prod. manag. eng. vol. 01 (2013): 27-37 mrp-based negotiation in collaborative supply chains creative commons attribution-noncommercial 3.0 spain too much risks (for instance by ordering parts on the base of the flexible period of forecasts, point ⑦). depending on additional (and sometimes subjective) information on his customers and suppliers (e.g. can they work faster? do they have financial stability?), he decides whether these risks are acceptable or not (point ⑧). if he considers that he takes more risks than his partners (customers and suppliers), he may ask for negotiation (point ⑨). the customer performs the same evaluation: he makes his assessment of both internal risks and risks he thinks the supplier takes (point ⑩). this assessment of course considers the horizon of the firm period received from his own customer, the horizon of the firm period he sends to his supplier, his internal cycle time, his supplier’s cycle time, etc. it should also include his opinion on additional information, like the cycle time from supplier’s suppliers, the real costs of his supplier, etc. it is clear that this information may be subjective or imprecise, since it is usually not provided by the supplier, who normally does not accept to communicate his real costs to his customer. the risk taken by the customer can be different for each of his suppliers, since two different suppliers do not need the same protection, or in other terms do not deserve that the customer takes the same risk. it is for instance acceptable to take risks for protecting a critical supplier, but not a “common” one. the assessment of the risk will so denote the customer’s vision on the allocation of risks between him and his suppliers. the next step is to decide on the acceptability of the risks he takes (point ⑪), by balancing the customer’s own strength (precisely known) and its supplier’s strength and weakness (supposed). this assessment is subjective, but is indeed implicitly done daily in real situations, within less formalized processes. if, from the customer’s vision, his risks are not acceptable, he will ask for a negotiation process (point ⑫). otherwise, the customer will accept the current plans (point ⑬). after the negotiation process, a new agreed horizon will be integrated into customer’s mrp plan. in all these cases, sharing real information instead of trying to guess the situation of the partner could facilitate to reach a consensus, but would certainly lead to other problems, especially linked to confidentiality and trust. the exchanges of information between actors when negotiation is requested by the customer have been modeled using an uml sequence diagram (omg, 32 int. j. prod. manag. eng. vol. 01 (2013): 27-37 grabot, b., ming y., & houé r. creative commons attribution-noncommercial 3.0 spain   figure 2. negotiation of the periods of forecasts figure 2. negotiation of the periods of forecast. 2011) in figure 3. the corresponding figures are not provided for the other negotiations, but bring quite similar information. 3.2.2. negotiation on load variations the second item we suggest to discuss, load variation, can generate problems at both customer and supplier’s sides. the capacity of the suppliers is usually limited, especially because low prices are poorly consistent with extra capacity, but also because of the present increasing workload in the aeronautic sector. therefore, it is certainly dangerous for the customer to send an irregular load to his suppliers if he is himself more likely able to cope with variations than his suppliers. we have seen in previous sections that some customers try to limit the load variation between two consecutive periods even if it would be allowed by their contractual agreement. nevertheless, this protective attitude is perhaps not always necessary since the supplier can occasionally be able to cope with this variation, especially if the price paid by the customer covers his extra costs, linked to a temporary increase of its capacity or to sub-contracting. therefore, instead of considering that the supplier has to answer to an overload if it is consistent with the contract, or cannot answer to an overload (in consistence with the contract or not), overloads (but also lacks of loads) could also be negotiated, including setting into question the corresponding price paid by the customer. on the customer’s side, the negotiation on load variation is proposed after the mrp step has been performed (see the top part of figure 4). after integrating the forecasts in the sop, then processing the mps, the customer begins the mrp calculation (point ① in figure 4). the customer may then consider the supply plan for each of his suppliers in order to identify high load variations (by comparing the load on the current and previous periods) (point ②). for dealing with high load variation, the flexibility of the mid-term capacity of the supplier is essential. therefore, the customer has to estimate the mid-term capacity on the supplier’s side (point ③), as well as the costs to manage such capacity (point ④). as a consequence, additional information on the supplier’s capacity, including internal regular and overtime capacity, externally accessible capacity (through subcontracting) (point ⑤), and additional information on related costs (point ⑥) are important inputs for this estimation. again, depending on the closeness of the relationship, this information can be known or estimated. based on the estimated results, the customer needs to assess the feasibility of the load variation expected in the current period (point ⑦). from the customer’s vision, if the supplier is capable to manage this load variation, the current plan is considered as feasible and the mrp result is accepted (point ⑧). otherwise, the customer requests for a negotiation process, considering as doubtful the supplier’s capability to perform on time delivery when facing the considered load variation (point ⑨). on the supplier’s side (bottom of figure 4), the detection of capacity problems is not based on estimation, but on the actual capacity/load situation. according to the result of the load planning (point ⑩), the supplier identifies a possible capacity problem (point ⑪) and checks the feasibility (point ⑫) to address this problem (by extra hours or subcontracting in case of increase, by other solutions aiming at decreasing his capacity in the opposite case). in that purpose, two important factors have to be taken into account: price paid by customer (point ⑬) cost for changing capacity (point ⑭). from the supplier’s vision, if the capacity change is considered as feasible, the current plans are accepted (point ⑮). otherwise, the supplier will request a 33int. j. prod. manag. eng. vol. 01 (2013): 27-37 mrp-based negotiation in collaborative supply chains creative commons attribution-noncommercial 3.0 spain   figure 3. sequence of activities for periods of forecast negotiation (customer requested). negotiation process and communicate his capacity problems to the customer (point ⑯). again, the negotiation process will be triggered either by a customer request, a supplier request or a double request (not considered here). of course, a problem detected by one of the partners should be validated by the other before negotiation. for instance, a customer may detect a high overload that may have no consequence for a supplier, if other customers of this supplier have decreased their orders during the same period. 3.2.3. negotiation on prices vs. cycle times as already mentioned, urgent orders are quite usual in the supply and demand process of the aeronautical industry, even if the demand is supposed to follow long term programs. we distinguish this point from the previous one in the sense that overloads can be detected quite early, in the flexible period of the forecasts for instance, whereas urgencies have to be handled at short term, often in the firm period. urgencies are of course firstly detected at the customer’s side, but the supplier is the one challenged through its flexibility and adjustment of capacity. therefore, it can be considered as in figure 5 that the problem of the cycle time for quick delivery is detected at the operational level of the supplier. usually, the urgencies are processed by the supplier depending on the influence of the customer over him, with the result of possible disturbances on the short term planning propagated to other customers. two cooperative behaviors could help to mitigate these problems: the first one would deal with the price, allowing the supplier to find extra capacity for processing the urgent parts, whereas the second one would deal with a better negotiation on the priorities between the supplier and his customers (this point is addressed in next section). concerning the first point, we shall consider here that the cycle time of urgent orders is partially negotiable, as well as their cost. when an urgent demand occurs, the customer should pay for the cycle time he expects according to the situation of his supplier; for instance, no increase of price would be required if the supplier is in an under loaded period but in other cases, a negotiation process on price and cycle time is suggested to cope with the constraints due to the supplier capacity. the negotiation on price and cycle time is considered here for a small number of urgent orders. at the customer’s side, the mrp calculation is based on the sop and mps, also taking into account the urgent orders sent by the customer’s customer (point ① of figure 5), at the level consistent with their 34 int. j. prod. manag. eng. vol. 01 (2013): 27-37 grabot, b., ming y., & houé r. creative commons attribution-noncommercial 3.0 spain   figure 4. negotiation on load variation figure 4. negotiation on load variation. degree of anticipation. the results of the mrp step provide a clear view on the changes in the material requirements induced by these urgent orders for the supplier (point ②): they may have no effects on the current supply plan, or urgent material orders may be necessary. after the load planning, the required due dates of the materials are confirmed (point ③), then the customer needs to estimate the feasibility of urgent orders on supplier’s side (point ④), as well as the possible extra cost for the supplier (point ⑤). according to customer’s vision, if the urgent orders are considered as feasible, meaning that the supplier is supposed to be capable to deal with such urgency, the current plan is accepted (point ⑥) and the urgent orders are sent to the supplier (point ⑦). otherwise, negotiation is requested (point ⑧). at the supplier’s side, the urgent orders are usually known at the load planning or detailed scheduling levels (point ⑨). based on the allocation of load/ capacity towards each customer, the supplier needs to check whether it is feasible to deliver the urgent order(s) (point ⑩) in the conditions required by the customer (including price) (point ⑪). if the actual situation allows the supplier to adjust his load/ capacity for fulfilling the urgent orders, the current plan is acceptable and the production process is launched (point ⑫). otherwise, the supplier sends a request for negotiation (point ⑬), and notifies his customer that delivery as required is questionable in the present situation. after negotiation on the urgent orders, the new agreed due date will be integrated in both customer and supplier’s plans (point ⑭, ⑮). 3.2.4. negotiation or priority of orders and lot sizes the final item that we suggest to put into the negotiation process groups two operational degrees of freedom, namely the orders priority and lot sizes. from the interviews, we have seen real cases where smes try to regroup orders having common characteristics, usually in order to decrease the setup times by increasing the lot sizes (but other reasons may exist). such regrouping, performed at the mrp level on the supplier’s side, could possibly lead to early or delayed orders if not done properly. if all the orders cannot be fulfilled in time, and without additional information from their customers, it is also common that the suppliers use an internal priority for scheduling the orders at the operational level. as a consequence, tardy orders for one or several customers may occur. time margins or safety stocks may allow the customer to face delayed delivery on some of the orders, but this information is not always shared with the suppliers. the negotiation on orders priority and lot sizes occurs at the operational levels, and is mainly related to constraints of capacity or cost (see figure 6). at the customer’s side, depending on the lot sizing policy, the lot size is either an input (for instance, if an economical lot size has been defined) (point ①) or a result (if a lot-for-lot policy is used) (point ②). the customer may in the last case need to check whether the supplier’s constraints on lot sizes are consistent with his actual requirements (point ③). if, from the customer’s point of view, there is no possible problem, the current mrp calculation is acceptable 35int. j. prod. manag. eng. vol. 01 (2013): 27-37 mrp-based negotiation in collaborative supply chains creative commons attribution-noncommercial 3.0 spain   figure 5. negotiation on prices and cycle times figure 5. negotiation on prizes and cycle times. and a load planning and detailed scheduling can be performed (point ④, ⑤). if the customer considers that the current lot size is not feasible, due to the constraints of the supplier, a request for negotiation on lot size will be sent (point ⑥). at the supplier’s side, there are two major tasks: one is to check the feasibility of lot sizes based on the results of the mrp calculation (point ⑦); the other is to check the respect of the due dates based on the load planning and detailed scheduling (point ⑧). if the supplier considers that increasing the contractual lot size could possibly lead to some benefits (point ⑨), a request for negotiation on lot sizes can be sent to the customer (point ⑩). similarly, if meeting all the due dates of the orders in process is not possible, and instead of defining internal priorities linked to the importance of each customer (point ⑪), the supplier can ask for a negotiation on the real priorities of the orders (point ⑩), which would allow him to define a schedule possibly acceptable by all the customers. after the negotiation process, the new agreed lot sizes will be integrated into the mrp calculation of both customer and supplier (points ⑫, ⑬), and the order priorities will be entered into the load planning and scheduling (points ⑭, ⑮). it can be noticed that these two negotiations are quite different from the previous ones, since they may involve several customers at the same time, and would so be certainly more difficult to handle in practice. we have suggested in this section several negotiation processes which can be added to a classical mrp2 process, aiming at discussing issues linked to local (but sometimes hidden) practices either by customers of suppliers of the aeronautical industry identified during our interviews. because they turn some points included in the contracts into negotiable items, these practices can be considered as unrealistic. nevertheless, we think that their possible interest should be assessed. 4. conclusion despite the fact that our suggestions may seem to be inconsistent with common industrial habits (including e.g. continuous negotiation on prices and cycle times), the suggested negotiation processes are quite consistent with some practices identified during the industrial interviews. in any case, our goal is not to suggest a so-called “optimal” negotiation process, but to take some real empirical situations from case studies as examples, and try to include them into a consistent formal negotiation process, in order to check their real potential. in that purpose, numerical simulations are in progress in order to better identify the situations in which such negotiatios would be pertinent. 36 int. j. prod. manag. eng. vol. 01 (2013): 27-37 grabot, b., ming y., & houé r. creative commons attribution-noncommercial 3.0 spain   figure 6. negotiation on priority of orders and lot sizes figure 6. negotiation on priority of orders and lot sizes. references adams, f. p., & cox, j.f., (1985). manufacturing resource planning: an information systems model. long range planning, 18(2), 86-92. doi:10.1016/0024-6301(85)90026-3 appiah-adu, k., & singh, s., (1998). customer orientation and performance: a study of smes, management decision, 36(6), 385-394. doi:10.1108/00251749810223592 blomqvist, k. (2002). partnering in the dynamic environment: the role of trust in asymmetric technology partnership formation, phd thesis, lappeenranta university of technology, finland. chen, h., & chen, t.j. (2002). asymmetric strategic alliance: a network view. journal of business research, 55(12), 1007-1013. doi:10.1016/ s0148-2963(02)00284-9 esposito, e., & passaro, r. (2009). the evolution of supply chain relationships: an interpretative framework based on the italian inter-industry experience. journal of purchasing & supply management, 15(2), 114-126. doi:10.1016/j.pursup.2009.03.002 ghobadian, a., & gallear, d.n. (1996). total quality management in smes. omega, 24(1), 83-106. doi:10.1016/0305-0483(95)00055-0 grabot, b., & mayère, a. (2009). sme and planning in supply chains: a socio-technical view. iesm’09, may 13-15, montréal, canada. krause, d. r., hanfield, r. b., & tyler, b. b., (2007). the relationships between supplier development, commitment, social capital accumulation and performance improvement. journal of operations management, 25(2), 528-545. doi:10.1016/j.jom.2006.05.007 lenny koh, s.c., & simpson, m. (2005). change and uncertainty in sme manufacturing environments using erp. journal of manufacturing technology management, 16(6), 629-653. doi:10.1108/17410380510609483 ming, y. (2011). models for customer-supplier negotiation in a collaborative supply chain, phd thesis, university of toulouse, inpt-enit. quayle, m. (2003). a study of supply chain management practice in uk industrial smes. supply chain management: an international journal, 8(1), 79-86. doi:10.1108/13598540310463387 sahin, f., & powell robinson, e. (2005). information sharing and coordination in make-to-order supply chains. journal of operations management, 23(6), 579-598. doi:10.1016/j.jom.2004.08.007 stadtler, h., & kilger, c. (2007) supply chain management and advanced planning. berlin, springer eds. thakkar, j., kanda, a., & deshmukh, s. g. (2008). supply chain management in smes: development of constructs and propositions. asia pacific journal of marketing and logistics, 20(1), 97-131. doi:10.1108/13555850810844896 tyson, l. (1992). who’s bashing whom? trade conflict in high technology industries institute for international economics. washington dc. van donselaar, k., van den nieuwenhof, j., & visschers, j. (2000). the impact of material coordination concepts on planning stability in supply chains. international journal of production economics, 68(2), 169-176. doi:10.1016/s0925-5273(00)00033-5 37int. j. prod. manag. eng. vol. 01 (2013): 27-37 mrp-based negotiation in collaborative supply chains creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.1016/0024-6301(85)90026-3 http://dx.doi.org/10.1108/00251749810223592 http://dx.doi.org/10.1016/s0148-2963(02)00284-9 http://dx.doi.org/10.1016/s0148-2963(02)00284-9 http://dx.doi.org/10.1016/j.pursup.2009.03.002 http://dx.doi.org/10.1016/0305-0483(95)00055-0 http://dx.doi.org/10.1016/j.jom.2006.05.007 http://dx.doi.org/10.1108/17410380510609483 http://dx.doi.org/10.1108/13598540310463387 http://dx.doi.org/10.1016/j.jom.2004.08.007 http://dx.doi.org/10.1108/13555850810844896 http://dx.doi.org/10.1016/s0925-5273(00)00033-5 http://polipapers.upv.es/index.php/ijpme pme i j international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.14734 received: 2020-12-03 accepted: 2021-04-12 a comparison of topsis, grey relational analysis and copras methods for machine selection problem in the food industry of turkey özcan, s.a1, çelik, a.k.a2 ardahan university, faculty of economics and administrative sciences, department of quantitative methods, 75002, ardahan, turkey. a1 samiozcan@ardahan.edu.tr, a2 alikemalcelik@ardahan.edu.tr abstract: the paper aims to compare the results of the selection/choice of cream separators by using multi-criteria decisionmaking methods in an integrated manner for an enterprise with a dairy processing capacity of 80 to 100 tons per day operating in the turkish food sector. a total of 7 alternative products and 7 criteria for milk processing were determined. criterion weights were calculated using entropy method and then integrated into topsis (technique for order preference by similarity to ideal solutions), gra (grey relational analysis) and copras (complex proportional assessment) methods. sensitivity analyses were carried out on the results obtained from the three methods to check for their reliability. at the end of the study, similar alternative and appropriate results were found from the topsis and copras methods. however, different alternative but appropriate or suitable results were obtained from the gra method. sensitivity analysis of the three methods showed that all the methods used were valid. in the review of available and related literature, very few studies on machine selection in the dairy and food sector in general were found. for this reason, it is thought that the study will contribute to the decision-making process of companies in the dairy sector in their choice of machinery selections. as far as is known, this paper is the first attempt in extant literature to compare in an integrated manner the results of topsis, copras and gra methods considered in the study. key words: machine selection, decision making, topsis, grey relational analysis, copras. 1. introduction national or international manufacturing companies are investing heavily in their product portfolios so as to be successful in highly competitive markets. companies make these investments not only to enter new markets, but also to transform or rejuvenate already existing markets within which they serve. this is important because it is vital for manufacturing companies to develop their skills to survive fierce competition from other competitors. in order for manufacturing companies to achieve this, they must work with the right machines. however, choosing the right machine is always critical, difficult and complex (aloini et al., 2014). choosing an unsuitable or inappropriate machine will negatively affect the entire production system. owing to this, selecting a suitable machine for a particular production system is vital for the sustainability of the entire production system. in addition, outputs of the production system such as ratio, quality and cost are generally directly proportional to the choice of machine selected and applied in the production system (ayağ and özdemir, 2006). since the cost of investment in machinery is a huge burden on companies, they are very sensitive and careful in their choice of machine to spend on. today, the tendency of companies to make evaluations based on scientific methods instead of making intuitive decisions with only knowledge and experience is increasing rapidly. in recent years, to cite this article: özcan, s., çelik, a.k. (2021). a comparison of topsis, grey relational analysis and copras methods for machine selection problem in the food industry of turkey. international journal of production management and engineering, 9(2), 81-92. https://doi.org/10.4995/ijpme.2021.14734 int. j. prod. manag. eng. (2021) 9(2), 81-92creative commons attribution-noncommercial-noderivatives 4.0 international 81 https://orcid.org/0000-0002-2605-6526 mailto:samiozcan@ardahan.edu.tr mailto:alikemalcelik@ardahan.edu.tr http://creativecommons.org/licenses/by-nc-nd/4.0/ companies have been using multi-criteria decisionmaking methods, with its inherently complex structures, for machine selection decisions. multicriteria decision-making can generally be defined as a collection of methods used to choose, sort or classify two or more alternatives, taking into account the quantitative and/or qualitative criteria that often conflict each other. according to the 2019 dairy market research published by the united nations food and agriculture organization (united nations fao), in 2018 global milk production in india, turkey, the european union, pakistan, the united states and due to production expansions in argentina, reached 843 million tons which represents an increment of 2.2% compared to 2017. the report also emphasized the widespread use of integrated dairy production systems in turkey as one of the reasons for the increased productivity. according to the organization for economic cooperation and development (oecd) and fao’s 2019 agricultural outlook, the experienced increase in dairy production is estimated to reach 981 million tons by 2028 at an increasing rate of 1.7%. the increase in production is directly related to the increase in consumption. again, in the report of oecd and fao, milk consumption is predicted to increase by more than 1%. according to estimates from fao’s 2019 food outlook report, turkey with an estimated production output of about 24 million tons is ranked 8th on the world dairy production ranking. accordingly, it is possible to say that both national and international dairy processing plants in turkey occupy an important position in the international dairy processing competition. the study aimed to compare the different mcdm methods considered to solve problems of selecting cream separator for a plant with dairy processing capacity of 80 to 100 tons per day operating in turkey. the cream separator is used in the standardization phase, which is one of the most important stages in dairy pre-processing. in the standardization phase, milk is separated into two forms as cream and skimmed milk (chandan, 2008). the separated cream and skimmed milk are processed according to the oil content of other products produced in the facility. for the purpose of this study, the topsis (technique for order preference by similarity to ideal solutions), gra (grey relational analysis) and copras (complex proportional assessment) methods were selected due to the similarities in the basic underlying ideas of these methods. however, gra method was preferred in the normalization process as it has a different approach compared to the other two methods. in the second section of the study is the review of the relevant literature in relation to the study and brief information about the methods used in the analysis is provided in the third section. analysis made within the scope of the study are included in the fourth section and the results of the analysis are evaluated in the conclusion section. 2. literature review there are many studies in the literature on using multi-criteria decision making (mcdm) methods to determine the choice of machine selections. when these studies are examined, it is possible to see that machine selection problems in different sectors are addressed. on the one hand, earlier studies (özgen et al., 2011; kumru and kumru 2015; özceylan et al., 2016; wu et al., 2016; kabak and dağdeviren, 2017; camcı et al., 2018) confirm that companies operating in the manufacturing sector have benefited from mcdm methods for various machine selection problems. on the other hand, some other studies (clarke et al., 1990; samanta et al., 2002, alpay and ihpar, 2018; štirbanović et al., 2019) also provided evidence that mcdm techniques are applied in the selection of machines used in the mining industry. ulubeyli and kazaz (2009), yazdani-chamzini and yakhchali (2012), temiz and çalış (2017) and uğur (2017) demonstrated usage of mcdm methods in the construction industry. similarly ertuğrul and güneş (2007), vatansever and kazançoğlu (2014), and ertuğrul and öztaş (2015) also showed that the method is preferred in solving machine selection problems in the textile industry. aloini et al. (2014), özdağoğlu et al. (2017) and çakır (2018) evaluated the machine selection decisions of companies operating in different areas within the food industry using mcdm methods. in addition, prior studies (yılmaz and dağdeviren, 2010, 2011; paramasivam et al., 2011; taha and rostam, 2011; datta et al., 2013; karim and karmaker, 2016) also used different mcdm methods for the selection of machines with applicability in more than one sector. very few studies addressing multi-criteria decision making methods not only for machine selection problems but also in other areas in the food sector int. j. prod. manag. eng. (2021) 9(2), 81-92 creative commons attribution-noncommercial-noderivatives 4.0 international özcan & çelik 82 http://creativecommons.org/licenses/by-nc-nd/4.0/ can be found. for instance, gurmeric et al. (2013) investigated multi-criteria decision making methods in determining the optimum aroma level in terms of vanilla, strawberry and cocoa for prebiotic pudding. karaman et al. (2014) also using multicriteria decision making methods considered the evaluation of different ratios of ice cream mixes in terms of physicochemical, bioactive and sensory terms. in a similar study, ozturk et al. (2014) also applied multi-criteria decision making methods to determine physiochemical properties of the mixtures in mellorine dessert and the functional and sensory properties of mellorine enriched with vegetable juices in different concentrations. doğan et al. (2016) used multi-criteria decision making techniques in determining the fat content in hot chocolate and increasing the biofunctional properties of butter using fiber concentrates. 3. methodology the presence of more than one criterion in mcdm problems causes different perspectives and complex information to emerge. the main purpose of the mcdm method is to help decision makers organize and synthesize such information more comfortably in decision making, to minimize the potential for post-decision remorse by being satisfied with all criteria (belton and stewart, 2002). many methods have been developed for solving multi-criteria decision making problems. brief information about the methods considered in this study is presented below. 1.1 weighting with entropy method the entropy method is expressed as a measure of uncertainty about a random variable (zhang et al., 2011). the primary steps of the entropy method can be briefly described as follows (deng et al., 2000). 1: each criterion in the decision matrix in equation (1) is normalized as specified in equation (2). (1) (2) in equation (2), xij denotes the real value of each alternative, while pij denotes the normalized form values for each criterion. a normalized decision matrix in the shape of the equation specified below is obtained after solving equation (2). (3) 2: in the light of information contained in the normalized decision matrix in equation (3), entropy values (ej) for each criterion are calculated using equation (4) as specified as follows. (4) in equation (4), k is calculated as 1/ln(n) and is a constant which guarantees 0 ≤ ej ≤ 1. in equation (4), ej denotes the amount of information for a certain criterion. when the entropy value is smaller, then the importance of the criterion on decision making process becomes higher (wu et al., 2011). in other words, entropy value shows the uncertainty of information on a criterion. the uncertainty decreases when the information values for the criteria are close to each other. therefore, the entropy value takes a small value accordingly. 3: the degree of divergence (dj) of the average information contained in each criterion can be calculated as follows: dj = 1-ej (5) the degree of differentiation (dj) refers to the contrast intensity of the information in the criterion. accordingly, as the value of the dj criterion increases, the importance of the criterion in problem solving increases (wang and lee, 2009). in other words, it is the degree of difference between the information belonging to the criteria. it has an inverse relationship with its entropy value. 4: the last stage in the entropy method is where criterion weights are calculated. this calculation can be accomplished in equation (6): (6) in equation (6), wj shows the weight values of the criteria. the important point to be considered here is the rule that the sum of all wj values (w1 + w2 + … + wn) should be equal to 1. int. j. prod. manag. eng. (2021) 9(2), 81-92creative commons attribution-noncommercial-noderivatives 4.0 international a comparison of topsis, grey relational analysis and copras methods for machine selection problem in the food industry of turkey 83 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.1. topsis method hwang and yoon (1980) developed the topsis method based on the concept that the chosen alternative should have the shortest possible distance from the positive ideal solution and the farthest distance away from the negative ideal solution. the steps of the topsis method (seçme et al., 2009) are briefly presented below. 1: the decision matrix is normalized using equation (7). (7) in equation (7), rij captures the normalized value and i in xij is the numerical value of the alternative in accordance to the criteria j. 2: a weighted normalized decision matrix is obtained by multiplying the normalized matrix by the weights of the criteria (wj). (8) 3: ideal solution (maximum value, a*) and negative ideal solution (minimum value, a–) are determined. a*={v1*, v2*,…, v3*} (9) a– ={v1-, v2-,…, v3-} (10) 4: the distance between each alternative is calculated by using n-dimensional euclidean distance as follows: (11) (12) where, di* symbolizes the positive ideal separation measure and disymbolizes the negative ideal separation measure. 5: the closeness coefficient (cci) of each alternative is calculated with the following equation: (13) 6: at the end of the analysis, the alternatives are ranked by comparing the cci values and subsequently a decision is made. in the last step, the calculated cci values are listed in ascending order. the alternatives are ranked such that the alternative with the largest cci value is the optimum alternative. 3.2. gra method gra is a method of analysis that measures the relationship among matrix elements based on the difference of similarity or difference of development trends among these elements (feng and wang, 2000). the calculation procedures and steps of the gra method can be described as follows (wu and peng, 2016). 1: after the decision matrix is created as in equation (1), the series that make up the matrix in the decision problem with i rows (i = 1, 2, …, m) and j columns (j = 1, 2,…, n), according to the status of benefit, cost and nominality is normalized as follows. benefit-oriented criterion: if the criterion in the series has the property “the larger value is better”, the i, rows (i = 1, 2,…., m) and j, the columns (j = 1, 2,…, n) the following normalization procedure is applied. (14) cost-oriented criterion: if the criterion in the series has the property “the smaller value is better”, i, rows (i = 1, 2,…., m) and j columns (j = 1, 2,…., n) the following normalization procedure is applied. (15) nominality criterion: if the criterion in the series has a value such as x0b (ie if it has the property “the nominal value is better”) the following normalization procedure is applied to i lines (i = 1, 2,… columns, m) and j (j = 1, 2,…, n) columns. (16) at the end of the calculation processes and steps, a normalized matrix of the form specified below is obtained. int. j. prod. manag. eng. (2021) 9(2), 81-92 creative commons attribution-noncommercial-noderivatives 4.0 international özcan & çelik 84 http://creativecommons.org/licenses/by-nc-nd/4.0/ (17) 2: for each criterion, a reference is determined using the normalization matrix x'(0). (18) in equation (18), x'1j(0) expresses the jth reference value and for each criteria it is obtained by the largest normalization value. 3: the difference ∆ij(0) between the reference series x'(0) and normalized values is calculated using equation (19) and as in equation (20), an absolute value matrix is created. ∆ij(0)=|x'(0)-x'ij| (19) (20) 4: the grey relational coefficients γij(0) are calculated with the help of the equation specified below in accordance to the absolute value matrix. (21) where δ is expressed as the distinguished coefficient. δ∈[0,1] bound, however, it is generally accepted to be 0.5. 5: with the help of the equation (22), the grey relational degrees (γi) are calculated. subject to (22) wj in equation (22) indicates the weight of the jth criterion. the condition in the equation states that the sum of the weights of all criteria should be 1. the grey relationship degrees obtained as a result of equation (22) are ranked in descending order and the alternative with the greatest grey relationship degree is determined as the optimum alternative. 3.3. copras method an examination of the basic underyling idea of the copras method where preference for alternatives are based on ideal and negative ideal solutions can be thought of as similar to that of the topsis method (feizabadi et al., 2017). the calculation steps of the copras method can generally be explained as follows (zavadskas et al., 2004): 1: creation of weighted normalized decision making matrix. (23) where wj in equation (23) represents the weight of the j criterion. 2: weighted normalized indices are summed up. at this stage, the maximization s+j or minimization s–j aspects of the criteria are taken into account. index totals, m; to show the number of criteria are calculated as follows. (24) 3: relative significance values of alternatives (qi), s–min; minimum s–j is obtained with the help of the following equation. (25) alternatives are then ranked according to their relative significance. the alternative of highest relative importance is determined as the optimum alternative. 4. results in addressing the decision problem, first, the most important criteria to consider when choosing a cream separator were determined. in this context, a total of 20 companies engaged in the manufacturing of cream separator in turkey were selected. sales managers of the selected companies were contacted via e-mail and/or phone call and feedback was received from a total of 9 companies. the criteria were created by blending both criteria presented by the sales managers and the opinions of the production manager of the dairy processing plant. in this way, 7 criteria were int. j. prod. manag. eng. (2021) 9(2), 81-92creative commons attribution-noncommercial-noderivatives 4.0 international a comparison of topsis, grey relational analysis and copras methods for machine selection problem in the food industry of turkey 85 http://creativecommons.org/licenses/by-nc-nd/4.0/ determined: cream separation performance (c1), drum discharge volume (c2), drum turnover (c3), energy consumption (c4), weight (c5), price (c6) and number of cream separators (c7). after determining the criteria proposals were sent to cream separator manufacturing firms in turkey and also to the nine firms from whom feedbacks were received requesting for cream separators in their product portfolios. in the portfolios, offers for 7 alternative machines from 5 companies that have cream separators suitable for application in their dairy processing plant were submitted. the offers received and the features of the cream separator in the portfolio of the suppliers are summarized in table 1. 4.1. calculation of criterion weights by entropy method entropy method was used to determine the criterion weights. calculation of the criterion weights by the entropy method enables more reliable results by using objective weightings instead of weighting criteria subjectively. the criteria weights obtained from calculations using the entropy method were presented in table 2. a thorough look at the criterion weights reveals that the most important criterion in choosing cream separator is c2, drum discharge volume, criterion with a weight of 34.42%. this criterion is followed by c6 with a weight of 21.48%, c7 with a weight of 16.72% and c1 with a weight of 11.86%. in the ranking of criterion weights, the last three criteria were c4 with a weight of 8.6%, c5 with a weight of 3.68% and c3 with a weight of 3.23%. table 1. alternatives and their associated properties according to criteria (decision matrix). criteria alternatives c1 c2 c3 c4 c5 c6 c7 (+) (+) (+) (-) (-) (-) (-) a1 10 3.5 7.7 15 1.3 39 2 a2 20 6 6.8 20 1.6 65 1 a3 15 9 5.05 30 1.6 60 2 a4 18 15 6.2 18.5 1.3 68 1 a5 10 15 5.7 18.5 1.5 87 2 a6 20 18 6.2 18.5 1.8 120 1 a7 12.5 15 5.1 15 1.1 42 2 criteria with (+) sign have beneficial characteristics while those with the (-) sign are determined as criteria with cost characteristics. table 2. criterion weights calculated by entropy method. c1 c2 c3 c4 c5 c6 c7 wj 0.1186 0.3442 0.0323 0.086 0.0368 0.2148 0.1672 4.2. calculations by topsis method topsis method was initially used to sort the alternatives taken from cream separator suppliers for the dairy processing plant where the application was made and to make the final decision. the criteria contained in the decision matrix as shown in table 1 were calculated using topsis method the results the calculation processes are presented in the tables below. in the first stage of the topsis calculation process, data collected from different sources were normalized using equation (7). then, normalized weighted values were calculated by multiplying the normalized values by the criterion weights using equation (8). in the next stage of the topsis method, ideal (a*) and negative ideal (a–) solutions were determined with elements in the weighted normalized decision matrix, depending on whether the criteria were of benefit or cost oriented. after this calculation process, the n-dimensional euclidean distance between each alternative was calculated using equation (11) and equation (12) following that positive separation (d*) and negative separation (d-) measurements were determined. table 3. normalized values in topsis method. c1 c2 c3 c4 c5 c6 c7 a1 0.2422 0.1042 0.4718 0.2847 0.3334 0.2009 0.4588 a2 0.4843 0.1786 0.4166 0.3795 0.4104 0.3348 0.2294 a3 0.3632 0.2679 0.3094 0.5693 0.4104 0.3090 0.4588 a4 0.4359 0.4466 0.3799 0.3511 0.3334 0.3502 0.2294 a5 0.2422 0.4466 0.3492 0.3511 0.3847 0.4481 0.4588 a6 0.4843 0.5359 0.3799 0.3511 0.4617 0.6180 0.2294 a7 0.3027 0.4466 0.3125 0.2847 0.2821 0.2163 0.4588 table 4. weighted normalized values in topsis method. c1 c2 c3 c4 c5 c6 c7 a1 0.0287 0.0359 0.0152 0.0245 0.0123 0.0432 0.0767 a2 0.0574 0.0615 0.0135 0.0327 0.0151 0.0719 0.0384 a3 0.0431 0.0922 0.0100 0.0490 0.0151 0.0664 0.0767 a4 0.0517 0.1537 0.0123 0.0302 0.0123 0.0752 0.0384 a5 0.0287 0.1537 0.0113 0.0302 0.0142 0.0963 0.0767 a6 0.0574 0.1844 0.0123 0.0302 0.0170 0.1328 0.0384 a7 0.0359 0.1537 0.0101 0.0245 0.0104 0.0465 0.0767 int. j. prod. manag. eng. (2021) 9(2), 81-92 creative commons attribution-noncommercial-noderivatives 4.0 international özcan & çelik 86 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 5. ideal (a*) and negative (a-) ideal solution values in topsis method. c1 c2 c3 c4 c5 c6 c7 a* 0.0574 0.1844 0.0152 0.0245 0.0104 0.0432 0.0384 a– 0.0287 0.0359 0.0100 0.0490 0.0170 0.1328 0.0767 table 6. positive (d*) and negative (d-) separation measures in topsis method. a1 a2 a3 a4 a5 a6 a7 d* 0.0574 0.1844 0.0152 0.0245 0.0104 0.0432 0.0384 d– 0.0287 0.0359 0.0100 0.0490 0.0170 0.1328 0.0767 4.3. calculations by gra method in the gra method normalisation processes are just as those observed in the topsis method. however, in addition to the purpose of normalizing the data collected from different sources, it was more convenient to normalize the data after standardizing data in a small range since the elements in the decision matrix were values drawn from data in wide ranges. normalization processes in the gra method were carried out with equation (14) and equation (15). after the normalization matrix was obtained, the reference series and absolute value matrix were created with the help of equation (18) in accordance with the benefit or cost characteristics of the criteria. after this calculation process, the gra relational coefficients matrix was calculated with the help of equation (21). in this calculation process, the separator coefficient (δ) was taken as 0.5, as in many other studies in the literature (tosun, 2006; sharma and yadava, 2011; guo and sun, 2016; sun, 2014). the results of the calculations made by the gra method are presented in table 7 to table 9 as shown below. table 7. normalization values in gra method. c1 c2 c3 c4 c5 c6 c7 a1 0.0000 0.0000 1.0000 1.0000 0.7143 1.0000 0.0000 a2 1.0000 0.1724 0.6604 0.6667 0.2857 0.6790 1.0000 a3 0.5000 0.3793 0.0000 0.0000 0.2857 0.7407 0.0000 a4 0.8000 0.7931 0.4340 0.7667 0.7143 0.6420 1.0000 a5 0.0000 0.7931 0.2453 0.7667 0.4286 0.4074 0.0000 a6 1.0000 1.0000 0.4340 0.7667 0.0000 0.0000 1.0000 a7 0.2500 0.7931 0.0189 1.0000 1.0000 0.9630 0.0000 table 8. reference series and absolute value table in gra method. c1 c2 c3 c4 c5 c6 c7 a1 1.0000 1.0000 0.0000 1.0000 0.7143 1.0000 0.0000 a2 0.0000 0.8276 0.3396 0.6667 0.2857 0.6790 1.0000 a3 0.5000 0.6207 1.0000 0.0000 0.2857 0.7407 0.0000 a4 0.2000 0.2069 0.5660 0.7667 0.7143 0.6420 1.0000 a5 1.0000 0.2069 0.7547 0.7667 0.4286 0.4074 0.0000 a6 0.0000 0.0000 0.5660 0.7667 0.0000 0.0000 1.0000 a7 0.7500 0.2069 0.9811 1.0000 1.0000 0.9630 0.0000 table 9. gra relational coefficients matrix. c1 c2 c3 c4 c5 c6 c7 a1 0.0395 0.1147 0.0323 0.0287 0.0152 0.3333 1.0000 a2 0.1186 0.1296 0.0192 0.0369 0.0234 0.4241 0.3333 a3 0.0593 0.1535 0.0108 0.0860 0.0234 0.4030 1.0000 a4 0.0847 0.2434 0.0151 0.0340 0.0152 0.4378 0.3333 a5 0.0395 0.2434 0.0129 0.0340 0.0198 0.5510 1.0000 a6 0.1186 0.3442 0.0151 0.0340 0.0368 1.0000 0.3333 a7 0.0474 0.2434 0.0109 0.0287 0.0123 0.3418 1.0000 finally, the grey relation degrees (γi) are calculated and presented in table 11 along with the results of other methods. 4.4. calculations by copras method in relation to the purpose of the study, the results obtained with the copras method were found as follows. the first step in copras application is the creation of a weighted normalized matrix. the weighted normalized matrix created as a result of the calculations made with the copras method applied in the study is given in table 10. table 10. weighted normalized values in copras method. c1 c2 c3 c4 c5 c6 c7 a1 0.0112 0.0148 0.0058 0.0095 0.0047 0.0174 0.0304 a2 0.0225 0.0253 0.0051 0.0127 0.0058 0.0290 0.0152 a3 0.0169 0.0380 0.0038 0.0190 0.0058 0.0268 0.0304 a4 0.0202 0.0633 0.0047 0.0117 0.0047 0.0304 0.0152 a5 0.0112 0.0633 0.0043 0.0117 0.0054 0.0389 0.0304 a6 0.0225 0.0760 0.0047 0.0117 0.0065 0.0536 0.0152 a7 0.0141 0.0633 0.0039 0.0095 0.0040 0.0188 0.0304 after the normalization process, weighted normalized indexes (s+j and s–j) were summed according to the maximization and minimization criteria and int. j. prod. manag. eng. (2021) 9(2), 81-92creative commons attribution-noncommercial-noderivatives 4.0 international a comparison of topsis, grey relational analysis and copras methods for machine selection problem in the food industry of turkey 87 http://creativecommons.org/licenses/by-nc-nd/4.0/ the relative importance (qi) of the alternatives was calculated. qi values as well as the ranked order of alternatives are presented in table 11. 4.5. ranking of alternatives the closeness coefficients calculated according to the three methods used in the study, their grey relational degrees, their relative importance and the ranked order of the alternatives accordingly are summarized in table 11 below. table 11. cci values and ranked alternatives in topsis method. topsis gra copras cci sıralama γi sıralama qi sıralama a1 0.3738 7 0.4692 7 0.1137 7 a2 0.3968 6 0.4746 6 0.1340 6 a3 0.4529 5 0.5869 3 0.1206 5 a4 0.7554 1 0.5422 5 0.1702 1 a5 0.6146 4 0.6352 2 0.1377 4 a6 0.6357 3 0.8193 1 0.1615 3 a7 0.7331 2 0.5834 4 0.1623 2 at the end of the analysis of topsis and copras, it was concluded that the most suitable cream separator for a facility with a dairy processing capacity of 80 to 100 tons per day is the a4 alternative. the order of other alternatives is in the form a4 > a7 > a6 > a5 > a3 > a2 > a1. the result of gra method demonstrated that the most suitable cream separator is the a6 alternative. the order of other alternatives following the gra method is of the form a6 > a5 > a3 > a7 > a4 > a2 > a1. 4.6. sensitivity analysis in order to analyze the sensitivity of the results, the binary replacement method as used in the literature by önüt et al. (2009), kang et al. (2012), pang and bai (2013), nguyen et al. (2014), ahmed et al. (2019) was used. alternative sequences obtained by changing the weights of each criterion were examined. with a total number of 7 criteria and pairwise comparisons, 21 (7!/((7-2)!×2!)) different results were calculated. the graphs obtained by changing the criterion weights are shown as follows: when the sensitivity analysis graphs for topsis and copras methods are examined together, it can be seen that there are no serious variations in the ranking of the a4 alternative. in figure 4.2, when the part of the a6 alternative is examined, it is seen that the order of the a6 alternative is generally the same as the criteria weights change. with this result, it is possible to state that the most suitable alternative found using the gra method is valid. when the results of the sensitivity analysis are evaluated together, it is possible to say that the results obtained for all three methods are consistent within themselves. figure 4.1. sensitivity analysis of alternatives according to topsis method. figure 4.2. sensitivity analysis of alternatives according to gra method. int. j. prod. manag. eng. (2021) 9(2), 81-92 creative commons attribution-noncommercial-noderivatives 4.0 international özcan & çelik 88 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5. conclusion alternative is the most suitable alternative according to topsis and copras methods, however, the most suitable alternative in the gra method is a6. differences in normalization processes can be thought of as the main reason why the results found with the gra method are different from the results found with the topsis and copras methods. the topsis method is heavily influenced by the choice of normalization techniques used (pavličić, 2001; shih et al., 2007; çelen, 2014; vafaei et al., 2018). in addition, study conducted by chatterjee and chakraborty (2014) demonstrated that the two methods operate with different normalization techniques and pointed that out as the reason for topsis and gra techniques giving different results. while the topsis method uses the vector normalization technique, the gra method operates with the max-min normalization technique, which is one of the linear normalization techniques. in addition, another reason why gra and topsis and copras methods give different results is that gra takes into account the criterion aspects (positive, negative, nominal) in the normalization process. antucheviciene et al. (2012) stated that even if the normalization methods affect the final ranking results, the results of topsis and copras methods are very close to each other. stanujkic et al. (2013) explained this situation as being more affected by the criteria weights of both methods. although topsis and copras methods use different normalization techniques, in both methods calculations are made on alternative results, unlike gra, where calculations are made in the overlaps of alternatives. a clear recommendation of cream separator for dairy processing plant as discussed within the scope of the study can be made in the light of the power of the results obtained from the analyses conducted. the main reason for this is that, it can be demonstrated to companies producing machinery for the dairy sector to produce machines according to customer demands and also to equip standard machines with similar technical features. however, considering the fact that topsis and copras methods take the alternatives into consideration, it can be said that a4 alternative will be preferred. in such cases, it is suggested that experts’ opinions be sought in the evaluation of the two alternatives. however, when the criteria discussed in the study are examined, it can be said that between a4 and a6 alternatives, a4 alternative is more suitable for the business. drum speed, energy consumption and number of machines to be purchased are the same for both alternatives. it can be suggested that the results obtained by topsis and copras methods can be applied because the investment that the enterprise will make for the cream separator is a significant limitation for the enterprise. in this case, the advantage of topsis and copras methods will be used to evaluate over alternatives. apart from the criteria determined in this study, new applications of the methods such as recently proposed range target-based criteria and interval data model of topsis (jahan et al. 2021) can be made for machine selection according to different technical features. however, the same technical features can be used by changing the methods used in the study. also, ahp, saw, expert opinion etc. techniques can be used to re-determined criteria weights and analyzes can be performed. references ahmed, m., qureshi, m.n., mallick, j., kahla, n.b. (2019). selection of sustainable supplementary concrete materials using osm-ahptopsis approach. advances in materials science and engineering, 2019, 1-12. https://doi.org/10.1155/2019/2850480 aloini, d., dulmin, r., mininno, v. (2014). a peer if-topsis based decision support system for packaging machine selection. expert systems with applications, 41(5), 2157-2165. https://doi.org/10.1016/j.eswa.2013.09.014 figure 4.3. sensitivity analysis of alternatives according to copras method. int. j. prod. manag. eng. (2021) 9(2), 81-92creative commons attribution-noncommercial-noderivatives 4.0 international a comparison of topsis, grey relational analysis and copras methods for machine selection problem in the food industry of turkey 89 https://doi.org/10.1155/2019/2850480 https://doi.org/10.1016/j.eswa.2013.09.014 http://creativecommons.org/licenses/by-nc-nd/4.0/ alpay, s., ihpar, m. (2018). equipment selection based on two different fuzzy multi criteria decision making methods: fuzzy topsis and fuzzy vikor. open geosciences, 10(1), 661-677. https://doi.org/10.1515/geo-2018-0053 antucheviciene, j., zavadskas, e.k., zakarevičius, a. (2012). ranking redevelopment decisions of derelict buildings and analysis of ranking results. economic computation and economic cybernetics studies and research, 46(2), 37-63. retrieved june 08, 2020 from http:// www.ecocyb.ase.ro/22012/edmundas%20zavadskas%20_da_.pdf ayağ, z., özdemir, r.g. (2006). a fuzzy ahp approach to evaluating machine tool alternatives. journal of intelligent manufacturing, 17(2), 179-190. https://doi.org/10.1007/s10845-005-6635-1 belton, v., stewart, t.j. (2002). multiple criteria decision analysis: an integrated approach. berlin: kluwer academic publishers. camcı, a., temur, g.t., beşkese, a. (2018). cnc router selection for smes in woodwork manufacturing using hesitant fuzzy ahp method. journal of enterprise information management, 31(4), 529-549. https://doi.org/10.1108/jeim-01-2018-0017 çakır, s. (2018). an integrated approach to machine selection problem using fuzzy smart-fuzzy weighted axiomatic design. journal of intelligent manufacturing, 29(7), 1433-1445. https://doi.org/10.1007/s10845-015-1189-3 çelen, a. (2014). comparative analysis of normalization procedures in topsis method: with an application to turkish deposit banking market. informatica, 25(2), 185-208. https://doi.org/10.15388/informatica.2014.10 chandan, r.c. (2008). dairy processing and quality assurance: an overview. ramesh c. chandan, arun kilara, nagendra shah (eds.), in dairy processing and quality assurance (pp. 1-40). new jersey: wiley-blackwell. chatterjee, p., chakraborty, s. (2014). investigating the effect of normalization norms in flexible manfacturing sytem selection using multicriteria decision-making methods. journal of engineering science and technology review, 7(3), 141-150. https://doi.org/10.25103/ jestr.073.23 clarke, m.p., denby, b., schofield, d. (1990). decision making tools for surface mine equipment selection. mining science and technology, 10(3), 323-335. https://doi.org/10.1016/0167-9031(90)90530-6 datta, s., sahu, n., mahapatra, s. (2013). robot selection based on grey-multimoora approach. grey systems: theory and application, 3(2), 201-232. https://doi.org/10.1108/gs-05-2013-0008 deng, h., yeh, c.h., willis, r. j. (2000). inter-company comparison using modified topsis with objective weights. computers and operations research, 27(10), 963-973. https://doi.org/10.1016/s0305-0548(99)00069-6 doğan, m., aslan, d., aktar, t., sarac, m.g. (2016). a methodology to evaluate the sensory properties of instant hot chocolate beverage with different fat contents: multi-criteria decision-making techniques approach. european food research and technology, 242(6), 953-966. https://doi.org/10.1007/s00217-015-2602-z ertuğrul, i̇., güneş, m. (2007). fuzzy multi-criteria decision making method for machine selection. p. melin, o. castillo, e.g. ramirez, j. kacprzyk and w. pedrycz (eds.), in analysis and design of intelligent systems using soft computing techniques (pp. 638-648). berlin, germany: springer. https://doi.org/10.1007/978-3-540-72432-2_65 ertuğrul, i̇., öztaş, t. (2015). the application of sewing machine selection with the multi-objective optimization on the basis of ratio analysis method (moora) in apparel sector. textile and apparel, 25(1), 80-85. retrieved may 17, 2020 from https://dergipark.org.tr/tr/pub/ tekstilvekonfeksiyon/issue/23647/251887 fao. (2019a). dairy market review. fao publishing, rome. fao. (2019b). food outlook biannual report on global food markets. fao publishing, rome. feizabadi, a., doolabi, m.s., sadrnezhaad, s.k., zafarani, h.r., doolabi, d.s. (2017). mcdm selection of pulse parameters for best tribological performance of cr–al2o3 nano-composite co-deposited from trivalent chromium bath. journal of alloys and compounds, 727, 286-296. https://doi.org/10.1016/j.jallcom.2017.08.098 feng, c.m., wang, r.t. (2000). performance evaluation for airlines including the consideration of financial ratios. journal of air transport management, 6(3), 133-142. https://doi.org/10.1016/s0969-6997(00)00003-x guo, x., sun, z. (2016). a novel evaluation approach for tourist choice of destination based on grey relation analysis. scientific programming, 2016, 1-10. https://doi.org/10.1155/2016/1812094 gurmeric, v.e., dogan, m., toker, o.s., senyigit, e., ersoz, n.b. (2013). application of different multi-criteria decision techniques to determine optimum flavour of prebiotic pudding based on sensory analyses. food and bioprocess technology, 6(10), 2844-2859. https://doi. org/10.1007/s11947-012-0972-9 hwang, c.l., yoon, k. (1980). multiple attribute decision making methods and applications: a state-of-the-art survey. new york: springerverlag. jahan, a., yazdani, m., edwards, k.l. (2021). topsis-rtcid for range target-based criteria and interval data. international journal of production management and engineering, 9(1), 1-14. https://doi.org/10.4995/ijpme.2021.13323 kabak, m., dağdeviren, m. (2017). a hybrid approach based on anp and grey relational analysis for machine selection. technical gazette, 24(supplement 1), 109-118. https://doi.org/10.17559/tv-20141123105333 kang, h.y., lee, a.h.i., yang, c.y. (2012). a fuzzy anp model for supplier selection as applied to ic packaging. journal of intelligent manufacturing, 23(5), 1477-1488. https://doi.org/10.1007/s10845-010-0448-6 int. j. prod. manag. eng. (2021) 9(2), 81-92 creative commons attribution-noncommercial-noderivatives 4.0 international özcan & çelik 90 https://doi.org/10.1515/geo-2018-0053 http://www.ecocyb.ase.ro/22012/edmundas%20zavadskas%20_da_.pdf http://www.ecocyb.ase.ro/22012/edmundas%20zavadskas%20_da_.pdf https://doi.org/10.1007/s10845-005-6635-1 https://doi.org/10.1108/jeim-01-2018-0017 https://doi.org/10.1007/s10845-015-1189-3 https://doi.org/10.15388/informatica.2014.10 https://doi.org/10.25103/jestr.073.23 https://doi.org/10.25103/jestr.073.23 https://doi.org/10.1016/0167-9031(90)90530-6 https://doi.org/10.1108/gs-05-2013-0008 https://doi.org/10.1016/s0305-0548(99)00069-6 https://doi.org/10.1007/s00217-015-2602-z https://doi.org/10.1007/978-3-540-72432-2_65 https://dergipark.org.tr/tr/pub/tekstilvekonfeksiyon/issue/23647/251887 https://dergipark.org.tr/tr/pub/tekstilvekonfeksiyon/issue/23647/251887 https://doi.org/10.1016/j.jallcom.2017.08.098 https://doi.org/10.1016/s0969-6997(00)00003-x https://doi.org/10.1155/2016/1812094 https://doi.org/10.1007/s11947-012-0972-9 https://doi.org/10.1007/s11947-012-0972-9 https://doi.org/10.4995/ijpme.2021.13323 https://doi.org/10.17559/tv-20141123105333 https://doi.org/10.1007/s10845-010-0448-6 http://creativecommons.org/licenses/by-nc-nd/4.0/ karaman, s.,toker, ö.s., yüksel, f., çam, m., kayacier, a., dogan, m. (2014). physicochemical, bioactive, and sensory properties of persimmon-based ice cream: technique for order preference by similarity to ideal solution to determine optimum concentration. journal of dairy science, 97(1), 97-110. https://doi.org/10.3168/jds.2013-7111 karim, r., karmaker, c.l. (2016). machine selection by ahp and topsis methods. american journal of industrial engineering, 4(1), 7-13. https://doi.org/10.12691/ajie-4-1-2 kumru, m., kumru, p.y. (2015). a fuzzy anp model for the selection of 3d coordinate-measuring machine. journal of intelligent manufacturing, 26(5), 999-1010. https://doi.org/10.1007/s10845-014-0882-y nguyen, h.t., dawal, s. z. md., nukman, y., aoyama, h. (2014). a hybrid approach for fuzzy multi-attribute decision making in machine tool selection with consideration of the interactions of attributes. expert systems with applications, 41(6), 3078-3090. https://doi. org/10.1016/j.eswa.2013.10.039 oecd/fao. (2019). oecd-fao agricultural outlook 2019-2028. oecd publishing, paris. önüt, s., kara, s.s., işik, e. (2009). long term supplier selection using a combined fuzzy mcdm approach: a case study for a telecommunication company. expert systems with applications, 36(2), 3887-3895. https://doi.org/10.1016/j.eswa.2008.02.045 özceylan, e., kabak, m., dağdeviren, m. (2016). a fuzzy-based decision making procedure for machine selection problem. journal of intelligent and fuzzy systems, 30(3), 1841-1856. https://doi.org/10.3233/ifs-151895 özdağoğlu, a., yakut, e., bahar, s. (2017). machine selection in a dairy product company with entropy and saw methods integration. faculty of economics and administrative sciences journal, 32(1), 341-359. https://doi.org/10.24988/deuiibf.2017321605 özgen, a., tuzkaya, g., tuzkaya, u.r., özgen, d. (2011). a multi-criteria decision making approach for machine tool selection problem in a fuzzy environment. international journal of computational intelligence systems, 4(4), 431-445. https://doi.org/10.1080/18756891 .2011.9727802 ozturk, g., dogan, m., toker, o.s. (2014). physicochemical, functional and sensory properties of mellorine enriched with different vegetable juices and topsis approach to determine optimum juice concentration. food bioscience, 7, 45-55. https://doi.org/10.1016/j. fbio.2014.05.001 pang, b., bai, s. (2013). an integrated fuzzy synthetic evaluation approach for supplier selection based on analytic network process. journal of intelligent manufacturing, 23(5), 163-174. https://doi.org/10.1007/s10845-011-0551-3 paramasivam, v., senthil, v., ramasamy, n.r. (2011). decision making in equipment selection: an integrated approach with digraph and matrix approach, ahp and anp. the international journal of advanced manufacturing technology, 54(9-12), 1233-1244. https://doi. org/10.1007/s00170-010-2997-4 pavličić, d.m. (2001). normalisation affects the results of madm methods. yugoslav journal of operations research, 11(2), 251-265. retrieved may 6, 2020 from http://scindeks.ceon.rs/article.aspx?artid=0354-02430102251p samanta, b., sarkar, b., mukherjee, s.k. (2002). selection of opencast mining equipment by a multi-criteria decision-making process. mining technology, 111(2), 136-142. https://doi.org/10.1179/mnt.2002.111.2.136 seçme, n.y., bayrakdaroğlu, a., kahraman, c. (2009). fuzzy performance evaluation in turkish banking sector using analytic hierarchy process and topsis. expert systems with applications, 36(9), 11699-11709. https://doi.org/10.1016/j.eswa.2009.03.013 sharma, a., yadava, v. (2011). optimization of cut quality characteristics during nd:yag laser straight cutting of ni-based superalloy thin sheet using grey relational analysis with entropy measurement. materials and manufacturing processes, 26(12), 1522-1529. https:// doi.org/10.1080/10426914.2011.551910 shih, h. s., shyur, h.j., lee, e.s. (2007). an extension of topsis for group decision making. mathematical and computer modelling, 45(78), 801-813. https://doi.org/10.1016/j.mcm.2006.03.023 stanujkic, d., đorđević, b., đorđević, m. (2013). comparative analysis of some prominent mcdm methods: a case of ranking serbian banks. serbian journal of management, 8(2), 213-241. https://doi.org/10.5937/sjm8-3774 štirbanović, z., stanujkić, d., miljanović, i., milanović, d. (2019). application of mcdm methods for flotation machine selection. minerals engineering, 137, 140-146. https://doi.org/10.1016/j.mineng.2019.04.014 sun, c.c. (2014). combining grey relation analysis and entropy model for evaluating the operational performance: an empirical study. quality and quantity, 48(3), 1589-1600. https://doi.org/10.1007/s11135-013-9854-0 taha, z., rostam, s. (2011). a fuzzy ahp-ann-based decision support system for machine tool selection in a flexible manufacturing cell. international journal of advanced manufacturing technology, 57(5-8), 719-733. https://doi.org/10.1007/s00170-011-3323-5 temiz, i., çalış, g. (2017). selection of construction equipment by using multi-criteria decision making methods. procedia engineering, 196, 286-293. https://doi.org/10.1016/j.proeng.2017.07.201 tosun, n. (2006). determination of optimum parameters for multi-performance characteristics in drilling by using grey relational analysis. the international journal of advanced manufacturing technology, 28(5-6), 450-455. https://doi.org/10.1007/s00170-004-2386-y uğur, l.o. (2017). application of the vikor multi-criteria decision method for construction machine buying. journal of polytechnic, 20(4), 879-885. https://doi.org/10.2339/politeknik.369058 int. j. prod. manag. eng. (2021) 9(2), 81-92creative commons attribution-noncommercial-noderivatives 4.0 international a comparison of topsis, grey relational analysis and copras methods for machine selection problem in the food industry of turkey 91 https://doi.org/10.3168/jds.2013-7111 https://doi.org/10.12691/ajie-4-1-2 https://doi.org/10.1007/s10845-014-0882-y https://doi.org/10.1016/j.eswa.2013.10.039 https://doi.org/10.1016/j.eswa.2013.10.039 https://doi.org/10.1016/j.eswa.2008.02.045 https://doi.org/10.3233/ifs-151895 https://doi.org/10.24988/deuiibf.2017321605 https://doi.org/10.1080/18756891.2011.9727802 https://doi.org/10.1080/18756891.2011.9727802 https://doi.org/10.1016/j.fbio.2014.05.001 https://doi.org/10.1016/j.fbio.2014.05.001 https://doi.org/10.1007/s10845-011-0551-3 https://doi.org/10.1007/s00170-010-2997-4 https://doi.org/10.1007/s00170-010-2997-4 http://scindeks.ceon.rs/article.aspx?artid=0354-02430102251p https://doi.org/10.1179/mnt.2002.111.2.136 https://doi.org/10.1016/j.eswa.2009.03.013 https://doi.org/10.1080/10426914.2011.551910 https://doi.org/10.1080/10426914.2011.551910 https://doi.org/10.1016/j.mcm.2006.03.023 https://doi.org/10.5937/sjm8-3774 https://doi.org/10.1016/j.mineng.2019.04.014 https://doi.org/10.1007/s11135-013-9854-0 https://doi.org/10.1007/s00170-011-3323-5 https://doi.org/10.1016/j.proeng.2017.07.201 https://doi.org/10.1007/s00170-004-2386-y https://doi.org/10.2339/politeknik.369058 http://creativecommons.org/licenses/by-nc-nd/4.0/ ulubeyli, s., kazaz, a. (2009). a multiple criteria decision-making approach to the selection of concrete pumps. journal of civil engineering and management, 15(4), 369-376. https://doi.org/10.3846/1392-3730.2009.15.369-376 vafaei, n., ribeiro, r.a., camarinha-matos, l.m. (2018). data normalisation techniques in decision making: case study with topsis method. international journal of information and decision sciences, 10(1), 19-38. https://doi.org/10.1504/ijids.2018.090667 vatansever, k., kazançoğlu, y. (2014). integrated usage of fuzzy multi criteria decision making techniques for machine selection problems and an application. international journal of business and social science, 5(9), 12-24. https://doi.org/10.1504/ijids.2018.090667 wang, t.c., lee, h.d. (2009). developing a fuzzy topsis approach based on subjective weights and objective weights. expert systems with applications, 36(5), 8980-8985. https://doi.org/10.1016/j.eswa.2008.11.035 wu, j., sun, j., liang, l., zha, y. (2011). determination of weights for ultimate cross efficiency using shannon entropy. expert systems with applications, 38(5), 5162-5165. https://doi.org/10.1016/j.eswa.2010.10.046 wu, w., peng, y. (2016). extension of grey relational analysis for facilitating group consensus to oil spill emergency management. annals of operations research, 238(1-2), 615-635. https://doi.org/10.1007/s10479-015-2067-2 wu, z., ahmad, j., xu, j. (2016). a group decision making framework based on fuzzy vikor approach for machine tool selection with linguistic information. applied soft computing, 42, 314-324. https://doi.org/10.1016/j.asoc.2016.02.007 yazdani-chamzini, a., yakhchali, s.h. (2012). tunnel boring machine (tbm) selection using fuzzy multicriteria decision making methods. tunnelling and underground space technology, 30, 194-204. https://doi.org/10.1016/j.tust.2012.02.021 yılmaz, b., dağdeviren, m. (2010). comparative analysis of promethee and fuzzy promethee methods in equipment selection problem. journal of the faculty of engineering and architecture of gazi university, 25(4), 811-826. retrieved may 6, 2020 from https://avesis. gazi.edu.tr/yayin/989e528e-9184-4d8e-8970-fccfabbbed73/comparative-analysis-of-promethee-and-fuzzy-promethee-methods-inequipment-selection-problem yılmaz, b., dağdeviren, m. (2011). a combined approach for equipment selection: f-promethee method and zero–one goal programming. expert systems with applications, 38(9), 11641-11650. https://doi.org/10.1016/j.eswa.2011.03.043 zavadskas, e.k., kaklauskas, a., banaitis, a., kvederyte, n. (2004). housing credit access model: the case for lithuania. european journal of operational research, 155(2), 335-352. https://doi.org/10.1016/s0377-2217(03)00091-2 zhang, h., gu, c.l., gu, l. w., zhang, y. (2011). the evaluation of tourism destination competitiveness by topsis and information entropy: a case in the yangtze river delta of china. tourism management, 32(2), 443-451. https://doi.org/10.1016/j.tourman.2010.02.007 int. j. prod. manag. eng. (2021) 9(2), 81-92 creative commons attribution-noncommercial-noderivatives 4.0 international özcan & çelik 92 https://doi.org/10.3846/1392-3730.2009.15.369-376 https://doi.org/10.1504/ijids.2018.090667 https://doi.org/10.1504/ijids.2018.090667 https://doi.org/10.1016/j.eswa.2008.11.035 https://doi.org/10.1016/j.eswa.2010.10.046 https://doi.org/10.1007/s10479-015-2067-2 https://doi.org/10.1016/j.asoc.2016.02.007 https://doi.org/10.1016/j.tust.2012.02.021 https://avesis.gazi.edu.tr/yayin/989e528e-9184-4d8e-8970-fccfabbbed73/comparative-analysis-of-promet https://avesis.gazi.edu.tr/yayin/989e528e-9184-4d8e-8970-fccfabbbed73/comparative-analysis-of-promet https://avesis.gazi.edu.tr/yayin/989e528e-9184-4d8e-8970-fccfabbbed73/comparative-analysis-of-promet https://doi.org/10.1016/j.eswa.2011.03.043 https://doi.org/10.1016/s0377-2217(03)00091-2 https://doi.org/10.1016/j.tourman.2010.02.007 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.16494 received: 2021-10-19 accepted: 2022-07-25 quantitative supply chain segmentation model for dynamic alignment rafael alves ferreira a1*, lucas a. s. santos a2, kleber f. espôsto a3 a production engineering department, são carlos school of engineering, university of são paulo, brazil. a1* rafael.alves.ferreira@alumni.usp.br, a2 lucasalves02@gmail.com, a3 kleberesposto@usp.br abstract: companies deal with different customer groups which makes it important to define the service level precisely and improve customer service through different supply chain strategies for each group. an alternative to deal with imprecision related to the segmentation processes suggested by either the leagile or the dynamic alignment schools is the application of fuzzy set theory. the objective of this work is to develop a quantitative model that uses the fuzzy set theory and, based on sales data, assesses the company’s supply chain(s) to facilitate managers’ decision-making processes to achieve dynamic alignment. it was possible to identify the supply chains that serve the client groups evaluated, providing answers faster than the analysis proposed by the models found in the literature. the application in two real situations validated the model. the results obtained were able to indicate managerial actions such as the establishment of clear processes for agile and campaign supply chains in one case or the improvement in the information sharing with a group of clients and thus moving from a fully flexible to the agile supply chain. to the best of our knowledge, this study is the first that aims to segment quantitatively supply chains in a company applying fuzzy set theory, providing a novel approach to align operations and supply chain strategy dynamically. key words: supply chain segmentation, supply chain management, fuzzy inference system. 1. introduction supply chain management is a powerful approach to achieving competitive advantages and superior business performance in companies. the significance of supply chain management has increased due to the breaking of national and international borders, which grew the global movement of products and services (routroy & shankar, 2015). the set of relationships that typify the connections between organizations in a chain enables competitive advantages through cost reduction and market differentiation (christopher & towill, 2002). it is vital to align the supply chain strategy with the countless demands of the markets. the “one size fits all” concept, that is, a unique strategy for the entire supply chain, does not apply to today’s competitive environment (gattorna, 2015; godsell et al., 2011; simchi-levi et al., 2013). yet, it is still possible to find examples of companies working in competitive and diversified markets that assume that the demands and purchasing behaviors are homogeneous (hjort et al., 2016). there are two schools for segmenting the supply chain (godsell et al., 2011). the first school, called lean-agile, is based mainly on the characteristics of the products. at this school, christopher and towill (2000) introduce a supply chain segmentation model based on the combination of five market characteristics: duration of life cycle, the time window for delivery, volume, variety, and demand variability. based on these characteristics, strategies to cite this article: ferreira, r.a., santos, l.a.s., espôsto, k.f. (2022). quantitative supply chain segmentation model for dynamic alignment. international journal of production management and engineering, 10(2), 99-113. https://doi.org/10.4995/ijpme.2022.16494 int. j. prod. manag. eng. (2022) 10(2), 99-113creative commons attribution-noncommercial-noderivatives 4.0 international 99 https://orcid.org/0000-0001-6139-4523 http://creativecommons.org/licenses/by-nc-nd/4.0/ for the supply chain are suggested. the authors advocate the existence of the three different types of chains: “lean”, “agile”, and the combination of these two, called “lean-agile”. for the authors of the first school, lean supply chains aim to serve the customer efficiently and at a low cost by removing all waste from the processes (mason-jones et al., 2000). agile supply chains seek to integrate reconfigurable resources in an environment of intense exchange of information to meet customers’ needs in uncertain markets (yusuf et al., 2014). lean-agile chains are the combination of the two previous options. the second school, called dynamic alignment, aims to formalize the connection between marketing and supply chain strategy with a strategic approach. for this school, supply chain segmentation is the result of understanding the customer it serves. consequently, the suggested segmentation method is based on customer behavior (christopher & gattorna, 2005; gattorna, 2015; gattorna & walters, 1996). a member of this school, (gattorna, 2015) proposed a model called dynamic alignment, which suggests five types of supply chains that work concurrently in companies, creating an environment of multiple supply chains. although the two schools of thought are consolidated in the scientific community and have a vast theoretical content, they still have a large gap concerning models that are not abstract and normative, with little empirical evidence (godsell et al., 2011). the supply chain segmentation literature focuses on proposing segmentation criteria (e.g., volume, demand variability, variety). thus, it lacks a robust solution for determining the segmentation parameters of these criteria (fichtinger et al., 2019). an approach to cope with the vagueness of criteria proposed in the literature is the fuzzy set theory, proposed by (zadeh, 1965). it has been applied in the decision-making process, as it can extract relevant results even based on uncertain or inaccurate input data (banaeian et al., 2016; fu et al., 2017; lima junior et al., 2014; santos et al., 2017). supply chain performance management applies fuzzy logic inference rules to build models based on leading metrics and if-then scenarios (ganga & carpinetti, 2011). analogously, supply chain segmentation is affected by uncertainty mainly due to the vagueness intrinsic to the evaluation of qualitative criteria and the imprecise weighing of different criteria by different decision-makers. according to renganath and suresh (2017), decision-making techniques based on fuzzy logic are considered one of the best ways to work with imprecise or vague problems, in which the selection of alternatives is abstract. the fuzzy set theory provides appropriate language wherein imprecise criteria can be handled, and it can integrate the qualitative and quantitative factors in the evaluation process (lima junior et al., 2013). using fuzzy inference in decision-making has an advantage, the assumption of the approximate reasoning concept in the inference process that models human reasoning. another advantage is capturing the experts’ judgments in the knowledge base (zimmermann, 1987). gattorna (2015) proposes a very detailed supply chain segmentation model based on customer buying behavior, supported by qualitative nature information, resulting in five possible supply chain types. on the other hand, christopher and towill (2000) suggest quantitative variables can be used as input to the segmentation process. this paper proposes a model that parametrizes a mamdani fuzzy inference system (mamdani & assilian, 1975) to translate the input variables proposed by christopher and towill (2000) (time window for delivery, volume, and demand variability) into gattorna’s (2015) supply chain types: collaborative supply chain, lean supply chain, agile supply chain, campaign supply chain, and fully flexible supply chain. in this context, this paper’s main contribution is the development of a model that, combining the two supply chain segmentation schools, and supported by a fuzzy inference system, aims to reduce the abstraction and improve applicability. the present work uses sales data as input to assess and characterize the company’s supply chains and, thus, facilitate the management decision-making processes towards the achievement of its supply chain’s dynamic alignment. our research follows the empirical analytic modeling proposed by bertrand and fransoo (2002) ensuring a model that fits the observations and actions in the chosen environment. we use empirical data for the numerical analysis supplied by a multinational industrial machinery manufacturer and a brazilian fertilizer mixing industry that serves two very distinct customer segments, thus, allowing the model assessment using real-world delivery lead-time, volume, and demand variability. int. j. prod. manag. eng. (2022) 10(2), 99-113 creative commons attribution-noncommercial-noderivatives 4.0 international ferreira et al. 100 http://creativecommons.org/licenses/by-nc-nd/4.0/ this paper has been divided into six sections. in sect. 2, we elaborate on the existing literature that addresses two schools of supply chain segmentation and the application of fuzzy set theory to it. sect. 3 touches on the details of the methodology used, including the novel model proposal. the model application is explored in sect. 4 and it is followed by the discussion in sect 5. finally sect. 6 focuses on our conclusions and plans for future research. 2. literature review 2.1. the dwv3 criteria a significant contribution of the lean-agile school was the introduction of the dwv3 market classification criteria by christopher and towill (2000) since the five criteria – duration of life cycle, time window for delivery, volume, variety, and variability – can be applied to target the supply chain strategy (godsell et al., 2011). christopher et al. (2009) suggest the use of dwv3 variables to classify products with similar characteristics. the principal output of this group is a clear definition of the requirements of each demand channel, together with the specific objectives to maximize competitiveness in each segment. godsell et al. (2011) suggest a model that uses the variables volume and variability to define which supply chain strategy would be ideal for products, lean or agile. with these variables, together with the application of filters, the authors claim that it is possible to define the demand profile. 2.2. the dynamic alignment changing the focus from product to market, gattorna (2015) suggests the concept of dynamic alignment, which is, that supply chains are constantly changing and require dynamism to remain aligned with the customers’ needs. christopher and gattorna (2005) advocate the advantages of segmenting the supply chain along with buyer behavior, unfortunately, most organizations use internal parameters that provide little indication of how the customers want to buy products and services. the key point for developing a supply chain that satisfactorily meets customers’ needs is to understand the mix of the behavioral segments for a given market (gattorna, 2015). this task, once performed, provides the possibility to segment customers so that the appropriate value propositions are made to meet this scenario of multiple supply chains (christopher & gattorna, 2005). from these five purchasing behaviors, gattorna (2015) proposed a model with five types of chain: continuous replenishment, lean, agile, campaign, and fully flexible. the model proposed by gattorna (2015), despite being very normative, is inaccurate in the context of decision-making. the factors that are evaluated for choosing the best strategy for a given supply chain are based on opinions, therefore burdened with uncertainty and imprecision. the theory of fuzzy sets is an effective method for dealing with linguistic variables and decision-making in complex situations (celik et al., 2015; giri et al., 2022) and will be described in the next topic. 2.3. application of fuzzy set theory to the supply chain segmentation process the bibliographic reference reveals that the task of segmenting the supply chain is not trivial. there are many normative models and few studies with applications in real environments (godsell et al., 2011). the lean-agile school has three possibilities for configuring the supply chain, proposing variables related to products as a source of information for segmenting the supply chain. the school of dynamic alignment proposes a more robust supply chain segmentation model, nevertheless, the application of the model to the reality of companies is complex, mainly because it is based on questions that result in vague and inaccurate answers. this work sought to combine the best characteristics of the two supply chain segmentation schools. a viable option was sought in the literature to deal with imprecision, applying the theory of fuzzy sets to the dwv3 variables and taking advantage of the quality of the characterizations of the chains proposed by the dynamic alignment school. 2.3.1. fuzzy theory conventional system analysis techniques are inherently inadequate for dealing with humanistic systems or any system in which complexity can be compared to a humanistic system (zadeh, 1973). the fuzzy sets theory (zadeh, 1965) has been used to model the decision-making process based on uncertain or inaccurate information, such as the judgments of managers or decision-makers (lima int. j. prod. manag. eng. (2022) 10(2), 99-113creative commons attribution-noncommercial-noderivatives 4.0 international quantitative supply chain segmentation model for dynamic alignment 101 http://creativecommons.org/licenses/by-nc-nd/4.0/ junior et al., 2014; banaeian et al., 2016). many applications of fuzzy sets theory can be found in the literature in the fields of supply chain management, including production management, quality, and cost-benefit analysis, in which the unavailability of complete information, accurate references, and reliable data make them even more interesting (kumar et al., 2013). rule-based models play a central role in fuzzy modeling while it captures relationships among fuzzy variables and provides a mechanism to link linguist descriptions of systems with their computational realizations (pedrycz & gomide, 2007), these are called the fuzzy inference systems, discussed in the next topic. 2.3.2. fuzzy inference system the objective of a fuzzy inference system (fis) is to control complex processes through human experience (zimmermann, 2001). complex systems involve various types of inaccuracies and represent a huge challenge for the development of models. this is especially true for the areas of business, finance, and management systems, which involve a large number of factors, some with socio-psychological nature (bojadziev & bojadziev, 2007). the inference rules connect the input variables with the output variables and are based on the description of the terms of the fuzzy linguistic variable. the input variables represent the conditions and the output variables represent the consequences of the control rule (zimmermann, 2001; bojadziev & bojadziev, 2007). mamdani and assilian (1975) proposed a system capable of making boolean logic more flexible when describing the states of processes through linguistic variables and using these variables as inputs to the inference rules. the mamdani fuzzy logic controller can be used as a decision support system (garcía et al., 2013) and has been used in a variety of problems such as life cycle analysis, supplier selection, and supplier performance evaluation (lima junior et al., 2013; santos et al., 2017; limajunior & carpinetti, 2020). the inference process begins with the assignment of terms for the input variables. in a system for evaluating customer satisfaction, for example, possible base variables are product delivery time, percentage of discount, and satisfaction. the “delivery time” variable could consist of the terms “low”, “medium” and “high”. the rules connect the input variables with the output variables and are based on the description of the state of the variable, obtained by defining the terms of the linguistic variables. 3. methodology in this work, we propose a quantitative model developed based on fuzzy logic to, based on selected input variables, enable the segmentation of the supply chain. bertrand and fransoo (2002) define that quantitative models are based on a set of variables that vary in a specific domain, whereas quantitative and causal relationships are defined between these variables. to the authors, the main concern of the researcher is to obtain solutions within a defined model and to make sure that these solutions can provide a greater understanding of the problem structure, as defined in the model. for the development of the work, it was necessary to adapt the conceptual models dwv3, proposed by christopher et al. (2009), and the dynamic alignment model, proposed by gattorna (2015). the combination of these models served as a reference for modeling the rule base of the fis. the variables from the dwv3 model were selected as the fis input variables since they are numerical information highly available in most of the enterprise resource planning systems (erp) available in the market. the criteria selected for the evaluation of the supply chain to be input variables were: the individual volume of each sku sold to each customer, the variability of demand for each sku per customer, and the average delivery time for the customer to be evaluated. the other criteria, although they have an impact on the supply chain strategy, were not considered of primary importance. the stage of the life-cycle of a particular product does not directly influence demand planning, however, the volume and variability that the product presents at a given point in the life-cycle do influence (godsell et al., 2011). for the model, the variety of items is also of secondary importance, since the inference system assesses the demand for each item individually. the identification of the variables can be seen in table 1. the inference system was developed to recognize the customer service patterns that have been being int. j. prod. manag. eng. (2022) 10(2), 99-113 creative commons attribution-noncommercial-noderivatives 4.0 international ferreira et al. 102 http://creativecommons.org/licenses/by-nc-nd/4.0/ implemented. therefore, the output variable of the selected fuzzy inference systems was the type of supply chain that is serving a specific customer or group of selected customers, based on the dynamic alignment model supply chain types, proposed by gattorna (2015). five tables (table 2 to table 6) were proposed showing the characteristics’ relationship of the time window for delivery, volume, and variability variables, based on the dwv3, applied to each of the five types of the supply chain of the dynamic alignment model. the linguistic variables selected to model the problem were “low”, “medium” and “high”. 3.1. fuzzy inference system the inference system adopted in the model is the mamdani type since it is widely used and tested in the literature, according to bojadziev and bojadziev (2007) and zimmermann (2001). the mamdani fuzzy logic controller was used because it is indicated for decision support systems (garcía et al., table 1. explanation of inclusion or exclusion of dwv3 variables (source: adapted from godsell et al. (2011)). variable explanation duration of product life cycle (not included) for reasons of simplification of the model, the product life cycle has not been included. time window for delivery (included) the delivery time of the product is relevant to the need to use resources to meet the established deadline. volume (included) it has a direct impact on the supply chain strategy. variety (not included) the proposed model measures at the sku level, so it is not relevant. variability (included) it has a direct impact on the supply chain strategy. table 2. collaborative supply chain characteristics (source: proposed by the authors). strategic dimension characteristics source time window for delivery medium to high, the main value of the customer is delivery on the agreed date and not specifically the length of delivery time gattorna (2015, p. 204) christopher and towill (2000, p. 116) volume medium to low, since mature products tend to drop consumption gattorna (2015, p. 203) christopher and towill (2000, p. 117) variability low, highly predictable through the communication channel between supplier and customer. the product mix tends to be composed of mature products gattorna (2015, p. 203) christopher and towill (2000, p. 117) table 3. lean supply chain characteristics (source: proposed by the authors). strategic dimension characteristics source time window for delivery medium to high, the customer seeks a pre-established delivery time, despite not sharing information on demand gattorna (2015, p. 243) christopher and towill (2000, p. 116) volume medium to high, since customers looking for the lowest cost tend to seek gains of scale gattorna (2015, p. 243) christopher and towill (2000, p. 117) variability low, transactional-minded customers tend to buy mature, established products gattorna (2015, p. 241) christopher and towill (2000, p. 117) table 4. agile supply chain characteristics (source: proposed by the authors). strategic dimension characteristics source time window for delivery low, the demand nature requires a quick response gattorna (2015, p. 281) christopher and towill (2000, p. 116) volume low, lack of planning shrinks delivery times required to meet demand and shrinks lot sizes gattorna (2015, p. 283) christopher and towill (2000, p. 117) variability high, dynamic minded customers tend to increase the range of choice. this large range generates high variability gattorna (2015, p. 283) christopher and towill (2000, p. 117) int. j. prod. manag. eng. (2022) 10(2), 99-113creative commons attribution-noncommercial-noderivatives 4.0 international quantitative supply chain segmentation model for dynamic alignment 103 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2013) and has been applied in an endless number of cases in companies (lima junior et al., 2013). the operator gamma was chosen in the implication relation because it allows a compensatory effect in the implication, getting closer to the result found by the human decision system. for aggregation, the mamdani model uses the maximum operator in aggregation. for the implementation of the model’s inference system, the fuzzytech 8.30c software was used. 3.2. identification of input variables the terms of the input fuzzy variables in the inference system were represented by triangular and trapezoidal membership functions since they are functions with great computational efficiency and are simpler to be parameterized (zimmermann, 2001). the number of terms for each variable is three, following the suggestion of (von altrock, 1997) who states that three terms are capable of simulating human thinking without generating the need for an excessive number of rules in the knowledge base of the inference system. the selected terms used were “low”, “medium” and “high”. the parameterization of the inference system must be performed based on the analysis of the input data and the experience of selected specialists in the company. the variable time window for delivery is represented by the average delivery time for a given item to the customer to be evaluated. the difference between the item’s billing date and the date of receipt of the customer’s order is used as the delivery time. the volume variable is represented by the total items purchased by the customer during the period to be evaluated. the variability was represented by the variation coefficient of each item purchased by the customer to be evaluated. the coefficient of variation is represented by the equation below, where σ represents the population standard deviation and μ represents the mean. cv=σ/µ (1) 3.3. identification of output variable the terms used as output variables were the names of the supply chains proposed by gattorna (2015): collaborative, lean, agile, campaign, and fully flexible. considering the pattern recognition characteristic, it was decided not to defuzzify the results. as a result, the highest degree of membership found in the output vector was considered, defining the greatest similarity with a given type of chain (von altrock, 1997; simões & shaw, 2007). table 5. campaign supply chain characteristics (source: proposed by the authors). strategic dimension characteristics source time window for delivery high, since there is the entire product design, purchasing, and manufacturing process, making it impossible to anticipate activities gattorna (2015, p. 324) christopher and towill (2000, p. 116) volume low, since the project's sale, in general, occupies a large part of the manufacturing capacity for every single project. gattorna (2015, p. 322) christopher and towill (2000, p. 117) variability high, each project has a specific customer and a different product design. gattorna (2015, p. 322) christopher and towill (2000, p. 117) table 6. fully flexible chain characteristics (source: proposed by the authors). strategic dimension characteristics source time window for delivery low, for crisis solution the sooner the demand is met, the better gattorna (2015, p. 355) christopher and towill (2000, p. 116) volume low, since demand will be met only once. in some cases, the prototype is the expected delivery gattorna (2015, p. 354) christopher and towill (2000, p. 117) variability high, it is impossible to predict what will be needed to resolve a possible crisis. gattorna (2015, p. 355) christopher and towill (2000, p. 117) int. j. prod. manag. eng. (2022) 10(2), 99-113 creative commons attribution-noncommercial-noderivatives 4.0 international ferreira et al. 104 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.4. definition of the inference rules base the rule base of the inference system was based on the propositions of characteristics of volume, variability, and time window for delivery for each type of supply chain proposed by (gattorna, 2015). tables 2 to 6 were used to support the definition of the knowledge base. the full inference system rule block can be found in the appendix. 3.5. the analytical model the proposed model can be represented analytically in figure 1. in the first step, the customer or group of customers to have its predominant supply chain type assessed is selected and it is defined which supply chain type has a better suit for them. in the second step, the customer or group of customers who will be evaluated for which type of supply chain are currently submitted is filtered. it is necessary to obtain the sales information of all the items supplied in the evaluated period, generating a table with the items and their respective data of volume, variability, and average delivery time. in the model’s third step, the list is processed by the fuzzy inference system and outputs the predominant type of supply chain that currently serves the selected customer. the fourth step of the proposed model is the analysis of the results obtained. at this point, decision-makers must identify the gap between the customer’s current supply chain and the type of chain most aligned with the company’s strategy. in this step, it is possible to recognize customers who are being over-served and customers who are not receiving the expected service. input variables can be used as an indication of actions to be taken to realign customers with the service strategy. 4. model application to evaluate the proposed model, it was applied to two companies, from different fields. empirical data for the numerical analysis and expert opinion was supplied by a multinational industrial machinery manufacturer, case 1, and a brazilian fertilizer mixing industry, case 2. 4.1. case 1 the model was applied to a multinational company in the field of manufacturing machinery and industrial equipment. the company operates in several business models, one of which is the manufacture of equipment for sale, developing and adapting projects according to customers’ needs, using the engineering to order (eto) strategy. it also caters to equipment maintenance applications that are already running on the clients using the make to order (mto) typology. the company attends sales orders for spare parts sold from stock using the make to stock (mts) typology and has machine rental businesses, providing services in the product-service system (pss) model. figure 1. supply chain alignment assessment model (source: proposes by the authors). int. j. prod. manag. eng. (2022) 10(2), 99-113creative commons attribution-noncommercial-noderivatives 4.0 international quantitative supply chain segmentation model for dynamic alignment 105 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.1.1. model parametrization to proceed with the input variables membership functions parameterization, specialists (see table 7) were chosen. those selected are experienced employees and have extensive knowledge of the company’s processes and customers. initially, the supply chain alignment assessment model proposed in the work was presented to the experts. the concepts of fuzzy logic were also briefly presented to enable them to collaborate in the parameterization of the membership functions shown in figures 2, figure 3, and figure 4. during this process, sales data for the past two years were consulted on the company’s erp system and tabulated on a spreadsheet. after analyzing the data collected in the system, a forum was opened for discussion and consensual definition of the parameters of the membership functions. 4.1.2. data collection and assessment of supply chains the group of experts suggested three groups of clients to be analyzed, considering data from the previous year. the first group of clients analyzed was the one that serves customers who bought machines in the previous year. the second is the group of clients that serves customers who consume spare parts and, finally, the group of clients that serves customers in the machine rental business. all data obtained were pre-validated by the company’s experts. a search was made for possible incomplete fields or with a large discrepancy in results. outliers have been removed to avoid distortion of the database. 4.1.3. evaluation of machinery manufacturing supply chain the evaluation of the machinery manufacturing supply chain started with the application of filters in sales orders to locate all orders for machines billed in the previous year. a list was obtained with all the units sold, with varied construction complexities. from the list of machines sold, the table was created with data on average delivery time, volume, and variability to serve as input data for the fuzzy figure 3. time window for delivery membership function for case 1 (source: proposed by the authors). figure 4. variability membership function for case 1 (source: proposed by the authors). table 7. expert description (source: proposed by the authors). education experience role field bachelor over 10 years manager purchase bachelor over 10 years purchaser purchase bachelor over 10 years planner planning bachelor 5-10 years supervisor sales figure 2. volume membership function for case 1 (source: proposed by the authors). int. j. prod. manag. eng. (2022) 10(2), 99-113 creative commons attribution-noncommercial-noderivatives 4.0 international ferreira et al. 106 http://creativecommons.org/licenses/by-nc-nd/4.0/ inference system. the result obtained by the system is shown in table 8. table 8. results from machines fis (source: proposed by the authors). supply chain type quantity of items agile 41 % campaign 50 % fully flexible 9 % to define the ideal strategy for the machinery manufacturing supply chain, experts concluded that a second segmentation should be carried out, between complex machines and simple machines. for the experts, equipment that requires many weeks of equipment design would have the ideal supply chain strategy as the campaign type, given that equipment design and manufacturing activities are concomitant for many weeks. the simpler equipment, on the other hand, has a short design time, so the agile supply chain would be the ideal strategy. in this case, by having a shorter design time, actions may be taken in a more organized way, since decisions regarding manufacturing processes are made based on a completed project. 4.1.4. evaluation of supply chain for spare parts sales in evaluating the supply chain for selling spare parts, the specialists selected a group with the five main customers of this supply chain. all orders for parts billed in the previous year were filtered, obtaining a list of 108 items sold. sales data were then processed to obtain a table with the input variables proposed in the model. the results obtained when processing the data in the fis are shown in table 9. table 9. results from spare parts fis (source: proposed by the authors). supply chain type quantity of items agile 81 % campaign 8 % fully flexible 11 % in the experts’ assessment, the ideal supply chain strategy for this group of customers is the agile supply chain, since the variability of ordering items is high. according to them, customer demand for specific replacement parts is intermittent, making it impossible to maintain large safety stocks, both because of the high cost and the risk of obsolescence. 4.1.5. evaluation of rental business supply chain in the evaluation of the supply chain for the rental business, all orders for parts billed as maintenance items from the previous year were filtered, obtaining a list with 1122 different items sent to the field. sales data were then processed to obtain a table with the input variables proposed in the model. the results obtained when processing the data in the fis are shown in table 10. the supply chain that serves the spare parts of the leased machines showed a predominantly agile service profile. this result is consistent, given that there is a large installed base of machines for customers with many variations of equipment models, consequently different spare parts. the company has as a competitive advantage a high overall equipment effectiveness (oee), thus there is an effort to quickly replace parts with customers. table 10. results from rental business fis (source: proposed by the authors). supply chain type quantity of items lean 1 % agile 72 % campaign 12 % fully flexible 15 % 4.1.6. case 1 discussion according to the company’s experts, the results obtained by the fis proved to be consistent. in the supply chain of the machinery manufactured by the company, two predominant types of the supply chain were found: the agile supply chain and the campaign supply chain. the machines served by the campaign-type supply chain are machines with greater complexity and long delivery times. the machines served by the agile supply chain are less complex. the engineering processes to which they are submitted are minor adjustments to adapt them to the needs of the customers. it was noticed that some replacement items were served by a supply chain of the fully flexible type due to a temporary movement in the type of supply chain. we also found some items that were served by a campaign-type supply chain that were items used in machinery refurbishing, to perform maintenance on key machine components. int. j. prod. manag. eng. (2022) 10(2), 99-113creative commons attribution-noncommercial-noderivatives 4.0 international quantitative supply chain segmentation model for dynamic alignment 107 http://creativecommons.org/licenses/by-nc-nd/4.0/ when analyzing the results of data processing by fis, it was noticed that 8 items, approximately 1%, had lean supply chain characteristics. from an investigation of the nature of these items, it was realized that they are items of consumption of equipment, such as lubricating oil and springs that suffer wear and tear, which represent a considerable cost for the company. 4.2. case 2 the second application of the model was in a fertilizer mixing industry located in a brazilian northeastern state. the fertilizer commercialized by the company is sold regionally and has as its main customers small and medium farmers, local resellers, and sugar cane processing companies. the company operates in two business models, one is fertilizer sales through agreements previously firmed and the other one is through direct sales to customers on a first-come-first-serve basis. both production strategies are mto. the company erp system was designed exclusively for the company by a local provider, and implemented eight years ago, with an extensive reliable database. this system controls all sales and buying orders, storing data on customers and suppliers, delivery time, and volume negotiated. the sales process is divided between direct sales and supply contracts. after the closing of the sale or the agreement, the orders are directed to the planning department, which verifies if the company can meet all the requirements of the closed sale. 4.2.1. model parametrization for the parametrization of the membership functions, the same process for parameters definition used in case 1 was followed. the selected team for this assessment was composed of one sales representative, one commercial manager, and the industrial director (see table 11). all selected employees were highly experienced with extensive knowledge of all the processes of the companies and their clients. table 11. expert description (source: proposed by the authors). education experience role field bachelor 5-10 years supervisor sales specialist over 10 years manager sales specialist over 10 years director production the membership functions are presented in figure 5, figure 6, and figure 7. figure 5. volume membership function for case 2 (source: proposed by the authors). figure 6. time window for delivery membership function for case 2 (source: proposed by the authors). figure 7. variability membership function for case 2 (source: proposed by the authors). 4.2.2. data collection and assessment of supply chains the team decided that two groups of clients should be analyzed to evaluate the supply chains. the first 0.0 0.5 1.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 µ(x) quantity low medium high 0.0 0.5 1.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 µ(x) quantity low medium high 0.0 0.5 1.0 0 0.3 0.5 0.7 µ(x) cv low medium high int. j. prod. manag. eng. (2022) 10(2), 99-113 creative commons attribution-noncommercial-noderivatives 4.0 international ferreira et al. 108 http://creativecommons.org/licenses/by-nc-nd/4.0/ group was composed of clients that firmed supply agreements in the previous year. the second is the group of clients who did not have firm supply agreements, composed of smaller clients with a buying behavior based on demand variation, most of them, local resellers. all obtained data were validated by the company’s experts before running the model. 4.2.3. evaluation of agreements’ supply chain the agreements’ supply chain is composed of clients who need high levels of predictability in terms of supply and price. these agreements are usually made with clients who need a high supply volume and certainty that they will receive their product in the exact time needed. the product is delivered within an agreed lead-time between the order input and the product dispatch. for the analyzed year, supply agreements were 37% of the company’s total demand. the evaluation of the agreements supply chain started with the application of filters to locate all orders that were originated from supply agreements from the previous years. the obtained list had 42 different fertilizers formulas sold during the period. from this list, the table was created with the data with a lead time for each product, volume, and variability to serve as input data for the fis. the result obtained is shown in table 12. based on the results, the ideal strategy for the agreements supply chain only can be tailored after subsequent segmentation. as the experts’ analysis, when supply agreements, which by its nature have higher volume, are agreed upon with few deliveries, campaign would be the ideal supply chain strategy type, given that the company must redirect its equipment almost exclusively to fulfill these orders, which are largely representative of the company’s total demand. the other agreements, in which the deliveries are divided into smaller orders and consequently more deliveries, the supply chain type based on the way that the demand was fulfilled would be the agile supply chain. 4.2.4. evaluation of resellers’ supply chain the resellers’ supply chain is composed of smaller clients, mostly resellers and small farmers, and clients who did not want to firm the supply agreements. these clients have a buying behavior based mostly on the demand variation and need the product on a much lower scale than the clients of the agreements supply chain. the evaluation of the resellers’ supply chain started with the application of filters to identify the orders from resellers and small buyers from the previous years. all the clients that were not served by agreements were defined as part of this supply chain since their buying behavior is mostly the same. the obtained list had 100 different fertilizers formulas sold during the period. from this list, the table was created with the data with a lead time for each product, volume, and variability to serve as input data for the fis. the result obtained is shown in table 13. table 13. results from resellers’ fis (source: proposed by the authors). supply chain type quantity of items agile 52 % campaign 1 % fully flexible 47 % based on the expert’s assessment, the results were consistent but further segmentation is needed to understand the results for this supply chain, between common formulas and seasonal formulas. common formulas are sold all over the year, with lower variability. seasonal formulas are sold in highly specific time windows, which are linked to the rainfall regimen and the specific crop for that period. 4.2.5. case 2 discussion the results provided by the fis were consistent according to the company’s experts. for the agreements’ supply chain, two predominant types of supply chain were found: the agile supply chain and the campaign supply chain. as the model proposes, the results of the supply chain are based on how the company attends the clients from different groups. for the company, the campaign supply chain is the right strategy for one part of the clients in the agreements’ supply chain, because table 12. results from agreement fis (source: proposed by the authors). supply chain type quantity of items agile 36 % campaign 42 % fully flexible 22 % int. j. prod. manag. eng. (2022) 10(2), 99-113creative commons attribution-noncommercial-noderivatives 4.0 international quantitative supply chain segmentation model for dynamic alignment 109 http://creativecommons.org/licenses/by-nc-nd/4.0/ of its characteristic small numbers of deliveries. however, the other part of the agreements’ supply chain is being served using the agile supply chain strategy although, according to the experts, using a collaborative supply chain strategy which would extend the delivery time to match the agreements. for the resellers’ supply chain, two main strategies were identified: the agile supply chain and the fully flexible supply chain. according to the experts, both strategies are correct and further segmentation is needed to understand these results, into common formulas, with an agile supply chain strategy, and seasonal formulas, with a fully flexible supply chain strategy. 5. discussion in both cases, the experts agreed that the model was able to detect precisely and segment the multiple supply chains within the companies’ systems according to their dwv3 data. in case 1, the company segmented its supply chain considering its business models and in a second step by client expenditure. they also have a list of key customers, defined by top management that has a preference in order fulfillment. in the company, all processes share the same resources and, the model detected two main supply chain types: agile and campaign. one of the purposes of supply chain segmentation is operational efficiency (wen et al., 2019) so the company could benefit from having a pattern to establish a clearer process for these two supply chain types. another opportunity is on the lean supply chain detected. although just a few skus, their value in the budget is high and they are products that don’t suffer from obsolescence. an alternative for them is to improve information sharing in the chain and move from a lean supply chain to a collaborative one. in case 2 the company segmented its supply chain based on the business models. the agreement supply chain has a big portion, 22%, of its orders classified on the fully flexible supply chains, the most expensive way of dealing with client orders. as the business model is based on formal agreements with clients the company has an opportunity to improve information sharing to increase demand predictability, making it possible to move to agile or campaign supply chains. another opportunity is finding the customers that could have their variability reduced to move them to a collaborative supply chain. on the resellers’ supply chain, due to the market conditions, the model pointed out that the segmentation in common or seasonal formulas is, at this moment, a good way to deal with orders. 6. conclusion the purpose of this work was to fill the gap regarding the complexity of applying supply chain segmentation models. the lean-agile school, despite suggesting practical ways of evaluating the supply chain for decision-making, only suggests three possibilities for the supply chain: lean; agile; or the leagile combination. this feature simplifies the view of the supply chain to the point of not segmenting it sufficiently, condensing types of chains with different characteristics within the same segment. on the other hand, the supply chain segmentation proposal of the dynamic alignment school is more robust, with more segmentation possibilities. however, this school suffers from excessive regulation, in addition to the imprecision inherent in its evaluation process, which is primarily qualitative and difficult to apply. according to godsell et al. (2011) the reasoning of the model in the two schools of thought on the supply chain design, lean and agile, and that of dynamic alignment, provides a holistic approach to the development of supply chain strategies. the model proposed in this work aimed to group the strengths of the two supply chain segmentation schools and sought, as an alternative to cope with the imprecision related to the segmentation process, the application of the theory of fuzzy sets using fuzzy inference and the development of an expert system. the system uses the perception of experts to create a knowledge base that is used in the data processing. the process of presenting the model to the companies’ experts was an important factor in the parameterization of the fuzzy inference system. it was necessary to carry out training so that the results obtained in the parameterization were assertive. the disadvantage of this process is that the lack of expert training could compromise the results. the query of data in the erp system, tabulation, and previous analysis favored the consensus on the parameterization of the fis. despite the apparent int. j. prod. manag. eng. (2022) 10(2), 99-113 creative commons attribution-noncommercial-noderivatives 4.0 international ferreira et al. 110 http://creativecommons.org/licenses/by-nc-nd/4.0/ complexity of the fuzzy logic, the specialists felt comfortable with the definition of the parameters, using the steps proposed by von altrock (1997) to choose the values of the membership function. future studies could investigate the inclusion of the neurofuzzy technique to automate membership functions definition. contemplating the proposal of this study, the model was able to assess the current service standards of a selected group of customers, based on a set of quantitative data available in most erp systems: the delivery time, volume, and variability. the quality and reliability of the erp database proved to be an important factor in the consistency of the results obtained. the evaluation of a customer’s current service level through a computational model collaborates with the seek for dynamic alignment. with the model, it was possible to identify the supply chains that serve the groups of clients evaluated, providing answers much faster than the analysis proposed by gattorna (2015). the model indicated managerial actions to realign the supply chain, for example, establishing a clearer process for agile and campaign supply chain, or fostering a collaborative supply chain in case 1. another possible managerial action detected by the model was to improve information sharing with a specific client group could increase demand predictability and create a collaborative supply chain in case 2. references banaeian, n., mobli, h., fahimnia, b., nielsen, i.e., & omid, m. (2016). green supplier selection using fuzzy group decision making methods: a case study from the agri-food industry. computers & operations research, 89, 337– 347. https://doi.org/10.1016/j.cor.2016.02.015 bertrand, j.w.m., & fransoo, j.c. (2002). operations management research methodologies using quantitative modeling. in international journal of operations & production management, 22(2), 241–264. https://doi.org/10.1108/01443570210414338 bojadziev, g., & bojadziev, m. (2007). fuzzy logic for business, finance, and management. (2 ed.). world scientific publishing co. pte. ltd. https://doi.org/10.1142/6451 celik, e., gul, m., aydin, n., gumus, a.t., & guneri, a.f. (2015). a comprehensive review of multi criteria decision making approaches based on interval type-2 fuzzy sets. knowledge-based systems, 85, 329–341. https://doi.org/10.1016/j. knosys.2015.06.004 christopher, m., & gattorna, j. (2005). supply chain cost management and value-based pricing. industrial marketing management, 34(2 spec. iss.), 115–121. https://doi.org/10.1016/j.indmarman.2004.07.016 christopher, m., & towill, d.r. (2000). marrying the lean and agile paradigms. proceedings of the euroma 2000 conference, 114–121. christopher, m., & towill, d.r. (2002). developing market specific supply chain strategies. the international journal of logistics management, 13(1), 1–14. https://doi.org/10.1108/09574090210806324 christopher, m., towill, d.r., aitken, j., & childerhouse, p. (2009). value stream classification. journal of manufacturing technology management, 20(4), 460–474. https://doi.org/10.1108/17410380910953720 fichtinger, j., wan-chuan, c., & yates, n. (2019). a joint network design and multi-echelon inventory optimisation approach for supply chain segmentation. international journal of production economics, 209(august 2017), 103–111. https://doi.org/10.1016/j.ijpe.2017.09.003 fu, x., zeng, x.j., luo, x., wang, d., xu, d., & fan, q.l. (2017). designing an intelligent decision support system for effective negotiation pricing: a systematic and learning approach. decision support systems, 96, 49–66. https://doi.org/10.1016/j.dss.2017.02.003 ganga, g.m.d., & carpinetti, l.c. r. (2011). a fuzzy logic approach to supply chain performance management. international journal of production economics, 134(1), 177–187. https://doi.org/10.1016/j.ijpe.2011.06.011 garcía, n., puente, j., fernández, i., & priore, p. (2013). supplier selection model for commodities procurement. optimised assessment using a fuzzy decision support system. applied soft computing journal, 13(4), 1939–1951. https://doi.org/10.1016/j.asoc.2012.12.008 gattorna, j. (2015). dynamic supply chains: how to design, build and manage people-centric value networks (3rd ed.). pearson education limited. gattorna, j., & walters, d.w. (1996). managing the supply chain: a strategic perspective (1st editio). palgrave. https://doi.org/10.1007/978-1-349-24841-4_1 int. j. prod. manag. eng. (2022) 10(2), 99-113creative commons attribution-noncommercial-noderivatives 4.0 international quantitative supply chain segmentation model for dynamic alignment 111 https://doi.org/10.1016/j.cor.2016.02.015 https://doi.org/10.1108/01443570210414338 https://doi.org/10.1142/6451 https://doi.org/10.1016/j.knosys.2015.06.004 https://doi.org/10.1016/j.knosys.2015.06.004 https://doi.org/10.1016/j.indmarman.2004.07.016 https://doi.org/10.1108/09574090210806324 https://doi.org/10.1108/17410380910953720 https://doi.org/10.1016/j.ijpe.2017.09.003 https://doi.org/10.1016/j.dss.2017.02.003 https://doi.org/10.1016/j.ijpe.2011.06.011 https://doi.org/10.1016/j.asoc.2012.12.008 https://doi.org/10.1007/978-1-349-24841-4_1 http://creativecommons.org/licenses/by-nc-nd/4.0/ giri, b.c., molla, m.u., & biswas, p. (2022). pythagorean fuzzy dematel method for supplier selection in sustainable supply chain management. expert systems with applications, 193. https://doi.org/10.1016/j.eswa.2021.116396 godsell, j., diefenbach, t., clemmow, c., towill, d., & christopher, m. (2011). enabling supply chain segmentation through demand profiling. international journal of physical distribution & logistics management, 41(3), 296–314. https://doi.org/10.1108/09600031111123804 hjort, k., lantz, b., ericsson, d., & gattorna, j. (2016). customer segmentation based on buying and returning behaviour: supporting differentiated service delivery in fashion e-commerce. in developments in logistics and supply chain management (pp. 153–169). palgrave macmillan uk. https://doi.org/10.1057/9781137541253_14 kumar, d., singh, j., singh, o.p., & seema. (2013). a fuzzy logic based decision support system for evaluation of suppliers in supply chain management practices. mathematical and computer modelling, 57(11–12), 2945–2960. https://doi.org/10.1016/j.mcm.2013.03.002 lima junior, f.r., osiro, l., & carpinetti, l.c. r. (2013). a fuzzy inference and categorization approach for supplier selection using compensatory and non-compensatory decision rules. applied soft computing journal, 13(10), 4133– 4147. https://doi.org/10.1016/j.asoc.2013.06.020 lima junior, f.r., osiro, l., & carpinetti, l.c. r. (2014). a comparison between fuzzy ahp and fuzzy topsis methods to supplier selection. applied soft computing journal, 21, 194–209. https://doi.org/10.1016/j.asoc.2014.03.014 lima junior, f.r., & carpinetti, l.c. r. (2020). an adaptive network-based fuzzy inference system to supply chain performance evaluation based on scor® metrics. computers and industrial engineering, 139(may 2019), 106191. https://doi.org/10.1016/j.cie.2019.106191 mamdani, e.h., & assilian, s. (1975). an experiment in linguistic synthesis with a fuzzy logic controller. international journal of man-machine studies, 7(1), 1–13. https://doi.org/10.1016/s0020-7373(75)80002-2 mason-jones, r., naylor, b., & towill, d.r. (2000). engineering the leagile supply chain. international journal of agile management systems, 2, 54–61. https://doi.org/10.1108/14654650010312606 pedrycz, w., & gomide, f. (2007). fuzzy systems engineering—toward human-centric computing. wiley. https://doi. org/10.1002/9780470168967 renganath, k., & suresh, m. (2017). supplier selection using fuzzy mcdm techniques: a literature review. 2016 ieee international conference on computational intelligence and computing research, iccic 2016. https://doi.org/10.1109/iccic.2016.7919590 routroy, s., & shankar, a. b. (2015). performance analysis of agile supply chain. international journal of manufacturing technology and management, 29(3–4), 180–210. https://doi.org/10.1504/ijmtm.2015.069255 santos, l.f. de o.m., osiro, l., & lima, r.h. p. (2017). a model based on 2-tuple fuzzy linguistic representation and analytic hierarchy process for supplier segmentation using qualitative and quantitative criteria. expert systems with applications, 79, 1339–1351. https://doi.org/10.1016/j.eswa.2017.02.032 simchi-levi, d., clayton, a., & raven, b. (2013). when one size does not fit all. mit sloan management review, 54(2), 15–17. simões, m.g., & shaw, i.s. (2007). controle e modelagem fuzzy (2 ed.). editora blucher. von altrock, c. (1997). fuzzy logic & neurofuzzy applications in business & finance. prentice hall. wen, x., choi, t.m., & chung, s.h. (2019). fashion retail supply chain management: a review of operational models. international journal of production economics, 207, 34–55. https://doi.org/10.1016/j.ijpe.2018.10.012 yusuf, y.y., gunasekaran, a., musa, a., dauda, m., el-berishy, n.m., & cang, s. (2014). a relational study of supply chain agility, competitiveness and business performance in the oil and gas industry. international journal of production economics, 147(part b), 531–543. https://doi.org/10.1016/j.ijpe.2012.10.009 zadeh, l.a. (1965). fuzzy sets. information and control, 8(3), 338–353. https://doi.org/10.1016/s0019-9958(65)90241-x zadeh, l.a. (1973). outline of a new approach to the analysis of complex systems and decision processes. ieee trans. syst., 3, 28–44. https://doi.org/10.1109/tsmc.1973.5408575 zimmermann, h.j. (1987). fuzzy sets, decision making and expert systems. kluwer academic publishers. https://doi.org/10.1007/978-94-009-3249-4 zimmermann, h.j. (2001). fuzzy set theory and its applications. in kluwer academic publishers, usa (4th ed.). kluwer academic. https://doi.org/10.1007/978-94-010-0646-0 int. j. prod. manag. eng. (2022) 10(2), 99-113 creative commons attribution-noncommercial-noderivatives 4.0 international ferreira et al. 112 https://doi.org/10.1016/j.eswa.2021.116396 https://doi.org/10.1108/09600031111123804 https://doi.org/10.1057/9781137541253_14 https://doi.org/10.1016/j.mcm.2013.03.002 https://doi.org/10.1016/j.asoc.2013.06.020 https://doi.org/10.1016/j.asoc.2014.03.014 https://doi.org/10.1016/j.cie.2019.106191 https://doi.org/10.1016/s0020-7373(75)80002-2 https://doi.org/10.1108/14654650010312606 https://doi.org/10.1002/9780470168967 https://doi.org/10.1002/9780470168967 https://doi.org/10.1109/iccic.2016.7919590 https://doi.org/10.1504/ijmtm.2015.069255 https://doi.org/10.1016/j.eswa.2017.02.032 https://doi.org/10.1016/j.ijpe.2018.10.012 https://doi.org/10.1016/j.ijpe.2012.10.009 https://doi.org/10.1016/s0019-9958(65)90241-x https://doi.org/10.1109/tsmc.1973.5408575 https://doi.org/10.1007/978-94-009-3249-4 https://doi.org/10.1007/978-94-010-0646-0 http://creativecommons.org/licenses/by-nc-nd/4.0/ appendix i – inference system rule block (source: proposed by the authors) antecedents consequents rule volume variability time window for delivery supply chain type 1 low low low collaborative 2 low low medium collaborative 3 low low high collaborative 4 low medium low agile 5 low medium medium agile 6 low medium high lean 7 low high low fully flexible 8 low high medium agile 9 low high high campaign 10 medium low low lean 11 medium low medium lean 12 medium low high lean 13 medium medium low agile 14 medium medium medium agile 15 medium medium high lean 16 medium high low agile 17 medium high medium agile 18 medium high high campaign 19 high low low lean 20 high low medium lean 21 high low high lean 22 high medium low agile 23 high medium medium lean 24 high medium high lean 25 high high low agile 26 high high medium agile 27 high high high campaign int. j. prod. manag. eng. (2022) 10(2), 99-113creative commons attribution-noncommercial-noderivatives 4.0 international quantitative supply chain segmentation model for dynamic alignment 113 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.16617 received: 2021-11-12 accepted: 2022-01-08 rediscovering scientific management – the evolution from industrial engineering to industrial data science jochen deusea,b* , nikolai westa & marius syberga a institute of production systems, technical university dortmund, leonhard-euler-str. 5, dortmund, 44227 germany. b centre for advanced manufacturing, university of technology sydney, 11 broadway, ultimo nsw 2007, australia. a1 jochen.deuse@uts.edu.au, a2 nikolai.west@ips.tu-dortmund.de, a3 marius.syberg@ips.tu-dortmund.de abstract: industrial engineering, through its role as design, planning and organizational body of the industrial production, has been crucial for the success of manufacturing companies for decades. the potential, expected over the course of industry 4.0 and through the application of data analytic tools and methods, requires a coupling to established methods. this creates the necessity to extend the traditional job description of industrial engineering by new tools from the field of data analytics, namely industrial data science. originating from the historic pioneers of industrial engineering, it is evident that the basic principles will remain valuable. however, further development in view of the data analytic possibilities is already taking place. this paper reviews the origins of industrial engineering with reference to four pioneers, draws a connection to current day usage, and considers possibilities for future applications of industrial data science. key words: scientific management, industrial engineering, industrial data science, data science, data analytics, process chain. 1. introduction ever since the first industrial revolution in the 18th century, optimization measures and operational decisions in the manufacturing industry rely on quantitative and fact-based assessments. fact-based decision-making has always been the cornerstone within the field of engineering, extensively taught by technical education and widely practiced in realworld operations. modern advancements due to the still on-going digitalization and globalization of nowadays world of work represent a logical continuation of the observable movements in science and technology. against the background of this natural and inevitable development, emerging potentials through data science do not necessarily represent a paradigm shift, but rather a continuation of the development of industrial engineering (ie). this development may greatly extend the established tools of today‘s engineers, but still draws on traditional ie principles that have been known for decades if not centuries. in view of the inflated expectations with regard to data science‘s problem solving capabilities and its promises of economic rationalization, this paper draws references to some of the main representatives and pioneers of ie, such as frank and lillian to cite this article: deuse, j., west, n., & syberg, m. (2022). rediscovering scientific management – the evolution from industrial engineering to industrial data science. international journal of production management and engineering, 10(1), 1-12. https://doi.org/10.4995/ijpme.2022.16617 int. j. prod. manag. eng. (2022) 10(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international 1 https://orcid.org/0000-0003-4066-4357 https://orcid.org/0000-0002-3657-0211 https://orcid.org/0000-0001-8832-4293 http://creativecommons.org/licenses/by-nc-nd/4.0/ gilbreth, john burbidge, william e. deming and eliyahu m. goldratt. these five pioneers are among the most prominent contributors to the field of industrial engineering and sparked novel research directions. by examining the respective field of research of these four pionieers of ie, this paper provides a comprehensive review of the historical development of ie to date. based on the patterns of this historical development, this paper outlines a broad view of contemporary movements, before also considering emerging trents for future research objectives within these four branches of research. summarizing, the paper embraces these pioneers of ie and it aims to help rediscovering established ideas and principles of industrial engineering at times when data science is permeating the manufacturing domain. 2. fundamentals 2.1. origin of scientific management the roots of ie stem from frederick w. taylor, whose main work principles of scientific management was instrumental in shaping the course of industrial manufacturing (taylor, 1911). by many, taylor is considered to be one of the most influential scientists of the 19th and 20th century. in addition to his contribution to the invention of high-speed steel, taylor is best known for working in the area of labor studies. his principles build on the assumption that work needs guidance by precise instructions given by management. this is based on the postulate that there is just one safest and most efficient way to accomplish a given work task. to identify this sequence, taylor proposed a five-step process: 1. select 10-15 worker from varied factories/ backgrounds, trained in a targeted activity 2. observe each elementary movement the tool usage during execution of the activity 3. measure the execution time of each element and select the fastest method for each 4. eliminate all incorrect, slow or unnecessary movements from the best practice 5. list the fastest method and the best tools for performing the activity in a table these five focusing steps help to identify and document optimal movement sequences and suitable tool usage in a standardized manner, using time recordings of any activity. with standard processes, best practice workflows are established; the efforts of the continuous improvement process are secured and made available across plants (deuse et al., 2020). along the belief in one best way to perform a given task, scientific management strictly enforces fact-based decision-making based on quantitatively measurable data. thus, quantifiably optimal solutions take the place of practices previously determined by formerly used rule-of-thumb methods (merkle, 1980). the onset of a worldwide adoption of scientific management principles defines a starting point for the continuous evolution of taylor‘s vision. it led to the emergence of ie and it is well established in modern manufacturing. the term ie includes all tasks concerned with ‘the design, improvement and installation of integrated systems of people, materials, information, equipment and energy. it draws upon specialized knowledge and skills in the mathematical, physical and social sciences together with the principles and methods of engineering analysis and design, to specify, predict and evaluate the results to be obtained from such systems’, as defined by the institute of industrial and systems engineers (iise, 2021). as such, industrial engineers lead continuous improvement processes and provide system understanding, method knowledge and problemsolving-competences, along the ever-evolving requirements of various other skills (richter & deuse, 2011). in this context, we consider the ability to utilize industrial data science as the latest addition to this catalogue of competences. 2.2. emergence of industrial data science the domain of data analytics experienced an increase in attention over the past years. industrial data science (ids) refers to the use of data analytics in industrial applications (mazarov et al., 2019). early applications of structural data analysis date back almost 70 years before the work of taylor. a u.s. naval officer and hydrographer, named matthew f. maury, recommended supplementing the previously undescribed nautical charts with information, such as longitude, latitude and other notes that seafarers collected on their routes (maury, 1963). this served to shorten the time at sea, since each seafarer could profit from the experiences of the others. he supplemented the pure information of the route with many additional variables, which all have an influence on the target variable ‘journey time’. he found patterns in this data, int. j. prod. manag. eng. (2022) 10(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international deuse et al. 2 http://creativecommons.org/licenses/by-nc-nd/4.0/ derived structures and generated knowledge from it. in doing so, he piloted a process, which we nowadays interpret as the basic problem solving approach of industrial data science (wierse & riedel, 2017). in manufacturing, taylor was among the first to identify a demand of data for fact-based decision-making within an industrial production environment. due to scale and complexity of the ever-increasing data acquisition, manual methods for data processing become uneconomical. hence, manufacturing companies seek the use of ids for the efficient evaluation and utilization of implicitly available knowledge. as for knowledge discovery in databases, ids includes all non-trivial measures to identify valid, novel, potentially useful, and ultimately understandable patterns in industrial datasets (fayyad et al., 1996). it draws on methods from multiple disciplines: machine learning, responsible for generation and generalization of knowledge by computers and statistics, the science of collecting, organizing and deriving conclusions from data, being the most relevant (awad & khanna, 2015). this basic idea behind the analysis of data is methodically carried out in the industrial environment today according to the cross-industry standard process for data mining (crisp-dm). with this five-step procedure, todays data scientists perform projects in a structured way from business understanding to deployment (chapman et al., 2000). we interpret crisp-dm as a formalization of the basic procedures of maury and taylor with a view to today’s conditions and challenges. the application of these methods and procedures in today’s industrial environment is both effective and unavoidable. this is explained by the increased complexity in making decisions, as in nowadays systems more variables have to be considered. simultaneously digitalization has also provided the infrastructure needed to record more data on these variables. the integration of computer technology in the form of embedded systems in industrial processes is standard today and enables the recording of a wide variety of data on products and processes. with the help of these cyber-physical systems, all steps from data acquisition with sensors to data storage in databases can be carried out in order to use ids methods to make intelligent, targeted decisions based on the analysis of every variable required (lee, 2006). some consider the use of ids tools and methods as part of the fourth industrial revolution, often called industry 4.0. other state that it has traits of a more gradual development of quantitative approaches that extend the traditional tools of ie. the emergence of ids shows distinct characteristics of an evolutionary process and is in line with the development trend of the past century. following approaches such as lean thinking or agile manufacturing, ids represents the latest facet of this traditional evolution of production principles. the advent of the internet of things and the availability of big data storage systems support the need for data-driven decisions. the approach for data-based decision-making has a predeceasing model in time management and is presented in the following section on the basis of the process chain of data analytics. 2.3. process chain of industrial data science ie and ids involve closely related tasks for a fact-based decision-making processes. with the process chain of time management, all activities for fact-based decision-making are broken down into subtasks and handled consecutively. this has led to the development of the process chain of data analytics, which follows a similar approach. in four stages, tasks of data collection, analysis, use and administration are carried out (figure 1). figure 1. process chain of industrial data science. since all fact-based decisions require a data basis for quantification, the first step in the process chain is to access all data sources that are necessary and related to the current task or project. this often includes the identification of relevant sources for a given analysis task, recoding missing data with suitable collection methods, and providing the data to an analysing system. the step is closely related to the initial phases of the well-established crisp-dm and required not only a good data understanding but also a solid business understanding. after ensuring a reliable end-to-end access to relevant data, the next step of the process chain is to analyse the provided data. for this, a suitable method for the analysis task must be selected from the wide range of available options. a number of pre-processing data administration data analysis data application data access int. j. prod. manag. eng. (2022) 10(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international rediscovering scientific management – the evolution from industrial engineering to industrial data science 3 http://creativecommons.org/licenses/by-nc-nd/4.0/ steps and transformations for subsequent usage may accompany the application of the analysis. while the first two steps are arguably necessary, a monetary gain will only arise through the operational use of the information generated during the analysis. hence, the third step is to apply the results of the analysis to the industrial use case, creating economic value. this includes both the application of selective analyses for specifically targeted questions as well as the implementation of continuous monitoring systems. finally, to administrate these tasks a wealth of supporting duties needs to be fulfilled. among others, this includes assigning a long-term data stewardship, allocating a clear data governance, ensuring an end-to-end data security or securing an ethical data usage. 3. pioneer case studies 3.1. time and motion study – frank and lillian gilbreth past. temporal measurements of industrial process increments form the basis for operational planning and decision-making processes. for ie, time data is essential for the analysis, design, modelling and simulation of production systems as well as for the design of workplaces and the control of manufacturing and assembly systems. time and motion study enables an industrial engineer to control and plan from a quantitative basis, according to taylor’s basic idea. frank and lillian gilbreth, whom we consider pioneers of time and motion study, recognized at the beginning of the 20th century that the conversion of movements into time data is essential for work analyses (gilbreth, 1912). the gilbreth clock allowed a detailed analysis with regard to a tasks duration and usefulness, so that value-adding and non-value-adding elements become distinctive. the stopwatch was the device for manual data recording.gilbreth determined that the time to carry out an activity for a singular sequence with equal practice, equal aptitude and equal effort of the workers within realistic limits depends solely on the method used. asa b. segur assigned standardized time values to the standard elements in industrial processes devised by gilbreth by studying numerous workers of different skill levels (maynard & zandin, 2001). this enabled analyzing work processes using a standardized scheme. thus, motion time analysis is considered the first predetermined motion time system. present. still, time recordings hold a fundamentally vital role for process management. predetermined motion time systems are still widespread today. numerous companies leverage the mtm method, developed between 1940 and 1950, to analyze manual work processes today. in german-speaking countries, over 250 companies belong to the mtm association, thus representing over 2 million employees (mtm association e. v., 2021). this underlines the relevance of precise data recording and attention to detail. today, time data from production forms the core of strategic and operational process planning. the analysis of work processes via video recordings and motion capturing approaches makes the use of systems of predetermined times even more direct and universal (bortolini et al., 2020). future. ids research deals with automatic data access through image recognition via machine learning or the use of sensor technology. the ease of access to the technology underlying vision-based and sensor-based analysis serves as enabler for further development. it is equally conceivable to apply gilbreth’s visionary ideas to the use of robots and to the optimization of industrial human-machine interaction (wang et al., 2011). motion data in the form of human silhouettes or human skeletons elaborates human movement. additionally, the movement of robots can be captured by the control system. the combination of such data as well as an enrichment by other data sources and the necessary forces for the different tasks forms the basis for the analysis and optimization of the technologydetermined workstations (figure 2). the usage of machine learning is a possible potential to be explored for automatically analyzing recorded motions. it allows suggestions of process improvements based on mtm studies using motion capturing data (deuse et al., 2019). the integration of mtm approaches with virtual reality (vr) provides the advantage of preventing suboptimal workplace designs during the planning phase without the need for physical mock-ups (cardboard engineering) for data collecting (gorobets et al., 2021). 3.2. group technology – john burbidge past. for ie, group technology is essential for production optimization. the term describes the approach of grouping objects and resources according to their similarity. organizing processes and structures is often more efficient and effec-tive based on such groupings. sergei p. mitrofanov was int. j. prod. manag. eng. (2022) 10(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international deuse et al. 4 http://creativecommons.org/licenses/by-nc-nd/4.0/ the first to research the idea of classifying process methods based on the shape of the resulting products using the research results of a. p. sokolowski (sokolowski, 1938; mitrofanov, 1946). he established a classification system that structured work pieces according to function, shape and technological features based on his finding to work on similar parts of a group with the same equipment of a lathe (burbidge, 1991). john burbidge, whom we consider as pioneer of group technology, based his research on mitrofanov and sokolowski and successfully applied the grouping on a larger scale (burbidge, 1975). thereby, he validated the prior research and coined the terms ‘group’ as a set of machines, and ‘family’ as a set of parts. burbidge introduced a production analysis method with the production flow analysis. according to his research, products passing through the same machines should also be manufactured in one machine group (burbidge, 1963). present. searching and finding similarities in industrial processes is essential for economically successful enterprises and possible with four different procedures: classification systems, productionanalytical methods, cluster-analytical methods and artificial intelligence methods. classification is the key method used by sokolowski and mitrofanov. production-analytical methods use the frequency of the production sequences or work piece-resourcematrices for part family formation. cluster analytical methods use methods of multivariate statistics, such as regression, to analyze different characteristic values for similarities and identify homogeneous groups between which there is a possibility of dissimilarity. last, it is possible to use knowledgebased approaches and artificial neural networks, both methods of artificial intelligence for part family formation (eversheim & deuse, 1997; kusiak & dagli, 1994). those approaches help with different challenges, such as building structures in low volume and high mix productions. levelling taking into account product families helps serving transparency, calming of variability and leads to output improvement (bohnen et al., 2013). research in practical pattern recognition by using sensors and software for analyzing images, characters and text show high potential (feng & hua, 2020). for subsequent analyses, it is necessary to record different data which is then analyzed for similarities. visualization using flowcharts allows easy access the subject matter. future. as manufacturing processes and products become more complex, simplification becomes more difficult. the search for similarities based on shape or machine group is also no longer efficient in variable production environments due to the high complexity of products and increased amount of different manufactured variants. for the homogenization of the product routes of a production the use of statistical methods like modified jaccard index is feasible (maschek et al., 2014). other methods for identifying process-routes weaknesses and subsequent their improvement, such as value stream mapping as a tool from the field of lean management are also evolving with the tools currently available to map, predict and control dynamic effects on value figure 2. convolutional neuronal networks (cnn) with 2 graph layers, 2 fully connected layers as one machine learning method to enhance mtm (own figure, mtm summit 2021). labelled motion capture data basic motions classifier reach grasp moveposition release current setting: 2 graph-layers 2 fully connected layers … hidden layer hidden layer … input output int. j. prod. manag. eng. (2022) 10(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international rediscovering scientific management – the evolution from industrial engineering to industrial data science 5 http://creativecommons.org/licenses/by-nc-nd/4.0/ streams. emerging technologies provide completely integrated production environments with realtime data gathering and transmission (valamede & akkari, 2020). the cloud connects all resources on the manufacturing floor to supervisory and control terminals and and combines data from different sources. this is combined with ontology-driven modeling-based graphical database technology or a multi-agent system based on cyber-physical systems to visualize the productivity of customerdriven dynamic manufacturing processes. (huang et al., 2019) another approach to simplify complex process systems is process mining. based on business process management, van der aalst developed a method for analyzing the data of event logs of processes that uses the process knowledge implicit in these events to graphically represent process sequences and information of process steps on the basis of paths, thus making potentials visible (van der aalst et al., 2012). process mining techniques can help increasing the management productivity by modelling production planning processes in a manufacturing company (er et al., 2018). with this information it is possible to identify the commonly unrecorded operations implemented to adapt the production plan to any changes in demand. this improves the ability to optimize its production process by balancing production efficiency and flexibility (corallo et al., 2020). in combination with different it systems of a company, process mining enables the visualization and optimization of both value-added production processes and their management and planning processes (knoll et al., 2019). considering burbidge‘s ideas, we see the consistent use of all data as a logical step to streamline processes by searching for similarities and forming groups, especially in highly complex systems. for ids, using process mining in industrial systems is the logical next step in the development of group technology. 3.3. quality management – william e. deming past. quality management (qm) holds an integral part of continuous improvement in ie, since it serves as a prominent starting point for process improvement and often acts as significant driver of costs. early on, taylor acknowledged qm as a crucial factor for the maximization of productivity in industrial processes. in the 1920s, walter a. shewhart recognized that preventing quality related issues is significantly more economical than sorting out defective parts or repairing them, as taylor had suggested. his invention of quality control charts serves as a static method for process control that allows for scientifically based and economically founded decisions based on recorded process metrics. for this purpose, it was necessary to develop target metrics and a tolerance range for all processes recorded on the control chart. the observed deviation between actual system performance and target metrics allowed for unprecedented levels of process monitoring and control (shewhart, 1931). we consider william e. deming, a student of shewhart, as pioneer in the field of qm, for he specified and propagated the early ideas. while initially unnoticed in the western world, japanese companies adopted deming’s methods in the 1950s and established his status as a visionary in qm. particularly successful was the application of statistical process control, which aimed for efficient process operation by producing more specificationconfirmative products while causing less rework or scrap. consideration of the temporal course of statistical values, such as current the range between actual and target metric, paired with the visualization on control charts allowed for detection of negative deviation as well as short-term adjustments during production (deming, 1950). present. several of deming’s approaches can be found in widely used standards, such as the prominent pdca-cycle that is included in the iso 9001. the principles of statistical process control lead to the emergence of the ideal of zero-defect manufacturing (zdm) that modern manufacturing companies still seek to achieve. zdm aims to reduce defects through prevention and targets the development of workers desire to perform a job correctly at all times (wang, 2013). to quantifiably record the occurrence of any defects, manufacturing companies utilize different strategies or platforms to bundle the wide range of potential sources for quality data. such collections allow using newer approaches, such as machine learning, to detect different types of defects on a large scale (schulte et al., 2020). a mayor task is building autonomous qm systems that achieve trustworthy results within an industry 4.0 setting, while remaining economically viable. future. quality control is usually reactionary and can only detect defects, not proactively prevent them. as an ongoing research subject, modern qm is primarily concerned with predicting future quality. the increased data availability allows drawing int. j. prod. manag. eng. (2022) 10(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international deuse et al. 6 http://creativecommons.org/licenses/by-nc-nd/4.0/ conclusions about the quality of products that are still in production, only using the recorded process data. utilizing supervised and unsupervised machine learning models, allows for advanced quality-based process control (lieber et al., 2013). additionally, it enables the prediction of quality-related features and identification of ideal process parameters (schmitt et al., 2019) (figure 3). random forest integrated inside the bayesian optimization approach are one option to enable organizations to manage large-scale product quality prediction in process industrial cyber–physical systems (wang et al., 2020). in another example, measurements such as tightening data of screw driving processes help to predict the final condition of engines without requiring end-of-line testing (west et al., 2021). identifying process anomalies and predicting likely assembly defects with ids enables early initiation of corrections, such as a partial deconstruction. 3.4. production control – eliyahu m. goldratt past. with the theory of constraints (toc), eliyau m. goldratt introduced a novel approach to production control in 1984. according to toc, the output of every production system is inevitably constrained by one single limiting factor. similar to the weakest link in a chain, such a factor poses a bottleneck for the entire system. all improvement activities must target that bottleneck, since optimizations of nonbottleneck stations do not improve the performance of the system but cause increased work in process (goldratt & cox, 1984). with regard to production control, this meant that the control system must also primarily account for the bottleneck. to implement bottleneck-oriented production control, goldratt proposed the drum-buffer-rope (dbr) method. in the toc, dbr is a method for process scheduling that increases production flow by leveraging the system’s bottleneck (goldratt & fox, 1986). through dbr, only a bottleneck needs scheduling, which is easier than scheduling every job at every station. in addition, the bottleneck’s capacity provides a simple way to plan due dates, since it matches the system’s overall output. through the development of toc and its application using dbr, goldratt made an innovative contribution that shaped the further course of ie. present. while toc and dbr were primarily aiming to manage static bottlenecks, modern production systems often encounter shifting bottlenecks. due to variability-related factors, such bottlenecks move between workstations over time. adaptations of the toc, led to new methods for real-time bottleneck identification. such methods require continuous monitoring of the production system, which only became possible in the last decade due to the increasing digitalization. whereas goldratt, for example, suggested interviewing employees to identify bottlenecks, these methods utilized quantified metrics. a prominent example is the active period method (apm) that identifies a bottleneck as the station working the longest without interruption. apm assumes that stations in interconnected production systems starve or block each other. an active machine running for extended periods is more likely to block or starve other machines. hence, the machine with the longest uninterrupted active period has the largest effect on the overall output and acts as the current bottleneck (roser et al., 2002). near realtime knowledge of bottleneck locations, as well as figure 3. framework for predictive model-based quality inspection (schmitt et al., 2020). technical integration . technical implementation . physical process data collection and processing 1 data storage model deployment model training and scoring 2 3 4 5 int. j. prod. manag. eng. (2022) 10(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international rediscovering scientific management – the evolution from industrial engineering to industrial data science 7 http://creativecommons.org/licenses/by-nc-nd/4.0/ knowledge of the relative frequency of occurrence, enables a targeted production control and realtime fault repair prioritization (wedel et al., 2015). however, identification methods do not manage to avoid productivity losses at shifting bottlenecks, but only help to mitigate the effects. future. in modern manufacturing systems, identifying bottlenecks will continue to be the focus of research due to the prevailing dynamics of increasingly complex systems. to minimize bottleneck-related losses of potential outputs, anticipatory knowledge about future system behavior is required. this thought led to the idea of bottleneck prediction, as a current subject of research. predicting a bottleneck shifting before it occurs, allows a production control system to counteract this change and prevent a shift. while established approaches for bottleneck detection require a measurable effect, their nature is similar to fire-fighting strategies known in maintenance. only by anticipating an emerging bottleneck, the effect can be controlled without effecting the overall output. in addition, the recent idea of bottleneck prescription proposes a novel type of system that fully subjects a system‘s control mechanism to the predictions of future bottleneck occurrences (west et al., 2022). since a bottleneck’s existence is inevitable, this approach will not eliminate a bottleneck, but it can avoid or reduce the adverse influence of shifting bottlenecks. predicting bottlenecks requires a realtime, databased bottleneck identification capability (deuse et al., 2016; roser et al., 2017). while theoretical approaches to bottleneck prediction are emerging in the scientific literature, practical implementation represents a future need for action in ids (figure 4). 4. discussion the development of the individual fields of industrial engineering shows the strong connection to data analysis from the beginning. the pioneers mentioned, gilbreth, burbidge, deming and goldratt, proved the dependence of optimisation and improvement on data analysis. the present solutions prove the relevance of this approach equally. industrial data science is therefore the logical evolutional step of working with data in industrial engineering. many manufacturing companies face much more complex and complicated problems in parallel. for these companies, we consider industrial data science as a novel tool that enables them to utilize the original ideas of the four pioneers in a more efficient, largescale and goal-oriented fashion. the different fields of ie need to be differentiated in this context. every company uses the ideas of gilbreth, burbidge and deming, in many productions they are even the basis of optimisations and the continuous improvement process in different sectors. the methods of industrial data science extend the previous procedures making them reach the next stage of their development. image recognition enables fast, accurate time recordings, machine learning enables the automatic creation of work plans of products of the same group and product and process quality can be dynamically detected, predicted and thus predictively improved in real time. industrial data science is influencing the field of bottleneck analysis, pioneered by goldratt, in a different way. although goldratt‘s ideas have a fundamental character for all production systems, figure 4. methodology for bottleneck analysis with corresponding research based on west et al. (2022). b en ef it effort bottleneck diagnosis bottleneck prediction bottleneck prescription bottleneck detection methods available forthcoming fieldon-going research (roser et al., 2002) (deuse et al., 2016) (west et al., 2022) (wedel et al., 2015) int. j. prod. manag. eng. (2022) 10(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international deuse et al. 8 http://creativecommons.org/licenses/by-nc-nd/4.0/ the detection of a bottleneck is not possible for every company with the developed methods. this is a result of the dynamic behavior of the different variables affecting the processes and consequently the occurrence of shifting bottlenecks. in addition, different methods for bottleneck detection do not always have the same result. this complicates the interpretation of the analyses and their target-oriented use. the application of industrial data science is an enabler in this field, as it allows industrial users to generate more knowledge about their processes and to analyse a larger amount of data more precisely. this opens up the field of bottleneck analysis to a variety of other companies. these two functions, enable and extend, can be transferred to other fields of ie. kingman and little, for example, are pioneers in the field of operations research (kingman, 1961). they have mathematically proven correlations between queue length, arrival time and utilisation of a system (little, 1961). this dependencies are used to material flow and buffer research questions in production environments in the literature (lödding, 2013). this works under certain boundary conditions. in this field, industry 4.0 and machine learning approaches can be both enabler and extender by enhancing the current limits of application of their ideas by capturing further influencing parameters, making them measurable and providing tools to recognise even more complex patterns (gallina et al., 2021). what applies to this practical field of ie can be transferred to the organisational field. projects have existed for thousands of years, but it was not until the 1940s that us military industry-academic-research project management was formalized in institutional processes (johnson, 2013). getting combined with systems engineering and operations research later, it became an essential aspect of the industrial engineer‘s tasks (johnson, 1997). as of today, the iiot and different ai approaches connect project management directly to events the shopfloor. with lean principles as a pillar of ie and selfoptimisation equally as a pillar of lean principles, the question of effort and benefit in ie has always arisen. with the value-creating processes at the centre of its own basic idea, industrial engineering, as a staff unit only indirectly involved in value creation, constantly questions itself. sustainable economic successes by finding the right balance between good work preparation and targeted continuous improvement have proven the role of ie in the past (deuse et al., 2006). the expansion through and development towards industrial data science poses the question of effort and benefit again. the initial effort to enable its production to use modern approaches to data analysis seems large. in addition to increasing the knowledge of the workforce involved, physical resources must be digitised or replaced, hardware must be purchased and installed, and software licences must be acquired. in addition to increasing the knowledge of the workforce involved, a company must digitise or replace physical resources, purchase and install hardware and acquire software licences. in addition, sensors support the former manual data acquisition, artificial intelligence helps with decisions or even relieves the industrial engineer. the rapid development of recent years refutes those arguments in different ways. first, the initial costs on the hardware side are declining due to the high demand and the resulting sharp increase in availability. open source solutions also make simplify starting with interface management and iiot in order to be able to analyse data using different methods (strauß et al., 2018). at the same time, the range of educational opportunities in the industrial sector has grown considerably, so employees can easily be empowered. second and most relevant, digitalization and the application of ids has a direct impact on a company‘s financial performance (eller et al., 2020). the development of the individual areas of ie and the markets as such shows that a company without targeted digitization and the application of data science will not be marketable in the future. strategic and operational implementation is essential for success (dold & speck, 2021). ids helps the industrial engineer in multiple ways and sometimes replaces some decisions, but brings new challenges to the job profile. domain knowledge is still indispensable for industrial issues, the industrial engineer must select and connect the right data sources as well as manage the targeted application of hardware, ai approaches and employee‘s datascience-education. the benefits overcome the effort of implementing ids midand long-term by a multiple. 5. conclusion the case studies of the pioneers of ie have shown the development of scientific management in four different domains as a rather evolutionary process. the contemporary trend towards a more widespread int. j. prod. manag. eng. (2022) 10(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international rediscovering scientific management – the evolution from industrial engineering to industrial data science 9 http://creativecommons.org/licenses/by-nc-nd/4.0/ application of industrial data science is an inevitable result of a decade-spanning development process. leveraging the growing data sources is merely the next logical step in an environment that relies on fact-based and quantified decision-making. thus, the application of data science in industrial engineering under the umbrella of industrial data science will continue to grow in importance in the coming decades. at their core, manufacturing companies will continue to use the original concepts of the discussed pioneers, but increase the effectiveness through the addition of digital and data-driven methods and tools, even in other fields of industrial engineering. accessing, analyzing, applying and administrating data is going to be vital for future applications of industrial data science. the pioneers presented in the paper, as well as their associated research areas, were selected primarily due to their high relevance to ie. nevertheless, these representatives have to be called a selection of pioneers. in the continuing development of scientific management since taylor, many scientists have distinguished themselves. as research limitations, we would therefore emphasize the small number of pioneers studied and the selection of application examples. for future competitiveness, an industrial engineer‘s collection of applicable methods and tools has to be expanded to accommodate for the capabilities of ids. at the same time, companies must create the technical and educational basis for applying ids in order to be able to assert themselves in the market. in addition, the multitude of requirements for an integrated and networked application of industrial data analysis in dynamic value creation networks will shape the further course of ids research. acknowledgement the work on this paper has been supported by the german federal ministry of education and research (bmbf) as part of the funding program ‘industry 4.0 collaborations in dynamic value networks (inkowe)’ in the project akkord (02p17d210). references awad, m., & khanna, r. (2015). efficient learning machines: theories, concepts, and applications for engineers and system designers. apress media. https://doi.org/10.1007/978-1-4302-5990-9 bohnen, f., buhl, m., & deuse, j. (2013). systematic procedure for leveling of low volume and high mix production. cirp journal of manufacturing science and technology, 6(1), 53–58. https://doi.org/10.1016/j.cirpj.2012.10.003 bortolini, m., faccio, m., gamberi, m., & pilati, f. (2020). motion analysis system (mas) for production and ergonomics assessment in the manufacturing processes. computers & industrial engineering, 139(1–4). https://doi.org/10.1016/j.cie.2018.10.046 burbidge, j. l. (1963). production flow analysis. production engineer, 42(12), 742. https://doi.org/10.1049/tpe.1963.0114 burbidge, j. l. (1975). the introduction of group technology. w. h. burbidge, j. l. (1991). production flow analysis for planning group technology. journal of operations management, 10(1), 5–27. https://doi.org/10.1016/0272-6963(91)90033-t chapman, p., clinton, j., kerber, r., khabaza, t., reinartz, t., shearer, c., & wirth, r. (2000). crisp-dm 1.0. spss inc. corallo, a., lazoi, m., & striani, f. (2020). process mining and industrial applications: a systematic literature review. knowledge and process management, 27(3), 225–233. https://doi.org/10.1002/kpm.1630 deming, w. e. (1950). elementary principles of the statistical control of quality: a series of lectures. nippon kagaku gijutsu remmei. deuse, j., dombrowski, u., nöhring, f., mazarov, j., & dix, y. (2020). systematic combination of lean management with digitalization to improve production systems on the example of jidoka 4.0. international journal of engineering business management, 12(3), 184797902095135. https://doi.org/10.1177/1847979020951351 deuse, j., lenze, d., klenner, f., & friedrich, t. (2016). manufacturing data analytics to identify dynamic bottlenecks in production systems with high value-added variability (in german). in c. schlick (ed.), megatrend digitalisierung (pp. 11–26). gito. deuse, j., stankiewicz, l., zwinkau, r., & weichert, f. (2019). automatic generation of methods-time measurement analyses for assembly tasks from motion capture data using convolutional neuronal networks. international conference on applied human factors and ergonomics, 141–150. https://doi.org/10.1007/978-3-030-20040-4_13 int. j. prod. manag. eng. (2022) 10(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international deuse et al. 10 https://doi.org/10.1007/978-1-4302-5990-9 https://doi.org/10.1016/j.cirpj.2012.10.003 https://doi.org/10.1016/j.cie.2018.10.046 https://doi.org/10.1049/tpe.1963.0114 https://doi.org/10.1016/0272-6963(91)90033-t https://doi.org/10.1002/kpm.1630 https://doi.org/10.1177/1847979020951351 https://doi.org/10.1007/978-3-030-20040-4_13 http://creativecommons.org/licenses/by-nc-nd/4.0/ deuse, j., wischniewski, s., & fischer, h. (2006). rediscovering industrial engineering – methods for applying lean production principles (deutsch: renaissance des industrial engineering methoden für die umsetzung ganzheitlicher produktionssysteme.). werkstattstechnik online, 96(1/2), 57–60. dold, l., & speck, c. (2021). resolving the productivity paradox of digitalised production. international journal of production management and engineering, 9(2), 65. https://doi.org/10.4995/ijpme.2021.15058 eller, r., alford, p., kallmünzer, a., & peters, m. (2020). antecedents, consequences, and challenges of small and medium-sized enterprise digitalization. journal of business research, 112, 119–127. https://doi.org/10.1016/j. jbusres.2020.03.004 er, m., arsad, n., astuti, h. m., kusumawardani, r. p., & utami, r. a. (2018). analysis of production planning in a global manufacturing company with process mining. journal of enterprise information management, 31(2), 317–337. https://doi.org/10.1108/jeim-01-2017-0003 eversheim, w., & deuse, j. (1997). formation of part families based on product model data. production engineering (2), 97–100. fayyad, u., piatetsky-shapiro, & smyth, p. (1996). knowledge discovery and data mining: towards a unifying framework. international conference on knowledge discovery and data mining, 2(1), 82–88. feng, z., & hua, x. (2020). pattern recognition and its application in image processing. journal of physics: conference series, 1518, 12071. https://doi.org/10.1088/1742-6596/1518/1/012071 gallina, v., lingitz, l., breitschopf, j., zudor, e., & sihn, w. (2021). work in progress level prediction with long short-term memory recurrent neural network. procedia manufacturing, 54, 136–141. https://doi.org/10.1016/j. promfg.2021.07.047 gilbreth, f. b. (1912). primer of scientific management. d. van nostrand c. goldratt, e. m., & cox, j. (1984). the goal. north river press. goldratt, e. m., & fox, r. e. (1986). the race. north river press. gorobets, v., holzwarth, v., hirt, c., jufer, n., & kunz, a. (2021). a vr-based approach in conducting mtm for manual workplaces. the international journal of advanced manufacturing technology, 117(7), 2501–2510. https://doi.org/10.1007/s00170-021-07260-7 huang, z., kim, j., sadri, a., dowey, s., & dargusch, m. s. (2019). industry 4.0: development of a multiagent system for dynamic value stream mapping in smes. journal of manufacturing systems, 52, 1–12. https://doi.org/10.1016/j.jmsy.2019.05.001 iise. (2021). what is industrial and systems engineering? (iise official definition). institute of industrial & systems engineers. https://iise.org/details.aspx?id=282 johnson, s. b. (1997). three approaches to big technology: operations research, systems engineering, and project management. technology and culture, 38(4), 891. https://doi.org/10.2307/3106953 johnson, s. b. (2013). technical and institutional factors in the emergence of project management. international journal of project management, 31(5), 670–681. https://doi.org/10.1016/j.ijproman.2013.01.006 kingman, j. f. c. (1961). the single server queue in heavy traffic. mathematical proceedings of the cambridge philosophical society, 57(4), 902–904. https://doi.org/10.1017/s0305004100036094 knoll, d., waldmann, j., & reinhart, g. (2019). developing an internal logistics ontology for process mining. procedia cirp, 79(1), 427–432. https://doi.org/10.1016/j.procir.2019.02.116 kusiak, a., & dagli, c. h. (1994). artificial neural networks for intelligent manufacturing. springer netherlands. lee, e. (2006). cyber-physical systems are computing foundations adequate? lieber, d., stolpe, m., konrad, b., deuse, j., & morik, k. (2013). quality prediction in interlinked manufacturing processes based on supervised & unsupervised machine learning. cirp conference on manufacturing systems, 46, 193–198. https://doi.org/10.1016/j.procir.2013.05.033 little, j. d. c. (1961). a proof for the queuing formula: l = λ w. operations research, 9(3), 383–387. https://doi.org/10.1287/opre.9.3.383 lödding, h. (2013). handbook of manufacturing control: fundamentals, description, configuration. springerlink bücher. springer. https://doi.org/10.1007/978-3-642-24458-2 maschek, t., heuser, c., hasselmann, v.-r., deuse, j., & willats, p. (2014). variability-based classification of production systems.: basis for individual design and management concepts. zeitschrift für wirtschaftlichen fabrikbetrieb, 109(9), 591–594 ((in german)). https://doi.org/10.3139/104.111204 maury, m. f. (1963). the physical geography of the sea, and its meteorology. harvard university press. https://doi.org/10.4159/harvard.9780674865280 maynard, h. b., & zandin, k. b. (2001). maynard’s industrial engineering handbook (5th ed.). mcgraw-hill. int. j. prod. manag. eng. (2022) 10(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international rediscovering scientific management – the evolution from industrial engineering to industrial data science 11 https://doi.org/10.4995/ijpme.2021.15058 https://doi.org/10.1016/j.jbusres.2020.03.004 https://doi.org/10.1016/j.jbusres.2020.03.004 https://doi.org/10.1108/jeim-01-2017-0003 https://doi.org/10.1088/1742-6596/1518/1/012071 https://doi.org/10.1016/j.promfg.2021.07.047 https://doi.org/10.1016/j.promfg.2021.07.047 https://doi.org/10.1007/s00170-021-07260-7 https://doi.org/10.1016/j.jmsy.2019.05.001 https://iise.org/details.aspx?id=282 https://doi.org/10.2307/3106953 https://doi.org/10.1016/j.ijproman.2013.01.006 https://doi.org/10.1017/s0305004100036094 https://doi.org/10.1016/j.procir.2019.02.116 https://doi.org/10.1016/j.procir.2013.05.033 https://doi.org/10.1287/opre.9.3.383 https://doi.org/10.1007/978-3-642-24458-2 https://doi.org/10.3139/104.111204 https://doi.org/10.4159/harvard.9780674865280 http://creativecommons.org/licenses/by-nc-nd/4.0/ mazarov, j., wolf, p., schallow, j., nöhring, f., deuse, j., & richter, r. (2019). industrial data science in value creation networks (in german). zeitschrift für wirtschaftlichen fabrikbetrieb, 114(12), 874–877 (concept of a service platform for data integration and analysis, competence development and novel business models). https://doi.org/10.3139/104.112205 merkle, j. a. (1980). management and ideology: the legacy of the international scientific management movement. uc press. mitrofanov, s. p. (1946). scientific principles of group technology. mtm association e. v. (2021). brand history. mtm association e. v. https://mtm.org/en/brands/brand-history richter, r., & deuse, j. (2011). industrial engineering in modern production (in german). zeitschrift f. angew. arbeitswissenschaft, 207(1), 6–13. roser, c., lorentzen, k., lenze, d., deuse, j., klenner, f., richter, r., schmitt, j., & willats, p. (2017). bottleneck prediction using the active period method in combination with buffer inventories. ifip-apms, 374–381. https://doi. org/10.1007/978-3-319-66926-7_43 roser, c., nakano, m., & tanaka, m. (2002). shifting bottleneck detection. proc. of the winter simulation conference(1), 1079–1086. https://doi.org/10.1109/wsc.2002.1166360 schmitt, j., bönig, j., borggräfe, t., beitinger, g., & deuse, j. (2020). predictive model-based quality inspection using machine learning and edge cloud computing. advanced engineering informatics, 45, 1–10. https://doi.org/10.1016/j.aei.2020.101101 schmitt, j., hahn, f., & deuse, j. (2019). practical framework for advanced quality-based process control in interlinked manufacturing processes. ieee-ieem, 511–515. https://doi.org/10.1109/ieem44572.2019.8978870 schulte, l., schmitt, j., meierhofer, f., & deuse, j. (2020). optimizing inspection process severity by machine learning under label uncertainty. advances in human factors and systems interaction, 3–9. https://doi.org/10.1007/978-3-030-51369-6_1 shewhart, w. a. (1931). economic control of quality of manufactured product. martino publishing. sokolowski, a. p. (1938). problems of typification of technological processes. lenitomasch. strauß, p., schmitz, m., wöstmann, r., & deuse, j. (2018). enabling of predictive maintenance in the brownfield through low-cost sensors, an iiot-architecture and machine learning. in 2018 ieee international conference on big data (big data). https://doi.org/10.1109/bigdata.2018.8622076 taylor, f. w. (1911). the principles of scientific management. harper & brothers publishers. valamede, l. s., & akkari, a. c. s. (2020). lean 4.0: a new holistic approach for the integration of lean manufacturing tools and digital technologies. international journal of mathematical, engineering and management sciences, 5(5), 851–868. https://doi.org/10.33889/ijmems.2020.5.5.066 van der aalst, w., adriansyah, a., de medeiros, ana karla alves, arcieri, f., baier, t., blickle, t., bose, j. c., van den brand, p., brandtjen, r., buijs, j., burattin, a., carmona, j., castellanos, m., claes, j., cook, j., costantini, n., curbera, f., damiani, e., leoni, m. de,... wynn, m. (2012). process mining manifesto. in f. daniel, k. barkaoui, & s. dustdar (eds.), business process management workshops (pp. 169–194). springer berlin heidelberg. https://doi.org/10.1007/978-3-642-28108-2_19 wang, k.-s. (2013). towards zero-defect manufacturing (zdm)—a data mining approach. advances in manufacturing, 1(1), 62–74. https://doi.org/10.1007/s40436-013-0010-9 wang, l., zhao, g., cheng, l., & pietikäinen, m. (2011). machine learning for vision-based motion analysis. springer london. https://doi.org/10.1007/978-0-85729-057-1 wang, t., wang, x., ma, r., li, x., hu, x., chan, f. t. s., & ruan, j. (2020). random forest-bayesian optimization for product quality prediction with large-scale dimensions in process industrial cyber–physical systems. ieee internet of things journal, 7(9), 8641–8653. https://doi.org/10.1109/jiot.2020.2992811 wedel, m., hacht, m., hieber, r., metternich, j., & abele, e. (2015). real-time bottleneck detection and prediction to prioritize fault repair in interlinked production lines. procedia cirp, 37, 140–145. https://doi.org/10.1016/j. procir.2015.08.071 west, n., schlegl, t., & deuse, j. (2021). feature extraction for time series classification using univariate descriptive statistics and dynamic time warping in a manufacturing environment. ieee-icbaie, 2(1), 762–768. https://doi.org/10.1109/icbaie52039.2021.9389954 west, n., syberg, m., & deuse, j. (2022). a holistic methodology for successive bottleneck analysis in dynamic value streams of manufacturing companies. in a.-l. andersen, r. andersen, t. d. brunoe, m. s. s. larsen, k. nielsen, a. napoleone, & s. kjeldgaard (eds.), lecture notes in mechanical engineering. towards sustainable customization: bridging smart products and manufacturing systems (pp. 612–619). springer international publishing. https://doi.org/10.1007/978-3-030-90700-6_69 wierse, a., & riedel, t. (2017). smart data analytics: mit hilfe von big data zusammenhänge erkennen und potentiale nutzen. walter de gruyter gmbh. https://doi.org/10.1515/9783110463958 int. j. prod. manag. eng. (2022) 10(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international deuse et al. 12 https://doi.org/10.3139/104.112205 https://mtm.org/en/brands/brand-history https://doi.org/10.1007/978-3-319-66926-7_43 https://doi.org/10.1007/978-3-319-66926-7_43 https://doi.org/10.1109/wsc.2002.1166360 https://doi.org/10.1016/j.aei.2020.101101 https://doi.org/10.1109/ieem44572.2019.8978870 https://doi.org/10.1007/978-3-030-51369-6_1 https://doi.org/10.1109/bigdata.2018.8622076 https://doi.org/10.33889/ijmems.2020.5.5.066 https://doi.org/10.1007/978-3-642-28108-2_19 https://doi.org/10.1007/s40436-013-0010-9 https://doi.org/10.1007/978-0-85729-057-1 https://doi.org/10.1109/jiot.2020.2992811 https://doi.org/10.1016/j.procir.2015.08.071 https://doi.org/10.1016/j.procir.2015.08.071 https://doi.org/10.1109/icbaie52039.2021.9389954 https://doi.org/10.1007/978-3-030-90700-6_69 https://doi.org/10.1515/9783110463958 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.16084 received: 2021-08-10 accepted: 2021-09-30 solving stochastic multi-manned u-shaped assembly line balancing problem using differential evolution algorithm mohammad zakaraia a1, hegazy zaher a2, naglaa ragaa a3 a faculty of graduate studies for statistical research, cairo university, 5 ahmed zewail, ad doqi, dokki, giza governorate, egypt. a1 zicooo82@gmail.com, a2 hgsabry@gmail.com, a3 naglaa777subkiii@yahoo.com abstract: the u-shaped assembly lines help to have more flexibility than the straight assembly lines, where the operators can perform tasks in both sides of the line, the entrance and the exit sides. having more than one operator in any station of the line can reduce the line length and thereby affects the number of produced products. this paper combines the u-shaped assembly line balancing problem with the multi-manned assembly line balancing problem in one problem. in addition, the processing times of the tasks are considered as stochastic, where they are represented as random variables with known means and variances. the problem is formulated as a mixed-integer linear programming and the cycle time constraints are formulated as chance-constraints. the proposed algorithm for solving the problem is a differential evolution algorithm. the parameter of the algorithm is optimized using experimental design and the computational results are done on 71 adapted problems selected from well-known benchmarks. key words: metaheuristics, differential evolution algorithm, u-shaped assembly line balancing problem, multi-manned assembly line balancing problem, chance-constrained programming. 1. introduction the assembly lines play an important role in industry. they reduce the learning aspects by dividing the assembly work into a set of stations that move in some kind of transportation system, such as conveyer belt, and they help to produce products in a fixed time called the cycle time. the assembly line balancing problem is the problem that is related to optimizing the assignment of the tasks to the stations in order to achieve some specific objectives, such as minimizing the number of stations and minimizing the cycle time. the simplified assumptions of the problem consist of three set of constraints. the first set of constraints is the set of assignment constraints, which ensures that each task is assigned in only one station. the second set of constraints is the cycle time constraints, which ensures that the total processing time of any station doesn’t exceed the cycle time. the third set of constraints ensures that each task has to be assigned after its predecessors. the problem in research is classified into two categories, which are the simple assembly line balancing problems (salbp) that only cover the simplified assumptions, and the general assembly line balancing problems (galbp) which contain some other constraints related to the practical relevancies. the u-shaped assembly line balancing problem (ualbp) represents one of galbp. its additional practical constraints are related the shape of the line, where the modification here is done on the precedence constraints by considering the successors beside the predecessors in the assignment procedure. the reason of having the successors in the assignment procedure is the shape of the line to cite this article: zakaraia, m., zaher, h., ragaa, n. (2022). solving stochastic multi-manned u-shaped assembly line balancing problem using differential evolution algorithm. international journal of production management and engineering, 10(1), 13-22. https://doi.org/10.4995/ijpme.2022.16084 http://polipapers.upv.es/index.php/ijpme creative commons attribution-noncommercial-noderivatives 4.0 international int. j. prod. manag. eng. (2022) 10(1), 13-22 13 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.4995/ijpme.2021.16084 allows to have the operators inside the line, where they can perform tasks either in the entrance or the exist sides of the line. the multi-manned assembly line balancing problem (malbp) is another form of galbp. it adds another set of constraints to the problem by considering the sequencing of the tasks that may restrict assign tasks to the additional operators in the same station. the contribution of this paper is that it combines both of ualbp and malbp in one problem and presents a new mixedinteger programming model for such new problem. in addition, the problem is solved under uncertainty by having the stochastic processing times of the tasks. therefore, the mathematical model shows the cycle time constraints as chance-constraints. due to the combinatorial nature of the problem that makes it one of the np-hard problems, the selected approach for solving it is an efficient metaheuristic called differential evolution algorithm (de). the paper is organized as follows. the second section shows a literature review that covers some of word presented in both of ualbp and malbp. the third section presents the mathematical model of the new problem. the fourth section illustrates the developed de for solving the problem. the fifth section shows a numerical example. the sixth section discusses the parameters of the proposed de and optimizing its parameters using design of experiments (doe). eventually, the seventh section is the conclusion. 2. literature review this paper discusses a combination between ualbp and malbp. therefore, this section covers some the previous works which are presented in both problems. the first work in ualbp is presented by miltenburg and wijngaard (1994). they showed the advantages of using ualbp instead of using salbp. ualbp in research can be classified into three categories according to objective functions: type-1, type-2, and type-e (rabbani, kazemi, and manavizadeh, 2012). type-1 is concerned with minimizing the number of stations for a given cycle time (yilmaz et al., 2020). type-2 objective is to minimize the cycle time with a given number of stations. type-e is to maximize the line efficiency when the number of stations and the cycle time are unknown (oksuz, buyukozkan, and satoglu, 2017). this paper considers multiobjectives for the new problem. the first objective is to minimize the number of stations and the second is to minimize the number of operators. the reason of having the number of operators as an objective is due to having the multi-manned concept of malbp. in terms of minimizing the number of stations, there are a lot of papers that were presented in ualbp. for example, ajenblit and wainwright (1998) developed an ordered-based genetic algorithm for solving ualbp type-1 (ualbp-1). scholl and klein (scholl and klein, 1999) proposed a branch and bound procedure for solving many types of ualbp. gökçen et al. (2006) presented a shortest route formulation for ualbp. sabuncuoglu et al. (2009) developed ant colony optimization for solving ualbp. kara et al. (2011) presented a resource dependent mathematical model for ualbp. hamzadayi and yildiz (2012) presented a genetic algorithm for solving ualbp in case of having parallel stations and mixed models. hamzadayi and yildiz (2013) developed a simulated annealing for solving ualbp in case of assembling mixed models. jayaswal and agarwal (2014) proposed a simulated annealing approach the resource dependent ualbp. kucukkoc and zhang (2015) proposed a hybrid design for the assembly line balancing problem that combines ualbp with the parallel assembly line balancing problem. they developed a heuristic procedure for solving such hybrid design. fathi et al. (2016) developed a simulated annealing based proposed heuristic for solving ualbp. li et al. (2017) presented a rulesbased heuristic for solving ualbp. sresracoo et al. (2018) developed de algorithm for solving ualbp-1. nourmohammadi et al. (2019) proposed a water flow inspired algorithm for solving ualbp. zhang and xu (2020) considered the energy cost in objective functions and solved ualbp using an improved flower pollination algorithm. yılmaz (2020a) produced a robust optimizatoin for the u-shaped assembly line balancing problem with worker assignment in case that the processing times of the tasks are uncertain. ö. f. yılmaz (2020b) developed a mathematical model for an integrated bi-objective u-shaped assembly line balancing problem that considers heterogenity inhert of workers. he aimed to minimize the operational cost and workload balance. li et al. (2021) developed an enhanced beam search heuristic for solving both of type-1 and type-2 of ualbp. the first mathematical model of malbp showed up in (fattahi, roshani, and roshani, 2011a). they solved the problem using ant colony optimization algorithm. fattahi et al. (2011b) developed an improved simulated annealing for solving malbp. their objectives are to minimize the number of zakaraia et al. creative commons attribution-noncommercial-noderivatives 4.0 internationalint. j. prod. manag. eng. (2022) 10(1), 13-2214 http://creativecommons.org/licenses/by-nc-nd/4.0/ stations and to maximize the line efficiency. kellegöz and toklu (2015) proposed a genetic algorithm for solving malbp. their objective is to minimize the number of stations. kellegöz (2017) another mathematical model for malbp. he proposed a simulated annealing that uses a gantt-based heuristic for solving the problem. michels (2018) proposed genetic algorithm for solving malbp. his objective is to minimize the costs per production unit. michels et al. (2019) presented a mixed integer programming model for malbp that can be solved using benders’ decomposition algorithm. abidin çil and kizilay (2020) proposed a constrainted programming approach to solve malbp. their objectives are to minimize the cycle time as primary objective and to minimize the number of stations as secondary objective. zhang et al. (2020) developed an ant colony optimization algorithm that solves malbp in case of having space constraints. to the best of knowledge, the only work that considers the combination of ualbp and malbp is presented by zakaraia et al. (2021), which the problem was sovled using stochastic local search (sls). the proposed de algorithm herein is more intellegent than sls, where the proposed de contains better priority structure for constructing feasible solutions and it contains learning procedures to ignore the worse solutions from the search space by replacing them with new random ones to increase exploration. 3. the optimization model as aforementioned, this study concerns with combining ualbp and malbp under uncertainty by having the processing times of the tasks as stochastic random variables with known means and variances and they are normally distributed. therefore, the cycle time constraints are formulated using probabilistic constraints that are restricted by predetermined chance probability. the mathematical model can be formulated as follows: 3.1. notations i={1,…, n} the set of tasks j={1,…, m} the set of stations k={1,…,l} the set of operators ct cycle time kmax the maximum number of operators in any station ti the processing time of task i (random variable) e(ti) the expected processing time of task i var(ti) the variance of task i ip(ti) the immediate predecessors of task i is(ti) the immediate successors of tasks i 3.2. decision variables x ijk 1 0 otherwise if task i assigned to operator k in station j y j 1 if n i 1 x ijk 0 0 otherwise the opened station decision variable r k 1 if n i 1 x ijk 0 0 otherwise the operator assignment decision variable p i 1 if m j 1 j x ijk m j 1 j x hjk h ip i 0 otherwise the immediate predecessor’s assignment decision variable s i 1 if m j 1 j x ijk m j 1 j x hjk h is i 0 otherwise the immediate successors assignment decision variables 3.3. the objective function min m j 1 y j 1 l k 1 r k (1) 3.4. the constraints m j 1 x ijk 1 i {1 n } (2) p n i 1 t i x ijk k m ax ct α j {1 m } (3) solving stochastic multi-manned u-shaped assembly line balancing problem using differential evolution algorithm creative commons attribution-noncommercial-noderivatives 4.0 international int. j. prod. manag. eng. (2022) 10(1), 13-22 15 http://creativecommons.org/licenses/by-nc-nd/4.0/ p ti xijk km ax k 1 g ip i tg xgjk kmax k 1 h i s i th x hjk ct α j {1 m } (4) p n i 1 ti xijk ct α k {1 l } (5) p i s i 1 i {1 n } (6) the objective function (1) seeks to minimize the number of stations and the number of operators. the set of constraints (2) is the set of assignment constraints, which ensures that each task is assigned in only one station. the set of constraints (3) shows that the maximum total processing time of the assigned tasks of any station must be less than or equal the cycle time multiplied by the maximum number of operators. the set of constraints (4) is the sequencing constraints that ensures that if any of the immediate predecessors or successors of any task are assigned on its station, then their processing time added to the task processing time must not exceed the cycle time. this set of constraints helps to avoid assigning tasks to more than one operator without considering the processing times of the predecessors and successors. the set of constraints (5) is the cycle time constraints for operators, which ensures that the total processing time of any assigned tasks to an operator must not exceed the cycle time. the sets of constraints (3), (4), and (5) are sets of chance-constraints, which are restricted by a predetermined chance probability. the set of constraints (6) is the set of predecessors and successors constraints, which ensures that each task has to be assigned either before its immediate predecessors or before immediate successors. the mathematical model contains some chanceconstraints that can be converted into deterministic in order to be solved. the processing times herein are normally distributed random variables with known means and variance. taha (2017) shows how to overcome the chance-constraints through converting them into non-linear deterministic constraints. therefore, the set of constraints (3), (4), and (5) can be converted into the non-linear deterministic form as shown in (7), (9), and (8) respectively. n i 1 e(tl) xijk kα n i 1 var(ti)x ijk kmax ct j { }1 m where kα is the standard normal value of α (7) e(ti)xijk (ti)xijk kmax k 1 kmax k 1 g ip(i) e(tg)xgjk kmax k 1 h is(i) e(th)xhjk kα var ct j {1 m } g ip(i) var(tg)xgjk var(th)xhjk kmax k 1 h is(i) (8) n i 1 e ti x ijk kα n i 1 var ti x ijk ct k { }1 l (9) 4. the proposed de algorithm de is one of the population-based metaheuristics that consists of four phases. the first phase is the initialization, which concerns with generating the initial population of solutions. the second phase concerns with the mutation procedure. the third phase is concerned with the crossover procedure. the fourth phase is the selection procedure. all of these phases work iteratively until reaching the stopping criterion, which is herein the number of iterations. 4.1. the initialization phase in the initialization phase, a set of random solutions is to be generated in order to cover diversified areas of the solution space. the problem here can have random solution by generating random sequence of the tasks. such random sequence (t) represents the priority of the tasks. so, any opened station will have the top priority tasks that satisfy the problem constraints through using the following heuristics: algorithm 1: the heuristic procedure j=1, station sj=ø, and solution =ø while t≠ø do: find the assignable tasks (as) that ensure the problem constrains if as≠ø then: assign the highest priority (p) task in t to sj t=t–{p} zakaraia et al. creative commons attribution-noncommercial-noderivatives 4.0 internationalint. j. prod. manag. eng. (2022) 10(1), 13-2216 http://creativecommons.org/licenses/by-nc-nd/4.0/ else: solution = solution ∪ sj j = j+1 and sj=ø return solution de algorithm uses vectors in its search methodology. therefore, the random sequence of tasks can be generated using a vector that its length equal to the number of tasks and its values are randomly generated using the following equation: vi= rand (0,1) (10) such random vector represents the position of the solution. the next step of generating the random sequence using such generated vector is to use the bubble sort algorithm as follows: algorithm 2: bubble sort for creating a random sequence of tasks t=the tasks vector arranged by index number in ascending order n=the number of tasks continue=1 i=1 while continue=1 do: continue=0 i'=1 while i'≤ n–i do: if vi' ≤vi' +1 then temporary1=vi' temporary2=ti' vi'=vi'+1 ti'=ti'+1 vi'+1=temporary1 ti'+1=temporary2 continue=1 i'=i'+1 return t by using equation (10), algorithm 1, and algorithm 2, the initial population can have a set of randomly generated candidate solutions c(s). the next steps of the algorithm are to update the solutions of the population through mutation and crossover procedures and either select to keep solutions or replace them. all of these steps are to be done in iterative manner until reaching a stopping criterion, which is herein the number of iterations. 4.2. the mutation phase in the mutation phase, each solution in the population is to be mutated using its positional vector. the mutation procedure uses three positions to generate new position posnew. the first position is the position of the best solution found posbest. the second position is the position of the current solution poscurrent . the third position posother is a position of randomly selected solution from the population that isn’t equal to the current solution. posnew=posbest–β(poscurrent–posother) (11) algorithm 1 now is ready to be used to obtain new solution using posnew. the new solution is to be compared with the best solution found and replaces it if it is better. the parameter β is called the differential weight, which helps to define how far the new position from the three used positions. 4.3. the crossover phase the crossover procedure uses both of posnew and posother. in such process a new position is to be generated by having properties from t posnew and poscurrent. the process is controlled by a new parameter cr, which is the crossover probability. so, each value of the crossover position poscross can be generated using the following equation: poscross i i if rand 0 1 cr i otherwise poscurrent posnew (12) the solution of poscross is to be generated and it will replace the best solution if it is better. 4.4. the selection phase the selection phase determines the new position of the new solution in the next iteration. so, it can be replaced by posnew if it produces a better solution than the current solution, or it will be replaced by another solution that is generated randomly using a random position posrandom by using equation (10). solving stochastic multi-manned u-shaped assembly line balancing problem using differential evolution algorithm creative commons attribution-noncommercial-noderivatives 4.0 international int. j. prod. manag. eng. (2022) 10(1), 13-22 17 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5. numerical example this section shows a numerical example to further illustrate the model and the proposed de algorithm. the numerical example consists of 6 tasks. its cycle time is 8 and its precedence graph is shown in figure 1. the number of operators is 2 and the chance probability is 0.95. figure 1. the precedence graph of the numerical example. in the initialization phase, the population of solutions is to be generated using algorithm 1 and 2. so, the generation of one of these solutions can be illustrated as follows. firstly, the number of tasks herein is 6. therefore, one of the random vectors (rv) can be shown as follows: tasks 1 2 3 4 5 6 rv 0.3 0.8 0.9 0.4 0.2 0.5 after using the proposed bubble sort, algorithm 2, the arrangement of the tasks that should be used with the heuristic algorithm is as follows: tasks 3 2 6 4 1 5 rv 0.9 0.8 0.5 0.4 0.3 0.2 now algorithm 1, the heuristic procedure, is ready to be used. table 1 shows the heuristic solution using rv vector. table 1. the heuristic solution using rv. task 3 6 1 2 4 5 processing time 4 6 1 5 3 5 station 1 1 1 2 2 3 operator 1 2 2 3 3 4 completion time 4 6 7 5 8 5 the mutation phase is about to find another solution in the local space of the current solution using the linear combination found in equation (11), where it uses the position vector of the best solution and a position vector of a random solution. if it considered that the solution found in table 1 is the best solution, then its position vector, which is the sorted rv vector (srv), is to be used along with another position vector of a random solution to find a neighbor of the current solution. for illustration, vc represents the position vector of the current position, vcr represents the position vector of a random solution, and vcn represents the position vector of the neighbor solution. table 2 shows the vcn after using β=0.3. table 2. generating neighbour using mutation phase. tasks 1 2 3 4 5 6 vc 0.9 0.6 0.5 0.4 0.2 0.1 srv 0.9 0.8 0.5 0.4 0.3 0.2 vcr 0.9 0.8 0.7 0.3 0.2 0.1 vcn 0.9 0.9 0.6 0.35 0.3 0.2 the crossover phase is about to generated new position vector using the position vector of the current solution and the position of the random solution, where this process selects characteristics from both solutions. table 3 shows the crossover process, where the highlighted numbers show the selected characteristics from each solution. table 3. generating new solution using crossover process. tasks 1 2 3 4 5 6 vc 0.9 0.6 0.5 0.4 0.2 0.1 vcr 0.9 0.8 0.7 0.3 0.2 0.1 vcn 0.9 0.6 0.5 0.3 0.2 0.1 6. experimental design the proposed algorithm is developed using python programming in pc that has 2.93 ghz core2duo cpu and 4 gb rams. it has five parameters, which are the population size popsize, the number of iterations maxit, the minimum values of the differential weight βmin, the maximum value of the differential weight βmax, and the crossover probability cp. each parameter has four levels shown in table 4. the number of experiments required to make the full factorial design is 45=1024 experiments. such number of experiments can be radically reduced using the taguchi method by having l16 orthogonal array, which only have 16 experiments. the corresponding l16 orthogonal array for the current experimental design is in shown in table 5. table 4. the parameter levels of the experimental design. popsize maxit βmin βmax cp 25 25 -1 0 0.1 50 50 -0.5 0.25 0.2 75 75 0 0.5 0.3 100 100 0.5 1 0.4 zakaraia et al. creative commons attribution-noncommercial-noderivatives 4.0 internationalint. j. prod. manag. eng. (2022) 10(1), 13-2218 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 5. the required orthogonal array for the experimental design. trail popsize maxit βmin βmax cp 1 25 25 -1 0 0.1 2 25 50 -0.5 0.25 0.2 3 25 75 0 0.5 0.3 4 25 100 0.5 1 0.4 5 50 25 -0.5 0.5 0.4 6 50 50 -1 1 0.3 7 50 75 0.5 0 0.2 8 50 100 0 0.25 0.1 9 75 25 0 1 0.2 10 75 50 0.5 0.5 0.1 11 75 75 -1 0.25 0.4 12 75 100 -0.5 0 0.3 13 100 25 0.5 0.25 0.3 14 100 50 0 0 0.4 15 100 75 -0.5 1 0.1 16 100 100 -1 0.5 0.2 the selected problems for design of experiments are taken from well-known benchmarks can be found in https://assembly-line-balancing.de/salbp/. the problems included in such benchmarks are deterministic and need to be adapted to fit the mathematical model of this paper. therefore, the processing times of tasks in the selected problems must have expected values and variances. in order to adapt the problems, the expected values of the processing times are considered the same as the values of the original processing times and the variances are calculated by subtracting each processing time from the expected value and divide the output by 1000. table 6 shows the selected problems for the experimental design. table 6. the selected problems for experimental design. serial problem cycle time number of tasks 1 jackson 7 11 2 jackson 9 11 3 mitchell 14 21 4 mitchell 15 21 5 heskia 138 28 6 heskia 205 28 7 sawyer30 25 30 8 sawyer30 27 30 9 arc83 5048 83 10 arc83 5853 83 11 arc111 5755 111 12 arc111 8847 111 the response value for the experimental design includes the value of the objective function and the cpu time where that leads to better solutions with a little time consumption. equation (13) shows the required response value for each trail in the experimental design. response m j 1 y j 1 l k 1 r k 1 cpu time (13) the selected problems are different and each has different solution and response value. therefore, the response values are normalized as shown in table 7. the analysis of variance for the parameter levels is done in order to study main effects. table 8 shows the f-value and p-value for each parameter. table 7. the normalized values for the responses for each selected problem. trail 1 2 3 4 5 6 7 8 9 10 11 12 1 0.062 0.063 0.063 0.062 0.062 0.069 0.061 0.062 0.061 0.062 0.063 0.062 2 0.062 0.063 0.063 0.063 0.060 0.061 0.062 0.059 0.061 0.058 0.063 0.064 3 0.062 0.063 0.062 0.063 0.063 0.057 0.058 0.060 0.063 0.063 0.063 0.063 4 0.062 0.062 0.062 0.062 0.061 0.063 0.062 0.067 0.064 0.063 0.063 0.064 5 0.062 0.063 0.062 0.063 0.065 0.073 0.060 0.061 0.064 0.064 0.063 0.059 6 0.062 0.063 0.062 0.062 0.061 0.056 0.063 0.061 0.067 0.058 0.063 0.064 7 0.063 0.063 0.062 0.062 0.064 0.069 0.063 0.067 0.067 0.064 0.063 0.063 8 0.063 0.062 0.062 0.062 0.064 0.061 0.058 0.067 0.061 0.063 0.063 0.061 9 0.062 0.063 0.062 0.062 0.061 0.069 0.067 0.061 0.061 0.058 0.063 0.063 10 0.062 0.062 0.063 0.063 0.062 0.057 0.061 0.059 0.061 0.064 0.063 0.062 11 0.063 0.063 0.062 0.063 0.065 0.060 0.063 0.065 0.061 0.065 0.059 0.063 12 0.062 0.063 0.062 0.063 0.066 0.055 0.063 0.066 0.061 0.065 0.063 0.062 13 0.062 0.063 0.063 0.063 0.065 0.058 0.063 0.059 0.065 0.063 0.063 0.062 14 0.063 0.063 0.062 0.062 0.060 0.060 0.063 0.059 0.060 0.065 0.059 0.061 15 0.062 0.063 0.062 0.062 0.061 0.072 0.069 0.066 0.061 0.064 0.063 0.063 16 0.062 0.063 0.063 0.063 0.061 0.062 0.061 0.060 0.061 0.059 0.061 0.065 solving stochastic multi-manned u-shaped assembly line balancing problem using differential evolution algorithm creative commons attribution-noncommercial-noderivatives 4.0 international int. j. prod. manag. eng. (2022) 10(1), 13-22 19 https://assembly-line-balancing.de/salbp/ http://creativecommons.org/licenses/by-nc-nd/4.0/ table 8. the analysis of variance for each parameter. parameter f-value p-value popsize 0.59 0.62 maxit 3.54 0.016 βmin 1.42 0.24 βmax 1.60 0.19 cp 0.45 0.71 the null hypothesis is accepted in all parameters except in the number of iterations. therefore, the tukey’s honest significant difference test is applied to show which parameter levels differ. figure 2 shows that the worst parameter level for the number of iterations is 75 iterations and there is no significant difference between the remaining levels. figure 2. tukey’s interval plot for the number of iterations parameter. 7. computational results this section shows the results of applying the proposed de algorithm on 71 adapted problems for the same benchmarks found in the experimental design section. table 9 shows the computational results. table 9. the computational results. problem problem size cycle time result cpu time mertens 7 6 3.83 0.001 mertens 7 7 2.83 0.001 mertens 7 8 2.83 0.001 mertens 7 10 2.75 0.001 mertens 7 15 1.5 0.001 mertens 7 18 0.5 0.005 bowman8 8 17 4.83 0.001 bowman8 8 20 3.8 0.003 bowman8 8 21 3.8 0.005 bowman8 8 24 3.8 0.001 bowman8 8 28 2.8 0.003 bowman8 8 31 1.67 0.001 jaeschke 9 6 5.88 0.001 jaeschke 9 7 5.86 0.001 jaeschke 9 8 5.86 0.002 problem problem size cycle time result cpu time jaeschke 9 10 3.8 0 jaeschke 9 18 2.67 0 jackson 11 7 5.88 0.029 jackson 11 9 4.86 0.001 jackson 11 10 3.83 0.003 jackson 11 13 2.75 0.005 jackson 11 14 2.75 0.001 jackson 11 21 1.67 0.001 mansoor 11 45 2.8 0.019 mansoor 11 54 2.75 0.001 mansoor 11 63 1.67 0.026 mansoor 11 72 1.67 0.001 mansoor 11 81 1.67 0.001 mitchell 21 14 7.9 0.002 mitchell 21 15 6.9 0.004 mitchell 21 21 3.86 0.023 mitchell 21 26 2.8 2.247 mitchell 21 35 2.75 0.003 mitchell 21 39 1.67 2.439 heskia 28 138 4.88 0.192 heskia 28 205 2.83 0.833 heskia 28 216 2.8 0.086 heskia 28 256 2.8 0.009 heskia 28 324 1.75 0.013 heskia 28 342 1.75 0.014 sawyer30 30 25 7.94 6.457 sawyer30 30 27 7.93 0.51 sawyer30 30 30 7.92 0.132 sawyer30 30 36 5.9 0.016 sawyer30 30 41 4.89 8.426 sawyer30 30 54 3.86 0.027 sawyer30 30 75 2.8 0.008 kilbrid 45 57 5.9 1.401 kilbrid 45 79 3.88 0.016 kilbrid 45 92 3.86 0.01 kilbrid 45 110 2.83 0.277 kilbrid 45 138 2.8 0.009 kilbrid 45 184 1.67 15.791 tonge70 70 176 11.96 0.447 tonge70 70 364 5.91 0.092 tonge70 70 410 4.89 1.524 tonge70 70 468 3.88 4.191 tonge70 70 527 3.86 0.018 arc83 83 5048 8.94 18.726 arc83 83 5853 7.93 3.26 arc83 83 6842 6.92 32.885 arc83 83 7571 5.91 9.666 arc83 83 8412 5.9 1.057 arc83 83 8998 4.89 7.431 arc83 83 10816 3.88 1.209 arc111 111 5755 15.97 31.141 arc111 111 8847 9.95 0.43 arc111 111 10027 8.94 13.533 arc111 111 10743 7.93 88.503 arc111 111 11378 7.93 3.89 arc111 111 17067 4.89 8.653 zakaraia et al. creative commons attribution-noncommercial-noderivatives 4.0 internationalint. j. prod. manag. eng. (2022) 10(1), 13-2220 http://creativecommons.org/licenses/by-nc-nd/4.0/ 8. conclusion the problem handled in this paper considers a combination between ualbp and malbp under uncertainty. such combination leads to minimize the line length through having more than one operator in any station and utilizing the flexibility of the task’s assignment in u-shaped lines. the processing times of the tasks differ from operator to another, where that leads to uncertain values of them. thus, the processing times of the tasks are represented as random variables with known means and variances. therefore, the cycle time constraints of the mathematical model for such combined problem are represented as chance-constraints. the proposed approach for solving the problem is de algorithm. the algorithm parameters are optimized and 71 adapted problems have been solved as a computational result. the future points of research may include the following: formulating the same problem with another type of uncertainty such as fuzzy and rough programming. including space constraints. including worker assignment. proposing other approaches for solving the same problem. references abidin çil, zeynel, & damla kizilay. 2020. constraint programming model for multi-manned assembly line balancing problem. computers and operations research, 124, 105069. https://doi.org/10.1016/j.cor.2020.105069 ajenblit, debora a., & roger l. wainwright. 1998. applying genetic algorithms to the u-shaped assembly line balancing problem. proceedings of the ieee conference on evolutionary computation, icec, 96–101. https://doi. org/10.1109/icec.1998.699329 fathi, masood, maría jesús álvarez, & victoria rodríguez. 2016. a new heuristic-based bi-objective simulated annealing method for u-shaped assembly line balancing. european journal of industrial engineering, 10(2), 145– 169. https://doi.org/10.1504/ejie.2016.075849. fattahi, parviz, abdolreza roshani, & abdolhassan roshani. 2011a. a mathematical model and ant colony algorithm for multi-manned assembly line balancing problem. international journal of advanced manufacturing technology, 53(1–4), 363–378. https://doi.org/10.1007/s00170-010-2832-y gökçen, hadi, kürşad ağpak, & recep benzer. 2006. balancing of parallel assembly lines. international journal of production economics, 103(2), 600–609. https://doi.org/10.1016/j.ijpe.2005.12.001 hamzadayi, alper, & gokalp yildiz. 2012. a genetic algorithm based approach for simultaneously balancing and sequencing of mixed-model u-lines with parallel workstations and zoning constraints. computers and industrial engineering, 62(1), 206–215. https://doi.org/10.1016/j.cie.2011.09.008 hamzadayi, alper, & gokalp yildiz. 2013. a simulated annealing algorithm based approach for balancing and sequencing of mixed-model u-lines. computers and industrial engineering, 66(4), 1070–1084. https://doi. org/10.1016/j.cie.2013.08.008 jayaswal, sachin, & prashant agarwal. 2014. balancing u-shaped assembly lines with resource dependent task times: a simulated annealing approach. journal of manufacturing systems, 33(4), 522–534. https://doi.org/10.1016/j. jmsy.2014.05.002 kara, yakup, cemal özgüven, neşe yalçin, & yakup atasagun. 2011. balancing straight and u-shaped assembly lines with resource dependent task times. international journal of production research, 49(21), 6387–6405. https://doi. org/10.1080/00207543.2010.535039 kellegöz, talip. 2017. assembly line balancing problems with multi-manned stations: a new mathematical formulation and gantt based heuristic method. annals of operations research, 253(1), 377–404. https://doi.org/10.1007/s10479016-2156-x kellegöz, talip, & bilal toklu. 2015. a priority rule-based constructive heuristic and an improvement method for balancing assembly lines with parallel multi-manned workstations. international journal of production research, 53(3), 736–756. https://doi.org/10.1080/00207543.2014.920548 kucukkoc, ibrahim, & david z. zhang. 2015. balancing of parallel u-shaped assembly lines. vol. 64. virginia tech. li, ming, qiuhua tang, qiaoxian zheng, xuhui xia, & c. a. floudas. 2017. rules-based heuristic approach for the u-shaped assembly line balancing problem. applied mathematical modelling, 48(2017), 423–439. https://doi. org/10.1016/j.apm.2016.12.031 solving stochastic multi-manned u-shaped assembly line balancing problem using differential evolution algorithm creative commons attribution-noncommercial-noderivatives 4.0 international int. j. prod. manag. eng. (2022) 10(1), 13-22 21 https://doi.org/10.1016/j.cor.2020.105069 https://doi.org/10.1109/icec.1998.699329 https://doi.org/10.1109/icec.1998.699329 https://doi.org/10.1504/ejie.2016.075849 https://doi.org/10.1007/s00170-010-2832-y https://doi.org/10.1016/j.ijpe.2005.12.001 https://doi.org/10.1016/j.cie.2011.09.008 https://doi.org/10.1016/j.cie.2013.08.008 https://doi.org/10.1016/j.cie.2013.08.008 https://doi.org/10.1016/j.jmsy.2014.05.002 https://doi.org/10.1016/j.jmsy.2014.05.002 https://doi.org/10.1080/00207543.2010.535039 https://doi.org/10.1080/00207543.2010.535039 https://doi.org/10.1007/s10479-016-2156-x https://doi.org/10.1007/s10479-016-2156-x https://doi.org/10.1080/00207543.2014.920548 https://doi.org/10.1016/j.apm.2016.12.031 https://doi.org/10.1016/j.apm.2016.12.031 http://creativecommons.org/licenses/by-nc-nd/4.0/ li, zixiang, mukund nilakantan janardhanan, & humyun fuad rahman. 2021. enhanced beam search heuristic for u-shaped assembly line balancing problems. engineering optimization, 53(4), 594–608. https://doi.org/10.1080/03 05215x.2020.1741569 michels, adalberto sato, tiago cantos lopes, celso gustavo stall sikora, & leandro magatão. 2018. with practical extensions the robotic assembly line design ( rald ) problem: model and case studies with practical extensions. computers & industrial engineering, 120, 320–333. https://doi.org/10.1016/j.cie.2018.04.010 michels, adalberto sato, thiago cantos lopes, celso gustavo stall sikora, & leandro magatão. 2019. a benders’ decomposition algorithm with combinatorial cuts for the multi-manned assembly line balancing problem. european journal of operational research, 278(3), 796–808. https://doi.org/10.1016/j.ejor.2019.05.001 miltenburg, g.j., & j. wijngaard. 1994. u-line line balancing problem. management science, 40(10), 1378–1388. https://doi.org/10.1287/mnsc.40.10.1378 nourmohammadi, amir, masood fathi, mostafa zandieh, & morteza ghobakhloo. 2019. a water-flow like algorithm for solving u-shaped assembly line balancing problems. ieee access, 7, 129824–129833. https://doi.org/10.1109/ access.2019.2939724 oksuz, mehmet kursat, kadir buyukozkan, & sule itir satoglu. 2017. u-shaped assembly line worker assignment and balancing problem: a mathematical model and two meta-heuristics. computers and industrial engineering, 112, 246–263. https://doi.org/10.1016/j.cie.2017.08.030 rabbani, masoud, seyed mahmood kazemi, & neda manavizadeh. 2012. mixed model u-line balancing type-1 problem: a new approach. journal of manufacturing systems, 31(2), 131–138. https://doi.org/10.1016/j.jmsy.2012.02.002 sabuncuoglu, ihsan, erdal erel, & arda alp. 2009. ant colony optimization for the single model u-type assembly line balancing problem. international journal of production economics, 120(2), 287–300. https://doi.org/10.1016/j.ijpe.2008.11.017 scholl, a., & r. klein. 1999. ulino: optimally balancing u-shaped jit assembly lines. international journal of production research, 37(4), 721–736. https://doi.org/10.1080/002075499191481 sresracoo, poontana, nuchsara kriengkorakot, preecha kriengkorakot, & krit chantarasamai. 2018. u-shaped assembly line balancing by using differential evolution algorithm. mathematical and computational applications, 23(4), 79. https://doi.org/10.3390/mca23040079 taha, hamdy. 2017. operations research an introduction. 10th ed. edited by r. horton, marcia; partridge, julian; stark, holly; brands, amanda; agarwal, aditee; raheja. harlow: pearson education limited. yilmaz, ö.f., ö.f. demirel, s. zaim, & s. sevim. 2020. assembly line balancing by using axiomatic design principles: an application from cooler manufacturing industry. international journal of production management and engineering, 8(1), 31–43. https://doi.org/10.4995/ijpme.2020.11953 yılmaz, faruk. 2020a. robust optimization for u-shaped assembly line worker assignment and balancing problem with uncertain task times¨omer times¨ times¨omer. crorr, 11(2), 229–239. https://doi.org/10.17535/crorr.2020.0018 yılmaz, ömer faruk. 2020b. an integrated bi-objective u-shaped assembly line balancing and parts feeding problem: optimization model and exact solution method. annals of mathematics and artificial intelligence, 1–18. https://doi.org/10.1007/s10472-020-09718-y zakaraia, mohammad, hegazy zaher, & naglaa ragaa. 2021. stochastic local search for solving chance-constrained multi-manned u-shaped assembly line balancing problem with time and space constraints. journal of university of shanghai for science and technology, 23(04), 278–295. https://doi.org/10.51201/jusst/21/04242 zhang, beikun, & liyun xu. 2020. an improved flower pollination algorithm for solving a type-ii u-shaped assembly line balancing problem with energy consideration. assembly automation, 40(6), 847–856. https://doi.org/10.1108/aa-07-2019-0144 zhang, zikai, qiuhua tang, & manuel chica. 2020. multi-manned assembly line balancing with time and space constraints: a milp model and memetic ant colony system. computers and industrial engineering, 150, 106862. https://doi.org/10.1016/j.cie.2020.106862 zakaraia et al. creative commons attribution-noncommercial-noderivatives 4.0 internationalint. j. prod. manag. eng. (2022) 10(1), 13-2222 https://doi.org/10.1080/0305215x.2020.1741569 https://doi.org/10.1080/0305215x.2020.1741569 https://doi.org/10.1016/j.cie.2018.04.010 https://doi.org/10.1016/j.ejor.2019.05.001 https://doi.org/10.1287/mnsc.40.10.1378 https://doi.org/10.1109/access.2019.2939724 https://doi.org/10.1109/access.2019.2939724 https://doi.org/10.1016/j.cie.2017.08.030 https://doi.org/10.1016/j.jmsy.2012.02.002 https://doi.org/10.1016/j.ijpe.2008.11.017 https://doi.org/10.1080/002075499191481 https://doi.org/10.3390/mca23040079 https://doi.org/10.4995/ijpme.2020.11953 https://doi.org/10.17535/crorr.2020.0018 https://doi.org/10.1007/s10472-020-09718-y https://doi.org/10.51201/jusst/21/04242 https://doi.org/10.1108/aa-07-2019-0144 https://doi.org/10.1016/j.cie.2020.106862 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.15217 received: 2021-03-06 accepted: 2021-05-02 geometric and harmonic means based priority dispatching rules for single machine scheduling problems ahmad, s.a1 , khan, z.a.a2, ali, m.b, asjad, m.a3 adepartment of mechanical engineering, faculty of engineering and technology, jamia millia islamia, new delhi, india. bdepartment of mechanical engineering, aligarh muslim university, aligarh, uttar pradesh, india. a1 shafiahmad.amu@gmail.com, a2 zakhanusm@yahoo.com, b mohdali234@rediffmail.com, a2 masjad@jmi.ac.in abstract: this work proposes two new priority dispatching rules (pdrs) for solving single machine scheduling problems. these rules are based on the geometric mean (gm) and harmonic mean (hm) of the processing time (pt) and the due date (dd) and they are referred to as gmpd and hmpd respectively. performance of the proposed pdrs is evaluated on the basis of five measures/criteria i.e. total flow time (tft), total lateness (tl), number of late jobs (tnl), total earliness (te) and number of early parts (tne). it is found that gmpd performs better than other pdrs in achieving optimal values of multiple performance measures. further, effect of variation in the weight assigned to pt and dd on the combined performance of tft and tl is also examined which reveals that for deriving optimal values of tft and tl, weighted harmonic mean (whmpd) rule with a weight of 0.105 outperforms other pdrs. the weighted geometric mean (wgmpd) rule with a weight of 0.37 is found to be the next after whmpd followed by the weighted pdt i.e. wpdt rule with a weight of 0.76. key words: job sequencing, priority dispatching rule, single machine scheduling, geometric mean of the processing time and due date (gmpd), harmonic mean of the processing time and due date (hmpd). 1. introduction production scheduling refers to the planning of the manufacturing process and it is an important activity as it leads to enhancement in the productivity of the system, reduction in job lateness, increased utilization of the machine etc. (doh et  al., 2013; geiger and uzsoy, 2008; pinedo, 2009). apart from manufacturing processes, scheduling finds its application in various other areas like operating systems where it is used for memory allocation of processor, in service industries for operator allotment etc. (baharom et al., 2015; lee et al., 2020; munir et al., 2008; rafsanjani and bardsiri, 2012). however, it is imperative in production scheduling due to its undeviating impact on the profitability of a company (kadipasaoglu et al., 1997; prakash et al., 2011). the effectiveness of a schedule is very much affected by the priority dispatching rule (pdr) used for job sequencing (hussain and ali, 2019; krishnan et al., 2012; waikar et al., 1995). these rules offer priority to one or more jobs over other jobs to improve certain performance measures of the system. the well-known pdrs found in the literature are as follows (forrester, 2006; pinedo, 2009): first come first served (fcfs): a job that arrives first will be given the highest priority for processing. shortest processing time (spt): a job having the least processing time (pt) will be processed first. earliest due date (edd): a job with a least due date (dd) will attain the highest priority and will be processed first. to cite this article: ahmad, s., khan, z.a., ali, m., asjad, m. (2021). geometric and harmonic means based priority dispatching rules for single machine scheduling problems. international journal of production management and engineering,  9(2), 93-102. https://doi.org/10.4995/ijpme.2021.15217 int. j. prod. manag. eng. (2021) 9(2), 93-102creative commons attribution-noncommercial-noderivatives 4.0 international 93 https://orcid.org/0000-0002-2193-8984 http://creativecommons.org/licenses/by-nc-nd/4.0/ slack: a slack time of each job is computed which is the difference between the remaining time and the pt. subsequently, a job with the least slack value will be processed first. critical ratio (cr): a ratio of time remaining until the dd andpt is calculated which is identified as a critical ratio. consequently, a job with the least critical ratio will be processed first. the choice of pdr has a significant effect on the overall performance of the system (waikar et al., 1995). however, it was observed that no single pdr is capable of optimizing all the performance measures (chan et al., 2003; dominic et al., 2004; ðurasević and jakobović, 2018). hence, the current era of research in scheduling is focused on the development of new pdrs to achieve optimal values of more than one performance measures (da silva et al., 2019; kanet and li, 2004; lu et al., 2012). holthaus and rajendran (1997) presented two new pdrs using additive and alternative strategies to combine process time and work content in the next queue and based on their simulation study they reported that this rule performed well to minimize the average flow time. jayamohan and rajendran (2007) proposed five new pdrs by combining different rules and showed that the proposed rules performed well for minimizing mean flow time, mean tardiness and percentage of tardy jobs. dominic et al. (2004) examined the effectiveness of three rules developed by combining fifo, spt and most work remaining (mwk) pdrs using arena 4.0 simulation software and found that the combined rule provided better results as compared to other rules. vinod and sridharan (2008) conducted a simulation study to evaluate five new set up oriented pdrs under different experimental conditions. the results of their study showed that the proposed rules provided better results over seven standard rules. hamidi (2016) proposed a priority dispatching rule where in the author took average of the pt and the dd of the jobs and termed this rule as pdt and suggested that the job with the least pdt value should be processed first. it is found from the literature that the pdt rule gives optimal results for multiple performance measures as compared to fcfs, spt, edd, slack and cr (hamidi, 2016). it is observed from literature that researchers have made attempt to combine two or more standard pdrs to develop new pdrs and based on the simulation studies it has been established that the combined rules perform well as compared to standard pdrs. further, the presented literature demonstrated that pdrs based on combination of pt and dd have been developed by using additive strategy i.e. pdt rule. however, there is lack of studies or no study which has combined the pt and dd using geometric and harmonic means or using similar methods. with this motivation, this work attempts to combine the pt and dd using multiplicative and reciprocal strategy and proposes two new pdrs i.e. gmpd and hmpd. to the best of authors knowledge, these rules have not been proposed yet in literature. further, researchers have used single machine scheduling problem (smsp) and simulation based study to compare the effectiveness of the newly developed pdrs over standard rules (cheng and kahlbacher, 1993; tyagi et al., 2016). therefore, a simulation study is also conducted in this work to evaluate the effectiveness of the proposed rules over fcfs, spt, edd, slack, cr and pdt on the basis of five performance measures viz. total flow time (tft), total lateness (tl), total number of late parts (tnl), total earliness (te) and total number of early parts (tne). hamidi (2016) also proposed weighted pt and dd total (wpdt) rule in which weights are assigned to pt and dd to obtain better performance measures. keeping this in view, the weighted geometric mean of pt and dd (wgmpd) and weighted harmonic mean of pt and dd (whmpd) are also proposed in this paper. it is found from the simulation study that the proposed rules give better results as compared to the standard rules as well as pdt rule. rest of the paper is organized as follows: section 2 describes how the proposed gmpd and hmpd rules are derived. section 3 presents a case study in which a sample problem using the standard rules as well as the proposed rules is solved. section 4 puts forth the simulation study in which eleven thousand randomly generated problems are solved and the results so obtained are compared for different pdrs. section 5 illustrates the effect of weight variation in the proposed wgmpd and whmpd on tft and tl. finally, section 6 presents the conclusion of the present study. to provide a better understanding of the various symbols used in this manuscript, a nomenclature is provided in table 1. 2. proposed rules a single machine scheduling problem (smsp) requires effective scheduling of ‘n’ jobs say j1, j2, j3,.........., jn on a single machine. several pdrs can be used to schedule these jobs. among int. j. prod. manag. eng. (2021) 9(2), 93-102 creative commons attribution-noncommercial-noderivatives 4.0 international ahmad et al. 94 http://creativecommons.org/licenses/by-nc-nd/4.0/ them, the most preferable are spt and edd as they possess advantages over others in terms of specific performance measures. spt rule prioritizes the jobs based on pt which results in small flow times. whereas, in edd rule, jobs with least dd are processed first which results in reduced lateness. as spt rule is purely based on the pt and edd rule is solely based on the dd of the job, priorities defined on the combined basis of pt and dd might be more effective for optimal results of multiple performance measures. in this regard, hamidi (2016) made an attempt and proposed the pdt rule, where additive strategy was used to combine the pt and dd. since, multiplicative and reciprocal strategy has not been examined to combine the different pdrs, this work proposes two new pdrs to solve a smsp based on the multiplicative and reciprocal strategies. in the first rule, geometric mean of the pt and ddof the jobs is taken, and it is termed as gmpd which is based on the multiplicative strategy. according to the gmpd, top priority is given to a job for which the value of gmpd is the minimum. the second rule considers harmonic mean of the pt and dd of the jobs and it is referred to as hmpd that is based on the reciprocal strategy. a job with the least value of the hmpd is processed first according to this rule. 2.1. the geometric mean of processing time and due date (gmpd) for two numbers say ‘a’ and ‘b’ the mathematical formula used to calculate geometric mean is √(ab). considering an smsp of ‘n’ jobs j1, j2, j3,.........., jn with deterministic processing time pti and due date ddi of job ji, the priority function for gmpd rule is defined according to equation (1). gm i pt i dd i (1) hence, for each job, the value of priority function is calculated and the job with the least gmi value will be processed first. 2.2. the harmonic mean of processing time and due date (hmpd) the harmonic mean is related to the arithmetic mean in a manner that it is reciprocal of the arithmetic mean. pdr based on the arithmetic mean of pt and dd i.e. pdt has already been developed and established to provide better performance than other standard pdrs (hamidi, 2016). hence, it was supposed that pdr based on harmonic mean might be able to give better results. in this regard, the harmonic mean of pt and dd (hmpd) rule is proposed in this work. the priority function for hmpd rule is used as defined by equation (2). hm i 2 pt i dd i pt i dd i (2) hence, priorities to jobs are defined based on hmi value. a job with a minimum value of hmi will be most preferable for processing over other jobs. 3. case study in this section, a sample problem is solved using the proposed pdrs along with other rules to examine table 1. nomenclature. symbol meaning i job number n total number of jobs ji name of the job i pti processing time of job ji ddi due date of job ji fi flow time of job ji li lateness of job ji ei earliness of job ji pdr priority dispatching rule gm geometric mean hm harmonic mean pt processing time dd due date gmpd geometric mean of processing time and due date hmpd harmonic mean of processing time and due date tft total flow time tl total lateness tnl number of late jobs te total earliness tne number of early parts wgmpd weighted geometric mean whmpd weighted harmonic mean fcfs first come first served spt shortest processing time edd earliest due date cr critical ratio mwk most work remaining int. j. prod. manag. eng. (2021) 9(2), 93-102creative commons attribution-noncommercial-noderivatives 4.0 international geometric and harmonic means based priority dispatching rules for single machine scheduling problems 95 http://creativecommons.org/licenses/by-nc-nd/4.0/ the effectiveness of the proposed pdrs for solving smsp. the sample problem consists of ten jobs that need to be sequenced for processing on a machine. each job has a deterministic pt in the interval of 1 to 15 days which is determined using uniform distribution. the dd of each job is also considered to be deterministic;in the interval ranging from 3×pt of the job and 60 days, and it is determined using the same distribution (uniform distribution). the pt and the dd of each job for the sample problem considered in this work are shown in table 2. table 2. pt and dd of jobs. job pt (days) dd (days) j1 9 34 j2 5 53 j3 15 50 j4 1 28 j5 8 56 j6 14 45 j7 3 15 j8 10 30 j9 9 32 j10 4 36 the effectiveness of the pdrs is examined on the basis of commonly used performance measures i.e. flow time, lateness, number of late parts, earliness and number of early parts (oyetunji, 2009). it may be noted that either maximum or average or total values of these performance measures may be used. in the present work, total values of the performance measures are considered for the evaluation of the pdrs. the performance measures considered in this work are described as follows: total flow time (tft): it represents the sum of time each job spends in the system and it is determined using equation (3). tft n i 1 f i (3) where, fi=fi-1+pti. pti and fi indicate the pt and flow time of job ji respectively. total lateness (tl): it measures the total amount of lateness to the jobs and it is computed using equation (4). tl n i 1 l i (4) where, li=max(0,fi – ddi). ddi and li represents the dd and lateness of job ji respectively. total number of late parts (tnl): if a job is completed after its dd, it is said to be late. tnl is the measure that indicates the count of the late parts i.e. total earliness (te): the difference between the dd and flow time when a job is completed before its dd is regarded as the earliness of a job otherwise earliness of a job is defined as 0. total earliness is defined by equation (5). te n i 1 e i (5) where, ei=max(0,ddi–fi). ddi and ei represent the dd and earliness of job ji respectively. total number of early parts (tne): the count of the jobs which are early is regarded as tne i.e., where n is the number of jobs. the sample problem shown in table 1 is solved using different pdrs viz. fcfs, spt, edd, slack, cr, pdt, gmpd and hmpd. the sequence of jobs for processing on a machine so obtained is depicted in table 3. table 3. sequence of jobs using different pdrs. fcfs spt edd slack cr pdt gmpd hmpd j1 j4 j7 j7 j8 j7 j4 j4 j2 j7 j4 j8 j7 j4 j7 j7 j3 j10 j8 j9 j9 j8 j10 j10 j4 j2 j9 j1 j1 j10 j2 j2 j5 j5 j1 j4 j4 j9 j9 j5 j6 j1 j10 j6 j6 j1 j8 j9 j7 j9 j6 j10 j10 j2 j1 j1 j8 j8 j3 j3 j3 j6 j5 j8 j9 j6 j2 j2 j2 j5 j6 j6 j10 j3 j5 j5 j5 j3 j3 j3 it may be noted that for pdt rule, the jobs are sequenced based on the minimum value of the sum of pt and dd. for gmpd rule job sequence is defined based on the minimum value of priority function defined in equation (1) and for hmpd rule equation (2) is used to define the job sequence. further, the values of performance measures are computed when the jobs are sequenced with these pdrs. the performance so obtained is shown in table 4. it can be seen from table 4 that (i) the least tft (306 days) is obtained when the jobs are sequenced using spt rule or hmpd rule, (ii) the minimum value of tl (47 days) is obtained when the jobs int. j. prod. manag. eng. (2021) 9(2), 93-102 creative commons attribution-noncommercial-noderivatives 4.0 international ahmad et al. 96 http://creativecommons.org/licenses/by-nc-nd/4.0/ are sequenced using pdt rule, (iii) the number of late jobs is the least i.e. 4 when either of the spt, edd, pdt, gmpd or hmpd rule is used, (iv) maximum te (145 days) is observed when spt rule is used,and (v) the number of jobs which are completed before the dd i.e. tne is maximum when either of the spt, edd, pdt, gmpd or hmpd is utilized. these results are in line with the results reported in the literature that no single rule is the best for all the performance measures (dominic et al., 2004). however, an attempt has been made to identify a pdr which might be effective in achieving optimal values of all performance measures. for this purpose, the best performance value for a measure is identified among the considered pdrs. based on the best performance value, the percentage deviation in the performance of each pdr is calculated. it is to be mentioned that for tft, tl and tnl, maximum values whereas for te and tne minimum values is regarded as the best performance value. in this way, a pdr with the best value for a specific performance measure will get a percentage deviation of 0% and a pdr with least average percentage over all the performance measure will be the best pdr. the percentage deviation of each performance measure for each pdr are depicted in table 5. table 5. percentage deviation of the performance measures. pdr tft (%) tl (%) tnl (%) te (%) tne (%) average % deviation fcfs 45.10 257.45 50.00 28.97 33.33 82.97 spt 0.00 53.19 0.00 0.00 0.00 10.64 edd 22.55 25.53 0.00 56.55 16.67 24.26 slack 33.99 55.32 50.00 71.03 33.33 48.73 cr 36.27 55.32 50.00 75.86 33.33 50.16 pdt 10.78 0.00 0.00 40.00 0.00 10.16 gmpd 1.63 17.02 0.00 15.17 0.00 6.77 hmpd 0.00 48.94 0.00 1.38 0.00 10.06 it is observed from table 5 that gmpd, hmpd and pdt are the top three rules with least deviations. hence, it can be inferred that the mean based pdrs outperform the standard pdrs. further, among the considered pdrs, the deviation from the best is the least for gmpd rule. hence, it can be concluded that the best pdr for deriving optimal values of the five performance measures i.e. tft, tl, tnl, te, and tne is gmpd. the next in the row is found to be hmpd. subsequently, the next best rule is observed to be the pdt rule. pdt rule has also been reported to perform better than fcfs, edd, slack, and cr for obtaining optimal values of the multiple performance measures (hamidi, 2016). based on the values of average percentage deviation (table 5) the importance of rest of the pdrs in decreasing order is edd>slack>cr>fcfs. 4. simulation study it is likely that results obtained for the sample problem solved in the previous section might change when the parameters of the sample problem i.e. pt and dd is changed. therefore, it is indeed vital to examine the changes that may occur in the results when these parameters are altered. for this purpose, a simulation study was conducted in which eleven thousand problems were generated randomly and in each problem ten jobs were considered. both pt (ranging from 1 to 10 days) and dd (in the interval of 3 x pt to 60 days) of each job was deterministically generated using uniform distribution. the performance measures were computed for each problem and the average results obtained for the eleven thousand problems were used to compare different pdrs. the average of tft taken over eleven thousand solved problems using different pdrs is shown in figure 1. 440,77 328,67 389,87 423,93 441,08 362,59 333,23 328,82 0 50 100 150 200 250 300 350 400 450 500 fc fs sp t ed d sl ac k cr pd t gm pd hm pd avg_tft figure 1. average tft of eleven thousand problems. table 4. performance measures for different pdrs. pdr tft (days) tl (days) tnl te (days) tne fcfs 444 168 6 103 4 spt 306 72 4 145 6 edd 375 59 4 63 5 slack 410 73 6 42 4 cr 417 73 6 35 4 pdt 339 47 4 87 6 gmpd 311 55 4 123 6 hmpd 306 70 4 143 6 int. j. prod. manag. eng. (2021) 9(2), 93-102creative commons attribution-noncommercial-noderivatives 4.0 international geometric and harmonic means based priority dispatching rules for single machine scheduling problems 97 http://creativecommons.org/licenses/by-nc-nd/4.0/ it is evident from figure 1 that the least value of tft (328.67 days) is obtained when job sequencing is done using spt rule and this result is in line with the result reported in literature (lu et  al., 2012; tyagi et al., 2016). further, mean based pdrs i.e. gmpd, hmpd and pdt perform better than other pdrs except spt. among the three mean based pdrs, hmpd is found to give better results for tft followed by gmpd and pdt. 123,76 60,50 57,81 74,79 71,05 51,25 54,00 58,01 0 20 40 60 80 100 120 140 fc fs sp t ed d sl ac k cr pd t gm pd hm pd avg_tl figure 2. average tl of eleven thousand problems. figure 2 depicts average value of tl for the eleven thousand solved problems using different pdrs and it reveals that pdt rule performs better than other pdrs as the average tl value for this rule is minimum i.e. 51.25 days. the proposed gmpd rule is next in the row as it results in average total lateness of 54 days. performance of rest of the pdrs in descending order as observed from figure 2 is edd>hmpd>spt>cr>slack>fcfs. 5,05 3,18 3,59 4,34 6,04 3,07 3,02 3,14 0 1 2 3 4 5 6 7 fcfs spt edd slack cr pdt gmpd hmpd avg_no_late figure 3. average tnl of eleven thousand problems. results for the number of late parts for different pdrs are depicted in figure 3 which clearly shows that performance of the gmpd is the best as it provides minimum number of late parts i.e. 3.02 and this rule is followed by hmpd, pdt, spt, edd, slack, fcfs, and cr. 103,31 152,15 88,26 71,17 50,28 108,97 141,08 149,50 0 20 40 60 80 100 120 140 160 fc fs sp t ed d sl ac k cr pd t gm pd hm pd avg_te figure 4. average te of eleven thousand problems. figure 4 depicts results of the simulation study for te for different pdrs which clearly show that maximum te is observed when jobs are sequenced using spt rule. next to spt rule is the hmpd and then the other pdrs follow. the decreasing order of the pdrs is spt>hmpd>gmpd>pdt>fcfs> edd>slack>cr. 4,83 6,69 6,26 5,51 3,64 6,79 6,87 6,74 0 1 2 3 4 5 6 7 8 fcfs spt edd slack cr pdt gmpd hmpd avg_no_early figure 5. average tne of eleven thousand problems. the performance measure tne is used to examine the count over all the parts which are completed on or before the dd. figure 5 depicts the average value of the tne for different pdrs obtained from simulation study. it is evident from figure 5 that gmpd rule is better than other pdrs as the average value of tne for this rule is maximum i.e. 6.87. the other two mean based pdrs i.e. pdt and hmpd come next to gmpd. the sequence of pdrs in decreasing order is gmpd>pdt>hmpd>spt> edd>slack>fcfs>cr. int. j. prod. manag. eng. (2021) 9(2), 93-102 creative commons attribution-noncommercial-noderivatives 4.0 international ahmad et al. 98 http://creativecommons.org/licenses/by-nc-nd/4.0/ the results obtained from the simulation study show that no single rule gives the best results for all the performance measures. however, it is observed that the results derived from using mean based pdrs are best for a few performance measures and for others although the results are not the best but they are close to the best one. thus, based on the results of the simulation study it is suggested that mean based pdrs should be used to obtain optimal results for all the performance measures. to investigate the effect of pdrs on all performance measures taken together, all pdrs are ranked separately based on the results of the simulation study. the ranking results for different pdrs with respect to the performance measures are shown in table 6. table 6. ranking results for pdrs. pdr rank based on performance measures average ranktft tl tnl te tne fcfs 7 8 7 5 7 6.8 spt 1 5 4 1 4 3 edd 5 3 5 6 5 4.8 slack 6 7 6 7 6 6.4 cr 8 6 8 8 8 7.6 pdt 4 1 2 4 2 2.6 gmpd 3 2 1 3 1 2 hmpd 2 4 3 2 3 2.8 it is evident from table 6 that average rank of the mean based pdrs i.e. gmpd, pdt, and hmpd is higher than other pdrs. further, among the three mean based pdrs, the proposed gmpd rule is found to be the best (as its average rank is 2) followed by pdt and hmpd for achieving optimal results for multiple performance measures. thus, the proposed gmpd rule is the most promising rule as compared to other pdrs and it should be implemented to obtain compromised or optimal results for multiple performance measures of single machine scheduling problems. next to gmpd is the pdt and then hmpd and therefore, it is found that mean based pdrs perform better as far as scheduling of jobs on a single machine is concerned. 5. weighted gmpd and hmpd rule it has been shown in the previous section that the pdrs based on three different strategies i.e. additive, multiplicative and reciprocal provides better results, in terms of the optimal performance measures of the system, as compared to other pdrs. it is realized that in these strategies, the weight component of both numbers is same. it is very likely that the weighted mean based pdrs where the weight components of the pt and dd of jobs are different may provide a better sequence of jobs and therefore, it is matter of investigation. consequently, the performance of weighted form of pdt, gmpd, and hmpd is compared in this section. the weighted pt and dd (wpdt) rule has already been developed and reported in the literature (hamidi, 2016). hence, the weighted geometric mean of pt and dd (wgmpd) and weighted harmonic mean of pt and dd (whmpd) rules are proposed in this work. the function used to determine the priority of the jobs using wpdt, wgmpd and whmpd is given in equation (6), equation (7), and equation (8) respectively. pdi=w · pti + (1–w)ddi (6) wgmi=ptiw · ddi1-w (7) whm i 1 w pt i 1 w dd i (8) where, w represents weight a job with the least value of wgmi and whmi is processed first when the priority is determined using wgmpd and whmpd respectively. the motivation behind the development of weighted pdrs is to combine spt and edd rules to obtain optimal multiple performance measures. spt and edd rules have been considered as they possess advantage over other pdrs in terms of small flow times and reduced lateness respectively. therefore, in this section, the weight used in wpdt, wgmpd, and whmps is varied from 0 to 1 with an increment of 0.1 and the values of two performance measures i.e. tft and tl are compared. as the unit of measurement tft and tl is different, it is necessary to normalize their values so as to bring their values on a common scale. among the several methods of normalization, the min-max normalization method is used. the mathematical formula used in min-max normalization method is given in equation (9). x i n x i min i x i max i x i min i x i (9) where, xi and xin represent the original and normalized value of the ith attribute. the normalized value of tft and tl with varying weights for wpdt rule is shown in figure 6. it int. j. prod. manag. eng. (2021) 9(2), 93-102creative commons attribution-noncommercial-noderivatives 4.0 international geometric and harmonic means based priority dispatching rules for single machine scheduling problems 99 http://creativecommons.org/licenses/by-nc-nd/4.0/ may be noted that when the value of weight is 0, it represents edd rule as the job sequencing is done based on dd only. when the weight is set at 1, job sequencing is done based on spt rule and when the weight is 0.5 it represents pdt rule. it is evident from figure 6 that for edd rule (w = 0), the values of both performance measures i.e. tft and tl are high which is not desirable. when w = 1 i.e.for spt rule, value of tft is minimum (almost zero) but the value of tl issignificantly high. further, for pdt rule i.e. when w = 0.5, the value of tl is minimum, and also the value of tft is smaller than that of edd but higher than spt. as the wpdt rule was developed with an aim to obtain optimal values of tft and tl, the graph shown in figure 6 supports the fact that wpdt rule can give optimal results for both the performance measures i.e. tft and tl. it can also be observed from figure 6 that better results can be obtained if a weight of 0.76 is considered in the wpdt as the two performance measures intersect at this point which suggests that their values are optimal. the normalized value of tft and tl when weight is 0.76 is found to be 0.20. 0,0 0,2 0,4 0,6 0,8 1,0 1,2 ed d 0.1 0.2 0.3 0.4 pd t 0.6 0.7 0.8 0.9 sp t n or m al iz ed v al ue s tft tl figure 6. variation in normalized values of tft and tl with weight for wpdt rule. the variation of normalized values of tft and tl for wgmpd with varying weights is shown in figure 7. the pattern observed in this case is similar to that of the wpdt rule as shown in figure 6. it can be seen from figure 7 that for w = 0, tft and tl values are high, but for spt (w = 1), tft value is minimum but tl value is reasonably high. for gmpd rule (w = 0.5), the values of tft and tl are in between that of spt and edd. further, intersection of tft and tl is obtained when the weight of 0.37 is assigned. hence, it can be concluded that for gmpd rule, the weight of 0.37 will result in optimal values of tft and tl which is 0.16. 0,0 0,2 0,4 0,6 0,8 1,0 1,2 ed d 0.1 0.2 0.3 0.4 gm pd 0.6 0.7 0.8 0.9 sp t n or m al iz ed v al ue s tft tl figure 7. variation in normalized values of tft and tl with weight for wgmpd rule. figure 8 shows variation in the values of tft and tl with varying weights of whpd rule. once again, a similar pattern of variation as that of wpdt and wgmpd is observed in this case too. it is evident from figure 8 that a weight of 0.105 results in the optimum values i.e. 0.135 of both tft and tl as the two curves intersects at this point. 0,00 0,20 0,40 0,60 0,80 1,00 1,20 ed d 0.1 0.2 0.3 0.4 hm pd 0.6 0.7 0.8 0.9 sp t n or m al iz ed v al ue s tft tl figure 8. variation in normalized values of tft and tl with weight for whmpd rule. from the analysis of the weight variation in the wpdt, wgmpd, and whpd presented above, it is found that a weight of 0.76 used in the wpdt results in the optimum values of both tft and tl which is 0.20. further, a weight of 0.37 employed in the wgmpd rule and a weight of 0.105 used in the whmpd rule leads to the optimum values of tft and tl which are 0.16 and 0.135 respectively. further, it is also found that among the three weighted mean based rules, whmpd with a weight of 0.105 produces results better than wgmpd with the weight of 0.37 and wpdt with weight 0.76 as optimum values of tft and tl are minimum i.e. 0.135. int. j. prod. manag. eng. (2021) 9(2), 93-102 creative commons attribution-noncommercial-noderivatives 4.0 international ahmad et al. 100 http://creativecommons.org/licenses/by-nc-nd/4.0/ 6. conclusion scheduling is imperative in manufacturing systems as it directly affects the systems’ performance. for a single machine scheduling problem, priority dispatching rules are used to define the sequence of jobs to be processed as they help in enhancing the performance of the system. among the standard pdrs i.e. fcfs, spt, edd, slack and cr, spt performs better as far as minimum flow time is required and edd rule is found to be promising for minimizing the lateness of the jobs. further, it was realized that combining these rules might be able to perform better and hence, several combined rules were developed and proposed. however, it is found that most of these rules are based on the additive strategy. therefore, an attempt was made in this study to examine the effectiveness of the combined rules based on the multiplicative and reciprocal strategy. in this regard two pdrs, first considering multiplicative strategy and second with the reciprocal strategy have been proposed in this work. the first rule named as geometric mean of pt and dd (gmpd) is based on multiplicative strategy and the other i.e. harmonic mean of pt and dd (hmpd) is based on reciprocal strategy. five performance measures viz. total flow time (tft), total lateness (tl), number of late parts (tnl), total earliness (te) and number of early parts (tne) were used to examine the effectiveness of the proposed rules. this study demonstrates the application of heuristic and metaheuristic algorithms to deal with best pdr. the major conclusions drawn from the present work are as follows: spt rule is found to the best rule for minimizing the tft of the jobs. pdt rule performs better than other pdrs as far as minimum tl is required. the proposed rule gmpd results in the least number of late parts compared to other pdrs. for maximizing the total earliness, the spt rule is found to give better results. the maximum number of early parts is observed when the gmpd rule is used. for optimal performance,it is found that pdrs based on different combination strategies, i.e. additive, multiplicative or reciprocal perform better than other pdrs. further, among the three strategy based pdrs, gmpd rule performs better than others followed by pdt and hmpd. in weighted pdrs, optimal values of multiple performance measures are found when a weight of 0.105 is used in whmpd followed by wgmpd with a weight of 0.37 and wpdt with a weight of 0.76. the findings of this study suggest that gmpd rule is the best rule among the considered pdrs. however, there are shortcomings of this work which could be considered in future work. the comparison of the new rules has been done with six rules. however, there are various other rules which can also be compared to find the effectiveness of the proposed rules. further, in this work the multiplicative strategy and reciprocal strategy were used to combine pt and dd which can also be used to combine process time and work content as done by holthaus and rajendran (1997). the results of the study are limited for a single machine scheduling problems. hence, it can be extended further for multiple machines and flexible manufacturing systems. references baharom, m.z., nazdah, w., hussin, w. (2015). scheduling analysis for job sequencing in veneer lamination line. journal of industrial and intelligent information, 3(3), 181-185. https://doi.org/10.12720/jiii.3.3.181-185 chan, f.t.s., chan, h.k., lau, h.c.w., ip, r.w.l. (2003). analysis of dynamic dispatching rules for a flexible manufacturing system. journal of materials processing technology, 138(1), 325-331. https://doi.org/10.1016/s0924-0136(03)00093-1 cheng, t.c.e., kahlbacher, h.g. (1993). single-machine scheduling to minimize earliness and number of tardy jobs. journal of optimization theory and applications, 77(3), 563-573. https://doi.org/10.1007/bf00940450 da silva, n.c.o., scarpin, c.t., pécora, j.e., ruiz, a. (2019). online single machine scheduling with setup times depending on the jobs sequence. computers & industrial engineering, 129, 251-258. https://doi.org/10.1016/j.cie.2019.01.038 doh, h.h., yu, j.m., kim, j.s., lee, d.h., nam, s.h. (2013). a priority scheduling approach for flexible job shops with multiple process plans. international journal of production research, 51(12), 3748-3764. https://doi.org/10.1080/00207543.2013.765074 dominic, panneer d.d., kaliyamoorthy, s., kumar, m.s. (2004). efficient dispatching rules for dynamic job shop scheduling. the international journal of advanced manufacturing technology, 24(1), 70-75. https://doi.org/10.1007/s00170-002-1534-5 int. j. prod. manag. eng. (2021) 9(2), 93-102creative commons attribution-noncommercial-noderivatives 4.0 international geometric and harmonic means based priority dispatching rules for single machine scheduling problems 101 https://doi.org/10.12720/jiii.3.3.181-185 https://doi.org/10.1016/s0924-0136(03)00093-1 https://doi.org/10.1007/bf00940450 https://doi.org/10.1016/j.cie.2019.01.038 https://doi.org/10.1080/00207543.2013.765074 https://doi.org/10.1007/s00170-002-1534-5 http://creativecommons.org/licenses/by-nc-nd/4.0/ ðurasević, m., jakobović, d. (2018). a survey of dispatching rules for the dynamic unrelated machines environment. expert systems with applications, 113, 555-569. https://doi.org/10.1016/j.eswa.2018.06.053 forrester, p. (2006). operations management: an integrated approach. international journal of operations & production management, 26(4), 449-450. https://doi.org/10.1108/01443570610650585 geiger, c.d., uzsoy, r. (2008). learning effective dispatching rules for batch processor scheduling. international journal of production research, 46(6), 1431-1454. https://doi.org/10.1080/00207540600993360 hamidi, m. (2016). two new sequencing rules for the non-preemptive single machine scheduling problem. the journal of business inquiry, 15(2), 116-127. holthaus, o., rajendran, c. (1997). new dispatching rules for scheduling in a job shop —an experimental study. the international journal of advanced manufacturing technology, 13(2), 148-153. https://doi.org/10.1007/bf01225761 hussain, m.s., ali, m. (2019). a multi-agent based dynamic scheduling of flexible manufacturing systems. global journal of flexible systems management, 20(3), 267-290. https://doi.org/10.1007/s40171-019-00214-9 jayamohan, m.s., rajendran, c. (2000). new dispatching rules for shop scheduling: a step forward. international journal of production research, 38(3), 563-586. https://doi.org/10.1080/002075400189301 kadipasaoglu, s.n., xiang, w., khumawala, b.m. (1997). a comparison of sequencing rules in static and dynamic hybrid flow systems. international journal of production research, 35(5), 1359-1384. https://doi.org/10.1080/002075497195371 kanet, j.j., li, x. (2004). a weighted modified due date rule for sequencing to minimize weighted tardiness. journal of scheduling, 7(4), 261-276. https://doi.org/10.1023/b:josh.0000031421.64487.95 lee, d.k., shin, j.h., lee, d.h. (2020). operations scheduling for an advanced flexible manufacturing system with multi-fixturing pallets. computers & industrial engineering, 144, 106496. https://doi.org/10.1016/j.cie.2020.106496 lu, c.c., lin, s.w., ying, k.c. (2012). robust scheduling on a single machine to minimize total flow time. computers & operations research, 39(7), 1682-1691. https://doi.org/10.1016/j.cor.2011.10.003 krishnan, m., chinnusamy, t.r., karthikeyan, t. (2012). performance study of flexible manufacturing system scheduling using dispatching rules in dynamic environment. procedia engineering, 38, 2793-2798. https://doi.org/10.1016/j.proeng.2012.06.327 munir, e.u., li, j., shi, s., zou, z., yang, d. (2008). maxstd: a task scheduling heuristic for heterogeneous computing environment. information technology journal, 7(4), 679-683. https://doi.org/10.3923/itj.2008.679.683 oyetunji, e.o. (2009). some common performance measures in scheduling problems. research journal of applied sciences, engineering and technology, 1(2), 6-9. pinedo, m.l. (2009). planning and scheduling in manufacturing and services (2nd ed.). springer-verlag. https://doi.org/10.1007/978-14419-0910-7 prakash, a., chan, f.t.s., deshmukh, s.g. (2011). fms scheduling with knowledge based genetic algorithm approach. expert systems with applications, 38(4), 3161-3171. https://doi.org/10.1016/j.eswa.2010.09.002 rafsanjani, m.k., bardsiri, a.k. (2012). a new heuristic approach for scheduling independent tasks on heterogeneous computing systems. international journal of machine learning and computing, 371-376. https://doi.org/10.7763/ijmlc.2012.v2.147 tyagi, n., tripathi, r.p., chandramouli, a.b. (2016). single machine scheduling model with total tardiness problem. indian journal of science and technology, 9(37), 1-14. https://doi.org/10.17485/ijst/2016/v9i37/97527 vinod, v., sridharan, r. (2008). dynamic job-shop scheduling with sequence-dependent setup times: simulation modeling and analysis. the international journal of advanced manufacturing technology, 36(3), 355-372. https://doi.org/10.1007/s00170-006-0836-4 waikar, a.m., sarker, b.r., lal, a.m. (1995). a comparative study of some priority dispatching rules under different shop loads. production planning & control, 6(4), 301-310. https://doi.org/10.1080/09537289508930284 int. j. prod. manag. eng. (2021) 9(2), 93-102 creative commons attribution-noncommercial-noderivatives 4.0 international ahmad et al. 102 https://doi.org/10.1016/j.eswa.2018.06.053 https://doi.org/10.1108/01443570610650585 https://doi.org/10.1080/00207540600993360 https://doi.org/10.1007/bf01225761 https://doi.org/10.1007/s40171-019-00214-9 https://doi.org/10.1080/002075400189301 https://doi.org/10.1080/002075497195371 https://doi.org/10.1023/b:josh.0000031421.64487.95 https://doi.org/10.1016/j.cie.2020.106496 https://doi.org/10.1016/j.cor.2011.10.003 https://doi.org/10.1016/j.proeng.2012.06.327 https://doi.org/10.3923/itj.2008.679.683 https://doi.org/10.1007/978-1-4419-0910-7 https://doi.org/10.1007/978-1-4419-0910-7 https://doi.org/10.1016/j.eswa.2010.09.002 https://doi.org/10.7763/ijmlc.2012.v2.147 https://doi.org/10.17485/ijst/2016/v9i37/97527 https://doi.org/10.1007/s00170-006-0836-4 https://doi.org/10.1080/09537289508930284 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.1862 received 2013-11-11 accepted 2013-11-25 h2020 opportunities for research and innovation in production management and engineering. poler, r. director of the research centre on production management and engineering centro de investigación de gestión e ingeniería de producción (cigip) universitat politècnica de valència camino de vera s/n, ed. 8b, acc. l. ciudad politécnica de la innovación 46022 valencia – spain rpoler@cigip.upv.es horizon 2020 is the new financial instrument implementing the european union, a global initiative aimed at securing europe’s competitiveness and creating new growth and jobs in europe. this new eu programme for research and innovation will run from 2014 to 2020 with a budget over €70 billion. what’s new compared with previous framework programs? horizon 2020 major goal is helping to bridge the gap between research and the market. research remains essential but innovation plays now an important role. this market-driven approach include creating public-private partnerships to bring together the resources needed. on the other hand, a two year perspective will be provided in order to allow researchers more time to prepare proposals and more scope to make innovative proposals. the horizon 2020 calls are structured in three priorities: excellent science, industrial leadership and societal challenges which cluster different thematic areas with funding allocation (figure 1). the proposed support for research and innovation under horizon 2020 aims i) to strengthen the eu’s position in science (priority 1) by boosting to toplevel research in europe, ii) to strengthen industrial leadership in innovation (priority 2) with major investment in key technologies, greater access to capital and support for smes and iii) to help address major societal challenges (priority 3) shared by all europeans such as climate change, developing sustainable transport and mobility, making renewable energy more affordable, ensuring food safety and security, or coping with the challenge of an ageing population. horizon 2020 aims to raise the level of excellence in europe’s science base and ensure a steady stream of world-class research to secure europe’s long-term competitiveness, supporting the best ideas, developing talent within europe, providing researchers with access to priority research infrastructure and making europe an attractive location for the world’s best researchers. the competitive industries objective aims at making europe a more attractive location to invest in research and innovation. it will provide major investment in key industrial technologies, maximise the growth potential of european companies helping innovative smes to grow into world-leading companies. leadership in enabling and industrial technologies will support the development of technologies underpinning innovation across a range of sectors with a strong focus on developing european industrial capabilities in key enabling technologies (kets). what are the opportunities in horizon 2020 for the production management and engineering (pme) domain? priority 2 “leadership in enabling and industrial technologies” (leit) contains six domains: i) information and communication technologies, ii) nanotechnologies, iii) advanced materials, iv) biotechnology, v) advanced manufacturing and processing and vi) space. the research and innovation in pme is mainly related with advanced manufacturing (management and engineering) but also with information and communication technologies (as support of management). 3int. j. prod. manag. eng. (2014) 2(1), 3-6creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2014.1862 the most interesting challenges in the leit-nmp work programme are “factories of the future” with topics as: global energy and other resources efficiency in manufacturing enterprises: development of new value-chain approaches oriented to reducing energy and resource consumption along the whole value chain, including final users and recycling or reprocessing companies through the implementation of common resources optimization. developing smart factories that are attractive to workers: creation of new methods and technologies for an optimised use of workers’ knowledge to enhance work related satisfaction, taking into consideration safety and ergonomics of the working areas. innovative product-service design using manufacturing intelligence: development of tools and methodologies to effectively involve customers and suppliers across the value chain through the collaborative management of engineering knowledge and its exchange between product design, service design and manufacturing. manufacturing of custom made parts for personalised products: development of advanced design and manufacturing technologies able to transform product-service data descriptions and protocols into manufacturing operations and processes, seamless data integration across the process and supply chains for the fast production and distribution of personalised products. sustainable product life cycle management focused on reuse, remanufacturing and recycling related to advanced materials: development of new methods and technologies for the replacement or reduction of critical raw materials through new equipment concepts, design and components for remanufacturing and recycling. integrated design and management of production machinery and processes: development of accurate simulation models and algorithms for model-based control of production machinery and demonstration of the reliability of modelbased machines with respect to production accuracy, quality, maintainability and lifecycle return-on-investment. regarding leit-ict, also in “factories of the future”, three interesting topics appear: process optimisation of manufacturing assets: cyber-physical systems based processes optimisation for adaptive and smart manufacturing systems with advanced control and new modelling and simulation technologies. development of agile cloud-enabled collaboration tools for priority 1: excellent science european research council 13,0k m€ future and emerging technologies 2,6k m€ marie skłodowska-curie actions 6,1k m€ research infrastructures 2,4k m€ priority 2: industrial leadership leadership in enabling and industrial technologies 13,5k m€ access to risk finance 2,8k m€ innovation in smes 0,6k m€ priority 3: societal challenges health, demographic change and wellbeing 7,4k m€ food security, sustainable agriculture, marine and maritime and inland water research & the bioeconomy 3,8k m€ smart, green and integrated transport 6,3k m€ climate action, resource efficiency and raw materials 3,0k m€ europe in a changing world-inclusive, innovative and reflective societies 1,3k m€ secure societies-protecting freedom and security of europe and its citizens 1,6k m€ secure, clean and efficient energy 5,9k m€ figure 1. h2020 priorities, thematic areas and funding (proposal). 4 int. j. prod. manag. eng. (2014) 2(1), 3-6 creative commons attribution-noncommercial 3.0 spain poler, r. process optimisation of manufacturing assets across the supply chain. ict-enabled modelling, simulation, analytics and forecasting technologies: development of modelling and simulation methods involving multiple phenomena, including multi-scale and integrated discrete/continuous models taking a holistic approach. development of integrated knowledge-based systems covering the complete product life-cycle with advanced analytics and self-learning capabilities exploiting the availability of “big data” from smart sensors, historical process files, or human-authored data. and development of integrated information management systems for product-processproduction systems embedded into their social, environmental and economic context. innovation for manufacturing smes: highperformance computing cloud-based modelling, simulation and analytics services for modelling multiple interconnected phenomena, integrating novel mobile interfaces for data management and decision support, achieving real-time response. integration of cyber-physical-system modules in manufacturing processes to increase sophistication and automation in production smes. finally, there are a couple of interesting topics at challenge “nanotechnology, advanced materials and production” (nmp): business models with new supply chains for sustainable customer-driven small series production: development of integrated business model solutions for customer-driven supply chain management for novel distributed manufacturing, sourcing and design solutions linking individual “home-based” designers and manufacturers to the supply-chain. transformation of the innovation process in industrial value chains by the introduction of open innovation networks: analysis of existing network innovation solutions and current best practise. knowledge integration of the value chain and of technology and service providers in the process of innovation. priority 3 “smart, green and integrated transport” contains “logistics” with three topics: transport advantages and implications of mutualisation of the supply chain and e-commerce: new concepts for the design, development and application of consolidation and distribution centres in cities for last-mile distribution and reverse logistics, co-operative intelligent transport systems (c-its) and cloud based services, integrated into an on-line planning platform offering new means of communication amongst vehicles, between delivery vehicles & traffic management and to end users. de-stressing the supply chain: the potential for slow steaming and synchromodality to improve business efficiency and sustainability: development of a transnational logistics information platform and a well-defined core network of hinterland connections and information systems including e-freight tools, infrastructures and smart coordination mechanisms to be able to use different transportation modes flexibly to deliver maximum value to the shipper or end customer. european logistics information sharing architecture: mobile communications for secured information exchange among users, service providers and operators through the deployment of web-based open and platform to enable information exchange across suppliers, manufacturers, logistics providers and retailers without necessitating costly interfaces. as novelty, the technologies addressed in the litnmp work programme make use of the concept of technology readiness levels (trls) as a measure to assess the maturity of developed technologies during its development at the projects: trl 1: basic principles observed. trl 2: technology concept formulated. trl 3: experimental proof of concept. trl 4: technology validated in laboratory. trl 5: technology validated in industrial environment. trl 6: technology demonstrated in industrial environment. trl 7: system prototype demonstration in operational environment. trl 8: system complete and qualified. trl 9: actual system proven in competitive manufacturing environment. europe 2020 is the eu’s growth strategy for the coming decade. european citizens want the eu to become a smart, sustainable and inclusive economy in order to achieve high levels of employment, 5int. j. prod. manag. eng. (2014) 2(1), 3-6creative commons attribution-noncommercial 3.0 spain h2020 opportunities for research and innovation in production management and engineering. productivity and social cohesion. eu has set five ambitious objectives on employment, innovation, education, social inclusion and climate/energy, to be reached by 2020. the innovation union plan contains over thirty actions points, with the aim of i) making europe into a world-class science performer, ii) remove obstacles to innovation which currently prevent ideas getting quickly to market and iii) revolutionise the way public and private sectors work together through innovation partnerships between the european institutions, national and regional authorities and business. the production management and engineering (pme) domain of research and innovation is ready to face these challenges. 6 int. j. prod. manag. eng. (2014) 2(1), 3-6 creative commons attribution-noncommercial 3.0 spain poler, r. pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.2004 in memoriam dario franco. the ijpme editorial committee and the cigip members are deeply sad by the death of our colleague and friend dario franco on november 20th, 2013 in valencia (spain). dario developed an important activity in research and teaching in the universitat politècnica de valència. those of us, who had the luck of knowing him, will also miss his extraordinary human qualities and friendship. dario, you will always be in our hearts. 1int. j. prod. manag. eng. (2014) 2(1), 1-1creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2014.2004 pme i j international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.16719 received: 2021-11-23 accepted: 2022-01-24 evolution of servitization: new business model opportunities aitor ruiz de la torre a , & david sanchez b a university of the basque country ehu/upv, spain. aitor.ruizdelatorre@ehu.eus d mondragon university, spain. dsanchez@mondragon.edu abstract: the concept of servitization has been constantly developing since its outset, but in the last decade due to the irruption of industry 4.0, the complexity of the concept and its typologies of value propositions have evolved considerably, opening up endless opportunities. in this sense, the main objective of this research is to show a summary review of the evolution of servitization since its beginnings and the new typologies that are emerging due to the digitalization that arises through industry 4.0. for this purpose, a systematic review of the leading databases in the field of services has been conducted. the results of the literature review show the potential of servitization and the need to understand each reality in order to adapt to new capabilities that help the companies who become service-oriented benefit from major advantages. ultimately, it can be concluded that, in the short term, industry 4.0 and its new business models are the key, however, servitization will continue to evolve to a point where all organizations will need to adapt to new trends. key words: servitization, service business model, industry 4.0, digitalization, service oriented. 1. introduction in recent years, all manufacturing companies have faced various challenges due to the high competitiveness of a market affected by globalization. for this reason, the need to offer greater value through services rather than the traditional unique selling points (price and quality) is undoubtedly essential to survive in the long term. the impact of service and manufacturing industries is frequently considered immeasurable. therefore, they tend to be considered separately, due to their potential influence over national economies, the classification of enterprises and employment, to name a few (bigdeli et al., 2017; bustinza et al., 2013; baines & lightfoot, 2014). competing strategically through service delivery is becoming a characteristic feature of innovative manufacturing firms (baines et al., 2009), boosting manufacturers’ competitive strategies and the process through which this is achieved is commonly known as servitization (baines & lightfoot, 2014; oliva & kallenberg, 2003; reim et al., 2015). this strategy can consolidate long-term customer loyalty (vandermerwe & rada, 1988; verstrepen et al., 1999), generate new, safe, steady sources of revenue (chesbrough & rosenbloom, 2002; lay, 2014) and establish major hurdles for competitors (kinnunen, 2018; lay, 2014). in recent years, interest in servitization has continued to grow exponentially due to the innumerable benefits it brings to its users. the key to success in today’s market has shifted towards services, away from the single production model employed by manufacturers (habegger, 2010). there is no doubt to cite this article: ruiz de la torre, a., & sanchez, d. (2022). evolution of servitization: new business model opportunities. international journal of production management and engineering, 10(1), 77-90. https://doi.org/10.4995/ijpme.2022.16719 http://polipapers.upv.es/index.php/ijpme int. j. prod. manag. eng. (2022) 10(1), 77-90creative commons attribution-noncommercial-noderivatives 4.0 international 77 https://orcid.org/0000-0002-9635-9063 https://orcid.org/0000-0002-9059-4898 http://creativecommons.org/licenses/by-nc-nd/4.0/ that the benefits of servitization are increasing, e.g., rolls-royce earns more than 50% of its income from services (smith, 2013); environmental energyefficiency arguments are also decisive, along with the huge opportunities offered by servitization, such as improved processes and training to mention two (cohen et al., 2006), continuing to raise industry’s interest in servitization. (in 2015, services’ value added accounted for 74% of gpd in high income countries (buckley & majumdar, 2018). due to this increasing interest in servitization, several key research challenges that require special attention for advanced servitization need to be faced. these include opportunities to build the right organizational capabilities and culture (benedettini & visnjic, 2011; brax, 2005; vandermerwe & rada, 1988); improve the understanding of how to integrate new business models (sandström et al., 2008; tukker, 2015), services, economic profitability (anderson et al., 2006; neely, 2008), how to provide solutions (galbraith, 2002a; windahl & lakemond, 2006) how to innovate and design successful offers (gebauer & friedli, 2005; jong & vermeulen, 2003), the necessary relationships with partners (galbraith, 2002a; sandström et al., 2008) and transformation challenges faced by manufacturers seeking to serve (oliva & kallenberg, 2003; roscitt, 1990; windahl & lakemond, 2006). manufacturing companies that adopt a serviceoriented strategy have to develop the necessary organizational structures and processes (gebauer & fleisch, 2007; mathieu, 2001; oliva & kallenberg, 2003) and possess different capacities to those of production (ceci & masini, 2011; dachs et al., 2012; datta & roy, 2011; gebauer & friedli, 2005; oliva & kallenberg, 2003). the lack of implementation of these service-related aspects shows why manufacturing companies have not been able to take advantage of the benefits that servitization strategies offer sooner. this article presents a qualitative review of the key aspects and new business models that focus on the concept of servitization to help understand the concept correctly, as well as its management and applicability. 2. research methodology the main aim is to present a summary review of how servitization is evolving in order to understand the beginnings and implementation of this concept, as well as the relevance of its implementation, figure 1 shows the methodology applied: figure 1. methodological model. in this article the main business for services databases including the articles indexed in scopus, web of science, engineering village have been analyzed, as these are the reference data-bases for the topic in question, allowing sufficient critical analysis of collected data to be extracted and, subsequently, certain conclusions and future research opportunities. 3. theoretical background analyzing industry in general, historically it is possible to define 4 different industrial revolutions where the degree of complexity is seen to increase (bartodziej, 2017; deloitte, 2015; vuksanović et al., 2016) over the years up to today. finally, the latest industrial revolution currently happening in all industrial businesses refers to a revolution based on a cyber-physical production system, better known as “industry 4.0”, one of the most popular topics drawing attention from both professional and academic fields (liao et al., 2017; nicolae et al., 2019). cyber-physical systems (cps) are defined as technologies to manage interconnected systems between physical assets and digital systems (lee et al., 2015; leitão et al., 2016; luthra et al., 2020), being a fundamental basis of industry 4.0 (kim, 2017; varghese & tandur, 2014; xu et al., 2018). by integrating cps in different company departments (production, logistics, services, etc.) in today’s industrial companies, the aim is to transform the current factory into a smarter factory generating int. j. prod. manag. eng. (2022) 10(1), 77-90 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz de la torre & sanchez 78 http://creativecommons.org/licenses/by-nc-nd/4.0/ significant economic potential (luthra et al., 2020; negri et al., 2017). this is achieved through easy information exchange and integrated control of products and manufacturing machines acting simultaneously and intelligently in interoperability (lu, 2017; ślusarczyk, 2018). industry 4.0 is an ongoing revolution and therefore a lot of thinking is necessary to strengthen competitiveness in a more complex environment unknown until recently, where players have to adapt to this type of industry and move away from the classic manufacturing value chain. there are many areas where companies can benefit enormously by digitizing their business. firstly, streamlined supply chains and smart factories can boost efficiency in the organization (frank et al., 2019; nagy et al., 2018; stock & seliger, 2016). secondly, corporate decision-making processes can be improved (deloitte, 2015; kamble et al., 2018) and, third and lastly the possibility of developing new businesses (kans & ingwald, 2016; lee et al., 2014; prause, 2015). in this article we focus on the last of these opportunities. we then explore the four business model typologies that can generate this type of opportunity. 3.1. industry 4.0 business models industry 4.0 is a general change of the model established a few years ago, starting from the optimization of physical assets to a totally revolutionary scenario based on the cyber-physical system as a transforming technology in order to manage interconnected systems through advanced data and information gathering, contributing an improved product lifecycle. mckinsey & company (2015), in its study “industry 4.0: how to navigate digitization of the manufacturing sector”, identifies this data optimization as an end-to-end digital stream, briefly: a “digital thread” running through the entire product lifecycle as its digital representation. to advance this digitalization process, it starts with the digital design of the product, through the transfer of the digitally controlled production process, leading to the digital monitoring of the final product during operation (e.g. for productivity improvement purposes), closing the cycle with the recycling of the product. in each of the phases, the aim of the digital information structure is to enable: the easy exchange of data, the visualization of the controlled processes via digital interfaces/tools (e.g. tablets, virtual glasses) and permit interconnection via digital channels (e.g. teleservice). in addition, the exploitation and exchange of information across this stream will benefit from greater cross-functional integration and closer collaboration across the entire product lifecycle, including different stakeholders, such as suppliers, partners or clients. the focus is evolving from a single production site to production networks spanning multiple sites belonging to the company, including the entire supply chain. therefore, the goal of digital thread optimization is to make the best use of information. all industry 4.0 technologies are similar to each other in that they offer ways of harnessing data to unlock its value potential (mckinsey&company, 2015). for example, turning information into valuable results through advanced analytics that help decision-makers. at present, many companies are still at a nascent stage of this revolution. this type of technology has shown that oil rig companies, for example, are losing up to 99% of their data before reaching operational decision makers through information loss. consequently, it is important to manage data effectively (and incorporate it into the dynamics of business management), as every information leak causes inefficiencies, which would otherwise be valuable in many places along the value chain. in short, actively managing information to avoid information leakage is the key to seizing the new opportunities offered by digitalization. mckinsey & company (2015) therefore proposes four basic activities to generate value from data: 1. data capture and recording. 2. information transfer. 3. information processing and synthesis. 4. converting information into results. with the correct application of the four tasks described above, industry 4.0 offers opportunities that can maximize performance and generate profits in traditional manufacturing companies. on the one hand, there is the so called “smart factory” that focuses on the production process itself, using digital tools to make production more efficient and of a higher quality (bag et al., 2018; luthra et al., 2020). on the other hand, there is the use of these same technologies to generate new business models, int. j. prod. manag. eng. (2022) 10(1), 77-90creative commons attribution-noncommercial-noderivatives 4.0 international evolution of servitization: new business model opportunities 79 http://creativecommons.org/licenses/by-nc-nd/4.0/ from innovative proposals to new or potential customers (ayala et al., 2019; bartodziej, 2017; deloitte, 2015; frank et al., 2019; ibarra et al., 2018; müller et al., 2021; ślusarczyk, 2018). we often say that digital technology enables the emergence of new business models, but we are actually referring more to the creation of innovative value propositions rather than to business models as a whole, of which the first concept is a part. however, it is not at all easy to move from this level of abstraction to more identifiable lines of action. accordingly, the contribution of schaeffer (2017) is quite enlightening. certainly not for discovering absolutely new things, but for visualizing them in a more clarifying way. for this author, as shown in figure 2 and applicable to any sector, there are three business models (value propositions in our language) enabled by the irruption of digitalization: “platforms/ marketplaces”, “information value add business model” and “as a service business model”. the first of these, “platform/marketplace” is already very recognizable in the consumer world (airbnb, uber, etc.), however, it is gaining traction in the industrial world, where six of the world’s largest companies (amazon, ibm, cisco, general electric, microsoft and ptc) are already actively vying to lead this space of opportunity. the second of the models, the “information value add business model” (data as a revenue generator), has to do with everything related to predictive maintenance services, for example, those most talked about in our manufacturing environment. nonetheless, in general it concerns everything that arises from the analysis of data captured from products, services, customer experiences, etc. that allow us to anticipate and personalize value-added services in the marketplace. finally, we have perhaps the most specific of all, “as a service business model”, which revolves around pay-per-use. with some variations, all sectors are incorporating this logic. thus, to give just two examples, mckinsey & company (2015) speaks of four business models, in addition to the three previously mentioned, the possibility of selling knowledge in a consultancy model or selling any type of licenses, as shown in the figure 3. in the second example, we can talk about another very important area of opportunity: the circular economy. experts in this field, such as könnölä (2017), point to digitalization as a “master pillar” in the take-off of circular economy strategies. circular economy in the coming years. here again, a recent report by accenture strategy (2015) identifies digital business models for the circular economy that will be familiar to us (see figure 4): figure 2. business model in the digital age (source: schaeffer (2017)). int. j. prod. manag. eng. (2022) 10(1), 77-90 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz de la torre & sanchez 80 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 3. digital business models for industry 4.0 (source: mckinsey & company (2015)). figure 4. technologies used by companies pioneering the adoption of circular business models (source: accenture strategy (2015)). in this case, the authors highlight the platform model and the product-as-a-service model that encompasses the “as a service business model” and the “information value add business model”. these new business models are leading to a major change from revenues from physical product sales to revenues from more service-based platforms and developer ecosystems, subsequently creating a shift in the ways value is created for both manufacturers and suppliers. while in the manufacturing industry, sales of actual products have traditionally been the largest value group in terms of share of total revenue, this share is likely to decline in favor of new business models in the coming years. for car manufacturers, for example, sources of value are shifting from initial revenues from vehicle sales to recurring revenues based on usage. this is primarily driven by interoperability, with the potential to unleash a significant shift in revenue distribution through five main groups in the sector: vehicle price, connectivity hardware, driver time and attention, maintenance and insurance. hardware will become more accessible, reducing barriers for new market entrants. traditional value chains will be dismantled, creating new sources of revenue from value and, therefore, new opportunities. one of the keys to the future is related to being able to offer the new business models described above to harness the potential to create additional value and redistribute existing value sets that industry 4.0 offer. therefore, all businesses that want to make the leap into digital should consider whether they are developing or planning to develop initiatives in any of the digital models presented here, regardless of the sector in which they operate. a digital transformation effort that does not incorporate a change in the business model is not very credible. in addition to model change, there is a more complex and uncertain landscape. currently, many of the breakthrough technologies are driven by small, innovative companies that have specialized in a particular field (mckinsey&company, 2015). these companies are often more agile than large, established firms, consequently, smaller firms can generally implement new business models more easily, while larger firms need to consider about how to become more agile. moreover, the number of players is likely to rise, increasing the complexity and multiplicity of interfaces. a likely outcome is the increasing emergence of highly specialized players (rüßmann et al., 2015). another consequence of changes in the value chain may be the entry of established operators outside traditional manufacturing, such as telecommunications enterprises providing solutions for machine to machine connectivity or data security. consequently, traditional value chains are undergoing a radical transformation (mckinsey & company, 2015; kohnová et al., 2019). instead of a company developing and producing a complete product, a higher degree of specialization (value chain disintegration) is likely to occur (barreto et al., 2017; gebauer et al., 2013). this can already be seen in the semiconductor industry, for example, where foundries are manufacturing products for other semiconductor companies, also known as “fabless” manufacturing where the focus is on developing and commercializing the technology. this is especially of interest to manufacturing companies with high investment needs for manufacturing workshops and a high level of complexity, such as the aviation int. j. prod. manag. eng. (2022) 10(1), 77-90creative commons attribution-noncommercial-noderivatives 4.0 international evolution of servitization: new business model opportunities 81 http://creativecommons.org/licenses/by-nc-nd/4.0/ aftermarket. companies can apply uninterrupted monitoring in order to enhance their maintenance and repair business, reduce the cost generated by services, improve the use of their facilities and their spare parts planning. with such a significant disruption of the value chain expected, there are still many unknown fields (xu et al., 2018), cybersecurity, for example, where which type of company has the best chance of becoming dominant player remains open to conjecture. will it be telecommunication companies, it companies or microchip suppliers, or will a completely new player or supplier develop around the new demand? next, due to their importance in the current business context, we would like to briefly review the “platforms/marketplaces” and “as a service business model” business models following the terminology previously proposed by eric schaeffer (2017). the “information value add business model” model, which we call servitization. the platform concept has been defined as “business scale powered by the ability to leverage and orchestrate a global connected ecosystem of producers and consumers toward efficient value creation and exchange”. (choudary et al., 2015). therefore, in the near future, the scenario in which we may find ourselves is one in which industrial components are connected through the cloud on a platform, where they can dump all the information they collect and also interact with other industrial objects. customers, suppliers and other partners also interact in this virtual space in order to optimize and make the value chain more valuable. according to the latest reports, in 2018 more than 50% of large companies and 80% of companies with advanced digital transformation strategies would be associated with this type of industrial platforms. the most foreseeable scenario is not that every company will have its own cloud, but that a few winning platforms will connect the vast majority of industrial objects, as is the case today with mobile phones, which are basically connected through two or three platforms. take the example of “predix”, the iot platform that general electric has been rolling out in recent years. the official definition of the platform is already quite striking: “a cloud-based operating system for industrial application [...] the world’s first industrial internet platform”. predix is a strategic commitment by general electric, as mentioned at the beginning, to position itself in a privileged position in the digital revolution of the industry. in this cloud, predix has created applications that serve its objects or machines of various types, but the interesting thing is that it has created a development environment so that not only predix but any developer can create services and applications for their particular case. this in turn means that a market of applications, algorithms, etc. will be created for the members of this ecosystem. in fact, there is already a “predix app showcase” where you can buy these applications or start developing your own. although we have used the example of predix throughout, this same orientation is repeated in the rest of the companies that want to build this type of platform. seeing how this platform dynamic has worked in other sectors, it is highly likely that, ultimately, a few of these platforms will account for a very high percentage of all connected industrial objects. in this context, we can identify two types of value proposition that will emerge. firstly, those companies that build and succeed with these platforms will be able to monetize their infrastructure through different service channels (by connection, use and sale of applications, etc.), however, due to the very nature of this value proposition, we believe it is very difficult for basque industrial companies to do so. smes, generally detached from the world’s large technology conglomerates, have opportunities to play a leading role in this scenario. nonetheless, and secondly, there is the part of new value propositions that will be offered to customers once machines and other industrial components of a company’s value proposition are within this ecosystem. by connecting to them, companies can have access to entire data networks, find new customers, use or develop new applications, pay-peroutcome value propositions etc. and, above all, build product and service chains with third parties more easily and faster than ever before. we can highlight two aspects in this challenge. the first has to do with interconnecting this new information ecosystem with the organization’s existing systems (erp, crm, etc.), which is why the platform should offer openness and ease of integration through modular systems, apis, etc. and, secondly, selecting the platform to which we will connect our products, a vitally important issue as the range of possibilities that will open up to us will depend to a large extent on this. int. j. prod. manag. eng. (2022) 10(1), 77-90 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz de la torre & sanchez 82 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.2. the road to servitization 3.2.1. definition in order to understand servitization it is necessary to have a clear definition. the following table summarizes some of the most important definitions according to this research. within the 4 business models that are generated through industry 4.0, the business model we will refer to below is the “information value added business model” and more specifically the possibilities that open up in the form of servitization and, as a more concrete example, predictive maintenance. capital goods companies that want to evolve towards “service-oriented” business models will require fundamental changes in the company, posing them a major challenge. however, there is no doubt that the most competitive and value-added companies will make this leap sooner rather than later. in a survey of 600 manufacturing managers in 13 countries, 86% of them say that the transition from a product-based to a service-based strategy is a key part of their growth (macaulay et al., 2015). however, when asked about their concrete expectations on the topic over the next five years, these are quite low, as shown in figure 5 (macaulay et al., 2015). obviously, the situation is explained by the fact that all managers perceive the wave of change, but are not yet very clear on how they are going to “surf” it. another recent study analyzed how despite the fact that almost all managers understood that digitization was key to the future of their organizations, only 45% of respondents indicated that this topic was a top-level element of importance on their boards (bradley et al., 2015). figure 5. growth of the service business model (source: macaulay (2015)). these data towards servitization can also be observed in macroeconomic magnitudes. for example, service-related jobs within manufacturing companies are on the rise, making it increasingly difficult for many manufacturing companies to say whether they are manufacturers of products or providers of services, or in other words “manufacturing is no longer the same as the production of goods”. another very significant fact is that approximately 1/3 of the value of manufactured products consists of services. seen in reverse, it is estimated that 40% of the output created by the service sector is used as intermediary input by manufacturing firms (stehrer et al., 2014). digitalization is set to boost this trend towards servitization exponentially in the coming years. table 1. servitization definition (source: own elaboration). authors & year definition levitt, 1981 “servitization, which entails adding extra service components to core products”. vandermerwe & rada, 1988 “the increased offering of fuller market packages or ‘bundles’ of customer focused combinations of goods, services, support, self-service and knowledge in order to add value to core product offerings”. bart et al., 2003 “a trend in which manufacturing firms adopt more and more service components in their offerings”. baines et al., 2007 “the innovation of a manufacturing organization’s capabilities and processes to shift from selling product to selling an integrated product and service offering that delivers value in use”. schmenner, 2009 “the innovations in the supply chains of companies in the latter half of the nineteenth century lead us straight to the servitization innovations of today; it was then in history where service begins to be bundled with goods and controlled by the same company”. baines & lightfoot, 2014 “the servitization phenomenon that has pervaded manufacturing has resulted in organizations offering complex packages of both product and service to generate superior customer exchange value and thus enhance competitive edge”. kowalkowski et al., 2017 “the transformational process of shifting from a product-centric business model and logic to a service-centric approach”. int. j. prod. manag. eng. (2022) 10(1), 77-90creative commons attribution-noncommercial-noderivatives 4.0 international evolution of servitization: new business model opportunities 83 http://creativecommons.org/licenses/by-nc-nd/4.0/ as shown in table 1, the concept of servitization has been constantly developing over the last 40 years. although the first references to this term date back to the early 1980s regarding this concept developed by levitt in the usa, most authors have based their work on the vandermerwe & rada (1988) definition, helping servitization to evolve and acquire different nuances that reflect its importance within different scopes. the term ‘product’ is generally internalized in the manufacturing industry, however, when defining ‘services’, many tacitly define this based on what is not a product (baines et al., 2009). in this paper, we will consider servitization as “the intangible economic activities that add value to core product”. on the basis of the vandermerwe & rada (1988) definition, services began to be considered intangible, beyond production yet needed. after the division of products and services, many authors have presented similar definitions with different connotations over the years. in addition to the main definitions of the servitization concept, it has also had variants and contributions implicitly linked to the definition, as shown in the table 2. among the different definitions, variants and contributions it is important to highlight a key factor that was missing in most of the previous cases: taking the customer’s needs through services into account. this concept was integrated by a productservice system (pss) which consists of “tangible products and intangible services designed and combined so that jointly they are capable of fulfilling specific customer needs” (tukker, 2004). the pss business model allows organizations to create new sources of added value propositions focused on end users (baines et al., 2007), increase competitiveness through satisfying customer needs, build stronger relationships and, in turn, provide innovative solutions. the latest evolution of servitization is related to industry 4.0. today, beyond any doubt, advanced services, digitalization and iot are considered to be of high added value in the manufacturing industry as a defense against other lower cost economies (baines et al., 2009; baumgartner, 1999; tukker, 2004; vandermerwe & rada, 1988), mainly in sectors with a high saturation of marketed products (baines et al., 2009; baumgartner, 1999; windahl et al., 2004). many authors confuse industry 4.0 with servitization, however, servitization goes further because once industry 4.0 is established and stabilized in the market, servitization will continue to evolve due to new competitive strategies. 3.3. typologies of servitization the new classification of formulae to offer services proposed in the work of adrodegari et al. (2015), is an important starting point to help better understand the shift towards service-oriented business models in manufacturing companies. truly, one of the innovative keys in the coming years will be the difference in the monetization of certain models from others, above many points to be developed. for these authors, there are five basic forms of service value proposition (see figure 1) that, starting with the most basic and ending with the most advanced, would be what is shown in figure 6. product focused: the supplier separately sells the product or system and the needs of the customers for services during the product use phase (for example: repairs, maintenance contract, etc.). table 2. servitization variants and contributions (source: own elaboration). variants & contributions authors & year performance economy stahel, 2010. product-service-systems tukker, 2004. service business expansion gebauer et al., 2005; oliva & kallenberg, 2003. service business performance fang et al., 2008; gebauer et al., 2005. services growth strategies gebauer et al., 2010; oliva & kallenberg, 2003; ulaga et al., 2011. service profitability kwak & kim, 2016; gebauer & fleisch, 2007. solution delivery davies et al., 2007; galbraith, 2002b. solution marketing tuli et al., 2007 solutions provision davies et al., 2006; galbraith, 2002b. int. j. prod. manag. eng. (2022) 10(1), 77-90 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz de la torre & sanchez 84 http://creativecommons.org/licenses/by-nc-nd/4.0/ focused on product processes: the main difference with the previous one is that the company offers services, both in pre-sale and postsale phases, aimed at increasing the efficiency and effectiveness of customer operations, e.g., system customization, support of use processes, full risk maintenance contracts. focused on access (availability): the customer does not buy the product, but pays a regular flat fee to have access to it. the fee is not related to the actual use of the product and may include additional services (for example, maintenance and insurance costs). focused on use: the customer does not buy the product or system, but pays a variable rate that depends on the use of the product (pay-per-use time, pay-per-use unit...). focused on results: the customer does not buy the product or system, but pays a fee that depends on achieving a contractually established result in terms of product/system performance or the result of its use. in figure 7, baines and lightfoot, (2014) also explain this same idea of increasing the complexity of an offer of advanced services in a very graphic way. therefore, according to fleisch et al. (2015) we are faced with a scenario in which a connected product is capable of generating data which, when properly analyzed, is capable of being the basis on which to build digital services that add value to the customer. for example, a sensor-based connected machine can generate huge amounts of data which, with proper analysis, are capable of predicting the failure of a machine in advance. this data processing allows for planned maintenance and therefore increases the efficiency of the entire system (simply put, we call this process predictive maintenance). a further step would be pay-per-use or per manufactured unit, i.e., the last stage of servitization. figure 7. from basic to advanced services (source: baines et al. (2014)). figure 8. digital services development process (source: fleisch et al. (2015)). figure 6. typologies of service value proposition models (source: adrodegari et al. (2015)). int. j. prod. manag. eng. (2022) 10(1), 77-90creative commons attribution-noncommercial-noderivatives 4.0 international evolution of servitization: new business model opportunities 85 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. conclusion as a conducted literature review, it is clear that servitization has evolved considerably since its creation, ever increasing in potential with respect to products. it is also essential to develop the necessary capabilities to reach the advanced services required to compete in current situations. on the one hand, nobody doubts the improvements and benefits associated with servitization that strengthen and retain customer relationships, generate new and constant income streams or establish differentiating competitive advantages with respect to competitors. on the other hand, obstacles are currently encountered when implementing servitization. primarily, furnishing organizations with the necessary capacities to face all the new advanced services linked to industry 4.0 and the amount of information available due to digitalization and ultimately managing these new proposed product and services business model typologies. in conclusion, there is a need for further research into new business models derived from the arrival of industry 4.0, with the aim of managing new value propositions based on models focused on customer needs. managing the correct applicability of these models will be the key to success when implementing advanced services in servitization. to close, it is important to state that despite industry 4.0 being key in the short term, servitization will continue evolving and all organizations will have to adapt to new streams. references accenture strategy (2015). la ventaja circular: tecnologías y modelos de negocio innovadores para generar valor en un mundo sin límites de crecimiento. adrodegari, f., alghisi, a., ardolino, m., & saccani, n. (2015). from ownership to service-oriented business models: a survey in capital goods companies and a pss typology. procedia cirp, 30, 245–250. https://doi.org/10.1016/j. procir.2015.02.105 anderson, j. c., narus, j. a., & van rossum, w. (2006). customer value propositions in business markets. harvard business review, 84(3), 1–8. ayala, n. f., gerstlberger, w., & frank, a. g. (2019). managing servitization in product companies: the moderating role of service suppliers. international journal of operations and production management, 39(1), 43–74. https://doi.org/10.1108/ijopm-08-2017-0484 bag, s., telukdarie, a., pretorius, j. h. c., & gupta, s. (2018). industry 4.0 and supply chain sustainability: framework and future research directions. benchmarking, 28(5), 1410–1450. https://doi.org/10.1108/bij-03-2018-0056 baines, t., lightfoot, h., peppard, j., johnson, m., tiwari, a., shehab, e., & swink, m. (2009). towards an operations strategy for product-centric servitization. international journal of operations and production management, 29(5), 494–519. https://doi.org/10.1108/01443570910953603 baines, t., & lightfoot, h. w. (2014). servitization of the manufacturing firm: exploring the operations practices and technologies that deliver advanced services. international journal of operations and production management, 34(1), 2–35. https://doi.org/10.1108/ijopm-02-2012-0086 baines, t. s., lightfoot, h. w., evans, s., neely, a., greenough, r., peppard, j., roy, r., shehab, e., braganza, a., tiwari, a., alcock, j. r., angus, j. p., basti, m., cousens, a., irving, p., johnson, m., kingston, j., lockett, h., martinez, v., … wilson, h. (2007). state-of-the-art in product-service systems. proceedings of the institution of mechanical engineers, part b: journal of engineering manufacture, 221(10), 1543–1552. https://doi.org/10.1243/09544054jem858 barreto, l., amaral, a., & pereira, t. (2017). industry 4.0 implications in logistics: an overview. procedia manufacturing, 13, 1245–1252. https://doi.org/10.1016/j.promfg.2017.09.045 bart et al. (2003). servitization: or why services management is relevant for manufacturing environments. services management: an integrated approach, 39. bartodziej, c. j. (2017). the concept industry 4.0. the concept industry 4.0, 27–50. https://doi.org/10.1007/978-3-65816502-4_3 baumgartner, p. (1999). go downstream: the new. harvard business review, 77(5), 1–18. int. j. prod. manag. eng. (2022) 10(1), 77-90 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz de la torre & sanchez 86 https://doi.org/10.1016/j.procir.2015.02.105 https://doi.org/10.1016/j.procir.2015.02.105 https://doi.org/10.1108/ijopm-08-2017-0484 https://doi.org/10.1108/bij-03-2018-0056 https://doi.org/10.1108/01443570910953603 https://doi.org/10.1108/ijopm-02-2012-0086 https://doi.org/10.1243/09544054jem858 https://doi.org/10.1016/j.promfg.2017.09.045 https://doi.org/10.1007/978-3-658-16502-4_3 https://doi.org/10.1007/978-3-658-16502-4_3 http://creativecommons.org/licenses/by-nc-nd/4.0/ benedettini, o., & visnjic, i. (2011). the servitization of manufacturing: abstract. european operations management association conference, july, 1–9. retrieved from https://www.researchgate.net/profile/andy-neely/ publication/265006912_the_servitization_of_manufacturing_further_evidence/links/5474eaad0cf29afed60ffc20/ the-servitization-of-manufacturing-further-evidence.pdf bigdeli, a., baines, t., … o. b.-… r. a. i., & 2017, u. (2017). organisational change towards servitization: a theoretical framework. competitiveness review, 27(1), 12–39. https://doi.org/10.1108/cr-03-2015-0015 bradley, j., loucks, j., macaulay, j., noronha, a., & wade, m. (2015). digital vortex. retrieved from http://gtmtecno.com/ wp-content/uploads/2015/11/20151109-2395xx_digital_vortex_06182015.pdf brax, s. (2005). a manufacturer becoming service provider challenges and a paradox. managing service quality, 15(2), 142–155. https://doi.org/10.1108/09604520510585334 buckley & majumdar. (2018). the services powerhouse: increasingly vital to world economic growth. deloitte insights, 12. bustinza, o. f., parry, g. c., & vendrell-herrero, f. (2013). supply and demand chain management: the effect of adding services to product offerings. supply chain management, 18(6), 618–629. https://doi.org/10.1108/scm-05-2013-0149 ceci, f., & masini, a. (2011). balancing specialized and generic capabilities in the provision of integrated solutions. industrial and corporate change, 20(1), 91–131. https://doi.org/10.1093/icc/dtq069 chesbrough, h., & rosenbloom, r. s. (2002). the role of the business model in capturing value from innovation: evidence from xerox corporation’s technology spin-off companies. academic. oup.com. industrial and corporate change, 11(3), 529-555. https://doi.org/10.1093/icc/11.3.529 choudary et al. (2015). platform scale: how an emerging business model helps startups build large empires with minimum investment. platform thinking labs. cohen, m., … n. a.-h., & 2006, undefined. (2006). winning in the aftermarket. fishmandavidson.wharton.upenn …. retrieved from https://fishmandavidson.wharton.upenn.edu/wp-content/uploads/2016/06/winning-in-theaftermarket_hbr_may_06.pdf dachs, b., biege, s., borowiecki, m., lay, g., jäger, a., & schartinger, d. (2012). the servitization of european manufacturing industries. datta, p. p., & roy, r. (2011). operations strategy for the effective delivery of integrated industrial product-service offerings: two exploratory defence industry case studies. international journal of operations and production management, 31(5), 579–603. https://doi.org/10.1108/01443571111126337 davies, a., brady, t., & hobday, m. (2006). charting a path toward integrated solutions. in mit sloan management review (vol. 47, issue 3, pp. 39–48). https://www.researchgate.net/publication/215915574_charting_a_path_toward_ integrated_solutions davies, a., brady, t., & hobday, m. (2007). organizing for solutions: systems seller vs. systems integrator. industrial marketing management, 36(2), 183–193. https://doi.org/10.1016/j.indmarman.2006.04.009 deloitte. (2015). industry 4.0. challenges and solutions for the digital transformation and use of exponential technologies. deloitte, 1–30. https://www.academia.edu/download/42978372/deloittes_study_on_industry_4.0.pdf fang, e. (er), palmatier, r. w., & steenkamp, j.-b. e. m. (2008). effect of service transition strategies on firm value: journal of marketing, 72(5), 1–14. https://doi.org/10.1509/jmkg.72.5.001 fleisch, e., weinberger, m., & wortmann, f. (2015). business models and the internet of things (extended abstract). lecture notes in computer science (including subseries lecture notes in artificial intelligence and lecture notes in bioinformatics), 9001, 6–10. https://doi.org/10.1007/978-3-319-16546-2_2 frank, a. g., dalenogare, l. s., & ayala, n. f. (2019). industry 4.0 technologies: implementation patterns in manufacturing companies. international journal of production economics, 210, 15–26. https://doi.org/10.1016/j.ijpe.2019.01.004 galbraith, j. r. (2002a). center for effective organizations, organizing to deliver solutions. galbraith, j. r. (2002b). organizing to deliver solutions. organizational dynamics, 31(2), 194–207. https://doi.org/10.1016/ s0090-2616(02)00101-8 gebauer, h,. fleisch, e., & friedli, t. (2005). overcoming the service paradox in manufacturing companies. european management journal, 23(1), 14-26. https://doi.org/10.1016/j.emj.2004.12.006 gebauer, h., edvardsson, b., gustafsson, a., & witell, l. (2010). match or mismatch: strategy-structure configurations in the service business of manufacturing companies. journal of service research, 13(2), 198–215. https://doi.org/10.1177/1094670509353933 gebauer, h., & fleisch, e. (2007). an investigation of the relationship between behavioral processes, motivation, investments in the service business and service revenue. industrial marketing management, 36(3), 337–348. https://doi.org/10.1016/j.indmarman.2005.09.005 int. j. prod. manag. eng. (2022) 10(1), 77-90creative commons attribution-noncommercial-noderivatives 4.0 international evolution of servitization: new business model opportunities 87 https://www.researchgate.net/profile/andy-neely/publication/265006912_the_servitization_of_manufacturing_further_evidence/links/5474eaad0cf29afed60ffc20/the-servitization-of-manufacturing-further-evidence.pdf https://www.researchgate.net/profile/andy-neely/publication/265006912_the_servitization_of_manufacturing_further_evidence/links/5474eaad0cf29afed60ffc20/the-servitization-of-manufacturing-further-evidence.pdf https://www.researchgate.net/profile/andy-neely/publication/265006912_the_servitization_of_manufacturing_further_evidence/links/5474eaad0cf29afed60ffc20/the-servitization-of-manufacturing-further-evidence.pdf https://doi.org/10.1108/cr-03-2015-0015 http://gtmtecno.com/wp-content/uploads/2015/11/20151109-2395xx_digital_vortex_06182015.pdf http://gtmtecno.com/wp-content/uploads/2015/11/20151109-2395xx_digital_vortex_06182015.pdf https://doi.org/10.1108/09604520510585334 https://doi.org/10.1108/scm-05-2013-0149 https://doi.org/10.1093/icc/dtq069 https://doi.org/10.1093/icc/11.3.529 https://fishmandavidson.wharton.upenn.edu/wp-content/uploads/2016/06/winning-in-the-aftermarket_hbr_may_06.pdf https://fishmandavidson.wharton.upenn.edu/wp-content/uploads/2016/06/winning-in-the-aftermarket_hbr_may_06.pdf https://www.researchgate.net/publication/215915574_charting_a_path_toward_integrated_solutions https://www.researchgate.net/publication/215915574_charting_a_path_toward_integrated_solutions https://doi.org/10.1016/j.indmarman.2006.04.009 https://www.academia.edu/download/42978372/deloittes_study_on_industry_4.0.pdf https://doi.org/10.1509/jmkg.72.5.001 https://doi.org/10.1007/978-3-319-16546-2_2 https://doi.org/10.1016/j.ijpe.2019.01.004 https://doi.org/10.1016/s0090-2616(02)00101-8 https://doi.org/10.1016/s0090-2616(02)00101-8 https://doi.org/10.1016/j.emj.2004.12.006 https://doi.org/10.1177/1094670509353933 https://doi.org/10.1016/j.indmarman.2005.09.005 http://creativecommons.org/licenses/by-nc-nd/4.0/ gebauer, h., & friedli, t. (2005). behavioral implications of the transition process from products to services. journal of business and industrial marketing, 20(2), 70–78. https://doi.org/10.1108/08858620510583669 gebauer, h., paiola, m., & saccani, n. (2013). characterizing service networks for moving from products to solutions. industrial marketing management, 42(1), 31–46. https://doi.org/10.1016/j.indmarman.2012.11.002 habegger, b. (2010). strategic foresight in public policy: reviewing the experiences of the uk, singapore, and the netherlands. futures, 42(1), 49–58. https://doi.org/10.1016/j.futures.2009.08.002 ibarra, d., ganzarain, j., & igartua, j. i. (2018). business model innovation through industry 4.0: a review. procedia manufacturing, 22, 4–10. https://doi.org/10.1016/j.promfg.2018.03.002 jong, j. p. j., & vermeulen, p. a. m. (2003). organizing successful new service development: a literature review. management decision, 41(9), 844–858. https://doi.org/10.1108/00251740310491706 kamble, s. s., gunasekaran, a., & gawankar, s. a. (2018). sustainable industry 4.0 framework: a systematic literature review identifying the current trends and future perspectives. process safety and environmental protection, 117, 408– 425. https://doi.org/10.1016/j.psep.2018.05.009 kans, m., & ingwald, a. (2016). business model development towards service management 4.0. procedia cirp, 47, 489–494. https://doi.org/10.1016/j.procir.2016.03.228 kim, j. h. (2017). a review of cyber-physical system research relevant to the emerging it trends: industry 4.0, iot, big data, and cloud computing. journal of industrial integration and management, 02(03), 1750011. https://doi. org/10.1142/s2424862217500117 kinnunen, r. (2018). creating customer value using digital tools in scaffolding industry. retrieved from https://lutpub.lut. fi/handle/10024/157114 kohnová, l., papula, j., practice, n. s.-b. t. and, & 2019, undefined. (2019). internal factors supporting business and technological transformation in the context of industry 4.0. ceeol.com. https://doi.org/10.3846/btp.2019.13 könnölä, t. (2017). totti könnölä: la digitalización como pilar maestro de la economía circular. el mundo: https://www.elmundo.es/economia/innovadores/2017/10/27/59f2f245e2704e5a458b4894.html. kowalkowski, c., gebauer, h., kamp, b., & parry, g. (2017). servitization and deservitization: overview, concepts, and definitions. industrial marketing management, 60, 4–10. https://doi.org/10.1016/j.indmarman.2016.12.007 kwak, k., & kim, w. (2016). effect of service integration strategy on industrial firm performance. journal of service management, 27(3), 391–430. https://doi.org/10.1108/josm-03-2014-0088 lay, g. (2014). servitization in industry. in servitization in industry (vol. 9783319069). https://doi.org/10.1007/978-3319-06935-7 lee, j., bagheri, b., & kao, h. a. (2015). a cyber-physical systems architecture for industry 4.0-based manufacturing systems. manufacturing letters, 3, 18-23. https://doi.org/10.1016/j.mfglet.2014.12.001 lee, j., kao, h. a., & yang, s. (2014). service innovation and smart analytics for industry 4.0 and big data environment. procedia cirp, 16, 3–8. https://doi.org/10.1016/j.procir.2014.02.001 leitão, p., colombo, a. w., & karnouskos, s. (2016). industrial automation based on cyber-physical systems technologies: prototype implementations and challenges. computers in industry, 81, 11-25. https://www.sciencedirect.com/science/ article/pii/s0166361515300348 https://doi.org/10.1016/j.compind.2015.08.004 levitt, t. (1981). marketing intangible products and product intangibles. cornell hotel and restaurant administration quarterly, 22(2), 37–44. https://doi.org/10.1177/001088048102200209 liao, y., deschamps, f., rocha loures, e., felipe ramos, l., de freitas rocha loures, e., & felipe pierin ramos, l. (2017). past, present and future of industry 4.0-a systematic literature review and research agenda proposal. international journal of production research, 55(12), 3609–3629. https://doi.org/10.1080/00207543.2017.1308576 lu, y. (2017). industry 4.0: a survey on technologies, applications and open research issues. journal of industrial information integration, 6, 1–10. https://doi.org/10.1016/j.jii.2017.04.005 luthra, s., kumar, a., zavadskas, e. k., mangla, s. k., & garza-reyes, j. a. (2020). industry 4.0 as an enabler of sustainability diffusion in supply chain: an analysis of influential strength of drivers in an emerging economy. international journal of production research, 58(5), 1505–1521. https://doi.org/10.1080/00207543.2019.1660828 macaulay, j., connell, k. o., & delaney, k. (2015). the digital manufacturer resolving the service dilemma. cisco, november, 1–20. mathieu, v. (2001). product services: from a service supporting the product to a service supporting the client. journal of business and industrial marketing, 16(1), 39–53. https://doi.org/10.1108/08858620110364873 mckinsey&company. (2015). how to navigate digitization of manufacturing sector. https://www.mckinsey.com/~/ media/mckinsey/business functions/mckinsey digital/our insights/getting the most out of industry 4 0/mckinsey_ industry_40_2016.pdf int. j. prod. manag. eng. (2022) 10(1), 77-90 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz de la torre & sanchez 88 https://doi.org/10.1108/08858620510583669 https://doi.org/10.1016/j.indmarman.2012.11.002 https://doi.org/10.1016/j.futures.2009.08.002 https://doi.org/10.1016/j.promfg.2018.03.002 https://doi.org/10.1108/00251740310491706 https://doi.org/10.1016/j.psep.2018.05.009 https://doi.org/10.1016/j.procir.2016.03.228 https://doi.org/10.1142/s2424862217500117 https://doi.org/10.1142/s2424862217500117 https://lutpub.lut.fi/handle/10024/157114 https://lutpub.lut.fi/handle/10024/157114 https://www.elmundo.es/economia/innovadores/2017/10/27/59f2f245e2704e5a458b4894.html https://doi.org/10.1016/j.indmarman.2016.12.007 https://doi.org/10.1108/josm-03-2014-0088 https://doi.org/10.1007/978-3-319-06935-7 https://doi.org/10.1007/978-3-319-06935-7 https://doi.org/10.1016/j.mfglet.2014.12.001 https://doi.org/10.1016/j.procir.2014.02.001 https://www.sciencedirect.com/science/article/pii/s0166361515300348 https://www.sciencedirect.com/science/article/pii/s0166361515300348 https://doi.org/10.1016/j.compind.2015.08.004 https://doi.org/10.1177/001088048102200209 https://doi.org/10.1080/00207543.2017.1308576 https://doi.org/10.1016/j.jii.2017.04.005 https://doi.org/10.1080/00207543.2019.1660828 https://doi.org/10.1108/08858620110364873 https://www.mckinsey.com/~/media/mckinsey/business%20functions/mckinsey%20digital/our%20insights/getting%20the%20most%20out%20of%20industry%204%200/mckinsey_industry_40_2016.pdf https://www.mckinsey.com/~/media/mckinsey/business%20functions/mckinsey%20digital/our%20insights/getting%20the%20most%20out%20of%20industry%204%200/mckinsey_industry_40_2016.pdf https://www.mckinsey.com/~/media/mckinsey/business%20functions/mckinsey%20digital/our%20insights/getting%20the%20most%20out%20of%20industry%204%200/mckinsey_industry_40_2016.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ müller, j. m., buliga, o., & voigt, k. i. (2021). the role of absorptive capacity and innovation strategy in the design of industry 4.0 business models a comparison between smes and large enterprises. european management journal, 39(3), 333–343. https://doi.org/10.1016/j.emj.2020.01.002 nagy, j., oláh, j., erdei, e., máté, d., & popp, j. (2018). the role and impact of industry 4.0 and the internet of things on the business strategy of the value chain—the case of hungary. sustainability, 10(10), 3491. https://doi.org/10.3390/ su10103491 neely, a. (2008). exploring the financial consequences of the servitization of manufacturing. operations management research, 1(2), 103–118. https://doi.org/10.1007/s12063-009-0015-5 negri, e., fumagalli, l., & macchi, m. (2017). a review of the roles of digital twin in cps-based production systems. procedia manufacturing, 11, 939–948. https://doi.org/10.1016/j.promfg.2017.07.198 nicolae, a., korodi, a., & silea, i. (2019). an overview of industry 4.0 development directions in the industrial internet of things context. romanian journal of information science and technology, 22(4), 183–201. oliva, r., & kallenberg, r. (2003). managing the transition from products to services. international journal of service industry management, 14(2), 160–172. https://doi.org/10.1108/09564230310474138 prause. (2015). security and sustainability issues. ministry of national defence republic of lithuania journal of. journal of security & sustainability issues. https://doi.org/10.9770/jssi.2015.5.2 reim, w., parida, v., & örtqvist, d. (2015). product–service systems (pss) business models and tactics – a systematic literature review. journal of cleaner production, 97, 61–75. https://doi.org/10.1016/j.jclepro.2014.07.003 roscitt, r. (1990). strategic service management. journal of business & industrial marketing, 5(1), 27–40. https://doi.org/10.1108/eum0000000002735 rüßmann, m., lorenz, m., gerbert, p., waldner, m., justus, j., engel, p., & harnisch, m. (2015). industry 4.0: the future of productivity and growth in manufacturing industries. inovasyon.org. retrieved from https://inovasyon.org/images/ haberler/bcgperspectives_industry40_2015.pdf sandström, s., edvardsson, b., kristensson, p., & magnusson, p. (2008). value in use through service experience. managing service quality: an international journal, 18(2), 112–126. https://doi.org/10.1108/09604520810859184 schaeffer, e. (2017). industry x.0: realizing digital value in industrial sectors. kogan page publishers. schmenner, r. w. (2009). manufacturing, service, and their integration: some history and theory. international journal of operations and production management, 29(5), 431–443. https://doi.org/10.1108/01443570910953577 ślusarczyk, b. (2018). industry 4.0: are we ready? polish journal of management, 17(1). https://doi.org/10.17512/ pjms.2018.17.1.19 smith, d. j. (2013). power-by-the-hour: the role of technology in reshaping business strategy at rolls-royce. technology analysis and strategic management, 25(8), 987–1007. https://doi.org/10.1080/09537325.2013.823147 stahel, w. (2010). the performance economy. springer. stehrer, r., baker, p., foster-mcgregor, n., & koenen, j. (2014). study on the relation between industry and services in terms of productivity and value creation. https://madoc.bib.uni-mannheim.de/45990/ stock, t., & seliger, g. (2016). opportunities of sustainable manufacturing in industry 4.0. procedia cirp, 40, 536–541. https://doi.org/10.1016/j.procir.2016.01.129 tukker, a. (2004). eight types of product-service system: eight ways to sustainability? experiences from suspronet. business strategy and the environment, 13(4), 246–260. https://doi.org/10.1002/bse.414 tukker, a. (2015). product services for a resource-efficient and circular economy – a review. journal of cleaner production, 97, 76–91. https://doi.org/10.1016/j.jclepro.2013.11.049 tuli, k. r., kohli, a. k., & bharadwaj, s. g. (2007). rethinking customer solutions: from product bundles to relational processes. journal of marketing, 71(3), 1–17. https://doi.org/10.1509/jmkg.71.3.001 ulaga, w., reinartz, w. j., gartmeier, v., lévèque, l., orbach, d., van putten, v., & lobschat, l. (2011). hybrid offerings: how manufacturing firms combine goods and services successfully. journal of marketing, 75(6), 5–23. https://doi.org/10.1509/jm.09.0395 vandermerwe, s., & rada, j. (1988). servitization of business: adding value by adding services. european management journal, 6(4), 314–324. https://doi.org/10.1016/0263-2373(88)90033-3 varghese, a., & tandur, d. (2014). wireless requirements and challenges in industry 4.0. proceedings of 2014 international conference on contemporary computing and informatics, ic3i 2014, 634–638. https://doi.org/10.1109/ ic3i.2014.7019732 verstrepen, s., deschoolmeester, d., & berg, r. j. (1999). servitization in the automotive sector: creating value and competitive advantage through service after sales. in global production management (pp. 538–545). https://doi. org/10.1007/978-0-387-35569-6_66 int. j. prod. manag. eng. (2022) 10(1), 77-90creative commons attribution-noncommercial-noderivatives 4.0 international evolution of servitization: new business model opportunities 89 https://doi.org/10.1016/j.emj.2020.01.002 https://doi.org/10.3390/su10103491 https://doi.org/10.3390/su10103491 https://doi.org/10.1007/s12063-009-0015-5 https://doi.org/10.1016/j.promfg.2017.07.198 https://doi.org/10.1108/09564230310474138 https://doi.org/10.9770/jssi.2015.5.2 https://doi.org/10.1016/j.jclepro.2014.07.003 https://doi.org/10.1108/eum0000000002735 https://inovasyon.org/images/haberler/bcgperspectives_industry40_2015.pdf https://inovasyon.org/images/haberler/bcgperspectives_industry40_2015.pdf https://doi.org/10.1108/09604520810859184 https://doi.org/10.1108/01443570910953577 https://doi.org/10.17512/pjms.2018.17.1.19 https://doi.org/10.17512/pjms.2018.17.1.19 https://doi.org/10.1080/09537325.2013.823147 https://doi.org/10.1016/j.procir.2016.01.129 https://doi.org/10.1002/bse.414 https://doi.org/10.1016/j.jclepro.2013.11.049 https://doi.org/10.1509/jmkg.71.3.001 https://doi.org/10.1509/jm.09.0395 https://doi.org/10.1016/0263-2373(88)90033-3 https://doi.org/10.1109/ic3i.2014.7019732 https://doi.org/10.1109/ic3i.2014.7019732 https://doi.org/10.1007/978-0-387-35569-6_66 https://doi.org/10.1007/978-0-387-35569-6_66 http://creativecommons.org/licenses/by-nc-nd/4.0/ vuksanović, d., ugarak, j., & korčok, d. (2016). industry 4.0: the future concepts and new visions of factory of the future development. portal.sinteza.singidunum.ac.rs, 293–298. https://doi.org/10.15308/sinteza-2016-293-298 windahl, c., andersson, p., berggren, c., & nehler, c. (2004). manufacturing firms and integrated solutions: characteristics and implications. european journal of innovation management, 7(3), 218–228. https://doi. org/10.1108/14601060410549900 windahl, c., & lakemond, n. (2006). developing integrated solutions: the importance of relationships within the network. industrial marketing management, 35(7), 806–818. https://doi.org/10.1016/j.indmarman.2006.05.010 xu, l. d., xu, e. l., & li, l. (2018). industry 4.0: state of the art and future trends. international journal of production research, 56(8), 2941-2962. https://doi.org/10.1080/00207543.2018.1444806 int. j. prod. manag. eng. (2022) 10(1), 77-90 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz de la torre & sanchez 90 https://doi.org/10.15308/sinteza-2016-293-298 https://doi.org/10.1108/14601060410549900 https://doi.org/10.1108/14601060410549900 https://doi.org/10.1080/00207543.2018.1444806 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.15894 received: 2021-07-06 accepted: 2021-10-18 conceptual model for assessing the lean manufacturing implementation maturity level in machinery and equipment of small and medium-sized enterprises jia yuik chong a1, & puvanasvaran a. perumal a2 a faculty of manufacturing engineering, universiti teknikal malaysia melaka, jalan hang tuah jaya, 76100 durian tunggal, melaka, malaysia. a1 chongjiayuik@gmail.com, a2 punesh@utem.edu.my abstract: the adoption of lean manufacturing (lm) in small and medium-sized enterprises (smes) is not as vigorous as in large organizations. this purpose of this study is to assess the maturity level of lm implementation in the machinery and equipment (m&e) smes. the close-ended survey questionnaire method was adopted in three malaysian manufacturing m&e smes, and data was collected for the descriptive analysis. the findings showed that these case companies are generally at a low-to-moderate level in terms of lm understanding. meanwhile, the extent of lm implementation and the success level is still moderate. the proposed lm conceptual model provides valuable perspectives and establishes a holistic understanding of the phenomena in lm maturity status for m&e smes. the proper synchronization of lm understanding, implementation, and success are vital to building the strong lm maturity foundation for lean organizational transformation. it serves as useful guidance and strategic framework to other companies in dealing with the operational excellence challenges. the significance of this study will help m&e smes to identify their current position and promote progress in the lean application journey. this will benefit the management team and lean practitioners in decision-making and enhance tactics to attain a higher level of success. key words: lean manufacturing, small and medium-sized enterprises, lean maturity, machinery and equipment, conceptual model. 1. introduction small and medium-sized enterprises (smes) are always constantly searching for new chances to continuously develop their business sector transformation in the globally competitive market. nowadays, the manufacturing industry faced various challenges in business sustainability, operational efficiency, and cost-saving. lean manufacturing (lm) is one of the systematic management systems or tools that can help firms provide value-added processess to customers and minimize unnecessary waste (achanga et al., 2006; driouach et al., 2019; womack et al., 1990). lean production implementation leads to operational excellence and enhances product quality (driouach et al., 2019; liker, 2004; shah and ward, 2002; ulewicz and kucȩba, 2016; womack et al., 1990; yahya et al., 2019). in malaysia’s manufacturing sector, smes are classified as companies with sales of rm50 million or not exceeding 200 full-time employees (sme corp. malaysia, 2020). smes are the core of the economy and contribute to the nation’s development. malaysian smes gross domestic product (gdp) grew by 5.8% in 2019 compared to 6.2% in 2018. the share of gdp contributed by smes rose to 38.9% in 2019 from 38.3% in the previous year (dosm, to cite this article: chong, j.y., perumal, a.p. (2022). conceptual model for assessing the lean manufacturing implementation maturity level in machinery and equipment of small and medium-sized enterprises. international journal of production management and engineering, 10(1), 23-32. https://doi.org/10.4995/ijpme.2022.15894 http://polipapers.upv.es/index.php/ijpme creative commons attribution-noncommercial-noderivatives 4.0 international 23int. j. prod. manag. eng. (2022) 10(1), 23-32 https://orcid.org/0000-0003-3822-415x http://orcid.org/0000-0002-2144-6277 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2020). although lm is commonly applied in large enterprises, many existing studies still illustrated the inconsistency of lm adoption in different industries and countries, especially in smes (abu et al., 2019; khusaini et al., 2014; nordin et al., 2013). achanga et al. (2006) stated that many smes had not adopted lm. driouach et al. (2019) claimed that many smes are still struggled to introduce lm to their processes. in the malaysian 11th programme, the machinery and equipment (m&e) sector serves as a driving force to shift to higher economic growth. the sub-sectors of m&e had contributed rm41.5 billion to exports in 2019 compared to rm40.5 billion in 2018 (mida, 2020). this reflects malaysia’s efficient and fastexpanding economy in achieving towards regional production hub for m&e. malaysian m&e sectors are categorized into four main sub-sectors (mida, 2020): i. specialized process machinery or equipment for specific industry; ii. metalworking machinery; iii. power generating machinery and equipment; iv. general industrial machinery & equipment, components, and parts. shah and hussain (2016) found that the textile sector in pakistan just begin to implement lm, and more than half are in-transitional due to insufficient understanding of the lean concept. kherbach et al. (2019) indicated that 87% of the surveyed small manufacturers in romania need lean training, especially for their engineers and supervisors. based on the study conducted by antosz and stadnicka (2017), 55% of the automotive smes in poland did not implement lm, and for those who have, 29% are used the 5s lean tool solely. sumantri (2017) recorded low lean implementation in logistic operations among smes in east java of indonesia, attributed to internal resistance, unavailability of resources, lack of leadership, and inadequate training. from the survey of 84 manufacturing smes in north africa, belhadi et al. (2018) concluded that despite their great need for lean, its implementation is very low due to the lack of resources and cultural issues. nordin et al. (2013) revealed weak lm implementation in their preliminary survey at 30 malaysian automotive component manufacturing firms. the lm practices implementation in the malaysian food and beverage industry is at its infancy based on the questionnaire survey of 53 organizations (khusaini et al., 2014); besides, the majority of surveyed firms considered the negative perception towards lm as the most critical barrier. abu et al. (2019) studied the lm implementation in 148 malaysian furniture companies, found that employee-related matters such as lack of labour resources, poor application of know-how, and employee resistance to change are the barriers in lean organizations. ali maasouman and demirli (2015) presented a lean maturity model, which is crucial in achieving a sustainable lean status but is limited to manufacturing cells. yadav et al. (2019b) highlighted that smes are frequently overlooked by researchers when it comes to lean adoption in comparison to large enterprises. even though there are no statistically significant differences between smes shown in the maturity of lean practices adoption in brazilian manufacturing companies, the findings vary when smaller companies are compared to large enterprises (bento and tontini, 2019). accordingly, the above studies indicated that lm implementation remains surprisingly low and lacks attention by the researchers. although it was introduced in several industries in many countries other than the automotive industry, most smes still struggle to advance their lean practices. furthermore, the present literature research on lm adoption for malaysian m&e is just 2.3%, much less than the automobile industry’s 37.1% (osman et al., 2020). as a result, this study aims to assess the maturity level of lm implementation in m&e smes to close the current gaps. according to literature reviewed by zanon et al. (2020), a greater lean maturity status is linked to better operational performance. the created lm conceptual model provides useful perspectives into understanding phenomena in lm maturity progress in m&e smes. it serves as useful guidance and strategic framework to other companies in dealing with operational challenges. the significance of this study will help m&e smes to identify their current position in the lean application journey. this paper is outlined into six sections: the first section mentions the overall lm implementation maturity level at different countries or industries. it highlights the problem statement and gap analysis. the second section reviews the current lm maturity assessment frameworks or models application. the third section contains the research methodology and overview of the case companies background. the fourth section presents the survey analysis. the lm maturity conceptual model is developed and discussed in the fifth section. lastly, the sixth section explains the conclusions, research implications, and future works recommendations. chong & perumal creative commons attribution-noncommercial-noderivatives 4.0 international24 int. j. prod. manag. eng. (2022) 10(1), 23-32 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. review of lm maturity assessment frameworks in smes driouach et al. (2019) stated that smes, especially very small business (vsbs), still have difficulties adapting lm to their process within the organizations based on the various lean implementation frameworks presented in their literature review. carvalho et al. (2019) revealed that the large organizations in brazil have already demonstrated some reluctance in implementing lm. still, small businesses are unaware that lm exists or is too cozy to rethink and implement it practically. majava and ojanperä (2017) stated that the lean philosophy application is a workable and famous approach in developing and enhancing production activities from a case study analysis in finnish-based manufacturing smes. the findings from sahoo and yadav (2018) suggested overall positive effects on a firm’s operational performance as a result of adopting lean tools and philosophy in indian smes. rose et al. (2010) recommended fifteen feasible lean practices for smes split into three category levels: basic, intermediate, and advanced. basic lean techniques are fundamental tools and soft skills executed at the operational level, such as 5s, visual control, quality circle, and teamwork. intermediate lean practices would require cross-department initiatives and include cellular manufacturing, setup time reduction, continuous improvement, etc. finally, advanced lean methods use the first two categories to carry out more intricate lean functions, such as small lot sizes, kanban, and continuous flow. belhadi et al. (2016) proposed an implementation framework for lean production in smes, comprising of the necessary components (process, tools, success factors) in three phases. afonso and cabrita (2015) developed a lean supply chain framework based on the performance measurement perspective, but it lacks process performance indicators and other soft lean practices (human elements). leite et al. (2016) formulated a roadmap for applying lean techniques within smes’ product development teams; hemilä and vilko (2015) developed a model in manufacturing industry supply chain smes, but both lean methodologies are not applicable in terms of production operation perspective. sousa et al. (2018) implemented project management good practices with lean production methodologies to enhance the process efficiencies in a smes portuguese innovation company via an action research methodology. five phases of implementation were developed to enhance the production from the “as-is” model to the future “to-be” model by applying lean principles and tools. belhadi et al. (2017b) presented a significant improvement in the lean performance indicators with the successful deployment of the framework of lean principles through a set of success factors. ali maasouman and demirli (2015) proposed a visual maturity model to evaluate the overall leanness for lm cells through analysis of lean in seven axes. zanon et al. (2020) developed the integrated lean maturity model showed that the performance assessment practices through the application of a performance measurement system (pms) could foster lean practices and promote improvements in the organizations. kolla et al. (2019) stated that there are many evaluation models which are available to assess the performance of an organization in accordance with either lean production system (lps) or industry 4.0; however, these models are also complicated and do not meet the specific needs of smes for further progression. this is one of the key issues that prohibit the effectiveness and advancement of lm applications in smes. ramadas and satish (2018) showed that employee barriers in lm implementation for smes are substantial inadequate training and experienced staff, lack of knowledge of experienced specialists, and cultural resistance to change. shah et al. (2019) proposed the lean manufacturing in small-sized engineering organizations (lemseo) model for implementing lm successfully in three phases which consists of 20 steps for smes, namely lm awareness, the confidence of successful implementation, and full-scale sustainable lm projects. abu et al. (2021) presented that culture and human behaviour, knowledge, and resources are three main factors that have a synergetic impact on lm implementation in manufacturing industries. based on the presented literature review, there is an apparent lack of investigation of the detail regarding lm implementation maturity level, specifically mapping with the m&e smes. so far, this subject’s research does not furnish a comprehensive lm framework for enhancing lean maturity among employees and establishing a robust lean culture in smes. most of the existing frameworks are focused on the lm implementation operational steps, lean principles enhancement, lean tools application, and lean performance measurement. before proceeding with the actual lm implementation, the current lm maturity level assessment is still less highlighted conceptual model for assessing the lean manufacturing implementation maturity level in machinery and equipment of small and medium-sized enterprises creative commons attribution-noncommercial-noderivatives 4.0 international 25int. j. prod. manag. eng. (2022) 10(1), 23-32 http://creativecommons.org/licenses/by-nc-nd/4.0/ in the early beginning stage. besides that, there is a lack of synchronization and proper alignment for the thorough lm understanding, implementation, and success level to enhance the effectiveness and efficiency of the lean application in smes. this always resultant in poor lm implementation and caused the lean initiative to fail eventually. therefore, there is a strong need to develop a practical lm maturity assessment model specifically created to provide a valuable guide for lm adoption in m&e smes to bridge this gap. 3. research methodology this study was carried out by using a multiple case studies approach. the purposive sampling approach was used to choose three smes that shown high interest and desire to partake. the case study explored a holistic understanding of the phenomenon for lm implementation experiences in m&e smes. the selection criteria are m&e companies that fulfilled the malaysian manufacturing sme’s definitions are eligible for this study. table 1 depicts the overall background of the three case companies. all organizations have been in operation for more than 10 years and have varying degrees of lm implementation. the targeted respondents were questioned about their current extent of lm implementation maturity in the company in terms of three different dimensions of assessments (understanding, implementation, and success). the interview questionnaires structures are divided into two sections as following: i. the first part is about the general company information and the background of respondents. ii. the second part is to assess lm maturity in terms of understanding, implementation, and success level. all case companies were informed three weeks in advance before the visit. the survey questionnaires were distributed on site. the questionnaires developed were checked by the two local university academicians who are specialist in the lm area and good manufacturing practices for pilot study before ready for field data collection. this validated that the smes’ respondents comprehended the questions’ context and ensured the results’ trustworthiness. a total of 40 respondents from m&e smes were selected from executive-level and manager-level staff to answer the close-ended survey in the company premises. in the survey questionnaires, there is a total of three questions. these questions were formed based on a five-point likert scale in order to gauge each variable’s level of implementation. the scale was ranged from 1 to 5 where 1 = very low, 2 = low, 3 = moderate, 4 = high, and 5 = very high. the respondents were briefed and requested to rate against the questions (variables) in assessing their agreement by given values ranging from 1 (lowest) to 5 (highest). during the site plant tours, the operation process on the production floor was examined to verify the respondents’ replied. this is followed by descriptive analysis of the data collected using microsoft excel, and the results were tabulated for discussion. table 1. case studies company background information. company name a b c year of establishment 1990 2006 1997 company ownership family own joint venture joint venture no. of full-time employees 32 60 40 year sales turnover (rm) within 5 million to 10 million range within 5 million to 10 million range within 10 million to 15 million range main products rubber machinery surface treatment industrial wires certifications/ achievements smes score 4 star (2019) iso 9001:2015; as9001; nadcap; sme award 2015 iso 9001:2015; iso 14001:2015 no. of years lm implementation ≈3 years ≈7 years ≈15 years manufacturing type high mix low volume high mix low volume low mix high volume type of m&e sub-sector general industrial m&e parts specialized process in m&e agriculture specialized process in m&e aerospace chong & perumal creative commons attribution-noncommercial-noderivatives 4.0 international26 int. j. prod. manag. eng. (2022) 10(1), 23-32 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. survey analysis of lm maturity level in the case m&e smes most respondents of the case smes stated that the level of understanding towards lm in their company was only moderate (73%), while 13% indicated that it was at a high level. likewise, the extent of lm implementation in the case companies was around 58% at a moderate level and 30% at a high level. half of the total respondents claimed that the level of success towards lm implementation projects was at an intermediate level, and 30% agreed it was high. none of the respondents replied that the status of lm implementation maturity was very high. the extent of lm maturity of the three m&e cases was evaluated, as shown in figure 1. in summary, the overall mean score achieved for the level of lm understanding in the m&e sme cases was found at low-to-moderate level (2.90), while the extent of lm implementation (3.15) and present success (3.08) is at a moderate level. employee involvement and the lack of appropriate lean subject matter knowledge and personnel skills, especially in smes, need to be emphasized as a key issue. in each case, the top management leadership teams support the lean initiative programs and are willing to allocate relevant resources to the project to ensure the lm implementation is successful. however, fundamental lean training on lean principles must be provided to all employees at the beginning stage of the project management and encourage them to practice the lean principles in the workplace. this will make the employees feel that what they have learned is valuable and can be applied directly to achieve successful outcomes. however, the correlation for the respondents’ perception in the context of the current lean adoption versus the actual real-life lm application in the case companies still needs to be practically validated via in-depth empirical study. from the results analysed, there was still much room for improvement to achieve a better performance in these three case companies towards mature lean organizations and readiness of lm implementation from end-to-end in a higher level for the long-term sustainability. the study found a similar perception of lm in terms of understanding, implementation, and success level in these cases, echoing the literature review (nordin et al., 2013; punnakitikasem et al., 2012; shah and hussain, 2016). in general, lean applications in the case of smes are still fragmented and imbalanced. the current extent of lm implementation in these smes was far from being world-class. human factors (soft lean methods) are critical for smes (mamat et al., 2015). despite sufficient training provided, smes are usually unable to achieve the intended results of lm adoption ultimately. staff needs to put their newly acquired abilities into action as fast as possible in order to build up and deepen their grasp of lm. this should be incorporated into the employee’s performance management plan. employees will value lean concepts more if they are connected to and convinced that lm will make their jobs easier. smes are often found to be overconfident in their abilities and capabilities to apply lean practices in one go (rose et al., 2017). however, smes are unable to sustain massive losses in business due to their financial capacities. the figure 1. a survey analysis of lm maturity level. conceptual model for assessing the lean manufacturing implementation maturity level in machinery and equipment of small and medium-sized enterprises creative commons attribution-noncommercial-noderivatives 4.0 international 27int. j. prod. manag. eng. (2022) 10(1), 23-32 http://creativecommons.org/licenses/by-nc-nd/4.0/ lm adoption is obviously still not broad but rather restricted to specific parts of the organization with minimal success. the issue is common among smes that are dealing with more acute resource and lm skills limitations. 5. development of lm maturity conceptual model the lm implementation maturity in m&e smes was assessed in the three critical dimensions: the level of lm understanding, the extent of lm implementation, and the lm implementation success level. all these three dimensions were investigated independently rather than together in previous studies. the analysis shows that the overall lm implementation maturity level in the m&e smes is low to moderate. this finding aligns with the research presented by yadav et al. (2019b), which indicates that the lean implementation in smes are not explicit, as many smes have minimal understanding and knowledge in lm. the m&e smes respondents perceived that they have obtained the basic lm understanding and adopted the fundamental lean tools practices in a particular process application. however, the extent of lm implementation is still very limited or covered in specifically selected areas but not expand on a large scale at the entire organizational level. figure 2 shows the combination of these three dimensions and their integrated elements that formed the lm maturity conceptual model for enhancement. from analysis, the success level achieved was fragmented, as there is still inconsistency of lm implementation in smes. smes management and lean practitioners can utilize this proposed model to gauge the status of lm maturity and define the right strategy to move on. abdallah et al. (2021) stated that both social (human aspects) and technical (tools and techniques) lm were discovered to have a beneficial impact on operational performance in manufacturing smes. puvanasvaran et al. (2015) performed a study of lean behavior in business development and information technology (it) found out that the employees were lack of substantial lean implementation expertise and skillsets. the results revealed that employees’ lean behaviour practices had been improved after adopting the lean tools. the strong connection of these dimensions, with the total commitment of employees from all organization levels, play an essential role to create a sustainable lm maturity culture for continuous improvement as follows: i. lm thinking improvement mindset – right lean thinking and positive attitude can speed up the improvement efforts. ii. lm theory learning and application – the fundamental knowledge about lean concepts or principles of learning needs to be appropriately applied in the workplace for problem-solving. iii. lm key performance measurement – the relationship of key performance input variables figure 2. conceptual model of lm maturity level. chong & perumal creative commons attribution-noncommercial-noderivatives 4.0 international28 int. j. prod. manag. eng. (2022) 10(1), 23-32 http://creativecommons.org/licenses/by-nc-nd/4.0/ with process approach adoption aligned with key performance output variables (cost, quality, time, delivery, etc.) in lm result assessment. amrani et al. (2018) conducted the lean maturity assessment by calculating the leanness level divided into three categories: structural, organizational, and human metrics. the continuous improvement cycle is another essential component of this lm maturity model. lean maturity evaluation process steps are an iterative cycle and an ongoing improvement journey. thus, smes must practice the feasible lean tools to grasp the in-depth understanding and expand the lm implementation to other critical areas in order to achieve tremendous success at entire organizational level. 5.1. lm thinking improvement mindset the lm change management is essential as early readiness for all employees towards the cultural changes in smes. the creation of the lm environment and culture will build a positive lean thinking mindset towards a continuous improvement journey. the end in mind concept shall be instilled in every employee’s mindset to visualize the lm success result in the future. critical lean thinking has been introduced by liker (2004) with emphasized the five key lean principles, which are: specify the value, mapping the value stream, creating flow, establishing pull, and pursuing perfection. the lm thinking involves eliminating the waste and creating value-added activities in fulfilling the customer’s requirements. the application of lean thinking in smes can significantly improve organizational efficiency in india (yadav et al., 2019a). the processbased approach and risk-based thinking can assist in identifying the top pareto of waste areas. the top management shall inform the strategic planning in the company and support the employees in working towards the common lean goals successfully. the current state of “as-is” needs to define and the future state “to-be” state in a lean roadmap established. the right lm thinking mindset will affect the employee’s attitude and strive to achieve excellent job performance in lm implementation. management is crucial in deciding an organization’s success or failure. the management should always promote the benefits of lm and convince their employees of the significant implication of lm if they execute systematically. the employer can develop their employee and nurture the talents to unleash their workforce’s limitless potential by instilling them with correct lean behaviours. 5.2. lm theory learning and application the lm theory learning is still very poor, especially in smes; this is due to the lack of an in-depth understanding of the real meaning of the lean concept. belhadi et al. (2017a) stated that the true meaning of the lean theory and principles does not seem to be readily grasped or perceived by smes, which could be a major obstacle to achieving anticipated results. it is claimed that if stakeholders understand the lean benefits, they will be more inspired to implement the initiative (yadav et al., 2019c). many smes claimed that they already apply lean but without knowing that in detail. this caused them to apply the wrong lean tool and techniques in their workplace. the common barrier in lm implementation is the misuse or wrong application of lm tools practices. belhadi et al. (2017a) claimed that the deficiency of tailored lean implementation approaches in the improvement programs always leads smes to a major problem in adapting the implementation process to suit their requirements. this always caused the lm initiative to fail and not able to deliver the intended results. smes should adopt suitable approaches in the lm application as not all the lean tools are appropriate to apply in smes. puvanasvaran et al. (2010) proposed the people development system (pds) to improve the problem-solving capabilities among its staff while applying lean process management. the management should allocate the necessary resources and encourage the employee to upgrade their skills expertise in problem-solving. the management shall lead as a role model in cultivating the applied learning environment in the organization. the engagement of lean consultants will transfer the elementary knowledge for the employees in order to for them to be well prepared for the lean task assigned. the lean principles learned should have cross-link to their respective daily jobs perform in the workplaces and apply it directly to assess the effectiveness in problem-solving. 5.3. lm key performance measurement the proper selection of key performance measurement matrices is crucial to determine and assess the success level of lm implementation in smes. the more remarkable lm maturity level will create better operational performances. it has been found that after lean practices have been completely applied in the targeted areas with established key indicators, the intended results will be attained (santos bento and tontini, 2018). the active involvement from all the employees will motivate each other and move conceptual model for assessing the lean manufacturing implementation maturity level in machinery and equipment of small and medium-sized enterprises creative commons attribution-noncommercial-noderivatives 4.0 international 29int. j. prod. manag. eng. (2022) 10(1), 23-32 http://creativecommons.org/licenses/by-nc-nd/4.0/ forward with teamwork. the individual or team performance in job competency and capability is strongly tied with the performance measurement in providing the rewards and promotion to higher-level jobs. amrani et al. (2018) analyzed the leanness metrics across three elements: consistency, frequency, and relevancy to evaluate the advancement of lean implementation. the lean assessment can be carried out to assess the various key performances such as people, process, and technology for continuous improvement. the high involvement of employees in lean will increase lm implementation at the entire organizational level and create a solid lean culture. according to the findings from galeazzo (2019), the degree of leanness is unrelated to financial performance, while lean maturity favorable influence on financial performance. therefore, the key focuses on performance outcomes are cost, quality, and time, creating value-added to customers. the critical success factors for enhancing the lm implementation in smes need to be identified to improve the possibility of success. smes need to place the top pareto of waste areas with prioritization, as knowing that the smes have limited resource capability in lean project implementation. the selection of key performance input variables (process parameter setting) is vital to match with its corresponding key performance output variables (cost, quality, time, delivery, etc.), producing the quantifiable result in fulfilling the customer requirements with good product quality. 6. conclusions this study aims to assess the current maturity level of lm implementation, specifically in m&e smes. the findings from the three case studies revealed that the lm understanding level in the m&e smes is low to moderate. in contrast, the present lm implementation level and success level is still moderate. there are some critical challenges and inhibitors encountered during the lm adoption, especially on the human-related factors such as lean understanding knowledge adoption in enhancing employee’s lean skills expertise application that needs to be focused thoroughly by top management. the lm maturity level in m&e smes is assessed in three dimensions combined with three essential integrated elements to form a conceptual model. the implication of the proposed model provides the strategy guidelines for m&e sme’s lean practitioner and management level staff in prioritizing the available resources in enhancing the lm maturity sustainability to a higher level of success. the proper synchronization of lm understanding, implementation, and success are vital to building the strong lm maturity foundation towards the lean transformation in industry 4.0 for m&e smes. the limitation of this study is this conceptual model is still not yet practically validated. it is suggested that it be applied in the actual case study to verify the effectiveness further. the correlation of the key elements presented can be expanded via statistical analysis to test the significance level of the relationship in future studies. acknowledgements the authors would like to thank the smes case companies and universiti teknikal malaysia melaka (utem) as the research institution for their strong support in completing this study. references abdallah, a. b., alkhaldi, r. z., & aljuaid, m. m. (2021). impact of social and technical lean management on operational performance in manufacturing smes: the roles of process and management innovations. business process management journal, 27(5), 1418-1444. https://doi.org/10.1108/bpmj-06-2020-0252 abu, f., gholami, h., mat saman, m. z., zakuan, n., & streimikiene, d. (2019). the implementation of lean manufacturing in the furniture industry: a review and analysis on the motives, barriers, challenges, and the applications. journal of cleaner production, 234, 660–680. https://doi.org/10.1016/j.jclepro.2019.06.279 abu, f., gholami, h., saman, m. z. m., zakuan, n., streimikiene, d., & kyriakopoulos, g. l. (2021). an sem approach for the barrier analysis in lean implementation in manufacturing industries. sustainability, 13(4), 1978. https://doi.org/10.3390/su13041978 achanga, p., shehab, e., roy, r., & nelder, g. (2006). critical success factors for lean implementation within smes. journal of manufacturing technology management, 17(4), 460–471. https://doi.org/10.1108/17410380610662889 afonso, h., & cabrita, m. d. r. (2015). developing a lean supply chain performance framework in a sme: a perspective based on the balanced scorecard. procedia engineering, 131, 270–279. https://doi.org/10.1016/j.proeng.2015.12.389 chong & perumal creative commons attribution-noncommercial-noderivatives 4.0 international30 int. j. prod. manag. eng. (2022) 10(1), 23-32 https://doi.org/10.1108/bpmj-06-2020-0252 https://doi.org/10.1016/j.jclepro.2019.06.279 https://doi.org/10.3390/su13041978 https://doi.org/10.1108/17410380610662889 https://doi.org/10.1016/j.proeng.2015.12.389 http://creativecommons.org/licenses/by-nc-nd/4.0/ ali maasouman, m., & demirli, k. (2015). assessment of lean maturity level in manufacturing cells. ifac-papersonline, 28(3), 1876–1881. https://doi.org/10.1016/j.ifacol.2015.06.360 amrani, a., possik, j. j., & ducq, y. (2018). contribution to a lean maturity evaluation: leanness metrics calculation contribution to a lean maturity evaluation: leanness metrics calculation. september 2020. antosz, k., & stadnicka, d. (2017). lean philosophy implementation in smes study results. procedia engineering, 182, 25–32. https://doi.org/10.1016/j.proeng.2017.03.107 belhadi, a., touriki, f. e., & el fezazi, s. (2016). a framework for effective implementation of lean production in small and medium-sized enterprises. journal of industrial engineering and management, 9(3), 786–810. https://doi.org/10.3926/ jiem.1907 belhadi, a., touriki, f. e., & el fezazi, s. (2017a). prioritizing the solutions of lean implementation in smes to overcome its barriers: an integrated fuzzy ahp-topsis approach. journal of manufacturing technology management, 28(8), 1115–1139. https://doi.org/10.1108/jmtm-04-2017-0066 belhadi, a., touriki, f. e., & el fezazi, s. (2018). lean implementation in small and medium-sized enterprises in less developed countries: some empirical evidences from north africa. journal of small business management, 56(00), 132–153. https://doi.org/10.1111/jsbm.12396 belhadi, a., touriki, f. e., & fezazi, s. el. (2017b). lean deployment in smes, performance improvement and success factors: a case study. proceedings of the international conference on industrial engineering and operations management, 928–945. bento, g. dos s., & tontini, g. (2019). maturity of lean practices in brazilian manufacturing companies. total quality management and business excellence, 30(sup1), s114–s128. https://doi.org/10.1080/14783363.2019.1665827 carvalho, c. p., carvalho, d. s., & silva, m. b. (2019). value stream mapping as a lean manufacturing tool: a new account approach for cost saving in a textile company. international journal of production management and engineering, 7(1), 1. https://doi.org/10.4995/ijpme.2019.8607 department of statistics, malaysia (dosm). (2020, july 29). small and medium enterprises (smes) performance 2019. retrieved from https://www.dosm.gov.my/v1/index.php?r=column/cthemebycat&cat=159&bul_id=vjm1enz2rmlv rdvtnfawrwzizus3qt09&menu_id=te5cruzcblh4ztzmodzibmk2awrrqt09 driouach, l., zarbane, k., & beidouri, z. (2019). literature review of lean manufacturing in small and medium-sized enterprises. international journal of technology, 10(5), 930–941. https://doi.org/10.14716/ijtech.v10i5.2718 galeazzo, a. (2019). degree of leanness and lean maturity: exploring the effects on financial performance. total quality management and business excellence, 0(0), 1–19. https://doi.org/10.1080/14783363.2019.1634469 hemilä, j., & vilko, j. (2015). the development of a service supply chain model for a manufacturing sme. international journal of logistics management, 26(3), 517–542. https://doi.org/10.1108/ijlm-01-2014-0001 kherbach, o., mocan, m. l., & dumitrache, c. (2019). implementation of the lean manufacturing in local small and medium sized enterprises. journal of innovation management in small and medium enterprises, 1–10. https://doi. org/10.5171/2017.799859 khusaini, n. s., jaffar, a., & noriah, y. (2014). a survey on lean manufacturing practices in malaysian food and beverages industry. applied mechanics and materials, 564, 632–637. https://doi.org/10.4028/www.scientific.net/amm.564.632 kolla, s., minufekr, m., & plapper, p. (2019). deriving essential components of lean and industry 4.0 assessment model for manufacturing smes. procedia cirp, 81, 753–758. https://doi.org/10.1016/j.procir.2019.03.189 leite, m., baptista, a. j., & ribeiro, a. m. r. (2016). a road map for implementing lean and agile techniques in smes product development teams. international journal of product development, 21(1), 20–40. https://doi.org/10.1504/ ijpd.2016.076922 liker, j. k. (2004). the toyota way: 14 management principles from the world’s greatest manufacturers, new york: mcgraw-hill. majava, j., & ojanperä, t. (2017). lean production development in smes: a case study. management and production engineering review, 8(2), 41–48. https://doi.org/10.1515/mper-2017-0016 malaysian investment development authority (mida). (2020). malaysia’s machinery & equipment and engineering supporting industries. retrieved from https://www.mida.gov.my/wp-content/uploads/2020/07/20200707115544_ me2020.pdf mamat, r. c., md deros, b., ab rahman, m. n., omar, m. k., & abdullah, s. (2015). soft lean practices for successful lean production system implementation in malaysia automotive smes: a proposed framework. jurnal teknologi, 77(27), 141–150. https://doi.org/10.11113/jt.v77.6910 nordin, n., deros, b. m., & wahab, d. a. (2013). a study on lean manufacturing implementation in malaysian automotive component industry. international journal of automotive and mechanical engineering, 8(1), 1467–1476. https://doi.org/10.15282/ijame.8.2013.33.0121 conceptual model for assessing the lean manufacturing implementation maturity level in machinery and equipment of small and medium-sized enterprises creative commons attribution-noncommercial-noderivatives 4.0 international 31int. j. prod. manag. eng. (2022) 10(1), 23-32 https://doi.org/10.1016/j.ifacol.2015.06.360 https://doi.org/10.1016/j.proeng.2017.03.107 https://doi.org/10.3926/jiem.1907 https://doi.org/10.3926/jiem.1907 https://doi.org/10.1108/jmtm-04-2017-0066 https://doi.org/10.1111/jsbm.12396 https://doi.org/10.1080/14783363.2019.1665827 https://doi.org/10.4995/ijpme.2019.8607 https://www.dosm.gov.my/v1/index.php?r=column/cthemebycat&cat=159&bul_id=vjm1enz2rmlvrdvtnfawrwzizus3qt09&menu_id=te5cruzcblh4ztzmodzibmk2awrrqt09 https://www.dosm.gov.my/v1/index.php?r=column/cthemebycat&cat=159&bul_id=vjm1enz2rmlvrdvtnfawrwzizus3qt09&menu_id=te5cruzcblh4ztzmodzibmk2awrrqt09 https://doi.org/10.14716/ijtech.v10i5.2718 https://doi.org/10.1080/14783363.2019.1634469 https://doi.org/10.1108/ijlm-01-2014-0001 https://doi.org/10.5171/2017.799859 https://doi.org/10.5171/2017.799859 https://doi.org/10.4028/www.scientific.net/amm.564.632 https://doi.org/10.1016/j.procir.2019.03.189 https://doi.org/10.1504/ijpd.2016.076922 https://doi.org/10.1504/ijpd.2016.076922 https://doi.org/10.1515/mper-2017-0016 https://www.mida.gov.my/wp-content/uploads/2020/07/20200707115544_me2020.pdf https://www.mida.gov.my/wp-content/uploads/2020/07/20200707115544_me2020.pdf https://doi.org/10.11113/jt.v77.6910 https://doi.org/10.15282/ijame.8.2013.33.0121 http://creativecommons.org/licenses/by-nc-nd/4.0/ osman, a. a., othman, a. a., & abdul rahim, m. k. i. (2020). lean manufacturing adoption in malaysia: a systematic literature review. international journal of supply chain, operation management and logistics, 1(1), 01-35. https://doi.org/10.35631/ijscol.11001 punnakitikasem, p., buavarapor, n., maluesri, p., & leelartapin, k. (2012). health care service quality: case example of a hospital with lean implementation. poms 23 rd annual conference, chicago, illinois, u.s.a april 20 to april 23, 2012, 2–6. puvanasvaran, a.p., norazlin, n., suk fan, c. (2015). a study of lean behavior in business development and it department. appl. mech. mater. 761, 566–570. https://doi.org/10.4028/www.scientific.net/amm.761.566 puvanasvaran, p., megat, h., sai hong, t., mohd razali, m., abdel magid, s.h. (2010). lean process management implementation through enhanced problem solving capabilities. j. ind. eng. manag. 3, 447–493. https://doi. org/10.3926/jiem.2010.v3n3.p447-493 ramadas, t., & satish, k. p. (2018). identification and modeling of employee barriers while implementing lean manufacturing in smalland medium-scale enterprises. international journal of productivity and performance management, 67(3), 467–486. https://doi.org/10.1108/ijppm-10-2016-0218 rose, a. m. n., deros, b. m., & rahman, m. n. a. (2010). development of framework for lean manufacturing implementation in smes. the 11th asia pacific industrial engineering and management systems conference the 14th asia pacific regional meeting of international foundation for production research, december, 1–5. rose, a. n. m., ab rashid, m. f. f., nik mohamed, n. m. z., & ahmad, h. (2017). similarities of lean manufacturing approaches implementation in smes towards the success: case study in the automotive component industry. matec web of conferences, 87, 02024. https://doi.org/10.1051/matecconf/20178702024 sahoo, s., & yadav, s. (2018). lean implementation in smalland medium-sized enterprises: an empirical study of indian manufacturing firms. benchmarking, 25(4), 1121–1147. https://doi.org/10.1108/bij-02-2017-0033 santos bento, g. dos, & tontini, g. (2018). developing an instrument to measure lean manufacturing maturity and its relationship with operational performance. total quality management and business excellence, 29(9–10), 977–995. https://doi.org/10.1080/14783363.2018.1486537 shah, r., & ward, p. t. (2002). lean manufacturing: context, practice bundles, and performance. journal of operations management, 43221(august), 129–149. https://doi.org/10.1016/s0272-6963(02)00108-0 shah, s. m., murugesh, r., & devadasan, s. r. (2019). model for implementing lean manufacturing in small and mediumsized engineering organisations. international journal of indian culture and business management, 19(2), 244. https://doi.org/10.1504/ijicbm.2019.101737 shah, z. a., & hussain, h. (2016). an investigation of lean manufacturing implementation in textile industries of pakistan. proceedings of the international conference on industrial engineering and operations management, 668–677. sme corp. malaysia. (2020, april). guideline for sme definition. retrieved from https://www.smecorp.gov.my/images/ pdf/2020/guideline-smedefinition_updated.pdf sousa, p., tereso, a., alves, a., & gomes, l. (2018). implementation of project management and lean production practices in a sme portuguese innovation company. procedia computer science, 138, 867–874. https://doi.org/10.1016/j. procs.2018.10.113 sumantri, y. (2017). lean logistics implementation level in small and medium enterprises (smes) sector. journal of engineering and applied sciences, 12(2), 195–198. https://doi.org/10.3923/jeasci.2017.171.175 ulewicz, r., & kucȩba, r. (2016). identification of problems of implementation of lean concept in the sme sector. engineering management in production and services, 8(1), 19–25. https://doi.org/10.1515/emj-2016-0002 womack, j. p., jones, d. t., & roos, d. (1990). the machine that changed the world, new york: rawson associates. yadav, v., jain, r., mittal, m. l., panwar, a., & lyons, a. (2019a). the impact of lean practices on the operational performance of smes in india. industrial management and data systems, 119(2), 317–330. https://doi.org/10.1108/ imds-02-2018-0088 yadav, v., jain, r., mittal, m. l., panwar, a., & lyons, a. c. (2019b). the propagation of lean thinking in smes. production planning and control, 30(10–12), 854–865. https://doi.org/10.1080/09537287.2019.1582094 yadav, v., jain, r., mittal, m. l., panwar, a., & sharma, m. k. (2019c). an appraisal on barriers to implement lean in smes. journal of manufacturing technology management, 30(1), 195–212. https://doi.org/10.1108/jmtm-12-20170262 yahya, m. s., mohammad, m., omar, b., & sulistyo, b. (2019). factors influencing selection of lean tools and techniques in malaysian organisations. journal of advanced manufacturing technology, 13(special issue 1), 51–61. zanon, l. g., ulhoa, t. f., & esposto, k. f. (2020). performance measurement and lean maturity: congruence for improvement. production planning and control, 0(0), 1–15. https://doi.org/10.1080/09537287.2020.1762136 chong & perumal creative commons attribution-noncommercial-noderivatives 4.0 international32 int. j. prod. manag. eng. (2022) 10(1), 23-32 https://doi.org/10.35631/ijscol.11001 https://doi.org/10.4028/www.scientific.net/amm.761.566 https://doi.org/10.3926/jiem.2010.v3n3.p447-493 https://doi.org/10.3926/jiem.2010.v3n3.p447-493 https://doi.org/10.1108/ijppm-10-2016-0218 https://doi.org/10.1051/matecconf/20178702024 https://doi.org/10.1108/bij-02-2017-0033 https://doi.org/10.1080/14783363.2018.1486537 https://doi.org/10.1016/s0272-6963(02)00108-0 https://doi.org/10.1504/ijicbm.2019.101737 https://www.smecorp.gov.my/images/pdf/2020/guideline-smedefinition_updated.pdf https://www.smecorp.gov.my/images/pdf/2020/guideline-smedefinition_updated.pdf https://doi.org/10.1016/j.procs.2018.10.113 https://doi.org/10.1016/j.procs.2018.10.113 https://doi.org/10.3923/jeasci.2017.171.175 https://doi.org/10.1515/emj-2016-0002 https://doi.org/10.1108/imds-02-2018-0088 https://doi.org/10.1108/imds-02-2018-0088 https://doi.org/10.1080/09537287.2019.1582094 https://doi.org/10.1108/jmtm-12-2017-0262 https://doi.org/10.1108/jmtm-12-2017-0262 https://doi.org/10.1080/09537287.2020.1762136 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.1744 received 2013-10-08 accepted 2014-05-22 textile industry can be less pollutant: introducing naturally colored cotton solimar garciai*and irenilza de alencar nääsii universidade paulista unip rua dr. bacelar, 1212, 4º andar, vila clementino, são paulo, sp, cep: 04026-002, brazil. i solimargarcia10@gmail.com ii irenilza@gmail.com abstract: studies in agribusiness and textile industry, both involved with the production of manufacturing fashion presents insufficient development for new products that could represent water savings or reduction of chemical effluents, making this production chain a a sustainable business. this paper introduces the colored and organic cotton as an alternative to foster colored cotton producing farmers and improving the concept of sustainability in the textile sector. results show that the increase in the production of colored and organic cotton may result in reduction of water use, and consequent reduction in the disposal of effluents in nature. as the colored and organic cotton is produced by small farmers, governmental agencies are required to participate in the effort of improving its production and distribution, providing the needed infrastructure to meet the increasing business. this action would slowly encourage the reduction of white cotton consumption in exchange for this naturally colored product. the water used, and consequent polluted discharge in the use of colored cotton in the textile industry might be reduced by 70%, assuming a reduction of environmental impact of 5% per year would represent expressive numbers in the next ten years. key words: agribusiness chain, supply chain, sustainability in the fashion industry. 1. introduction worldwide the production of cotton fibers reach 70 million tons per year, occupying near 10% of arable land, reaching 34 million ha. the consumption of cotton in the textile fiber chain reaches 24.6 million tons per year transforming into various garments for apparel, hospitals linen, uniforms, etc. the textile chain has a turnover of us$ 400 billion per year (mapa, 2012). brazil is the fourth largest producer of apparel, and it is the world’s fifth largest producer, with near 30 thousand industries, which manufactures annually, near 10 billion pieces, and employing (directly and indirectly) 8 million workers, representing 16.4% of formal employment in the country. the sector revenues reached u.s. $ 60.5 billion in 2011, representing 3.5% of brazilian total gdp, and 5.5% of the manufacturing industry in the country’s gdp, a growth that exceeded 10% compared to 2010 (abit, 2012). brazil has the highest productivity rates among leading cotton producer countries. as a result from the investigation for almost 20 years of embrapa cotton (brazilian enterprise in agriculture, a governmental agency) a genetic seed of colored cotton was developed, along with special experience in its production, processing and marketing. available in different shades of brown, the agroecological cotton does not need to be tinted, and because of that it does not impact either in peoples’ health or the environment. the cotton cultivated variety was developed to minimize the use of chemical during production and processing, being a product that can be used by the organic market (table 4, figure 1). according to beltrão et al. (2009), this production has been developed mainly in the northeast of brazil, especially in the state of paraiba and put brazil in the global market of organic cotton. to keep the business at a fair price, however, quality and reliability of production needs to be dependable for ensuring the availability of jobs in family farming (cartaxo et al., 2008). from 2007, 265,517 bales of organic cotton were produced in 24 countries, and global production grew almost 50% per year (ota, 2009). this paper deals with the production of agro85int. j. prod. manag. eng. (2014) 2(2), 85-91creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2014.1744 http://creativecommons.org/licenses/by-nc-nd/4.0/ ecological and colored cotton as a real possibility of sustainable product for brazilian agribusiness, showing the benefits in terms of water saving and effluent reduction in relation to the production of white cotton, over a period of ten years. 2. methodology a literature review was done on the subject of cotton and colored cotton production. data were collected in former studies on the production of colored and organic cotton including the quantification of industries’ use of water, quantity and quality of effluents, and chemicals involved in the various phases of processing (galindo et al., 2001; sauer, 2002; beltrão and carvalho, 2004; forgiarini, 2006; beltrão et al., 2009; abit, 2012). data on sustainability principles and social responsibility regarding family farming was also searched (refosco et al., 2011; abreu et al., 2012; baskaran et al., 2012). this study was approved by the ethics committee of the universidade paulista. the figures obtained were subjected to a descriptive analysis to present a projection on possible savings in water use when colored cotton is used when compared to white cotton. reduction of effluent was also accounted when considering the textile industrial processes. 3. results and discussion 3.1. textile production chain chemical components output fashion production chain is represented by several industries such as agroindustry and the chemical industry fueling its main stream (spinning, weaving, knitting and non-woven materials (mendes, 2010). these cotton fibers supply the clothing industry to cater to the customer for different types of retail, and this value chain is highly polluting and aggressive by nature excessive water consumption (table 3), as well as by dumping effluents in nature (table 1). studying ten cotton processing textile industries, martins (1997) identified the reject the industry produces. these processes occur at all stages of the textile industry (manufacturing synthetic fibers and natural, spinning, weaving and knitting, pretreatment of fabrics, dyeing, printing, finishing, manufacturing, retail) (forgiarini, 2006). the use of water in the processing phases for producing 1 ton of knitting is also very high. table 2 shows that the reduction on the use of water is approximately 70% for processing the colored and ecologic cotton, as indicated by baskaran et al. (2012). the study of the chemical processes involved in the textile industry uses large amounts of water, and the environment around it becomes contaminated by chemicals and dyes after the industrial process. the waste in all its states (solid, liquid and gaseous), and the operations of bleaching dyeing and finishing effluents emit various chemical, which can cause problems for people and the environment when improperly discarded. the processing of cotton yarn is also to various processes for transforming raw materials of the textile articles in a natural state white, dyed, printed and finished as well as downsizing, produce polluting effluents. in the complexity surrounding the textile chain, there are still wet finishing processes in which prepares the fabric to be dyed, print by color or design or receive finishing. substances such as water, resins, dyes and surfactants are used in this phase of the process (forgiarini, 2006). the main chemical products and dyes found in textile effluent are from various origins (martins, 1997), and the author highlights the most polluting table 1. dejects generated in the production process (source: adapted from beltrão et al. (2004); galindo et al. (2001) and martins (1997)). phase components of the dejects pre-ironing humectants, salts, caustic soda, and peroxide ironing starch and synthetic gums based on poly-acrylate bleaching humectants, salts, caustic soda, sequestrates, peroxide and / or chlorine and neutralizers dyeing colorants, sequestrates, salts, caustic soda and / or kelp stamping dyes, caustic soda and gums washing detergents softening softeners and sliding table 2. basic water consumption for producing 1 ton of processed 1 ton of white cotton (source: the authors). water consumption for processing (1 ton/month) unit (l/month) white cotton 30 103 colored or ecologic cotton 9 103 86 int. j. prod. manag. eng. (2014) 2(2), 85-91 creative commons attribution-noncommercial-noderivatives 4.0 international garcia, s. & de alencar nääs, i. http://creativecommons.org/licenses/by-nc-nd/4.0/ dyes and the most commonly found in the industries visited for their study. the author indicates that products, which are derived from sulfur dyes, exhibit more contaminated effluents, and presents the most chemicals used in textile production (table 3). a projection for the use of mesh chemicals for producing approximately 1000 tons of textile/month is projected in table 3. representing 25% of the brazilian industrial sector, the textile segment of the state of santa catarina is one which somehow neglects the impact this activity may cause to the environment, producing excessive amounts of toxic effluents without treatment (sauer, 2002). estimation for the use of mesh chemicals for producing approximately 1000 tons of textile/month is projected showing a significant number of effluents, agreeing with galindo et al. (2001) and herrmann et al. (2011) and representing a significant impact on environmental contamination and aquatic life. the authors state that about 1% to 15% of the dyes used by the textile industries are lost in the dyeing process and, in overall, released in effluents, these findings also agree with other authors (herrmann et al., 2011), and table 4 shows data related to this issue. these results on the reduction of polluted wastewater and water consumption would be an attractive argument for improving production of colored and agroecological cotton, or investing more in research and development of this product. however, the government’s investment in new technologies is primarily directed to large agricultural enterprises, leaving the small farmers with little if none support. table 4 shows the use of water and waste dumped in nature in the past ten years, during the production of cotton. the production of cotton designed with the same consumption for the next 10 years is shown in table 5. data on table 6 show cotton production with a reduction of 5%, which would be a possibility of increasing production of colored cotton, reducing water usage, and also projecting 5% reduction in waste over the next ten years. in table 6, it can be observed the development of the next ten years, from one production of white cotton 5% lower than expected, which could be transformed table 3. basic consumption of chemical products used in usual textile industry and the effluent which pollute the environment, compared to naturally colored cotton (source: adapted from galindo et al. (2001) and martins (1997)). chemical products basic consumption (t/month) basic consumption (t/month/year) assumption (av. of 10% t/month) effluents (av. of 10% t/month/year) salt 60 720 103 12 144 103 peroxide 8 96 103 1,6 19 103 kelp 15 180 103 3 36 103 acetic acid 3 36 103 0,6 7 103 other acids 60 720 103 12 144 103 reactive dyes 3.2 38 103 0,64 7,680 sulfur dyes 6.9 82 103 1,38 16,560 total 312.2 t/month 1,873 103 31.22 t/month 374,640 table 4. total of production, water consumption and chemical effluents in ten year ago with cotton production (source: iemi (2012); mapa (2012)). year production 1000 tons h2o consumption 1000 tons chemical effluents 1000 tons 2000/2001 1.511 543.960 106 566.081.106 2001/2002 1.245 448.200 106 466.426. 106 2002/2003 1.365 491.400 106 511.383. 106 2003/2004 2.099 755.640 106 786.369. 106 2004/2005 2.129 766.440 106 797.608. 106 2005/2006 1.038 373.680 106 388.876. 106 2006/2007 1.524 548.640 106 570.951. 106 2007/2008 1.602 576.720 106 600.173. 106 2008/2009 1.411 507.960 106 528.617. 106 2009/2010 1.194 429.840 106 447.320. 106 2010/2011 2.052 738.720 106 768.761. 106 total 17.170 6.181.200 106 6.432.568. 106 87int. j. prod. manag. eng. (2014) 2(2), 85-91creative commons attribution-noncommercial-noderivatives 4.0 international textile industry can be less pollutant: introducing naturally colored cotton http://creativecommons.org/licenses/by-nc-nd/4.0/ in cotton production and organic colored. it can also be noted the result in the reduction of water use of 70% (for these 5% less), and also assume a drop of 5% per year in the fall of effluents dumped in nature in the same period (mapa, 2012). the intensive use of the raw material for textile industries (damiano, 2003), and the increasing interest by consumers in the use of green products (demir et al., 2010), has led to the augment in the research on colored fibers. the focus has been mainly to reduce the use of dyes, as it is a carcinogenic material (ramalho et al., 2010). however, the colors of the fibers influence the technological nature of the final product (carvalho and santos, 2003). according to pan et al. (2010 ) genotypes, with colored fibers produce a higher amount of wax, which has a negative effect on the production of cellulose, reducing the quality of the fibers produced. for these reasons, search the cotton fiber color shifts to other regions not only in northeast brazil and embrapa cotton led researchers to paraná, because of the recent price increases of the product on site (bellettini, 2011). the researchers sought to examine technological features of colored cotton with the potential to be grown in the northern state of paraná, and according to the methodologies used, the colored cultivars obtained inferior in length (uniformity index and short fiber length and strength). a study of carvalho et al. (2005) also shows that the average colored cotton is 10% lower than white, as well as the characteristics of its fibers, are inferior. the author showed that genetic selection for stronger color intensity may results in negative effects on the fiber. limited results on this subject are due to lack of studies using the colored cottons, which has been subjected to intense research and development as it occurred with the white fiber. table 7 presents the chronological steps towards the development of the colored cotton in brazil. there are cultivars being planted in the northeastern of brazil, in the state of paraiba, in small family farms. figure 1 shows various kinds of colors cotton fibers. fashion has been innovative in the use of synthetic fibers (composed by artificial, natural and mixed), but these innovations do not necessarily translate into environmental concerns. they only combine changes in relation to the style, the design, the artificial tissues, adding unique features for use table 5. projection, water consumption and chemical effluents in next ten years (source: adapted iemi (2012), mapa (2012)). year projection 1000 tons h2o consumption 1000 tons chemical effluents 1000 tons 2011/2012 2.155 775.800 106 807.349 106 2012/2013 1.563 562.680 106 585.562 106 2013/2014 1.543 555.480 106 578.069 106 2014/2015 2.309 831.240 106 865.043 106 2015/2016 2.504 901.440 106 938.098 106 2016/2017 1.912 688.320 106 716.311 106 2017/2018 1.892 681.120 106 708.818 106 2018/2019 2.658 956.880 106 995.793 106 2019/2020 2.853 1.027.080 106 1.068.847 106 2020/2021 2.261 813.960 106 847.061 106 2021/2022 2.241 806.760 106 839.568 106 total 23.891 8.600.760 106 8.950.524 106 table 6. next 10 years with production 5 percent of colored and organic cotton and suppose reduction in use of water and 5% reduce year per year in chemical effluents (source: adapted iemi (2012), mapa (2012)). year projection 1000 tons consumption h20 reduce 70% chemical effluents (mil/t) reduction of 5% per year 2011/2012 2.047 736.918 103 766.888 106 2012/2013 1.484 534.238 103 555.965 106 2013/2014 1.465 527.398 103 548.847 106 2014/2015 2.193 789.478 103 821.585 106 2015/2016 2.378 856.078 103 890.893 106 2016/2017 1.816 653.758 103 680.346 106 2017/2018 1.797 646.918 103 673.228 106 2018/2019 2.525 908.998 103 945.966 106 2019/2020 2.710 975.598 103 1.015.274 106 2020/2021 2.147 772.918 103 804.352 106 2021/2022 2.128 766.078 103 797.233 106 total 22.690 8.168.384 103 8.500.581 106 table 7. timeline of the development of colored fibers from 2001 to 2010 (source: adapted from beltrão e carvalho (2004), embrapa algodão (2012)). year event 2000 development of the variety brs 200 marron 2001 fiber color begins commercial scale in paraíba state by small farmers 2001 fiber color reaches 30 to 40% higher price per pound relative to white fiber in 2002 2002 development of the variety brs verde 2002 cultivation of organic fiber begins (without chemicals or fertilizers) 2002 fiber color reaches 200% higher price per pound relative to white fiber 2005 development of the variety brs rubi 2010 development of the variety brs topázio 88 int. j. prod. manag. eng. (2014) 2(2), 85-91 creative commons attribution-noncommercial-noderivatives 4.0 international garcia, s. & de alencar nääs, i. http://creativecommons.org/licenses/by-nc-nd/4.0/ by health, sports and leisure, being supported by the chemical industry, electronics and even by nanotechnology. one feature that has been noted is the increasing use of chemical fibers at the expense of natural fibers, which have as a consequence more pollutants in industrial processes. the chemical fiber represented 62% of total consumption in 2006 to 39% in 1970, 44% in 1980 and 48% in 1990 (garcia, 2009; mariano, 2011). a study of socially responsible consumption of clothing by norum and ha-brookshire (2011) examined the effect of fiber origin, method of production and the price in consumer preference for cotton clothing in the usa. results showed that the price appeared as the most important criteria for the purchase of cotton products (58.5%), transparency (30%) and fibers grown with sustainable methods were cited by only 11.5% of survey 3.2. sustainability on fashion/apparel supply chain sustainability concepts have been discussed worldwide in several areas of knowledge and industrial production (baskaran et al., 2012; gri, 2010; mariano, 2011, wced, 1987). the annual global sales of organic cotton products, for example, grew by over 40% between 2001 and 2009 (ota, 2009). the agroecologic cotton is produced in sustainable systems, with proper management and protection of natural resources, without the use of pesticides, genetically modified organisms, chemical fertilizers or other inputs harmful to human health, animal and the environment (beltrão et al., 2009, cartaxo et al., 2008). the cotton prices are defined internationally by the yarn quality which are related to the plant fibers, their length and reflectance. produced under irrigation, cultivars with medium length fiber, cotton is not naturally white, and that makes the fiber is white and chemical methods are highly polluting, with bleaching products. if not for the action of man, there would be this level of whiteness in cotton yarn (embrapa algodão, 2012). there is a mutation of the cotton plant which makes possible the production of colored cotton. this production avoids dyeing process and saves water. the procedure was developed by embrapa cotton, brazilian government agency which is specialized in technology transfer, and support family farming. being organic or with conventional management grown cotton plants (not genetically modified) are certified to be produced without the use of synthetic chemicals, the cultivation of cotton colored fiber in the northeast of brazil through family farming has developed in former years. it also modified the behavior of farmers, seeking more efficient ways of producing with reduced waste and avoiding chemicals. the farmers also get a good price on colored cotton, when compared to white fiber (carvalho et al., 2011). thus, these products have attracted the attention of companies who care about environmental problems, valuing the clothing business and adding more value to the finished product (refosco et al., 2008). the yarns produced from naturally colored fiber, undergoes fewer chemical processes, and it does not pollute the environment, besides representing a decrease of about 70% in the use of water, in the finishing process of the fabric. currently, the social point of view, this production presents itself as a source of income for about a thousand farmers from the states of paraíba, pernambuco, rio grande do norte and ceará. no transgenic techniques are figure 1. colorful cotton fibers produced from the manual choice of seedlings. the colors vary according to the planting site (source: embrapa algodão (2012)). 89int. j. prod. manag. eng. (2014) 2(2), 85-91creative commons attribution-noncommercial-noderivatives 4.0 international textile industry can be less pollutant: introducing naturally colored cotton http://creativecommons.org/licenses/by-nc-nd/4.0/ used in this development (embrapa algodão, 2012). a study by embrapa cotton, which analyzes the performance of commercial brs brown, a research of over 15 years in the field and the laboratory, showed that the technological characteristics of fiber and yarn are produced to meet the processing improvement (organic cotton, 2008; refosco et al., 2008; mariano, 2011). 4. final remarks apparently 5% reduction in the production of white cotton, reverted to the production of organic cotton and colored, seems to represent little economy in every way, even to the general development of the country. instead, a production of only 5% organic cotton and colorful world, could represent a drive revenue in small areas of family farming, spinning craft, producing handmade clothes and many other possibilities as objects of decoration and fashion accessories. furthermore, the water savings produced by this 5% of color cotton production and organic represent at least 70% total water used. assuming that the development of the chemical industry would provide a 5% reduction in effluent thrown in nature in standard process, which does not exist in the production of colored cotton and organic, the result would be even more positive as the results of this study. getting into the era of sustainability has been a requirement for large companies that want to stay in business. the agroecological production of cotton is the embryo of an idea that could have an effect on this business network that moves millions worldwide. it still could provide brazil the pioneer in this area, and improve the income of small and medium farmers, processing companies and clothing making, the entire chain of fashion. the general awareness of people cause this change and new deeper studies on economic feasibility of production on a larger scale and colorful organic cotton, which is not yet available, one will be a pressing need to be addressed in future work on the topic. acknowledgements the authors wish to thank capes, for the support of this study. references abit. associação brasileira da indústria têxtil. (2012). available at www.abit.org.br. accessed on april 2012. abreu, m.c.s. de, castro, f. de, soares, f. de a., silva filho, j.c.l. da. (2012). a comparative understanding of corporate social responsibility of textile firms in brazil and china. journal of cleaner production, 20(1): 119-126. doi:10.1016/j.jclepro.2011.08.010 baskaran, v., nachiappan, s., rahman, s. (2012). indian textile suppliers’ sustainability evaluation using the grey approach. international journal of production economics, 135(2): 647-658. doi:10.1016/j.ijpe.2011.06.012. bellettini, s., santos, k.h. dos, bellettini, n.m.t., miglioranza, é., chimbo junior, a., silva, d.p. da, carvalho, f.k., nagashima, g.t. (2011). características tecnológicas de fibra de algodão colorido no paraná. 8º congresso brasileiro de algodão & i cotton expo 2011, são paulo, 2011. pp. 704. evolução da cadeia para construção de um setor forte: anais. campina grande, pb: embrapa algodão, 2011. pp. 701-705. (cd-rom). available at http://ainfo.cnptia.embrapa.br/digital/bitstream/item/51240/1/fit-069poster.225.pdf. accessed on april 2013. beltrão, n.e.m., carvalho, l.p. (2004). algodão colorido no brasil, e em particular no nordeste e no estado da paraíba. campina grande, pb. issn 0102-0205. agosto, 2004. available at http://ainfo.cnptia.embrapa.br/digital/bitstream/cnpa/16943/1/doc128.pdf. accessed on april 2012. beltrão, n.e.m., silva, c.a.d., suinaga, f.a., arriel, n.h.c., ramalho, f.s. (2009). algodão agroecológico: opção de agronegócio para o semiárido do brasil. campina grande: embrapa algodão. cartaxo, w.v., guimarães, f.m., soares, j.j., beltrão, n.e.m. (2008). potencialidades da produção de algodão pela agricultura familiar do nordeste. embrapa. centro nacional de pesquisa de algodão. documentos 202. campina grande, pb. carvalho, l.p., andrade, f.p., silva filho, j.l. (2011). cultivares de algodão colorido no brasil. rev. bras. ol. fibros., 15(1): 37-44. available at http://www.alice.cnptia.embrapa.br/bitstream/doc/910145/1/488rbof15127362011.pdf. accessed on july 2014. carvalho, l.p., barroso, p.a.v., santos, j.a.t., alves, h.s. (2005). seleção massal e porcentagem de fibra em cultivar de algodoeiro colorido. pesquisa agropecuária brasileira, 40(9): 895-898. available at http://www.scielo.br/pdf/pab/v40n9/a09v40n9.pdf. accessed on april 2013. carvalho, l.p., santos, j.w. (2003). respostas correlacionadas do algodoeiro com a seleção para a coloração de fibras. pesquisa agropecuária brasileira, 38: 79-83. available at http://www.scielo.br/scielo.php?script=sci_arttext&pid=s0100-204x2003000100011&lng=en. accessed on april 2013. 90 int. j. prod. manag. eng. (2014) 2(2), 85-91 creative commons attribution-noncommercial-noderivatives 4.0 international garcia, s. & de alencar nääs, i. http://www.abit.org.br http://dx.doi.org/10.1016/j.jclepro.2011.08.010 http://dx.doi.org/10.1016/j.ijpe.2011.06.012 http://ainfo.cnptia.embrapa.br/digital/bitstream/item/51240/1/fit-069poster.225.pdf http://ainfo.cnptia.embrapa.br/digital/bitstream/cnpa/16943/1/doc128.pdf http://www.alice.cnptia.embrapa.br/bitstream/doc/910145/1/488rbof15127362011.pdf http://www.scielo.br/pdf/pab/v40n9/a09v40n9.pdf http://www.scielo.br/scielo.php?script=sci_arttext&pid=s0100-204x2003000100011&lng=en http://creativecommons.org/licenses/by-nc-nd/4.0/ damiano, v.b. (2003). as indústrias de materiais lignocelulósicos em busca de tecnologias amigas da natureza. revista unorp, 3(2): 49-61. demir, a., ozdogan, e., osdil, n., gurel, a. (2010). ecological materials and methods in the textile industry: atmospheric-plasma treatments of naturally colored cotton. journal of applied polymer science, 119(3): 1410-1416. doi:10.1002/app.32575 embrapa algodão. (2012). informações diversas sobre algodão e algodão colorido. available at http://www.cnpa.embrapa.br/. accessed on april 2012. forgiarini, e. (2006). degradação de corantes e efluentes têxteis pelaenzima horseradish peroxidase (hrp). dissertação de mestrado do programa de pós–graduação em engenharia química do centro tecnológico, da universidade federal de santa catarina. florianópolis – santa catarina. galindo, c., jacques, p., kalt, a. (2001). photooxidação of the phenylazonaphtol ao20  on tio 2 : kinetic and mechanistic investigations. chemosphere, 45(6-7): 997-1005. http://dx.doi.org/10.1016/s0045-6535(01)00118-7 garcia, r, oliveira, a, madeira, p. (2009). bndes. banco nacional de desenvolvimento social. projeto pib perspectivas do investimento no brasil. documento setorial: textil, vestuário e calçados. available at http://www3.eco.unicamp.br/neit/images/stories/arquivos/06_ds_ benssalario_textil_vestuario_calcados.pdf. accessed on july 2014.. gri. global reporting initiative. (2010). g3 sustainable reporting guidelines. available at http://www.globalreporting.org/. accessed on april 2012. herrmann, j.m., vautier, m., guillard, c. (2001). photocatalytic degradation of dyes in water: case study of indigo and of indigo carmine. journal of catalysis. 201(1): 46-59. doi:10.1006/jcat.2001.3232 iemi. instituto de estudos de marketing industrial. (2012). brasil têxtil 2012. available at http://www.iemi.com.br/biblioteca/textil/brasiltextil-2011/. accessed on march 2013. mapa. ministério da agricultura, pecuária e abastecimento. (2012). available at http://www.agricultura.gov.br/vegetal/culturas/algodao. accessed on april 2013. mariano, m. (2011). brasil é hoje o único player têxtil mundial fora da ásia. relatório brasil têxtil. apoio do texbrasil programa de exportação da indústria da moda brasileira desenvolvido com a apex-brasil. relatório setorial anual produzido pelo instituto de estudos e marketing industrial (iemi), associação brasileira da indústria têxtil e de confecção (abit). textilia. 18/10/2011. available at http:// www.textilia.net/materias/ler/textil/conjuntura/brasil-_e_hoje_o_unico_player_textil_mundial_fora_da_asia. accessed on june 2012. martins, g.b.h. (1997). práticas limpas aplicadas às indústrias têxteis de santa catarina. dissertação de mestrado em engenharia de produção da universidade federal de santa catarina. available at https://repositorio.ufsc.br/handle/123456789/77280. accessed on july 2014. mendes, f.d. (2010). um estudo comparativo entre as manufaturas do vestuário de moda do brasil e da índia. tese de doutorado. universidade paulista. são paulo: 2010. norum, p.s., ha-brookshire, j.e. (2011). consumer trade-off analysis and market share estimation for selected socially responsible product attributes for cotton apparel. clothing and textiles research journal, 29(4): 348-362. http://dx.doi.org/10.1177/0887302x11425956 organic cotton farm and fiber report. (2008). organic exchange. ota. organic trade association. (2009). associação do comércio orgânico. pan, n.z., sun, d., sun, j., zhou, z., jia, y., pang, b., ma, z., du, x. (2010). effects of fiber wax and cellulose content on colored cotton fiber quality. euphytica, 173(2): 141-149. doi:10.1007/s10681-010-0124-0 ramalho, f.s., azeredo, t.l., fernandes, f.s., nascimento júnior, j.l., malaquias, j.b., nascimento, a.r.b., silva, c.a.d., zanuncio, j.c. (2010). food intake and utilization of alabama argillacea (hübner) (lepidoptera: noctuidae) fed on cotton cultivars with colored fibers. journal of pest science, 84(2): 199-205. doi:10.1007/s10340-010-0341-2 refosco, e., mazzotti, k., sotoriva, m., broega, a.c. (2011). o novo consumidor de moda e a sustentabilidade. vii colóquio de moda. universidade do minho. portugal. available at http://hdl.handle.net/1822/14946. accessed on april 2013. sauer, t. (2002). degradação fotocatalítica de corante e efluente têxtil. dissertação apresentada ao curso de pós-graduação em engenharia química do centro tecnológico da universidade federal de santa catarina. florianópolis, sc. available at https://repositorio.ufsc.br/ handle/123456789/83511. accessed on july 2014.. wced. (1987). our common future. oxford: oxford university press, 1987. 91int. j. prod. manag. eng. (2014) 2(2), 85-91creative commons attribution-noncommercial-noderivatives 4.0 international textile industry can be less pollutant: introducing naturally colored cotton http://dx.doi.org/10.1002/app.32575 http://www.cnpa.embrapa.br http://dx.doi.org/10.1016/s0045-6535(01)00118-7 http://www3.eco.unicamp.br/neit/images/stories/arquivos/06_ds_benssalario_textil_vestuario_calcados.pdf http://www3.eco.unicamp.br/neit/images/stories/arquivos/06_ds_benssalario_textil_vestuario_calcados.pdf http://www.globalreporting.org http://dx.doi.org/10.1006/jcat.2001.3232 http://www.iemi.com.br/biblioteca/textil/brasil-textil-2011 http://www.iemi.com.br/biblioteca/textil/brasil-textil-2011 http://www.agricultura.gov.br/vegetal/culturas/algodao http://www.textilia.net/materias/ler/textil/conjuntura/brasil-_e_hoje_o_unico_player_textil_mundial_fora_da_asia http://www.textilia.net/materias/ler/textil/conjuntura/brasil-_e_hoje_o_unico_player_textil_mundial_fora_da_asia https://repositorio.ufsc.br/handle/123456789/77280 http://dx.doi.org/10.1177/0887302x11425956 http://dx.doi.org/10.1007/s10681-010-0124-0 http://dx.doi.org/10.1007/s10340-010-0341-2 http://hdl.handle.net/1822/14946 https://repositorio.ufsc.br/handle/123456789/83511 https://repositorio.ufsc.br/handle/123456789/83511 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j international journal of production management and engineering institutions for collaboration in industrial clusters: proposal of a performance and change management model carpinetti, luiz c. r., & lima, rafael h. p. department of production engineering – university of sao paulo (usp). av. trabalhador são-carlense, 400, são carlos-sp – brazil. carpinet@sc.usp.br rhlima@sc.usp.br abstract: this paper proposes a performance and change management model for institutions for collaboration (ifcs) in industrial clusters to assist them while planning, conducting and evaluating joint actions. a three-stage implementation scheme and a self-assessment tool that helps an ifc determine its compliance with the proposed model are also introduced. the self-assessment tool was applied in three brazilian ifcs from different clusters. it was found that the ifcs researched face major difficulties in designing and implementing performance measures to evaluate the results and impacts of joint actions. however, ifcs have been successful in identifying local infrastructure and devising informal strategic plans. key words: industrial clusters, performance management, collaborative networks, institutions for collaboration. 1. introduction geographic agglomerations of firms have been widely studied in the literature and are often referred to as industrial clusters. the topic has become top priority in the agenda of many regional development agencies and national governments, especially due to the competitive advantages and economic development they enable to firms and to the regions they are located (porter, 1998; mytelka and farinelli, 2000; sölvell et al., 2003). companies located in these regions can benefit greatly from external economies, collaboration, and exchange of knowledge between organizations (unido, 2001; karaev et al., 2007; capo-vicedo et al., 2008). such benefits may be extended if firms act together through joint actions that stimulate them to collaborate for the common good (bititci et al., 2004). however, some authors argue that the execution of joint actions requires some sort of local coordination, whose role is to intermediate the interests of companies and coordinate the execution of activities related to the joint action (schmitz and nadvi, 1999; sölvell et al, 2003; gerolamo et al., 2008). the definitions found in the literature for industrial clusters often emphasize the importance of related institutions that do not participate directly in the supply chain, but are fundamental for the cluster activities (porter, 1998). examples of such institutions are universities, research centres, training centres and specialized service providers. karaev et al. (2007) highlights the existence of local supporting institutions that are oriented to the particular needs of the cluster participants. in brazil it is common to observe local associations that promote initiatives that seek to satisfy some needs common to a subset of the local firms. some authors have reported on cases of such institutions, such as schmitz (1998), sölvell et al. (2003) and gerolamo et al. (2008). the term institution for collaboration (ifc) is used in this paper to refer to such supporting associations. this paper studies how ifcs manage joint actions in industrial clusters. for this purpose, a performance and change management model is proposed to guide such institutions in the conduction of joint actions. the model is strongly influenced by the pdca cycle and continuous improvement principles. the model outlines a series of performance and change management practices to assist ifcs in the planning, implementation and assessment of joint actions. therefore, the model stresses the importance of performance measurement systems as a way of demonstrating the benefits gained through joint http://dx.doi.org/10.4995/ijpme.2013.1502 received: 2013-04-29 accepted: 2013-05-20 https://ojs.upv.es/index.php/ijpme 13int. j. prod. manag. eng. (2013) 1(1), 13-26creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2013.1502 https://ojs.upv.es/index.php/ijpme actions. later on the model is organized in three implementation stages and a self-assessment tool is devised to evaluate the institution’s management practices. the self-assessment tool was used to evaluate the management practices of three brazilian ifcs from different industrial clusters – sertaozinho (metal-mechanic cluster), arapongas (furniture cluster) and londrina (information and communication technology cluster). the article is organized as follows: section 2 describes the research method and section 3 presents the theoretical background used to support the management model. section 4 introduces the performance and change management model, the implementation stages and the self-assessment tool. section 5 describes the application of the selfassessment tool in the three ifcs, whilst section 6 discusses the data collected. finally, section 7 concludes the paper by summarizing the findings and suggesting topics for future research. 2. research method figure 1 shows the procedure adopted in this research. given the objective of this paper, it was deemed necessary to review the literature on industrial clusters, performance management in industrial clusters, and institutions for collaboration. the knowledge gained during this activity served as background for the proposition of the performance and change management model. after that, three implementation stages and a self-assessment tool were derived from the management model. this tool aims to quantify the extent to which an ifc complies with the model, that is, it points what management practices outlined in the model the institution has or has not implemented. the self-assessment tool, which is described in detail in section 4, consists of 18 requirements derived from the management model. a supporting question was associated to each of the requirements, so that the person conducting the assessment understands what should be assessed in a given requirement. during the self-assessment, a score ranging from 0 to 10 is associated to each requirement indicating the level of compliance of the ifc with the management model. the scores obtained are then grouped according to the implementation stages to which they are associated to determine the areas that need improvement with respect to the institution’s management practices. the items listed in the theoretical contribution served as guidance during the self-assessment carried out in the three cases, which were based on interviews with the coordinators of each institution. one of the authors acted as a facilitator during the application of the self-assessment tool and in the determination of the scores to each requirement. the answers given by the coordinator were transcribed and qualitatively evaluated to determine such scores. the scores were then arranged in a table, in such a way to enable a cross-case analysis. the findings were based upon the comparison of similarities and differences between the cases. 3. literature review 3.1. industrial clusters the term industrial cluster was made popular in the late nineties by porter (1998), who defined it as geographic concentrations of interconnected companies and institutions in a particular field. his definition also encompasses other entities that are important to competition, such as suppliers of specialized inputs, service providers, specialized infrastructure, and governmental and private institutions as universities, training providers or trade associations. in brazil the government, funding institutions and some research centres refer to industrial clusters by the acronym apl, which stands for local productive arrangements, as defined 14 int. j. prod. manag. eng. (2013) 1(1), 13-26 carpinetti, luiz c.r., & lima, rafael h.p. creative commons attribution-noncommercial 3.0 spain figure 1. research method. by the brazilian ministry for development, industry and foreign trade (mdic, 2010). the fundamentals that seek to explain the competitive advantages of industrial clusters were set out in the 19th century by alfred marshal, who noted that geographical agglomerations of firms might ultimately result in three types of externalities – pool of specialized workers, specialized providers of inputs and services, and the technological spillovers that flow easily among co-located firms (krugman, 1991; plummer and taylor, 2001; cortright, 2006). these externalities are called by marshal as external economies. schmitz and nadvi (1999) added that, aside of the unplanned benefits of external economies, companies in industrial clusters may gain competitive advantage through planned joint actions, which are enabled by cooperation and collaboration among firms. two topics are often associated with collaboration in industrial clusters: social capital: refers to the set of intangible factors that exist in a community, such as values, norms, attitudes, trust and networks that facilitate coordination and collaboration for the common good (cohen and prusak, 2001); collective efficiency: competitive advantage gained through external economies and joint actions (schmitz, 1995). some authors contend that concentrations of firms foster network formation, since companies can take advantage of the proximity to strengthen the bonds with local firms and establish new partnerships (camarinha-matos and affsarmanesh, 2005). there is some empirical evidence in the literature linking social networks and the knowledge exchange among co-located firms, which in turn may facilitate innovation among companies (morosini, 2004; steiner and hartmann, 2006). for example, baptista and swann (1998) found that the concentration of specialized workers of a given sector facilitate knowledge spillovers, which in turn may lead to increased levels of innovation. steinle and schiele (2002) argue that companies must interact with each other in order to benefit from locating within a cluster, which in turn requires a climate that stimulates cooperation and intensive knowledge exchange. 3.2. performance management in industrial clusters performance management has been a central topic in organizational and operations management (neely, 2005). this has led many authors to develop frameworks that assist companies in designing their performance measurement systems (pms) (neely, 1995; 2005; kaplan, 1996; kennerley, 2002; radnor and barnes, 2007). as pointed out by neely (1998), a pms can be used to four purposes: check current position, communicate performance, confirm priorities and compel progress. this reinforce the role of performance management in strategic issues, such as setting priorities, targets and deploying strategies by cascading down actions that will ultimately make the company achieve its planned objectives. due to the apparent importance of performance management to individual organizations, many authors have tried to apply this theory to measure multi-firm relationships as supply chains (chow et al., 1994), organizational networks (camarinha-matos and affsarmanesh, 2007) and industrial clusters (carpinetti et al., 2008). indeed, performance management in industrial clusters has drawn considerable attention from several authors and has been viewed as a basis for the management of a cluster (sölvell et al. 2003; dti, 2005; gerolamo et al., 2008; carpinetti et al., 2008). furthermore, the use of numerical means to demonstrate the benefits of collaboration in organizational networks would motivate companies to collaborate more and establish new partnerships (camarinha-matos and affsarmanesh, 2007). according to a report written to the english department of trade and industry, measuring a cluster’s performance may be useful to evaluate the appropriateness, effectiveness and efficiency of interventions (dti, 2005). collaboration, on its turn, can be viewed as a metric composed of three measurable dimensions: information sharing, decision synchronization and incentive alignment (simatupang and sridharan, 2005). it is thus clear that, for companies to collaborate with each other in joint initiatives, it becomes necessary a coordination mechanism that balances the interests and serves as a communication channel among the parties involved. this means communicating the status of ongoing activities and the impacts of joint initiatives, which should be done by means of performance measures. there does not seem to be a sound approach or framework in the literature that fulfils the need of a performance management model to facilitate the conduction and assessment of collaborative initiatives in industrial clusters. there are though some contributions that try to fill this gap. sölvell et al. (2003) developed the cluster initiative performance model (cipm), in which the performance of a cluster initiative is measured in terms of innovation, international competitiveness, cluster growth, and achievement of goals. according 15int. j. prod. manag. eng. (2013) 1(1), 13-26 institutions for collaboration in industrial clusters: proposal of a performance and change management model creative commons attribution-noncommercial 3.0 spain to the authors, cluster’s performance is ultimately affected by three elements: the social, political and economic setting within the nation; the objectives of the cluster initiative; and the processes by which the cluster initiative develops. carpinetti et al. (2008) proposed a framework to design performance measurement systems for industrial clusters. the authors divided performance measures into four perspectives: economic and social results; company’s performance; collective efficiency; and social capital. a similar contribution was made by bortolotti and amato neto (2006), who developed a 6-dimension framework to characterize and evaluate industrial clusters. the six dimensions included in the framework were related to geographic, economic, institutional, social, technological and environmental aspects of the industrial cluster as a whole. the european commission carried out seven case studies in networks from the automotive sector and proposed a five-step method to the management of networks (ec, 2001): step 1 – goal, potential and strategy: consists of drawing together the key stakeholders of the network; step 2 – starting the network: setting of rules between partners and preparation of the operational background; step 3 – implementation of the network: establishment of an appropriate information and communication platform to connect all network participants; step 4 – management of the network: management of the network activities by focusing on information and communication, training, innovative projects, internationalization, and marketing and advertising; step 5 – evaluation of the network: consists of auditing the network actions and gathering feedback from network members to continuously improve the management of the network; an analogous contribution was made by gerolamo et al. (2008), who developed a performance management model for industrial clusters and cooperation networks. their model is divided into five steps: step 1 – identification of the stakeholders: the first step is to identify the stakeholders directly or indirectly related to the cluster activities (large enterprises, smes, local partners, local associations, the local chamber of commerce and industry, universities, public authorities, financial institutions and so forth; step 2 – strategic orientation and definition of objectives: formulation of a strategic plan that balances the interests of local companies as well as other interested parties, such as local authorities and the community. step 3 – implementation of improvement and innovation projects: based on the elements formulated in steps 1 and 2, a series of joint initiatives should be developed to take advantage each partner’s capabilities as well as to increase trust among companies; step 4 – performance evaluation and measurement: definition of a process to measure and evaluate the performance and impact of joint actions; step 5 – supporting infrastructure for the management process: establishment of the infrastructure necessary to support cooperation projects, such as a regional office or a regional development agency. 3.3. institutions for collaboration besides the external economies that naturally emerge in agglomerations, companies my benefit greatly from local supporting institutions oriented towards satisfying specific needs of the cluster participants (karaev et al., 2007). as pointed by seliger et al. (2008), such institutions are vital for the coordination of joint actions and diffusion of specific knowledge inside the cluster. schmitz and nadvi (1999) contend that local agencies should mediate conflicts of interest that may arise between companies within the cluster. the existence of local coordination may differentiate a mere agglomeration of companies from a comprehensive local innovation system that aims to improve local competitiveness through joint actions and network formation (gerolamo et al., 2008). several terms can be found in the literature to refer to these supporting institutions, such as cluster initiatives (sölvell et al., 2003; ketels and sölvell, 2006), institutions for collaboration (sölvell et al., 2003; 2008), industry associations (itd, 2009), regional development agencies (seliger et al., 2008) or institutional thickening (andriani et al., 2005). in this paper we refer to local supporting institutions in clusters as institutions for collaboration (ifc). these institutions may assume a variety of forms, such as private organizations, public agencies or industry associations. the literature reports a number of cases concerning institutions for collaboration (see some examples in schmitz, 1998; sölvell et al., 2003; gerolamo et al., 2008). it seems that cultural 16 int. j. prod. manag. eng. (2013) 1(1), 13-26 carpinetti, luiz c.r., & lima, rafael h.p. creative commons attribution-noncommercial 3.0 spain issues, the type of business, the economic setting, and the existing social capital may influence the form of local ifcs. in spite of that, it seems to be consensus in the literature that regardless of the way the ifc comes about in a cluster, it plays a vital role in managing interactions, sharing of knowledge and in providing a cognitive framework for transforming information into useful knowledge (audretsch and lehmann, 2006; steiner and ploder, 2008). as to the emergence of these institutions, sölvell et al. (2003) argue that after a cluster’s formation, the region tends to accumulate resources and commitment of its participants, which may culminate in the establishment of local ifcs. the authors describe ifcs as formal institutions maintained with fees paid by local firms that seek to balance the interests of the different actors involved with the industrial cluster. ifcs may act upon several issues by coordinating joint actions concerning quality of life, education, infrastructure (transportation, energy, and communication), tax regulation, export strategy, quality standards, research and training, and so forth (sölvell, 2008; itd, 2009). 4. the performance and change management model the model described in this section aims to help ifcs in industrial clusters in the planning, implementation and evaluation of joint actions. one such institution may encompass representatives from the various cluster actors, such as the companies, the government, research centres, universities and so forth. figure 2 shows the proposed model, which is divided into three dimensions – infrastructure, strategic planning and implementation and assessment. the model emphasizes the implementation of joint actions that seek to improve some aspect of the industrial cluster’s performance. it does not necessarily mean that all companies in the cluster should be involved in every joint action supported by the institution, but rather that each joint action should benefit at least a subset of the local companies. the execution of these actions requires strong coordination, especially due to their collective nature. at this point, the support provided by institutions becomes crucial, since they can serve as mediators between the several parties involved in the joint action. 4.1. infrastructure the model in figure 2 suggests that the ifc should identify the existing infrastructure prior to the definition and execution of joint actions. by identifying the local infrastructure, the institution will be able to put forth joint actions that optimize infrastructure utilization and improve existing facilities. local infrastructure can be analyzed at two major levels – institutional and regional. at the institutional level, the model points to the establishment of a statute to formalize its management hierarchy and the roles assigned to each member. a local office should also be set up with the management tools and information systems necessary for the institution to conduct its activities and manage joint actions. additionally, a coordinator or a coordination team should be chosen to act directly upon the joint actions and innovation programs supported by the institutions. the role of 17int. j. prod. manag. eng. (2013) 1(1), 13-26 institutions for collaboration in industrial clusters: proposal of a performance and change management model creative commons attribution-noncommercial 3.0 spain   figure 2. the three dimensions of the proposed model. this coordination body is to mediate the interests and assignments between the parties involved with a given joint action. the infrastructure elements at the regional level refer to cluster actors and facilities that can contribute to the execution of joint actions. for example, cluster actors such as technical schools, specialized service providers and universities can provide specific knowledge to the coordination team and to the companies involved in a joint action. besides these actors, the institution should identify the local facilities that could be exploited by local companies. examples of such infrastructure elements are roads, railroads, warehouses, intermodal ports, communication lines, power supplies, sources of raw material and so forth. these need to be mapped by the institution so that actions can be taken, both to use and to improve the existing infrastructure. 4.2. strategic planning from the standpoint of individual firms, strategic planning is the process by which leaders formulate their vision of future and develop the procedures and operations to achieve such vision (goodstein et al, 1993). in this sense, strategic planning can be viewed as a tool to help organizations set priorities and allocate resources to achieve them (allison and kaye, 2006). from the perspective of an institution promoting joint actions in industrial clusters, the main strategic objectives differ from the traditional profit and productivity objectives of regular organizations. the strategic objectives of an ifc should be related to the improvement of the various performance dimensions of the industrial cluster. hence, the objective of the second dimension of the model is to motivate the institution to formulate its strategic planning. however, the plurality of actors involved with the cluster raises several challenges to the formulation of short and long term goals that balance the desires of local businesses that often compete with each other. for that reason, joint actions need not necessarily involve all firms, but rather those whose goals match the purpose of the initiative being planned. the following sources can be used to inspire the formulation of joint actions: needs of businesses: the institution may conduct a diagnosis of local businesses needs and devise joint actions based upon the needs common to certain groups of firms; public policies: on the one hand, the institution may search for public policies that favour local economic activities and increase awareness of local firms about them. on the other hand, the institution may act as a representative of companies before the governmental agencies to suggest policies that would incentive the local economic activities; existing infrastructure: the diagnosis conducted in the first dimension of the model can reveal opportunities with respect to the use and improvement of the local infrastructure. moreover, joint actions may combine the skills of a subset of the actors in the cluster to promote innovation among businesses. in order to formulate the strategic planning, the model also suggests that the institution should characterize the local productive system and identify ongoing joint actions, so that their goals can be reassessed for the next management cycle. with all this information in hand, the institution will be able to determine more appropriate short and long term goals and set out the joint actions to help achieve these goals. later on, these actions will have to be deployed to all the parties involved to determine their roles and activities. the institution can refer to the hoshin kanri technique to this end (akao, 2004). finally, three additional aspects should be taken into account when formulating the institution’s strategic planning: the sources of funding for implementing joint actions; the means by which the results of actions will be communicated to businesses and other stakeholders; the performance measures that need to be implemented to evaluate the results of actions in numerical terms. 4.3. implementation and assessment the third dimension of the model consists of implementing the items designed in the strategic plan. 18 int. j. prod. manag. eng. (2013) 1(1), 13-26 carpinetti, luiz c.r., & lima, rafael h.p. creative commons attribution-noncommercial 3.0 spain figure 3. the proposed model and the pdca cycle. in other words, the institution should execute the joint actions foreseen in the strategic plan and gather data to calculate performance measures as a means of assessing the impacts of such actions. it becomes now clear that the model is strongly influenced by continuous improvement principles and the pdca (plan, do, check and act) cycle. figure 3 illustrates this by associating the dimensions of the model with the phases of the pdca cycle. as depicted in figure 3, the first two dimensions of the model correspond to the plan phase of the pdca cycle, during which the institution should characterize the local infrastructure, set short and long term objectives, devise the joint actions to be implemented and define the performance measurement system to be used in the remainder of the cycle. the third dimension of the model stretches across the do, check and act phases of the pdca cycle. during the do phase, joint actions should be implemented as planned and data for performance measures should be collected. during the check phase, performance measures and the results of joint actions need to be assessed in order to determine the degree to which the objectives have been achieved. the last phase of the pdca cycle corresponds to the communication of performance and action results, benchmarking with other clusters and the identification of further opportunities for improvement. 4.4. model implementation the implementation of the proposed model will hardly occur at once. instead, it is expected that institutions will develop some of the practices concurrently, regardless of the dimension to which they belong. it is thus pointless to devise a series of steps to implement the model, because each institution will choose different paths to implement it. it is however necessary to understand the dynamics of the model, that is the structure required so that continuous improvement may flow throughout the model. figure 4 illustrates the dynamics of the model by dividing it into four stages, by which the institution should gradually implement and improve its management practices. the dynamics proposed in figure 4 enables the institution to learn from experience and encourages continuous improvement of its planning, execution and assessment capabilities. the stages in figure 4 are associated with the operation of each dimension of the management model. stage zero (s-0) is the initial stage of implementation and corresponds to an ifc that has no formal planning and control capabilities in place to manage joint actions. the first stage (s-1) encompasses the characterization of the regional infrastructure and the establishment of the management tools, information systems and supporting facilities to coordinate joint actions. the second stage (s-2) covers the strategic planning and the implementation and assessment dimensions. it does not regard both dimensions separately, but rather the closed planning, execution and assessment loop, which was depicted in figure 3. thus, when the institution reaches this stage, it will have successfully developed practices to formulate strategic plans, design performance measures, execute joint actions and assess its results and outcomes. stage three (s-3) is achieved when the ifc has put in place a mature management system that fully covers the practices from the three dimensions of the model. at this stage the institution will have learned from experience and improved its managerial capabilities in a way that future joint actions will be better coordinated between the companies, the institution and other cluster actors. additionally, the experience gained after several management cycles may teach the institution how to better choose and formulate joint actions that will ultimately meet the real needs of local companies and actors. these aspects together may increase the success rate of actions and hence increase trust between companies and the institution. 4.5. self-assessment tool as mentioned earlier, it is very unlikely that an ifc will implement the management model at once. moreover, even though institutions may not be aware of the model proposed in this paper, many of them have already implemented managerial practices that satisfy some of the model requirements. it becomes thus necessary a tool to help such institutions evaluate their management practices in relation to those required by the model as a way to determine what areas need improvement. to this end, a set of requirements were devised to characterize each stage of the management model and organized as a 19int. j. prod. manag. eng. (2013) 1(1), 13-26 institutions for collaboration in industrial clusters: proposal of a performance and change management model creative commons attribution-noncommercial 3.0 spain figure 4. dynamics between the dimensions of the model diagnostic tool. tables 1, 2 and 3 list the requirements and questions to evaluate each of the requirements from the stages s-1, s-2 and s-3, respectively. consistent with the stages in figure 4, the questions in table 1 are related to the characterization of the infrastructure at the regional and institutional levels. table 2 puts forth questions to evaluate how the institution plans, executes and assesses joint actions, which is done by verifying the existence of strategic plans, performance measures, communication with stakeholders and benchmarking mechanisms. finally, the questions in table 3 address the effectiveness of the institution’s management practices and the joint actions it has carried out. the requirements and questions from tables 1, 2 and 3 can be used as a self-assessment tool so that institutions can evaluate their management practices and determine to which extent they comply with the management model. this can be helpful in pointing areas for improvement in the management of ifcs. to this end, the institution should assign scores using integral numbers ranging from 0 to 10 to indicate the extent to which the requirement is met. the following reference scale can be used to help determine scores: 0 to 3 points: indicate that the institution has no adherence to the requirement or at best it has plans of meeting the requirement, but no effective results have been achieved yet; 4 to 6 points: the institution has conducted activities that indicate partial compliance with the requirement, that is the activities have been reasonably effective but still can be performed better; 7 to 10 points: the practices being carried out by the institution demonstrate high or total adherence to the requirement. 5. application of the self-assessment tool the tool described in the previous section was used to evaluate the management practices of three brazilian industrial clusters against the proposed model. table 4 lists the three clusters researched and the institutions in which the self-assessment tool was applied. the prevalent economic activity in the city of sertaozinho (c1) is the production of equipment to the ethanol industry. there is in the city an above average concentration of metal-mechanic firms plus a number of companies that provide supporting services, such as automation and maintenance (sebrae, 2007). the self-assessment tool was applied in the apl metaltec, which is an institution supported by the local association of entrepreneurs (ceise) that aims to foster cooperation and improvement among local firms. the acronym apl is commonly used in brazil to refer to industrial clusters. apl metaltec was founded in 2008 and since then it has been promoting joint actions, especially among small and medium-sized firms, to promote continuous improvement and innovation. examples of such initiatives are the free consulting services provided to smes to teach entrepreneurs about best management practices and the creation of a local seal of quality. coordination of joint actions is done by a sebrae (brazilian micro and small business support service) consultant fully devoted to the promotion and management of joint actions. the city of arapongas (c2) is renowned by its high concentration of furniture producers. the cluster covers also the surrounding cities of apucarana, cambe, rolandia and sabaudia, totalling 545 firms and about 12,000 employees (ipardes, 2006a). the furniture industry association of arapongas (sima) started off in 2005 the furniture apl of arapongas as a side project to support and coordinate some joint actions that were being conducted at that time. the 20 int. j. prod. manag. eng. (2013) 1(1), 13-26 carpinetti, luiz c.r., & lima, rafael h.p. creative commons attribution-noncommercial 3.0 spain table 1. requirements for stage 1. requirement question r1 – establishment of the local office does the institution have a local office that allows its operation? r2 – management tools and information systems are there appropriate management tools and information systems in place to assist the institution in its operation and in the coordination of joint actions? r3 – coordination does the institution have a coordinator or a coordination team to manage joint actions and innovation programs? r4 – identification of the local infrastructure has the local infrastructure (facilities, communication, transportation and so forth) been formally identified? r5 – identification of the cluster actors have the actors involved with the cluster been formally identified? coordinator and vice coordinator of the initiatives are local entrepreneurs who dedicate part of their time to the management of the cluster’s joint actions. they operate from within sima by using its infrastructure to promote meetings among companies and seminars about subjects of interest to local firms. among the ongoing joint actions are the annual furniture trade fair, business missions to international fairs as a way of bringing new ideas to local designers, the construction of a quality lab to measure the quality of local products as well as courses to improve local managers’ capabilities. the information and communication technology (ict) cluster found in the city of londrina (c3) was identified by ipardes (2006b) and comprises software developers, automation firms and a range of other ict service providers. according to the coordinator of the ict apl of londrina, the city has approximately 140 ict companies, among which 60 have signed the participation agreement so far. the ict apl of londrina was started off in 2006 as a joint initiative of local entrepreneurs and the state government. joint actions are managed by a coordinator, a vice coordinator and a secretary, who are also company owners in the city. they dedicate part of their time to hold meetings with local companies in order to identify their needs and suggest actions that should be taken to improve competitiveness and performance of local firms. noteworthy joint actions are the identification of 21int. j. prod. manag. eng. (2013) 1(1), 13-26 institutions for collaboration in industrial clusters: proposal of a performance and change management model creative commons attribution-noncommercial 3.0 spain table 2. requirements for stage 2. requirement question r6 – characterization of the local productive system has the institution carried out a diagnosis of the local productive system? r7 – awareness of local companies and other local actors how effective has the work of the institution been towards the awareness of companies and other local actors with respect to collaboration as a means of improvement and innovation? r8 – formulation of the strategic plan does the institution periodically formulate its strategic plan with short and long term objectives that aim to improve the cluster’s performance as a whole? r9 – formulation of joint actions are joint actions derived from the strategic plan and appropriately formulated? (that is with an execution team, determination of responsibilities, associated performance measures, sources of funding and the like) r10 – existence of a pms is there a pms in place that covers all the performance dimensions of the cluster and that enables the institution to assess the impacts of joint actions? r11 – assessment of joint actions does the institution periodically assess the results of joint actions as a way to: (i) determine the level of compliance with predetermined goals, (ii) readjust the plan if necessary or (iii) identify new opportunities for improvement? r12 – performance communication are the results of joint actions and performance measures communicated to all cluster stakeholders? r13 – benchmarking has the institution implemented mechanisms to benchmark its performance measures and practices against those from other industrial clusters? table 3. requirements for stage 3. requirement question r14 – existence of a mature pms is there a stable and mature pms with historical data stored for at least two years? r15 – learning from experience has the institution learned from experience with past joint actions so that the formulation and implementation of new actions that involve local companies and actors is facilitated? r16 – long term initiatives has the institution formulated and conducted long term initiatives that aim to improve local infrastructure, both at the regional and institutional levels? r17 – involvement of small, medium and large firms have the initiatives started off by the institution drawn interest from small, medium and large firms? r18 – impact on performance have the joint actions conducted by the institution been successful in improving the overall cluster’s performance? common training needs to specialize local workforce and the establishment of a local business centre that can be used both for joint purchasing and for selling local products and services to private and public organizations. the self-assessment was conducted with assistance of the researchers, who used the questions from table 1, 2 and 3 to interview the coordinators of each institution. the responses given to each question were transcribed to determine the level of compliance to each of the requirements. table 5 presents the scores obtained after the interviews. the scores in each requirement, as shown in table 5, were grouped to determine the average score in relation to the three implementation stages. these results are shown in table 6. 6. discussion the use of three cases of industrial clusters enabled not only an evaluation of the management practices at the cluster level, but also a cross-case investigation of the practices to establish similarities and differences between them. the line graph in figure 5 shows the scores for each of the 18 requirements in the three clusters researched and the mean score for each requirement. it is visually noticeable in the line graph that the lines for each cluster tend to follow the mean line, which indicates little variation in many of the requirements. there is though great variety between the scores of some other requirements. a better measure to quantify this variation is the column range in table 5. a great range between the scores of a certain requirement indicates that there is significant difference between the management practices adopted in the three cases. the average range observed is 2,67. we will thus consider that a requirement has little variation in the cross-case analysis if its range is lower or equal to 2. great variation in a requirement is characterized by a range equal or greater than 4. because scores were defined in integer numbers, we defined an intermediate classification of variability when the range is equal to 3. by using these criteria, the requirements r5, r6, r9, r10, r11, r12, r13 and r18 showed low variation, whereas the requirements r4, r8, r15, r16 and r17 showed high variation between the cases. requirements r1, r2, r3 and r7 fell in the intermediate group. the observation of the means obtained for each requirement allows the determination of the overall level of compliance with the practices specified in the management model. the scale described in section 4.5 was used to classify and discuss the means observed. however, relying solely on the means or on the ranges may lead to wrong conclusions. for example, if a given requirement showed a low mean, it does not necessarily mean that all the three clusters did not perform the practices specified for that requirement, because there may be high variability between the cases, which is indicated by the range. in order to reach more precise conclusions with respect to each requirement, it is necessary to analyze 22 int. j. prod. manag. eng. (2013) 1(1), 13-26 carpinetti, luiz c.r., & lima, rafael h.p. creative commons attribution-noncommercial 3.0 spain table 4. description of the industrial clusters researched. industrial cluster economic sector institution re-searched type of coordination sertaozinho (c1) metal-mechanic industry ceise and apl metaltec the coordinator is a full-time sebrae consultant arapongas (c2) furniture producers sima and furniture apl of arapongas the coordinator and the vice-coordinator are company owners in the city londrina (c3) information and communication technology ict apl of londrina the coordinator and the vice-coordinator are company owners in the city table 5. scores obtained in eah cluster. req. scores (0 to 10) avg. rangec1 c2 c3 r1 7 10 7 8,0 3 r2 5 5 2 4,0 3 r3 10 7 7 8,0 3 r4 5 9 6 6,7 4 r5 9 9 10 9,3 1 r6 6 5 7 6,0 2 r7 6 9 6 7,0 3 r8 3 6 10 6,3 7 r9 6 7 6 6,3 1 r10 1 3 2 2,0 2 r11 4 4 5 4,3 1 r12 5 6 5 5,3 1 r13 1 1 3 1,7 2 r14 0 0 0 0,0 0 r15 3 6 7 5,3 4 r16 3 6 7 5,3 4 r17 3 5 8 5,3 5 r18 3 5 5 4,3 2 both their means and ranges. table 7 classifies the requirements according to their means and ranges, in which the rows represent the categories for the mean adapted to real numbers (see section 4.5) and the columns represent the three classifications for the range, as previously described in this section. one important conclusion that can be drawn from table 7 is that the scores did not differ significantly in 9 out of the 18 requirements (see first column in table 7). nevertheless, only r5 had a mean score above 7, which indicates that the identification of the actors involved with the cluster is a common practice in the three clusters researched. the diagnosis of the local productive system (r6), the formulation (r9) and evaluation (r11) of joint actions, performance communication (r12) and the impact of joint actions (r18) achieved a partial level of compliance. in such cases, either the management practices were still being implemented or they still needed improvement. for example, all the clusters had some qualitative mechanism to evaluate the results of joint actions (r11), which was done mainly in meetings with the institution staff and companies’ representatives, but none had performance measures to quantify the efficiency and effectiveness of the joint action. it is seemingly a consequence of the inexistence of a formal pms in the three clusters, which is demonstrated by the low scores recorded in r10 and r14. besides, the requirement r13 achieved low scores in all clusters because there were no benchmarking mechanisms in place to assist the institution in comparing the cluster’s performance with that of other clusters. the ict cluster of londrina was the only one that was planning to take part in a sebrae benchmarking initiative that seeks to compare the performance of companies according to the criteria from the brazilian national quality award. such initiative, however, is still at an early stage of implementation and concrete results have not been observed as of the time of this research. the establishment of a local office (r1), the existence of a coordination team (r3) and the awareness of companies (r7) also seem to be common practices, though the observed range for these requirements was equal to 3. in fact, the scores to r1 and r3 were equal or greater than 7 in the three clusters, and the range equalled 3 because one of the clusters scored 10 in these requirements. as for r7, londrina and sertaozinho scored below 7 because they were facing difficulties in formulating joint actions that draw the attention from small and large companies at the same time. in the case of londrina, the cluster coordinator reported that many companies have not realized the benefits of taking part in the cluster initiatives regardless of their size, which partly explains why so many companies do not participate in the meetings periodically held in the institution. among the requirements with range greater than 3, the formulation of strategic plans (r8) and the involvement of small and large companies (r17) were the ones that most differed in the cross-case analysis. the cluster of londrina has steadily formulated strategic plans since 2006, whilst in sertaozinho a formal strategic plan has never been written. arapongas obtained an intermediate score because the practice was discontinued in 2009 to be resumed only in 2011. as for the requirement r17, londrina achieved the highest score because the institution has successfully carried out initiatives that benefit companies regardless of their sizes, even though many of the local companies have not participated in these initiatives. the cluster of arapongas, on its 23int. j. prod. manag. eng. (2013) 1(1), 13-26 institutions for collaboration in industrial clusters: proposal of a performance and change management model creative commons attribution-noncommercial 3.0 spain table 6. scores grouped according to the implementation stages. stage c1 c2 c3 score avg. score avg. score avg. s-1 36 7,20 40 8,00 32 6,40 s-2 32 4,00 41 5,13 41 5,13 s-3 12 2,40 22 4,40 27 5,40 overall 80 4,44 103 5,72 103 5,72 table 7. requirements classified according to the means and ranges observed. r ≤ 2 r = 3 r ≥ 4 mean ≤ 3 r10, r13, r14 3 < mean < 7 r6, r9, r11, r12, r18 r2 r4, r8, r15, r16, r17 mean ≥ 7 r5 r1, r3, r7 figure 5. line graph with scores in each case     figure  5.  line  graph  with  scores  in  each  case   turn, reported that large companies are participating in the initiatives and meetings, though they act as observers rather than proactive agents. another important analysis that can be made concerns the average scores obtained by grouping the requirements according to the implementation stages. the line graph in figure 6 was based on the data from table 6. it is apparent in this line graph that the best scores were obtained in the first implementation stage. this is an indication that the clusters researched have established their local offices, coordination teams, and have identified the local infrastructure. a considerable drop can be noted in the second stage, which is caused mainly by the requirements r10, r11 and r13. this shows that designing performance measures, assessing the results of joint actions and establishing benchmarking mechanisms are still challenges in all the clusters. with exception of londrina, the lowest scores were observed in the third implementation stage. it is apparently a consequence of the nature of this stage, whose requirements demand that the management practices implemented in s-1 and s-2 become more mature and effective. moreover, in order to achieve the third stage the cluster needs to learn from experience with past initiatives. this means that new joint actions should not only be well managed, but also that they should encompass the real interests of the parties involved so that their performance is impacted positively. a hypothesis derived from this reasoning is that an industrial cluster can achieve high scores in s-3 only after several iterations of s-2, which is the continuous improvement cycle from figure 3. 7. conclusions institutions for collaboration in industrial clusters have played a vital role in improving the capabilities of local companies and in carrying out joint actions that extend the benefits of agglomeration beyond external economies. this was the motivation of this research, which aimed to contribute to the body of knowledge on industrial clusters by putting forth a performance and change management model to guide ifcs in the planning, implementation and assessment of joint actions. the model was divided into three dimensions – infrastructure; strategic planning; and implementation and assessment. to each of these dimensions, a number of management practices were associated. based on this model, three implementation stages were identified, which served as ground to the formulation of a self-assessment tool that help the cluster determine its level of compliance with the proposed management model. the tool was used in three industrial clusters to evaluate their management practices according to the model, which led to important insights and findings. first, the highest scores were observed in the first implementation stage (s-1), which is strongly related to the infrastructure dimension of the model. this is an indication that the clusters have not faced great barriers in establishing the infrastructure at the institutional level and identifying the local infrastructure and actors at the regional level. as for the strategic planning, some positive practices could be found in all the three clusters, though they have not been able to design performance measures to assess the results of joint actions in numerical terms. this may prevent future joint actions from drawing more interest of local companies, mainly because companies will not be able to measure precisely the benefits of taking part in such actions. additionally, the inexistence of a performance measurement system hinders the benchmarking with other industrial clusters. based on the scores obtained by each cluster, it becomes apparent that issues related to performance measurement prevented them from scoring better in s-3, since this stage requires that the institution establishes more mature management practices to plan, implement and assess joint actions. although the findings of this paper cannot be extended to all industrial clusters, they serve as empirical evidence that, in general, measuring the benefits of joint actions numerically is not a common practice yet. future research on ifcs should seek ways to overcome the barriers to performance measurement, strategic planning, and the assessment of joint actions, since no widely accepted solutions for these issues have been proposed so far. 24 int. j. prod. manag. eng. (2013) 1(1), 13-26 carpinetti, luiz c.r., & lima, rafael h.p. creative commons attribution-noncommercial 3.0 spain   figure 6. compliance to the requirements according to the implementation stages. acknowledgements: this should always be a run-in heading and not a section or subsection heading. it should not be assigned a number. the acknowledgements may include reference to grants or supports received in relation to the work presented in the paper. references akao, y. (2004). hoshin kanri: policy deployment for successful tqm. new york: productivity press. allison, m., & kaye, j. (2005). strategic planning for non-profit organizations: a practical guide and workbook. new jersey: john wiley & sons. andriani, p., jones, c., perkmann, m., propris, l., sena, v., delbridge, r., möslein, k. m., & neely, a. (2005). challenging clusters – the prospects and pitfalls of clustering for innovation and economic development. summary report from an aim management research forum in cooperation with the welsh economy research unit, advanced institute of management research, june. audretsch, d., & lehmann, e. (2006). the role of clusters in knowledge creation and diffusion: an institutional perspective, in asheim, b., cooke, p., & martin, r. (eds). clusters and regional development: critical reflections and explorations, (pp. 188-198). abingdon: routledge. baptista, r., & swann, p. (1998). do firms in clusters innovate more? research policy, 27(1), 535-540. bititci, u., martinez, v., albores, p., & parung, j. (2004). creating and managing value in collaborative networks. international journal of physical distribution and logistics management, 34(3/4), 251-268. doi:10.1108/09600030410533574 bortolotti, f., amato neto, j. (2006). development of a set of matrices for the evaluation and classification of regional clusters, in: proceedings of the euroma conference, glasgow, scotland. brazilian micro and small business support service sebrae (2007). perfil econômico municipal. http://www.sebraesp.com.br/ conhecendo_mpe/setoriais_regionais/perfil_economico camarinha-matos, l., & afsarmanesh, h. (2005). collaborative networks: a new scientific discipline. journal of intelligent manufacturing, 16(1), 439-452. doi:10.1007/s10845-005-1656-3 camarinha-matos, l., & afsarmanesh, h. (2007). performance indicators for collaborative networks based on collaboration benefits. production planning and control, 18(7), 592-609. doi:10.1080/09537280701546880 capó-vicedo, j., expósito-langa, m., & molina-morales, f. x. (2008). improving sme competitiveness reinforcing interorganisational networks in industrial clusters. international entrepreneurship and management journal, 4(2), 147-169. doi:10.1007/s11365-0070059-7 carpinetti, l., cardoza, e., & gerolamo, m. (2008). a measurement system for managing performance of industrial clusters: a conceptual model and research cases. international journal of productivity and performance management, 57(5), 405-419. doi:10.1108/17410400810881854 chow, d., heaver, t., & henriksson, l. (1994). logistics performance: definition and measurement. international journal of physical distribution & logistics management, 24(1), 17-28. doi:10.1108/09600039410055981 cohen, d., & prusak, l. (2001). in good company: how social capital makes organizations work. boston: harvard business school press. cortright, j. (2006). making sense of clusters: regional competitiveness and economic development. washington: the brookings institution. english department of trade and industry dti (2005). a practical guide to cluster development: report to the department of trade and industry and the english rdas by ecotec research & consulting. european commission ec (2001). methodology for regional and transnational clusters: learning with european best practices – promotion of entrepreneurship and smes. enterprise directorate general, promotion of entrepreneurship and smes, business co-operation and community business support network development, march. gerolamo, m., carpinetti, l., seliger, g., & galdamez, e. (2008). performance management of regional clusters and sme cooperation networks. international journal of business excellence, 1(4), 457-483. doi:10.1504/ijbex.2008.018844 goodstein, l., nolan, t., & pfeiffer, j. (1993). applied strategic planning: how to develop a plan that really works. new york: mcgraw-hill. institute for economic and social development of paraná – ipardes (2006a). arranjo produtivo local de móveis de arapongas – nota técnica, curitiba, brasil. institute for economic and social development of paraná ipardes (2006b). arranjo produtivo local de software de londrina – estudo de caso, curitiba, brasil. international trade department itd (2009). clusters for competitiveness – a practical guide & policy implications for developing cluster initiatives. the world bank. kaplan, r., & norton, d. (1996). the balanced scorecard: translating strategy into action. boston: harvard business school press. 25int. j. prod. manag. eng. (2013) 1(1), 13-26 institutions for collaboration in industrial clusters: proposal of a performance and change management model creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.1108/09600030410533574 http://www.sebraesp.com.br/conhecendo_mpe/setoriais_regionais/perfil_economico http://www.sebraesp.com.br/conhecendo_mpe/setoriais_regionais/perfil_economico http://dx.doi.org/10.1007/s10845-005-1656-3 http://dx.doi.org/10.1080/09537280701546880 http://dx.doi.org/10.1007/s11365-007-0059-7 http://dx.doi.org/10.1007/s11365-007-0059-7 http://dx.doi.org/10.1108/17410400810881854 http://dx.doi.org/10.1108/09600039410055981 http://dx.doi.org/10.1504/ijbex.2008.018844 karaev, a., koh, s., & szamosi, l. (2007) the cluster approach and sme competitiveness: a review. journal of manufacturing technology management, 18(7), 818-835. doi:10.1108/17410380710817273 kennerley, m., & neely, a. (2002). a framework of the factors affecting the evolution of performance measurement systems. international journal of operations and production management, 22(11), 1222-1245. doi:10.1108/01443570210450293 ketels, c., & sölvel, ö. (2006). clusters in the eu-10 new member countries. europe innova cluster mapping. krugman, p. (1991). geography and trade. cambridge: mit press. ministry of development, industry and foreign trade mdic (2010). arranjos produtivos locais. http://www2.desenvolvimento.gov.br/sitio/ sdp/proacao/arrprolocais/arrprolocais.php morosini, p. (2004). industrial clusters, knowledge integration and performance. world development, 32(2), 305-326. doi:10.1016/j. worlddev.2002.12.001 mytelka, l., & farinelli, f. (2000). local clusters, innovation systems and sustained competitiveness, discussion paper, united nations university, the netherlands. neely, a., gregory, m., & platts, k. (1995). performance measurement system design: a literature review and research agenda. international journal of operations and production management, 15(4), 80-116. doi:10.1108/01443579510083622 neely, a. (1998). measuring business performance. london: the economist. neely, a. (2005). the evolution of performance measurement research developments in the last decade and a research agenda for the next. international journal of operations & production management, 25(12), 1264-1277. doi:10.1108/01443570510633648 plummer, p., & taylor, m. (2001). theories of local economic growth (part 1): concepts, models, and measurement. environment and planning, 33(2), 219-236. doi:10.1068/a339a porter, m. (1998). clusters and the new economics of competition. harvard business review, nov-dec, 77-90. radnor, z., & barnes, d. (2007). historical analysis of performance measurement and management in operations management. international journal of productivity and performance management, 56(5/6), 384-396. doi:10.1108/17410400710757105 schmitz, h. (1995). collective efficiency: growth path for small-scale industry. journal of development studies, 31(4), 529-566. doi:10.1080/00220389508422377 schmitz, h. (1998). responding to global competitive pressure: local cooperation and upgrading in the sinos valey, brazil. institute of development studies, working paper, 82. schmitz, h., & nadvi, k. (1999). clustering and industrialization: introduction. world development, 27(9), 1503-1514. doi:10.1016/s0305750x(99)00072-8 seliger, g., carpinetti, l., & gerolamo, m. (2008). promoting innovative clusters and cooperation networks: the european commission observatories of smes and the context of berlin-brandenburg. international journal of networking and virtual organizations, 5(2), 204223. doi:10.1504/ijnvo.2008.017011 simatupang, t., & sridharan, r. (2005). the collaboration index: a measure for supply chain collaboration. international journal of physical distribution and logistics management, 35(1), 44-62. doi:10.1108/09600030510577421 sölvell, ö., lindqvist, g., & ketels, c. (2003). the cluster initiative greenbook. stockholm: bromma tryck ab. sölvell, ö. (2008). clusters – balancing evolutionary and constructive forces. stockholm: ivory tower publishers. steiner, m., & hartmann, c. (2006). organizational learning in clusters: a case study on material and immaterial dimensions of cooperation. regional studies, 40(5), 493-506. doi:10.1080/00343400600757494 steiner, m., & ploder, m. (2008). structure and strategy within heterogeneity: multiple dimensions of regional networking. regional studies, 42(6), 793-815. doi:10.1080/00343400701861310 steinle, c., & schiele, h. (2002). when do industries cluster? a proposal on how to assess an industry’s propensity concentrate at a single region or nation. research policy, 31(6), 849-858. doi:10.1016/s0048-7333(01)00151-2 united nations industrial development organization unido (2001). development of clusters and networks of smes, vienna. 26 int. j. prod. manag. eng. (2013) 1(1), 13-26 carpinetti, luiz c.r., & lima, rafael h.p. creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.1108/17410380710817273 http://dx.doi.org/10.1108/01443570210450293 http://www2.desenvolvimento.gov.br/sitio/sdp/proacao/arrprolocais/arrprolocais.php http://www2.desenvolvimento.gov.br/sitio/sdp/proacao/arrprolocais/arrprolocais.php http://dx.doi.org/10.1016/j.worlddev.2002.12.001 http://dx.doi.org/10.1016/j.worlddev.2002.12.001 http://dx.doi.org/10.1108/01443579510083622 http://dx.doi.org/10.1108/01443570510633648 http://dx.doi.org/10.1068/a339a http://dx.doi.org/10.1108/17410400710757105 http://dx.doi.org/10.1080/00220389508422377 http://dx.doi.org/10.1016/s0305-750x(99)00072-8 http://dx.doi.org/10.1016/s0305-750x(99)00072-8 http://dx.doi.org/10.1504/ijnvo.2008.017011 http://dx.doi.org/10.1108/09600030510577421 http://dx.doi.org/10.1080/00343400600757494 http://dx.doi.org/10.1080/00343400701861310 http://dx.doi.org/10.1016/s0048-7333(01)00151-2 pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.1609 received 2013-07-23 accepted 2013-11-19 reduction in repair rate of welding processes by determination & controlling of critical kpivs. yousaf, f. i and ikramullah butt, s.ii national university of science and technology pakistan, school of mechanical and manufacturing engineering iengr_faheem37@yahoo.com iihodmech@smme.nust.edu.pk abstract: six sigma is being implemented all over the world as a successful quality improvement methodology. many companies are now days are using six sigma as an approach towards zero defects. this article provides a practical case study regarding the implementation of six sigma project in a welding facility and discusses the statistical analysis performed for bringing the welding processes in the desired sigma limits. dmaic was chosen as potential six sigma methodology with the help of findings of this methodology, six sigma team first identified the critical factors affecting the process yield and then certain improvement measures were taken to improve the capability of individual welding processes and also of overall welding facility. cost of quality was also measured to validate the improvement results achieved after conducting the six sigma project. key words: statistical process control (sqc), process capability indices, six sigma, variable control chart, pareto charts, standard deviation, critical to quality (ctq), analysis of variance (anova), dmaic, total quality management. 1. introduction in this era of changing customer needs and demand of highly reliable products have pushed many manufacturing companies to adopt total quality management (tqm) principles. globalization and extension of product market has also increased the need of quality products at reasonable cost to customers. to respond to these demands many companies are implementing different quality management principles at their manufacturing facilities such as iso 9000, just in time (jit), lean manufacturing, and kaizen etc. a new and improved quality improvement approach called six sigma is also becoming popular in controlling the defect rate and managing the quality as overall process function. 2. six sigma as an improvement approach it is the set of practices originally developed by motorola to systematically improve process by eliminating defects. defect is defined as nonconformity of a product or service to its specification. like its previous quality improving methodologies six sigma focuses on the following points. a continuous effort to reduce variation in process outputs is essential to business success. manufacturing and business processes can be measured, analyzed, improved and controlled. in order to achieve best quality improvement results, role of upper management is very critical. the term six sigma refers to a highly capable process that can produce products within specifications. process that achieves six sigma levels produces only 3.4 defective products per million opportunities. main focus of six sigma is to improve all key processes of manufacturing setup and takes quality as a function of processes capability to produce items with in specification. 2.1. dmaic overview define, measure, analyze, improve and control (dmaic) is a six sigma methodology mainly used for improving quality of already established processes and manufacturing systems. basically 23int. j. prod. manag. eng. (2014) 2(1), 23-36creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2014.1609 mailto:engr_faheem37@yahoo.com mailto:hodmech@smme.nust.edu.pk this methodology comprises of following five key points. define the process improvement goals that are aligned with the customer demands and company’s strategy. measure the current process and make a strategy for making further improvement. analyze to verify the relationship and causality of factors. determine what the relationship is and attempts to ensure that all the factors have been considered. improve and optimize process based on findings of analysis phase using different techniques. control to ensure that any variances are corrected before they result in defects. in this research dmaic is used as potential six sigma methodology to bring quality improvements in manufacturing company. 2.2. case study 2.2.1. company profile the pakistan welding institute (pwi) is a pakistan based professional institution devoted to maintain and promote standards of excellence in welding technology. pwi provides industry with technical support through advice & information, consultancy, research and development (r & d) and training & qualification. its services and expertise cover all areas of welding & joining technology and materials engineering for metals and non-metals alike. pwi has the capacity of welding almost all commercially available engineering materials ranging in thickness from 0.1mm to 300mm. 2.2.2. problem statement the head manufacturing at pakistan welding institute was not satisfied from the current welding repair rate. from the last few months he was receiving the complaints from the asme authorized inspector and the client’s inspector that in a number of welding jobs due to a higher repair rate the quality of the product is suffering; and there level of confidence is decreasing on the production-welding process. head manufacturing also showed the concern with reference to the last financial review; showing that the manufacturing is bearing a larger amount due to the welding repair work. 2.2.3. research methodology to overcome the welding problem defined previously, companies upper management decided to launch a six sigma quality improvement study. as six sigma further comprises of different methodologies so by studying the nature of problem it was decided to choose dmaic methodology which consists of sequential identification and controlling of root causes of problem to bring the process under control and in desired quality level. 3. data and results 3.1. define phase (what the problem is and what customer wants) define phase of the project helps to identify the problem according to the demand of customers. in this phase of project, quality problems and future roadmap for the project are defined. the project started with the investigation of the problem. this was evaluated in greater depth with the help of process map and other tools. findings of define phase are given as. table 1. dmaic project charter. project title minimization of welding repair rate by using dmaic approach business case welding is one of the most critical processes in the pwi equipment manufacturing area. higher repair rate increases the cost and decreases the productivity. by decreasing the welding repair rate overall project quality and productivity would be improved and cost will be saved moreover all the interested parties including internal/external (execution team, authorized inspectors, clients inceptors) customer satisfaction level will be improved. goal repair rate to be minimized up to 0.25% age. metrics (ctq’s) primary metric (% age of repair rate), cost of quality (rupees) project scope welding section, ndt section, procurement, quality control and store department should involve during different phases of the project 24 int. j. prod. manag. eng. (2014) 2(1), 23-36 creative commons attribution-noncommercial 3.0 spain yousaf, f. and ikramullah butt, s. 3.1.1. project charter a project charter is established by visiting welding facility. production and quality departments helped in understanding current performance of facility. table 1 gives details of project`s charter. 3.1.2. welding processes flow chart at pwi to understand the details of welding processes and to identify the root causes efficiently welding process flow chart was established by six sigma team. figure 1 describes the welding flow chart. 3.1.3. supplier input process output customer (sipoc) to understand the relationship between different departments at pwi the sipoc diagram was made. table 2 shows the findings of sipoc diagram. 3.1.4. voice of customers the needs of customers have been identified by coordinating with six sigma black belt and quality engineering department after elaborated discussion with the internal and external customers. from the 25int. j. prod. manag. eng. (2014) 2(1), 23-36creative commons attribution-noncommercial 3.0 spain reduction in repair rate of welding processes by determination & controlling of critical kpivs. reject accept reject reject accept drawing/) detail)of) weld!joints! electrode/wire)selection) pqr testing visual) inspection)rectification) testing)) of)welds) send)to)next) shop) issuance)of)electrodes) form)store) ! wps)preparation) ! pqr$preparation$ ! finalization)of)wps) ! distribution)of)wps)on)shop)) ! welding)execution) ! ndt)as)per) specification) ! rectification) welding)sequence) finalization! welding)process) selection) • optimum!process! selection! • welding!parameters! selection! ! customer) requirements) • quality!demands! • contract! • voice!of!customer! • cost!limitations! ! ! figure 1. welding processes flow chart at pwi. view point of customers it is clear that proper welds made according to the specific standards and codes is key to the customer satisfaction. figure 2. elaborates quality of welds according to the view point of customers. 3.1.5. define phase outcomes welding is one of the most critical processes in the pakistan welding institute equipment manufacturing area. higher repair rate increases the cost and decreases the productivity. by decreasing the welding repair rate overall project quality and productivity would be improved and cost will be saved moreover all the interested parties including internal/external (execution team, authorized inspectors, clients inceptors) customer satisfaction level will be improved. project goal is to reduce repair rate up to 0.25 %age. the welding facility has up to date and a well-controlled quality management system to ensure proper quality of welds. the welding company follows codes and standards of american welding society and american society of mechanical engineers for proper execution and documentation of welding different projects. the welding facility is well equipped with modern welding technologies and welds testing labs. 26 int. j. prod. manag. eng. (2014) 2(1), 23-36 creative commons attribution-noncommercial 3.0 spain yousaf, f. and ikramullah butt, s. table 2. sipoc diagram. supplier input process output customer engineering department latest drawings/ specifications weld map preparation details of weld joints drawing production department production department details of weld joints and design requirements preparation of weld matrix (wps & pqr) weld matrix welding department & quality control welding engineer materials and welding requirements electrodes/ filler wires selection & requirements electrodes/ filler wires compatible with base metal according to the wps welding department welding engineer all welding parameters and their qualification reports lab testing (mechanical testing) according to design / code requirements test reports welding department & quality control project engineer detail welding requirements welder selection selected welder capable of welding sound weld & wpqr welding department & quality control welding engineer welding requirements, joint number and applicable wps welding execution welding of job according to weld matrix fabrication engineer/ area supervisor project engineer/ welding engineer production test plate requirement and process lab testing (production test plate) test reports values according to design/ code welding department & quality control project engineer all ndt requirements visual testing ndt (r.t., u.t., d.p.t. & m.p.t.) inspection reports, nde reports quality control/ third party inspector/ client good quality consumable proper storage facility for consumables good welding equipment qualified welding procedures certified testing labs trained testing personnel implementation of latest codes and standard proper identification of welds voice of customers qualified welder good appearance and cleanliness of welds figure 2. voice of customers. 3.2. measure phase (establishing the base line for the dmaic project) after studying the nature of problem in the define phase, the six sigma team started collecting the data in order to measure project outputs in more detail and from different angles. the measure phase now focuses to get a bit more information about the welding processes by measuring the yield of different projects performed in past and calculating current sigma levels. this will help to identify areas of improvement and bench mark the quality levels to be achieved by bringing improvements. some tools of measure phase are given in the following. 3.2.1. defining project inputs and outputs (x`s and y`s) for defining the critical inputs and outputs of the six sigma project variables a brainstorming session involving the six sigma team, authorized inspector, clients inspectors and internal quality control inspectors was conducted. sipoc diagram was also used as an input for this session. after conducting many sessions with different stake holders the cause and effect analysis was made. cause and effect diagram is shown in figure 3. here wps stands for welding procedure specification a wps is a written procedure prepared to provide direction for making production welds according to code requirements. pqr is an abbreviation of procedure qualification record a pqr lists what was used in qualifying the wps and test results. 3.2.2. cause and effect (c & e) matrix based on the findings’ of process x’s and y’s and rating of importance to customers the cause and effect matrix is developed. table 3 describes the cause and effects of different input process variables on the critical process output variables in form of highest and lowest scores. on the basis of the cause and effect matrix; project team concluded three critical x(s) that influence the output variables the most. these three variables are defined as: welder skill: capability of the welder to produce sound weld (i.e. weld according to wps &should be defect free) tool & equipment: tool & equipment includes welding machines, welding holders, welding torches. consumables: consumables includes electrodes and filler wires used for welding purpose. 27int. j. prod. manag. eng. (2014) 2(1), 23-36creative commons attribution-noncommercial 3.0 spain reduction in repair rate of welding processes by determination & controlling of critical kpivs. figure 3. causes and effect diagram of welding defects. 3.2.3. measurement system analysis and gage r&r study to identify the repair rate, defect length is the most important factor. quality control personnel (ndt level ii) are responsible to review the ndt (radiography) report to identify the defects length of the respective type of the defect. six sigma team selected the three radiographic films and three quality inspectors (ndt levelii). each inspector viewed the radiographic films three times and then collected data is used to perform the following analysis. table 4 shows that gage r&r % is 0.59% which is less than 1%. according to msa standard if total gauge r&r is between 1 and 9 the measurement system is acceptable and if it is less than 1 the system is highly acceptable. total study variation is 7.70% which is less than 30% of the msa standard the distinct category is 18, which is greater than the minimum requirement 5 of the msa standard. therefore according to above conclusions the six sigma team agreed that the measurement system for the welding repair work is acceptable. 28 int. j. prod. manag. eng. (2014) 2(1), 23-36 creative commons attribution-noncommercial 3.0 spain yousaf, f. and ikramullah butt, s. table 3. cause and effect matrix. key process output variables (kpov`s) rating of importance 6 6 3 9 9 9 visual appearance welding size proper cleaning internal defects external defects mechanical &chemical properties of weld total sr. no. key process input variables (kpiv`s) priority rating 1 drawing/specifications 3 0 3 0 0 0 3 45 2 weld matrix 6 0 0 0 3 0 6 81 3 wps/pqr 6 0 0 0 3 3 6 108 4 welder skill 9 9 9 9 9 9 9 378 5 tool & equipment 6 3 0 6 3 6 3 126 6 consumable 9 9 3 3 9 9 9 324 reduction in repair rate of welding processes by determination & controlling of critical kpivs 4 | int. j. prod. manag. eng. (2014) 2(1), ppp-ppp creative commons attribution-noncommercial 3.0 spain quality inspectors (ndt levelii). each inspector viewed the radiographic films three times and then collected data is used to perform the following analysis. table 4 shows that gage r&r % is 0.59% which is less than 1%. according to msa standard if total gauge r&r is between 1 and 9 the measurement system is acceptable and if it is less than 1 the system is highly acceptable. total study variation is 7.70% which is less than 30% of the msa standard the distinct category is 18, which is greater than the minimum requirement 5 of the msa standard. therefore according to above conclusions the six sigma team agreed that the measurement system for the welding repair work is acceptable. table 4. two-way anova table with interaction source df ss ms f p part 2 1352.29 676.143 511.95 0.000 operator 2 2.77 1.384 1.05 0.431 part * operator 4 5.28 1.321 2743.03 0.000 repeatability 18 0.01 0.000 total 26 1360.34 gage repeatability and reproducibility (r&r) %contribution source varcomp (of varcomp) total gage r&r 0.4476 0.59 repeatability 0.0005 0.00 reproducibility 0.4471 0.59 operator 0.0070 0.01 operator*part 0.4401 0.58 part-to-part 74.9802 99.41 total variation 75.4278 100.00 study var %study var %tolerance source stddev (sd) (6 * sd) (%sv) (sv/toler) total gage r&r 0.66901 4.0140 7.70 20.07 repeatability 0.02194 0.1317 0.25 0.66 reproducibility 0.66865 4.0119 7.70 20.06 operator 0.08371 0.5022 0.96 2.51 operator*part 0.66338 3.9803 7.64 19.90 part-to-part 8.65911 51.9547 99.70 259.77 total variation 8.68492 52.1095 100.00 260.55 number of distinct categories = 18 3.2.4 welding defects data collection data was collected for all the projects which were executed during november 2012 to april 2013.total projects are 20 in number. the 100% of the last 6 months projects data was collected by the six sigma team. from figure 4 it is clear that slag inclusions have the highest frequency of occurrence. defect (cm) 706 325 98 67 62 52 21 7 percent 52.8 24.3 7.3 5.0 4.6 3.9 1.6 0.5 cum % 52.8 77.1 84.4 89.4 94.0 97.9 99.5 100.0 defect type tu ng ste n in clu sio nslo f ot he r d ef ec ts lo p un de rcu ts cr ac ks po ro sit y sla g in clu sio n 1400 1200 1000 800 600 400 200 0 100 80 60 40 20 0 d e fe ct ( cm ) p e rc e n t pareto chart of defect type figure 4. pareto chart of defect type repair (note: lof = lack of fusion, lop = lack of penetration, other defects = root concavity etc.) table 4. two-way anova table with interaction. 3.2.4. welding defects data collection data was collected for all the projects which were executed during november 2012 to april 2013.total projects are 20 in number. the 100% of the last 6 months projects data was collected by the six sigma team. from figure 4 it is clear that slag inclusions have the highest frequency of occurrence. defect (cm) 706 325 98 67 62 52 21 7 percent 52.8 24.3 7.3 5.0 4.6 3.9 1.6 0.5 cum % 52.8 77.1 84.4 89.4 94.0 97.9 99.5 100.0 defect type tu ng ste n in clu sio nslo f ot he r d ef ec ts lo p un de rcu ts cr ac ks po ro sit y sla g in clu sio n 1400 1200 1000 800 600 400 200 0 100 80 60 40 20 0 d e fe ct ( cm ) p e rc e n t pareto chart of defect type figure 4. pareto chart of defect type repair (note: lof=lack of fusion, lop=lack of penetration, other defects=root concavity etc.). repair (cm) 931 270 66 56 15 percent 69.6 20.2 4.9 4.2 1.1 cum % 69.6 89.8 94.7 98.9 100.0 welding tecnique sawgmawfcawgtawsmaw 1400 1200 1000 800 600 400 200 0 100 80 60 40 20 0 r e p a ir ( cm ) p e rc e n t pareto chart of welding tecnique figure 5. pareto chart of welding processes (note: smaw=shielded metal arc welding, gtaw=gas tungsten arc welding, saw=submerged arc welding, fcaw=flux cored arc welding, gmaw=gas metal arc welding). figure 5 shows that shielded metal arc welding has also highest contribution in defect or repair rate where gas tungsten arc welding has the second most impact. it is hence cleared that major improvements can be brought in quality of welds by targeting shielded metal arc welding and gas tungsten arc welding and factors contributing to the occurrence of slag inclusions and porosity. 3.2.5. calculation of sigma values six sigma team calculated the sigma values for the minimizing welding repair work project. table 5 represents the calculated sigma values of each welding process and overall welding facility. 3.2.6. measure phase outcomes major conclusions that can be drawn from measure phase of the project are: significant x(s) (kpivs) have been found. welder skills, consumables and welding equipment are found to be critical input variables that influence the quality of welding. shielded metal arc welding and flux cored arc welding are the processes with lowest sigma values so these processes are selected for further analysis. slag inclusions and porosity are the most frequently occurring defects so efforts will be made to minimize these defects. base materials welded in previous projects by shielded metal arc welding process are mostly different grades of stainless steel and carbon steel and plate and pipe welding were usually performed in those projects. so these types of welding are chosen for experimental scheme in further project phases. target is to minimize the welding repair rate up to 0.25%. 29int. j. prod. manag. eng. (2014) 2(1), 23-36creative commons attribution-noncommercial 3.0 spain reduction in repair rate of welding processes by determination & controlling of critical kpivs. table 5. calculation of sigma values of welding processes. sr. no. welding technique weld length (cm) repair defects (cm) defect % dpmo yield sigma cpk 1 smaw 21778 931 4.2749 42749.5 95.72 3.2 1.07 2 gtaw 123853 270 0.2180 2180 99.78 4.3 1.43 3 fcaw 1921 66 3.4357 34357 96.56 3.3 1.1 4 saw 7415 15 0.2022 2022 99.79 4.3 1.43 5 gmaw 41921 56 0.1335 1336 99.86 4.5 1.5 total 196888 1338 0.70 7049.69 99.30 4 1.33 3.3. analyze phase (analyze source of variation) after measurement phase and establishing the baseline and target level, the team analyzed the causal relationships in detail. this phase involved identifying and validating possible x’s and prepare for the design of experiment for the improve phase. 3.3.1. analysis of welding processes with low sigma values the findings of measure phase show that two welding processes i.e. shielded metal arc welding and flux cored arc welding have the low sigma values of 3.2 and 3.3 consecutively. based on the facts shown in measure phase, flux cored arc welding is a semi-automated arc welding process that is rarely used in the execution of projects at the welding facility. table 5 describes that fcaw technique is used to weld only 1921 cm of welding length, reasons behind this fact is limitations of this welding technique because of high cost associated with its operation. so the decision here is to remove flux cored arc welding from the investigation list and focus of improvement will now be on shielded metal arc welding due to its lowest sigma values and high repair rate. reduction in repair rate of welding processes by determination & controlling of critical kpivs 6 | int. j. prod. manag. eng. (2014) 2(1), ppp-ppp creative commons attribution-noncommercial 3.0 spain 3.3.1 analysis of welding processes with low sigma values the findings of measure phase show that two welding processes i.e. shielded metal arc welding and flux cored arc welding have the low sigma values of 3.2 and 3.3 consecutively. based on the facts shown in measure phase, flux cored arc welding is a semi-automated arc welding process that is rarely used in the execution of projects at the welding facility. table 5 describes that fcaw technique is used to weld only 1921cm of welding length, reasons behind this fact is limitations of this welding technique because of high cost associated with its operation. so the decision here is to remove flux cored arc welding from the investigation list and focus of improvement will now be on shielded metal arc welding due to its lowest sigma values and high repair rate. figure 6. experimental setup for fillet weld by shielded metal arc welding 3.3.2 screening experiments to analyze the sources of variation in shielded metal arc welding it is necessary to first ensure the smooth welding process that is not influenced or affected by the process parameters. for this purpose a brain storming session was conducted with the welding engineer and welding literature was consulted to identify the primary source of variation in shielded metal arc welding process. few factors that were identified are welding electrode diameter, welding electrode length (size), welding arc length, welding travel speed. a multilevel factorial experiment was designed to analyze effect of different values of these factors on response variable. the response variables selected are the defect % of slag inclusion or porosity. experimental scheme is given in figure 6 is shown, it describes testing plate of 3/8 inches stainless steel with fillet weld joint was tested against different input variable settings. table 6 shows the different variables values used for experimental scheme. welding material used here is aisi 304l stainless steel and electrode type used is 308l. table 6. factors settings for screening experiment s.no. factors levels 01 x1= electrode diameter 3/32, 5/32 inches 02 x2= electrode length 9, 12 inches 03 x3= arc length buried, 1/4 inches 04 x4= welding travel speed of electrode 20, 40 inches/min the analysis of variance results are shown in table 7 and figure 7; it becomes clear that electrode thickness and arc length are the significant factors with p-value of 0.007 and 0.069. other two factors have not the significant effect and can be treated as redundant factors for further analysis. thus it is recommended to use thin electrode with proper arc length to reduce slag inclusions and porosity. table 7. anova results for screening experiments analysis of variance for defect %, using adjusted ss for tests source df seq ss adj ss adj ms f p x1 1 0.254898 0.254898 0.254898 45.31 0.007 x2 1 0.000480 0.000480 0.000480 0.09 0.789 x3 1 0.043218 0.043218 0.043218 7.68 0.069 x4 1 0.003960 0.003960 0.003960 0.70 0.463 error 3 0.016877 0.016877 0.005626 total 7 0.319434 s = 0.0750033 r-sq = 94.72% r-sq. (adj) = 87.67% figure 6. experimental setup for fillet weld by shielded metal arc welding. 3.3.2. screening experiments to analyze the sources of variation in shielded metal arc welding it is necessary to first ensure the smooth welding process that is not influenced or affected by the process parameters. for this purpose a brain storming session was conducted with the welding engineer and welding literature was consulted to identify the primary source of variation in shielded metal arc welding process. few factors that were identified are welding electrode diameter, welding electrode length (size), welding arc length, welding travel speed. a multilevel factorial experiment was designed to analyze effect of different values of these factors on response variable. the response variables selected are the defect % of slag inclusion or porosity. experimental scheme is given in figure 6 is shown, it describes testing plate of 3/8 inches stainless steel with fillet weld joint was tested against different input variable settings. table 6 shows the different variables values used for experimental scheme. welding material used here is aisi 304l stainless steel and electrode type used is 308l. table 6. factors settings for screening experiment. s.no. factors levels 01 x1= electrode diameter 3/32, 5/32 inches 02 x2= electrode length 9, 12 inches 03 x3= arc length buried, 1/4 inches 04 x4= welding travel speed of electrode 20, 40 inches/min the analysis of variance results are shown in table 7 and figure 7; it becomes clear that electrode thickness and arc length are the significant factors with p-value of 0.007 and 0.069. other two factors have not the significant effect and can be treated as redundant factors for further analysis. thus it is recommended to use thin electrode with proper arc length to reduce slag inclusions and porosity. table 7. anova results for screening experiments. analysis of variance for defect %, using adjusted ss for tests source df seq ss adj ss adj ms f p x1 1 0.254898 0.254898 0.254898 45.31 0.007 x2 1 0.000480 0.000480 0.000480 0.09 0.789 x3 1 0.043218 0.043218 0.043218 7.68 0.069 x4 1 0.003960 0.003960 0.003960 0.70 0.463 error 3 0.016877 0.016877 0.005626 total 7 0.319434 s=0.0750033 r-sq=94.72% r-sq. (adj)=87.67% 30 int. j. prod. manag. eng. (2014) 2(1), 23-36 creative commons attribution-noncommercial 3.0 spain yousaf, f. and ikramullah butt, s. 5/32in3/32in 0.8 0.7 0.6 0.5 0.4 12in9in 1/4inburried 0.8 0.7 0.6 0.5 0.4 40in/min20in/min x1 m e a n x2 x3 x4 main effects plot for defect % fitted means figure 7. main effect plots for factors used in screening experiments. 3.3.3. analysis of variance of critical kpiv`s with reference to the short listed process input variables i.e. x(s), sigma team designed the experimental scheme by using design of experiment concepts. three factors chosen for analysis are welder skills, tool and equipment and consumables used for welding sample welding plates of stainless steel grade 304l materials by shielded arc welding process. test plates of 30mm thickness were welded in butt weld profile and were tested by visual inspection and radiographic tests. inspector of quality, ndt level ii was appointed to view test reports. the response variable here is slag inclusions and porosity whose defect rate is measured against different settings of input variables. three level of operator skills and two levels of other two factors were used for variation analysis. table 8 shows the data obtained from experimental settings of different variables 3.3.3.1. analysis of variance results using the results shown in table 8 a dot plot figure was created which showed greater variation in welder skills and consumables and showed lesser variation in tool and equipment. figure 8 shows the dot plot. tool and equipment`s are removed from further investigation. defects % 6.35.44.53.62.71.80.9-0.0 operator consumable tool & equipment x y z hyundai miller hyundai miller hyundai miller bohular hallirus bohular hallirus bohular hallirus bohular hallirus bohular hallirus bohular hallirus dotplot of defects % vs operator, consumable, tool & equipment figure 8. dot plot of defect% versus three factors table 8. experimental data for analysis of effect of different variables sr. no operator consumable tool & equipment defects % of slag inclusion and porosity 1 z miller bohular 5.10 2 z miller hallirus 4.393 3 y miller hallirus 3.358 4 x hyundai bohular 0.2022 5 x miller hallirus 2.388 6 x hyundai hallirus 0.128 7 x hyundai hallirus 0.310 8 z miller hallirus 5.45 9 z miller bohular 6.66 10 x hyundai bohular 0.199 11 x hyundai hallirus 0.147 12 y hyundai hallirus 2.651 13 y miller bohular 2.922 14 z miller bohular 4.916 15 y hyundai hallirus 2.708 16 y hyundai bohular 2.74 17 z miller bohular 5.513 18 x hyundai hallirus 0.1995 19 z miller hallirus 4.623 20 x hyundai bohular 0.170 31int. j. prod. manag. eng. (2014) 2(1), 23-36creative commons attribution-noncommercial 3.0 spain reduction in repair rate of welding processes by determination & controlling of critical kpivs. in figure 9, interaction plot shows that welder x with hyundai consumable is producing least defect% as compare to the welder y and z with miller consumable. it is hence clear that hyundai company manufactured consumables are the most appropriate for decreasing the defect %, thus decision here is to use hyundai consumables in further welding and to improve welder skills in improvement phase. welder skill m e a n zyx 2.0 1.5 1.0 0.5 0.0 consumable hy undai miller interaction plot (data means) for defect% figure 9. interaction plots of two factors versus response variable. 3.3.4. analyze phase outcomes from the results obtained by analyze phase analysis it is clear that arc length used during welding and thickness of electrode highly affect the defect rate of slag inclusions and porosity, so it is recommended to use ¼ inches arc length with less diameter electrode for reducing the defect percentage. furthermore project team has short listed the following two kpivs: welder skill consumable in consumables the hyundai manufactured consumables are producing good quality of welds while the reason why variation is being caused by welder skills will be analyzed and improved in next phase of the project. 3.4. improve phase (making changes) this phase involved identifying solutions, select best choice, and carrying out experimentations to validate solutions and relations between the effects and causes. 3.4.1. analysis of variance for finding factors affects for further improvement in shielded metal arc welding process a multi-level factorial experiment as designed to analyze variance of different factors suggested by six sigma team that can cause variation. for this purpose three factors were selected with two levels of each. the three factors selected are: factor 1 = shift timings, level 1 = morning, level 2 = evening factor 2 = heating time of electrode in electrode oven, level 1 = 3 hours (normal), level 2 = 5 hours (suggested) at 250 degrees centigrade temperature. factor 3 = electrode composition, level 1 = electrode with low flux deposition rate (flux deposition rate of 2 lb/hour), level 2 = electrode with high flux deposition rate (flux deposition rate of 4lb/hour). table 9 shows the data collected for the before mentioned experimental scheme. table 9. experimental scheme used for analysis of variance. run order shift electrode heating time defect % 1 2 3 4 5 6 7 8 evening morning evening evening evening morning morning morning high flux deposition high flux deposition low flux deposition high flux deposition low flux deposition high flux deposition low flux deposition low flux deposition 3 hours 3 hours 5 hours 5 hours 3 hours 5 hours 3 hours 5 hours 0.90 0.87 0.10 0.30 0.20 0.28 0.19 0.12 in table 10 and figure 10, 11 the effect of each factor is are shown. the experiments were performed on 304l pipes with 30 mm thickness in 6g position by shielded metal arc welding process. 32 int. j. prod. manag. eng. (2014) 2(1), 23-36 creative commons attribution-noncommercial 3.0 spain yousaf, f. and ikramullah butt, s. ac a ab bc c b 9080706050403020100 te rm standardized effect 12.71 a s hift timings b e lectrode ty pe c h eating time f actor n ame pareto chart of the standardized effects (response is defect%, alpha = 0.05) figure 10. pareto chart of standardized effect. morningevening 0.6 0.5 0.4 0.3 0.2 low flux depositionhigh flux deposition 5 hours3 hours 0.6 0.5 0.4 0.3 0.2 shift timings m e a n electrode type heating time main effect plots for factors figure 11. main effect plots for defect %. 3.4.2. anova conclusions from the values obtained by analysis of variance of three before mentioned factors it is clear that shift timings of welders have little or no effect on defect rate of welding process, p value of 0.295 is high enough to support this claim. electrode and heating timings of electrode in oven have p values of 0.007 and 0.009 respectively, so conclusion can be drawn that both of these factors have significant effect on the defect rate. interaction effect of both these factors is significant because p value of 0.012 is much lesser then the alpha value of 0.05. from the factorial plots it is clear that by increasing the heating time of electrode in oven the defect rate drops significantly and using low flux deposition rate electrode also cause reduction in defect rate of welding. shift timings effect is not significant and remains almost constant over the range of morning and evening as shown in plots. interaction plot of the three factors also support the fact that interaction of shift timings with other two factors do not bring significant changes in the defect rate, while interaction of heating time along with electrode type gives reduced defect rate of welding. from all these results it can be conclude that using low flux deposition electrode along with the more heating time will be set as final setting for the shielded metal arc welding process. 3.4.3. further improvement changes for improvement purpose two main changes were suggested by the welding engineer in the general welding process of the welding facility. welder skill is a strong factor identified previously in analyze phase to bring quality improvement in the welding process. for this purpose proper testing of the 33int. j. prod. manag. eng. (2014) 2(1), 23-36creative commons attribution-noncommercial 3.0 spain reduction in repair rate of welding processes by determination & controlling of critical kpivs. table 10. results of multilevel factorial experiments. analysis of variance for defect %, using adjusted ss for tests term effect coef se coef t p constant 0.3700 0.002500 148.00 0.004 0.000 shift timings –0.0100 –0.0050 0.002500 –2.00 0.295 electrode type 0.4350 0.2175 0.002500 87.00 0.007 heating time –0.3400 –0.1700 0.002500 –68.00 0.009 shift timings×electrode type –0.0150 –0.0075 0.002500 –3.00 0.205 shift timings×heating time 0.0100 0.0050 0.002500 2.00 0.295 electrode type×heating time –0.2550 –0.1275 0.002500 –51.00 0.012 s=0.00707107, press=0.0032, r-sq = 99.99%, r-sq(pred) = 99.57%, r-sq. (adj) = 99.95% welders before execution of any new welding project was necessary to be done. in most of the welding companies in the word this testing of welders is being done and called welding operator performance qualification test (wpq). hiring of the welders maintain record of each welder regular monitoring of the welder performance and training selection of welders for welding project calibration of equipment, test plates preparation, assignment of codes to welders and work pieces no yesis performance of welder according to the required quality level? deployment of welder according to project requirement welder performance qualification test figure 12. welder performance qualification process in figure 12 details of the processes inducted to bring quality improvement in welder skill area are given. this process explains that how welder ability to perform satisfactory welds will be enhanced and the welder best in performance will be chosen to perform welding on a specific project. according to changes implemented, only the best performance giving welder would be chosen regardless the capability of the welder and his reputation. the record of test plates would be used to analyze the performance and selection of welder for further projects. 3.4.4. improvements from the six sigma project table 11 shows the results from two of the recently completed jobs by shielded metal arc welding process. slag inclusions and porosity were taken as responses variable to be calculated. cost of quality was also calculated based upon the factors identified in the measure phase. clearly here sigma value of shielded metal arc welding given in table 11 is 4.30. improved sigma value of overall facility is summed up in table 12. it is clear that smaw process has improved from 3.3 sigma to 4.3 sigma which has also improved combined sigma value of overall facility from 4.0 to 4.3 sigma level. from the data shown in figure 13 it is clear that by improving sigma value of shielded metal arc welding process from 3.3 sigma to 4.3 sigma a cost of rs. 1,000,000 is saved initially and company will continue to save cost in future projects depending upon the length of welding performed by smaw process. 34 int. j. prod. manag. eng. (2014) 2(1), 23-36 creative commons attribution-noncommercial 3.0 spain yousaf, f. and ikramullah butt, s. table 12. process capability calculations. sr. no weldingtechnique defect % dpmo yield sigma cpk 1 smaw 0.27 (improved) 2689 99.73 4.3 1.43 2 gtaw 0.2180 2180 99.78 4.3 1.43 3 fcaw 3.4357 34357 96.56 3.3 1.10 4 saw 0.2022 2022 99.79 4.3 1.43 5 gmaw 0.1335 1336 99.86 4.5 1.50 total 0.2235 2235 99.74 4.3 1.43 table 11. results of quality improvement. sr. no. project no. project description weld length (cm) slag (cm) porosity (cm) repair (defects) (cm) defects (%) dpmo yield sigma 1 tk-25 rwst tank 5055 7 3 10 0.20 1978 99.8 4.30 2 s-925 generator cooler 4233 5 3 08 0.19 1890 99.81 4.30 total 7065 13 6 19 0.27 2689 99.73 4.30 figure 13. welding processes cost of quality analysis 3.5. control phase (control the improved process) table 13 shows the welding process control plan that was developed to ensure the consistent and to effectively implement the control measures. after satisfaction from the project outcome and achievement of its main objectives, the project was closed. conclusions that can be drawn from six sigma project are the following 4. conclusions pwi is the welding facility that is equipped with modern and up to date welding technologies. a quality of welds being produced in the facility are the prime concern for the upper management of the company, because that defines the overall quality of welding facility and also explains how reliable are the welds. from the past one year this company is facing quality defects in its welding projects, due to which a six sigma project was selected for implementation. the five phases of six sigma were implemented and results were obtained to bring quality improvement in welding processes. shielded metal arc welding was found to be at lowest sigma level so efforts were made to analyze source of variation for smaw process. after obtaining optimum process settings for smaw process these were implemented and results were analyzed. references chen, k.s., huang, m.l., li, r.k. (2001). process capability analysis for an entire product. international journal of production research, 39(17), 4077-4087. doi:10.1080/00207540110073082 de mast, j., roes, k.c.b, does, r.j.m.m. (2001). the multi-vary chart: a systematic approach. quality engineering, 13(3), 437-447. doi:10.1080/08982110108918672 douglas, c.m. (2003). introduction to statistical quality control. new york, ny: john wiley publications. 35int. j. prod. manag. eng. (2014) 2(1), 23-36creative commons attribution-noncommercial 3.0 spain reduction in repair rate of welding processes by determination & controlling of critical kpivs. table 13. six sigma project control plan. sub process specification characteristic specification requirement measurement method who measures where recorded decision rule/ corrective action/reference documents qualification of wps asme secix & code of construction asme sec viii div.-i & supplementary requirements mechanical and radiography results approved laboratory pwi performance record sheets must be approved by the nde level-iii & client qualification of welder asme sec ix & code of construction asme sec viii div-i & supplementary requirements radiography results radiography lab test reports welder certificate hrd/sop-06 selection of wps & welders specification of the material asme sec viii div.-i & supplementary requirements radiography results nde leveliii personnel radiography test reports verification by level-iii or level-ii selection of consumable asme sec-ii part-c wps & qpr chemical & mechanical results internal inspector & testing lab accepted material reports qa&qc/ms-01 welding execution asme sec v-iii div.-i & asme sec-vi drawings & client specifications visual & radiography results welding engineer, nde level-ii & iii welder performance sheet inspection reports http://dx.doi.org/10.1080/00207540110073082 http://dx.doi.org/10.1080/08982110108918672 flaig, j.j. (2006). selecting optimal specification limits. quality technology & quantitative management, 3(2), 207-216. flaig, j.j. (2009). a unifying process capability metric. journal of industrial engineering and management, 2(1), 48-59. jacobson, j.m., johnson, m.j. (2006). lean and six sigma: not for amateurs. labmedicine 37(4):140-145. doi:10.1309/9lhb-9g96ahmt-9xg2 laureani, a., antony, j., douglas, a. (2010). lean six sigma in a call center: a case study. international journal of productivity and performance management, 59(8), 757-768. doi:10.1108/17410401011089454 mahesh, b.p., prabhuswamy, m.s. (2010a). improvement of quality awareness using six sigma methodology for achieving higher cmmi level. international journal of advance research in management, 1(1), 20-41. mahesh, b.p., prabhuswamy, m.s. (2010b). process variability reduction through statistical process control for quality improvement. international journal for quality research, 4(3), 193-203. plecko, a., vujica, h.n., polajnar, a. (2009). an application of six sigma in manufacturing company. advances in production engineering and management, 4, 243-254. ricardo, c., allen, t. t. (2003). an alternative desirability function for achieving ‘six sigma’ quality. quality and reliability engineering, 19 (3), 227-240. doi:10.1002/qre.523 linn, r.j., tsung, f., ellis, l.w.c. (2006). supplier selection based on process capability and price analysis. quality engineering, 18(2), 123-129. doi:10.1080/08982110600567475 sivasamya, r., santhakumaranb, a., subramanianc, c. (2000). control chart for markov-dependent sample size. quality engineering, 12(4), 593-601. doi:10.1080/08982110008962624 36 int. j. prod. manag. eng. (2014) 2(1), 23-36 creative commons attribution-noncommercial 3.0 spain yousaf, f. and ikramullah butt, s. http://dx.doi.org/10.1309/9lhb-9g96-ahmt-9xg2 http://dx.doi.org/10.1309/9lhb-9g96-ahmt-9xg2 http://dx.doi.org/10.1108/17410401011089454 http://dx.doi.org/10.1002/qre.523 http://dx.doi.org/10.1080/08982110600567475 http://dx.doi.org/10.1080/08982110008962624 pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering received: 2021-11-14 accepted: 2022-04-07 packaging design for competitiveness. contextualizing the search and adoption of changes from a sustainable supply chain perspective jesus garcía-arca a1*, a. trinidad gonzález-portela garrido a2, j. carlos prado-prado a3, iván gonzález-boubeta a4 a universidade de vigo, grupo de ingeniería de organización (gio), escuela de ingeniería industrial, 36310 vigo, spain. a1* jgarca@uvigo.es, a2 tgonzalez-portela@uvigo.es, a3 jcprado@uvigo.es, a4 ivangonzalezboubeta@uvigo.es abstract: the “sustainable packaging logistics” (spl) approach seeks sustainable integration of the combined “packaging-product-supply chain” system orientated to increase competitiveness. however, characterizing which changes make it possible to guide such design in each company and supply chain is an aspect that has not been covered in the literature from different supply chain perspectives. the main goal of this paper is to identify and justify the main actions for improvement in spl, combined with a proposal of methodology for contextualizing, selecting and implementing each of these potential actions, applying the “action research” approach. likewise, this paper illustrates the interest of this methodology with its adoption in four different companies and supply chains. this paper opens up new avenues of applied research in packaging design, generating knowledge that contributes to sustainable and competitive improvement. key words: packaging, sustainability, logistics, supply chain. 1. introduction the rational, fair and balanced use of the planet’s scarce resources is a growing concern in today’s society, a concern that companies should reconcile with their natural and legitimate search for efficiency and profitability. thus, the adoption in the management of business activities of a sustainable approach (in its triple perspective, economic, environmental and social) is no longer a voluntary matter but has become an unavoidable requirement of competitiveness. logically, in this general context, among all these business activities, productive and logistical ones stand out and a true reflection of this is that the concept of a sustainable supply chain arouses ever more interest in business and academic forums (nilsson & christopher, 2018). in the promotion of this sustainability in companies and supply chains, one of the elements that cuts across the board to generate more impacts is packaging design. thus, depending on how this design is developed, a greater or lesser impact can be generated, not only at the commercial level, but also at the level of inefficiencies or losses (“waste” in terms of lean management or kaizen cultures) in the different productive and logistic processes throughout the supply chain. this waste includes product breakages, setups and rejects in packing processes, large clearances in packaging (including load units) or excessive consumption of raw materials (and excessive generation of waste) (hellström & saghir, 2007; svanes et al., 2010; azzi et al., 2012; sohrabpour et al., 2012, 2016; wikström et al., 2019; garcíaarca & prado-prado, 2020). to cite this article: garcía-arca, j., gonzález-portela garrido, a.t, prado-prado, j.c., gonzález-boubeta, i. (2022). packaging design for competitiveness. contextualizing the search and adoption of changes from a sustainable supply chain perspective. international journal of production management and engineering, 10(2), 115-130. https://doi.org/10.4995/ijpme.2022.16659 https://doi.org/10.4995/ijpme.2022.16659 int. j. prod. manag. eng. (2022) 10(2), 115-130creative commons attribution-noncommercial-noderivatives 4.0 international 115 http://creativecommons.org/licenses/by-nc-nd/4.0/ to complete this strategic importance, it should also be noted that, in many cases, the packaging itself is a relevant source of innovations (garcía-arca et al., 2019; hellström & nilsson, 2011; lindh et al., 2016; verghese & lewis, 2007; vernuccio et al., 2010). in this context, two decades ago the term “packaging logistics” was coined in academic circles. it sought efficient integration of the combined system “packaging-product-supply chain” orientated to gain competitive advantages (saghir, 2002; regattieri et al., 2019); that framework was later extended to promote improved behavior in sustainability in firms and supply chains with the “sustainable packaging logistics” approach or spl (garcía-arca et al., 2014a). beyond the definition of its fundamental pillars, recent literature has identified some examples of actions for improvement related to packaging design. some of these sources illustrate these actions through the analysis of case studies (for example, hellström & nilsson, 2011; pålsson et al., 2013; kye et al., 2013; bertoluci et al., 2014; garcía-arca et al., 2014b; molina-besch & pålsson, 2020) or exploratory studies (for example, pålsson & hellström, 2016; garcía-arca et al., 2017, 2019; coelho et al., 2020). however, in the literature there is not a methodology that allows the interest and priority of each improvement action to be contextualized in a more general and universal perspective. therefore, in this paper, the first goal is to identify and justify the main potential actions for improvements in the realm of spl. to achieve this goal, a literature review is carried out, combined with a proposal of methodology for contextualizing, selecting and implementing each of these potential actions, by using the “action research” approach (näslund et al., 2010; coughlan et al., 2016; prado-prado et al., 2020). the paper also goes beyond the purely theoretical to show that potential by developing this methodology in four different companies and supply chains. 2. systematizing the adoption of improvements in packaging design the many faces and the impacts of the design restrictions and aspects that packaging must satisfy can be understood under a sustainable perspective (economic, environmental, and social). those design restrictions and aspects are placed on very different planes and have very different needs. they include, for example, marketing, protective considerations, production, logistics, purchasing, the environment, ergonomics, legal aspects, or communication (azzi et al., 2012; garcia-arca et al., 2019, 2021b; lindh et al., 2016; molina-besch & pålsson, 2020; pålsson & sandberg, 2020, 2021; rundh, 2016). nevertheless, it is not enough just to understand and internalize the many requirements that packaging must satisfy, because it is also important to know the structure of the packaging system, which is often organized in three interconnected layers: primary, secondary, and tertiary (garcía-arca et al., 2020). in this way, it is possible to distinguish a first layer (primary packaging) that is directly in contact with the product; a secondary packaging, which groups several units of primary packaging (for example, a box); and a tertiary packaging, which groups together several units of secondary packaging in order to typically facilitate handling, transport and storage operations (for example, a pallet). this structure endows the packaging with a systemic character, which means that its performance must be analyzed from an integrated point of view of the whole product-packaging-supply chain, since the design requirements are not distributed homogeneously in each one of the layers, affecting each productionlogistic process, company or organization throughout the supply chain in a different way. simultaneously, the decisions of the packaging design process (dimensions, materials, aesthetic, etc.) are linked to the area, department or company that is especially affected by them, highlighting the importance of a suitable organizational structure based on multifunctional teams for applying an integrated vision of design. therefore, this organizational structure, based on multifunctional teams, should include people from different departments (commercial, productive, logistic, quality and so on) and companies (packaging suppliers, factories, retailers, 3pl or third party logistics, packing machines suppliers, etc.) affected by packaging design. likewise, the organizational structure should develop a “dynamic” perspective in this design process (olander-roese & nilsson, 2009). when such “dynamic” integration is envisaged, not only is it easier for numerous design options to appear, but also the decision-making method becomes more int. j. prod. manag. eng. (2022) 10(2), 115-130 creative commons attribution-noncommercial-noderivatives 4.0 international garcía-arca et al. 116 http://creativecommons.org/licenses/by-nc-nd/4.0/ objective, agile and consensual. for this reason, this coordinated and dynamic organizational approach could be considered a critical aspect that contributes actively to the implementation of a culture of continuous improvement in companies and supply chains, which undoubtedly also contributes to increasing competitiveness. this is what is proposed in the “sustainable packaging logistics” approach. thus, garcía-arca et al. (2014a) fit and extend the “packaging logistics” concept proposed by saghir (2002) to the scope of sustainability, developing the term “sustainable packaging logistics” (spl) with the following definition: “the process of designing, implementing, and controlling the integrated packaging, product and supply chain systems in order to prepare goods for safe, secure, efficient and effective handling, transport, distribution, storage, retailing, consumption, recovery, reuse or disposal, and related information, with a view to maximizing social and consumer value, sales, and profit from a sustainable perspective, and on a continuous adaptation basis”. what that means, therefore, is the integration of packaging design in the first stages of the product and supply chain design, with a “dynamic” perspective so that it can adapt at all times not only to the different perspectives of the design restrictions that each part of the supply chain appraises but also to the changes that can arise in the design restrictions themselves or in the environment (technological, legal, etc.). the new spl approach would radically shift how people see the way packaging-product-supply chain design is done towards a vision more in tune with “ecodesign” or the “circular economy”, in line with the un’s agenda 2030 aims (de koeijer et al., 2017a,b; wikström et al., 2019; molina-besch & pålsson, 2020). in order to deploy a spl strategy with impact in the three pillars of sustainability, three aspects appear as significant (garcía-arca et al., 2017): the identification of restrictions in the packaging design process; the adoption of an assessment system with a set of kpis (key performance indicators) for valuing and comparing different alternatives in packaging design; the implementation of an organizational structure (based on multifunctional teams) for identifying, testing and validating these alternatives. thanks to the adoption of these three pillars linked to spl, an innovative and “dynamic” vision of packaging design is implemented. this area of changes, innovations or “best practices” would include the lines for improvement which have already been successfully implemented by other companies (garcía-arca & prado-prado, 2008). in this context, the scientific literature provides these improvement lines, innovations or “best practices”. thus, a content analysis (seuring & gold, 2012) was used to carry out a bibliographic review to identify these improvements. this process involved an initial systematic search among papers in english from academic journals published from 2011 to 2021 and ranked in scopus. the search criteria were based on different combinations of terms related with three main ideas: first, sustainable packaging design; second, sustainable logistics or supply chains; third, improvement, innovation or “best practice”. the results of this initial search were refined through the abstract review. after selecting the papers with the greatest connection with the objective, a forward and backward snow-balling strategy was applied to prevent any relevant paper not being identified. this process led to a selection of 35 papers. after applying this strategy, a detailed analysis of these 35 papers was developed in order to define the main improvements, changes or “best practices” related to sustainable packaging design. table 1 summarizes nine alternatives of improvement with the references extracted from literature review. on the other hand, table 2 indicates the main potential advantages achieved with their implementation from a sustainable point of view; later, some of the most popular actions will be implemented in the four case studies and these changes are also indicated in table 2. however, these changes, improvements, innovations or “best practices”, although they may have been successfully implemented by some companies, cannot be adopted indiscriminately without prior understanding of the constraints and requirements of the products and the supply chain (contextualization). in this context, the authors propose a four-step methodology that adapts the proposal of garcíaarca et al. (2021a) for selecting the best range of boxes in a company, and includes the three pillars of spl. this methodology persues the suitable int. j. prod. manag. eng. (2022) 10(2), 115-130creative commons attribution-noncommercial-noderivatives 4.0 international packaging design for competitiveness. contextualizing the search and adoption of changes from a sustainable supply chain perspective 117 http://creativecommons.org/licenses/by-nc-nd/4.0/ contextualization of the needs, restrictions and requirements of each company and supply chain in order to value the interest and priority of each potential implementation. in this process of analysis and implementation the researchers participate directly under an “action research” approach, coordinating the adoption of the actions through multifunctional work teams. through this direct involvement, researchers can not only know first-hand the problem to be solved, but are also in a better position to provide an external, clean and different vision of the needs and possibilities of each potential improvement. along these lines, the role of these researchers is also to generate reflection and knowledge that can be extrapolated or adapted to other problems and environments. thus, the four steps of the methodology are: step 1. structuring the process of packaging design. this step includes the definition of a multifunctional work team for contextualizing needs and priorities and identifying design requirements. this team is coordinated by the researchers following the “action research” approach. this team is also responsible for proposing an assessment system for evaluating design alternatives from a sustainable perspective (particularly, economic and environmental; for example, costs, filling rate in packaging and load units, environmental impact, waste reduction). table 1. different changes/innovations for sustainable improvement in packaging design (source: authors). change or innovation main mentions in references according to literature review 1. dimensions wever (2011); grönman et al. (2013); wever & vogtländer (2013); accorsi et al. (2014); bertoluci et al. (2014); garcía-arca et al. (2014, a, b); gámez-albán et al. (2015); pålsson & hellström (2016); garcía-arca et al. (2017); zhao et al. (2017); mumani & stone (2018); garcía-arca et al. (2019); lu et al. (2020); coelho et al. (2020); molina-besch & pålsson (2020); garcía-arca et al. (2020); garcía-arca et al. (2021, a, b) 2. amount of product per packaging grönman et al. (2013); pålsson & hellström (2016); garcía-arca et al. (2017); krishna et al. (2017); mumani & stone (2018); garcía-arca et al. (2019) 3. packaging materials hellström & nilsson (2011); sohrabpour et al. (2012); albretch et al. (2013); pålsson et al. (2013); wever & vogtländer (2013); accorsi et al. (2014); bertoluci et al. (2014); garcía-arca et al. (2014, a, b); regattieri et al. (2014); pålsson & hellström (2016); garcía-arca et al. (2017); licciardello (2017); pålsson et al. (2017); garcía-arca et al. (2019); lu et al. (2020); coelho et al. (2020); molinabesch & pålsson (2020); garcía-arca et al. (2020) 4. change in the way of packing hellström & nilsson (2011); pålsson et al. (2013); wever & vogtländer (2013); bertoluci et al. (2014); faccio et al. (2015); garcía-arca et al. (2017); mumani & stone (2018); garcía-arca et al. (2019) 5. number of primary units per secondary or tertiary packaging hellström & nilsson (2011); sohrabpour et al. (2012); grönman et al. (2013); kye et al. (2013); pålsson et al. (2013); garcía-arca et al. (2014, a, b); gámez-albán et al. (2015); mcdonald (2016); pålsson & hellström (2016); garcía-arca et al. (2017); pålsson et al. (2017); garcía-arca et al. (2019); lu et al. (2020); molinabesch & pålsson (2020); garcía-arca et al. (2020); garcía-arca et al. (2021, b); sternberg & denizel (2021) 6. standardization (dimensions, formats, qualities) grönman et al. (2013); kye et al. (2013); garcía-arca et al. (2014, a, b); regattieri et al. (2014); faccio et al. (2015); mcdonald (2016); garcía-arca et al. (2017); zhao et al. (2017); garcía-arca et al. (2019); garcía-arca et al. (2020); garcíaarca et al. (2021, a, b); sternberg & denizel (2021) 7. elimination of “overpackaging” (excesive materials, size and/or protection) hellström & nilsson (2011); garcía-arca et al. (2014, a, b); regattieri et al. (2014); krishna et al. (2017); licciardello (2017); pålsson et al. (2017); garcía-arca et al. (2019) 8. returnable packaging levi et al. (2011); albretch et al. (2013); kye et al. (2013); pålsson et al. (2013); accorsi et al. (2014); garcía-arca et al. (2017); bortolini et al. (2018); accorsi et al. (2019); garcía-arca et al. (2019); accorsi et al. (2020); coelho et al. (2020) 9. new graphic design gelici-zeco et al. (2012); sohrabpour et al. (2012); grönman et al. (2013); pålsson & hellström (2016); garcía-arca et al. (2017); krishna et al. (2017); mumani & stone (2018); garcía-arca et al. (2019) int. j. prod. manag. eng. (2022) 10(2), 115-130 creative commons attribution-noncommercial-noderivatives 4.0 international garcía-arca et al. 118 http://creativecommons.org/licenses/by-nc-nd/4.0/ step 2. searching for potential changes and improvements in packaging design. the team explores new packaging alternatives and these alternatives are selected from the general list of actions mentioned previously (see table 1). logically, the interest and priority of each action is based on the fulfillment of the packaging design requirements and the understanding of the supply chain and market needs (contextualization). step 3. proofs, selection and adoption. to ensure a correct selection and development of design options, the following four phases are followed: an evaluation of the different alternatives thanks to the kpis system adopted. “artisanal” proofs for new design options (with internal “artisanal” packaging samples) including new ways of product placement inside them.“industrial” proofs with packaging samples; these samples are supplied by packaging suppliers; after these two phases of proofs, the work team decides the best changes in packaging design to implement. step 4. monitoring and improvement; due to potential changes in initial design requirements (for commercial, technological, social, logistic or legal reasons, for example), it would be recommendable to establish a monitoring system to improve the initial packaging design for adapting it to the new needs. this step supports the “dynamic” perspective table 2. impact of changes/innovations regarding packaging design in sustainability (source: authors). change or innovation main impact on the economic axis of sustainability main impact on the environmental axis of sustainability main impact on the social axis of sustainability implementation in case studies 1. packaging purchase cost. logistics costs. packaging waste management costs consumption of raw materials and waste generation. energy consumption and contamination (transport) help in use, ergonomics, and dosage for the needs of different final customers a, b, c 2. d 3. packaging purchase cost. packaging waste management costs recyclability. valorization. consumption of raw materials, energy, and waste generation. energy consumption and contamination (transport) help in use. perception of product quality a, b, c 4. packing cost. cost of deteriorated products or rejections. waste management costs waste generation (in production). energy consumption in packing activities help in use and lengthening useful life of product. perception of product quality a, b 5. packaging purchase cost. logistics costs. packaging waste management costs. consumption point handling cost. consumption of raw materials and waste generation. energy consumption and contamination (transport) help in use, ergonomics, and dosage of product for the needs of different customers on the supply chain a, b 6. packaging purchase cost. production cost. waste management costs energy use (production). consumption of raw materials and waste generation. energy consumption and contamination (transport) communication (“visibility”). perception of product quality b, c, d 7. packaging purchase and logistics costs. packaging waste management costs. handling cost. help in use. perception of product quality a, b, c, d 8. packaging purchase cost. logistics costs. packaging waste management costs help in use. perception of product quality d 9. packaging purchase cost consumption of raw materials and waste generation help in use. communication (“visibility”). perception of quality a, b int. j. prod. manag. eng. (2022) 10(2), 115-130creative commons attribution-noncommercial-noderivatives 4.0 international packaging design for competitiveness. contextualizing the search and adoption of changes from a sustainable supply chain perspective 119 http://creativecommons.org/licenses/by-nc-nd/4.0/ commented previously, closing a pdca cycle. in this final step should participate again the initial work team, including the researchers. in the next four sections, the most relevant aspects of the implementation of the proposed methodology in four companies are presented. in order to obtain a broad view of the applicability of the methodology, the authors have selected companies from different sectors and supply chains, including two in the retail sector (case a, a pizza maker and case b, a fishery product manufacturer) and two in the industrial sector (case c, a street furniture supplier and case d, an automotive supplier). in these teams, at least one of the authors participated directly, although with different involvement and duration depending on the company and the target of each project. 3. case a. pizza maker this food company is devoted to producing frozen products. its workforce is around 100 people and its yearly turnover is around 50 million euros. its main activity is frozen pizza production. the work team included two of the authors, the production manager, the logistics manager and a purchasing technician. the intervention process was divided into three different periods of six months each, spread over 6 years. the contextualization of the problem can be summarized in that the pizza has little added value and little density (a full pallet can weigh less than 200 kilos), which means logistics costs can be up to 25% of overall costs. likewise, a better pallet efficiency implies a reduction in environmental impact of transport. improvement of the pallet efficiency is thus a key factor for competitiveness for this reason, the basic kpi adopted to evaluate the improvement is the number of kg per pallet. initially, the packaging system at the firm was based on an individual cardboard box (35 mm. height; primary packaging), grouped 12 by 12 in corrugated cardboard boxes (secondary packaging); these boxes were palletized on eur pallets (49 boxes per pallet; 176.4 kg. per pallet; tertiary packaging). in order to improve this pallet efficiency some conceptual changes could be adopted, such as dimensions (1), packaging materials (3), packing system (4), number of primary packaging per each secondary packaging (5), elimination of “overpackaging” (7). additionally, any change in primary packaging could support changes in image and graphic design (9). however, other conceptual changes were discarded: the amount of product per primary packaging (2), as it affects the needs identified by the customers; the standardization (6), as the initial level was good and a greater level of standardization could affect the improvement of pallet efficiency; the adoption of a returnable secondary packaging (8), as the high number of collecting points in the retail supply chain of the firm makes it difficult to implement without increasing the reverse logistics costs. in the first period of intervention, after an analysis of product positioning in the primary packaging and some preliminary tests, it was concluded that its height could be lowered from 35 mm. to 31. this modification meant the secondary box dimensions could be adjusted to add another layer on the pallet (from 7 to 8 layers; see figure 1) without exceeding the height restrictions set by the supply chain. figure 1. initial situation (left) and improvement (right) in the pizza packaging system (source: authors). 31 int. j. prod. manag. eng. (2022) 10(2), 115-130 creative commons attribution-noncommercial-noderivatives 4.0 international garcía-arca et al. 120 http://creativecommons.org/licenses/by-nc-nd/4.0/ curiously, this height reduction demanded a change in some of the pizza topping ingredients; for example, whole olives were swapped for sliced olives. the combination of all these changes led to a better pallet capacity of 14.28% (from 176.4 kg. per pallet to 201.6 kg.), achieving an important logistics saving (including savings in the cost of both box types as they were smaller, as well as waste reduction from the boxes). some years after, in a second period of intervention, the work team focused on eliminating the initial cardboard box (secondary packaging) and substituting it for a shrink-wrap plastic pack. the alternative was to use the shrink-wrap pack as a grouping agent to substitute the secondary box, using a stronger individual box (primary packaging) of heavier cardboard (more expensive). at the same time, eliminating the initial cardboard box brought added economic benefits, not only because of the better use of pallets but also because the costs associated with it were eliminated (savings that were greater than the cost of the shrink-wrap pack, amortization of the shrink-wrapping equipment and even the increased environmental costs of the plastic used). at first, a study of the plastic pack was carried out with the individual boxes placed horizontally in the palletization pattern (the initial distribution; boxes stacked in the plastic pack, see figure 1). however, improvements in pallet efficiency were not found. later, the potential of vertical palletization of these individual boxes was studied. this study enabled improvement in not only the efficiency of pallet (13.67%) and strength in the individual box compression (which allows changes in the type of raw material in the primary packaging and its cost), but also the stability of the pallet. however, this last proposal presented, a priori, two major drawbacks: the pizzas could bend if the temperature in any step of the supply chain dropped too much and the topping could fall off because of the movement and handling during transport. both incidents could lead to deterioration in the image of the product on the market. after tests with individual boxes placed vertically, it was seen that the new layout was possible as long as the cold chain was maintained throughout the whole distribution process and the tension in the shrinkwrap plastic was also maintained in the packing process. once these two key factors were ensured, it was feasible to implement the newly proposed format with an overall saving in materials and logistics costs (with respect to the initial packaging system) of 45%. a third period of intervention took place some years after with the approval of the commercial department. for some products with high rotation, the number of individual boxes (primary packaging) in each shrink-wrap pack was increased (depending on the product, it was increased from 6 to 8 boxes per pack or from 8 to 10 boxes per pack). simultaneously, in the individual box, the height was adjusted, by which pallet capacity was increased by up to 14% (on the improvements already obtained in the two previous stages). in addition, the company took advantage of each period of intervention to update the image and graphic design that appeared on the primary packaging. the evolution of the packaging system for this product family is a good example of how the design process must be tackled dynamically to adapt to the needs and opportunities that arise in the environment and in the market. 4. case b. fishery product manufacturer the second case focuses on a food firm, specialized in production of pre-cooked products derived from fish and cephalopods. this firm processes around 30 000 tons a year. the work team included two of the authors, the firm director, the industrial manager, the purchasing manager and a production technician. the intervention process was divided into three periods of six months each, spread over five years. initially, the packaging system was based on two types of boxes: an individual one (primary packaging) and a cardboard box (secondary packaging). the logistics systems used is based on eur pallets (800×1200 mm, tertiary packaging). the contextualization of the problem can be summed up as the general need to reduce production and logistics costs in order to be more competitive. this double target implies working simultaneously in two directions: higher packing process efficiency and int. j. prod. manag. eng. (2022) 10(2), 115-130creative commons attribution-noncommercial-noderivatives 4.0 international packaging design for competitiveness. contextualizing the search and adoption of changes from a sustainable supply chain perspective 121 http://creativecommons.org/licenses/by-nc-nd/4.0/ higher pallet efficiency. likewise, both directions lead to a reduction in environmental impact of production and transport (energy consumption and pollution generation). for this reason, the basic kpis adopted to evaluate the improvement were the number of kg. per pallet and the reduction of hours in setups. in order to move in these two directions some conceptual changes could be adopted. from a pallet efficiency perspective, most of these changes were related to dimensions (1), number or primary packaging per each secondary packaging (5) or elimination of “overpackaging” (7). on the other hand, from a productive perspective, most changes could be related to the packing system (4), with or without changes in materials (3) and, particularly, standardization (6), as a reduction of packaging formats implies a reduction of setups in packing machines or lines. likewise, any change and update in primary packaging could support changes in image and graphic design (9). again, two conceptual changes were discarded by the work team for similar reasons as case a: the amount of product per primary packaging (2), and the adoption of a returnable secondary packaging (8). in the first period of intervention, the firm focused on improving pallet efficiency, substituting the cardboard boxes that grouped the primary boxes with shrink-wrapped plastic packs and, for some products, increasing the number of primary packaging per each secondary packaging. thanks to these actions, an average improvement in the quantity of kg per pallet of 22% was achieved. as a whole, this generated logistics savings (handling, storage, and transport) and materials savings (cardboard), which far exceeded the cost of the new wrapper (including the plastic material themselves and a stronger quality of individual boxes). some years after, in the second period of intervention, the work team proposed going forward in packing process efficiency with the standardization of the dimensions of the primary packaging bases (individual boxes) because, by doing so, they could reduce the number of tools (“dies”) used for conforming and closing in the packing process, in order to optimize the number of line setups. therefore, this standardization meant improvements in production performance as it reduced the number of stops on the manufacturing line (which had previously caused a bottleneck there) without decreasing, but rather increasing, the flexibility of the lines. initially, the firm had two production lines (a and b) devoted to producing six different products (two products in line a, and four products in line b) with six different dies (a different die base for each product). after the trials, standardizing the bases of the individual boxes (primary packaging) was seen to be viable. in order to carry out the standardization proposed, the possibility of rapidly adjusting the height of the box closure without changing the die was taken advantage of. however, the dimensional changes in primary packaging that allow a higher level of standardisation (with greater production efficiency by reducing the number of setups required), could also be associated with losses in the overall efficiency of palletisation (given that standard primary packaging may not adapt as well volumetrically to the product as a more specific one). logically, these dimensional design decisions also condition the total amount of cardboard consumed (and its associated waste) in each primary packaging. therefore, in this analysis, different dimensional alternatives arise with their pros and cons in terms of production and palletising efficiency. to illustrate this complexity, table 3 shows two of the alternatives contemplated by the work team in which the level of standardisation achieved (with its reduction in annual setup hours compared to the initial situation), the range of improvement in palletisation (compared to the initial situation) and the total savings achieved (productive and logistical) can be seen. after evaluating the different alternatives and carrying out the tests described in the methodology, the work team decided in favour of the second of the options described in table 3, which not only allows a higher level of overall savings, but also a higher level of production flexibility, as only one type of die base is needed to be able to package any of the six products involved, regardless of the line on which it is produced. in some cases, redimensioning the primary packaging could also involve changes in the placement of the product. figure 2 shows an example with hake slices int. j. prod. manag. eng. (2022) 10(2), 115-130 creative commons attribution-noncommercial-noderivatives 4.0 international garcía-arca et al. 122 http://creativecommons.org/licenses/by-nc-nd/4.0/ in which, thanks to the new positioning, logistics savings were achieved (improved pallet occupation by 12%) with additional savings in materials consumption (less cardboard was used and less waste generated). some years after the implementation of the standardization program, in the third period of intervention, the firm once again re-thought its packaging system to return to the corrugated cardboard box, adjusting primary packaging to avoid losses in pallet efficiency, but without affecting die standardization. figure 2. improvement in hake slice. initial situation (top) and improvement (bottom) (source: authors). this new approach responded to several considerations: on the one hand, to increased production line automation that allowed adjustment of slack in the boxes; and on the other hand, to increase automation in internal and external storage (the shrink-wrap pack presented more rejections when handled automatically); finally, it responded to increased pressure from the firm’s main customers to reduce plastic use for environmental reasons. as for the previous case, the company took advantage of each period of intervention to update the image and the graphic design that was visible on the primary packaging. the changes of the packaging system for these products illustrates again how the design process must be tackled dynamically. 5. case c. street furniture supplier the third case is developed in a firm that manufactures street furniture and games for children and adults with a wide diversity of components and sets, both standard and tailored. it presents a yearly turnover of around 12 million euros with a growing importance in international sales. the work team included one of the authors, the operations director, and a purchasing technician. the intervention process lasted four months. the initial packaging system included 10 formats of cardboard box and different types of large wooden crates used for exporting equipment and sets (see figure 3). the selection of packaging in each shipment depended on the type and number of products, as well as the final destination. inside these wooden crates, cardboard sheets were used to protect and separate the different components. transport used eur pallets (800×1200 mm.) and intermodal containers for export. the contextualization of the problem can be summed up as the general need to reduce packaging costs and logistics costs (improving space occupation) in order to be more competitive in an international context, but ensuring product protection. likewise, both costs strategies lead to a reduction in the environmental impact of each shipment (energy consumption and pollution generation). for all these reasons, the basic kpis adopted to evaluate the 2 1 0 .0 0 (d e ) 3 4 8 .0 0 (d e ) 1 2 7 .0 0 (d e ) 7 6 2 .0 1 1 7 1 .0 1 8 2 4 .0 2 1 0 .0 0 (d e ) 3 4 8 .0 0 (d e ) 1 2 7 .0 0 (d e ) 7 6 2 .0 1 1 7 1 .0 1 8 2 4 .0 1 9 7 .0 0 (d e ) 2 6 2 .0 0 (d e ) 1 8 7 .0 0 (d e ) 7 8 8 .0 1 1 8 0 .0 2 0 1 4 .0 1 9 7 .0 0 (d e ) 2 6 2 .0 0 (d e ) 1 8 7 .0 0 (d e ) 7 8 8 .0 1 1 8 0 .0 2 0 1 4 .0 table 3. comparison of annual savings according to the number of dies. productive line different dimensional bases in dies range of improvement in pallet efficiency (different products involved) annual reduction in hours of setups annual savings proposal 1 line a 1 10%-19% 300 73,000 euros line b 1 9%-21% 300 proposal 2 (final implementation line a 1 4%-15% 900 85,000 euros (only one dimensional base in dies on both lines)line b 5%-17% int. j. prod. manag. eng. (2022) 10(2), 115-130creative commons attribution-noncommercial-noderivatives 4.0 international packaging design for competitiveness. contextualizing the search and adoption of changes from a sustainable supply chain perspective 123 http://creativecommons.org/licenses/by-nc-nd/4.0/ improvement were the packaging material cost per shipment and the filling rate in boxes, pallets and containers. in order to go forward in these two directions some conceptual changes could be adopted: from dimensions (1) and materials (3) to standardization (6) and elimination of “overpackaging” (7). however, due to the different products involved in each shipment (type and number), no special analysis was needed to put a value on changes in the amount of product per packaging (2) or the number of primary packaging per each secondary packaging (5). likewise, due to the low number of shipments and the high diversity of products, no alternative for easy and efficient automation in the packing process was developed, although some new methods or criteria in manual packing could be considered (4). on the other hand, due to the various international destinations of shipments, the implementation of a returnable system for secondary packaging was discarded (8). finally, as the company operates in the industrial sector, image and graphic design are not so critical from a commercial perspective (9). figure 3. example of initial wooden crate (source: authors). an analysis made by the work team highlighted generalized poor use of volumetric space in the cardboard boxes (depending on the product being packaged, this wasted space could vary between 20 and 60 per cent; see an example in figure 4). simultaneously, it was noted that the wooden crates were too heavy and strong for the function and use the firm needed from them. finally, it was identified that organizationally there were no clear criteria for selecting the packaging and placing the components in it, which did not facilitate good use of the packaging resources. figure 4. examples of the poor use of volumetric space in boxes (source: authors). in this context, the firm has redimensioned the cardboard boxes and also reduced the number of formats (from 10 types to 5). this format standardization meant a 50% reduction. this twofold change allowed the firm to reduce the cost of purchasing boxes (increased buying volume and more efficient boxes). likewise, it has improved the criteria for selecting and for placing the components and sets, both in the cardboard boxes and in the wooden crates. the wooden boxes have been structurally redesigned to “lighten” both their weight and cost (adjusting the board type and number). in parallel, the size of the cardboard sheets has been adjusted to the size of the crate. logically, all these changes have been applied without affecting the performance of the packaging in terms of protection and logistics. with the implementation of all these actions, savings of 33% have been achieved in cardboard box purchases, 30% in wooden crates, and 12% in the cardboard sheets in each shipment. at the same time, the improved use of the space within the boxes and crates (15% in average) has also led to major savings in terms of handling and, particularly, transport. 6. case d. automotive supplier the last case was developed in a firm making components for the automotive sector, specialized in the manufacture and integration of specific plastic and metal parts that together form the framework for a vehicle seat or some of its components. the int. j. prod. manag. eng. (2022) 10(2), 115-130 creative commons attribution-noncommercial-noderivatives 4.0 international garcía-arca et al. 124 http://creativecommons.org/licenses/by-nc-nd/4.0/ spanish factory, where the project is developed, has a workforce of more than 500 and supplies components to 10 car assembly factories in spain. the work team included one of the authors, the production director, the quality director and a logistics technician. the intervention process was developed over a period of six months. the initial packaging system included returnable “box-pallets” (type unit, measuring 1000×1200 mm. the base and 980 mm. the height), plastic boxes and cardboard boxes (modular system 600×400 mm. in bases of boxes). the aim of rationalizing the packaging system was to improve the efficiency and sustainability of the operations undertaken during handling (wrapping and unwrapping), storage and transport. transport was based on the american pallet (1000×1200 mm.). likewise, any change in the packaging system should ensure product protection and quality. the basic kpi adopted to evaluate the improvement was the filling rate in boxes and pallets. in order to go forward in this aim, the company wanted to maintain the packaging system adopted previously (both returnable (8) and cardboard) without changing dimensions (1), materials (3) and packing process (4), as this initial packaging system was also used by other factories outside spain. so, in all factories of the company, a standardization program of the packaging system had been developed previously (6). in this context, other conceptual changes could be adopted: amount of product per packaging (2) and elimination of “overpackaging” (7) in terms of over protection. as there was no primary packaging in the system, the alternative 5 (increasing the number of primary packaging per secondary packaging) was not feasible. finally, as the company operates in the industrial sector, image and graphic design are not so critical from a commercial perspective (9). after the analysis and trials, in the packaging system for small pieces, improvements of between 20 and 50% of pieces per box were implemented, depending on the product. the main line of improvement work was to rethink how parts were placed in the box (from flat piles to vertically on edge; see an example in figure 5). this was done, furthermore, without questioning the dimensions of the parts or of the returnable plastics boxes (for local or regional customers) or the cardboard box (for international customers) with a base size of 600×400 mm. figure 5. improvement in placement of small component inside the box. initial situation (left) and improved situation (right) (source: authors). at the same time, in the more voluminous components (the main seat framework structure) there was also an increase in the quantity of components for each palletized unit (“box-pallet”). this improvement varied between 9% and 20%, depending on the product. again, it was only necessary to redesign how the products were placed within the “boxpallet” (improving the way the parts fitted together; see figure 6). these two examples were developed without altering the main dimensions of the products while at the same time, of course, ensuring product protection and quality. figure 6. improvement in placement of main structures inside the “box pallet”. initial situation (left) and improved situation (right) (source: authors). 7. discussion deploying the sustainable packaging logistics (spl) approach can actively contribute to competitive improvement in supply chains and, by extension, each and every company comprising them, regardless of whether they are located in the industrial or the consumer field. this deployment should be done from a sustainable and dynamic standpoint. when any supply chain (industrial or retail) is analyzed in detail, highly diverse initial situations int. j. prod. manag. eng. (2022) 10(2), 115-130creative commons attribution-noncommercial-noderivatives 4.0 international packaging design for competitiveness. contextualizing the search and adoption of changes from a sustainable supply chain perspective 125 http://creativecommons.org/licenses/by-nc-nd/4.0/ in the packaging system can appear. thus, in some companies the changes are produced in a conscious and structured way, founded on prior experience and knowledge of the repercussions of certain decisions when designing the packaging, motivated by changes in commercial, protection, productive and/ or logistics requirements. in this way, companies and supply chains adapt their packaging design process for deploying efficiency and sustainability, behaving like “learning organizations”. this learning process is the focus of the proposed methodology, i.e., to understand and contextualise the interest of adopting certain changes or innovations that do not always adapt to the needs of each company and supply chain. however, in many other companies, the relationship between packaging, product and supply chain may have been created in an anarchic, random, or unconscious fashion. logically, in such cases, it is difficult for all the proposed alternatives to contemplate a complete, multifunctional, and objective vision of the design requirements, or an overall evaluation of the effect of some resolutions on the sustainability (and efficiency) of the supply chain they form a part of. some of these changes, supposedly designed as improvements, may, depending on the case, be counterproductive; in fact, many of the potential changes are interrelated, in some cases fostering the emergence of synergies and, in other cases, the emergence of negative impacts. in order to promote the deployment of the spl approach in companies, it would be necessary for one of the parts in the supply chain to act as its leader. that is the role of the researchers, integrated in the work teams to act as agents of change (“action research”). as a result of this evolution in the perception and interpretation of spl, a working framework has been obtained in which solutions combining product, packaging and supply chain are identified and evaluated with a double perspective of efficiency and sustainability. logically, the diffusion of best practices or innovations (tables 1 and 2) that have already been implemented with good results in other firms or in other sectors will help to enlighten, incentivize, and motivate other companies and chains, which means swifter organizational learning. as commented previously, however, the diffusion of successful best practices, even though this may motivate the search for new alternatives in other organizations, is not, in itself, necessarily associated with their being implemented in a sure way. this claim is based on the fact that not all companies or chains are equal and, therefore, neither the design requirements nor the costs are equally important. in this context, it is critical to adopt a good system for measuring and evaluating the alternatives. nevertheless, one of the main difficulties when structuring and implementing a method for measuring and evaluating is associated with the problem of weighting the design requirements in an objective way and on the same scale (for example, costs, environmental impact through techniques such as life cycle assessment (lca), the customer’s perception of quality). in practice, the above problem implies a need to combine different methods simultaneously (qualitative and/or quantitative). this process of measurement and evaluation is another key resposibility of the work team in the proposed methodology. in this field, in the four cases described here, a measurement system based mainly on the objective measurement of costs has been used. however, this is in turn enriched indirectly by the results from other scales or metrics that are connected to the costs, particularly from an environmental perspective. for example, if the amount of material is reduced in the packaging system, (and, therefore, the quantity of waste is also reduced), the purchase cost of the material is improved; on the other hand, if palletizing efficiency is improved (with more product per pallet), there is an improvement in the firm’s environmental behavior as well as in the costs of transport, handling and storage, because fewer vehicles are used for distribution (less fuel is needed), and the contamination associated with those vehicles is also reduced. the challenge facing organizations and chains is to shift from a passive approach in their packaging system design (a packaging alternative is launched without contemplating all its potential impacts) to a more proactive approach, that is, a launch that is, from the outset, as efficient and sustainable as possible, at least as far as the baseline restrictions at the time are concerned. int. j. prod. manag. eng. (2022) 10(2), 115-130 creative commons attribution-noncommercial-noderivatives 4.0 international garcía-arca et al. 126 http://creativecommons.org/licenses/by-nc-nd/4.0/ to make that shift, it would be recommendable to join packaging design into the product design itself (including its supply chain design), as in the approach proposed by spl and the methodology. to illustrate this integrated design, it can be recalled that in three of the cases analyzed (a, b and d), the implemented improvements included changes in the arrangement of the products or even the components (for example, the type of pizza topping in case a). in addition, this vision that integrates the design process should also be complemented by a vision that is dynamic, flexible, or capable of adapting to new conditioning or restricting factors in the commercial, logistics, legal or technological environment. this dynamic and continuous monitoring is again one of the roles of the work team. for example, the sphere of materials and equipment suppliers is a continual source of technical solutions and novelties that can and should be considered when it comes to seeking alternatives. examples of this continual adaptability can be found in cases a and b. at the same time, this dynamic vision takes on more relevance when the packaging system configuration to be used with the product varies according to the makeup of the order itself (case c). thus, this example exposes the need to search for the best combination of packages in each order, which can be selected from a previously designed range of alternatives. this final question is related to a technical issue faced today by many companies operating in e-commerce: what is the most ideal range of packaging options and what dimensions and features are associated with it? the solution to this issue is a future challenge facing the retail market from a perspective of logistics efficiency and sustainability. from among all the potential changes described in tables 1 and 2, a special mention should be made of the impact of standardization on the packaging systems of some companies (cases b, c and d), which presented opportunities for improvement that are not only in production (such as those given in the packaging process in case b) but also in purchasing (case b again, but also with the packaging rationalization in case c). additionally, this rationalization is key for the proposal of other alternatives that have, a priori, less environmental impact, such as the returnable packaging in case d. 8. conclusions this paper has presented different changes or innovations related to packaging design that can contribute to improving the efficient and sustainable management of supply chains and, in short, improve their overall competitiveness. going beyond that, this paper has proposed a methodology for contextualizing, selecting and implementing each of these potential changes or innnovations, applying the “action research” approach. in order to illustrate its applicability, this methodology has been applied in four different case studies. these case studies present different perspectives of the retail and industrial sectors from a “dynamic” point of view. from a scientific point of view, this article is interesting because it opens up new avenues of research in the design (and management) of packaging. on the other hand, from an applied point of view, this article may be useful for companies and supply chains to understand and to apply actions related to packaging that promote sustainable and competitive improvement. likewise, the “action research” approach proposed in the methodology to develop, in a collaborative way, the packaging redesign in the four companies can also be mentioned as new and innovative. thanks to this collaboration, researchers and practitioners can generate useful knowledge in the context of packaging design. references accorsi, r., cascini, a., cholette, s., manzini, r., & mora, c. (2014). economic and environmental assessment of reusable plastic containers: a food catering supply chain case study. international journal of production economics, 152, 88-101. https://doi.org/10.1016/j.ijpe.2013.12.014 accorsi, r., baruffaldi, g., manzini, r., & pini, c. (2019). environmental impacts of reusable transport items: a case study of pallet pooling in a retailer supply chain. sustainability, 11, 3147. https://doi.org/10.3390/su11113147 int. j. prod. manag. eng. (2022) 10(2), 115-130creative commons attribution-noncommercial-noderivatives 4.0 international packaging design for competitiveness. contextualizing the search and adoption of changes from a sustainable supply chain perspective 127 https://doi.org/10.1016/j.ijpe.2013.12.014 https://doi.org/10.3390/su11113147 http://creativecommons.org/licenses/by-nc-nd/4.0/ accorsi, r., baruffaldi, g., & manzini, r. (2020). a closed-loop packaging network design model to foster infinitely reusable and recyclable containers in food industry. sustainable production and consumption, 24, 48-61. https://doi.org/10.1016/j.spc.2020.06.014 albrecht, s., brandstetter, p., beck, t., fullana-i-palmer, p., grönman, k., baitz, m., & fischer, m. (2013). an extended life cycle analysis of packaging systems for fruit and vegetable transport in europe. the international journal of life cycle assessment, 18(8), 1549-1567. https://doi.org/10.1007/s11367-013-0590-4 azzi, a., battini, d., persona, a. & sgarbossa, f. (2012). packaging design: general framework and research agenda. packaging technology and science, 25(8), 435-456. https://doi.org/10.1002/pts.993 bertoluci, g., leroy, y., & olsson, a. (2014). exploring the environmental impacts of olive packaging solutions for the european food market. journal of cleaner production, 64, 234-243. https://doi.org/10.1016/j.jclepro.2013.09.029 bortolini, m., galizia, f.g., mora, c., bottia, l., & rosano, m. (2018). bi-objective de-sign of fresh food supply chain networks with reusable and disposable packaging containers. journal of cleaner production, 184, 375-388. https://doi.org/10.1016/j.jclepro.2018.02.231 coelho, p.m., corona, b., ten klooster, r., & worrell, e. (2020). sustainability of reusable packaging-current situation and trends. resources, conservation & recycling: x, 6, 100037. https://doi.org/10.1016/j.rcrx.2020.100037 coughlan, p., draaijer, d., godsell, j. & boer, h. (2016). operations and supply chain management: the role of academics and practitioners in the development of research and practice. international journal of operations & production management, 36, 1673-1695. https://doi.org/10.1108/ijopm-11-2015-0721 de koeijer, b., wever, r. & henseler, j. (2017a). realizing product-packaging combinations in circular systems: shaping the research agenda. packaging technology and science, 30(8), 443-460. https://doi.org/10.1002/pts.2219 de koeijer, b., de lange, j. & wever, r. (2017b). desired, perceived, and achieved sustainability: trade-offs in strategic and operational packaging development. sustainability, 9, 1923. https://doi.org/10.3390/su9101923 faccio m., gamberi, m., pilati, p., & bortolini, f. (2015). packaging strategy definition for sales kits within an assembly system. international journal of production research, 53,(11), 3288-3305. https://doi.org/10.1080/00207543.2014.975862 gamez albán, h.m., soto cardona, o.c., mejía argueta, c., & tullio sarmiento, a. (2015). a cost-efficient method to optimize package size in emerging markets. european journal of operational research, 241(3), 917-926. https://doi.org/10.1016/j.ejor.2014.09.020 garcía-arca, j. & prado-prado, j.c. (2008): packaging design model from a supply chain approach". supply chain management: an international journal, 13(5), 375-380). https://doi.org/10.1108/13598540810894960 garcía-arca, j., prado-prado, j.c. & gonzález-portela garrido, a.t. (2014a). packaging logistics: promoting sustainable efficiency in supply chains. international journal of physical distribution and logistics management, 44(4), 325-346. https://doi.org/10.1108/ijpdlm-05-2013-0112 garcía-arca, j., gonzález-portela, a. t., & prado-prado, j. c. (2014b). packaging as source of efficient and sustainable advantages in supply chain management. an analysis of briks. international journal of production management and engineering, 2(1), 15-22. https://doi.org/10.4995/ijpme.2014.1860 garcía-arca, j., gonzález-portela garrido, a.t. & prado-prado, j.c. (2017). sustainable packaging logistics. the link between sustainability and competitiveness in supply chains. sustainability, 9(7), 1098. https://doi.org/10.3390/ su9071098 garcía-arca, j., trinidad gonzález-portela, a. & carlos prado-prado, j. (2019). organizational best practices in packaging design. an analysis in perfumery and cleaning products. dirección y organización, 68, 5-19. https://doi.org/10.37610/ dyo.v0i69.547 garcía-arca, j., comesaña-benavides, j.a., gonzález-portela garrido, a.t. & prado-prado, j.c. (2020). rethinking the box for sustainable logistics. sustainability, 12, 1870. https://doi.org/10.3390/su12051870 garcia-arca, j., comesaña-benavides, j.a., gonzalez-portela garrido, a.t. & prado-prado, j. c. (2021a). methodology for selecting packaging alternatives: an "action research" application in the industrial sector. central european journal of operations research, 29, 1427-1446. https://doi.org/10.1007/s10100-020-00724-3 garcia-arca, j., gonzalez-portela garrido, a.t., prado-prado, j.c. & gonzález-romero, i. (2021b). estructurando el diseño de envases y embalajes para mejorar la sostenibilidad. evidencias empíricas en el sector de menaje. dirección y organización, 7, 60-79. https://doi.org/10.37610/dyo.v0i73.593 gelici-zeko, m.m., lutters, d., klooster, t.r., & weijzen, p. l. g. (2012). studying the influence of packaging design on consumer perceptions (of dairy products) using categorizing and perceptual mapping. packaging technology and science, 26(4), 215-228. https://doi.org/10.1002/pts.1977 int. j. prod. manag. eng. (2022) 10(2), 115-130 creative commons attribution-noncommercial-noderivatives 4.0 international garcía-arca et al. 128 https://doi.org/10.1016/j.spc.2020.06.014 https://doi.org/10.1007/s11367-013-0590-4 https://doi.org/10.1002/pts.993 https://doi.org/10.1016/j.jclepro.2013.09.029 https://doi.org/10.1016/j.jclepro.2018.02.231 https://doi.org/10.1016/j.rcrx.2020.100037 https://doi.org/10.1108/ijopm-11-2015-0721 https://doi.org/10.1002/pts.2219 https://doi.org/10.3390/su9101923 https://doi.org/10.1080/00207543.2014.975862 https://doi.org/10.1016/j.ejor.2014.09.020 https://doi.org/10.1108/13598540810894960 https://doi.org/10.1108/ijpdlm-05-2013-0112 https://doi.org/10.4995/ijpme.2014.1860 https://doi.org/10.3390/su9071098 https://doi.org/10.3390/su9071098 https://doi.org/10.37610/dyo.v0i69.547 https://doi.org/10.37610/dyo.v0i69.547 https://doi.org/10.3390/su12051870 https://doi.org/10.1007/s10100-020-00724-3 https://doi.org/10.37610/dyo.v0i73.593 https://doi.org/10.1002/pts.1977 http://creativecommons.org/licenses/by-nc-nd/4.0/ grönman, k., soukka, r., järvi-kääriäinen, t., katajajuuri, j. m., kuisma, m., koivupuro, h. k., ollila, m., pitkänen, m., miettinen, o. silvenius, f., thun, r., wessman, h. & linnanen, l. (2013). framework for sustainable food packaging design. packaging technology and science, 26(4), 187-200. https://doi.org/10.1002/pts.1971 hellström, d. & saghir, m. (2007). packaging and logistics interactions in retail supply chains. packaging technology and science, 20, 197-216. https://doi.org/10.1002/pts.754 hellström, d. & nilsson, f. (2011): logistics-driven packaging innovation: a case study at ikea. international jounal of retailing and distribution management, 39, 638-657. https://doi.org/10.1108/09590551111159323 hellström, d.,& olsson, a (2017). managing packaging design for sustainable development. wiley, london. https://doi. org/10.1002/9781119151036 krishna, a., cian, l., & aydınoğlu, n. z. (2017). sensory aspects of package design. journal of retailing, 93(1), 43-54. https://doi.org/10.1016/j.jretai.2016.12.002 kye, d., lee, j. & lee, k. (2013). the perceived impact of packaging logistics on the efficiency of freight transportation (eot). international journal of physical distribution & logistic management, 43(8), 707-720. https://doi.org/10.1108/ijpdlm-08-2011-0143 levi, m., cortesi, s., vezzoli, c., & salvia, g. (2011). a comparative life cycle assess-ment of disposable and reusable packaging for the distribution of italian fruit and vegetables. packaging technology and science, 24(7), 387-400. https://doi.org/10.1002/pts.946 licciardello, f. (2017). packaging, blessing in disguise. review on its diverse contribution to food sustainability. trends in food science & technology, 65, 32-39. https://doi.org/10.1016/j.tifs.2017.05.003 lindh, h., williams, h., olsson, a., & wikström, f. (2016). elucidating the indirect contributions of packaging to sustainable development: a terminology of packaging functions and features. packaging technology and science, 29, 225-246. https://doi.org/10.1002/pts.2197 lu, s., yang, l., liu, w., & jia, l. (2020). user preference for electronic commerce overpackaging solutions: implications for cleaner production. journal of cleaner production, 258, 120936. https://doi.org/10.1016/j.jclepro.2020.120936 mcdonald, c.m. (2016). integrating packaging and supply chain decisions: selection of economic handling unit quantities. international journal of production economics, 180, 208-221. https://doi.org/10.1016/j.ijpe.2016.08.003 molina-besch, k. & pålsson, h. (2020). a simplified environmental evaluation tool for food packaging to support decisionmaking in packaging development. packaging technology and science, 33(4-5), 141-157. https://doi.org/10.1002/ pts.2484 mumani, a., & stone, r. (2018). state of the art of user packaging interaction (upi). packaging technology and science, 31(6), 401-419. https://doi.org/10.1002/pts.2363 näslund, d., kale, r. & paulraj, a. (2010). action research in supply chain management-a framework for relevant and rigorous research. journal of business logistics, 31, 331-355. https://doi.org/10.1002/j.2158-1592.2010.tb00155.x nilsson, f., & christopher, m. (2018). rethinking logistics management towards a strategic mind-set for logistics effectiveness and innovation. emergence: complexity and organization, 20(2), 1-24. olander-roese, m., & nilsson, f. (2009). competitive advantage through packaging design? propositions for supply chain effectiveness and efficiency. in ds 58-1: proceedings of iced 09, the 17th international conference on engineering design, vol. 1, design processes, palo alto, ca, usa, 24.-27.08. pålsson, h., finnsgård, c., & wänström, c. (2013). selection of packaging systems in supply chains from a sustainability perspective: the case of volvo. packaging technology and science, 26(5), 289-310. https://doi.org/10.1002/pts.1979 pålsson. h., & hellström, d. (2016). packaging logistics in supply chain practice current state, trade-offs and improvement potential. international journal of logistics research and applications, 19(5), 1-18. https://doi.org/10.1080/1367556 7.2015.1115472 pålsson, h., pettersson, f., & winslott hiselius, l. (2017). energy consumption in e-commerce versus conventional trade channels insights into packaging, the last mile, unsold products and product returns. journal of cleaner production, 16, 765-778. https://doi.org/10.1016/j.jclepro.2017.06.242 pålsson, h., & sandberg, e. (2020). paradoxes in supply chains: a conceptual framework for packed products. the international journal of logistics management, 31(3), 423-442. https://doi.org/10.1108/ijlm-12-2019-0338 pålsson, h., & sandberg, e. (2021). packaging paradoxes in food supply chains: exploring characteristics, underlying reasons and management strategies. international journal of physical distribution & logistics management, 52(11), 1-28. https://doi.org/10.1108/ijpdlm-09-2019-0270 prado-prado, j.c., garcía-arca, j., fernández-gonzález, a.j., mosteiro-añón, m. (2020). increasing competitiveness through the implementation of lean management in healthcare. international journal of environmental research & public health, 17, 4981. https://doi.org/10.3390/ijerph17144981 int. j. prod. manag. eng. (2022) 10(2), 115-130creative commons attribution-noncommercial-noderivatives 4.0 international packaging design for competitiveness. contextualizing the search and adoption of changes from a sustainable supply chain perspective 129 https://doi.org/10.1002/pts.1971 https://doi.org/10.1002/pts.754 https://doi.org/10.1108/09590551111159323 https://doi.org/10.1002/9781119151036 https://doi.org/10.1002/9781119151036 https://doi.org/10.1016/j.jretai.2016.12.002 https://doi.org/10.1108/ijpdlm-08-2011-0143 https://doi.org/10.1002/pts.946 https://doi.org/10.1016/j.tifs.2017.05.003 https://doi.org/10.1002/pts.2197 https://doi.org/10.1016/j.jclepro.2020.120936 https://doi.org/10.1016/j.ijpe.2016.08.003 https://doi.org/10.1002/pts.2484 https://doi.org/10.1002/pts.2484 https://doi.org/10.1002/pts.2363 https://doi.org/10.1002/j.2158-1592.2010.tb00155.x https://doi.org/10.1002/pts.1979 https://doi.org/10.1080/13675567.2015.1115472 https://doi.org/10.1080/13675567.2015.1115472 https://doi.org/10.1016/j.jclepro.2017.06.242 https://doi.org/10.1108/ijlm-12-2019-0338 https://doi.org/10.1108/ijpdlm-09-2019-0270 https://doi.org/10.3390/ijerph17144981 http://creativecommons.org/licenses/by-nc-nd/4.0/ regattieri a., santarelli g., gamberi m., & mora, c. (2014). a new paradigm for packaging design in web-based commerce. international journal of engineering business management, 6, 1-11. https://doi.org/10.5772/58825 regattieri a., santarelli g., & piana f. (2019). packaging logistics. in: zijm h., klumpp m., regattieri a., heragu s. (eds) operations, logistics and supply chain management. lecture notes in logistics (pp. 273-303). springer, cham. https://doi.org/10.1007/978-3-319-92447-2_13 rundh, b. (2016). the role of packaging within marketing and value creation", british food journal, 118(10), 2491-2511. https://doi.org/10.1108/bfj-10-2015-0390 saghir, m. (2002). packaging logistics evaluation in the swedish retail supply chain, lund university, lund. seuring, s., & gold, s. (2012). conducting content-analysis based literature reviews in supply chain management. supply chain management: an international journal, 17(5), 544-555. https://doi.org/10.1108/13598541211258609 sohrabpour, v., hellström, d., & jahre, m. (2012). packaging in developing countries: identifying supply chain needs. journal of humanitarian logistics and supply chain management, 2(2), 183205. https://doi.org/10.1108/20426741211260750 sohrabpour, v., oghazi, p., & olsson, a. (2016). an improved supplier driven packaging design and development method for supply chain efficiency. packaging technology and science, 29, 161-173. https://doi.org/10.1002/pts.2194 sternberg, h.s., & denizel, m. (2021). toward the physical internet-logistics service modularity and design implications. journal of business logistics, 42(1), 144-166. https://doi.org/10.1111/jbl.12261 svanes, e., vold, m., møller, h., pettersen, m.k., larsen, h., & hanssen, o.j. (2010). sustainable packaging design: a holistic methodology for packaging design. packaging technology and science, 23(3), 161-175. https://doi.org/10.1002/pts.887 verghese, k. & lewis, h. (2007). environmental innovation in industrial packaging: a supply chain approach. international journal of production research, 45(18-19), 4381-440. https://doi.org/10.1080/00207540701450211 vernuccio, m., cozzolino, a., & michelin, i.l. (2010). an exploratory study of marketing, logistics, and ethics in packaging innovation. european journal of innovation management, 13, 333-354. https://doi.org/10.1108/14601061011060157 wever, r. (2011). design for volume optimization of packaging for durable goods. packaging technology and science, 24(4), 211-222. https://doi.org/10.1002/pts.927 wever, r. & vogtländer, j. (2013). eco-efficient value creation: an alternative perspective on packaging and sustainability. packaging technology and science, 26(4), 229-248. https://doi.org/10.1002/pts.1978 wikström. f., verghese, k., auras, r., olsson, a., williams, h., wever, r., grönman, k., pettersen, m. k., møller, h., & soukka, r. (2019). packaging strategies that save food: a research agenda for 2030. journal of industrial ecology, 23(3), 532-540. https://doi.org/10.1111/jiec.12769 zhao, c., johnsson, m., & he, m. (2017). data mining with clustering algorithms to reduce packaging costs: a case study. packaging technology and science, 30(5), 173-193. https://doi.org/10.1002/pts.2286 int. j. prod. manag. eng. (2022) 10(2), 115-130 creative commons attribution-noncommercial-noderivatives 4.0 international garcía-arca et al. 130 https://doi.org/10.5772/58825 https://doi.org/10.1007/978-3-319-92447-2_13 https://doi.org/10.1108/bfj-10-2015-0390 https://doi.org/10.1108/13598541211258609 https://doi.org/10.1108/20426741211260750 https://doi.org/10.1002/pts.2194 https://doi.org/10.1111/jbl.12261 https://doi.org/10.1002/pts.887 https://doi.org/10.1080/00207540701450211 https://doi.org/10.1108/14601061011060157 https://doi.org/10.1002/pts.927 https://doi.org/10.1002/pts.1978 https://doi.org/10.1111/jiec.12769 https://doi.org/10.1002/pts.2286 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j international journal of production management and engineering hybrid control of supply chains: a structured exploration from a systems perspective grefen, p., & dijkman r. school of industrial engineering, eindhoven university of technology p.w.p.j.grefen@tue.nl r.m.dijkman@tue.nl abstract: supply chains are becoming increasingly complex these days, both in the structure of the chains and in the need for fine-grained, real-time control. this development occurs in many industries, such as manufacturing, logistics, and the service industry. the increasing structural complexity is caused by larger numbers of participating companies in supply chains because of increasing complexity of products and services. increasing requirements to control are caused by developments like mass-customization, pressure on delivery times, and smaller margins for waste. maintaining well-structured strategic, tactic, and operational control over these complex supply chains is not an easy task – certainly as they are pressured by end-to-end synchronization requirements and just-in-time demands. things become even more complex when chains need to be flexible to react to changing requirements to the products or services they deliver. to enable design of well-structured control, clear models of control topologies are required. in this paper, we address this need by exploring supply chain control topologies in an organized fashion. the exploration is based on integrating a supply chain model and a control model in two alternative ways to obtain two extreme models for supply chain control. these two models are next combined to obtain a hybrid chain control model in which control parameters can be adapted to accommodate different circumstances, hence facilitating agility in supply chains and networks. we apply the developed model to a number of case studies to show its usability. the contribution of this paper is the structured analysis of the design space for chain-level control models not the description of individual new models. key words: supply chain, control model, information system. 1. introduction in the modern economy, we see the development of ever more complex products. a good example can be found in the automotive domain. here, we see that the complexity of automobiles has increased in a dramatic way: the inclusion of new features like safety systems, driver guidance and support systems, mobile entertainment systems, and air conditioning systems has increased the number of components in an average car significantly (maxton and wormald, 2004). similar developments can be observed in many high-tech industries, like the airplane industry and the computer industry. next to the increasing complexity of products, we also see an increasing variety of products: many products cannot be delivered anymore as a standard product that uniformly suits every customer. these products are customized for specific markets, for specific customer groups or even individual customers – giving rise to the development of mass customization in the past decades (smith, jiao and chu, 2012). we can find a good example in the automotive industry again where each individual car is delivered to the requirements of an individual customer. apart from the increasing complexity and variety of products, we also see that the pressure on performance of supply chains increases. delivery times need to be compressed to stay competitive in modern markets that behave increasingly in a real-time fashion. increasing competition and the advent of ‘green’ chains increase the pressure on waste reduction, for example to decrease the carbon footprint of a chain (hoen et al., 2012). this requires tight integration of a complex chain, e.g., in multimodal logistics (jansen et al., 2004). http://dx.doi.org/10.4995/ijpme.2013.1544 received:2013-05-21 accepted: 2013-06-03 https://ojs.upv.es/index.php/ijpme 39int. j. prod. manag. eng. (2013) 1(1), 39-54creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2013.1544 on top of the complexity introduced by the above static aspects of supply chains, chains must also be able to react quickly to changes in their environments. a typical cause on the strategic or tactic decision level is the shortening life cycle of many products and services. changing requirements to products or services can imply reorganizations of supply chains (e.g., because a different composition of modules in a product is required). a typical cause on the operational level is found in exceptions occurring during supply chain execution (such as transport disruptions), giving rise to synchro-modal logistics in which alternative transportation modes can be used side-by-side. the above developments in markets have a profound influence on the way companies collaborate to serve their customers: 1. the complexity of products requires that the organizations that produce them need to collaborate in complex production supply chains, where each partner in a chain contributes to part of the complex product (corswant and frederiksson, 2002). chains can have a simple linear structure, but can also have complex network structures. in a survey by the aberdeen group, growing supply chain complexity is identified as the top business pressure (heaney, 2012). 2. the variety of products requires that a supply chain reacts as early as possible to the requirements of the customer, pushing customerspecific operations as far as possible towards the start of the chain. this transforms a supply chain into a demand chain (or a hybrid form in between). 3. pressure on delivery times and conditions requires that a supply chain must operate in a (near) real-time fashion such that delays can be avoided and exceptions can be handled on-thefly. 4. the short life cycle of products requires that supply chains must be able to change in a fast and efficient way to produce new (generations of) products to be able to serve markets. this may lead to highly dynamic chains and networks or even instant virtual enterprises (grefen et al., 2009; mehandjiev and grefen, 2010). these developments in business collaboration require a high level of control to enable effective and efficient operation of complex supply chains. in the past, operation control was centered on intra-organizational issues and inter-organizational issues were considered as simple, messagebased synchronization points between the ‘intraorganizational islands’ of control. in the modern situation, the emphasis needs to be on interorganizational control as well – in some situations perhaps even more than on the intra-organizational control: the operation of the chain as a whole often determines business success more than the operation of the individual links of the chain. this paper addresses the issue of chain control in the context explained above. the aim of the paper is to provide a simple conceptual framework that can serve as the basis for analysis and design of complex chain control. as information is the basis for control, the paper takes an information system perspective in its analysis. we take a qualitative approach in this paper, but this can be extended into a quantitative approach. the structure of this paper is as follows. first, we lay the basis for the framework in a simple supply chain model (section 2) and a simple business control model (section 3). next, we integrate these two models in two ways to arrive at two extreme chain control models. we first discuss a decentralized control model in section 4, which is in many cases the current default in practice. we show what the weaknesses of this model are in terms of transparency and synchronization. secondly, we present a model with centralized control in section 5. this model avoids the weaknesses of the model with decentralized control. centralized control, however, has its own weaknesses in terms of autonomy and vulnerability (which are in turn avoided by the decentralized approach). to address the weaknesses of both extreme models, we combine the decentralized and centralized approaches into a model with hybrid chain control in section 6. this model can be used in a parameterized way to be adapted to a specific context. in section 7, we focus on agility in this hybrid model, i.e., on ways to organize an implementation of this model such, that change becomes a structural, natural element of this implementation, instead of a disruptive factor. in section 8, we present case studies in supply chain developments to illustrate the developed idea from a more practical perspective and show its applicability. we end this paper with conclusions in section 9. 2. basic supply chain model in this section, we present a basic supply chain model that we use as the first ingredient for the supply chain control model that we develop in the sequel of this grefen, p., & dijkman r. 40 int. j. prod. manag. eng. (2013) 1(1), 39-54 creative commons attribution-noncommercial 3.0 spain paper. we base the discussion on linear chains for reasons of simplicity, but the approach applies to arbitrary business network topologies, as discussed for example by grefen et al. (2009), and mehandjiev and grefen (2010). 2.1. the basic model in figure 1, we show the basic supply chain model that we use in this paper (this model can be considered an abstracted variation on the model used in muckstadt et al. (2001)). we use a supply chain consisting of three links for reasons of presentation clarity, where each link represents an autonomous organization in the supply chain. three links include a start link, a middle link, and an end link. longer chains can be formed by having multiple middle links, but this does not change the approach of this paper. in the figure, the double arrows indicate the material flows. the supply chain receives material input from its environment (e.g., raw materials) and delivers its output to this environment (e.g., end products). in manufacturing and logistics industries, materials are typically of a physical nature (like car parts in the automotive industry). in the service industry, like the financial industry, materials may also be non-physical (like elements of complex financial products). the single arrows indicate information flows related to the execution of business processes. order (purchase) process info flows from end to start of the supply chain to initiate the delivery of materials. delivery process info flows from start to end to organize the delivery of materials. delivery process info covers both information on the nature of the products delivered (such as product identification and quantity) and information about their transport (such as delivery times). detailed material documentation is typically transferred as part of the material flow. 2.2. supply chains vs. demand chains in pure supply chains, large batch-wise purchase orders are typically issued periodically by supply chain links to replenish stocks at these links, such that operation of the individual links is decoupled as much as possible. in pure demand chains, all purchase orders in the chain are triggered by individual purchase orders originating from the environment. demand chain operation couples the operation of individual links in the chain in a fine-grained fashion, thereby introducing high demands on synchronization of activities between supply chain links. insufficient inter-link synchronization introduces inter-link problems with respect to the effectiveness and efficiency of the chain. in complex chains, we see hybrid situations where the upstream part of the chain operates in supply chain mode (make to stock) and the downstream part in demand chain mode (make to order). the point where upstream and downstream parts meet is called the customer order decoupling point (codp) (olhager, 2012). in modern markets, chains are increasingly changing from supply chain to demand chain operation by moving the copd more upstream. 2.3. internal structure of chain links each of the links in the supply chain model of figure 1 can be modeled by porter’s value chain model (porter, 1985). we see the result of this modeling in figure 2 with porter’s model slightly simplified as in grefen (2010). the material and process information flows in the figure coincide with those in figure 1. from figure 2, we can see that within a single link, four different business functions are involved hybrid control of supply chains: a structured exploration from a systems perspective 41int. j. prod. manag. eng. (2013) 1(1), 39-54creative commons attribution-noncommercial 3.0 spain environment sc link 1 sc link 2 sc link 3 material order process info delivery process info figure 1. basic supply chain model with material and information flow structure. procurement inbound logistics o perations o utbound logistics m arketing & s ales s ervice technology development human resource management firm infrastructure figure 2. supply chain link with internal structure. in supply chain control. this may create intra-link synchronization problems, where synchronization is not fine-grained enough. these intra-link problems come on top of the inter-link problems we have observed before in this section. note that porter’s model was originally developed for the manufacturing industry. it can be applied, however, with slight modification, to other industry domains. in case of a logistics company, for example, the operations function refers to transport and the inbound and outbound logistics to respectively loading and unloading. we will see different industry domains when we get to our application case studies in section 8. 3. basic control model after introducing our basic supply chain model in the previous section, we now introduce our second main ingredient: the basic control model. we first present the model itself. then, we focus on the two types of control loops in the model. next, we introduce control levels in terms of time horizons. 3.1. the basic model the basic control model we use is shown in figure 3. this model is a simple cybernetic model with an emphasis on information processing. double arrows are material (product) flows, single arrows information flows (as in the basic supply chain model introduced in the previous section). the model consists of an environment, a transformation system, a control system, and an information system. the latter three components form the business organization under study. the transformation system receives materials from the environment and transforms them into other materials which it feeds again to the environment. typically, the transformation adds value to the materials. the precise nature of the transformation depends on the industrial context: it can be manufacturing, transport or service integration, for example. the control system controls the transformation system, i.e., it takes decisions about its operation. to take these decisions, it receives information from the information system. the information system produces information based on data it retrieves from both the transformation system (internal data) and the environment (external data). 3.2. control loops in our basic control model, control loops are used to regulate the behavior of the transformation system. a control loop includes making a decision, executing it in the transformation system, observing its effects, and deciding about its effectiveness. we can distinguish between two control loops in the basic grefen, p., & dijkman r. 42 int. j. prod. manag. eng. (2013) 1(1), 39-54 creative commons attribution-noncommercial 3.0 spain transformation system environment ctrl system information system data data information decisions figure 3. basic control model. transformation system environment ctrl system information system figure 4. intern control loop. transformation system environment ctrl system information system figure 5. external control loop. control model: the internal loop and the external loop (as illustrated in figure 4 respectively figure 5). the internal loop is completed within the boundaries of the organization under study (it is an intraorganizational loop). the control loop is used to adjust the transformation system to meet internal goals. the cycle time of the loop is fully under the control of the organization. the minimum cycle time for the internal loop is equal to the sum of the time taken to process information in the information system, the time taken to make a decision in the control system, and the time to implement this decision in the transformation system. the external loop includes the environment: the information system retrieves data about the response of the environment to materials produced by the transformation system. the control loop is used to adjust the transformation system to meet external goals. the minimum cycle time of the external loop is dependent on the reaction time of the environment and is hence not fully under the control of the organization. the cycle time is also dependent on the time to feed the output of the transformation system to the environment and on the time to retrieve information from the environment into the information system. obviously, internal and external control loops can be combined to serve different purposes in the context of business management. 3.3. control levels the control system introduced in section 3.1 operates at three control levels (as illustrated in figure 6), each of which has its own time horizon. the lowest level is the operational level. at this level, processing of individual client orders is controlled. the time horizon is short. many individual decisions have to be taken, but these are typically of a routine character. only in collaborations that change on a per-client-order basis (dynamic collaborations according to grefen (2010)), decisions that affect the mode of collaboration are taken at this level. the middle level is the tactical level. this level is concerned with handling of batches of client orders. the time horizon is medium. the amount of decisions is also medium, but these decisions require more intelligence. often, these decisions imply setting the parameters for the operational level. also, the selection of tactical supply chain partners (in semi-dynamic collaborations according to grefen (2010)) belongs to this level. the highest level is the strategic level. here, decisions are taken that are not related to concrete orders – implementing the business strategy is at stake here. selecting strategic partners (in static collaborations according to grefen (2010)) is a main issue at the strategic level. here, the parameters for the tactical level are set. each level requires information at the appropriate level (as indicated by the three arrows in figure 6). information for the strategic level is most aggregated and abstract, that for the operational level most detailed and concrete. as such, the model shares characteristics with the viable system model proposed by beer (1984), in which also a hierarchy of control systems is used (applied in supply chain management for example by verdouw at al. (2011)). 4. chains with decentralized control in this section, we discuss the first and most common case of supply chain control: the case of fully decentralized control. this case is most common because it is based on autonomous links in the chain that collaborate on the basis of link-to-link (peer-topeer) message passing. we start this section with constructing our supply chain control model by combining the two basic models that we have introduced in the two preceding sections. 4.1. the supply chain control model with decentralized control to obtain a supply chain control model with decentralized control, we combine the basic supply chain model of figure 1 and the basic control model of figure 3. hybrid control of supply chains: a structured exploration from a systems perspective 43int. j. prod. manag. eng. (2013) 1(1), 39-54creative commons attribution-noncommercial 3.0 spain transformation system environment information system s tactic operational figure 6. control levels (time horizon). to do so, we first have to answer the question how the two basic models are combined. as all links in a chain with decentralized control are fully autonomous, they each have their own control system. this implies that the basic control model will be embedded in the basic chain model on a per-link basis. consequently, we construct the model by replacing each black box link of the basic supply chain model by the structure of the basic control model. in modeling terms, we substitute and refine each of the links in the supply chain. next, we have to answer the question where to place the inter-link flows of the basic chain model between the elements of the control model instances. to answer this question, we map porter’s model (see figure 2) to the basic control model. the transformation system module of the control system should be mapped to all business functions related to supply chain management and production, i.e., all primary functions plus procurement. the control system module of the control model is in control of the entire transformation system. hence, it must be placed ‘above’ these functions. the information system module is at the same level as the control system module. this implies that we map these two modules to the firm infrastructure function. these choices are illustrated in figure 7. after having answered the above two questions, we can construct the supply chain control model. the resulting model is shown in figure 8. for reasons of simplicity and clarity, we have not distinguished between the two kinds of information flow (purchase orders and product information, see figure 1) in this figure. note that in the model of figure 8, all inter-link information flow occurs between transformation systems of the individual links (conforming to figure 2 and figure 7). likewise, there are information flows between the environment and the transformation systems of the first and last links of the chain (to transport the order process information and delivery process information of figure 1). the information flows between the environment and the information systems of the chain links transport additional contextual business information. 4.2. lifting the level of control information passing the model of figure 8 implies that all inter-link information has to pass through various business functions of the individual links – as observed in the discussion of figure 2. as this is undesirable for reasons of efficiency, we can ‘lift’ the level of inter-link information passing to the level of the control and information systems of the chain links. the resulting model is shown in figure 9. note that we have included information flows from the control systems of the first and the last chain links to the environment. these information flows correspond with the information flows from the transformation systems of the first and last chain links to the environment in figure 8 (the upward direction of the two flows in the figure). the reverse information flows are merged with the flows from environment to information systems in figure 9. grefen, p., & dijkman r. 44 int. j. prod. manag. eng. (2013) 1(1), 39-54 creative commons attribution-noncommercial 3.0 spain information & control systems transformation system procurement inbound logistics o perations o utbound logistics m arketing & s ales s ervice technology development human resource management firm infrastructure figure 7. mapping of porter’s functions to basic control model. scl1 scl2 scl3 ts environment cs is cs is cs is ts ts figure 8. basic supply chain control model with control information passing at ts level. scl1 scl2 scl3 ts environment cs is cs is cs is ts ts figure 9. basic supply chain control model with explicit control information passing. the model of figure 9 explicitly distinguishes between the level of the primary process (the transformation systems) and the level of the control process (the control systems and information systems, including the intraand inter-organizational connections between them), comparable to other supply chain control models (e.g., verdouw et al. (2011)). in terms of distributed business service management, the topologies shown in figure 8 and figure 9 are cases of choreography (aalst 2009a): we have peerto-peer control to make a global business process work. 4.3. chain-level control loops in section 3.2, we have discussed the internal and external control loops in the basic control model. as the supply chain control model is constructed using this basic control model, we can identify corresponding control loops in this model, but now on the chain level. an internal chain-level control loop is used to manage the internal operation of the supply chain, i.e., without involving the environment. we can have variations on chain-level control loops. a full chainlevel internal control loop is shown in figure 10. in this loop, materials are fed forward from the start through the chain and information about requirements to materials (like ordering information) is fed backward from the end through the chain. the minimum cycle time of the loop depends on the sum of the internal processing times (both forward in ts and backward in is and cs) of the individual links plus the communication times between the links (both forward communication at the ts level and backward communication at the is/cs level). obviously, the longer a chain is, the longer the minimum cycle time is. an external chain-level control loop is used to manage the external operation of the entire chain, i.e., to have the chain react to responses from the environment. again, we can have several variations on control loops, two extremes of which we illustrate. figure 11 shows a configuration where feedback from the environment is fed back into the chain as soon as possible, i.e., at the tail of the chain. this configuration is used in a situation where the link that delivers the end products to the environment also monitors the reaction of the environment to these products (e.g., in a typical production chain). in this configuration, for the other links of the chain, the situation is as for the internal control loop. figure 12 shows an external control loop where the feedback from the environment is fed back into the head of the chain. this configuration is applicable in a situation where the producer of the initial materials monitors the reaction of the environment to these materials (e.g., in a typical logistics chain). in this configuration, the other links are not involved at all in the control loop at the cs and is level. 4.4. evaluation of chains with decentralized control for chains with decentralized control, minimal control cycle times are heavily dependent on the length of the chain. this means that for chains with non-trivial lengths, several problems can arise: hybrid control of supply chains: a structured exploration from a systems perspective 45int. j. prod. manag. eng. (2013) 1(1), 39-54creative commons attribution-noncommercial 3.0 spain ts environment cs is cs is cs is ts ts figure 10. full internal chain-level control loop in decentralized control model. ts environment cs is cs is cs is ts ts figure 12. external chain-level control loop in decentralized control model (maximal external scope). ts environment cs is cs is cs is ts ts figure 11. external chain-level control loop in decentralized control model (minimal external scope). the chain becomes non-reactive to internal events. the chain becomes non-reactive to external events. the chain becomes confused by concurrent events, the handling of which may ‘catch up’ with each other. the chain may miss important information because links are ‘skipped’ in information processing (as in figure 12). often, links in the chain have only limited visibility with respect to processes is the entire chain (schulte et al., 2012), which may lead to sub-optimization. because of the above problems, the principle of decentralized control may be departed from. if we depart from it completely, we arrive at chains with fully centralized control. we discuss these in the next section. 5. chains with centralized control given the drawbacks of decentralized chain control as discussed in the previous section, we now move to the opposite control paradigm: fully central chain control. we first discuss the chain control model with centralized control. next, we discuss the internal and external control loops in this model. we end this section with an evaluation of the model (like in the previous section). 5.1. supply chain control model with centralized control in a chain with centralized control, there is one control system that controls the entire composite transformation system, i.e., the combination of the transformation systems of the entire chain. to construct a control model for this situation, we embed the supply chain model into the control model (opposite to what we have done to construct the decentralized control model in the previous section): we substitute the black-box transformation system of figure 3 with the chain of figure 1. this means that in the supply chain control model with centralized control, we move the information system and control system functionality from the local level (per chain link) to the chain level. consequently, we need only one chain-level control system (ccs) and one chain-level information system (cis), as obtained by embedding the basic chain model into the basic control model. the result is shown in figure 13 (for reasons of clarity, data/ information links are represented by dashed lines, decision links by solid lines). note that a decision flow is included between ccs and environment to implement the order process information flow of figure 1. the cis provides the information in the chain. to do so, it performs two kinds of monitoring. in the first place, it monitors the internal operation of the entire chain. hence, it can create aggregated information that individual links cannot create. this can be used to improve the effectiveness of the chain. in the second place, it monitors the environment on behalf of the entire chain. this avoids replication of this task in the individual links. this yields an improvement of efficiency of the chain. the ccs controls the transformation systems of the individual links in the chain, thereby synchronizing their operation. this creates two possibilities. firstly, the ccs can synchronize activities of individual transformation systems with information at the chain level (which individual links do not have available). secondly, the ccs can manage business activities executed across individual links (which is hard to realize at the local level). in terms of distributed business service management, this is a case of business service orchestration (aalst 2009b): one central component controls the execution of a number of participants in a business process. 5.2. control loops in the centralized chain control model in the centralized chain control model, we can identify chain-level internal and external control loops as introduced in section 3.2. the internal chain-level control loop is shown in figure 14. grefen, p., & dijkman r. 46 int. j. prod. manag. eng. (2013) 1(1), 39-54 creative commons attribution-noncommercial 3.0 spain ts environment ts ts ccs cis figure 13. chain control model with fully centralized control. with external chain-level control loops, we can distinguish between minimal and maximal external scope, as we have done for the decentralized control model. the resulting loops in the centralized control model are shown in figure 15 and figure 16. in a concrete scenario, the ccs can decide where to send its decisions into the chain, i.e., to vary in the spectrum of minimal and maximal external scope of the control loop to optimize effectiveness or efficiency of the chain. this provides a higher level of information routing flexibility than in the decentralized control model of figure 11 and figure 12. 5.3. evaluation of the centralized chain control model the obvious advantage of the centralized model is that chain control becomes much easier and more transparent. we have also seen that information routing can be more flexible. the obvious disadvantage is that link autonomy completely disappears. this has important consequences: more information is shared between organizations (and processed by the cis), which may affect competitiveness of individual participants. the more dynamic chains are (the more their composition is changed depending in market circumstances), the more important this may be. there is a risk of ‘one-size-fits-all’ decision making, which may not adequately take into account specific characteristics of individual organizations in a chain. each of the organizations in the chain has to completely trust the ccs in its decision making, certainly where individual organizations in the chain may have conflicting interests. depending on the organizational nature of the combination of ccs and cis, the weight of the above issues varies. the ccs+cis system may be an independent party or may be owned by one of the organizations in the chain (typically the most powerful one). if the ccs+cis system is independent, it may either be a trusted third party (ttp, for example semi-government) or another commercial partner with interests of its own. the latter case becomes certainly interesting if the service-dominant business paradigm (lusch, fargo and o’brien, 2007) is applied in the formation of service-dominant business networks (lüftenegger, grefen and weisleder, 2012), where the orchestrator may actually be the most important player in the network. a compromise to address the last two issues outlined above can be found in a model with a centralized hybrid control of supply chains: a structured exploration from a systems perspective 47int. j. prod. manag. eng. (2013) 1(1), 39-54creative commons attribution-noncommercial 3.0 spain ts environment ts ts ccs cis figure 14. internal chain-level control loop in centralized control model. ts environment ts ts ccs cis figure 15. external chain-level control loop in centralized control model (minimal external scope). ts environment ts ts ccs cis figure 16. external chain-level control loop in centralized control model (maximal external scope). information system (cis), but with decentralized control systems (cs). this model is shown in figure 17. we have used two shades of gray to distinguish upward (to cis) and downward (from cis) information flows. the control flows from the css to the environment correspond with those in figure 9. 6. chains with hybrid control fully decentralized control (choreography) and fully centralized control (orchestration) have serious limitations. choreography severely limits overall transparency of chains. orchestration severely limits the autonomy of individual organizations in the chain. therefore, a hybrid approach often is most suitable (despite its complexity): hybrid control. 6.1. the hybrid control model to obtain a hybrid chain control model, we merge the decentralized and centralized control models that we have introduced in the previous two sections. in doing so, we superimpose centralized control on decentralized control. we see the result in figure 18 and figure 19. the control model in figure 18 has no explicit peer-topeer information passing (based on the model in figure 8). the model in figure 19 includes explicit peer-to-peer information passing (based on the model in figure 9). note that in both figures, we have omitted the decision flows from local control systems to the environment for the sake of clarity of the figures. 6.2. feedback loops in the hybrid chain control model with the hybrid control model, we distinguish between internal and external feedback loops (as we have done before with the other control models). for an internal feedback loop, we have in principle the option to handle things in the decentralized (as in figure 10) or the centralized way (as in figure 14). to obtain maximal efficiency in chain control, typically the centralized way is best as this involves the smallest number of communication steps. this is illustrated in figure 20. even for a chain of three links, there is an advantage for the centralized approach: centralized requires two external and three internal communication steps (as shown in the figure, excluding material flows), whereas decentralized would require two external and five internal steps. the longer the chain is, the greater the difference becomes between centralized and decentralized. external feedback loops in the hybrid control model always follow the centralized way in the grefen, p., & dijkman r. 48 int. j. prod. manag. eng. (2013) 1(1), 39-54 creative commons attribution-noncommercial 3.0 spain ts environment cs cs cs ts ts cis figure 17. control model with centralized is but decentralized cs. ts environment cs is cs is cs is ts ts ccs cis figure 18. hybrid control model without p2p information passing. ts environment cs is cs is cs is ts ts ccs cis figure 19. hybrid control model with p2p information passing. hybrid control model, as the cis is the only element monitoring the environment for chain control. 6.3. choosing the interface level between global and local control using a hybrid chain control model implies that some control decisions are taken at the global (central) level and some at the local (decentral) level. the two levels have to collaborate to make the chain work. this means that choosing the right interface level between global and local control is essential. there are two dimensions of decisions (and the underlying information) to be taken into account here: aggregation and abstraction. in the aggregation dimension, we determine the granularity of the information passed from is to cis and hence the granularity of control decisions of css respectively cs. in the granularity dimension, we can further distinguish between the sub-dimensions of time-granularity and goods-granularity (which are usually not completely orthogonal). time-granularity determines the time scope of information and decisions, e.g., an hour or a day. goods-granularity determines the granularity of the goods that are the basis for decision making, e.g., an individual goods item, a pallet, or a container. the more detailed the interface between the global and the local level, the more transparency exists in the chain and the more real-time global control becomes possible. in the abstraction dimension, we determine the concreteness of the information passed from is to cis and hence the concreteness of control from css to cs. abstraction only pertains to goods (time-abstraction is not useful). if the abstraction level is high, information only mentions abstract characteristics of goods (such as count, size or weight). if the abstraction level is low, concrete characteristics of goods are mentioned as well (e.g., type numbers of products). the more concrete the interface is between the global and the local level, the more transparency exists in the chain. both decision dimensions can be linked to the decision levels as shown in figure 6: the more abstract and aggregated the decisions (and the underlying information), the higher they are located in the decision pyramid. this is illustrated in figure 21. in this division space, we can try and make the divide between decisions that are taken in a centralized fashion in the hybrid control model and those that are taken in a decentralized fashion. depending on the overall chain control strategy that we take, different divides can be made. a possible model for this with an emphasis on operational global chain control (as for example discussed by muckstadt et al. (2001) to obtain tightlycoupled chains) is shown in figure 22. in this model, we see that the most strategic chain-level choices are always made by individual partners in a chain and the most operational chain-level choices always at the central level. strategic choices relate to partner autonomy in a chain, hence they are decentralized. operational choices relate to concrete, real-time chain coordination, hence they are centralized. obviously, partner-level operational choices are made in a decentralized fashion. the intermediate level choices can be divided in several ways, as shown by the three alternative divides. with divide 1, the central level only handles the most concrete and detailed information, i.e., only controls the real-time, low-level operations. with divide 3, the decentralized level is responsible for taking high-level decisions only and leaves the rest to the centralized level. in this situation, the decentralized level takes the strategic/ tactical decisions and parameterizes the centralized level with these to perform the operational tactical chain control. hybrid control of supply chains: a structured exploration from a systems perspective 49int. j. prod. manag. eng. (2013) 1(1), 39-54creative commons attribution-noncommercial 3.0 spain ts environment cs is cs is cs is ts ts ccs cis figure 20. internal chain-level feedback loop in hybrid control model. abstract concrete aggregateddetailed operational tactic strategic figure 21. decision making space. 6.4. evaluation of the hybrid chain control model the hybrid chain control model solves the efficiency problems of the decentralized model and the autonomy problems of the centralized model. the penalty for this is added complexity, both in design and execution. as we have illustrated with the discussion in the previous subsection, the hybrid model can be parameterized, creating additional chain design issues compared to the decentralized and centralized models. in the execution of a chain with hybrid control, we require the tight collaboration between the global and the local control levels, which implies run-time complexity. 7. the agility dimension in setting up supply chains and their control, built-in agility becomes increasingly important. this is caused by the fact that markets become increasingly dynamic, both in their requirements to products as in the ways they want their products delivered. building agility into a supply chain means moving from repeated intensive change processes to a flexible configuration that can easily adapt to changing requirements. we see the latter for example in the concept of virtual factories (upton and mcafee, 1996; schulte et al., 2012), in which flexible configurations of production units and the chains that connect them are a starting point. in realizing chain agility, a number of aspects have to be taken into account. in this section, we briefly address a few important aspects and relate them to the models we have developed in this paper so far. the idea is to briefly sketch a picture, not to be complete here. 7.1. the business network aspect an important aspect to chain agility is to be able to flexibly change the structure of a chain, i.e., to arrive at the concept of dynamic business networks (grefen et al., 2009; grefen et al, 2013). in supply chain terms, this means the ability to easily replace a link, add a link or delete a link from a chain. to dynamically create and change chains, the potential partners for chains are present in a collaborative business ecosystem, in which they can be found based on specific criteria and in which they can be easily coupled to collaborative networks (camarinha-matos, boucher and afsarmanesh, 2010). in other words, dynamic binding between organizations must be made possible on the business level. explicit trust management is an essential element in this. 7.2. the process and service aspect in the complexity that hybrid chain control (as discussed in the previous section) augmented with agility brings, explicit business process management becomes of paramount importance. in chain-level process management, the operation of the entire chain becomes a business process, in which the chain links are process steps or subprocesses, depending on the visibility of details of the processing in the chain links. in this context, the distinction between external and internal process and data models (grefen, ludwig and angelov, 2003) is important. this distinction is closely related to the interface issues between global and local control as discussed in section 6.3. in the context of service-orientation, the links in a chain can be seen as distributed business services, each of which provide part of the overall transformation process of the entire chain. decentralized chain control then becomes service choreography. centralized chain control becomes service orchestration. 7.3. the information systems and it aspect effectively and efficiently dealing with complex, agile supply chains requires the right levels of automation in advanced information systems. the importance of information systems for supply chain management has already been clearly established quite some time ago, for example by gunasekaran and ngai (2004). research has been performed into selecting the right information technology to grefen, p., & dijkman r. 50 int. j. prod. manag. eng. (2013) 1(1), 39-54 creative commons attribution-noncommercial 3.0 spain abstract concrete aggregateddetailed central decentral figure 22. central/decentral divides in decision space. embody these information systems, for example by sarkis and talluri (2004). the importance of information systems to implement dynamic binding mechanisms in chains (as discussed in section 7.1) has been clearly established. we see this for example in standards like ebxml (oasis ebxml joint committee, 2006) and rosettanet (alonso et al., 2004). information systems can (or even should) play an explicit role in business process management and business service management as discussed in the previous subsection. a typical example of this can be found in the crosswork approach, which we discuss in the next section. 8. cases in this section, we present three cases that show how the hybrid control model that we have developed in section 6 can be mapped to advanced real-world situations. the cases are intended to be illustrative in this respect – they are by no means intended to be fully representative of their domains. 8.1. crosswork: hybrid control in the automotive industry in the crosswork project, research has been performed towards automated support for flexible supply chains in the manufacturing industry, with case studies in the automotive domain (grefen et al., 2009; mehandjiev and grefen, 2010). flexible supply chains are realized in the form of instant virtual enterprises (ives). a simplified version of the crosswork architecture is shown in figure 23. in the architecture, we see that the approach explicitly distinguishes between the construction of ives (left-hand side of the figure) and the operation of ives (right-hand side of the figure). tactic chain control with respect to agility is centralized in the design-time part of the supply chain support system: here new chains are formed. operational business process control at the chain level is centralized in the architecture this is the responsibility of the global enactment module. the local enactment modules operate local business processes within the partners of an ive. in the simplest case of interaction, these local processes are black boxes to the global level. the crosswork architecture has an interface (mehandjiev and grefen, 2010), however, that allows more complex interaction following glass box, half-open box, and open box styles (grefen et al, 2006), in which global and local level interact in a more fine-grained fashion. following the above analysis, the crosswork approach is conceptually of the hybrid kind: tactical decision making (design time) is centralized, operational decision making (run time) is hybrid (with an emphasis, though, on the central side). technically, all chain-level synchronization takes place through a centralized orchestration engine. hence, we classify the crosswork approach as hybrid control without peer-to-peer information passing. the resulting control model is shown in figure 24. we see that the ccs module contains both design time and run time support, with corresponding levels in the cis (the knowledge bases correspond to those in figure 23). the local cs modules implement the local enactment (le) module functionality (among other things that have been omitted for reasons of clarity). links between le modules and environment have been left out for reasons of figure clarity (as in figure 18). hybrid control of supply chains: a structured exploration from a systems perspective 51int. j. prod. manag. eng. (2013) 1(1), 39-54creative commons attribution-noncommercial 3.0 spain back-end: ive operation goal decomp. team format. process compos. global enactm. local enactm. market kb products kb legacy integrat. patterns kb front-end: ive construction figure 23. simplified crosswork architecture. ts environment le is le is le is ts ts glob. enactm. enactm. info. knowl. bases ive design figure 24. crosswork hybrid control model. 8.2. get service: hybrid control in realtime logistics in the get service project1, automated support is developed for real-time management of multi-modal logistics. real-time information from various parties in the transportation supply chain is used to enable on-the-fly replanning of transport and synchromodality (being able to decide in real-time on the mode of transportation to use between two points). in the get service approach, the information system at the operational level is to a large extent centralized in a logistics platform, which functions as an information backbone for all organizations in a chain. the non-real-time perspective of the get service chain control model is a standard hybrid model as shown in figure 19. in figure 25, we see the real-time perspective of the get service chain control model. non-realtime elements have been left out for clarity, most notably the local information systems of the links in the chain. flows between local control systems and the environment have been omitted as well. local transformation systems (and the environment) feed their real-time process execution data directly into a centralized event warehouse. this event warehouse is used by centralized (chain-level) re-planning and process management decision mechanisms that feed their suggestions to the local control systems of the individual links in the chain. this way, the local information systems of the chain links are bypassed for two reasons: both because they are often not equipped with real-time information processing mechanisms (infrastructure) and because directly feeding data into the central warehouse allows fast integration of data on the chain level (performance). 1 http://www.getservice-project.eu the get service approach is a hybrid approach, but with different orientation for information management on the one hand and chain control on the other hand. information management is hybrid, but leans towards centralized, as real-time information management is highly centralized. chain control is hybrid, but leans towards decentralized, as the ccs mainly suggests the local control systems with respect to important decisions. 8.3. 4c: hybrid control across supply chains our third case is also from the logistics domain, but with an emphasis on logistics chain concurrence, i.e., interwoven logistics chains. this interweaving is caused by the fact that companies have to take part in multiple supply chain configurations concurrently (verdouw et al., 2011). having interwoven chains, the challenge is to try and optimize them in an overall fashion, e.g., by having chains share resources in a dynamic fashion such as to not optimize one chain and de-optimize another one in the same go. to do so, the concept of cross-chain control center (4c) has been introduced to integrate intra-chain and inter-chain control to achieve collaboration advantages (dinalog 2013). intra-chain control is hybrid at the chain level, but inter-chain control even hybrid above the chain level. hence, a three level control model would be possible, but we prefer to stick to two levels where the 4c also has the role of the ccs at the level of an individual chain. this is illustrated in figure 26. in the figure, we have abstracted supply chain links into black boxes and merged information and control links for reasons of clarity. in the figure, we see two chains sc1 and sc2, which each consist of three links (labeled l1, l2 and l3). one link is shared between the chains, so here grefen, p., & dijkman r. 52 int. j. prod. manag. eng. (2013) 1(1), 39-54 creative commons attribution-noncommercial 3.0 spain ts cs cs cs ts ts environment replan & manage event warehouse figure 25. real-time aspect of get service model. environment sc1.l2 + sc2.l2 sc1.l3 sc2.l1 sc2.l34c sc1.l1 cisccs figure 26. 4c chain integration model. the two chains concur. the cross-chain control center (c4) consists of the cis and ccs, which controls the two individual chains plus the concurrence of the two chains. cross-chain control can optimize the use of the resources of the shared chain link, and hence of both individual chains. 9. conclusions in this paper, we have explored the issue of supply chain control, where we have approached control from an information management perspective. we have combined a basic supply chain model and a cybernetic control model in two alternative ways: embedding the latter in the former and the other way around. this has resulted in two extreme chain control models (centralized and decentralized). we have ‘interpolated’ between these two models to arrive at a hybrid chain control model. the hybrid control model allows parameterization of control, but at the cost of additional complexity. we see this complexity both at design time and at run time of a chain. analysis of a number of cases shows that the hybrid control model is often used in advanced contexts as they are often found in r&d projects. in these projects, however, the parameterization of the hybrid model is typically more or less rigid, as it is statically defined in chain control approach of a project. this indicates that a truly flexible application of the hybrid model in practical, industrial scenarios still requires quite some development. flexible application of the model presented in this paper does provide good opportunities to have chain control in complex, dynamic markets adapted well to the requirements of those markets. references aalst, w. van der (2009a). choreography. in: liu, l., & özsu, m. (2009). encyclopedia of database systems (pp. 329). springer. doi:10.1007/978-0-387-39940-9_1198 aalst, w. van der (2009b). orchestration. in: liu, l., & özsu, m. (2009). encyclopedia of database systems (pp.2004-2005). springer. doi:10.1007/978-0-387-39940-9_1197 alonso, g., casati, f., kuno, h., machiraju, v. (2004). web services: concepts, architectures and applications. springer. beer, s. (1984). the viable system model: its provenance, development, methodology and pathology. journal of the operational research society, 35:7-25. camarinha-matos, l., boucher, x., afsarmanesh, h. (2010). collaborative networks for a sustainable world; in proceedings 11th ifip wg 5.5 working conference on virtual enterprises. st. etienne, france. doi:10.1007/978-3-642-15961-9 corswant, f. von, & fredriksson, p. (2002). sourcing trends in the car industry: a survey of car manufacturers’ and suppliers’ strategies and relations. international journal of operations & production management 22(7), 741-758. doi:10.1108/01443570210433526 dinalog (2013). 4c: cross chain control center. dutch institute for advanced logistics. available at: http://www.dinalog.nl/en/themes/4c__ cross_chain_control_center grefen, p., ludwig, h., & angelov, s. (2003). a three-level framework for process and data management of complex e-services. international journal of cooperative information systems, 12(4), 487-531. doi:10.1142/s0218843003000838 grefen, p., ludwig, h., dan, a., angelov, s. (2006). an analysis of web services support for dynamic business process outsourcing. information and software technology, 48(11):1115-1134. doi:10.1016/j.infsof.2006.03.010 grefen, p., mehandjiev, n., kouvas, g., weichhart, g., eshuis, r. (2009). dynamic business network process management in instant virtual enterprises. computers in industry, 60(2):86-103. doi:10.1016/j.compind.2008.06.006 grefen, p. (2010). mastering e-business. london: routledge. grefen, p., lüftenegger, e., van der linden, e., weisleder, c. (2013). base/x: business agility through cross-organizational service engineering. beta working paper 414. eindhoven university of technology. gunasekaran, a., ngai, e. (2004). information systems in supply chain integration and management. european journal of operational research, 159(2), 269-295. doi:10.1016/j.ejor.2003.08.016 heaney, b. (2012). supply chain visibility excellence: mastering complexity and landed costs. boston: aberdeen group. hoen, k., tan, t., fransoo, j., van houtum, j. (2012). effect of carbon emission regulations on transport mode selection under stochastic demand. flexible services and manufacturing journal (published online). doi:10.1007/s10696-012-9151-6 jansen, b., swinkels, p., teeuwen, g., van antwerpen de fluiter, b., fleuren, h. (2004). operational planning of a large-scale multi-modal transportation system. european journal of operational research, 156(1), 41-53. doi:10.1016/j.ejor.2003.02.001 luftenegger, e., grefen, p., weisleder, c. (2012). the service dominant strategy canvas: towards networked business models. in proceedings 13th ifip working conference on virtual enterprises (pp. 207-215). bournemouth, uk. doi:10.1007/978-3-642-32775-9_21 hybrid control of supply chains: a structured exploration from a systems perspective 53int. j. prod. manag. eng. (2013) 1(1), 39-54creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.1007/978-0-387-39940-9_1198 http://dx.doi.org/10.1007/978-0-387-39940-9_1197 http://dx.doi.org/10.1007/978-3-642-15961-9 http://dx.doi.org/10.1108/01443570210433526 http://www.dinalog.nl/en/themes/4c__cross_chain_control_center http://www.dinalog.nl/en/themes/4c__cross_chain_control_center http://dx.doi.org/10.1142/s0218843003000838 http://dx.doi.org/10.1016/j.infsof.2006.03.010 http://dx.doi.org/10.1016/j.compind.2008.06.006 http://dx.doi.org/10.1016/j.ejor.2003.08.016 http://dx.doi.org/10.1007/s10696-012-9151-6 http://dx.doi.org/10.1016/j.ejor.2003.02.001 http://dx.doi.org/10.1007/978-3-642-32775-9_21 lusch, r., vargo, s., o’brien, m. (2007). competing through service: insights from service-dominant logic. journal of retailing, 83(1), 5-18. doi:10.1016/j.jretai.2006.10.002 maxton, g.p., wormald, j. (2004). time for a model change: re-engineering the global automotive industry. new york: cambridge university press. doi:10.1017/cbo9780511488535 mehandjiev, n., grefen, p. (2010). dynamic business process formation for instant virtual enterprises. springer london. doi:10.1007/9781-84882-691-5 muckstadt, j., murray, d., rappold, j., collins, d. (2001). guidelines for collaborative supply chain system design and operation. information systems frontiers, 3(4), 427-453. doi:10.1023/a:1012824820895 oasis ebxml joint committee (2006). the framework for ebusiness. oasis. olhager, j. (2012). the role of decoupling points in value chain management. in: jodlbauer, h., olhager, j., schonberger, r. (eds.). modelling value. springer. doi:10.1007/978-3-7908-2747-7_2 porter, m. (1985). competitive advantage: creating and sustaining superior performance. new york: free press. sarkis, j., talluri, s. (2004). evaluating and selecting e-commerce software and communication systems for a supply chain. european journal of operational research, 159(2), 318-329. doi:10.1016/j.ejor.2003.08.018 schulte, s., schuller, d., steinmetz, r., abels, s. (2012). plug-and-play virtual factories. ieee internet computing, 16(5), 78-82. doi:10.1109/ mic.2012.114 smith, s., jiao, r., chu., c. (2012). editorial: advances in mass customization. journal of intelligent manufacturing (published online). doi:10.1007/s10845-012-0700-3 upton, d., mcafee, a. (1996). the real virtual factory. harvard business review, 74(4), 123-133. verdouw, c., beulens, a., trienekens, j., van der vorst, j. (2011). a framework for modelling business processes in demand-driven supply chains. production planning & control, 22(4), 365-388. doi:10.1080/09537287.2010.486384 grefen, p., & dijkman r. 54 int. j. prod. manag. eng. (2013) 1(1), 39-54 creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.1016/j.jretai.2006.10.002 http://dx.doi.org/10.1017/cbo9780511488535 http://dx.doi.org/10.1007/978-1-84882-691-5 http://dx.doi.org/10.1007/978-1-84882-691-5 http://dx.doi.org/10.1023/a:1012824820895 http://dx.doi.org/10.1007/978-3-7908-2747-7_2 http://dx.doi.org/10.1016/j.ejor.2003.08.018 http://dx.doi.org/10.1109/mic.2012.114 http://dx.doi.org/10.1109/mic.2012.114 http://dx.doi.org/10.1007/s10845-012-0700-3 http://dx.doi.org/10.1080/09537287.2010.486384 pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.1860 received 2013-11-11 accepted 2013-11-25 packaging as source of efficient and sustainable advantages in supply chain management. an analysis of milk cartons. garcía-arca, j.a,i, gonzález-portela garrido, a.t.b,iii and prado-prado, j.c.a,ii a grupo de ingeniería de organización (gio), departamento de organización de empresas y marketing. escuela de ingeniería industrial, universidad de vigo. campus lagoas-marcosende, c/ maxwell, 36310-vigo. spain. i jgarca@uvigo.es ii jcprado@uvigo.es b departamento de organización de empresas y marketing. escuela universitaria de estudios empresariales, universidad de vigo. campus ciudad, c/ torrecedeira, 105, 36208-vigo. spain. iii tgonzalez-portela@uvigo.es abstract: in higher competitive markets, the suitable supply chain management (particularly, in logistic and productive processes) and the adoption of sustainability programs are strategic points in companies. in this context, no many companies have devoted special attention to the impact of packaging design on logistic efficiency and sustainability. thus, the integration of logistics and the packaging design has been conceptualized in the term «packaging logistics», particularly emphasizing its operational and organizational impact on supply chain performance. going beyond, authors consider that a greater emphasis should be given to the important strategic connotations to do with packaging design, in many cases this being one of the supports of competitive advantages in the supply chain management from an overall perspective of efficiency and sustainability. to illustrate this statement, in this paper, not only the conceptual field of this concept is developed, but also its application, analysing different alternatives of products packed in cartons (“briks”) and based on case study methodology. key words: packaging, supply chain, logistics, sustainability, carton. 1. packaging and supply chain companies must face with the challenges, not only in terms of new products and processes, shorter life cycles or increased commercial range, but also in terms of the demand for ever lower prices, with increasingly improved quality and service standards. in this context, many organizations are searching for a more efficient management of their supply chains as a source of competitiveness (christopher, 2000). in the last few years, companies also have to deal with two situations of strong impact on supply chains’ efficiency: globalization of supply chains and the continuous increased costs of raw materials, particularly, the oil. the combination of these two phenomena is important because, strategically, underscores the urgency of action in pursuit of maximum performance in logistics activities undertaken across the supply chain (transport, handling, storage, production, ...), eliminating activities that do not add value to the market (in line with "lean manufacturing" approach) but also developing and implementing innovations in processes and products. on the other hand, the growing sensitivity in society as regards a responsible management should imply that the supply chain management should be enlarged to take in the concept of sustainability and its three pillars associated: environmental, economic and social (ciliberti et al., 2008). beyond the isolated vision of one company, this concept of sustainability should be extended to the other companies in the supply chain, where-by all their organizations should take an active part in 15int. j. prod. manag. eng. (2014) 2(1), 15-22creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2014.1860 designing and implementing logistic processes that can be considered as sustainable (ciliberti et al., 2008; carter and rogers, 2008; seuring and müller, 2008; andersen et al., 2009). thus, sustainability and efficiency should be considered as complementary (mejías-sacaluga et al., 2011). in the context commented previously, packaging arises as one of the key elements that makes it possible the combined implementation of efficiency and sustainability strategies (jahre and hatteland, 2004; klevas, 2005; verghese and lewis, 2007; azzi et al., 2012). beyond the traditional (but nonetheless important) view of packaging as a means of protecting products (williams et al., 2008), over the last few years, new design requirements have been added for packaging: on the one hand, to improve the differentiation capacity of the product (commercial function), and on the other, to improve the efficiency of the product at logistic and productive level (logistic function). this contribution of packaging to logistic and productive efficiency should be considered not only in terms of its direct view (in the processes of supplying, packing, handling, storing and transport), but also reversely (re-use, recycling and/or recovery waste from packaging). all this has, in practice, meant the development of specific legislations (e.g. european directive 94/62/ec; 1994 and its updated version 2004/12/ec) and introduces the environmental function, not only in reverse logistics, but also in direct logistics. authors such as saghir (2002), garcía-arca and prado-prado (2008a), bramklev (2009) and azzi et al., (2012) identify in packaging three main functions: the commercial function, the logistics function and the environmental function. also, in order to put these functions into practice, it is essential to consider the packaging as a system comprising three levels (saghir, 2002): the primary packaging (“consumer packaging”), the secondary packaging (“transport packaging”; tipically, boxes) and the tertiary packaging (several primary or secondary packages grouped together on a pallet). when considering packaging from a global perspective, the interaction among different levels would become manifest, depicting the dependence among them. in fact, the adaptation of a set level of packaging should not be contemplated if the integration of the set of all the levels of grouped form is not also considered the choice of the type of packaging is usually subject only to considerations involving cost reduction. thus, packaging design affects costs both directly (costs of purchasing and waste management) and indirectly (packing, handling, storage and transport). it is precisely this indirect way that makes difficult an adequate under-standing of the repercussions of certain decisions in packaging design (garcía-arca and prado-prado, 2008a). garcía-arca and prado-prado’s study (2008b) of more than 300 companies in the supply chain of the spanish food industry, shows that logistics costs (direct or indirect) due to packaging were approximately 40% of packing companies’ revenue (14% direct and 26% indirect), and 10% of distributors’ revenue. this percentage of distributors’ costs does not include the logistic costs at the point of sale. some studies calculate the handling cost at the point of sale to be 10% of the product’s price (saghir and jönson, 2001). in this line, azzi et al. (2012), in their literature review comment that approximately 9% of the cost of any product is likely to be the cost of its packaging. this study also shows that hidden costs associated with overpackaging in europe, seem to be 20 times higher than the cost of excessive packaging materials itself. on the other hand, garcía-arca and prado-prado’s study (2008b) shows that at least 18% pallets employed downstream in the supply chain were inefficient in terms of volume and/or weight. an european study (2009) on consumer markets, carried out in five european countries, points out that the wasted volume between the primary and secondary packaging varied between 34% and 50%. between the secondary packaging, which is typically a box, and the pallets, the unoccupied space varies between 46% and 64%. with this broader view of packaging, over the last few years, the integration of logistics and the packaging design has been conceptualized in the term “packaging logistics”, particularly emphasizing the operational and organizational repercussions (hellström and saghir, 2006; garcía-arca and prado-prado, 2008a). shagir considers “packaging logistics” as “the process of planning, implementing and controlling the coordinated packaging system of preparing goods for safe, efficient and effective handling, storage, retailing, consumption and recovery, reuse 16 int. j. prod. manag. eng. (2014) 2(1), 15-22 creative commons attribution-noncommercial 3.0 spain garcía-arca, j., gonzález-portela garrido, a.t. and prado-prado, j.c. or disposal and related information combined with maximizing consumer value, sales and hence profit”. as a result of the packaging logistics implementation, it is possible to deal with the search for packaging able to meet the needs of the companies based on the possibilities associated with the combinations in the packaging structure (primary, secondary and tertiary packaging) and with the four main decisions to be taken in design: selection of the materials, dimensions, groupings (the number of packs/ package) and “graphic artwork” (or the aesthetic design of the packaging). 2. research method with the conceptual definition of the “packaging logistics” in mind, the main objective of this paper is to illustrate the potential of applying this approach in the supply chain of milk cartons (“briks”). this paper is based on a previous paper presented at cio 2013 conference in valladolid (spain). the “brik” was developed by ruben rausing in 1951 in lund (sweden). it can be made for up to six different layers and for guidance, a brick pack would comprise 75% cardboard, 20% plastic (polyethylene) and 5% of aluminum. despite its usefulness to preserve perishable liquid foods (including milk) without refrigeration and preservatives and its good logistical efficiency (volumetric occupation), this package is still blames environmental misbehavior. however, this difficulty of recycling has improved as technology evolves in separation of layers. for theory testing, the authors have adopted the “action research” approach, directly participating in the “packaging logistics” implementation process in a dairy company. thanks to this approach (action research), the researchers have the opportunity to witness the process, not only as mere observers, but also as real “agents of change” in intervention and know-how compiling processes (maull et al., 1995; prado-prado, 2000). action research can be seen as a variant of case research (yin, 2002), but whereas a case researcher is an independent observer, an action researcher “…is a participant in the implementation, but simultaneously wants to evaluate a certain intervention technique...” (coughlan and coghlan, 2002). the analysis was complemented by a literature review and a field study of dairy products (based on cartons) in three supermarkets chains in galicia (northwest spain). 3. action research analysis the analyzed company, based in galicia (northwest spain), is one of the most important manufacturers in dairy spanish market (among the 12 main manufacturers), with an annual turnover of over 100 million euros and over 250 employees. this company produces and distributes various dairy products such as milk, liquid yoghurts, cream and butter, milkshakes and cheeses. in the analysis, the authors have focused on the products packaged in milk cartons. in this kind of product, the company packs more than 100 million liters/year). particularly, we have focused on the 1 liter milk carton with cap (primary packaging), grouped in packs of 6 cartons (secondary packaging) and palletized in eur pallet (tertiary packaging). the logistics of milk cartons does not demand special requirements of conservation (temperature) as it happens with other milky products like cheese, yoghurt and cream (with a specific supply chain, due to temperature-controlled conditions). so, the supply chain of the dairy company selected could summed up including processes from packaging purchases, packing and physical distribution to reverse logistics. thus, the cost associated to this last supply chain could be summarized in the following categories: packaging purchases, packing, handling, storage and picking in manufacturer´s warehouse, transport (mainly full truck) from manufacturer to distributors, handling, storage and picking in distributors´ warehouses, transport (mainly combined truck) from distributor and supermarkets, handling and storage in supermarkets and, finally, reverse logistics (green dot). furthermore, in the analysis 4 milk carton formats were selected (see figure 1; a, b, c and d) as well as 5 of the most widely used formats of packs (see table 1 at the end of the paper; a.1, a.2, b.1, c.1 and d.1). in the figure 2, different options of grouping cartons in pack are presented. among all these combinations the company used, initially, carton a and pack a.1. regarding the packaging process purchases, indicating that the final price depends on the type of carton format, material and weight (a, b, c and d), as well as purchase volume (economies of scale). as a simplification for the analysis, it was considered that for the same volume of purchase, the final cost of each carton depends on its individual weight (see 17int. j. prod. manag. eng. (2014) 2(1), 15-22creative commons attribution-noncommercial 3.0 spain packaging as source of efficient and sustainable advantages in supply chain management. an analysis of milk cartons. table 3 at the end of the paper), although, also would be affected by the number of layers and the type of material of each layer. also, the cost of pack (secondary packaging) is determined by the type of materials, their weight and by the number of cartons/pack. furthermore, the packing process is highly automated, although their flexibility and adaptation to different formats of cartons and packs is low (high impact of setup). figure 1. alternatives of carton formats analysed. figure 2. alternatives of grouping cartons in packs. table 3. improvements in a “wrap-around” box, reducing the flaps length. flaps size (mm) cardboard surface (m2) initial stage 85 0.445 first change 50 0.407 (–9%) second change 60 0.389 (–13%) this aspect limits, in general, the coexistence of various formats of cartons and packs in the same manufacturing line so that, in practice, these lines are specialized. in the last few years the amount of raw materials used in packs has been reduced thanks to technology and design improvements. for example, in table 3 and figure 3 the reduction of flaps in “wrap-around” boxes are presented (“wrap-around” box is a kind of box especially designed for automating the packing process). figure 3. development of a wrap-around box (source: afco). with regard to the physical distribution (handling, storage and transport), the efficiency of palletizing is conditioned by the type of carton but also by the part of the supply chain that focuses on the analysis (see table 1 at the end of the paper). in this sense, the milk cartons pallet has a high density and high consumption. a priori, this product could be distributed efficiently, optimizing the activities of handling, storage and transport, looking for a larger number of liters per pallet, within the constraints of strength of carton and pack. in this regard, the maximum number of layers per pallet is conditioned not only by the type of carton, but also by the location of the cap (other formats without cap can withstand more layers). however, the type of transport between manufacturer and distributors´ warehouses is “full load truck" (maximum load limit of 33 eur pallets and 24.4 tonnes). traditionally, manufacturers have not paid much interest in improving the volumetric efficiency 18 int. j. prod. manag. eng. (2014) 2(1), 15-22 creative commons attribution-noncommercial 3.0 spain garcía-arca, j., gonzález-portela garrido, a.t. and prado-prado, j.c. of pallet (although there are significant differences as shown in table 2), since the weight determines the maximum number of pallets per truck. in fact, in the company, the number of pallets per truck does not exceed 30. all this significantly affects, not only to the efficiency of handling and storage in the ware-houses of the manufacturer, distributors and supermarkets, but also in transport between distributors´ warehouses and supermarkets. in particular, in the latter transport, the type of truck changes, not only in the capacity (typically with less capacity vehicles), but also in the configuration of goods, due to mill pallets are combined with other pallets of products food with lower densities (monoreference and / or multi-reference pallets). by combining these different kinds of products, generally, the average weight in each pallet on this new truck is reduced, enabling a priori a better pallet volume. in figure 4, the filling rate in trucks (%) is presented according to density of products and the designed height in pallets. this brings additional advantages in supermarkets that have opted to present directly at the point of sale the milk pallet (minor handling and better occupancy in supermarkets). even, company can adopt alternatives that reduce the weight of traditional wooden pallet (25 kilograms per pallet) to gain useful load capacity on trucks. among these alternatives are: the plastic pallet (6 kg), the cardboard pallet (12 kg), the "loading ledge" (1.5 kg) or the “slip sheets” (1 kg approximately). in an initial full load truck (30 pallets), the analyzed manufacturer could earn more than 700 kilograms (option with loading ledge and slip sheets), equivalent to an additional pallet on each truck (3.33%). in the other options, also an additional pallet per truck is load except in the option d. however, any of these alternatives would require a change in the pool system (exchange of pallets or "loading ledges"). besides, the “loading ledges” and “slip sheets”, moreover, requires changes in the palletizing system at the manufacturer's premises and in the handling machines. therefore, it has not been considered all these options in the final analysis. finally, at the level of reverse logistics, the green dot cost also depends on the selection of the carton and the pack. table 2, at the end of the paper, summarizes the total costs of green dot per each alternative. the ecoembes fees in 2014 are: 0.323 €/ kg carton; 0.068 €/kg cardboard; 0.472 €/kg plastic. 4. results and discussion as a final decision, it was decided to choose the most efficient carton format in logistics (format d), since is the alternative with major level of total savings (see tables 2 and 3). this alternative involves no substantial changes in the system of packing in carton and pack and implies savings in handling and storage of over 16,000 pallets a year (a reduction of 11.7%). 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% 0.45  m 0.65  m 0.85  m 1.15  m 1.45  m 1.85  m 2.3  m 322  kg.*eur*1  m  height 401  kg.*eur*1  m  height 500  kg.*eur*1  m  height 624  kg.*eur*1  m  height 645  kg.*eur*1  m  height 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% 0.45  m 0.65  m 0.85  m 1.15  m 1.45  m 1.85  m 2.3  m 322  kg.*eur*1  m  height 401  kg.*eur*1  m  height 500  kg.*eur*1  m  height 624  kg.*eur*1  m  height 645  kg.*eur*1  m  height figure 4. analysis of filling rate of trucks according to density and height in pallets. 19int. j. prod. manag. eng. (2014) 2(1), 15-22creative commons attribution-noncommercial 3.0 spain packaging as source of efficient and sustainable advantages in supply chain management. an analysis of milk cartons. this improvement is also an annual reduction in the number of full trucks to 92 trucks (2% reduction). only in transport between the manufacturer and distributors, this change should involve saving of 60,000 euros/year (at least, total savings of 35,000 euros/year). additional savings could be achieved thanks to the reduction of handling and storage in manufacturers, distributors and supermarkets, but also of the transport costs between the distributors' warehouses and supermarkets. in summary, packaging design could be considered as a “microworld” where every millimeter and gram counts in the context of overall supply chain efficiency and sustainability. going beyond, redesigning packaging can lead not only to savings, but also increased sales. to achieve this, the global impact of design decisions must be measured in the whole chain. in this context, the key to achieving efficient and sustainable packaging is the coordination and collaboration between all areas, departments or companies throughout the supply chain. this last statement implies that packaging formats selected should not be considered ‘fixed’, but rather a solution that can be constantly improved. thereby, design decisions should be regularly revised in case commercial, logistic or environmental requirements change, or if new innovations in materials and/ or technical solutions occur within the packaging industry. on the other hand, the role of “change agents” (the authors of this paper), as promotors of “packaging logistics” implementation in line with the scientific approach “action research,” must be outlined. 5. conclusions in a competitive and global world, companies should face supply chain design from a sustainable and efficient perspective. the real challenge for companies is how to integrate, proactively and strategically, both concepts. in this scenario, redesigning packaging by applying the “packaging logistics” concept is an example of this integration as it was illustrated in the dairy company with cartons analysis. as described in this paper, the supply chain as a whole has also succeeded in making substantial savings at logistics and environmental level. table 1. logistics comparison among carton alternatives. type and carton dimensions (mm, w×l×h) type of pack and dimensions (mm, l×w×h) palletization pallet height (m, incl. pallet) pallet weight (kg, incl. pallet) transport efficiency type a (60×90×195) type a.1; card-board box (wrap-around); 282×127×216 720 cartons; 24 packs/layer; 5 layers/pallet 1.33 807 30 pallets/truck; 21,600 cartons/truck (initial solution) type a (60×90×195) type a.2; card-board tray and plastic cover; 282×128×210 720 cartons; 24 packs/layer; 5 layers/pallet 1.3 800 30 pallets/truck ; 21,600 cartons/truck; (no improvement) type b (65×70×252) type b.1; plastic cover; 219×130×265 768 cartons; (+6.66%); 32 packs/layer; 4 layers/pallet 1.31 848 28 pallets/truck; 21,504 cartons/truck; (-0.44%) type c (71×75×204) type c.1; plastic cover and car-board sheet; 227×150×205 864 cartons; (+20%); 24 packs/layer; 6 layers/pallet 1.4 951 25 pallets/truck; 21,600 cartons/truck; (no improvement) type d (62×70×239) type d.1; card-board tray and plastic cover; 228×128×245 816 cartons; (+13.3%); 34 packs/layer; 4 layers/pallet 1.23 903 27 pallets/truck; 22,032 car-tons/trailer; (+2%) 20 int. j. prod. manag. eng. (2014) 2(1), 15-22 creative commons attribution-noncommercial 3.0 spain garcía-arca, j., gonzález-portela garrido, a.t. and prado-prado, j.c. the path taken by the company in the implementation of the "packaging logistics" combining logistics efficiency and sustainability (strategy "lean and green"), could be assimilated and adapted by other companies, regardless of sector or size, as it would contribute to improving competitiveness through innovation of products and processes within the supply chain. in fact, the adoption of the packaging logistics reinforces the initial vision that efficiency and sustainability are not incompatible terms but, rather, are complementary terms. table 2. sustainable comparison among cartons alternatives. carton model carton weight (g; without cap) carton green dot (million €/year) pack model type of pack pack weight pack green dot (million €/year) total green dot (million €/year) a 38 1.227 a.1 cardboard box (wrap-around) 87 g 0.098 1.325 a 38 1.227 a.2 cardboard tray (c) and plastic cover (p) 44 g (c); 12 g (p) 0.149 1.376 (+3.44%) b 39 1.26 b.1 plastic cover 15 g 0.118 1.378 (+3.90%) c 36 1.162 c.1 plastic cover (p) and carboard sheet (c) 14 g (p); 22 g (c) 0.135 1.297 (–2.11%) d 36 1.162 d.1 cardboard tray (c) and plastic cover (p) 51 g (c); 13 g (p) 0.160 1.322 (–0.23%) references andersen, m., skjoett-larsen, t. (2009). corporate social responsibility in global supply chains. supply chain management: an international journal, 14(2), 75-86. doi:10.1108/13598540910941948 azzi, a., battini, d., persona, a., sgarbossa, f. (2012). packaging design: general framework and research agenda. packaging technology and science, 25(8), 435-456. doi:10.1002/pts.993 bramklev, c. (2009). on a proposal for a generic package development process. packaging technology and science, 22(3), 171-186. doi:10.1002/pts.850 carter, c.r., rogers, d.s. (2008). a framework of sustainable supply chain management: moving toward new theory. international journal of physical distribution & logistics management, 38(5), 360-387. doi:10.1108/09600030810882816 ciliberti, f., portrandolfo, p., scozzi, b. (2008). logistics social responsibility: standard adoption and practices in italian companies. international journal of production economics, 113(1), 88-106. doi:10.1016/j.ijpe.2007.02.049 christopher, m. (2005). logistics and supply chain management strategies for reducing cost and improving service. 3ª edition, financial times pitman publishing, london, uk. coughlan, p., coghlan, d. (2002). action research for operations management. international journal of operations & production management, 22(2), 220-240. doi:10.1108/01443570210417515 europen. (2009), the european shopping baskets, europen and packforsk, brussels, belgium. garcía-arca, j., prado-prado, j.c. (2008a). packaging as source of competitive advantages. universia business review, 17, 64-79. garcía-arca, j., prado-prado, j.c. (2008b). packaging design model from a supply chain approach. supply chain management: an international journal, 13(5), 375-380. doi:10.1108/13598540810894960 hellström, d. saghir, m. (2006). packaging and logistics interactions in retail supply chain. packaging technology and science, 20(3), 197216. doi:10.1002/pts.754 jahre, m., hatteland, c.j. (2004). packages and physical distribution: implications for integration and standardization. international journal of physical distribution and logistics management, 34(2), 123-139. doi:10.1108/09600030410526923 klevås, j. (2005). organization of packaging resources at a product-developing company. international journal of physical distribution and logistics management, 35(2), 116-131. doi:10.1108/09600030510590309 21int. j. prod. manag. eng. (2014) 2(1), 15-22creative commons attribution-noncommercial 3.0 spain packaging as source of efficient and sustainable advantages in supply chain management. an analysis of milk cartons. http://dx.doi.org/10.1108/13598540910941948 http://dx.doi.org/10.1002/pts.993 http://dx.doi.org/10.1002/pts.850 http://dx.doi.org/10.1108/09600030810882816 http://dx.doi.org/10.1016/j.ijpe.2007.02.049 http://dx.doi.org/10.1108/01443570210417515 http://dx.doi.org/10.1108/13598540810894960 http://dx.doi.org/10.1002/pts.754 http://dx.doi.org/10.1108/09600030410526923 http://dx.doi.org/10.1108/09600030510590309 maull, r.s., weaver, a.m., childe, s.j., smar, p.a., bennett, j. (1995). current issues in business process reengineering. international journal of operations and production management, 15(11), 37-52. doi:10.1108/01443579510102882 mejías-sacaluga, a., garcía-arca, j., prado-prado, j.c., fernández-gonzález, a.f., comesaña, j.a. (2011). modelo para la aplicación de la responsabilidad social corporativa en la gestión de la cadena de suministro. dirección y organización, 45, 20-31. prado-prado, j.c. (2000). el proceso de mejora continua en la empresa. pirámide, madrid, spain. saghir, m. (2002). packaging logistics evaluation in the swedish supply chain. lund university, sweden. saghir, m., jönson, g. (2001). packaging handling evaluation methods in the grocery retail industry. packaging technology and science, 14(1), 21-29. doi:10.1002/pts.523 seuring, s., müller, m. (2008). from a literature review to a conceptual framework for sustainable supply chain management. journal of cleaner production, 16(15), 1699-1710. doi:10.1016/j.jclepro.2008.04.020 verghese, k., lewis, h. (2007). environmental innovation in industrial packaging: a supply chain approach. international journal of production research, 45(18-19), 4381-4401. doi:10.1080/00207540701450211 williams, h., wikström, f., löfgren, m. (2008). a life cycle perspective on environmental effects of customer focused packaging development. journal of cleaner production, 16(7), 853-859. doi:10.1016/j.jclepro.2007.05.006 yin, r.k. (2002). applications of case study research, ed. sage publications, thousands oaks, ca, usa. 22 int. j. prod. manag. eng. (2014) 2(1), 15-22 creative commons attribution-noncommercial 3.0 spain garcía-arca, j., gonzález-portela garrido, a.t. and prado-prado, j.c. http://dx.doi.org/10.1108/01443579510102882 http://dx.doi.org/10.1002/pts.523 http://dx.doi.org/10.1016/j.jclepro.2008.04.020 http://dx.doi.org/10.1080/00207540701450211 http://dx.doi.org/10.1016/j.jclepro.2007.05.006 pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.17207 received: 2022-03-02 accepted: 2022-07-17 analyzing store features for online order picking in grocery retailing: an experimental study mar vazquez-noguerol a1* , sara riveiro-sanromán a2 , iago portela-caramés a3 , j. carlos prado-prado a4 a universidade de vigo, grupo de ingeniería de organización (gio), escuela de ingeniería industrial, 36310 vigo, spain. a1marfernandezvazquez@uvigo.es, a2sara.riveiro@uvigo.es, a3iagportela@uvigo.es, a4jcprado@uvigo.es abstract: the digital transformation is having a major impact on the consumer product market, pushing food retailers to foster online sales. to avoid large investments, e-grocers are tending to use their existing physical stores to undertake the online order picking process. in this context, these companies must choose in which traditional stores must prepare online orders. the aim of this study is to identify which store features affect order preparation times. the action research approach has been used at a spanish e-grocer to analyze the characteristics that differentiate picking stores from each other; furthermore, the preparation times for a sample of online orders have been measured. the data were analyzed statistically using one-way anova to define the optimal store in terms of size, assortment, backroom and congestion. the study shows that three of the four characteristics are significant on the preparation time. therefore, e-grocers using a store-based model can use this information to focus their efforts on optimizing this process, assigning online order picking to the most appropriate stores. the approach used allows the study to be suitable for different retail context. moreover, the results serve as support for strategic decision-making of researchers and e-grocers seeking to become more competitive in this continually growing market. key words: e-grocery, omnichannel, store-based, picking, food retail, fulfillment. 1. introduction online and traditional sales channels are increasingly connected, although it has been demostrated that the online sales are substituting the offline ones (suel et al., 2018). this shift in consumer preferences appears to have accelerated in recent years and an increase in the use of technology has been observed. this behavioral change among consumers has been noted to a greater extent in the food sector. e-commerce has been boosted by the effects of covid-19, as consumer food distribution had to be adapted quickly (forbes, 2020). on the one hand, border closures hindered the supply of some products. this issue became more critical in long-distance supply chains, where breaks in the chain provoked wholesale supply problems that led to stock-outs at retailers offering the online channel (mahajan and tomar, 2021). on the other hand, social restrictions resulted in the horeca channel (hotel, restaurant and catering) to close down. for this reason, the food sector underwent major growth in retail as consumers began to prepare more meals at home and eat out less (hailu, 2020). this sudden change in consumer behavior highlighted more than ever the need to create sustainable and resilient supply chains that enable the adoption of alternatives to guarantee business continuity (hendrickson, 2020). to cite this article: vazquez-noguerol, m., riveiro-sanromán, s., portela-caramés, i., prado-prado, j.c. (2022). analyzing store features for online order picking in grocery retailing: an experimental study. international journal of production management and engineering, 10(2), 183-193. https://doi.org/10.4995/ijpme.2022.17207 int. j. prod. manag. eng. (2022) 10(2), 183-193creative commons attribution-noncommercial-noderivatives 4.0 international 183 https://orcid.org/0000-0002-5319-7359 https://orcid.org/0000-0003-4674-960x https://orcid.org/0000-0001-9314-2021 https://orcid.org/0000-0003-2189-2100 mailto:marfernandezvazquez@uvigo. mailto:sara.riveiro@uvigo.es http://creativecommons.org/licenses/by-nc-nd/4.0/ as a consequence of these breaks in the supply chain and the increase in omnichannel demand brought about by covid-19, supermarkets encountered capacity problems in order to meet customer orders. that is why retailers who seek to increase omnichannel sales should pay special attention to their supply chain and new business models (rodríguezgarcía et al., 2016). within this scope, the digital transformation has become one of the most important strategies in order to mitigate the risks to the supply during the pandemic (kumar et al., 2021). for that reason, supermarket chains are seeking to optimize their processes, concentrating their efforts on activities with the most impact on costs. looking at the online channel of grocery sellers, order preparation and transport are the most important activities, accounting for a large part of overall costs of stores (tompkins et al., 2010). these activities are jointly known as e-fulfillment, and are especially important in this sector, which is based on handling food products that must meet a series of particular requirements (temperature control, expiry dates, traceability, etc.). authors such as (komijan and delavari, 2017; schoen et al., 2018) underscore the need to increase the proper conservation of sensitive products when it comes to managing the supply chain because there is increasing demand for greater speed. at the same time, online orders tend to be highly variable in terms of the number of items in them and often include products of a variety of sizes and weights, or that need to be kept at different temperatures (zhao et al., 2020). to this end, products are often categorized into three groups: dry products (at room temperature), fresh products (kept above freezing), and frozen products (kept below freezing). authors such as (vazquez-noguerol et al., 2020) state that order delivery is an activity that has been studied in detail by several authors, whereas there has been limited published research into improving the online order picking process. in the extant literature, a distinction is commonly made between two strategic models that an e-grocer can adopt in order to meet demand, based on where preparation takes place: the warehouse-based model and the store-based model (rai et al., 2019).within each of these alternatives, specific peculiarities can be found or even hybrid models that combine features of both. supermarket chains that choose the warehousebased option have two different ways of operating: they can do their online order picking at their existing warehouses, or they can devote specialized infrastructures (known as dark stores) to the online channel. in both cases, investment tends to be greater than that for the store-based option due to the automation of handling means. in this respect, warehouses usually gain efficiency in the picking process, but the flexibility of the installation is reduced because automatic systems generally operate under very specific conditions. in general, the viability of the warehouse-based model is usually conditioned by the volume of orders, particularly in those models that are exclusively devoted to the online channel (rodríguez-garcía et al., 2021). in both warehouseand store-based models, when the infrastructures are used to fulfil both onand offline channels, a major advantage can arise. this is because order preparation resources can be shared and online orders can be prepared when the workload for the traditional channel is low. owing to this synergy between both channels, many supermarkets have opted for a store-based model, preparing online orders from their existing stores. this model has several advantages including that it allows flexibility in the face of variable demand and could avoid any major initial investment. however, one of the main disadvantages is that stores are not designed for the picking process to be done efficiently, but instead are aimed at the traditional sales channel. (wollenburg et al., 2018). add to this the fact that there is a great variety of features differentiating stores from each other, and so the same process may be very different in each store. the great diversity existing among stores gives rise to discrepancies in preparation times, which constitutes a potentially interesting problem to be studied. to date, there are no studies on e-grocery aimed at optimizing picking tasks by looking at store features. nevertheless, knowing which features can favor or hinder picking tasks will provide information of great value that can help supermarket chains to decide which stores to use for online orders. up to now, this strategical decision has been based on the proximity of the store and the customer; however, the closest store may not be the best alternative. thus, it can be concluded that there is a gap in the literature. that gap is addressed by the current study and by answering the following research question: rq: how do the characteristics of stores affect the online order picking process? in order to answer that research question, a study has been developed a statistical study using the action int. j. prod. manag. eng. (2022) 10(2), 183-193 creative commons attribution-noncommercial-noderivatives 4.0 international vazquez-noguerol et al. 184 http://creativecommons.org/licenses/by-nc-nd/4.0/ research approach. a case study has been carried out at a spanish firm with a store-based model. the time needed to prepare a sample of online orders has been measured taking as a reference the most representative stores of that e-grocer. the aim is to improve the online order picking process based on store features because this process needs to simultaneously achieving low costs, high accuracy, and high velocity (hübner et al., 2019). the remainder of the paper is organized as follows. section 2 presents a review of the main publications about improving the online order picking process in stores. section 3 describes the methodology used in the research process. the empirical analysis carried out is presented in section 4. section 5 illustrates the findings of the statistical study. section 6 discusses the results with what exists so far in the literature and finally section 7 presents the conclusions of the study. 2. literature review with regard to the two existing strategic order picking models, warehouseor store-based, it should be noted that most e-grocers have opted for the storebased model (ishfaq and bajwa, 2019). this has mainly been due to the rapid growth in the online channel and the high costs of picking and transport. with the store-based model, e-grocers can make the most of this service to consolidate their market share without the need to make large initial investments by making good use of their existing physical sales points as order preparation centers (mishra and mukherjee, 2019; dias et al., 2020). in this strategic model, the work operation is as follows. a supermarket employee goes along the aisles of the store gathering the ordered products from the shelves to a trolley, just like any other client of the traditional channel. the main advantage of this model is that it allows the transformation to the online channel to be more gradual and flexible. moreover, stores can adapt as demand increases. this is because, in order to carry out picking from the store, little else is required beyond staff training and any electronic devices. however, as mentioned previously, one main drawback of this strategic model is that these sales points are not designed for online picking (rodríguez-garcía et al., 2021). firstly, stores tend to be designed on the basis of sales studies that encourage offline customers to buy more products as they go up and down the aisles. however, for an online order picker, the goal is to travel the minimum number of meters to complete an order, which is not something that is taken into account when designing traditional stores (pires et al., 2021). the way products are presented on the shelves is also different because stores have products grouped in “families”, whereas online picking is more efficient when the products are sorted in terms of their rotation, as they are in warehouses. because the traditional channel continues to predominate over the online one, changing store layout to optimize picking routes would not seem to be the best option. however, firms can act on other factors, such as choosing to assign online order picking to the optimum stores (pires et al., 2017). when supermarket chains decide to implement the store-based model, a strategic decision must be made to select the stores to be used for picking online orders. to date, picking has usually been done in the stores nearest the customer and in locations where there is a greater population density of online orders. however, each store has features that could hinder or favor the picking task. the chief differences between stores are found in size, in selling concepts, in backroom availability and in congestion (seidel et al., 2016). the same authors focused one study on categorizing stores by size and distinguished three types of stores. although there is no unanimity about the defining limits for size, supermarket chains can categorize their stores from the smallest to the largest surface area as: convenience stores (also known as discount stores), supermarkets, and hypermarkets (seidel, 2021). in terms of picking, convenience stores tend to be the least efficient as their grid layout will cause pickers to take longer routes thereby increasing the picking time (do and omdahl, 2018). similarly, it can also be reasoned that supermarket-sized stores would increase the travel distance as they have a greater surface area. this aspect would become even more relevant in the case of hypermarkets, which are not only the stores with the largest area but also stock other products not found in supermarkets, such as household appliances. it seems reasonable that the main objective of the stores is to minimize the travel distance of pickers and consolidation effort of orders (hübner et al., 2019). the aim of such a consolidation effort is to reduce the process time by preparing several orders simultaneously, either by grouping orders that must be sent in the same time slot or grouping products from several orders int. j. prod. manag. eng. (2022) 10(2), 183-193creative commons attribution-noncommercial-noderivatives 4.0 international analyzing store features for online order picking in grocery retailing: an experimental study 185 http://creativecommons.org/licenses/by-nc-nd/4.0/ by category (e.g., dry, fresh, and frozen). however, this is a little used model as it often gives rise to confusion and errors among the various orders being prepared at the same time. this travel distance also usually depends on the assortment. for that reason, it is very important to define the assortment in the store properly, as this will limit the online assortment because online orders will be picked from the shelves in the physical stores (gallino and moreno, 2019; rooderkerk and kök, 2019). thus, stores with a wider range of products will need a greater number of shelves to display them. as the number of shelves increases, so does the number of aisles in the store and, in turn, the distance in meters on the route of a picker. it is not clear whether it is better to offer a large or limited assortment. the former increases customer service, even though this lengthens search and route times. the latter reduces search times, although this may increase congestion of pickers and customers in the aisles; moreover, it may reduce the level of service (fernie et al., 2010). another question to bear in mind when integrating both channels is management of products on offer as this could trigger an increase in demand and affect availability of the products on the shelves. furthermore, those offers and promotions are not usually the same in both channels. another characteristic that tends to differentiate stores is backroom availability, as this space allows the store to be supplied when there are stock-outs as orders are being prepared (pires et al., 2017). some stores even use this space to speed up online order preparation and avoid contact with offline customers (mangiaracina et al., 2018). one of the greatest drawbacks of this model is that the time devoted to picking is usually increased as case packs are not divided into customer units (broekmeulen et al., 2017). as a result, the pickers have to divide the units in their packaging or even unpack parts of pallet loads, which can lead to increases in picking time. this problem becomes even clearer when multi-item pallets are used, where the pickers have to handle several products in order to collect the particular reference they require. add to this the fact that backroom pallets are not always at picking height. they can be stored high up and need handling equipment to be reached. in contrast, not having a backroom can lead to decreased customer service due to there being less capacity for replacement (paul et al., 2019). between the two options, there is an intermediate model, in which the store does have a backroom (and thus some capacity for replenishment but it is not used for online order picking. the handling problems are avoided with this picking option. the last factor identified in the literature that affects stores is congestion. this variable refers to the store traffic, specifically the aisle traffic. picking time is expected to slow down as customer traffic increases because in the hours of greater offline activity, the speed of the picker can be reduced owing to interaction with customers sharing the same space. however, the picking time does not have to increase significantly when the picker spends more time standing still picking up products from shelves and less time travelling between aisles (salgado, 2015). in summary, the extant literature has been used to identify the store features that appear to influence order picking time. these are: size, assortment, backroom availability, and congestion. this information can be used to determine which options are the most favorable for speeding up the picking process and thus define the optimum stores to serve the online channel. 3. methodology this section describes the methodology employed throughout the study. the first part contains a description of the research process and an explanation of the different stages followed in the proposed approach. subsequently, the second part presents the case study selected for the analysis of the order picking process. 3.1. research process to carry out this study, the researchers have used the well-known action research approach. authors such as shani & coghlan (2019) define it as a research process in which applied scientific knowledge is integrated with existing organizational learning and employed in real-world problems. a researcher using the action research approach is not just a mere observer of the process of change but a participant that is directly involved in it, acting as an agent of change (prado-prado et al., 2020). under these considerations, the research process begins with a contextualization phase. firstly, authors conduct a literature review with the aim of identifying which store features can affect the picking process. then, to undertake the study, we int. j. prod. manag. eng. (2022) 10(2), 183-193 creative commons attribution-noncommercial-noderivatives 4.0 international vazquez-noguerol et al. 186 http://creativecommons.org/licenses/by-nc-nd/4.0/ choose a sample of pilot stores classified according to the characteristics defined above. with the aim of determining the optimal store for the online picking process, studies of methods and times are carried out in the most representative omnichannel stores. the time measured for each order is that corresponding to carrying out all the tasks that form a part of the online order picking process, from downloading the orders to the devices pickers work with to the packing process for later dispatch to the end consumer. due to the large amount of collected data, it is necessary to use statistical tools in order to draw conclusions about the time studies. for this analysis, one way anova is used, which is a statistical technique for assessing how several explanatory variables affect a continuous response variable (hofmann and meyer-nieberg, 2018). in this respect, one-way anova is applied to determine the effect of each store characteristic (explanatory or independent variables) in the order picking time (response or dependent variable). since the objective of the study is to reduce the picking process time, those store characteristics that speed up the preparation will be considered as the best option. 3.2. case selection the firm of the study is a supermarket chain with stores throughout spain, which has an annual sales figure of €5bn. the case study has been limited to two regions where sales are more representative: galicia and país vasco. this firm is a reference in e-commerce as it has been offering the service for 10 years, although only 13% of its stores offer this sales channel. the traditional stores used for picking have been selected based on their location, all of them being situated in the greater population density, where the online demand is usually concentrated. in a first phase of analysis, the stores selected for the study were classified according to their size, product assortment, backroom and traffic or congestion; these differences are the product of nonstandardization. based on this information, the stores were categorized according to their characteristics. as a result, a representative sample of their online demand was selected. the sample obtained consisted of 200 orders, which corresponds to the average daily demand of the stores studied. because the firm is continually growing, and the online demand is steadily increasing, optimizing the online order picking process is vital to maintain leadership. 4. empirical analysis this section deals with the description of the problem. first, the initial considerations of the study are justified. then, the store features under observation that affect the online order picking process are described. 4.1. field problem the first stage of the case study focuses on investigating the problems linked to the stores and their work method. first, it should be pointed out that for this study only the preparation process has been taken into account, and any other activity related to replenishment and transport has been excluded. regarding the stores, we identified that the management is the same for all of them. in this respect, the online order picking process is always carried out with the store open to the public and the frequency and method for replenishment are similar at all stores. at a technological and operational level, the stores can be considered similar, as they all have the same working method and the same type of devices. at the beginning, pickers download online orders onto a mobile device. they carry the device with them during the process because it indicates the items and quantities of the products to pick up. just as if they were an offline customer in the store, the pikers walk through the aisles, stopping at the locations of the products indicated by the device, and placing them into a trolley. although the layout of the stores may vary depending on the size or the product range, the route of the pickers is very similar in all stores. they all start by picking up the dry products, continue with the fresh products, and finally complete the order with the frozen products. at the end of the process, the order is packed and ready for shipment. it should be noted that in all stores the order picking is done individually, this means that each order is prepared by a single picker and that each picker only prepares one order at a time. finally, it is worth remarking that all measurements were taken at stores located in city centers, where most online orders are prepared. moreover, these stores are the ones with the longest experience in preparing online orders, so all the pickers have received training and have a similar level of experience. taking all these considerations into account, and with the objective of collecting comparable int. j. prod. manag. eng. (2022) 10(2), 183-193creative commons attribution-noncommercial-noderivatives 4.0 international analyzing store features for online order picking in grocery retailing: an experimental study 187 http://creativecommons.org/licenses/by-nc-nd/4.0/ measurements, the researchers paid special attention to the day and time slot when preparation took place in order to avoid bias as a result of the weekly seasonality of the sector (hübner et al., 2015). for that reason, in addition to taking into account that the chosen stores were similar (in terms of demand, work method and location), all the measurements of the orders assigned to each store were made on a single day. this last factor is relevant because omnichannel and the substitution effect between channels cause discrepancies between online and offline sales. 4.2. design of intervention regarding the initial considerations, four store variables were identified for each order that affect the online order picking process (independent variables). analysis of this information, together with the measurements taken, made it possible to determine the type of store that provided efficient online order preparation. the first store variable was size. the firm itself classifies its stores in three groups: convenience stores have a surface area less than 700 m2, supermarkets covers 700 to 2000 m2, and hypermarkets cover a surface area of up to 2000 m2. the second variable was the product assortment on offer. after analyzing the distribution of the assortment size in the different stores, the following categories of assortment were established: small, fewer than 32000 items, medium, between 32 000 and 40 000 items, and large, over 40 000 items. the third variable was the availability of a backroom. if the store has a backroom, it is also analyzed whether this facility adjacent to the store is used for online order preparation. the difference between these last two options is that the one that does have a backroom will have a greater replenishment capacity, even though picking is only carried out in the aisles of the store. finally, the fourth variable defined corresponds to store customer traffic or congestion. for this, an indicator of daily orders per square meter of the store was defined. the orders considered in this indicator correspond to the customers of the offline channel, that is, the customers who physically visit the store to make their purchase. in the store-based model, the traffic of online order pickers can be distributed throughout the working day, taking advantage of the times when the store has less workload. however, the traffic of customers visiting the store is a variable that cannot be acted upon. this variable was also sorted into three ranges because of the distribution of the indicator values at the stores of the firm: low, when the indicator is below 0.3 daily orders/m2; medium, when it is between 0.3 and 0.5 daily orders/m2; and high, when it is over 0.5 daily orders/m2. once the identified variables have been described, table 1 shows the distribution of a sample of 200 online orders. in addition, the picking process time has been considered as the dependent variable to determine which alternative is more efficient. those characteristics that allow a shorter picking time will be more favorable. to this end, the preparation time of the orders that make up the sample has been measured. given that the time for preparing an order depends mainly on the number of items, the total time for the order has been divided to obtain a suitable indicator for preparation time (chintagunta et al., 2012). the number of items per order in the sample ranged from 4 to 69, with an average of 36 items per order. table 1. store features description and sample distribution. store features alternatives range sample (n=200) convenience < 700 m² 24 (12%) store size supermarket 700 < store < 2000 m² 95 (47.5%) hypermarket > 2000 m² 81 (40.5%) small assortment < 32000 items 56 (28%) assortment medium assortment 32000 < assortment < 40000 items 60 (30%) large assortment > 40000 items 84 (42%) no 36 (18%) backroom yes 95 (47.5%) yes and picking 69 (34.5%) low traffic < 0,3 daily orders/ m² 51 (25.5%) congestion medium traffic 0,3 < traffic < 0,5 daily orders/ m² 75 (37.5%) high traffic > 0,5 daily orders/ m² 74 (37%) int. j. prod. manag. eng. (2022) 10(2), 183-193 creative commons attribution-noncommercial-noderivatives 4.0 international vazquez-noguerol et al. 188 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5. findings after defining the sample and the variables, studies of methods and times were undertaken on the orders of the pilot sample. then, a one-way anova statistical analysis was carried out in order to determine the effect of each characteristic on the picking time per item. table 2 presents the findings obtained. as the results show, three of the four store features identified are statistically significant. on the one hand, store size and assortment, are significant at the 0.01 level (p-value<0.001), while backroom is significant at the 0.05 level (p-value=0.014). on the other hand, no interaction effect is detected for the store congestion feature (p-value=0.161). means and standard deviations for the dependent variable are reported in table 3. at this point, it is not possible to identify the alternatives that minimize the preparation time. although the mean time per item is known and very different, deviations must also be taken into account. this first stage of analysis only indicates that the differences among the alternatives of the store size, assortment and backroom are significant. in order to complete the study, it is necessary to carry out a second stage of pair analysis, comparing the results obtained for the alternatives of each group. means and standard deviations and anova results for pair analysis are shown in table 4. regarding the store size variable, it can be seen that the lowest values are obtained in convenience stores (m=61.1). these stores present significant differences at the 0.01 level, in times per item, com-pared to the supermarkets (m=81.2) and hypermarkets (m=85.4). the two larger store types do not show significant differences between them (p-value=0.465). thus, the statistical analysis shows that the option that minimizes preparation time is the assignment of convenience stores. for the assortment variable, the highest values are obtained with a large assortment (m=92.1), as table 2. individual anova results (n=200). store features sum sq df mean sq f-value p-value store size 11018 2 5509 10.07 < 0.001*** assortment 19636 2 9818 19.50 <0.001*** backroom 5012 2 2506.2 4.34 0.014* congestion 2184 2 1092.1 1.85 0.161 table 3. mean scores and standard deviations for the dependent variable (n = 200). store features alternatives mean (sd) store size convenience 61.1 (7.0) supermarket 81.2 (26.4) hypermarket 85.4 (22.7) assortment small assortment 73.7 (18.8) medium assortment 70.7 (19.9) large assortment 92.1 (26.0) backroom no 73.8 (16.1) yes 78.3 (26.4) yes and picking 87.0 (24.1) congestion low traffic 85.9 (26.2) medium traffic 77.7 (23.4) high traffic 79.6 (23.9) table 4. pair analysis results for the variable store size. store features pairs mean (sd) p-value store size convenience 61.1 (7.0) <0.001*** supermarket 81.2 (26.4) convenience 61.1 (7.0) <0.001*** hypermarket 85.4 (22.7) supermarket 81.2 (26.4) 0.465 hypermarket 85.4 (22.7) table 5. pair analysis results for the variable assortment. store features pairs mean (sd) p-value assortment small 73.7 (18.8) 0.741 medium 70.7 (19.9) small 73.7 (18.8) <0.001*** large 92.1 (26.0) medium 70.7 (19.9) <0.001*** large 92.1 (26.0) table 6. pair analysis results for the variable backroom. store features pairs mean (sd) p-value backroom no 73.8 (16.1) 0.592 yes 78.3 (26.4) no 73.8 (16.1) 0.021* yes and picking 87.0 (24.1) yes 78.3 (26.4) 0.059 yes and picking 87.0 (24.1) int. j. prod. manag. eng. (2022) 10(2), 183-193creative commons attribution-noncommercial-noderivatives 4.0 international analyzing store features for online order picking in grocery retailing: an experimental study 189 http://creativecommons.org/licenses/by-nc-nd/4.0/ there are significant differences at the 0.01 level, not only with the medium assortment but also the small assortment. however, the difference between the means of medium assortment (m=70.7) and low assortment (m=73.7) is not significant (p-value=0.741). therefore, it cannot be guaranteed which of the two options is better; although it can be pointed out that the most unfavorable of the three options is the large assortment. for the backroom variable, significant differences were only observed at the 0.05 level (p-value=0.021) between having or not having a backroom and, furthermore, undertaking picking in it. what seems to be clear is that the lowest values are obtained when picking is undertaken in the aisles of the store (m=73.8, without backroom; m=78.3, with backroom for replenishment). finally, the different alternatives of the congestion variable have not been studied by pair analysis, since in the first stage of the analysis this variable did not turn out to be significant. 6. discussion this statistical study presents valuable information for supermarket chains that meet online demand with a store-based model. authors have identified which store characteristics speed up the online order picking time. as a result, it seems that the most agile way to carry out this process is in a convenience store, with a small or medium assortment of products and using the aisles of the store for the picking process. with respect to the influence of customer congestion, no significant interactions with picking time have been observed. in this section, these results are compared and discussed with the previous literature. on the one hand, convenience stores seemed to be the least efficient stores for carrying out picking (do and omdahl, 2018). however, our results show that this type of store can lead to a reduction of up to 40% of the total order picking time, due to the reduction in the distance traveled by the pickers. furthermore, some of the doubts existing in the literature regarding the assortment size have been resolved, as this study shows that picking time can increase by up to 30% if the assortment is large. this result coincide with the study by wollenburg et al. (2018), in which the authors have identified that a wide range of products involves a greater travel of the pickers and, consequently, an increase in the costs of the activity. on the other hand, our statistical analysis shows that preparing in the aisles of the store is the most effective alternative. however, the literature highlighted that preparing in the backroom speeds up the preparation of online orders by avoiding contact with the offline clients (mangiaracina et al., 2018). even though this alternative can reduce the customer service due to possible stock-outs, the process is more agile because the products are collected in customer units, avoiding double handling. regarding the store congestion variable, it has been demonstrated that this is not a relevant variable. this result agrees with that presented by salgado (2015), who sustains that store congestion is less of a problem when picking density is high and the pickers have to stop more often because the pickers spend more time collecting and less time travelling. in summary, the results of our study show discrepancies with the extant literature in two aspects: on the one hand, in relation to the optimal size of the stores in which to prepare online orders; on the other hand, about whether or not to use the warehouse for this task. furthermore, authors confirm that congestion is not a variable that has a significant effect on the picking time, despite the fact that the literature on this subject is very limited. finally, the results are in line with previous studies defining low or medium product ranges as the best options to minimize picking time. 7. conclusions by undertaking this study, it has been possible to determine which features of traditional stores affect online order picking time. this information is of great value for supermarket chains that are opting for a store-based model and want to optimize picking times. in this sense, the study has important implications at a strategic level as it permits improvements in the decision to assign orders to preparation stores. to date, the only decision criterion used by stores have been the distance of the store from the customer and the population density, which may not be the best options. the results obtained by applying a one-way anova statistical analysis show that the most efficient stores for picking are those that have a smaller surface int. j. prod. manag. eng. (2022) 10(2), 183-193 creative commons attribution-noncommercial-noderivatives 4.0 international vazquez-noguerol et al. 190 http://creativecommons.org/licenses/by-nc-nd/4.0/ area (convenience stores) and a range that is not excessively large. stores of this type reduce the time and route length when searching. the results also indicate, although with a lower level of dependency, that carrying out picking in the aisles of the store itself is a better option than in the backroom. finally, the congestion variable linked to store traffic was not shown to be significant in the efficiency of the process. all these results have been contrasted and discussed with reference to the extant literature. thus, the proposed research question has been answered by defining the store features that will optimize the online order picking process. similarly, this publication makes it possible to resolve future lines for research presented in the literature (vazqueznoguerol et al., 2021) by identifying the best store to minimize picking time and determine how traditional sales affect the online order picking process. there are some limitations in this study, which also provide an opportunity for future research. first, in order to define the store characteristics, the ranges of values used were those defined by the firm on which the case study was based. however, these ranges of values could vary if another firm were analyzed and this could lead to slightly different results. second, regarding the online orders selected for measuring timings, these were only broken down in terms of the number of items. no distinction has been made as regards the type of products: dry, fresh, or frozen. future research could focus on determining how order characteristics can affect picking times. thus, the time and resources needed for e-fulfilment could be estimated and anticipated. once the approximate picking time for an order is known, demand predictions could be made in order to optimize order planning and increase control over picking costs. this would all be of great interest at a strategic level to analyze the growth in online demand and, if this reached a sizeable volume, to propose a change to the warehouse-based model. another important line of research would be to study how technological developments affect the efficiency of picking processes. in this sense, comparative studies could be undertaken on the timing and efficiency of order picking in several scenarios using distinct levels of automation. in conclusion, this study and the lines of future research proposed are intended to serve as a reference for the optimization of the processes in supermarket chains operating in the online channel. this information will be of great value for sustainable growth in an environment of digital transformation in which there is strong competition between companies and where consumer requirements are increasingly demanding. references broekmeulen, r.a., sternbeck, m.g., van donselaar, k.h., & kuhn, h. (2017). decision support for selecting the optimal product unpacking location in a retail supply chain. european journal of operational research, 259(1), 84-99. https://doi.org/10.1016/j.ejor.2016.09.054 chintagunta, p.k., chu, j., & cebollada, j. (2012). quantifying transaction costs in online/off-line grocery channel choice. marketing science, 31(1), 96-114. https://doi.org/10.1287/mksc.1110.0678 dias, f.f., lavieri, p.s., sharda, s., khoeini, s., bhat, c.r., pendyala, r.m., & srinivasan, k.k. (2020). a comparison of online and in-person activity engagement: the case of shopping and eating meals. transportation research part c: emerging technologies, 114, 643-656. https://doi.org/10.1016/j.trc.2020.02.023 do, v.c., & omdahl, k. (2018). efficiency of e-grocery: challenges and suggestions. (doctoral thesis). university of stavanger, norway. fernie, j., sparks, l., & mckinnon, a.c. (2010). retail logistics in the uk: past, present and future. international journal of retail & distribution management, 38(11/12), 894-914. https://doi.org/10.1108/09590551011085975 forbes techonology council homepage, david mounts, the digital transformation of retail grocery, https://www.forbes. com/sites/forbestechcouncil/2020/06/26/the-digital-transformation-of-retail-grocery/?sh=11e66b287300. accessed: february 2021. gallino, s., & moreno, a. (2019). operations in an omnichannel world. springer international publishing. https://doi.org/10.1007/978-3-030-20119-7 hailu, g. (2020). economic thoughts on covid-19 for canadian food processors. canadian journal of agricultural economics/revue canadienne d’agroeconomie, 68(2), 163-169. https://doi.org/10.1111/cjag.12241 hendrickson, m.k. (2020). covid lays bare the brittleness of a concentrated and consolidated food system. agriculture and human values, 37(3), 579-580. https://doi.org/10.1007/s10460-020-10092-y int. j. prod. manag. eng. (2022) 10(2), 183-193creative commons attribution-noncommercial-noderivatives 4.0 international analyzing store features for online order picking in grocery retailing: an experimental study 191 https://doi.org/10.1016/j.ejor.2016.09.054 https://doi.org/10.1287/mksc.1110.0678 https://doi.org/10.1016/j.trc.2020.02.023 https://doi.org/10.1108/09590551011085975 https://www.forbes.com/sites/forbestechcouncil/2020/06/26/the-digital-transformation-of-retail-grocery/?sh=11e66b287300 https://www.forbes.com/sites/forbestechcouncil/2020/06/26/the-digital-transformation-of-retail-grocery/?sh=11e66b287300 https://doi.org/10.1007/978-3-030-20119-7 https://doi.org/10.1111/cjag.12241 https://doi.org/10.1007/s10460-020-10092-y http://creativecommons.org/licenses/by-nc-nd/4.0/ hofmann, m., & meyer-nieberg, s. (2018). time to dispense with the p-value in or?. central european journal of operations research, 26(1), 193-214. https://doi.org/10.1007/s10100-017-0484-9 hübner, a., holzapfel, a., & kuhn, h. (2015). operations management in multi-channel retailing: an exploratory study. operations management research, 8(3), 84-100. https://doi.org/10.1007/s12063-015-0101-9 hübner, a., holzapfel, a., kuhn, h., & obermair, e. (2019). distribution in omnichannel grocery retailing: an analysis of concepts realized. operations in an omnichannel world. springer. https://doi.org/10.1007/978-3-030-20119-7_12 ishfaq, r., & bajwa, n. (2019). profitability of online order fulfillment in multi-channel retailing. european journal of operational research, 272(3), 1028-1040. https://doi.org/10.1016/j.ejor.2018.07.047 komijan, a.r., & delavari, d. (2017). vehicle routing and scheduling problem for a multi-period, multi-perishable product system with time window: a case study. international journal of production management and engineering, 5(2), 45-53. https://doi.org/10.4995/ijpme.2017.5960 kumar, a., mangla, s.k., kumar, p., & song, m. (2021). mitigate risks in perishable food supply chains: learning from covid-19. technological forecasting and social change, 166, 120643. https://doi.org/10.1016/j.techfore.2021.120643 mahajan, k., & tomar, s. (2021). covid-19 and supply chain disruption: evidence from food markets in india. american journal of agricultural economics, 103(1), 35-52. https://doi.org/10.1111/ajae.12158 mangiaracina, r., perego, a., seghezzi, a., & tumino, a. (2018). optimizing store-based picking in the e-grocery: a model to assess costs and benefits. industrial systems engineering. mishra, n., & mukherjee, s. (2019). effect of artificial intelligence on customer relationship management of amazon in bangalore. international journal of management, 10(4), 168-172. https://doi.org/10.34218/ijm.10.4.2019.016 paul, j., agatz, n., & savelsbergh, m. (2019). optimizing omni-channel fulfillment with store transfers. transportation research part b: methodological, 129, 381-396. https://doi.org/10.1016/j.trb.2019.10.002 pires, m., pratas, j., liz, j., & amorim, p. (2017). a framework for designing backroom areas in grocery stores. international journal of retail & distribution management. 45(3), 230-252. https://doi.org/10.1108/ijrdm-01-2016-0004 pires, m., silva, e., & amorim, p. (2021). solving the grocery backroom layout problem. international journal of production research, 59(3), 772-797. https://doi.org/10.1080/00207543.2019.1708990 prado-prado, j.c., garcía-arca, j., fernández-gonzález, a.j., & mosteiro-añón, m. (2020). increasing competitiveness through the implementation of lean management in healthcare. international journal of environmental research and public health, 17(14), 4981. https://doi.org/10.3390/ijerph17144981 rai, h.b., verlinde, s., macharis, c., schoutteet, p., & vanhaverbeke, l. (2019). logistics outsourcing in omnichannel retail: state of practice and service recommendations. international journal of physical distribution & logistics management.49(3), 267-286. https://doi.org/10.1108/ijpdlm-02-2018-0092 rodríguez-garcía, m., gonzález-romero, i., bas, á.o., & prado-prado, j.c. (2021). e-grocery retailing: from value proposition to logistics strategy. international journal of logistics research and applications, 1-20. https://doi.org/10.1080/13675567.2021.1900086 rodríguez-garcía, m., domínguez-caamano, p., & prado-prado, j.c. (2016) the new supply chain in the era of e-retailers: a state-of-the-art literature review dirección y organización, 59, 18-31. https://doi.org/10.37610/dyo. v0i59.491 rooderkerk, r.p., & kök, a.g. (2019). omnichannel assortment planning. operations in an omnichannel world. salgado, t.m.f.d.n. (2015). in-store order picking routing: a biased random-key genetic algorithm approach. (doctoral thesis). university of porto, portugal. schoen, q., sanchis, r., poler, r., lauras, m., fontanili, f., & truptil, s. (2018). categorisation of the main disruptive events in the sensitive products transportation supply chains. international journal of production management and engineering, 6(2), 79-89. https://doi.org/10.4995/ijpme.2018.10369 seidel, s. (2021). one goal, one approach? a comparative analysis of online grocery strategies in france and germany. case studies on transport policy, 9(4), 1922-1932. https://doi.org/10.1016/j.cstp.2021.10.013 seidel, s., blanquart, c., & ehrler, v. (2016). same–same but different? a comparison of food retail and distribution structures in france and germany. case studies on transport policy, 4(1), 29-37. https://doi.org/10.1016/j.cstp.2015.09.001 shani, a.b., & coghlan, d. (2021). action research in business and management: a reflective review. action research, 19(3), 518-541. https://doi.org/10.1177/1476750319852147 suel, e., daina, n., & polak, j.w. (2018). a hazard-based approach to modelling the effects of online shopping on intershopping duration. transportation, 45(2), 415-428. https://doi.org/10.1007/s11116-017-9838-3 tompkins, j.a., white, j.a., bozer, y.a., & tanchoco, j.m.a. (2010). facilities planning. john wiley & sons. int. j. prod. manag. eng. (2022) 10(2), 183-193 creative commons attribution-noncommercial-noderivatives 4.0 international vazquez-noguerol et al. 192 https://doi.org/10.1007/s10100-017-0484-9 https://doi.org/10.1007/s12063-015-0101-9 https://doi.org/10.1007/978-3-030-20119-7_12 https://doi.org/10.1016/j.ejor.2018.07.047 https://doi.org/10.4995/ijpme.2017.5960 https://doi.org/10.1016/j.techfore.2021.120643 https://doi.org/10.1111/ajae.12158 https://doi.org/10.34218/ijm.10.4.2019.016 https://doi.org/10.1016/j.trb.2019.10.002 https://doi.org/10.1108/ijrdm-01-2016-0004 https://doi.org/10.1080/00207543.2019.1708990 https://doi.org/10.3390/ijerph17144981 https://doi.org/10.1080/13675567.2021.1900086 https://doi.org/10.37610/dyo.v0i59.491 https://doi.org/10.37610/dyo.v0i59.491 https://doi.org/10.4995/ijpme.2018.10369 https://doi.org/10.1016/j.cstp.2021.10.013 https://doi.org/10.1016/j.cstp.2015.09.001 https://doi.org/10.1177/1476750319852147 https://doi.org/10.1007/s11116-017-9838-3 http://creativecommons.org/licenses/by-nc-nd/4.0/ vazquez-noguerol, m., comesaña-benavides, j., poler, r., & prado-prado, j.c. (2020). an optimisation approach for the e-grocery order picking and delivery problem. central european journal of operations research. https://doi.org/10.1007/s10100-020-00710-9 vazquez-noguerol, m., gonzález-boubeta, i., portela-caramés, i., & prado-prado, j.c. (2021). rethinking picking processes in e-grocery: a study in the multichannel context. business process management journal. 27(2), 565-589. https://doi.org/10.1108/bpmj-04-2020-0139 wollenburg, j., hübner, a., kuhn, h., & trautrims, a. (2018). from bricks-and-mortar to bricks-and-clicks: logistics networks in omni-channel grocery retailing. international journal of physical distribution & logistics management. https://doi.org/10.1108/ijpdlm-10-2016-0290 zhao, g., liu, s., lopez, c., chen, h., lu, h., mangla, s.k., & elgueta, s. (2020). risk analysis of the agri-food supply chain: a multi-method approach. international journal of production research, 58(16), 4851-4876. https://doi.org/10.1080/00207543.2020.1725684 int. j. prod. manag. eng. (2022) 10(2), 183-193creative commons attribution-noncommercial-noderivatives 4.0 international analyzing store features for online order picking in grocery retailing: an experimental study 193 https://doi.org/10.1007/s10100-020-00710-9 https://doi.org/10.1108/bpmj-04-2020-0139 https://doi.org/10.1108/ijpdlm-10-2016-0290 https://doi.org/10.1080/00207543.2020.1725684 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering received: 2021-12-03 accepted: 2022-06-27 sustainable and agile manufacturing outsourcing partner selection: a literature review mohammad akhtar galgotias university, greater noida, u.p., 203201, india. * mk71b@yahoo.com, mohammad.akhtar@galgotiasuniversity.edu.in abstract: outsourcing to third party to manage non-core activities helps the firm to focus on core activities. manufacturing firms are outsourcing product development, manufacturing, logistics, customer care etc. to enhance production capacity and flexibility, and to reduce operational costs, which in turn can improve profitability and competitive advantage of the enterprise. sustainability in operations and supply chain is gaining momentum due to increased global environmental concern, pressures from consumers and communities, and enforced regulations. volatile and uncertain business environment necessitates the adoption of agility and flexibility to effectively manage manufacturing and supply chain. globalisation has made the market very competitive and hence manufacturing firms are adopting manufacturing outsourcing to third parties. selecting a sustainable and agile manufacturing outsourcing partner (mps) is crucial as it will improve sustainability, efficiency, and effectiveness of the supply chain and competitive advantage to the firm. detailed literature review on sustainable and agile manufacturing outsourcing partner selection has been carried out from ebsco data base and goggle scholar. selection criteria used are classified into agile, operational, economic, environmental and social. the techniques use are mostly multi criteria decision making methods (mcdm) while few have adopted programming techniques. discussion, implication and the scope of future work is also provided. key words: manufacturing outsourcing partner selection, contract manufacturer selection, outsourcing partner selection, manufacturing outsourcing criteria, multi criteria decision making. 1. introduction supply chain starts with procurement of raw materials, storage, inventory control, to transportation, and distribution of finished goods to customers to meet their demand. outsourcing establishes a contract with an outside party to manage its non-core work in order to efficiency improvement, cost reduction, increase profits and to focus on core activities. outsourcing has become an important business approach to manufacture products more efficiently by a contract manufacturer to gain competitive advantages. logistics, finance, accounting, legal services, marketing and customer care are already outsourced to external service providers. new technologies, globalization and increasing demand from end-users provide extra scope for outsourcing activities (yang et al., 2007). ineffective outsourcing may lead to loss of core capabilities and competencies, and may cause to business failure (wang and yang, 2007). to attain effective manufacturing outsourcing, manufacturing outsourcing partner selection (mps) or contract manufacturer selection (cms) is essential. mps affects upstream, downstream and reverse supply to cite this article: akhtar, m. (2022). sustainable and agile manufacturing outsourcing partner selection: a literature review. international journal of production management and engineering, 10(2), 143-158. https://doi.org/10.4995/ijpme.2022.16807 https://doi.org/10.4995/ijpme.2022.16807 int. j. prod. manag. eng. (2022) 10(2), 143-158creative commons attribution-noncommercial-noderivatives 4.0 international 143 http://creativecommons.org/licenses/by-nc-nd/4.0/ chain operations, hence both qualitative and quantitative factors need to be examined carefully. with the increased global environmental awareness and pressure, firms’ decision-makers should consider environmental perspective in all commercial activities (afred and adam, 2009). sustainable supply chain will improve environmental efficiency and social responsibility, meet stakeholder requirements and enhance firm competitiveness and profit (gualandris et al., 2014). environmental factors in product design, material purchasing, and supply network design have become important (hervani et al., 2005; sarkis, 2005). manufacturing outsourcing partner selection (mps) directly affects the sustainable initiatives implementation. for successful sustainable business practices, companies need to consider economic, environmental, and social sustainable criteria for outsourcing partners performances evaluation (govindan et al., 2013). companies should share their capabilities and resources with outsourcing partners in areas of green and technological innovations, environmental management systems, social responsibilities and sustainable initiatives (luthra et al., 2017). agile supply chain is needed to efficiently and effectively respond to volatile business environment (christopher, 2000). to achieve necessary levels of sc agility, it is essential to align supply partners with firm operations to improve efficiency (wu and barnes, 2011). therefore, a sustainable and agile (susgile) supply chain is desired to meet the sustainability obligation and business volatility. manufacturing outsourcing partner selection process (mps) involves several criteria, number of outsourcing partners and multiple decision makers (dm) and it is said to be a multi-criteria group decision making (mcdm) problem. a structured approach should be adopted to select the right criteria, and technique to assess and select the manufacturing outsourcing partner (mp) or contract manufacturer (cm). in the literature, many selection criteria and techniques including mcdm have been applied by various researchers for mps. research questions has been framed that should be addressed in the literature review. rq1. what are the operational criteria used for manufacturing outsourcing partner selection? rq2. what are the economic, environmental and social sustainable criteria used for manufacturing outsourcing partner selection? rq3. what are the agile criteria used for manufacturing outsourcing partner selection? rq4. what are the techniques used for the assessment and ranking of manufacturing outsourcing partners? the paper is organised in following sections. section two presents the detailed review of the literature; section three provides the findings and discussion, and section four offers the conclusion and scope for future work. 2. review of literature to undertake literature review, guidelines provided by denyer and tranfield (2009) have been adopted, which help researchers to formulate a research question and define the boundaries of a literature review. the literature review is carried out by searching the word ‘manufacturing outsourcing partner selection’, ‘manufacturing outsourcing provider selection’, ‘contract manufacturer selection’, ‘manufacturing outsourcing’, ‘contract manufacturing’, ‘strategic outsourcing partner’, ‘strategic outsourcing provider’ ‘vendor selection’, ‘triple bottom line sustainability’, ‘agile criteria’, ‘sustainable criteria’, ‘economic criteria’, ‘environmental criteria’, and ‘social criteria’ in ebsco database and google scholar. outsourcing is the process where an external company takes the responsibility of certain activities and processes through a contract with the company (yang et al. 2007). outsourcing reduces operating costs and improve competitiveness of a firm. outsourcing manufacturing activities to mp or cm decreases operational cost, increases the flexibility of production capability and improves the profit. it is therefore, crucial for a company to select appropriate mp. supply chain managers have incorporated sustainability in manufacturing outsourcing partnership which improved firm’s competitiveness (govindan et al. 2013; luthra et al. 2017). suppliers and vendors selection affects supply chain operations and performances which is evident from the significant number of studies found in the literature (malviya et al., 2018). right supplier selection reduces purchasing costs, improves enduser’s satisfaction, and competitiveness in the market (onut et al., 2009), while wrong supplier selection can negatively impact the operational and int. j. prod. manag. eng. (2022) 10(2), 143-158 creative commons attribution-noncommercial-noderivatives 4.0 international akhtar 144 http://creativecommons.org/licenses/by-nc-nd/4.0/ financial performance of the firm (bhattacharya and singh, 2019). selection right supplier is an important part of sustainable supply chain (sen et al., 2018) and key concern for business firms (seuring, 2013; grimm et al., 2014). it is challenging for many firms to adopt a dependable approach for supplier selection (ortiz-barrios et al., 2017; sen et al., 2018) that can enhance cost savings, delivery, quality, flexibility, service level (govindan et al., 2013), and innovation (nair et al., 2015). ebrahimipour et al. (2016) suggested product characteristics, finished products reliability and product life cycle to choose the right vendor and suppliers for manufacturing firms. process of supplier selection is to select the right criteria and right technique (büyüközkan and göçer, 2019). fan et al. (2020) studied battery outsourcing for electric vehicle manufacturers considering battery production cost, electric vehicle manufacturing and assembly cost, government subsidy. helo et al. (2021) designed cloud ecosystem for cloud-based collaborative manufacturing portals for sheet metal manufacturing companies. lahiri et al. (2022) carried out meta-analysis to examine the effect of industrial nature of activity (manufacturing vs. services), value chain activity (core vs. non-core), and provider’s location (domestic vs. international) in sourcing on firm performance and found that effect is stronger for non-core and international outsourcing equally for manufacturing and service outsourcing. 2.1. criteria for manufacturing outsourcing partner evaluation it is crucial for decision-makers to identify effective evaluation criteria, as well as assess outsourcing partner’s feasibility and compatibility prior to outsourcing. a number of evaluation criteria for mps have been used in the literature. mummalaneni et al. (1996) adopted price, quality, on-time delivery, responsiveness, expertise, and supplier relationship. ho et al. (2010) suggested price, quality and delivery. liou et al. (2012) considered cost (flexibility in billing, cost saving), quality (on-time rate, customers’ satisfactions, knowledge skills), risk (management control loss, labor union, information security), and compatibility (information sharing, flexibility, relationship) criteria outsourcing provider selection in a taiwanese airline. garg and sharma (2020) adopted economic factors (service delivery and access, firm performance and reputation, resources capacity, financial capacity, outsourcing benefits, technical ability, and communication), environmental factors (green certification and accreditation, emission and pollution minimization, green purchasing and designing, green manufacturing and marketing, waste minimization, green practices, green packaging, energy efficiency, cleaner technology, and reverse logistics), and social factors (working conditions, health and safety, employee rights and fair wages, women specific issues, employee and community equity, social welfare, community connection and support, ethical and transparent practices) for sustainable outsourcing partner selection in indian electronics company. chen and hung (2010) used service criteria such as on-time delivery, process capacity, experience, response to changes, and reputation; financial criteria such as services cost, long-term viability, and finacial stability; quality criteria such as product quality, processes for quality control, and program for continuous improvement; compliance criteria such as goods manufacturing practices compliance, environmental health and safety, intellectual property infringement; and culture criteria such as lasting and mutually profitable relationships, interaction ability for mps in pharmaceutical r&d. for sustainable supplier and vendor selection, ulutas et al. (2016) used cost, supplier production capacity, financial position, order requirement, sectoral price compliance, percentage defective, percentage late delivered, volume flexibility, technological capability reputation, and communication issues. luthra et al. (2017) adopted product price, quality, delivery & service of product, transportation cost, product profit, lead time, production capacity, flexibility, green manufacturing, waste management, green product design, green purchasing, green packing and labeling, technological and financial capability, green management, green research and innovation, pollution prevention, environment management systems, environmental costs, environmental competencies, employees interests and rights, occupational health and safety, stakeholders rights, and information disclosure. fallahpour et al. (2017) considered cost (material cost, after-sales service cost, freight cost), delivery & service (on-time delivery, lead time, aftersales service, flexibility, complaint resolution time), quality (internal quality audit process, abnormal quality handling capability, rejection rate), flexibility (delivery time flexibility, ordering flexibility, discount flexibility) criteria. song et al. (2017) adopted ten criteria; cost, quality, delivery, int. j. prod. manag. eng. (2022) 10(2), 143-158creative commons attribution-noncommercial-noderivatives 4.0 international sustainable and agile manufacturing outsourcing partner selection: a literature review 145 http://creativecommons.org/licenses/by-nc-nd/4.0/ resource consumption, eco-design, recycling, environmental management system, occupational safety and health, rights and welfare of employees, and community training and development. awasthi et al. (2018) adopted five sustainable criteria; economic, quality, environment, social, and global risk. cheraghalipour and farsad (2018) used three main criteria of economic (delivery, quality, loyalty, cost, service, financial situation, technology), environmental (product performance, environmental commitment, environmental pollution, greenhouse gas, environmental management), and social (working hours, worker safety and health, freedom of association and wages, social commitment, social management). arabsheybani et al. (2018) used delivery, quality, green supply chain, suppliers of the supplier, environmental management system (ems), worker safety and health, worker safety cost, worker dismissal, and employee interests & rights criteria. sinha and anand (2018) adopted criteria such as cost, delivery reliability, quality, technology capability, green product, financial situation, pollution control, environmental management system, green image, health & safety contractual, social responsibility management, local community and stakeholder influence. vasiljević et al. (2018) considered price, delivery, quality, environmental and social criteria that help organizations in achieving longterm economic sustainability, ecological stability, and market position. sen et al. (2018) proposed price, on-time delivery, quality, flexibility, service, production facility, financial capability, innovation, green design, green product, resource consumption, waste, recycling, ems, work safety, stakeholders’ rights, and information disclosure. for agile contract manufacturer selection, adali and isik (2017) adopted product cost, on-time delivery, production capacity, equipment, material quality, geographic location and reliability criteria, hu and yu (2015) considered cost, delivery, quality, and flexibility. supply chain agility studies have been conducted across industries; manufacturing (blome et al., 2013; um, 2017; al-shboul, 2017; kim and chai, 2017); auto components (dubey et al., 2018); fashion and textiles (ngai et al., 2011; chan et al., 2017); electronics (tse et al., 2016; wu et al., 2017); telecommunications (collin and lorenzin, 2006), and oil and gas (yusuf et al., 2014). in moroccan manufacturing companies, barhmi (2019) studied the supply chain agility and resilience effect on the supply chain performance. 2.2. technique for manufacturing outsourcing partner selection manufacturing outsourcing partner selection involves multiple inter-related criteria, alternatives and decision makers and hence complex process. in the literature, various techniques including mcdm have been applied for mps by various scholars. vendor selection for outsourcing in taiwanese semiconductor company, lin et al. (2010) applied analytic network process (anp) method. parthiban et al. (2012) adopted fuzzy swot and dea for vendor selection. liou et al. (2012) adopted integrated dematel, fuzzy preference programming and anp method to select outsourcing provider in a taiwanese airline. for outsourcing provider selection, hsu et al. (2013) applied danp and gra method. festel et al. (2014) proposed an action research-based selection of strategic outsourcing partner in a global pharmaceutical manufacturing company. hu and yu (2015) proposed an integrated voting method and the goal programming (gp) model for electronic contract manufacturer selection. for vendor selection in the steel industry, kar (2015) adopted delphi and fuzzy ahp method. rezaeisaray et al. (2016) utilized dematel-fuzzy anp-dea to select outsourcing supplier in pipe manufacturing company. adali and isik (2017) applied critic and multi-attribute utility theory (maut) methods for cms. momeni and vandchali (2017) proposed a data envelopment analysis (dea) model using evidential reasoning (er) algorithm for strategic outsourcing in an iranian software company. ji et al. (2018) proposed neutrosophic linguistic sets based mabac–electre method for outsourcing provider selection. büyüközkan and göçer (2019) proposed pythagorean fuzzy ahp and copras for digital supply chain partner selection. song (2019) adopted ahp for selection of outsourcing partner in korean pharmaceutical r&d. fei et al. (2019) presented dempster–shafer evidence theory (ds theory) and vikor for supplier selection. ghorabaee et al. (2017) applied interval type-2 fuzzy critic– waspas for selection of third party logistics provider. rostamzadeh et al. (2018) proposed fcritic and ftopsis method, while abdel-basset and mohamed (2020) applied a plithogenic critic and topsis model for managing sustainable supply chain risk. awasthi and kannan (2016) proposed nominal group technique and fuzzy vikor for green supplier development program. chen et al. (2019) designed a model based on capability index and manufacturing time performance index for outsourcing partner selection. percin (2019) adopted int. j. prod. manag. eng. (2022) 10(2), 143-158 creative commons attribution-noncommercial-noderivatives 4.0 international akhtar 146 http://creativecommons.org/licenses/by-nc-nd/4.0/ fswara and fuzzy axiomatic design method for selecting outsourcing provider in turkish chemical manufacturing company. liaw et al. (2020) proposed dematel-critic method for criteria weight and classifiable topsis to classify green manufacturing outsourcing providers in taiwanese multinational machine tool manufacturing firm. for malaysian manufacturing smes, zulkiffli and padlee (2021) used confirmatory factor analysis to study sustainable outsourcing impact on the competitive capabilities and business performance. yazdani et al. (2021) assessed outsourcing risk using triangular fuzzy hesitant sets, failure mode and effect analysis (fmea) and combined compromise solution (cocoso) in iranian chemical company. singh and sarkar (2021) applied integrated ahp and vikor method for sustainable contract manufacturer selection in automotive industry. akhtar et al. (2021) applied stochastic fuzzy topsis to select sustainable vendor in the indian petroleum refining sector. haoues et al. (2021) presented genetic algorithm techniques under reliability maintenance constraint to minimize total cost for inhouse and outsourced manufacturing maintenance. teerasoponpong and sopadang (2022) applied genetic algorithm and artificial neural network (ann) for sourcing and inventory management, which reduced raw materials purchasing cost, order interval and on-hand inventory cost in medium-sized food company. for sustainable supplier and vendor evaluation, dobos and vörösmarty (2014) utilized dea with the common weights analysis (cwa) method; bai and sarkis (2014) adopted rough set theory based dea; zarbakhshnia and jaghdani (2018) adopted two-stage dea; mohammed et al. (2018) proposed hybrid ahp and topsis; sivakumar et al. (2015) used ahp and taguchi loss functions; trapp and sarkis (2016) applied integer programming technique; luthra et al. (2017) applied integrated ahp and vikor; and cheraghalipour and farsad (2018) used bwm, milp, and revised multi-choice goal programming. garg and sharma (2020) adopted integrated bwm-vikor method for selection of sustainable outsourcing partner in an indian electronics company. for agile supplier and vendor selection, hasan et al. (2008) adopted dea and anp methods; luo et al. (2009) proposed a radial basis function based ann; alimardani et al. (2014) demonstrated hybrid dematel, anp and topsis application; and beikkhakhian et al. (2015) applied ism for agile selection criteria and integrated topsis-ahp method for vendor ranking. lee et al. (2015) used fahp and ftopsis for agile supplier selection and studied the impact of agility criterion and order allocation strategy on business performance. matawale et al. (2016) presented fuzzy multi-level (fml) approach for selecting agile supplier and compared the result with ftopsis and fmoora. goker (2021) applied intuitionistic fuzzy cognitive map and copras method for agile outsourcing provider selection in turkish white goods industry. fuzzy variant of mcdm methods have also been reported by many researchers in the literature. chen and hung (2010) adopted fahp and ftopsis method for selecting outsourcing manufacturing partner for pharmaceutical r&d. akhavan et al. (2015) applied fuzzy quantitative strategic planning matrix-based swot for strategic alliance planning; f-aras, f-copras, f-topsis, and f-moora for strategic outsourcing partner evaluation in an iranian car manufacturer company. for supplier selection with triple bottom line sustainability, wang et al. (2019) proposed fuzzy ahp and topsis in vietnamese garment industry. dos santos et al. (2019) adopted shanon entropy and ftopsis in the furniture industry. for sustainable supplier selection, zhou et al. (2016) utilized type-2 multi-objective dea; fallahpour et al. (2017) proposed fahp and ftopsis; paydar et al. (2017) and arabsheybani et al. (2018) utilized integrated f-moora and failure mode and effects analysis (fmea); awasthi et al. (2018) proposed fahp and fvikor; and mohammed et al. (2018) adopted fahp, ftopsis and multi-objective programming model. rabbani et al. (2019) presented interval-valued fuzzy group decision model-based reference point systems with fuzzy possibilistic statistical concepts. 3. findings and discussion based on the literature review, discussions on the findings are summarized under following headings. 3.1. distribution of articles journal wise though large number of articles are published on supplier selection but few articles related to manufacturing outsourcing partner selection are available in the literature. the manufacturing outsourcing partner selection with sustainability consideration are rarest. the manufacturing outsourcing partner selection articles published in int. j. prod. manag. eng. (2022) 10(2), 143-158creative commons attribution-noncommercial-noderivatives 4.0 international sustainable and agile manufacturing outsourcing partner selection: a literature review 147 http://creativecommons.org/licenses/by-nc-nd/4.0/ journals are listed table 1. most of the articles are published once in any journal. 3.2. distribution of articles year wise the number of articles published year wise from 2010 to 2021 are 18 as shown in figure 1. the highest number of articles published are 3 in years 2019 and 2020. figure 1. manufacturing outsourcing partner selection articles publication year wise 3.3. selection criteria from the literature review, important criteria for manufacturing outsourcing partner selection are operational, agile, and economic, environmental and social sustainable criteria as displayed in table 2. the operational criteria are process and production capacity, product quality, on-time delivery, response to customer needs, and technological ability. the agile criteria are production capability & flexibility, service level, lead time minimisation, delivery flexibility, sourcing flexibility, multi-skilled & flexible workforce, collaboration with partners for innovation & capacity enhancement, and customer driven innovation. the economic criteria are product price, resource consumption, and financial stability. the environmental sustainability criteria are green product, green manufacturing process, green r&d, and environmental management system. the social sustainability criteria are worker’s wages and welfare, worker’s occupational health & safety, and social welfare & community development. most of the papers cover operational, economic and agile criteria. latest trend is to include environmental and social sustainability criteria in selection process. 3.4. evaluation methods techniques used for outsourcing partner selection in the literature are shown in table 3 and 4. for criteria weight determination ahp, anp, critic, swara, bwm, mabac, cognitive map and fuzzy variant, and voting method have been used. for outsourcing partner evaluation and selection, mcdm methods such as ahp, topsis, anp, maut, electre, vikor, copras, gra, and fuzzy variant have been adopted. other methods such as goal programming (hu and table 1. manufacturing outsourcing partner selection articles publication in journals. s. no. journal name year no. of publication 1 international journal of production research 2010 1 2 expert system with applications 2010 1 3 expert system with applications 2013 1 4 journal of business chemistry 2014 1 5 technological and economic development of economy 2015 1 6 resources policy 2015 1 7 omega 2016 1 8 journal of modelling in management, 2016 1 9 european journal of multidisciplinary studies 2017 1 10 computers & industrial engineering 2018 1 11 journal of manufacturing technology management 2019 1 12 international journal of reliability, quality and safety engineering 2019 1 13 journal of pharmaceutical innovation 2019 1 14 international transactions in operational research 2020 1 15 symmetry 2020 1 16 environment, development and sustainability 2020 1 17 international journal of industrial and system engineering 2021 1 18 soft computing 2021 1 int. j. prod. manag. eng. (2022) 10(2), 143-158 creative commons attribution-noncommercial-noderivatives 4.0 international akhtar 148 http://creativecommons.org/licenses/by-nc-nd/4.0/ (table 2 continues in the next page) table 2. important criteria for sustainable and agile manufacturing outsourcing partner evaluation from the literature. criteria type criteria benefit/ nonbenefit description references o pe ra tio na l p er fo rm an ce c ri te ri a process and production capacity benefit production and process capacity chen and hung (2010), ulutas et al. (2016), luthra et al. (2017), adali and isik (2017) product quality benefit product quality and reliability liou et al. (2012), chen and hung (2010), mummalaneni et al. (1996), ho et al. (2010), ulutas et al. (2016), awasthi et al. (2018), cheraghalipour and farsad (2018), gören (2018), sinha and anand (2018), arabsheybani et al. (2018), song et al. (2017), luthra et al. (2017), fallahpour et al. (2017), hu and yu (2015), adali and isik (2017). on-time delivery benefit on time delivery to customers garg and sharma (2020), chen and hung (2010), mummalaneni et al. (1996), ulutas et al. (2016), luthra et al. (2017), fallahpour et al. (2017 fallahpour et al. (2017), song et al. (2017), cheraghalipour and farsad (2018), arabsheybani et al. (2018), sinha and anand (2018), vasiljević et al. (2018), adali and isik (2017), hu and yu (2015) response to customer needs benefit responsiveness and order fulfilment mummalaneni et al. (1996), chen and hung (2010). technological ability benefit technology and technical ability garg and sharma (2020), ulutas et al. (2016), luthra et al. (2017), cheraghalipour and farsad (2018), sinha and anand (2018) a gi le p er fo rm an ce c ri te ri a production flexibility and capability benefit production ability for variety of products to meet customer’s demand. chen and hung (2010), ulutas et al. (2016), awasthi et al. (2018), gören (2018), luthra et al. (2017), fallahpour et al. (2017), hu and yu (2015), adali and isik (2017). service level benefit providing service without stockout situation garg and sharma (2018), chen and hung (2010), bhutta and huq (2002), ulutas et al. (2016), awasthi et al. (2018), cheraghalipour and farsad (2018), fallahpour et al. (2017). lead time benefit lead time and variability minimisation liou et al. (2012), gören (2018), luthra et al. (2017), fallahpour et al. (2017). delivery flexibility benefit the ability to exploit various dimensions of delivery garg and sharma (2018), chen and hung (2010), mummalaneni et al. (1996), ho et al. (2010), ulutas et al. (2016), awasthi et al. (2018), cheraghalipour and farsad (2018), sinha and anand (2018), arabsheybani et al. (2018), song et al. (2017), luthra et al. (2017), fallahpour et al. (2017), hu and yu (2015), adali and isik (2017). sourcing flexibility benefit the availability of range of sourcing options garg and sharma (2018), chen and hung (2010), luthra et al. (2017), hu and yu (2015). multi-skilled and flexible workforce benefit multi-skilled workforce will provide flexibility in scheduling workers liou et al. (2012), chen and hung (2010), mummalaneni et al. (1996), ulutas et al. (2016). collaboration with partners benefit collaboration with suppliers will enhance innovation and capability garg and sharma (2018), liou et al. (2012), chen and hung (2010), mummalaneni et al. (1996), ulutas et al. (2016), gören (2018), sinha and anand (2018), cheraghalipour and farsad (2018), arabsheybani et al. (2018), luthra et al. (2017), awasthi et al. (2018). customer driven innovation benefit customer need based innovation liou et al. (2012), mummalaneni et al. (1996), sinha and anand (2018). int. j. prod. manag. eng. (2022) 10(2), 143-158creative commons attribution-noncommercial-noderivatives 4.0 international sustainable and agile manufacturing outsourcing partner selection: a literature review 149 http://creativecommons.org/licenses/by-nc-nd/4.0/ yu, 2016) and genetic algorithm (haoues et al., 2021) have also been used in few cases. dematel (hsu et al., 2013; rezaeisaray et al., 2016; wu et al., 2017; liaw et al., 2020) and ism (beikkhakhian et al., 2015) are used to show interrelationship among criteria. there is no clear trend on usage of any particular method. however, integrated fuzzy models have been used by majority researchers and among them fuzzy ahp and fuzzy topsis are two most frequently used methods: chen and hund (2010), liaw et al. (2020), beikkhakhian et al. (2015), lee et al. (2015), singh and sarkar (2021). there is no justification given for using a particular mcdm method in any of the papers. fuzzy theory and its variant (intuitionistic fuzzy sets, neutrosophic linguistic sets), and grey theory have been applied to deal with imprecision and ambiguity in decision makers’ judgments. most of the papers have used integrated models. it has been observed that ahp or fuzzy ahp is highly used for evaluation while dematel is used to find the interrelationship among the criteria. criteria type criteria benefit/ nonbenefit description references e co no m ic p er fo rm an ce c ri te ri a product price nonbenefit product price garg and sharma (2018), liou et al. (2012), mummalaneni et al. (1996), chen and hung (2010), ho et al. (2010), ulutas et al. (2016), gören (2018), luthra et al. (2017), fallahpour et al. (2017), song et al. (2017), arabsheybani et al. (2018), hu and yu (2015), cheraghalipour and farsad (2018), adali and isik (2017), awasthi et al. (2018), sinha and anand (2018). resource consumption nonbenefit resource consumption in production process luthra et al. (2017), song et al. (2017). financial stability benefit financial position and stability garg and sharma (2020), chen and hung (2010), ulutas et al. (2016), sinha and anand (2018), cheraghalipour and farsad (2018). e nv ir on m en ta l p er fo rm an ce c ri te ri a green product benefit product requiring less physical resources and low environmental impacts garg and sharma (2018), awasthi et al. (2018), gören (2018), sinha and anand (2018), song et al. (2017), arabsheybani et al. (2018), luthra et al. (2017), cheraghalipour and farsad (2018). green manufacturing process benefit manufacturing process that minimise waste, pollution, and energy use. garg and sharma (2018), awasthi et al. (2018), sinha and anand (2018), gören (2018), luthra et al. (2017), arabsheybani et al. (2018), song et al. (2017), cheraghalipour and farsad (2018). green r & d benefit environmental sustainability in research and development garg and sharma (2018), awasthi et al. (2018), luthra et al. (2017). sinha and anand (2018), cheraghalipour and farsad (2018), arabsheybani et al. (2018). environmental management system (ems) benefit environmental planning, implementation, monitoring and controlling garg and sharma (2018), chen and hung (2010), gören (2018), sinha and anand (2018), arabsheybani et al. (2018), song et al. (2017), luthra et al. (2017), awasthi et al. (2018), cheraghalipour and farsad (2018). so ci al p er fo rm an ce c ri te ri a worker’s wages and welfare benefit workers’ wages and welfare at supplier’s firm garg and sharma (2018), liou et al. (2012), arabsheybani et al. (2018), luthra et al. (2017), song et al. (2017), cheraghalipour and farsad (2018). worker’s occupational health and safety benefit workers’ occupational health and safety at suppliers’ firm garg and sharma (2018), luthra et al. (2017). cheraghalipour and farsad (2018), gören (2018), sinha and anand (2018), arabsheybani et al. (2018), song et al. (2017). social welfare and community development benefit social welfare and community development garg and sharma (2020), luthra et al. (2017), song et al. (2017), cheraghalipour and farsad (2018), sinha and anand (2018), vasiljević et al. (2018) (table 2 continues from the previous page) int. j. prod. manag. eng. (2022) 10(2), 143-158 creative commons attribution-noncommercial-noderivatives 4.0 international akhtar 150 http://creativecommons.org/licenses/by-nc-nd/4.0/ based on literature review study, a proposed framework of manufacturing outsourcing partner selection based on five dimensions such as operational, agile, economic, environmental and social sustainability is shown in figure 2. the proposed model takes into consideration operational/ technical, agile as well as triple bottom line sustainability aspects. table 3. techniques for sustainable and agile outsourcing partner selection in the literature. author (s) methodology and techniques adopted issues addressed chen and hung (2010) fuzzy ahp + fuzzy topsis selection of outsourcing manufacturing partners lin et al. (2010) anp outsourcing vendor selection in taiwanese semiconductor company. hsu et al. (2013) dematel + anp + gra outsourcing provider selection festel et al. (2014) action research selection of strategic outsourcing partner in a global pharmaceutical manufacturing company akhavan et al. (2015) fuzzy quantitative strategic planning matrix (fqspm) swot and f-aras, f-copras, f-topsis, f-moora. strategic outsourcing partners evaluation in car manufacturer company in iran beikkhakhian et al. (2015) ism + fuzzy ahp + fuzzy topsis ism technique for agile supplier selection criteria evaluation and fuzzy ahp and topsis for suppliers ranking. lee et al. (2015) fuzzy ahp + fuzzy topsis selection of agile supplier, assessing business impacts and comparison of business cost under skewed order and even order strategy. hu and yu (2016) voting method +goal programming (gp) electronic contract manufacturer selection matawale et al. (2016) fuzzy multi-level (fml) mcdm approach. supplier selection in agile supply chain rezaeisaray et al. (2016) dematel + dea + fuzzy anp outsourcing supplier selection in pipe and fittings manufacturing company adali and isik (2017) critic + maut selection of agile contract manufacturer wu et al. (2017) delphi method + anp + dematel supply chain agility under uncertainty to achieve competitive advantage ji et al. (2018) neutrosophic linguistic sets based mabac + electre method outsourcing provider selection dubey et al. (2018) research based view (rba) agility, adaptability, and alignment in supply chain create sustainable competitive advantage percin (2019) fuzzy swara + fuzzy axiomatic design method selection of outsourcing provider in turkish chemical manufacturing company chen et al. (2019) capability index and manufacturing time performance index-based model outsourcing partner selection liaw et al. (2020) dematel + critic + classifiable topsis evaluate and classify green manufacturing outsourcing providers in taiwanese multinational machine tool manufacturing company garg and sharma (2020) bwm +vikor sustainable outsourcing partner selection in electronic firm haoues et al. (2021) genetic algorithm (ga) techniques to minimize total cost for inhouse and outsourced manufacturing maintenance. goker (2021) intuitionistic fuzzy cognitive map + copras method selection of agile outsourcing provider selection in turkish white goods industry. singh and sarkar (2021) integrated ahp + vikor sustainable contract manufacturer selection in automotive industry. int. j. prod. manag. eng. (2022) 10(2), 143-158creative commons attribution-noncommercial-noderivatives 4.0 international sustainable and agile manufacturing outsourcing partner selection: a literature review 151 http://creativecommons.org/licenses/by-nc-nd/4.0/ l in e t a l. (2 01 0) x x c he n an d h un g (2 01 0) x x x l io u et a l. (2 01 2) x x x x h su e t a l. (2 01 3) x x x fe st el e t a l. (2 01 4) x si va ku m ar e t a l. (2 01 5) x x h u an d y u (2 01 5) x x a kh av an e t a l. (2 01 5) x x x x x r ez ae is ar ay e t a l. (2 01 6) x x x x m om en i a nd v an dc ha li (2 01 7) x x a da li an d is ik (2 01 7) ji e t a l. (2 01 8) x x x so ng (2 01 9) x b üy ük öz ka n an d g öç er (2 01 9) x x x c he n et a l. (2 01 9) x pe rc in (2 01 9) x x x l ia w e t a l. (2 02 0) x x x g ar g an d sh ar m a (2 02 0) x x y az da ni e t a l. (2 02 1) x x x a kh ta r e t a l. (2 02 1) x x g ok er (2 02 1) x x x m et ho ds /t ec hn iq ue s ahp topsis anp copras vikor gra aras moora dea electre cocoso bwm critic swara mabac cognitive map dematel fmea taguchi loss function goal programming fuzzy neutrosophic linguistic sets pythagorean fuzzy sets fuzzy preference programming fuzzy axiomatic design triangular fuzzy hesitant sets stochastic fuzzy intuitionistic fuzzy action research voting method evidential reasoning capability index and manufacturing time performance index cfa ann & ga table 4. techniques used for outsourcing partners selection in the literature. int. j. prod. manag. eng. (2022) 10(2), 143-158 creative commons attribution-noncommercial-noderivatives 4.0 international akhtar 152 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. conclusion and scope for future work the selection of a suitable manufacturing outsourcing partner is strategic decision that is complex and uncertain due to involvement of multiple qualitative and quantitative criteria, alternatives and decision makers. it also involves imprecision and ambiguity in ratings by a group of decision makers. in this paper, an attempt has been made to review the literature on manufacturing outsourcing decisions from 2010-2021. various criteria have been used but there is increasing trend to use agile as well as environmental sustainability criteria. majority have used integrated fuzzy mcdm method, which shows that importance of integrated model with fuzzy logic. however, there is no trend on usage of any particular mcdm method. the proposed framework integrates the crucial dimensions of agility and triple bottom line sustainability in manufacturing outsourcing partner evaluation that would contribute to firm’s agility, economics, sustainability and competitiveness. the model also covers the impreciseness in decision makers rating by using fuzzy logic. the criteria listed in table 2 may be used in the proposed framework for future studies for mps. few criteria may be added or substituted depending upon industry and firms requirements. future research may use fuzzy critic or fuzzy swara for criteria weight and fuzzy topsis or fuzzy vikor method for alternative selection. to understand the causal relationship among criteria, fuzzy dematel or fuzzy danp method are good suggestion. future study should apply a suitable technique, compare the result with other techniques and also carry out sensitivity analysis to improve the accuracy and robustness of the framework. this study will contribute to better understanding of manufacturing outsourcing problem and scope for future studies. the study is also helpful to managers to understand different dimensions of assessment that will improve firm’s agility, sustainability and competitiveness. this study may not be exhaustive. the future studies may cover more databases and latest papers to get a broader picture of manufacturing outsourcing provider evaluation and selection. references abdel-basset, m., & mohamed, r. (2020). a novel plithogenic topsis-critic model for sustainable supply chain risk management. journal of cleaner production, 247, 119586. https://doi.org/10.1016/j.jclepro.2019.119586 adalı, e.a., & işık, a.t. (2017). critic and maut methods for the contract manufacturer selection problem. european journal of multidisciplinary studies, 2(5), 93-101. https://doi.org/10.26417/ejms.v5i1.p93-101 figure 2. proposed framework of manufacturing outsourcing partner selection. manufacturing outsourcing partner selection agile criteria social criteria operational/technical criteria environmental criteria economic criteria integrated fuzzy multi criteria decision making model int. j. prod. manag. eng. (2022) 10(2), 143-158creative commons attribution-noncommercial-noderivatives 4.0 international sustainable and agile manufacturing outsourcing partner selection: a literature review 153 https://doi.org/10.1016/j.jclepro.2019.119586 https://doi.org/10.26417/ejms.v5i1.p93-101 http://creativecommons.org/licenses/by-nc-nd/4.0/ akhavan, p., barak, s., maghsoudlou, h., & antuchevičienė, j. (2015). fqspm-swot for strategic alliance planning and partner selection; case study in a holding car manufacturer company. technological and economic development of economy, 21(2), 165-185. https://doi.org/10.3846/20294913.2014.965240 akhtar, m., & ahmad, m.t. (2021). a stochastic fuzzy multi-criteria group decision-making for sustainable vendor selection in indian petroleum refining sector. benchmarking: an international journal, 29(3), 963-996. https://doi. org/10.1108/bij-09-2020-0500 alfred, a.m., & adam, r.f. (2009). green management matters regardless. academy of management perspectives, 23(3), 17-26. https://doi.org/10.5465/amp.2009.43479261 alimardani, m., rabbani, m., & rafiei, h. (2014). a novel hybrid model based on dematel, anp and topsis for supplier selection in agile supply chains. international journal of services and operations management, 18(2), 179-211. https://doi.org/10.1504/ijsom.2014.062000 al-shboul, m.d.a. (2017). infrastructure framework and manufacturing supply chain agility: the role of delivery dependability and time to market. supply chain management: an international journal, 22(2), 172-185. https://doi. org/10.1108/scm-09-2016-0335 arabsheybani, a., paydar, m.m., & safaei, a.s. (2018). an integrated fuzzy moora method and fmea technique for sustainable supplier selection considering quantity discounts and supplier's risk. journal of cleaner production, 190, 577-591. https://doi.org/10.1016/j.jclepro.2018.04.167 awasthi, a., & kannan, g. (2016). green supplier development program selection using ngt and vikor under fuzzy environment.computers & industrial engineering, 91, 100-108. https://doi.org/10.1016/j.cie.2015.11.011 awasthi, a., govindan, k., & gold, s. (2018). multitier sustainable global supplier selection using a fuzzy ahpvikor based approach, international journal of production economics, 195, 106-117. https://doi.org/10.1016/j. ijpe.2017.10.013 bai, c.a., & sarkis, j. (2014). determining and applying sustainable supplier key performance indicators. supply chain management: an international journal, 19(3), 275-291. https://doi.org/10.1108/scm-12-2013-0441 bhattacharya, a., & singh, p.j. (2019). antecedents of agency problems in service outsourcing. international journal of production research, 57(13), 4194-4210. https://doi.org/10.1080/00207543.2018.1506179 barhmi, a. (2019). agility and responsiveness capabilities: impact on supply chain performance, european scientific journal, 15(7), 212-224. https://doi.org/10.19044/esj.2019.v15n7p212 beikkhakhian, y., javanmardi, m., karbasian, m., & khayambashi, b. (2015). the application of ism model in evaluating agile suppliers' selection criteria and ranking suppliers using fuzzy topsis-ahp methods. expert systems with applications, 42(15-16), 6224-6236. https://doi.org/10.1016/j.eswa.2015.02.035 blome, c., schoenherr, t., & rexhausen, d. (2013). antecedents and enablers of supply chain agility and its effect on performance: a dynamic capabilities perspective, international journal of production research, 51(4), 1295-1318. https://doi.org/10.1080/00207543.2012.728011 büyüközkan g., & göçer, f. (2019). a novel approach integrating ahp and copras under pythagorean fuzzy sets for digital supply chain partner selection. ieee transactions on engineering management, 68(5), 1486-1503. https:// doi.org/10.1109/tem.2019.2907673 chan, a.t., ngai, e.w., & moon, k.k. (2017). the effects of strategic and manufacturing flexibilities and supply chain agility on firm performance in the fashion industry, european journal of operational research, 259(2), 486-499. https://doi.org/10.1016/j.ejor.2016.11.006 chen, k.s., chang, t.c., & lin, y.t. (2019). developing an outsourcing partner selection model for process with two-sided specification using capability index and manufacturing time performance index. international journal of reliability, quality and safety engineering, 26(3), 1950015. https://doi.org/10.1142/s0218539319500153 chen, l.h., & hung, c.c. (2010). an integrated fuzzy approach for the selection of outsourcing manufacturing partners in pharmaceutical r&d, international journal of production research, 48(24), 7483-7506. https://doi. org/10.1080/00207540903365308 cheraghalipour, a., & farsad, s. (2018). a bi-objective sustainable supplier selection and order allocation considering quantity discounts under disruption risks: a case study in plastic industry, computers and industrial engineering, 118, 237-250. https://doi.org/10.1016/j.cie.2018.02.041 collin, j., & lorenzin, d. (2006). plan for supply chain agility at nokia: lessons from the mobile infrastructure industry, international journal of physical distribution and logistics management, 36(6), 418-430. https://doi. org/10.1108/09600030610677375 christopher, m. (2000). the agile supply chain competing in volatile markets, industrial marketing management, 29(1), 37-44. https://doi.org/10.1016/s0019-8501(99)00110-8 int. j. prod. manag. eng. (2022) 10(2), 143-158 creative commons attribution-noncommercial-noderivatives 4.0 international akhtar 154 https://doi.org/10.1108/bij-09-2020-0500 https://doi.org/10.1108/bij-09-2020-0500 https://doi.org/10.5465/amp.2009.43479261 https://doi.org/10.1504/ijsom.2014.062000 https://doi.org/10.1108/scm-09-2016-0335 https://doi.org/10.1108/scm-09-2016-0335 https://doi.org/10.1016/j.jclepro.2018.04.167 https://doi.org/10.1016/j.cie.2015.11.011 https://doi.org/10.1016/j.ijpe.2017.10.013 https://doi.org/10.1016/j.ijpe.2017.10.013 https://doi.org/10.1108/scm-12-2013-0441 https://doi.org/10.1080/00207543.2018.1506179 https://doi.org/10.19044/esj.2019.v15n7p212 https://doi.org/10.1016/j.eswa.2015.02.035 https://doi.org/10.1080/00207543.2012.728011 https://doi.org/10.1109/tem.2019.2907673 https://doi.org/10.1109/tem.2019.2907673 https://doi.org/10.1016/j.ejor.2016.11.006 https://doi.org/10.1142/s0218539319500153 https://doi.org/10.1080/00207540903365308 https://doi.org/10.1080/00207540903365308 https://doi.org/10.1016/j.cie.2018.02.041 https://doi.org/10.1108/09600030610677375 https://doi.org/10.1108/09600030610677375 https://doi.org/10.1016/s0019-8501(99)00110-8 http://creativecommons.org/licenses/by-nc-nd/4.0/ denyer, d., & tranfield, d. (2009). producing a systematic review, in buchanan, d and bryman, a. (eds.), the sage handbook of organizational research methods, sage publications, london, 671-689. dos santos, b.m., godoy, l.p., & campos, l.m. (2019). performance evaluation of green suppliers using entropytopsis-f. journal of cleaner production, 207, 498-509. https://doi.org/10.1016/j.jclepro.2018.09.235 dubey, r., altay, n., gunasekaran, a., blome, c., papadopoulos, t., & childe, s.j. (2018). supply chain agility, adaptability and alignment: empirical evidence from the indian auto components industry, international journal of operations and production management, 38(1), 129-148. https://doi.org/10.1108/ijopm-04-2016-0173 dobos, i., vörösmarty, g. (2014). green supplier selection and evaluation using dea-type composite indicators. international journal of production economics, 157, 273-278. https://doi.org/10.1016/j.ijpe.2014.09.026 ebrahimipour, v., shoja, b.m., & li, s. (2016). supplier selection considering product structure and product life cycle cost, international journal of quality & reliability management, 33(5), 654-675. https://doi.org/10.1108/ ijqrm-05-2015-0069 fallahpour, a., udoncy olugu, e., nurmaya musa, s., yew wong, k., & noori, s. (2017). a decision support model for sustainable supplier selection in sustainable supply chain management, computers and industrial engineering, 105, 391-410. https://doi.org/10.1016/j.cie.2017.01.005 fan, z.p., chen, z., & zhao, x. (2020). battery outsourcing decision and product choice strategy of an electric vehicle manufacturer. international transactions in operational research, 29(3), 1943-1969. https://doi.org/10.1111/itor.12814 fei, l., deng, y. & hu, y. (2019). ds-vikor: a new multi-criteria decision-making method for supplier selection, international journal of fuzzy systems, 21(1), 157-175. https://doi.org/10.1007/s40815-018-0543-y festel, g., de nardo, m., & simmen, t. (2014). outsourcing of pharmaceutical manufacturing-a strategic partner selection process. journal of business chemistry, 11(3), 117-132. garg, c.p., sharma, a. (2020). sustainable outsourcing partner selection and evaluation using an integrated bwmvikor framework. environment, development and sustainability, 22(2), 1529-1557. https://doi.org/10.1007/ s10668-018-0261-5 ghorabaee, m.k., amiri, m., zavadskas, e.k., & antuchevičienė, j. (2017). assessment of third-party logistics providers using a critic-waspas approach with interval type-2 fuzzy sets, transport, 32(1), 66-78. https://doi.org/10.3846/ 16484142.2017.1282381 goker, n. (2021). a novel integrated intuitionistic fuzzy decision aid for agile outsourcing provider selection: a covid-19 pandemic-based scenario analysis. soft computing, 25(24), 13723-13740. https://doi.org/10.1007/s00500-021-06037-0 gören, h.g. (2018). a decision framework for sustainable supplier selection and order allocation with lost sales, journal of cleaner production, 183, 1156-1169. https://doi.org/10.1016/j.jclepro.2018.02.211 govindan, k., khodaverdi, r., & jafarian, a. (2013). a fuzzy multi criteria approach for measuring sustainability performance of a supplier based on triple bottom line approach. journal of cleaner production, 47, 345-354. https://doi.org/10.1016/j.jclepro.2012.04.014 grimm, j.h., hofstetter, j.s., & sarkis, j. (2014). critical factors for sub-supplier management: a sustainable food supply chains perspective. international journal of production economics, 152, 159-173. https://doi.org/10.1016/j. ijpe.2013.12.011 gualandris, j., golini, r., & kalchschmidt, m. (2014). do supply management and global sourcing matter for firm sustainability performance? supply chain management: an international journal, 19(3), 258-274. https://doi. org/10.1108/scm-11-2013-0430 haoues, m., dahane, m., & mouss, k.n. (2021). capacity planning with outsourcing opportunities under reliability and maintenance constraints, international journal of industrial and system engineering, 37(3), 382-409. https://doi. org/10.1504/ijise.2021.113470 hasan, m.a., shankar, r., & sarkis, j. (2008). supplier selection in an agile manufacturing environment using data envelopment analysis and analytical network process, international journal logistics systems and management, 4 (5), 523-550. https://doi.org/10.1504/ijlsm.2008.017599 helo, p., hao, y., toshev, r., & boldosova, v. (2021). cloud manufacturing ecosystem analysis and design, robotics and computer-integrated manufacturing, 67, 102050. https://doi.org/10.1016/j.rcim.2020.102050 hervani, a., helms, m., sarkis, j. (2005). performance measurement for green supply chain management. benchmarking: international journal, 12(4), 330-353. https://doi.org/10.1108/14635770510609015 ho, w., xu, x., & de, p.k. (2010). multi criteria decision making approaches for supplier evaluation and selection: a literature review, european journal of operational research, 202(1), 16-24. https://doi.org/10.1016/j.ejor.2009.05.009 hsu, c.c., james, j.h., & chuang, y.c. (2013). integrating danp and modified grey relation theory for the selection of an outsourcing provider. expert system with applications, 40(6), 2297-2304. https://doi.org/10.1016/j.eswa.2012.10.040 int. j. prod. manag. eng. (2022) 10(2), 143-158creative commons attribution-noncommercial-noderivatives 4.0 international sustainable and agile manufacturing outsourcing partner selection: a literature review 155 https://doi.org/10.1016/j.jclepro.2018.09.235 https://doi.org/10.1108/ijopm-04-2016-0173 https://doi.org/10.1016/j.ijpe.2014.09.026 https://doi.org/10.1108/ijqrm-05-2015-0069 https://doi.org/10.1108/ijqrm-05-2015-0069 https://doi.org/10.1016/j.cie.2017.01.005 https://doi.org/10.1111/itor.12814 https://doi.org/10.1007/s40815-018-0543-y https://doi.org/10.1007/s10668-018-0261-5 https://doi.org/10.1007/s10668-018-0261-5 https://doi.org/10.3846/16484142.2017.1282381 https://doi.org/10.3846/16484142.2017.1282381 https://doi.org/10.1007/s00500-021-06037-0 https://doi.org/10.1016/j.jclepro.2018.02.211 https://doi.org/10.1016/j.jclepro.2012.04.014 https://doi.org/10.1016/j.ijpe.2013.12.011 https://doi.org/10.1016/j.ijpe.2013.12.011 https://doi.org/10.1108/scm-11-2013-0430 https://doi.org/10.1108/scm-11-2013-0430 https://doi.org/10.1504/ijise.2021.113470 https://doi.org/10.1504/ijise.2021.113470 https://doi.org/10.1504/ijlsm.2008.017599 https://doi.org/10.1016/j.rcim.2020.102050 https://doi.org/10.1108/14635770510609015 https://doi.org/10.1016/j.ejor.2009.05.009 https://doi.org/10.1016/j.eswa.2012.10.040 http://creativecommons.org/licenses/by-nc-nd/4.0/ hu, k-j., & yu, v.f. (2015). an integrated approach for the electronic contract manufacturer selection problem. omega, 62, 68-81. https://doi.org/10.1016/j.omega.2015.08.010 ji, p., zhang, h.y., & wang, j.q. (2018). selecting an outsourcing provider based on the combined mabac-electre method using single-valued neutrosophic linguistic sets. computers & industrial engineering, 120, 429-441. https://doi.org/10.1016/j.cie.2018.05.012 kar, a.k. (2015). reinvestigating vendor selection criteria in the iron and steel industry. international journal of procurement management, 8(5), 570-586. https://doi.org/10.1504/ijpm.2015.070902 kim, m., & chai, s. (2017). the impact of supplier innovativeness, information sharing and strategic sourcing on improving supply chain agility: global supply chain perspective, international journal of production economics, 187, 42-52. https://doi.org/10.1016/j.ijpe.2017.02.007 lahiri, s., karna, a., kalubandi, s.c., & edacherian, s. (2022). performance implications of outsourcing: a meta-analysis, journal of business research, 139, 1303-1316. https://doi.org/10.1016/j.jbusres.2021.10.061 lee, j., cho, h., & kim, y.s. (2015). assessing business impacts of agility criterion and order allocation strategy in multi-criteria supplier selection. expert systems with applications, 42(3), 1136-1148. https://doi.org/10.1016/j. eswa.2014.08.041 liaw, c.f., hsu, w.c. j., & lo, h.w. (2020). a hybrid mcdm model to evaluate and classify outsourcing providers in manufacturing. symmetry, 12(12), 1962. https://doi.org/10.3390/sym12121962 lin, y.t., lin, c.l., yu, h.c. & tzeng, g.h. (2010). a novel hybrid mcdm approach for outsourcing vendor selection: a case study for a semiconductor company in taiwan. expert systems with applications, 37(7), 4796-4804. https://doi. org/10.1016/j.eswa.2009.12.036 liou, j.j.h. (2012). developing an integrated model for the selection of strategic alliance partners in the airline industry, knowledge-based systems, 28, 59-67. https://doi.org/10.1016/j.knosys.2011.11.019 luo, x., wu, c., rosenberg, d., & barnes, d. (2009). supplier selection in agile supply chains: an information-processing model and an illustration. journal of purchasing and supply management, 15(4), 249-262. https://doi.org/10.1016/j. pursup.2009.05.004 luthra, s., govindan, k., kannan, d., mangla, s.k., & garg, c.p. (2017). an integrated framework for sustainable supplier selection and evaluation in supply chains, journal of cleaner production, 140, 1686-1698. https://doi.org/10.1016/j. jclepro.2016.09.078 malviya, r.k., kant, r., & gupta, a.d. (2018). evaluation and selection of sustainable strategy for green supply chain management implementation. business strategy and the environment, 27(4), 475-502. https://doi.org/10.1002/ bse.2016 matawale, c.r., datta, s., & mahapatra, s.s. (2016). supplier selection in agile supply chain: application potential of fmlmcdm approach in comparison with fuzzy-topsis and fuzzy-moora. benchmarking: an international journal, 23(7), 2027-2060. https://doi.org/10.1108/bij-07-2015-0067 mohammed, a., setchi, r., filip, m., harris, i., & li, x. (2018). an integrated methodology for a sustainable two-stage supplier selection and order allocation problem, journal of cleaner production, 192, 99-114. https://doi.org/10.1016/j. jclepro.2018.04.131 momeni, m., & vandchali, h.r. (2017). providing a structured methodology for supplier selection and evaluation for strategic outsourcing. international journal of business performance and supply chain modelling, 9(1), 66-85. https://doi.org/10.1504/ijbpscm.2017.083888 mummalaneni, v., dubas, k., & chao, c. (1996). chinese purchasing managers' preferences and trade-offs in supplier selection and performance evaluation, industrial marketing management, 25(2), 115-124. https://doi.org/10.1016/0019-8501(95)00070-4 nair, a., guldiken, o., fainshmidt, s., & pezeshkan, a. (2015). innovation in india: a review of past research and future directions. asia pacific journal of management, 32(4), 925-958. https://doi.org/10.1007/s10490-015-9442-z ngai, e.w.t., chau, d.c.k., & chan, t.l.a. (2011). information technology, operational, and management competencies for supply chain agility: findings from case studies, journal of strategic information systems, 20(3), 232-249. https://doi.org/10.1016/j.jsis.2010.11.002 önüt, s., kara, s.s., & işık, e. (2009). long term supplier selection using a combined fuzzy mcdm approach: a case study for a telecommunication company, expert systems with applications, 36(2), 3887-3895. https://doi.org/10.1016/j.eswa.2008.02.045 ortiz-barrios, m.a., kucukaltan, b., carvajal-tinoco, d., neira-rodado, d. & jiménez, g. (2017). strategic hybrid approach for selecting suppliers of high-density polyethylene. journal of multi-criteria decision analysis, 24(5-6), 296-316. https://doi.org/10.1002/mcda.1617 int. j. prod. manag. eng. (2022) 10(2), 143-158 creative commons attribution-noncommercial-noderivatives 4.0 international akhtar 156 https://doi.org/10.1016/j.omega.2015.08.010 https://doi.org/10.1016/j.cie.2018.05.012 https://doi.org/10.1504/ijpm.2015.070902 https://doi.org/10.1016/j.ijpe.2017.02.007 https://doi.org/10.1016/j.jbusres.2021.10.061 https://doi.org/10.1016/j.eswa.2014.08.041 https://doi.org/10.1016/j.eswa.2014.08.041 https://doi.org/10.3390/sym12121962 https://doi.org/10.1016/j.eswa.2009.12.036 https://doi.org/10.1016/j.eswa.2009.12.036 https://doi.org/10.1016/j.knosys.2011.11.019 https://doi.org/10.1016/j.pursup.2009.05.004 https://doi.org/10.1016/j.pursup.2009.05.004 https://doi.org/10.1016/j.jclepro.2016.09.078 https://doi.org/10.1016/j.jclepro.2016.09.078 https://doi.org/10.1002/bse.2016 https://doi.org/10.1002/bse.2016 https://doi.org/10.1108/bij-07-2015-0067 https://doi.org/10.1016/j.jclepro.2018.04.131 https://doi.org/10.1016/j.jclepro.2018.04.131 https://doi.org/10.1504/ijbpscm.2017.083888 https://doi.org/10.1016/0019-8501(95)00070-4 https://doi.org/10.1007/s10490-015-9442-z https://doi.org/10.1016/j.jsis.2010.11.002 https://doi.org/10.1016/j.eswa.2008.02.045 https://doi.org/10.1002/mcda.1617 http://creativecommons.org/licenses/by-nc-nd/4.0/ parthiban, p., zubar, h.a., & katakar, p. (2013). vendor selection problem: a multi-criteria approach based on strategic decisions. international journal of production research, 51(5), 1535-1548. https://doi.org/10.1080/00207543.2012. 709644 paydar, m.m., arabsheybani, a., & safaei, a.s. (2017). a new approach for sustainable supplier selection. international journal of industrial engineering & production research, 28(1), 47-59. percin, s. (2019). an integrated fuzzy swara and fuzzy ad approach for outsourcing provider selection. journal of manufacturing technology management, 30(2), 531-552 https://doi.org/10.1108/jmtm-08-2018-0247 rabbani, m., foroozesh, n., mousavi, s.m., & farrokhi-asl, h. (2019). sustainable supplier selection by a new decision model based on interval-valued fuzzy sets and possibilistic statistical reference point systems under uncertainty. international journal of systems science: operations & logistics, 6(2), 162-178. https://doi.org/10.1080/23302674. 2017.1376232 rezaeisaray, m., ebrahimnejad, s., & khalili-damghani, k. (2016). a novel hybrid mcdm approach for outsourcing supplier selection: a case study in pipe and fittings manufacturing, journal of modelling in management, 11(2), 536-559. https://doi.org/10.1108/jm2-06-2014-0045 rostamzadeh, r., keshavarz ghorabaee, m., govindan, k., esmaeili, a., & khajeh nobar, h.b. (2018). evaluation of sustainable supply chain risk management using an integrated fuzzy topsiscritic approach, journal of cleaner production, 175, 651-669. https://doi.org/10.1016/j.jclepro.2017.12.071 sarkis, j. (2005). performance measurement for green supply chain management, benchmarking: an international journal, 12, 330-353. https://doi.org/10.1108/14635770510609015 sen, d.k., datta, s., & mahapatra, s.s. (2018). sustainable supplier selection in intuitionistic fuzzy environment: a decision-making perspective. benchmarking: an international journal, 25(2), 545-574. https://doi.org/10.1108/bij11-2016-0172 seuring, s. (2013). a review of modeling approaches for sustainable supply chain management. decision support systems, 54(4), 1513-1520. https://doi.org/10.1016/j.dss.2012.05.053 singh, p.k. & sarkar, p. a (2021). multi-criteria decision approach to select contract manufacturer for sustainable development of automotive products: an integrated framework. process integration and optimization for sustainability, 5, 843-857. https://doi.org/10.1007/s41660-021-00181-8 sinha, a.k., & anand, a. (2018). development of sustainable supplier selection index for new product development using multi criteria decision making, journal of cleaner production, 197, 1587-1596. https://doi.org/10.1016/j. jclepro.2018.06.234 sivakumar, r., kannan, d., & murugesan, p. (2015). green vendor evaluation and selection using ahp and taguchi loss functions in production outsourcing in mining industry. resources policy, 46, 64-75. https://doi.org/10.1016/j. resourpol.2014.03.008 song, w., xu, z., & liu, h.c. (2017). developing sustainable supplier selection criteria for solar air conditioner manufacturer: an integrated approach, renewable and sustainable energy reviews, 79, 1461-1471. https://doi.org/10.1016/j.rser.2017.05.081 song, c.h. (2019). deriving and assessing strategic priorities for outsourcing partner selection in pharmaceutical r&d: an approach using analytic hierarchy process (ahp) based on 34 experts' responses from korean pharmaceutical industry, journal of pharmaceutical innovation, 14, 66-75. https://doi.org/10.1007/s12247-018-9336-y teerasoponpong, s., & sopadang, a. (2022). decision support system for adaptive sourcing and inventory management in smalland medium-sized enterprises, robotics and computer-integrated manufacturing, 73, 102226 https://doi. org/10.1016/j.rcim.2021.102226 trapp, a.c., & sarkis, j. (2016). identifying robust portfolios of suppliers: a sustainability selection and development perspective. journal of cleaner production, 112, 2088-2100. https://doi.org/10.1016/j.jclepro.2014.09.062 tse, y.k., zhang, m., akhtar, p., & macbryde, j. (2016). embracing supply chain agility: an investigation in the electronics industry, supply chain management an international journal, 21(1), 140-156. https://doi.org/10.1108/scm06-2015-0237 ulutas, a., shukla, n., kiridena, s., & gibson, p. (2016). a utility-driven approach to supplier evaluation and selection: empirical validation of an integrated solution framework, international journal of production research, 54(5), 1554-1567. https://doi.org/10.1080/00207543.2015.1098787 um, j. (2017). improving supply chain flexibility and agility through variety management, international journal of logistics management, 28(2), 464-487. https://doi.org/10.1108/ijlm-07-2015-0113 vasiljević, m., fazlollahtabar, h., stević, ž., & vesković, s. (2018). a rough multicriteria approach for evaluation of the supplier criteria in automotive industry. decision making: applications in management and engineering, 1(1), 82-96. https://doi.org/10.31181/dmame180182v int. j. prod. manag. eng. (2022) 10(2), 143-158creative commons attribution-noncommercial-noderivatives 4.0 international sustainable and agile manufacturing outsourcing partner selection: a literature review 157 https://doi.org/10.1080/00207543.2012.709644 https://doi.org/10.1080/00207543.2012.709644 https://doi.org/10.1108/jmtm-08-2018-0247 https://doi.org/10.1080/23302674.2017.1376232 https://doi.org/10.1080/23302674.2017.1376232 https://doi.org/10.1108/jm2-06-2014-0045 https://doi.org/10.1016/j.jclepro.2017.12.071 https://doi.org/10.1108/14635770510609015 https://doi.org/10.1108/bij-11-2016-0172 https://doi.org/10.1108/bij-11-2016-0172 https://doi.org/10.1016/j.dss.2012.05.053 https://doi.org/10.1007/s41660-021-00181-8 https://doi.org/10.1016/j.jclepro.2018.06.234 https://doi.org/10.1016/j.jclepro.2018.06.234 https://doi.org/10.1016/j.resourpol.2014.03.008 https://doi.org/10.1016/j.resourpol.2014.03.008 https://doi.org/10.1016/j.rser.2017.05.081 https://doi.org/10.1007/s12247-018-9336-y https://doi.org/10.1016/j.rcim.2021.102226 https://doi.org/10.1016/j.rcim.2021.102226 https://doi.org/10.1016/j.jclepro.2014.09.062 https://doi.org/10.1108/scm-06-2015-0237 https://doi.org/10.1108/scm-06-2015-0237 https://doi.org/10.1080/00207543.2015.1098787 https://doi.org/10.1108/ijlm-07-2015-0113 https://doi.org/10.31181/dmame180182v http://creativecommons.org/licenses/by-nc-nd/4.0/ wang, c.n., yang, c.y., & cheng, h.c. (2019). a fuzzy multicriteria decision-making (mcdm) model for sustainable supplier evaluation and selection based on triple bottom line approaches in the garment industry. processes, 7(7), 400. https://doi.org/10.3390/pr7070400 wang, j.j., & yang, d.l. (2007). using a hybrid multi-criteria decision aid method for information system outsourcing, computers & operations research, 34, 3691-3700. https://doi.org/10.1016/j.cor.2006.01.017 wu, c., & barnes, d. (2011). a literature review of decision-making models and approaches for partner selection in agile supply chains, journal of purchasing & supply management, 17(4), 256-274. https://doi.org/10.1016/j. pursup.2011.09.002 wu, k.j., tseng, m.l., chiu, a.s., & lim, m.k. (2017). achieving competitive advantage through supply chain agility under uncertainty: a novel multi-criteria decision-making structure, international journal of production economics, 190, 96-107. https://doi.org/10.1016/j.ijpe.2016.08.027 yang, d.h., kim, s., nam, c., & min, j.w. (2007). developing a decision model for business process outsourcing. computers & operations research, 34(12), 3769-3778. https://doi.org/10.1016/j.cor.2006.01.012 yazdani, m., mohammed, a., bai, c. & labib, a. (2021). a novel hesitant-fuzzy-based group decision approach for outsourcing risk, expert systems with applications, 184, 115517. https://doi.org/10.1016/j.eswa.2021.115517 yusuf, y.y., gunasekaran, a., musa, a., dauda, m., el-berishy, n.m., & cang, s. (2014). a relational study of supply chain agility, competitiveness and business performance in the oil and gas industry, international journal of production economics, 147, 531-543. https://doi.org/10.1016/j.ijpe.2012.10.009 zarbakhshnia, n., & jaghdani, t.j. (2018). sustainable supplier evaluation and selection with a novel two-stage dea model in the presence of uncontrollable inputs and undesirable outputs: a plastic case study, the international journal of advanced manufacturing technology, 97(5), 2933-2945. https://doi.org/10.1007/s00170-018-2138-z zhou, x., pedrycz, w., kuang, y., & zhang, z. (2016). type-2 fuzzy multi-objective dea model: an application to sustainable supplier evaluation, applied soft computing, 46, 424-440. https://doi.org/10.1016/j.asoc.2016.04.038 zulkiffli, n.a., & padlee, f. (2021). sustainable outsourcing decisions, competitive capabilities and business performance of malaysian manufacturing smes: a confirmatory factor analysis approach. journal of sustainability science and management, 16(1), 158-173. https://doi.org/10.46754/jssm.2021.01.014 int. j. prod. manag. eng. (2022) 10(2), 143-158 creative commons attribution-noncommercial-noderivatives 4.0 international akhtar 158 https://doi.org/10.3390/pr7070400 https://doi.org/10.1016/j.cor.2006.01.017 https://doi.org/10.1016/j.pursup.2011.09.002 https://doi.org/10.1016/j.pursup.2011.09.002 https://doi.org/10.1016/j.ijpe.2016.08.027 https://doi.org/10.1016/j.cor.2006.01.012 https://doi.org/10.1016/j.eswa.2021.115517 https://doi.org/10.1016/j.ijpe.2012.10.009 https://doi.org/10.1007/s00170-018-2138-z https://doi.org/10.1016/j.asoc.2016.04.038 https://doi.org/10.46754/jssm.2021.01.014 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2021.15058 received: 2021-09-02 accepted: 2021-07-16 resolving the productivity paradox of digitalised production dold, l. a*, speck, c.b a middlesex university, the burroughs hendon, london nw4 4bt, united kingdom. b kalaidos fachhochschule, departement wirtschaft jungholzstrasse 43, 8050 zürich, switzerland. a luciandold@aol.com abstract: although industry 4.0 and other initiatives predict widespread adoption of digitalised technology on the factory floor, few companies use new digitalised production technology holistically in their ecosystems; in practical implementation, companies often decide against digitalisation for financial reasons. this is due to a paradox (akin to the so called “productivity paradox”) caused by the complexity of value creation and value delivery within digitalised production. this article analyses and synthesises cross-disciplinary research using a grounded theory model, thus offering valuable insights for businesses considering investing in digitalised production. a qualitative model and an associated toolbox (complete with tools for practical application by business leaders and decision-makers) are presented to address organisational uncertainty and leadership disconnect that often contribute to the paradoxical gap between digital strategy and operational implementation. key words: digitalised production, digital transformation, industry 4.0, industrie 4.0, value creation, value capture, manufacturing strategy. 1. introduction digitalisation is expected to deliver wide ranging value to almost all areas of society and the business world. the german initiative “industrie 4.0” (kagermann et al., 2011) is the most prominent example of international initiatives to drive investments towards a “fourth industrial revolution” that will utilise cyber-physical systems (cps), cyber-physical production systems (cpps), horizontal integration of the value chain, and vertical integration of production systems (kagermann et al., 2013, p. 39) to unlock new added value for industrial production. high economical value is expected from applied digitalised technology. for example, additional revenue of 110 billion € and productivity gains of up to 18 % were predicted by pwc in 2014 (geissbauer et al., 2014, p. 10f). however, lerch et al. (2017, p. 6) underlines that only 15 % of german manufacturing companies show advanced “industry 4.0 readiness,” while over 50 % continue to rely on conventional – not digitalised production processes. actatech, the german national academy of science and engineering, therefore, assumes the importance to follow through the process to mature in competences related to digital capabilities in their “industrie 4.0 maturity index” (schuh et al., 2020). further a production planning focussed taxonomy of design principles proposed by cañas et al. (2021) becomes essential to get the concept of industry 4.0 better defined. this discrepancy between predicted benefit and real implementation is rooted in the lack of corporate experience with respect to purposeful and successful industry 4.0 implementation (veile et al., 2019, p. 2), as well in uncertainty caused by the complexity of a new technological landscape (magruk 2016, p. 278) and novel digital dimensions (fleisch et al., 2014, p. 816). digitalisation, however, is not pursued to cite this article: dold, l., speck, c. (2021). resolving the productivity paradox of digitalised production. international journal of production management and engineering, 9(2), 65-80. https://doi.org/10.4995/ijpme.2021.15058 int. j. prod. manag. eng. (2021) 9(2), 65-80creative commons attribution-noncommercial-noderivatives 4.0 international 65 https://orcid.org/0000-0002-6544-859x http://creativecommons.org/licenses/by-nc-nd/4.0/ for the sake of digitalisation. digitalisation offers competitive advantage by delivering new products and cost benefits (porter and heppelmann 2014). new business models for data-based value creation play an essential role in providing these benefits (arnold et al., 2017, p. 371) and challenge traditional business processes, as producing industries are more conservative than internet-based businesses. to encourage a wider scale of investment in new technology for production lines, a transition or extension of these companies’ understanding of value proposition, value network, and value architecture has to be at centre stage. the transformation of the corporation’s strategic necessity to go digital is creating an increasing “digital gap” from traditional operational reality. from technology perspective many companies appear well prepared but fall short in necessary structural adjustments (gürdür et al., 2019). this results in a paradoxical situation where companies must evaluate the possibility of digitalised production while profitability criteria remain unclear (kiel et al., 2017, p. 25). al-debei and avison (2010, p. 371) show that this “digital gap” can be understood and managed by applying business models that hold the notion of value at centre stage. this approach is in line with chesbrough and rosenbloom (2002, p. 532), who understand business models as “mediating construct[s] between technology and economic value”. industrial productivity is strongly related to measures for improving overall equipment efficiency (oee) (koch 2016, p. 49). the potential for flexibility and connectivity to improve productivity mandates that companies consider standardising communication within and between production equipment (sauer 2014, p. 295). opc-ua is identified as key enabler for the industrial internet of things and the related factory of the future (palm et al., 2014). further, in most industrialised countries, initiatives to drive digitalisation of manufacturing offer alliances and funding programs (kagermann et al., 2016). the digital gap is influenced by all these different dimensions related to digitalisation and production. to develop a practical model for investments in digitalised production, it is mandatory to create a holistic theory that covers all related disciplines of research. scholars typically work within their research domain and identify cross themes as new area for research. investment towards digitalised production often falls within these cross-disciplinary areas as it combines four scientific disciplines to understand the relations. the wide-ranging dependencies amongst these research disciplines forms the research question, how necessary investments in digitalization can be assessed in terms of their contribution to value generation within a manufacturing value chain. the scientific aim of the study is therefore to research and surface the underlaying structural constructs and interrelations determine the digital transformation processes of industrial production infrastructure. these results contribute to a holistic view on the novel research field of digital transformation of manufacturing companies. in practical terms, the empirical findings are furnished into an innovative toolbox supporting necessary assessment and validation that business enterprises undergo within their digital transformation processes. an intensive review of the research literature from these four disciplines offers a rich data source for understanding questions related to digitalised production. a grounded theory can be established through a qualitative study of the phenomena identified in the literature data (strauss and corbin 2010, p. 54; charmaz 2014, p. 45ff) combined with complementary expert interviews. the final coded grounded theory is elaborated upon to deliver a toolbox for gathering and processing conditions to then determine investment patterns. the paper presents the process for creating a suitable toolbox for production companies to manage paradox of digital investment. the first section presents an analysis of the relevant scientific disciplines to create a holistic model for the field of digitalised production. the second section explains the theoretical model of investment behaviour that bridges the gap of digitalised production. in the third section the toolbox is developed, based on the finding of the grounded theoretical model. finally, conclusions and recommendation for further research are outlined. 2. interrelation of research disciplines as is often the case, existing research around digitalised production is focused and narrow, following the specific interests of individual disciplines. for example, publications related to digitisation initiatives—like industry 4.0, ivi, or iic (kagermann et al., 2011; industrial value chain initiative 2018; lin et al., 2017)—are concept-oriented and focus on technology needs. technology-centric publications, on the other hand, take advantage of new development trends such as int. j. prod. manag. eng. (2021) 9(2), 65-80 creative commons attribution-noncommercial-noderivatives 4.0 international dold and speck 66 http://creativecommons.org/licenses/by-nc-nd/4.0/ opc-ua, mqtt, big data, or artificial intelligence, as the core of their research. business model research considers business issues with external partners, i.e., b2b, b2c and the aspects related to the “office floor.” in the same fashion, the work resulting from research related to production is dedicated to investigative tools for continuous productivity increases on the “shop floor.” since digitalised production touches all these disciplines, there is value in combining their individual results to generate a holistic view to guide company investment perspectives. figure 1 illustrates the four research disciplines that relate to investments towards digitalised production. figure 1. four research disciplines related to digitalised production (author’s illustration). currently, there is a lack of studies bridging these related disciplines, although the demand is often recognised and identified as area of further research (obermaier and schweikl 2019, p. 558). in most existing publications, cross-discipline aspects are mentioned superficially or remain fuzzy. to leverage our collective knowledge to inform investment decisions, a holistic model that incorporates all relevant insights from the scholars of these four disciplines is required. a search of the literature available on ebscohost, emerald and google scholar delivered 15,450 matches with keywords related to the domain of digitalised production within these four disciplines. table 1 contains the most relevant keywords and references. after deep review of the resulting keywords, abstracts, and full studies, 121 sources in total were identified as relevant to deliver data for a qualitative analysis, further described in section 4. table 1. leading keywords and references (author’s illustration). keywords and boolean conditions references (excerpt) automation and data arnold and voight 2017; sauer 2014 business model osterwalder 2004; al-debei and avison 2010; zott et al., 2011 digital transformation berghaus 2018; schuh et al., 2020; obermaier 2019 digitalization rachinger et al., 2019; bouwman et al., 2018 industry 4.0 or industrie 4.0 schuh et al., 2015; magruk 2016; kagermann et al., 2013; 2016 2016kagermann et al., 2013 internet of things or iot jesse 2016; imtiaz and jasperneite 2013; zuehlke 2010 overall equipment efficiency or oee ylipää et al., 2017; gibbons and burgess 2010 productivity and manufacturing hopp and spearman 2004; schmenner 2015 standardization cottyn et al., 2008; dorst 2016; eruvankai et al., 2017 value proposition osterwalder and pigneur 2010; rese et al., 2013 willingness to invest grebe et al., 2019; skilton et al., 2010 3. the paradox of digitalised production throughout history, technological innovation has often caused paradoxical situations. in the 1980’s, for example, business and society at large entered the era of computerisation. robert solow (1987) observed then that “you [could] see the computer age everywhere but in the productivity statistics.” this was solow’s paradox, also known as the “productivity paradox;” that in spite of rapid technological growth in every sector, productivity was down. since the solow paradox, we have observed increasing values of computer usage that additional benefits being generated with latency, over time. obermaier and schweikl (2019, p. 540ff) researched the relevance of the solow paradox in relation to germany’s industry 4.0 initiative. they demonstrated that several phenomena of the “fourth industrial revolution” appear to follow same patterns as observed in the “computer revolution.” the paradox of digitalised production is grounded in the ultimate need of a business to evaluate potential technology int. j. prod. manag. eng. (2021) 9(2), 65-80creative commons attribution-noncommercial-noderivatives 4.0 international resolving the productivity paradox of digitalised production 67 http://creativecommons.org/licenses/by-nc-nd/4.0/ investments, a process that many companies are illprepared for (porter 2010, 150,250ff). traditional investments can be evaluated by weighing the value generated by improvements within the horizontal value stream (porter 2010, p. 226). however, within the sphere of digital business, traditional judgment methods are challenged by complexity and uncertainty. figure 2 visualises how the digital gap emerges from insufficient knowledge of digital technology combined with a higher level of risks and dynamics of the digital business environment. figure 2. comparison between the world of traditional and digital business (source: al-debei and avison, 2010, p. 369). this gap between process and strategy related to digital production equipment originates from the interaction between horizontal integration (dorst 2016, p. 19) of the value stream and the related demand for vertical integration (dorst 2016, p. 28) of data and information. in order to optimise the value stream of a produced item, information acquired from devices in the production process— like machine cells, fixtures, motors, robots, vision systems, a single sensor, or a power supply—must be aggregated and analysed. consequently, to realise vertical integration, a heavy investment in hardware and infrastructure is borne on the factory floor. the monetizable value, however, is captured from the usage within the horizontal integration of the value chain. digital transformation requires either an explicitly defined digital business model or the implicit expectation that the company will benefit from digitalised technology. the ultimate need to ensure that an investment delivers a sufficient return and competitive advantage relates to the underlaying assumptions of a digital business model. as a consequence of organisational allocation of funds to install the value architecture, the assumed value proposition, and the mechanism to capture value, an effect of decoupling can be observed. figure 3 visualises this phenomenon of decoupled costs and value and illustrates the underlying effects that create the digital gap related to investments into digitalised production. the increasing complexity of digital transformation mandates new processes and criteria to evaluate return on investment. the decoupling of cost and value capture forms a gap between digital strategy and operational implementation in digitalised production. this gap hinders or biases traditional business decisions, especially in regard to digitaldriven investments. al-debei and avison propose to overcome such a gap using business models as a conceptual tool for “alignment” between corporate strategy and business processes (al-debei and avison, 2010, p. 371). figure 4 shows how business models intersect with digital business strategy and digital business processes. a business model can be understood as a “mediating construct between technology and economic value” (chesbrough and rosenbloom figure 3. instances of an industry 4.0 transformation and their relations (source: dold, 2020, p. 18). int. j. prod. manag. eng. (2021) 9(2), 65-80 creative commons attribution-noncommercial-noderivatives 4.0 international dold and speck 68 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2002, p. 532). thus, business models and their value-oriented approach offer a path for reducing the identified digital gap and for managing the paradoxical conundrum facing decisions about investing into digitalised production. 4. grounded theory the literature produced by scholars within the disciplines, technology, standardisation, productivity, and digitalisation initiatives, is a rich source of data for analysing the phenomena related to digitalised production, though the data cannot be easily compared as it is rooted in such disparate disciplines. to make use of such heterogenous sources within a grounded theory analysis model, charmaz (2014, p. 45) explicitly recognises the value of data that is drawn from all aspects of the literature; phenomena can be identified not only in the content of the research studies, but also in their target audiences, the backgrounds of the authors, and in the presentation of the text. drawing on this observation, a paradigmatic coding process was used to create a grounded theory model based on the identified literature sources. the applied paradigm was based on the work of strauss and corbin (2010) and delivers 76 phenomena from the process of open coding (strauss and corbin, 2010, p. 54). to understand the relationships between and the meaning of the observed phenomena, ten high-level categories have been defined to structure the coding results (table 2). table 2. results of the open coding process (source: dold 2020 p.78). discipline category (open coding) phenomena business model utilization of business models no. 1 to 7 layers of strategic decisions no. 8 to 14 impact of business model no. 15 to 20 business commercial value determinants of investment strategy and transformation no. 21 to 26 no. 27 to 33 no. 34 to 42 productivity maturity levels no. 43 to 49 lean production no. 50 to 57 technology technological economical relation no. 58 to 66 data-integration no. 67 to 76 digitalised production is a new research domain; hence the analysed literature derives from surrounding disciplines only and may not reflect the latest status of the practical application of digitalisation. to cover this possible bias, the research design integrates qualitative expert interviews (mayring 2008; jüttemann 1989; witzel 2000; döring and bortz 2016) to ensure applicable expert data is available to develop the theoretical model. the phenomena outlined in table 2 were evaluated by four experienced industry experts to define the path diagram visualised in figure 5. the path diagram utilises the causal and intervening conditions identified via axial coding (strauss and corbin, 2010, p. 75), thus fully incorporating data from literature and qualitative interviews. it shows that decision making is influenced by multiple phenomena centred around “maturity of company resources” and “digital penetration.” further, the prominent representation of value elements originating from business models confirms the moderation of value. figure 4. business model intersection points (source: al-debei and avison, 2010, p. 370). int. j. prod. manag. eng. (2021) 9(2), 65-80creative commons attribution-noncommercial-noderivatives 4.0 international resolving the productivity paradox of digitalised production 69 http://creativecommons.org/licenses/by-nc-nd/4.0/ the process of selective coding (strauss and corbin 2010, p. 94) integrates the identified phenomena and conditions to form the core category of the final grounded theory model. a single central phenomenon cannot directly be identified from the available results. three phenomena demonstrate similar dominance. further validation with the source data show that these three phenomena are interrelated in their influence over investment decisions. together they form a latent phenomenon that is the core category of the theoretical model. the core category is composed of phenomena related to “risk,” “people,” and “value,” and is given the name “associated balance” (see figure 6). this core category enables evaluation of the conditions, context, actions, interactions, and strategies that consequently determine investment patterns in digitalised production. the process of paradigmatic transition (strauss and corbin 2010, p. 101) is outlined in figure 7 and incorporates all available data and analysis results from the preceding steps. the characteristics of the core category are defined by the conditions of “risk awareness,” “consideration of persons,” “networks,” and “value proposition.” core category: associated balance people, risk, and value all influence investment decisions, and cross-influence each other. their interactions form a core construct: the "associated balance. “ balance is achieved when negative or positive influences from any one of these three factors are counteracted by influence from the others. figure 6. core category “associated balance” (source: dold 2020, p. 127). maturity of company resources (47) 14: value network (23) 18: situation of sme vs. large organization (29) 16: competencies of organization (48) 1: invest decision taking ( 27) 12: maturity level of digitalization (45) 18: situation of sme vs. large organization (29) digital penetration (6) 6: level of digital culture in organization (14) 3: vvlue creation (24) 16: competencies of organization (48) 9: data as production factor (32) 1: invest decision taking ( 27) 4: monetising of data( 64) 2: impact on profitability ( 17) interview 1-2: elements of value generation (16) interview 1-3: value proposition/value offer (21) interview 1-4: risk of digitalization (61) 8: value architecture (22) 4: monetising of data( 64) 6: level of digital culture in organization (14) 3: value creation (24) interview 1-4: risk of digitalization (61) interview 1-5: target to use technology (66) 7: process of digital change (36) 5: cost/invest of the digital enterprise(58) 9: data as production factor (32) 13: latency between invest and harvest (38) 5: : cost/invest of the digital enterprise(58) 13: latency between invest and harvest (38) figure 5. path model of analysed relations amongst phenomena (source: dold 2020, p. 101). int. j. prod. manag. eng. (2021) 9(2), 65-80 creative commons attribution-noncommercial-noderivatives 4.0 international dold and speck 70 http://creativecommons.org/licenses/by-nc-nd/4.0/ risk awareness is based on the uncertainties and risks connected with digitisation. across companies and individual decision makers, understanding and considerations of related risks vary (magruk 2016, p. 284). the deeper the understanding of risks, the clearer the risk level can be assessed and taken into account within a balanced investment evaluation. consideration of persons is the acknowledgment that people are responsible for successful implementation of and value generation through digital transformation. involving people is indispensable (burggräf et al., 2017, 2463). considerations of the range of the people involved, as well in their hierarchical and department roles, contribute to determining the balance and influence of this factor. networks consider that value is created by a coordination of activities. this effect requires widereaching networking between functions and partners, both internally and externally (al-debei and avison, 2010, p. 367). the more completely these networks are established, the more effectively risks and values can be balanced. value proposition for desired additional values from digitisation can manifest differently. a value proposition defines the benefit that is to be achieved or that has been achieved (al-debei and avison, 2010, p. 365). the strength of the additional values constitutes an opposite pole to compensate evaluated risks. thus, determined from the four conditions of risk awareness, consideration of persons, networks, and value proposition, the “associated balance” reflects a balanced understanding of the total situation, providing a basis from which to take actions, define strategies, and implement further interaction. the phenomenon of the associated balance (ab) is distinct in four states. if the interrelations between persons, risk and value are well balanced, a secured ab is considered. in case some, when discrepancies within the determining factors remain but the total effect is positive, an optimistic ab exists. a critical ab exists if, despite some positive influence, majority negative parameters predict instability. in the case when all conditions are shaped negative, a disrupted ab emerges. the status of the ab works in conjunction with the specific context of the project being evaluated, which encompasses “digital business model,” “digital culture,” “data usage as production factor,” and the “competitive situation” as conditions. each company’s digital business model utilises digitisation differently, with variations ranging from the use of simple digital tools for a conventional business model, to a holistic digitisation of the value chain (burmeister et al., 2016, p. 146). presence of a digital business model supports a well-developed transition towards action and decision for digitisation in production. digital culture addresses the possibility that significant digitisation may not be realised due to weak anchoring in the corporate culture (schuh et al., 2020, p. 11). readiness for change within the workforce, and the necessary communication of social aspects, are fundamental prerequisites for figure 7. paradigmatic transition (source: dold 2020, p. 129). int. j. prod. manag. eng. (2021) 9(2), 65-80creative commons attribution-noncommercial-noderivatives 4.0 international resolving the productivity paradox of digitalised production 71 http://creativecommons.org/licenses/by-nc-nd/4.0/ developing and implementing digital strategies. insufficient willingness within a company’s personnel counteracts a positive effect from ab. data usage as production factor covers the potential for additional value capture or monetization via data collection in production (maier and weber 2013, 17,38). if the contribution of data to capture value is not considered, an inhibiting effect on the formation of strategies is considered (legrenzi, 2017, p. 36). if the potential is recognised and included in the assessment of value, the implementation of digitised production technology will be supported (tantik and anderl, 2016). actions towards digitisation are influenced by a company’s specific competitive situation. small and medium enterprises usually have fewer available funds or resources to implement comprehensive measures (leyh and bley 2016; andulkar et al., 2018). this deficiency has a limiting influence on the effects of the status of the ab. the aggregate composed of the status of the ab and the contextual conditions impacts a variety of factors to establish an overarching attitude towards digital investments. these factors include leadership initiatives around digital transformation and developing the competence of people and organisations; the strategic management of latency of investments as well digital standardisation strategies; and the manifestation of value capture. value capture incorporates the creation of value by digitisation as is amplified through means of networked activities (arnold and voight, 2017, p. 100). increase in efficiency and or the use of synergies (zott et al., 2011, p. 1029) are simplified by networked communication to coordinate and capture value creation. the more the ab is balanced, the more a value-oriented decision on investments is enabled. there is a time lag between the realisation of value and execution of the investments necessitating that a latency of investment to be recognised during the decisions-making process. the strategy for dealing with this latency (maklan et al., 2015, p. 583) is formed by the ab and the given context as well as the understanding that production works as an overall process (see industrial value chain initiative 2018, p. 18). standardisation aims to cope with the increasing complexity of digitalised production through new and specific standards (henssen and schleipen 2014, p. 302). combined with an overall positive context and ab, strategic orientation towards full standardisation can be expected (buchholz et al., 2017, p. 32). with digital transformation leadership, the need to understand the process of digital transformation is counted in. the leadership style is determined by the level of digital maturity (berghaus et al., 2017, p. 22) which is reflected in observable behaviours. the development of people and organisational competences is necessary to expand value and to bridge uncertainties around adopting new technology (magruk, 2016, p. 279). in order to achieve economy of scale and synergy, broad digital skills are required throughout an organisation (remane et al., 2017, p. 7). figure 8. grounded theory model after paradigmatic transition (source: dold, 2020, p. 136). int. j. prod. manag. eng. (2021) 9(2), 65-80 creative commons attribution-noncommercial-noderivatives 4.0 international dold and speck 72 http://creativecommons.org/licenses/by-nc-nd/4.0/ these impact items underpin a behaviour toward investment decisions as they represent the measures taken to create value via innovation of technology (chesbrough, 2010, p. 359). the question of what investment is the most attractive (zennaro et al., 2018, p. 7) consequently follows from an attitude that ranges from being reluctant to being proactively entrepreneurial. the paradigmatic transition builds a grounded theory model as visualised in figure 8 incorporating all considerations and phenomena outlined in the qualitative analysis. this model enables structured evaluation of individual projects and enterprises with regards to planning and execution of investments in digitalised production. 5. the digital investment toolbox the transformation of conventional production processes towards a digitalised production paradigm promises to unlock further value generation and gains in competitiveness. however, many businesses struggle to start or scale up from their first light tower projects. this difficulty is due to the multiplicity of investment projects and technical innovations on the one hand, and uncertainty and a lack of best practices on the other. businesses would benefit from guidance and tools to facilitate solidified decisions on investment priorities, and to map the transformation process. in constructing the grounded theory model presented in the preceding sections, extensive analysis of the rich data and insights gleaned from the existing body of literature, plus expert interviews, was conducted. this analysis also enables the creation of a practical toolbox. the formative model, as outlined in figure 8, delivers relevant elements to assess the condition of a business in regard to digitalised investment conditions, context, action, strategy, interaction, and consequent investment patterns. this assessment enables a validation of the desired “to be” condition from companies’ strategic considerations and identifies areas to change from “as is.”. based on the model and data insights, a toolbox to assess and validate the determinants of digitalised production can be developed. this digitalinvestment-toolbox (dit) enables enterprises to analyse and navigate their digital transformation process by efficiently managing the aforementioned digital investment paradox. technically the logical connections are coded into excel documents and the relations, queries and evaluations are realized by vba-scripts. the dit (figure 9) offers four tools based on the paradigmatic elements developed within the selective coding process. the “associated balance check” (abc) is the first tool, and retrieves the observed status of the exogenous characteristics to determine the state of the ab. the second tool subsumes the contextual conditions and assigns a specific context pattern. this tool has been given the name “context compass” (cc). the output of the abc-tool and the cc-tool are preconditions for the third tool to provide guidance about related impacts. this “impact guide” (ig) provides a set of “to be” conditions based on the observed conditions and context. these “to be” conditions are compared with the “as is” situation within the organisation to visualise the potential gap within the business. the fourth tool, the “investment validator” (iv), given the gap identified with the ig-tool, delivers recommendations for corrective actions, interaction, and strategy to realise the desired transition to digital investments. figure 9. the digital-investment-toolbox (source: dold 2021, p. 92). 5.1. associate balance check (abc) – tool 1 the abc aims to determine the aggregated status of the relevant exogenous variables that form investment decisions for digitalised production. the variables consist of two-dimensional characteristics for each of the ab conditions identified within the qualitative model (risk level, networking, consideration of people, and value delivery). a numeric value can be assigned by taking into account the effect of both dimensional characteristics. the values are normalised within a range of +1 to -1 and correspond to a value related to the underlaying results from the grounded theory process. to calculate the ab value, eight characteristics have to be qualitatively evaluated and ranked as high or low, minor or complete, and isolated or holistic. for ease of use, four-field diagrams depicting the int. j. prod. manag. eng. (2021) 9(2), 65-80creative commons attribution-noncommercial-noderivatives 4.0 international resolving the productivity paradox of digitalised production 73 http://creativecommons.org/licenses/by-nc-nd/4.0/ corresponding value of each variable are determined, and finally the average of all four values is taken to define the state of the individual ab. figure 10 shows the abc calculation card to be utilised in workshops and practical investment evaluation. the abc-tool delivers an early prediction for business deciders regarding the solidity of the underlaying variables as they relate to successful digitalised investment. further, this easy tool enables identification of areas to address if the ab result turn outs to be lower than 0.5 (critical or disrupted). 5.2. context compass (cc) – tool 2 the specific context of the business to be digitalised has a significant impact on the actions, strategies and interactions to be established and considered. the context related to digitalised production is composed out of five variables; competitive pressure, resources, digital culture and digital business model, as outlined in section 4. the dimensional characteristics of these variables are more sophisticated than in the ab, and the constitution of the individual context is not based on a deterministic balance as is used to calculate the ab; there are a total of forty-eight possible combinations of the variables. however, based on the theoretical results, the number of different scenarios can be reduced to eighteen contexts. a headline has been defined for each identified context to describe the expected qualitative influence on digitalised investments. these headlines range from “missing substance” to “high risk in digital transformation,” and from “digital laggard” to “digital specialist” and “digital leader.” the cc-tool aims to underline the implications that may result from a business’s specific context, in which the intended investments are situated. table 3 summarises the composed contexts in a simplified format. the cc-tool offers a full selection table as well a detailed explanation of each context, outlining the theoretically based implications and opportunities. the cc-tool is uniquely valuable in its ability to offer evaluation of external factors which are rarely considered in traditional investment evaluation. the influence of the contextual elements on leverage of digitalisation can be highly supportive (e.g., contexts 7, 12, 16.17 and 18) or obstructing (e.g., contexts 1, 3, 8 and 13). once this contextual influence is understood, the cc also provides reflective insights to help identify blind spots and areas of improvement for the investment project. 5.3. impact guide (ig) – tool 3 while the abc and cc tools aim to assess exogenous elements and evaluate influences and possible blind spots within the decision-making process, the ig takes the related impacts into consideration and reveals potential gaps that may hinder the digitalised investment project. the first part of this tool surfaces risk level (rl) networking degree (nd) high . low external organized high . high minor . internal consideration of people (cp) value delivery (vd) partial holistic partial strong low . low low . low secured ab > 0,75 optimistic ab >0,5 0,74 critical ab > 0 0.49 disrupted ab ≤ 0 associated balance (ab)ris k kn ow le dg e ri s k i mpact knowl edge low hi gh lo w h ig h ex te rn al n et w or k internal network low high lo w h ig h in vo lv em en t of p eo pl e invol vement over hi erarchi es mi nor compl ete m in or co m pl et e va lu e st re ng th val ue i mpact is ol ated hol i s ti c lo w hi gh -1 -1 -1 +1 +1+0,5 0-1 +10 -1 -1 +10 -1 -1 𝐴𝐵 = 𝑅𝐿 + 𝑁𝐷 + 𝐶𝑃 + 𝑉𝐷 4 secured ab > 0,75 optimistic ab >0,5 0,74 critical ab > 0 0.49 disrupted ab ≤ 0 associated balance (ab)ris k kn ow le dg e ri s k i mpact knowl edge low hi gh lo w h ig h ex te rn al n et w or k internal network low high lo w h ig h in vo lv em en t of p eo pl e invol vement over hi erarchi es mi nor compl ete m in or co m pl et e va lu e st re ng th val ue i mpact is ol ated hol i s ti c lo w hi gh -1 -1 -1 +1 +1+0,5 0-1 +10 -1 -1 +10 -1 -1 𝐴𝐵 = 𝑅𝐿 + 𝑁𝐷 + 𝐶𝑃 + 𝑉𝐷 4 figure 10. abc-tool calculation card (author’s illustration). int. j. prod. manag. eng. (2021) 9(2), 65-80 creative commons attribution-noncommercial-noderivatives 4.0 international dold and speck 74 http://creativecommons.org/licenses/by-nc-nd/4.0/ the required “to be” in the areas of strategy, action, and interaction, as indicated by the results of abc and cc. the second part is a self-assessment of the “as is” situation. the final part compares the “to be” with the “as is” to visualise deficits and to prioritise corrective actions. figure 11 illustrate the structure and rational of the ig-tool. the “to be” evaluation utilises a table of impact requirements based on theory. this impact table indicates clearly applicable impacts from the ab and the business context. if the evaluation results in a “non-applicable” judgement, the underlaying theoretical conditions do not recommend proceeding with a digital investment until the ab or context improves or changes. for the “as is” status evaluation, the dimensional characteristics based on the paradigmatic transitions are surveyed, as described below. the interaction of value capture is judged as “high” or “low” based on the additional expected value that can be gathered from three streams of value. first, the intermediate value resulting from direct savings in cost or proven improvements or productivity measures. second, the mediate value delivered by improvements identified in the preand post-processes. third, the value that is expected to be delivered with time delay. in relation to the two strategical impacts, the perceived “as is” paradigm must be judged. the paradigm related to latency of investments is rated as either “conventional,” “clearly defined,” or “entrepreneurial.” similarly, the practical usage of standardisation is rated as “not defined,” “partial,” or “complete.” paradigms of leadership and competence development are the basis of the action-based impact items, in which digital transformation leadership style is categorised table 3. the cc-tool with defined context combinations and related headlines (author’s illustration). context number combined categories competi ti ve pres s ure resources data usage digital culture di gi tal bus i nes s model context headline 1 9 missing substance 2 3 utilized high m. c. neutral readiness 3 6 no conservative basis 4 3 utilized low m. c. basis for digital beginners 5 1 utilized high no basis for a digital start 6 1 utilized high partial digital fundament 7 1 utilized high holistic potential for digital leadership 8 6 no high risk in digital transformation 9 3 utilized low m. c. digital laggard 10 1 utilized high no digital set up with limited resources 11 1 utilized high partial digital extension with limited resources 12 1 utilized high holistic digital optimum with limited resources 13 3 no low m. c. neglected digital chance 14 3 no high m. c. ready for digital kick start 15 3 utilized low m. c. ready for digital leadership 16 1 utilized high no digital kick-starter 17 1 utilized high partial digital specialist 18 1 utilized high holistic digital leader multiple combinations multiple combinations low high low sufficient low sufficient multiple combinations figure 11. the impact guide tool (author’s illustration). int. j. prod. manag. eng. (2021) 9(2), 65-80creative commons attribution-noncommercial-noderivatives 4.0 international resolving the productivity paradox of digitalised production 75 http://creativecommons.org/licenses/by-nc-nd/4.0/ as “top-down,” “bottom-up,” “specialised,” or “innovative.” activities to develop people’s and the organisation’s competences are rated as “ad-hoc,” “project based,” or “holistic.” a gap analysis collates the values in a radar chart to provide the ig-tool output. the ig is therefore the central tool within the dit as it harnesses the theoretical results from grounded theory into a structural thought process that generates a practical gap analysis. the ig will clarify whether the intended digital investment considerations have theoretical stability or will identify the need for corrective measures that should be taken before wide-ranging investments in digitalisation are considered. 5.4. investment validator (iv) – tool 4 while the three former tools offer guidance and clarity for operationally involved people up to the middle management, the iv-tool targets senior leadership, who sponsor the investment projects. recall that the digital gap identified by al-debei and avison (figure 2) exists between the strategic considerations made by senior management and onthe-ground operations. the iv-tool aims to bridge the strategic agenda by identifying corrective actions to close the potential gap identified by the ig. first, the iv-tool requires senior management to reflect on investment paradigms that are necessary to best achieve the outlined digital transformation strategic targets. the tool describes the patterns of the investment paradigms, ranging from “cautious,” and “roi-oriented,” to “innovative” and “entrepreneurial,” as compiled in table 4. the impact table used within the ig-tool correlates investment patterns to both the “to be” requirements (to support the given ab-status and context), as well as to the “as is” pattern (resulting from the operational assessment). in figure 12, all three investment patterns are put into perspective and potential deviations with the strategic expectation of senior management are identified. the iv-tool delivers the fundamental understanding about whether the digital gap exists—and if so, how wide it is—as well as pointing to where the limitations of the digital paradox may restrict necessary strategic implementations to capture value via digitalising production. as the findings of the iv-tool are based on insights generated from the ig, abc and cc tools, this tool presents the overall digital awareness and capability to manage the effect of the decoupling of cost and value (presented in section 3). the iv-tool further delivers transparency on corrective measures to senior management, provides understanding about how the value architecture of the business is composed, and how to capture the value from digitalisation via a holistic view. table 4. understanding the strategic investment patterns within the iv-tool (author’s illustration) investment pattern description entrepreneurial (eip) the eip is clearly looking ahead and acts in an open, innovation-friendly manner. the paradigm is characterised by tangible entrepreneurial courage in the business. the characteristics of the eip are an innovative approach to the digital transformation, the entrepreneurial wide horizon of expectations for roi in future and the commitment to comprehensive standardisation amongst value chains and data streams. innovative (iip) an iip is characterised by the clear orientation towards technological benefits. the commitment towards a full implemented standardisation is characteristic. the prevailing management style is either bottom-up or specialised. roi orientation is considered but is handled flexible if the technological innovation promises to deliver a latent value. roi-oriented (rip) rip is characterised by its clear orientation towards a defined roi. rip-based decisions require tangible timewise assessment, whereby the latency must be considered either conventionally or with clearly defined hedging. the leadership stick to conventional judgement and investment criteria as top-down or specialised approach. the immediate direct value creation is the dominating evaluation criteria within the rip. cautious (cip) cip describes the cautious way of dealing with funds. the willingness to invest for rip, iip and eip is clearly recognisable by their specific characteristics and can be assigned as consequence of expected or proven value capture of digitalised investments. conversely, it can be concluded that in consequence, if there is dominating uncertainty to decide for the three other patterns the cautious investment pattern apply. int. j. prod. manag. eng. (2021) 9(2), 65-80 creative commons attribution-noncommercial-noderivatives 4.0 international dold and speck 76 http://creativecommons.org/licenses/by-nc-nd/4.0/ 6. conclusion and discussion similar to other areas of technological innovation, the digital transformation of production presents companies and people with new challenges. though a majority of publications are focused on the new technology that comes with industry 4.0, a digital gap between strategy and operational reality has created a paradoxical situation. this article promotes a holistic, cross-disciplinary perspective to resolve this decoupling between cost and value capture related to digitalised production. senior management, middle managers, and operational contributors need to understand the relationship between these core elements of a digitalised business. the researched grounded theory model delivers a solid understanding based on data from scholars and experts. the digital investment toolbox provides a method for applying this analysis to practical projects and is designed to offer a straight-forward assessment of the relevant parameters. this assessment enables organisations to understand where and how to tackle deficits and blind spots. the successful leveraging of digital investments will require sufficient maturity and full support from all layers. the toolbox is designed to indicate if digital maturity or leadership commitment appears insufficient. the parties involved, however, must heed these indications in order to benefit from a successful digital enterprise and in order to make the necessary paradigm shifts away from practices that have been internalised over years of traditional business. the ultimate exercise outlined in the iv-tool carries potential to open eyes and promote change towards new entrepreneurial spirit and innovation. 7. limitations and outlook towards further research the research described in this article integrates data from heterogenous scientific disciplines. since then, the impact of covid-19 is supposed to have significant impact in the described transformation processes. as the proposed toolbox is based on pre-covid data, a verification with post-covid data is recommended once societies do return to a new normal paradigm. the qualitative analysis demonstrates the value of more frequent border crossing between disciplines for further research into the topic of digitalisation. the value-centric view will help practitioners to build bridges between technology, productivity, and business aspects. other disciplines maybe be included within further research and expand the scope given by the limitations originated from selected disciplines. while the grounded theory delivers a formative model, further research is recommended to refine the relations and phenomena. the qualitative model delivers a solid base to combine a quantitative analysis for a mixed-methods-study. references al-debei, mutaz m., avison, david (2010). developing a unified framework of the business model concept. european journal of information systems, 19(3), 359–376. https://doi.org/10.1057/ejis.2010.21 andulkar, mayur; le, duc tho; berger, ulrich (2018). a multi-case study on industry 4.0 for sme’s in brandenburg, germany. proceedings of the 51st hawaii international conference on system sciences. hawaii, 2018. https://doi.org/10.24251/hicss.2018.574 figure 12. investment validator process to point towards corrective measures (auto’’s illustration). int. j. prod. manag. eng. (2021) 9(2), 65-80creative commons attribution-noncommercial-noderivatives 4.0 international resolving the productivity paradox of digitalised production 77 https://doi.org/10.1057/ejis.2010.21 https://doi.org/10.24251/hicss.2018.574 http://creativecommons.org/licenses/by-nc-nd/4.0/ arnold, christian; kiel, daniel; voight, kai-ingo (2017). innovative business models for the industrial internet of things. in berg huettenmaenn monatsh 162 (9), 371–381. https://doi.org/10.1007/s00501-017-0667-7. arnold, christian; voight, kai-ingo (2017). ecosystem effects of the industrial internet of things on manufacturing companies. acta infologica, 1(2), 99–108. https://doi.org/10.1007/s00501-017-0667-7 berghaus, sabine (2018). the fuzzy front end of digital transformation. activities and approaches for initiating organizational change strategies. universität st. gallen. available online at https://www1.unisg.ch/www/edis.nsf/syslkpbyidentifier/4704/$file/dis4704.pdf berghaus, sabine; back, andrea; kaltenrieder, bramwell. (2017). digital maturity & transformation report 2017. zürich: crosswalk ag,. in veröffentlichung zur studie der universität st. gallen in kooperation mit crosswalk. st. gallen, march 2017. bouwman, harry; nikou, shahrokh; molina-castillo, francisco j.; reuver, mark de (2018). the impact of digitalization on business models. digital policy, regulation and governance, 20(2), 105–124. https://doi.org/10.1108/dprg-07-2017-0039 buchholz, birgit; ferdinand, jan-peter; gieschen, jan-hinrich; seidel, uwe (2017). digitalisierung industrieller wertschöpfung. eine studie im rahmen der begleitforschung zum technologieprogramm autonomik für industrie 4.0 des bundesministeriums für wirtschaft und energie. berlin: iit-institut für innovation und technik der vdi/vde innovation + technik gmbh. burggräf, peter; dannapfel, matthias; voet, hanno; bök, patrick-benjamin; uelpenich, jérôme; hoppe, julian (2017). digital transformation of lean production. systematic approach for the determination of digitally pervasive value chains. in world academy of science, engineering and technology, international journal of social, behavioral, educational, economic, business and industrial engineering 11(10), 2462–2471. burmeister, christian; luettgens, dirk; piller, frank t. (2016). business model innovation for industrie 4.0. why the ‘industrial internet’ mandates a new perspective. die unternehmensführung; rwth-tim working paper, 70(2), 124–152. https://doi.org/10.2139/ ssrn.2571033 cañas, héctor; mula, josefa; díaz-madroñero, manuel; campuzano-bolarín, francisco (2021). implementing industry 4.0 principles. computers & industrial engineering, 158(1), p. 107379. https://doi.org/10.1016/j.cie.2021.107379 charmaz, kathy (2014). constructing grounded theory. 2nd edition. los angeles, london, new delhi, singapore, washington dc: sage. chesbrough, henry (2010). business model innovation: opportunities and barriers. opportunities and barriers. long range planning, 43 (2-3), 354–363. https://doi.org/10.1016/j.lrp.2009.07.010 chesbrough, henry; rosenbloom, richard s. (2002). the role of the business model in capturing value from innovation: evidence from xerox corporation’s technology spin-off companies. industrial and corporate change, 11(3), 529–555. https://doi.org/10.1093/icc/11.3.529. cottyn, johannes; stockman, kurt; hendrik, van landeghem (2008). the complementarity of lean thinking and the isa 95 standard. wbf 2008. wbf. barcelona, november 2008. available online at http://hdl.handle.net/1854/lu-524679 dold, luzian (2020). beurteilung von investitionen in die digitalisierte produktion. eine mixed-method-studie zur moderierenden wirkung von nutzenkonstrukten aus geschäftsmodellen an der lücke zwischen digitaler strategie und operativen prozessen. dissertation. middlesex university, london. dold, luzian (2021). a value centred paradigm to moderate the digital transformation of manufacturing. advanced journal of social science, 8(1), 86–95. available online at https://journals.aijr.in/index.php/ajss/article/view/3563 döring, nicola; bortz, jürgen (2016). forschungsmethoden und evaluation in den sozialund humanwissenschaften. with assistance of sandra pöschl. 5. vollständig überarbeitete, aktualisierte und erweiterte auflage. berlin, heidelberg: springer (springer-lehrbuch). https://doi.org/10.1007/978-3-642-41089-5 dorst, wolfgang (2016). implementation strategy industrie 4.0. report on the results of the industrie 4.0 platform. with assistance of w. dorst, c. glohr, t. hahn, u. loewen, rosen, r. schiemann, t., f. vollmar et al. edited by bitkom e.v., vdma e.v., zvei e.v. berlin, frankfurt am main. eruvankai, saju; muthukrishnan, murugesan; mysore, anantharamaiah kumar (2017). accelerating iiot adoption with opc ua. internetworking indonesia, 9(1), 3–8. available online at http://www.internetworkingindonesia.org/issues/vol9-no1-2017/iij_vol9_ no1_2017_eruvankai.pdf. fleisch, elgar; weinberger, markus; wortmann, ass felix; wortmann, felix (2014). geschäftsmodelle im internet der dinge. hmd praxis der wirtschaftsinformatik, 51(6), 812–826. https://doi.org/10.1365/s40702-014-0083-3. geissbauer, reinhard; schrauf stefan; koch volkmar; kuge simon (2014). industry 4.0: opportunities and challenges of the industrial internet. edited by pricewaterhousecooper aktiengesellschaft. münchen. gibbons, paul m.; burgess, stuart c. (2010). introducing oee as a measure of lean six sigma capability. lean six sigma journal, 1(2), 134–156. https://doi.org/10.1108/20401461011049511 grebe, michael; rüssmann, michael,leyh michael; franke, roman (2019). how digital champions invest. edited by boston consulting group. münchen. available online at http://image-src.bcg.com/images/bcg-how-digital-champions-invest-june-2019_tcm15223286.pdf, checked on 1/6/2020. gürdür, didem; el-khoury, jad; törngren, martin (2019). digitalizing swedish industry. what is next? computers in industry, 105(1), 153– 163. https://doi.org/10.1016/j.compind.2018.12.011 int. j. prod. manag. eng. (2021) 9(2), 65-80 creative commons attribution-noncommercial-noderivatives 4.0 international dold and speck 78 https://doi.org/10.1007/s00501-017-0667-7 https://www1.unisg.ch/www/edis.nsf/syslkpbyidentifier/4704/$file/dis4704.pdf https://doi.org/10.1108/dprg-07-2017-0039 https://doi.org/10.2139/ssrn.2571033 https://doi.org/10.2139/ssrn.2571033 https://doi.org/10.1016/j.lrp.2009.07.010 http://hdl.handle.net/1854/lu-524679 https://doi.org/10.1007/978-3-642-41089-5 http://www.internetworkingindonesia.org/issues/vol9-no1-2017/iij_vol9_no1_2017_eruvankai.pdf http://www.internetworkingindonesia.org/issues/vol9-no1-2017/iij_vol9_no1_2017_eruvankai.pdf https://doi.org/10.1108/20401461011049511 http://image-src.bcg.com/images/bcg-how-digital-champions-invest-june-2019_tcm15-223286.pdf http://image-src.bcg.com/images/bcg-how-digital-champions-invest-june-2019_tcm15-223286.pdf https://doi.org/10.1016/j.compind.2018.12.011 http://creativecommons.org/licenses/by-nc-nd/4.0/ henssen, robert; schleipen, miriam (2014). interoperability between opc ua and automationml. procedia cirp, 25, 297–304. https://doi. org/10.1016/j.procir.2014.10.042 hopp, wallace j.; spearman, mark l. (2004). to pull or not to pull. what is the question? manufacturing & service operations management, 6(2), 133–148. https://doi.org/10.1287/msom.1030.0028. imtiaz, jahanzaib; jasperneite, jürgen (2013). scalability of opc-ua down to the chip level enables “internet of things”. ieee intelligent systems, 500–505. https://doi.org/10.1109/indin.2013.6622935 industrial value chain initiative (2018). industrial value chain reference architecture -next. strategic implementation framework of industrial value chain for connected industries. edited by industrial value chain initiative. monozukuri nippon conference c/o. tokyo. jesse, norbert (2016). internet of things and big data – the disruption of the value chain and the rise of new software ecosystems. ifacpapersonline, 49(29), 275–282. https://doi.org/10.1016/j.ifacol.2016.11.079 jüttemann, gerd (ed.) (1989). qualitative forschung in der psychologie. grundfragen, verfahrensweisen, anwendungsfelder. 2. aufl. heidelberg: asanger. kagermann, henning; anderl, reiner; gausemeier, jürgen; schuh, günther; wahlster wolfgang (2016). industrie 4.0 im globalen kontext. strategien der zusammenarbeit mit internationalen partnern. acatech studie. münchen: herbert utz verlag. kagermann, henning; wahlster, wolfgang; helbig, johannes (2013). umsetzungsempfehlungen für das zukunftsprojekt industrie 4.0. abschlussbericht des arbeitskreises industrie 4.0. edited by prof. dr. henning kagermann. forschungsunion wirtschaft und wissenschaft, arbeitskreis industrie 4.0. frankfurt am main. kagermann, henning; wahlster, wolfgang; lukas, wolf-dieter (2011). industrie 4.0: mit dem internet der dinge auf dem weg zur 4. industriellen revolution. vdi nachrichten 2011, 4/1/2011 (13). kiel, daniel; müller, julian; arnold, christian; voight, kai-ingo (2017). sustainable industrial value creation. benefits and challenges of industry 4.0. international journal of innovation management, 21(8), 1–34. https://doi.org/10.1142/s1363919617400151 koch, arno (2016). oee für das produktionsteam. das vollständige oee-benutzerhandbuch oder wie sie die verborgene maschine entdecken. 3., korrigierte auflage. herrieden: cetpm publishing (operational excellence, nr. 5). legrenzi, christopher (2017). the digital paradox. information, informatics, and information system. ism journal of international business, 35–42. lerch, christian; jäger, angela; maloca, spomenka (2017). wie digital ist deutschlands industrie wirklich. arbeit und produktivität in der digitalen produktion. in mitteilungen aus der isi-erhebung modernisierung der produktion, 71. leyh, christian; bley, katja (2016). digitalisierung. chance oder risiko für den deutschen mittelstand? – eine studie ausgewählter unternehmen. hmd, 53(1), 29–41. https://doi.org/10.1365/s40702-015-0197-2 lin, shi-wan; crawford, mark; mellor, stephen (2017). the industrial internet of things volume g1: reference architecture. version 1.80. needham, ma. in industrial internet consortium (iic) tech. rep. magruk, andrzej (2016). uncertainty in the sphere of the industry 4.0 – potential areas to research. business, management & education/ verslas, vadyba ir studijos, 14(2), 275–291. https://doi.org/10.3846/bme.2016.332 maier, w.; weber, m. (2013). management von big-data-projekten. leitfaden. berlin: bundesverband informationswirtschaft,telekommuni kation und neue medien e. v. maklan, stan; peppard, joe; klaus, philipp (2015). show me the money. european journal of marketing, 49(3/4), 561–595. https://doi. org/10.1108/ejm-08-2013-0411 mayring, philipp (2008). einführung in die qualititative sozialforschung. eine anleitung zu qualitativem denken. 5. aufl. weinheim, basel: beltz (beltz studium). obermaier, robert (2019). industrie 4.0 und digitale transformation als unternehmerische gestaltungsaufgabe. in robert obermaier (ed.). handbuch industrie 4.0 und digitale transformation. betriebswirtschaftliche, technische und rechtliche herausforderungen. 1st ed. 2019. wiesbaden: springer fachmedien wiesbaden, 3–46. https://doi.org/10.1007/978-3-658-24576-4_1 obermaier, robert; schweikl, stefan (2019). zur bedeutung von solows paradoxon. empirische evidenz und ihre übertragbarkeit auf digitalisierungsinvestitionen in einer industrie 4.0. in robert obermaier (ed.). handbuch industrie 4.0 und digitale transformation. betriebswirtschaftliche, technische und rechtliche herausforderungen. 1st ed. 2019. wiesbaden: springer fachmedien wiesbaden, 529–564. https://doi.org/10.1007/978-3-658-24576-4_22 osterwalder, alexander (2004). the business model ontology: a proposition in a design science approach. dissertiation. universite de lausanne, lausanne. ecole des hautes etudes commerciales. available online at https://doc.rero.ch/record/4210/files/1_these_ osterwalder.pdf osterwalder, alexander; pigneur, yves (2010). business model generation: a handbook for visionaries, game changers, and challengers. hoboken: john wiley & sons. palm, florian; grüner, sten; pfrommer, julius; graube, markus; urbas, leon (2014). open62541–der offene opc ua stack. in onlinepublikation des fraunhofer iosb, lehrstuhl prozessleittechnik der rwth aachen; tu dresden, professur für prozessleitechnik, int. j. prod. manag. eng. (2021) 9(2), 65-80creative commons attribution-noncommercial-noderivatives 4.0 international resolving the productivity paradox of digitalised production 79 https://doi.org/10.1016/j.procir.2014.10.042 https://doi.org/10.1016/j.procir.2014.10.042 https://doi.org/10.1109/indin.2013.6622935 https://doi.org/10.1016/j.ifacol.2016.11.079 https://doi.org/10.1142/s1363919617400151 https://doi.org/10.1365/s40702-015-0197-2 https://doi.org/10.3846/bme.2016.332 https://doi.org/10.1108/ejm-08-2013-0411 https://doi.org/10.1108/ejm-08-2013-0411 https://doi.org/10.1007/978-3-658-24576-4_1 https://doi.org/10.1007/978-3-658-24576-4_22 https://doc.rero.ch/record/4210/files/1_these_osterwalder.pdf https://doc.rero.ch/record/4210/files/1_these_osterwalder.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ porter, michael e. (2010). wettbewerbsvorteile. spitzenleistungen erreichen und behaupten. 7. durchgesehene auflage auflage. frankfurt, new york: campus. porter, michael e.; heppelmann, james e. (2014). how smart, connected products are transforming competition. harvard business review, 92(11), 64–88. rachinger, michael; rauter, romana; müller, christiana; vorraber, wolfgang; schirgi, eva (2019). digitalization and its influence on business model innovation. journal of manufacturing technology management, 30(8), 1143–1160. https://doi.org/10.1108/jmtm-01-20180020 remane, gerrit; hanelt, andre; wiesboeck, florian; kolbe, lutz m. (2017). digital maturity in traditional industries an exploratory analysis. twenty-fifth european conference on information systems (ecis). guimarães, portugal, 2017. rese, mario; meier, horst; gesing, judith; boßlau, mario (2013). an ontology of business models for industrial product-service systems. in yoshiki shimomura, koji kimita (eds.). the philosopher’s stone for sustainability. proceedings of the 4th cirp international conference on industrial product-service systems, tokyo, japan, november 8th 9th, 2012. heidelberg: springer (lecture notes in production engineering), 191–196. sauer, olaf (2014). information technology for the factory of the future – state of the art and need for action. procedia cirp, 25, 293–296. https://doi.org/10.1016/j.procir.2014.10.041 schmenner, roger w. (2015). the pursuit of productivity. production and operations management, 24(2), 341–350. https://doi.org/10.1111/ poms.12230 schuh, g.; anderl, r.; dumitrescu, r.; hompel, m. ten; krüger, a. (eds.) (2020). industrie 4.0 maturity index. managing the digital transformation of companies – update 2020 – (acatech study). munich: herbert utz verlag. schuh, günther; reuter, christina; hauptvogel, annika; dölle, christian (2015). hypotheses for a theory of production in the context of industrie 4.0. advances in production technology: springer, 11–23. https://doi.org/10.1007/978-3-319-12304-2_2 skilton, mark; gordon, penelope; harding, chris (2010). building return on investment from cloud computing. cloud business artifacts project, cloud computing work group, the open group. edited by the open group. burlington, ma. solow, r. m. (1987). we’d better watch out. in new york times book review 36. strauss, anselm; corbin, juliet (2010). grounded theory. grundlagen qualitativer sozialforschung. unveränd. nachdr. der letzten aufl. weinheim: beltz. tantik, erdal; anderl, reiner (2016). industrie 4.0. using cyber-physical systems for value-stream based production evaluation. procedia cirp, 57, 207–212. https://doi.org/10.1016/j.procir.2016.11.036 veile, johannes; kiel, daniel; voight, kai-ingo; müller, julian marius (2019). lessons learned from industry 4.0 implementation in the german manufacturing industry. journal of manufacturing technology management, 31(5). https://doi.org/10.1108/jmtm-08-20180270 witzel, andreas (2000). das problemzentrierte interview. in forum qualitative sozialforschung / forum: qualitative social research 1(1). https://doi.org/10.1016/j.lrp.2015.04.001 ylipää, torbjörn; skoogh, anders; bokrantz, jon; gopalakrishnan, maheshwaran (2017). identification of maintenance improvement potential using oee assessment. international journal of productivity and performance management, 66(1), 126–143. https://doi.org/10.1108/ ijppm-01-2016-0028. zennaro, ilenia; battini, daria; sgarbossa, fabio; persona, alessandro; marchi, rosario de; van der wiele, ton (2018). micro downtime data collection, analysis and impact on oee in bottling lines the san benedetto case study. int j qual & reliability mgmt, 35(4), 965-995. https://doi.org/10.1108/ijqrm-11-2016-0202 zott, christoph; amit, raphael; massa, lorenzo (2011). the business model: recent developments and future research. journal of management, 37(4), 1019–1042. https://doi.org/10.1177/0149206311406265 zuehlke, detlef (2010). smartfactory—towards a factory-of-things. annual reviews in control, 34(1), 129–138. https://doi.org/10.1016/j. arcontrol.2010.02.008 int. j. prod. manag. eng. (2021) 9(2), 65-80 creative commons attribution-noncommercial-noderivatives 4.0 international dold and speck 80 https://doi.org/10.1108/jmtm-01-2018-0020 https://doi.org/10.1108/jmtm-01-2018-0020 https://doi.org/10.1016/j.procir.2014.10.041 https://doi.org/10.1111/poms.12230 https://doi.org/10.1111/poms.12230 https://doi.org/10.1007/978-3-319-12304-2_2 https://doi.org/10.1016/j.procir.2016.11.036 https://doi.org/10.1108/jmtm-08-2018-0270 https://doi.org/10.1108/jmtm-08-2018-0270 https://doi.org/10.1016/j.lrp.2015.04.001 https://doi.org/10.1108/ijqrm-11-2016-0202 https://doi.org/10.1177/0149206311406265 https://doi.org/10.1016/j.arcontrol.2010.02.008 https://doi.org/10.1016/j.arcontrol.2010.02.008 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.1859 received 2013-11-11 accepted 2014-06-27 application of fuzzy logic in performance management: a literature review verónica gurreaa*, juan-josé alfaro-saizb,i, raul rodríguezb,ii & maría-josé verdechob,iii a escuela de doctorado. universitat politènica de valència. camino de vera s/n. 46022 valencia. * veronicagurrea@gmail.com b research centre on production management and engineering. universitat politècnica de valència. camino de vera s/n, ed. 8b –acc. l –planta 2 (ciudad politécnica de la innovación). 46022 valencia. i jalfaro@cigip.upv.es ii raurodro@upvnet.upv.es iii mverdecho@omp.upv.es abstract: performance management has become in a key success factor for any organization. traditionally, performance management has focused uniquely in financial measures, mainly using quantitative measures, but two decades ago they were extended towards an integral view of the organization, appearing qualitative measures. this type of extended view and associated measures have a degree of uncertainty that needs to be bounded. one of the essential tools for uncertainty bounding is the fuzzy logic and, therefore,the main objective of this paper is the analysis of the literature about the application of fuzzy logic in performance measurement systems operating within uncertainty environments with the aim of categorizing, conceptualizing and classifying the works written so far. finally, three categories are defined according to the different uses of fuzzy logic within performance management concluding that the most important application of fuzzy logic that counts with a higher number of studies is uncertainty bounding. key words: fuzzy logic, uncertainty, performance measurement, performance management, supply chain. 1. introduction performance management has become in a key success factor for organizations because it favors a better understanding of enterprise, internal and external, scenarios and it offers a more realistic view, allowing to easily understand the current situation of the organization. in addition, it allows modeling future behaviors through objective establishment and actions to achieve them, always in coherence with enterprise strategy and towards competitive results. traditionally, performance measurement systems have focused uniquely in financial measures but two decades ago they were extended towards an integral view of the organization considering other factors such as product quality, effectiveness and efficiency of the organization (accomplishment of delivery times, adjusted stocks, etc.), customer satisfaction o measurement of enterprise sustainability. this new approach emerges to solve the gaps encountered in the traditional measurement systems that overlooked important elements of the organizations and the increasingly competitive environment of enterprises. precisely, competitiveness among companies generates environments where the uncertainty plays a fundamental role and emerge new concepts such as “green enterprises”, “e-business”, etc. thus, the scope of performance measurement increases as well as the information collection needed. enterprises and supply chains operate within uncertainty environments that are characterized by redundant and shortcoming information for decision-making causing trouble in the definition, measuring and follow-up of objectives and goals that allow to establish target levels for performance measurement. thus, there is a necessity to generate new performance measurement systems and approaches of increasing complexity able to deal different types of uncertainty (intra-organizational, inter-organizational) and able to manage the information collected from managers and experts that sometimes has vagueness and incompleteness and is frequently subjective. 93int. j. prod. manag. eng. (2014) 2(2), 93-100creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2014.1859 http://creativecommons.org/licenses/by-nc-nd/4.0/ attending to the literature, fuzzy logic is an important tool to manage different types of uncertainty because it uses reliable parameters that adapt easily to changes in uncertainty parameters. this is due to the fact that fuzzy logic is multivariate allowing in a practical manner to deal with problems in the way they are in reality. thus, the present paper focuses in analyzing, classifying and contextualizing the works of the literature that link fuzzy logic to performance management-measurement. the structure of this work is as follows: in the second section, the methodology followed to develop the work is described; in the third section, it is presented the essential characteristics and the contribution of the analyzed works; in the fourth section, the results and discussion is exposed, and finally, conclusions and future research lines and presented. 2. review methodology to develop the present paper, in the first place, a set of terms were selected to establish the scope to analyze. thus, firstly, two main terms were selected: fuzzy logic and performance management, as well as similar terms: performance management, performance indicators, fuzzy theory and fuzzy set theory. once the search was performed, it was seen that other terms emerge always linked to the main ones. therefore, these terms were also included in the search, limiting the scope of study. these terms were: balance scorecard f or performance management and ahp/anp and topsis for fuzzylogic. similarly, another set of terms was encountered that linked the two initial main terms aiding to finally determine the scope of the study: multi-criteriadecisionmaking (mcdm), manufacturingcompanies, supplychain, supplychain management, subjectivity, uncertainty and linguisticvariables with respect to the period to analyze, this was defined form the results obtained without any initial restriction, with the aim of accessing to to all the information of the topic of research. in table 1 shows the distribution of papers and their corresponding period. it is observed an increasing growth starting from 2008: table 1. distribution of the articles with respect to years of publication. year articles 1995 2000 2002 2003 2004 2006 2007 2008 2009 2010 2011 2012 2013 total 1 1 2 2 2 2 2 8 2 4 4 3 1 34 the databases used for the study are: emerald, ieee xplore, sciencedirect, scopus and web of knowledge. 3. basic characteristics and contribution of the works reviewed in order to have a full understanding of the relationship between fuzzy logic and performance measurementmanagement, an analysis and classification of the works encountered has been performed. the classification has been performed according to the aim of using fuzzy logic in the papers. this classification has been selected because it is the one that allows to develop a higher number of categories as well as makes evident the different applications of the fuzzy logic in the studied scope. thus, three main categories have been defined. the following sections describe their main characteristics: i. type a: uncertainty bounding (28 references) ii. type b: integration of qualitative and quantitative values (2 references) iii. type c: conversion of linguistics variables (5 references) 3.1. type a: uncertainty bounding the works within type a conform the most numerous group, with a total of 25 references, that come from mainly asia (16), followed by south america (4 references), north america (3 references) and europe (5 references). all the works within type a have the basis on achieving uncertainty bounding in performance 94 int. j. prod. manag. eng. (2014) 2(2), 93-100 creative commons attribution-noncommercial-noderivatives 4.0 international gurrea, v., alfaro-saiz, j.j., rodríguez, r., & verdecho, m.j. http://creativecommons.org/licenses/by-nc-nd/4.0/ measurement, sometimes in specific industry sector and other times, in a more general scope. thus, pilevaril et al., (2008) define a system based in the if-then rules of the fuzzy logic to evaluate supply chain agility, bounding uncertainty in characteristics such as responsibility, competence, flexibility and speed. mahnam et al., (2008), suggest a bi-objective approach to model supply chain and use fuzzy logic to bound uncertainty in product demand and in the method to determine supplier reliability. datta et al. (2011), use fuzzy logic to deal with vagueness and uncertainty in quality criteria to choose a logistic partner 3pl. behrouzi & wong (2010) propose the use of membership functions in fuzzy logic to bound uncertainty in selected variables that determine the level of implementation of lean manufacturing. chan et al. (2003) and chan & qi (2003), create an innovative method to measure supply chain performance, focusing in performance measures, where fuzzy logic is used to establish the real situation for judgment and evaluation of processes avoiding subjectivity. lu & li (2006) bound uncertainty of the time required for the processes, customer demand and supplier reliability by using fuzzy logic. cheng et al. (2007) provide performance estimations with more realistic results based on measuring vagueness on product quality by using fuzzy logic. unahabhokha et al. (2007) propose the use of fuzzy expert systems for developing predictive performance measurement systems that allow in an automatical manner predict performance of future distributions and identify, for instance, potential problems for companies. behesti & lollar (2007) suggest an alternative method of performance evaluation methods in comparison with traditional quantitative methods. they focus on estimating human resource performance and suggest a model of fuzzy logic to prevent subjectivity that sometimes occur in these type of evaluations. wei et al. (2008) create a framework to measure performance of erp’s (enterprise resourceplanning). they use a fuzzy performance index to deal with vagueness. for that purpose, the index is first translated in a simple score and then, it comes to linguistic terms. others interesting articles as the ones from campuzano et al. (2013), campuzano et al. (2010) and mula et al. (2013) who tackle the harmful bullwhip effect in supply chains using fuzzy logic for replenishment orders. tabrizi & razmi (2013) introduce fuzzy logic to control risks originated in the uncertainty in the design of the network of supply chains. amini & jochem (2011) suggest that in performance measurement it is very important the role of soft factors (such as friendship or worker competency). these soft factors should be necessarily evaluated by human judgement and therefore fuzzy logic is used to control this subjectivity. olugu & wong (2012) use fuzzy logic to measure performance of a closed-looped supply chain. serna et al. (2008) study the mathematical models that use the principles of fuzzy logic to measure supply chain performance in uncertainty environments and then, they are applied to a set of bakeries. nunes et al. (2011) use a model of fuzzy system for aiding decision-making in order to choose the better alternatives to deal with supply chain disturbances. ganga & carpinetti (2011) propose a model of performance measurement based on fuzzy logic to predict performance with respect to causal relations that exist among indicators of the scor model. muñoz et al. (2008), link financial performance and organization sustainability through a set of ratings that are determined using fuzzy logic. liao et al. (2010) propose a performance evaluation of processes based on an index for uncertainty measures. all the works have a similar structure: the problem is introduced; then, some definitions are presented (regarding fuzzy logic, performance measurement or the issue covered); develop the solution proposed; analyze the results y finally, present the extracted conclusions. in table 2, the main characteristics of the works analyzed are presented (mutual characteristics for all the types of works). 3.2. type b: integration of qualitative and quantitative values two recent works (xu et al., 2008; behrouzi et al., and 2011) point out fuzzy logic as a mechanism for the integration of both qualitative and quantitative values, whose main characteristics are shown in table 3. the integration of qualitative and quantitative variables favour the generation of more flexible methodologies for managing-measuring performance. since most of the models bring together only quantitative variables, those models that integrate both types of variables deliver a more complete vision of performance (xu et al., 2008). behrouzi et al. (2010) point out that through the integration of qualitative and quantitative factors, it is possible to achieve a final score that accounts for the result of all the metrics applied in the assessment system, employing the fuzzy logic to obtain also discrete numeric variables for the qualitative factors. 95int. j. prod. manag. eng. (2014) 2(2), 93-100creative commons attribution-noncommercial-noderivatives 4.0 international application of fuzzy logic in performance management: a literature review http://creativecommons.org/licenses/by-nc-nd/4.0/ table 2. basic characteristics of main papers of type a. paper application environment industry sector objective ammar & wright (1995) intra-organizational public sector elimination of uncertainty in public satisfaction surveys relating the score of each question with the importance of that question ammar & wright (2000) intra-organizational public sector fuzzy logic application in three different cases, where the evaluation comes from panels with integrated multiple criteria lau et al. ( 2002) supply chain toys industry supply chain management to analyze and monitor the suppliers performance based on the products quality and the delivery time amini & jochem (2011) intra-organizational --performance measuring and evaluating of service processes chan & qi (2003) supply chain --holistic performance measurement of complex supply chains chan et al. (2003) supply chain --integral performance measurement of complex supply chains, evaluating processes and the judgment actual situation wang & shu (2004) supply chain --managing the supply chain uncertainty and inventory strategies determining when there is a lack of historical data. lu & li (2006) supply chain --supply chain modelling based on the incorporation of the relationship between the customer satisfaction level and inventory investment. cheng et al. (2007) supply chain technological uncertainty dimensioning in measuring the product quality and selecting the best performing processes. unahabhokha et al. (2007) intra-organizational printing and textile creating a performance prediction system in manufacturing environments pilevaril et al. (2008) supply chain automotive determining the supply chains agility serna et al. (2008) supply chain --addressing uncertainty arising from the complex interrelationships taking place among the various supply chain levels behesti&lollar (2008) intra-organizational --developing a fuzzy model for decisions making, illustrated by the application that estimates the employees performance muñoz et al. (2008) intra-organizational --sustainability evaluation in organizations with a fuzzy focus mahnam et al. (2008) supply chain --uncertainty dimension from both the demand and the reliability of suppliers in an assembly line. wei et al. (2008) intra-organizational technological erp selection based on their performance theeranuphattana & tang (2008) supply chain --analysis of the model proposed by chan & qi (2003) to reduce their limitations arango et al. (2010) supply chain food industry supply chain uncertainty dimensioning by generating a performance indicators system behrouzi & wong (2010) intra-organizational --performance evaluation of lean manufacturing environments liao & wu (2010) intra-organizational --processes performance evaluation based on disability index for measurements with uncertainty datta et al.(2011) supply chain automotive evaluation and selection of 3pl logistics operators in fuzzy environments ganga & carpinetti (2011) supply chain --performance prediction in supply chains based on the causal relationships of the scor metrics nunes et al. ( 2011) supply chain --addressing disturbances that might arise in the supply chain environments olugu & wong (2012) supply chain automotive performance measurement of supply chain closed loop (clsc, closed-loop supply chain) tabrizi & razmi (2013) supply chain --incorporation of risk management in the supply chain networks design 96 int. j. prod. manag. eng. (2014) 2(2), 93-100 creative commons attribution-noncommercial-noderivatives 4.0 international gurrea, v., alfaro-saiz, j.j., rodríguez, r., & verdecho, m.j. http://creativecommons.org/licenses/by-nc-nd/4.0/ on the other hand, both works have as a goal to carry out suppliers selection processes according to their performance. then, behrouzi et al. (2010) build a highly flexible model, which integrates both qualitative and quantitative measures as well as it is able to join metrics measured in different units (time, €, etc), through the generation of nondimensional final scores. additionally, it fosters continuous improvement processes and it is easy to use, as it does not require a high number of metrics. besides, xu et al.(2008) develop a system whose first phase is the joint between the performance metrics and the organization’s strategy. this process facilitates the selection of the most relevant metrics. then, the weights for each criterion are worked out and both the qualitative and the quantitative measures are integrated through the application of fuzzy logic. the structure of both papers is similar from a methodological point of view, as both start with the metric/criteria selection process, they secondly develop the proposed solution and, finally, they apply the solution to a concrete example. the main limitations of both papers are by one hand that, firstly, it is still necessary the experts’ opinions in order to define the fuzzy logic rules, and by the another that it is a priori allocated more importance to some specific suppliers aspects such as cost. 3.3. type c: conversion of linguistic variables. five works compose the type c, where the conversion of linguistic variables into numerical variables is studied. the 80% of them refer to the supplier selection topic, where this type of fuzzy logic application is especially convenient because of the linguistic nature of the attributes associated to both the suppliers and the manufacturing units (jain et al. (2004)). the main characteristics of each one of these papers is presented in table 4. four of these five papers present a fuzzy system whereas the another establish a methodology where the fuzzy logic is incorporated. odhar & kumar (2004) and jain et al. (2004) incorporate to the system a genetic algorithm that develops the conjoint of basic fuzzy rules. the methodology’s structure of all the papers is similar: the evaluators express their preferences in linguistic terms; these preferences are then used as input variables for the selection process, where the selection criteria are weighted and the suppliers’ performance measured, using for this fuzzy logic. then, the fuzzy scores of each potential supplier is obtained. these fuzzy scored are then translated to a discrete value that allows comparison among the proposed suppliers. tabla 3. basic characteristics of papers of type b. paper application environment industry sector objective xu & lim (2008) supply chain --supplier selection based on the literature evaluation criteria and techniques behrouzi et al. (2010) supply chain automotive industry supplier performance measurement by evaluating lean attributes tabla 4. basic characteristics of papers of type c. paper application environment industry sector objective jain et al. (2004) supply chain --suppliers selection through an evolutionary fuzzy approach ohdar & kay (2004) supply chain metallurgical suppliers selection through an evolutionary fuzzy approach chen et al. (2005) supply chain --suppliers selection delimiting the vagueness and imprecision ordoobadi (2009) supply chain --suppliers selection incorporating fuzzy arithmetic ferreira et al. (2012) supply chain --supply chain performance evaluation by creating an larg index (lean, agile, resilient, green) 97int. j. prod. manag. eng. (2014) 2(2), 93-100creative commons attribution-noncommercial-noderivatives 4.0 international application of fuzzy logic in performance management: a literature review http://creativecommons.org/licenses/by-nc-nd/4.0/ the main conclusion of this group of papers is that the use of linguistic variables when measuring performance is highly beneficial when the performance values can not be expressed by numeric values, being the role of the fuzzy logic essential, as it allows the conversion from one type of variable to the another. 4. conclusions and future research directions through the analysis of the different selected papers, it has been proved the high variety of applications and uses that the fuzzy logic has within the performance measurement field, being these uses the base for the classification and conceptualization carried out. then, the next three types of papers have been determined regarding the application of fuzzy logic to the performance measurement field: uncertainty delimitation, integration of qualitative and quantitative values and linguistic variables conversion. then, the analysis has shown that the main application of fuzzy logic within the performance measurement field is the uncertainty delimitation through dealing with both inaccuracy and non clear characteristics from many of the information needed to carry out a good assessment. in most cases, the inaccuracy factor comes from human valuations and experts’ opinions that complete the measurement process, and it is added to the own need of dealing with the subjectivity inherent to human beings judgements. regarding the activity sectors where the application of fuzzy logic within the performance measurement area is more common, these are: automotive, industrial, technological and public. however, this is not a very precise categorization, as in most of the references it is not specified the type of sector, especially on these that cover the supply chain context. this is due to the fact that, in most occasions, it is only shown a general environment and, when examples are provided, these describe slightly the number of elements and how these are distributed, without presenting concrete examples. in this area, 60% of the papers are referred to the supply chain context and the another 40% to the intra-organizational context. this may due to the fact that in the last years the supply chain context is having more importance and relevance within the organizational environment. then, the study, analysis and improvement of this ambit is the objective of more and more researching works. regarding the fuzzy logic, the fact that there are more references related to the supply chain tan to the intra-organizational context is due to the own uncertain nature inherent to the supply chain, which makes the fuzzy logic to be an adequate option. on the other hand, a future research line is to delimit the uncertainty base on the unpredictable factors, both internal and external, that surround both the organizations and the supply chains and that are affecting to the performance measurement process. there a multitude of unpredictable factors of different nature that may modify the performance in different aspects and whose management is beyond the capabilities that usually can be found in any organization. for instance, an internal factor that modifies the performance could be the lack of historical data regarding important aspects of the organization; on the other hand, an external aspect could be the modification of any macroeconomic variable that surround the organization, for example the oscillations in the value of the euro. then, the application of fuzzy logic for delimiting the uncertainty is one of the key applications when applying a mechanism to face all these unpredictable factors. finally, other future research line, whose main objective were to determine in what phases of the management-measurement process is better to apply fuzzy logic, would be welcome. references amini, s. & jochem, r. (2011). a conceptual model based on the fuzzy set theory to measure and evaluate the performance of service processes. 15th international enterprise distributed object computing conference workshops. aug. 29-sept. 2, 2011, helsinki, finland. pp. 122-131. http://dx.doi.org/10.1109/edocw.2011.25 ammar, s. & wright, r. (1995). a fuzzy logic approach to performance evaluation. uncertainty modeling and analysis, 1995, and annual conference of the north american fuzzy information processing society. proceedings of isuma nafips ‘95., pp. 246 251 ammar, s. & wright, r. (2000). applying fuzzy-set theory to performance evaluation. socio-economic planning sciences, 34(4): 285-302. http://dx.doi.org/10.1016/s0038-0121(00)00004-5 arango, m.d., jaimes, w.a. & zapata, j.a. (2010). gestión cadena de abastecimiento logística con indicadores bajo incertidumbre, caso aplicado sector panificador palmira. ciencia e ingeniería neogranadina, 20(1): 97-115. 98 int. j. prod. manag. eng. (2014) 2(2), 93-100 creative commons attribution-noncommercial-noderivatives 4.0 international gurrea, v., alfaro-saiz, j.j., rodríguez, r., & verdecho, m.j. http://dx.doi.org/10.1109/edocw.2011.25 http://dx.doi.org/10.1016/s0038-0121(00)00004-5 http://creativecommons.org/licenses/by-nc-nd/4.0/ beheshti, h.m. & lollar, j.g. (2008). fuzzy logic and performance evaluation: discussion and application. international journal of productivity and performance management, 57(3): 237-246. http://dx.doi.org/10.1108/17410400810857248 behrouzi, f. & wong, k.y. (2010). lean performance evaluation of manufacturing systems: a dynamic and innovative approach. procedia computer science, 3: 388-395. http://dx.doi.org/10.1016/j.procs.2010.12.065 behrouzi, f., wong, k.y. & kuah, c.t. (2010). a fuzzy-based model to measure supplier performance with lean attributes. fourth asia international conference on mathematical/analytical modelling and computer simulation (ams), 26-28 may, 2010. kota kinabalu, malaysia. pp. 372 377. http://dx.doi.org/10.1109/ams.2010.81 campuzano, f., mula, j., peidro, d. (2010) fuzzy estimations and system dynamics for improving supply chains fuzzy sets and systems, 161(11): 1530-1542. http://dx.doi.org/10.1016/j.fss.2009.12.002 campuzano-bolarín, f., mula, j., peidro, d. (2013). an extension to fuzzy estimations and system dynamics for improving supply chains. international journal of production research, 51(10): 3156-3166. http://dx.doi.org/10.1080/00207543.2012.760854 chan, t.s., qi, h.j. (2003). an innovative performance measurement method for supply chain management. supply chain management: an international journal, 8(3): 209-223. http://dx.doi.org/10.1108/13598540310484618 chan, t.s., qi, h.j., chan, h.k., lau, c.w. & ip, w.l. (2003). a conceptual model of performance measurement for supply chains. management decision, 41(7): 635-642. http://dx.doi.org/10.1108/00251740310495568 chen, c.t., lin c.t. & huang s.f., (2006). a fuzzy approach for supplier evaluation and selection in supply chain management. int. j. production economics, 102: 289-301. http://dx.doi.org/10.1016/j.ijpe.2005.03.009 cheng, s.r., hsu, b.m. & shu, m.h. (2007). fuzzy testing and selecting better processes performance. industrial management & data systems, 107(6): 862-881. http://dx.doi.org/10.1108/02635570710758761 datta, s., samantra, c., sankar, s., mandal, g. & majumdar, g., (2011). appraisement and selection of third party logistics service providers in fuzzy environment. benchmarking: an international journal, 20(4): 537-548. http://dx.doi.org/10.1108/bij-11-2011-0087 ferreira, a., azevedo, s. & fazendeiro, p. (2012). a linguistic approach to supply chain performance assessment”. ieee international conference on fuzzy systems, 10-15 june, 2012. brisbane, australia. pp.1-5. http://dx.doi.org/10.1109/fuzz-ieee.2012.6251327 ganga, g.m.d. & carpinetti, l.c.r. (2011). a fuzzy logic approach to supply chain performance management. int. j. production economics, 134(1): 177-187. http://dx.doi.org/10.1016/j.ijpe.2011.06.011 jain, v., tiwari, m. & chan, e. (2004). evaluation of the supplier performance using an evolutionary fuzzy-based approach. journal of manufacturing technology management, 15(8): 735-744. http://dx.doi.org/10.1108/17410380410565320 lau, h.c.w., pang, w.k. & wong, c.w.y. (2002). methodology for monitoring supply chain performance: a fuzzy logic approach. logistics information management, 15(4): 271-280. http://dx.doi.org/10.1108/09576050210436110 lalmazloumian, m. & wong, k.y., (2012). a review of modelling approaches for supply chain planning under uncertainty. 9th international conference on service systems and service management (icsssm), 2-4 july, 2012, shanghai, china. pp. 197-203. http://dx.doi. org/10.1109/icsssm.2012.6252220 liao, m.y. & wu, c.w., (2010). evaluating process performance based on the incapability index for measurements with uncertainty. expert systems with applications, 37(8): 5999-6006. http://dx.doi.org/10.1016/j.eswa.2010.02.005 lu, c. & li, x.w. (2006). supply chain modeling using fuzzy sets and possibility theory in an uncertain environment. the 6th world congress on intelligent control and automation, pp. 3608-3612. http://dx.doi.org/10.1109/wcica.2006.1713042 mahnam, m., yadollahpour, m.r., famil-dardashti, v. & hejazi, s.r. (2008). supply chain modeling in uncertain environment with bi-objective approach. computers & industrial engineering, 56(4): 1535-1544. http://dx.doi.org/10.1016/j.cie.2008.09.038 mula, j., campuzano-bolarin, f., díaz-madroñero, m., carpio, k.m. (2013). a system dynamics model for the supply chain procurement transport problem: comparing spreadsheets, fuzzy programming and simulation approaches. international journal of production research, 51(13): 4087-4104. http://dx.doi.org/10.1080/00207543.2013.774487 muñoz, m., rivera, j. & moneva, j. (2008). evaluating sustainability in organisations with a fuzzy logic approach. industrial management & data systems, 108(6): 829-841. http://dx.doi.org/10.1108/02635570810884030 nunes, i.l., figueira, s. & machado, v.h. (2011). evaluation of a supply chain performance using a fuzzy decision support system. ieee international conference on industrial engineering and engineering management (ieem), 6-9 dec., 2011. singapore. pp. 1735-1739. http://dx.doi.org/10.1109/ieem.2011.6118213 ohdar r. & kumar p. (2004). performance measurement and evaluation of suppliers in supply chain: an evolutionary fuzzy-based approach. journal of manufacturing technology management,15(8): 723-734. http://dx.doi.org/10.1108/17410380410565311 olugu, e.u. & wong, k.y. (2012). an expert fuzzy rule-based system for closed-loop supply chain performance assessment in the automotive industry. expert systems with applications, 39(1): 375-384. http://dx.doi.org/10.1016/j.eswa.2011.07.026 ordoobadi s.m. (2009). development of a supplier selection model using fuzzy logic. supply chain management: an international journal, 14(4): 314-327. http://dx.doi.org/10.1108/13598540910970144 99int. j. prod. manag. eng. (2014) 2(2), 93-100creative commons attribution-noncommercial-noderivatives 4.0 international application of fuzzy logic in performance management: a literature review http://dx.doi.org/10.1108/17410400810857248 http://dx.doi.org/10.1016/j.procs.2010.12.065 http://dx.doi.org/10.1109/ams.2010.81 http://dx.doi.org/10.1016/j.fss.2009.12.002 http://dx.doi.org/10.1080/00207543.2012.760854 http://dx.doi.org/10.1108/13598540310484618 http://dx.doi.org/10.1108/00251740310495568 http://dx.doi.org/10.1016/j.ijpe.2005.03.009 http://dx.doi.org/10.1108/02635570710758761 http://dx.doi.org/10.1108/bij-11-2011-0087 http://dx.doi.org/10.1109/fuzz-ieee.2012.6251327 http://dx.doi.org/10.1016/j.ijpe.2011.06.011 http://dx.doi.org/10.1108/17410380410565320 http://dx.doi.org/10.1108/09576050210436110 http://dx.doi.org/10.1109/icsssm.2012.6252220 http://dx.doi.org/10.1109/icsssm.2012.6252220 http://dx.doi.org/10.1016/j.eswa.2010.02.005 http://dx.doi.org/10.1109/wcica.2006.1713042 http://dx.doi.org/10.1016/j.cie.2008.09.038 http://dx.doi.org/10.1080/00207543.2013.774487 http://dx.doi.org/10.1108/02635570810884030 http://dx.doi.org/10.1109/ieem.2011.6118213 http://dx.doi.org/10.1108/17410380410565311 http://dx.doi.org/10.1016/j.eswa.2011.07.026 http://creativecommons.org/licenses/by-nc-nd/4.0/ pilevari, n., seyedhosseini, s.m. & jassbi, j. (2008). fuzzy logic supply chain agility assessment methodology. ieee international conference on industrial engineering and engineering management, 8-11 dec., 2008, singapore. pp. 1113-1117. http://dx.doi.org/10.1109/ ieem.2008.4738043 serna, m., serna, c. & pérez, g. (2008). aplicaciones de lógica difusa a las cadenas de suministro. revista avances en sistemas e informática, 5(3), isnn 1657-7663 tabrizi, b.h. & razmi, j., (2013). introducing a mixed-integer non-linear fuzzy model for risk management in designing supply chain networks. journal of manufacturing systems, 32(2): 295-307. http://dx.doi.org/10.1016/j.jmsy.2012.12.001 theeranuphattana, a. & tang, j.c.s. (2008). a conceptual model of performance measurement for supply chains, alternative considerations. journal of manufacturing technology management, 19(1): 125-148. http://dx.doi.org/10.1108/17410380810843480 unahabhokha, c., platts, k. & tan, k.h. (2007). predictive performance measurement system, a fuzzy expert system approach. benchmarking: an international journal, 14(1): 77-91. http://dx.doi.org/10.1108/14635770710730946 van der vorst, j.g.a.j. & beulens, a.j.m. (2002). identifying sources of uncertainty to generate supply chain redesign strategies. international journal of physical distribution and logistics management, 32(6): 409-430. http://dx.doi.org/10.1108/09600030210437951 wang, j. & shu, y.f. (2009). fuzzy decision modeling for supply chain management. fuzzy sets and systems, 150(1): 107-127. doi:10.1016/j. fss.2004.07.005 wei, c.c., liou, t.s. & lee, k.l. (2008). an erp performance measurement framework using a fuzzy integral approach. journal of manufacturing technology management, 19(5): 607-626. http://dx.doi.org/10.1108/17410380810877285 xu, l.x.x., ma, b. & lim, r. (2008). supplier performance measurement in a supply chain. 6th ieee international conference on industrial informatics, 13-16 july, 2008, daejeon, south korea. pp. 877-881. http://dx.doi.org/10.1109/indin.2008.4618224 100 int. j. prod. manag. eng. (2014) 2(2), 93-100 creative commons attribution-noncommercial-noderivatives 4.0 international gurrea, v., alfaro-saiz, j.j., rodríguez, r., & verdecho, m.j. http://dx.doi.org/10.1109/ieem.2008.4738043 http://dx.doi.org/10.1109/ieem.2008.4738043 http://dx.doi.org/10.1016/j.jmsy.2012.12.001 http://dx.doi.org/10.1108/17410380810843480 http://dx.doi.org/10.1108/14635770710730946 http://dx.doi.org/10.1108/09600030210437951 http://dx.doi.org/10.1016/j.fss.2004.07.005 http://dx.doi.org/10.1016/j.fss.2004.07.005 http://dx.doi.org/10.1108/17410380810877285 http://dx.doi.org/10.1109/indin.2008.4618224 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.17620 received: 2022-04-27 accepted: 2022-07-18 supply chain network design: a case study of the regional facilities analysis for a 3d printing company brena carvalho de sá a, erick henrique dutra de souza b, luciana paula reis c, michael david de souza dutra d* a federal university of goiás, av. esperança, chácaras de recreio samambaia, goiânia go, 74690-900, brazil. b centro federal de educação tecnológica de minas gerais cefet-mg, av. amazonas, 5.253, nova suiça, belo horizonte, mg, brasil. cep: 30.421-169, brazil. c federal university of ouro preto, rua 36, joão monlevade mg, 35931-008, brazil. d polytechnique montréal, 2500 chem. de polytechnique, montréal, qc, canadá. a1 brenasa@discente.ufg.br, a2 20193018454@aluno.cefetmg.br, a3 lucianapaula@ufop.edu.br, a4 michaeldavid.dutra@polymtl.ca abstract: abstract: 3d printing supply chain network. the objective is to analyze regional facility configurations in order to lower investment risks for an organization that aims to provide additive manufacturing the growing 3d printing market can be an attraction for investment in new businesses, which may entail strategic planning for new ventures. this paper presents a case study of designing a services for orthopedic and dental prostheses production. to this end, the competitive environment, the aggregating factor and logistic costs, tariffs and tax incentives, regional demand, political factors, the value of currency, and the demand uncertainty are analyzed. the results indicate that the adopted framework for network design decisions effectively allows the analysis of regional facility configuration. it also suggests that there are no hindering factors to the implementation of a 3d printing service company. in the region studied, there are fiscal incentives of more than 60% for taxes on the movement of goods between municipalities, which can be an advantage when locating facilities outside the capital. competitors are well qualified, but there is room for new companies focused on quality and price, which may be a case for specialized products such as protheses. the estimated demand ranges from 146 to 509 units per month, which may be an opportunity for a new entrant given the few additive manufacturing ventures identified in the region. key words: supply chain, physical network design, 3d printing, production technology. 1. introduction it is not possible to stop investing in the modernization of machines and equipment, as this is a business necessity that generates competitive advantage for an organization (de souza dutra, 2015). this modernization brought the 3d printing, known in the industrial context as additive manufacturing, which is the process of transforming a digital model from a file into a physical object. that is, through a printer, with inputs such as resins and inks, a consumer can print an object using a digital file in his own home, producing, from this intangible digital model, a good for use and consumption (gasparino, 2021). in the context of industry 4.0, also called advanced manufacturing, or smart grid in the energy sector (souza dutra, 2019), additive manufacturing has been widely used. berman (2012), conducted a to cite this article: carvalho de sá, b., dutra de souza, e.h., reis, l.p., de souza dutra, m.d. (2022). supply chain network design: a case study of the regional facilities analysis for a 3d printing company. international journal of production management and engineering, 10(2), 211-223. https://doi.org/10.4995/ijpme.2022.17620 int. j. prod. manag. eng. (2022) 10(2), 211-223creative commons attribution-noncommercial-noderivatives 4.0 international 211 mailto:email@author3.com http://creativecommons.org/licenses/by-nc-nd/4.0/ study that proposed to evaluate the characteristics and applications of 3d printing, defining it as “mass customization” as it allows the manufacturing of small quantities of customized products at relatively low costs. the positive and negative impacts of additive manufacturing on production systems have repercussions on the competitiveness of organizations, affecting competitive criteria. according to the literature review conducted by veit (2018), the criterion most positively impacted by the use of additive manufacturing is cost. one of its consequences of its lower operating cost was reflected in the cost of the equipment itself. the cost of printers is decreasing over the years after the fall of patents (berman, 2012). with the advancement of this technology and the development of new software, the first mass production with 3d printers appeared and some manufacturers developed the first simulations for the formation of 3d “printing farms”. industries in the automotive, aerospace, health, and construction fields use this technology for both prototyping and components production using topology and optimization techniques to fit three-dimensional models, seeking the maximum performance and best design (nogueira, 2021). according to the german portal statista, the 3d printing market moved us $14.5 billion in 2019. in brazil, by 2022, 49% of companies intend to invest in this technology, in view of its price reduction in recent years and the wide competitive advantage provided to the production process (impressão, 2020). in the state of goiás, specifically in the metropolitan region of goiânia, this market is present in some establishments. with this, new entrants can expect fierce competition. a plan to deal with such competition can be developed by defining a competitive strategy, which is defined as how an organization competes in the market. in other words, competitive strategy defines how the organization will be perceived by customers vis-à-vis competitors through advantage over such competitors. therefore, according to porter (2004), competitive strategy refers to a strategic position to compete in the market, always thinking about a vision of the future. the study of the strategic plan for the insertion of a new entrant in a supply chain is one of the subjects discussed by chopra and meindl (2013). for example, these authors propose a framework, consisting of four phases, to assist in the design of the physical network of a supply chain. in particular, the objective of phase ii of this framework is to determine the regions where the facilities will be located, their functionality and capacity. this is done by studying the demand and its uncertainty, exchange rate, tariffs and policies related to regional markets, export or import incentives, and restrictions for each market. the 3d printing market in brazil was studied by almeida (2019). this author verified the existence of several challenges to be overcome, such as the transition from traditional production methods to the new industry 4.0 processes. the objective of almeida’s work is to identify the main barriers to the implementation of additive manufacturing in brazil. on the other hand, there is no work in the literature whose focus is on 3d printing service in the state of goiás. the lack of information can inhibit investments (de souza dutra, 2022). thus, in this paper, the general objective is to apply the aforementioned phase ii to assist the installation of a 3d printing service unit in the metropolitan region of goiânia. specifically, considering the current context of goiânia/aparecida de goiânia (brazil) and an intention to install a venture related to the 3d printing factory in the region, it is aimed to analyze the following factors for the venture: necessary production technologies, competitive environment, aggregating factor and logistics costs, tariffs and tax incentives, regional demand, political factors, value of currency and demand uncertainty. 2. literature review according to chopra and meindl (2011), the design of a physical network for a supply chain aims to maximize the company’s profits while satisfying customer needs with respect to demand and responsiveness. that said, a model is proposed by these authors to help in such a design. this model, composed of four phases, begins with strategic decisions, such as product and target market definition, market study, and the definition of regions as potential candidates to host a new business facility. in the second phase, the potential regions are studied in detail considering logistical factors of the enterprise. the second phase of the aforementioned model is illustrated in figure 1. int. j. prod. manag. eng. (2022) 10(2), 211-223 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho de sá et al. 212 http://creativecommons.org/licenses/by-nc-nd/4.0/ phase ii aims to identify which region best suits the needs of one or more facilities. for this, some decision elements must be studied. a first element to be analyzed for the definition of the network project is the demand per region. each region will have a demand, different proportions, and different customer requirements. for some regions, the requirements may be homogeneous, which favors large consolidated facilities; comparatively, for other regions, the requirements may be heterogeneous, which favors facilities with lower capacity and more centered on certain products (chopra and meindl, 2011). another element to be analyzed is to identify the availability of production technologies and their influence on cost reduction. if economies of scale or scope are not significant, a decision tends to be that each market will have its own facility. on the other hand, if the economies of scale or scope are significant, a few facilities are expected to serve many markets (chopra and meindl, 2011, p. 123). a third element of study is the identification in each of the regional markets about the uncertainties of demand, the exchange rate, and the political influence on the environment in which the markets are located. this decision element also requires studies about tariffs, tax incentives, import and export restrictions, and the legal regulations for producing in a particular region. finally, deciding whether facilities will be closer or farther from competitors is also necessary (chopra and meindl, 2011). in the following sections, the concepts of these elements for defining the configuration of regional facilities will be presented. 2.1. regional facility configuration according to bavaresco (2013), the success of a venture requires an analysis such as the location and the public for which the business is intended. for this, it is necessary the anticipated study of the macro environment in which the company will operate, which includes an analysis of the social, cultural, economic, technological, and legislative potentialities. entrepreneurs analyze the feasibility of setting up the new business based on environmental studies and research that can define the success factors of the enterprise. these factors play an important role in the decisions necessary to build a network project. the process of evaluating the environment and the market that is intended to be reached with the new business should be considered as one of these factors because in many cases the rush to start a business causes wrong choices to delay or even annul the profitability and the return on the expected investment (bavaresco, 2013). thus, an analysis of the competitive environment and an evaluation of regional facilities is as important as a business plan for the consolidation of any company’s future. the importance of this plan serves not only for the company’s future plans, but also as a detailed means of verifying the seriousness of the organization (bavaresco, 2013). 2.2. production technologies production technologies available in an environment or a market have a great importance on network design decisions. their characteristics relative to the production cost are significant factors that define figure 1. phase ii of chopra and meindl’s framework for designing the physical network of a supply network. source: chopra and meindl (2011). int. j. prod. manag. eng. (2022) 10(2), 211-223creative commons attribution-noncommercial-noderivatives 4.0 international supply chain network design: a case study of the regional facilities analysis for a 3d printing company 213 http://creativecommons.org/licenses/by-nc-nd/4.0/ even the amount of facilities a company needs. such technologies can provide economies of scale or scope, as well as influence production flexibility (chopra and meindl, 2011). economies of scale is understood as the cost reduction that occurs when companies increase production. there are many reasons why process cost decreases simultaneously with increased production volume. one example would be keeping production aligned and focused on just one product, providing cost savings associated with the exchange of raw materials and equipment between different productions (amadeo, 2021). in addition to economies of scale, economies of scope are also present. according to caetano (2014), economies of scope occur when some input resources are shared in the production of goods and services. 2.3. competitive environment the concept of competitive environment is linked to the conceptualization and importance of what the organizational environment is. this definition, which was initially linked only to internal aspects of organizations, after the influence of the open systems theory, in the 50’s and 60’s, was expanded to incorporate the idea of environment. according to lawrence and lorsch (1967), the organization can be seen as the coordination of different individual contributions for the purpose of carrying out planned transactions with the environment. in the same vein, research conducted by the biologist bertalanffy (1972) conceives organizations as open systems. thus, the author considers that organizations and their external environments are parts of a larger system that continuously interact. the organization exchanges resources with the environment, ensuring its survival, and modifies itself to adapt to the environmental contingencies that provide it access to these resources. as a result, it acquires new properties and characteristics. thus, organizations are open systems that relate to the external environment. the structural configurations theory, presented by mintzberg (2013), defines that organizational effectiveness is related to situational factors (linked to the organization itself and its external environment). the occasional factors linked to the organization’s external environment are: complexity and technological instability, sales instability, market diversity, the degree of hostility in competition, and power relationships involving external influencers, thus defining the scenario of a competitive environment. 2.4. aggregating factor and logistics costs for chopra and meindl (2013), key logistics factors (facilities, inventory, transportation, information) determine supply chain performance as a function of responsiveness and efficiency. similarly, ballou (2006) defines that logistics is about adding value to products and services offered to customers. it is expected that there is a directly proportional relationship between value and integrated cost into a network project, because the higher the responsiveness of the supply chain, the higher the operational cost and consequently, the lower the efficiency. in this context, logistics cost, which impacts decisions about the design of the physical supply chain network, is defined as equal to the sum of inventory, installation, transportation, and information system costs. the higher the logistics cost, the lower the efficiency. in general, no distribution network will outperform in responsiveness and efficiency at the same time. one way to counterbalance these performances is using aggregating factors, such as the use of warehouses for vehicle consolidation. 2.5. tariffs and tax incentives for chopra and meindl (2013), as global trade increases, macroeconomic factors such as tariffs and tax incentives contribute to the failure or success of supply chain networks. thus, the relevance of considering these factors in making an efficient network design decision is undeniable. the government has power to exercise when it comes to taxation. when the state exercises its power to tax, according to a specific average burden, which is applied to all without distinction, we observe fiscal activity. however, when the focus of the tax activity is to stimulate a particular activity, group or legally protected value, then there is a function that is conventionally called “extra fiscal”. it is in this last aspect that the theory of tax incentives is applied (catão, 2004, p.4). “extra fiscal” is characterized by nabais (2012) as a set of rules that, although formally integrate tax law, have as its main or dominant purpose the achievement of certain economic or social results through the use of the tax instrument and not the int. j. prod. manag. eng. (2022) 10(2), 211-223 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho de sá et al. 214 http://creativecommons.org/licenses/by-nc-nd/4.0/ collection of revenue to meet public expenditure. these tax incentives aim to influence business decisions taking into account the tax incidence. 2.6. regional demand in the economic field, “demand” means the quantity of a good or service that the market or a group of consumers wishes or wants to buy. thus, demand is the desire or need supported by the consumer’s ability and intention to buy. demand exists because there is, among other things, purchasing power. however, this same buying power oscillates according to the economic context of each country, region or market operation. in a crisis context, for example, it is crucial to analyze not only what the population is consuming but how consumption is being influenced (dino, 2018). this demand analysis impacts the design of the physical supply chain network. chopra and meindl (2013) cite that fluctuations in demand may exist depending on the political and economic context of a region, or even, demand may not be homogeneous across regions. for example, in 2017, datafolha published that 46% of brazilians surveyed believed that their economic situation would improve. however, even with falling inflation in january 2018, data released in the employment and unemployment survey, pointed to a 0.7% reduction in average income (purchasing power) (dino, 2018). thus, the volume of demand was affected by an inflation context experienced by consumers. as per chopra and meindl (2013), regional demand can influence the decision on the capacity of a facility, on the type of production technology to be used considering production flexibility vs. economies of scale and scope to meet the consumption desires of the local culture, among other decisions. 2.7. political factors, currency exchange rate and demand risk there are many factors that influence the planning of an efficient network project, among which political factors are present. according to chopra and meindl (2011), the political stability of the country plays a significant role in choosing the location of the network. politically stable countries with free, independent and clear legal systems, in which rules of the economy are well defined, are territories that companies prefer to locate their facilities. based on literature, political factors look at the degree of government interference in the economy and include tax policies, labor regulations, trade agreements between countries, change of government, wars, and conflicts. for the company, the greater the political stability, the more beneficial it is for its development (oliveira, 2020). as far as inflation is concerned, its increase decreases people’s purchasing power. however, one should not think that deflation (when inflation is below zero) is suitable for business. in deflation, if prices fall and this becomes a trend, people stop buying to save money, which decreases sales and the currency in circulation in the market, leading to an economic recession. therefore, the ideal for business is for inflation to be controlled (rocha, 2019). another relevant factor when choosing a network project is the value of the currency (foreign exchange). fluctuations in foreign exchange rates happen frequently and impact the supply chain profits. an example of this is the value of the dollar currency and its fluctuation in yen, japanese currency, during the years 2002 and 2004. the appreciation of the yen decreased revenues from major foreign markets. as a result, many japanese companies built their production facilities spread all over the world (chopra and meindl, 2011). the us dollar rate is an uncertainty in the global scenario that directly interferes with the price of some imported raw materials and equipment used in additive manufacturing (betim et al., 2019). fluctuations in demand caused by changes in the economy of each country also influence network design decisions. for example, during the period 1996 and 1998, asian economies contracted, consequently decreasing the purchasing power of society. this decreased the demand in the local market, implying increased idle capacity in production facilities in asia. toyota, a company that had assembly plants in asia that were only able to meet the local market, was motivated by the crisis in the continent to make its factories more flexible in order to supply the demand of other locations as well (chopra and meindl, 2011). 3. methodology as qualitative study with a case study approach, this research was conducted on the brazilian small and medium-sized company “pro3d”. data collection was carried out by observation, documentation, and empirical and bibliographic research. this study int. j. prod. manag. eng. (2022) 10(2), 211-223creative commons attribution-noncommercial-noderivatives 4.0 international supply chain network design: a case study of the regional facilities analysis for a 3d printing company 215 http://creativecommons.org/licenses/by-nc-nd/4.0/ focuses on analyzing regional facility configurations to decrease investment risks of an organization that aims to provide additive manufacturing services for orthopedic and dental prostheses. to this end, the competitive environment, aggregating factor, logistics costs, tariffs and tax incentives, regional demand, political factors, the value of currency, and the uncertainty of demand are analyzed. to identify the production technologies, an exploratory literature search was conducted with the topics related to 3d printers. furthermore, considering the product and the production layout, the necessary production technologies were defined. the competitive environment was analyzed via experimental research. this field research comprises the selection of establishments that provide 3d printing services through the tool “google maps”. then, from the listing of these companies, visits were made to such companies for a survey of data and information about the competitive environment of the additive manufacturing market in the metropolitan region of goiânia, capital of goiás. related to the aggregating factor and logistics costs, based on initial definitions proposed by the organization concerning the network model, the possibilities that exist in the region under study regarding the consolidation of vehicles for an operation to have satisfactory logistics costs are analyzed. to study tariffs and tax incentives in the goiás state, research was done on news portals and government websites about the current existence of tariffs and tax incentives in the state. interviews were conducted with potential customers in order to obtain information about their interest in the product defined by the organization, that is, prostheses. from this collected information, the demand for 3d prints of this product was approximated. a literature search was conducted in relation to brazil risk (credit default swap brazil) for a discussion of the influence of political factors on business. currency value was not analyzed in this paper, as there are no envisioned facilities and markets in countries with different currencies. 4. results and discussion 4.1. strategic description of the enterprise at a time prior to this work, the company’s strategic analysis was carried out based on phase i of chopra and meindl’s (2011) supply network project framework. the results of this analysis defined the strategies of a supply chain for a venture that aims to provide additive manufacturing services. to this end, it was considered data collected via applied and bibliographic research, the important aspects of the business and equating the theoretical strategic vision with the actual scenarios found in the metropolitan region of goiânia. the products defined by the organization are orthopedic and dental prostheses. the target market is patients in the public health system. it is estimated that potential customers will require such products being delivered to medical laboratories, dental clinics, or hospitals. initially, 12 sub-regions with potential customers were initially identified, whose locations are illustrated in figure 2. figure 2. potential customers. source: google maps. a benchmark performed by silva et al. (2022) is used in this paper to assist in the guidelines for defining the production flow and the layout for the enterprise. the latter was defined as a functional layout, which is illustrated in figure 3. figure 3. layout of the facility for manufacturing parts by 3d printing. source: (silva et al., 2022). the suggested production flow is illustrated in figure 4. int. j. prod. manag. eng. (2022) 10(2), 211-223 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho de sá et al. 216 http://creativecommons.org/licenses/by-nc-nd/4.0/ initially, a print request is received via sharepoint. then, it is checked if the part is for mandatory maintenance, in which case it is analyzed if the .stl file (computational model) exists. if the requested piece is not related to mandatory maintenance, it is considered a request for process improvement purposes. then, the urgency is analyzed. if it is not urgent, the request for this piece is placed in a queue for future analysis. however, if it is urgent, it is checked if the .stl file is already complete and ready for printing. when the.stl file is already complete and ready for printing, the request will be placed in the print queue, which is managed by kanban. otherwise, the modeling flow illustrated in figure 5 is executed, which aims to create the .stl file. figure 4. print flowchart. source: (silva et al., 2022). figure 5. modeling flowchart. source: (silva et al., 2022). int. j. prod. manag. eng. (2022) 10(2), 211-223creative commons attribution-noncommercial-noderivatives 4.0 international supply chain network design: a case study of the regional facilities analysis for a 3d printing company 217 http://creativecommons.org/licenses/by-nc-nd/4.0/ in the modeling flow, the first activity is to understand how the piece works in its real use context. it is desired that measurements and wear characteristics are understood in order to begin the computational modeling activity, that is, the computational design of the piece. when this design is finished, the side, top, and front projections of the designed piece are printed on a4 paper. this printout is used for visual comparison of dimensions with a physical part via overlay of the part on the printout. with this, it is possible to check whether the dimensions are correct, in which case the part is 3d printed. if the dimensions of the piece, compared to the a4 printout, are visually problematic, the process is restarted. once the piece has been 3d printed, finally a validation step is done. in this step, the part’s resistance and operating temperature characteristics are studied, which implies the choice of the material for final printing. as a result of this process, the printing settings for this piece are specified in the. stl file. finally, the proposed broadly physical network model for the supply chain is represented in figure 6. this network design is classified as “storage at the distributor with direct delivery” by chopra and meindl (2011). 4.2. production technologies in view of the technological evolution of industry 4.0, betim et al. (2019) conducted a prospective scenario analysis focused on additive manufacturing in brazil for 2024. the research was done using the momentum methodology, which identified three scenarios. they concluded that, although this market is growing moderately, it is necessary to encompass the leading countries in this type of manufacturing, with changes in the national scenario to encourage the additive manufacturing market. these results are explained on the basis that machinery and raw material costs directly impact production, which may be dependent on imports. already cost-effective, the alternatives for manufacturing processes are diverse (betim et al., 2019). there are three main technologies used for additive manufacturing operations, each with its particularities. the design features and requirements are listed in tables 1 and 2, respectively. for ryan et al. (2017), three types of manufacturing operation on 3d printers can be identified: 1. craft: low volume of products is produced by low-cost equipment. users typically operate equipment in this type of manufacturing; 2. job shop: the equipment, operated by specialists, has higher quality and cost, with low continuous production volume; 3. factory, where the equipment is not only specialized, but also used by operators trained in it. the production volumes are high, so manufacturing procedures are standardized. according to conner et al. (2014), conventional manufacturing may be less competitive than 3d printing when it comes to manufacturing products with higher levels of customization, complexity, or a combination of both. for products with low complexity, low volume, and low customization, additive manufacturing will be desired only if it provides lower lead time and lower cost compared to conventional methods. considering 3d printing technologies, sls is adopted since such technology has been used for the purpose of orthopedic prostheses, the product of the business under study. considering the production flow, it is believed that a place that provides 3d printing service only needs, in addition to printers, a computer to communicate with such printers. job shop may be the ideal process for the defined product. 4.3. competitive environment from the search conducted on google’s maps tool, it was obtained, as a result, 9 companies that provide 3d printing services and that are located in the city of goiânia, which are illustrated in figure 7. the search was conducted by companies that are active in the market and have a physical store to receive customers. figure 6. proposed physical network model. source: adapted from silva et al. (2022). int. j. prod. manag. eng. (2022) 10(2), 211-223 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho de sá et al. 218 http://creativecommons.org/licenses/by-nc-nd/4.0/ a focus group, consisting of 12 people, visited and used the services of these 9 competitors to comparatively evaluate service-related characteristics. table 3 shows in its first column these characteristics and in its first row a symbolic reference for the nine competitors (c1 to c9). the evaluation considers scores from 1 to 10, with 1 and 10 being the worst and the best, respectively. the result for each characteristic is given by the rounded average of the 12 individual scores. the last column of table 3 shows the sum of the scores of all the competing companies for a certain characteristic. based on table 3, there are competitors in the region with scores higher than 8 in relation to all features. based on the sum of the evaluations, the features “product quality” and “price” have the lowest scores. considering the specificity of the table 1. characteristics associated with each 3d printing method. feature fused deposition modelling (fdm) stereolithography (sla) selective laser sintering (sls) working principle extrusion of cast material light curing resin sintered microparticles type of compatible materials thermoplastics light cured resins thermoplastics quantity of compatible materials very high medium low price of materials mediumlow high medium complexity high medium medium production speed very high medium low minimum layer resolution 0.1 mm 0,05 mm 0.06 mm maximum xy resolution 0.25 mm 0.05 mm 0.08 0.08 mm accuracy low medium high application rapid prototyping. models with small details. functional prototypes. advantages education. casting negatives for jewelry and dentistry. short series. disadvantages model and tool making. splints. models and tools. source: filament2print (2020). table 2. design requirements associated with each 3d printing method. feature fdm sla sls minimum wall thickness 0.8 mm 0.5 mm in supported walls. 1 mm in unsupported walls 0.7 mm engravings 0.6 mm width 2 mm from top 0.4 mm 1 mm minimum hole diameter 2 mm 0.5 mm 1.5 mm tolerance on moving parts and connections 0.5 mm 0.5 mm 0.3 mm in moving parts 0.1 mm in connections holes in hollow parts no need 4 mm 5 mm minimum detail size 2 mm 0.2 mm 0.8 mm minimum column diameter 3 mm 0.5 mm 0.8 mm general tolerances ±0.5% (lower limit ±0.5 mm) ±0.5% (lower limit ±0.15 mm) ±0.3% (lower limit ±0.3 mm) source: filament2print (2020). figure 7. 3d printing companies in goiânia. source: google maps. int. j. prod. manag. eng. (2022) 10(2), 211-223creative commons attribution-noncommercial-noderivatives 4.0 international supply chain network design: a case study of the regional facilities analysis for a 3d printing company 219 http://creativecommons.org/licenses/by-nc-nd/4.0/ organization’s product (prostheses), a strategy in the face of the competitive environment is to focus on specialization of production as a way to target both features. it is believed that this specialization can also provide good response times and increase profitability. this decision corroborates conner et al. (2014). finally, a location within the metropolitan region of goiânia is indicated. although this decision puts the facility close to the competitors, it also puts it close to potential customers. thus, the facility can benefit from the flow of people. furthermore, according to the network model proposed by the company’s strategic plan, customers will receive the products at its facilities. therefore, a location in the metropolitan region of goiania can reduce transportation costs. 4.4. aggregating factor and logistics costs one of the important factors for the reduction of logistical costs is the aggregating factor, or even, freight consolidation. the suggested strategy is to perform vehicle consolidation for purchases. the network project classified as “storage at the distributor with direct delivery” foresees buying directly from manufacturers, which suggests buying according to a minimum order lot. with this, orders can be placed with a view to consolidating a vehicle. the objective is to guarantee the total occupation of the vehicle’s capacity and the consequent reduction of costs in the delivery routine. therefore, in the consolidation of vehicles, the company groups small quantities of different products, guaranteeing total occupancy of the vehicle, taking advantage of all its capacity. 4.5. tariffs and tax incentives all individuals and companies can apply for tax incentives. governments are in charge of evaluating each application. among the factors evaluated are compatibility of costs, the interest of the government, compliance with legislation, and technical capacity. obtaining tax incentives is a good solution for organizations, and one that positively benefits the image and visibility of a business. it is essential that the organization is up to date with the tax authorities for the requested project to be approved. the processing of a tax incentive is, in general, easy and not very bureaucratic, as long as the companies meet the requirements to obtain it. technology policy is currently a central part of the economic agenda in developed countries and in emerging countries such as brazil. starting in the 2000s, it was possible to observe the brazilian government’s initiatives to transform the discourse of innovation into effective actions. laws came into force that encourage technological production and the openness for the import of new production technologies. for instance, the innovation law no. 10,973, enacted on december 2, 2004, was created with the aim of providing tax incentives for innovation and scientific and technological research, and law no. 13,243, january 11, 2016, was created to represent the new legal basis for innovation. besides these, in brazil there is also the law of good, which is a policy to foster innovation. such a policy is part of the government’s strategic expansionist policies for the scientific and technological development of the country. table 3. competitor evaluation via focus group. features competitor 1 2 3 4 5 6 7 8 9 sum product quality 5 9 8 5 10 6 5 5 6 59 price 6 7 5 6 5 6 6 8 9 58 response time 9 6 9 6 7 9 9 8 7 70 location 8 10 9 8 7 8 10 5 7 72 customer experience (service, location, ease of payment) 8 7 8 6 7 9 7 6 6 64 product variety 7 10 10 10 7 7 6 8 8 73 returnability 7 8 5 6 5 9 10 9 6 65 order visibility 8 9 7 7 8 9 10 9 5 72 int. j. prod. manag. eng. (2022) 10(2), 211-223 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho de sá et al. 220 http://creativecommons.org/licenses/by-nc-nd/4.0/ in goiás, decree no. 9724 of october 7, 2020 decided for the adhesion of the state of goiás to the tax benefits provided in the legislation of the state of mato grosso do sul and also established procedures for the operationalization of these benefits. this legislation is part of the regional development program progoiás. the beneficiaries of progoiás can be the establishments that carry out industrial activities in goiás, that are framed in the referred program and that make investments corresponding to: i implantation of a new industrial establishment; ii expansion of an existing industrial establishment; and iii revitalization of a paralyzed industrial establishment. progoiás’ incentives vary from 64% to 67%, which are applicable over the positive value resulting from the confrontation between the debits and credits of the tax on operations related to the circulation of merchandise and the rendering of interstate and intercity transportation and communication services (imposto sobre operações relativas à circulação de mercadorias e sobre prestações de serviços de transporte interestadual e intermunicipal e de comunicação) known, in brazil, as icms. with this, facilities set up in municipalities neighboring the capital city can benefit from cheaper real estate costs and take advantage of the aforementioned incentives. thus, it may be a tradeoff between transportation cost and fixed cost. compared to its neighbors, the capital may face higher real estate rental or acquisition costs, but it will have lower transportation costs. 4.6. regional demand about 50 potential clients, among dental and medical offices, in the regions indicated by the organization were prospected. among these, 47 were interviewed about the usability of printed prostheses in their patients in order to estimate a demand for such impressions. the individual demands resulting from the interviews were estimated by the number of patients who consulted in their offices and whose diagnosis indicated the need for a prosthesis. the demands of each medical office were aggregated considering the 12 regions identified in figure 2, whose results are summarized in table 4. the second column of this table gives the minimum, mode and maximum values. adding the minimum values gives an aggregate minimum demand value. proceeding similarly with the maximum values, it can be stated that the aggregate demand ranges from 146 to 509. 4.7. political factors cds brazil, also known as credit default swap, is a derivative security in the financial market. it works as an insurance, with the objective of avoiding default in credit operations. thus, the cds offers some protection for those seeking to reduce the risks of a credit portfolio. this is done through the issuance of a bond by an insurance company. this institution assumes the function of guaranteeing the redemption of the amount invested in cases of default for any reason. the cds is measured in points. if the cds is high, this means that there is a greater risk that the investor will not receive payment. thus, this number is also often used as a risk indicator in the economy, particularly in international investments. thus, the cds price is directly related to the possibility of a country not paying its debts. therefore, the lower the cds, the more optimism there is about the economy. brazil’s cds between april 2017 and april 2022 is shown in figure 8. for brazil, the cds increased in 2020 probably due to the pandemic situation, since the cds for other regions such as eua, germany, united kingdom, australia had similar peak. since 2020, the cds brazil has fluctuated with the increase of vaccinated table 4. estimated demand for each potential customer subregion. subregion of potential customers monthly demand (min, mode, max) d1 (18,25,48) d2 (15,20,24) d3 (31,35,42) d4 (15,22,30) d5 (0,17,20) d6 (5,40,81) d7 (12,35,49) d8 (15,19,33) d9 (0,8,12) d10 (10,30,66) d11 (15,40,58) d12 (10,30,46) int. j. prod. manag. eng. (2022) 10(2), 211-223creative commons attribution-noncommercial-noderivatives 4.0 international supply chain network design: a case study of the regional facilities analysis for a 3d printing company 221 http://creativecommons.org/licenses/by-nc-nd/4.0/ people, relaxation of isolation measures, and discoveries of new covid-19 variants. however, before 2020, brazil’s cds showed a downward trend. with this, it is expected this trend in the long term. thus, brazil’s economy may be favorable to industry, including 3d printing enterprises. 5. conclusions based on the framework proposed by chopra and meindl, which is composed of four phases and assists in the projection of the physical network of a supply chain, this paper proposes a case study focused on the application of phase ii of the aforementioned framework to assist in the installation of a 3d printing company in the metropolitan region of goiânia. specifically, considering this region, and an interest to install a venture related to a 3d printing service facility aimed at the production of orthopedic and dental prostheses, the following factors were analyzed for the venture: required production technologies, competitive environment, aggregating factor and logistics costs, tariffs and tax incentives, regional demand, political factors, and uncertainty of demand. data collection was carried out by empirical and bibliographic research, observation, and documentation. the results indicate viability of the region for such a venture. in the region studied, there are fiscal incentives of more than 60% for taxes on the movement of goods between municipalities, which can be an advantage in locating facilities outside the capital. the estimated demand ranges from 146 to 509 units per month, which may be an opportunity for a new entrant, given the few additive manufacturing ventures identified in the region. competitors are well qualified, but there is room for new companies focused on quality and price, which may be a case for specialized products, as protheses. finally, job shop is indicated as the type of manufacturing operation, where recruited workers are trained by the company. as future work, phases iii and iv of the framework proposed by chopra and meindl should be developed for the company studied in this paper. this will spell out the logistics costs and infrastructure required and define specific locations that implement the company’s strategic plan. in addition, the final phase will choose a location for a facility. references almeida, a.m. de. (2019). principais barreiras para a implementação da manufatura aditiva no brasil. são paulo, 2019. (master in production engineering), universidade nove de julho, 2019. available at http://bibliotecatede.uninove.br/ handle/tede/2198#preview-link0. access: mar. 2022. amadeo, k. (2021) what are economies of scale?: economies of scale explained. the balance, 24 out. available at https://www.thebalance.com/economies-of-scale-3305926. access: mar. 2022. ballou, r. (2006) gerenciamento da cadeia de suprimentos: logística empresarial. porto alegre: bookman editora. bavaresco, s. (2013) a importância da localização para o sucesso do negócio. curitiba: 2013. (master in production engineering), universidade federal do paraná. available at https://www.acervodigital.ufpr.br/handle/1884/42099. access: mar. 2022. bertalanffy, l. von. (1972) the history and status of general systems theory. the academy of management journal, 15(4), 407–426. berman, b. (2012). 3-d printing: the new industrial revolution. business horizons, 55(2), 155–162. https://doi. org/10.1016/j.bushor.2011.11.003 betim, d.v., teixeira, l.f.h. de s. de b., gomes, c.f.s., quelhas, o.l.g., santos, m. dos. (2019). análise de cenários prospectivos: um estudo sobre a manufatura aditiva no brasil em 2024. in: simpósio de excelência em gestão e tecnologia, 16, out. rio de janeiro. anais do xvi simpósio de excelência em gestão e tecnologia: faculdades dom bosco, 2019. available at https://www.aedb.br/seget/arquivos/artigos19/16228143.pdf. access: mar. 2022. figure 8. brazil cds between april 2017 and april 2022. source: http://www.worldgovernmentbonds.com. int. j. prod. manag. eng. (2022) 10(2), 211-223 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho de sá et al. 222 http://bibliotecatede.uninove.br/handle/tede/2198#preview-link0 http://bibliotecatede.uninove.br/handle/tede/2198#preview-link0 https://www.thebalance.com/economies-of-scale-3305926 https://www.acervodigital.ufpr.br/handle/1884/42099 https://doi.org/10.1016/j.bushor.2011.11.003 https://doi.org/10.1016/j.bushor.2011.11.003 https://www.aedb.br/seget/arquivos/artigos19/16228143.pdf http://www.worldgovernmentbonds.com http://creativecommons.org/licenses/by-nc-nd/4.0/ caetano, m.a.b. (2014). economias de escala e escopo na previdência complementar fechada brasileira. texto para discussão. ipea instituto de pesquisa econômica aplicada. brasília: rio de janeiro. available at http://repositorio. ipea.gov.br/bitstream/11058/3017/1/td_1954.pdf. access: mar. 2022. catão, m.a.v. (2004). regime jurídico dos incentivos fiscais. rio de janeiro: renovar. chopra, s., meindl, p. (2011). gestão da cadeia de suprimentos: estratégia, planejamento e operações. 4th ed. são paulo: pearson prentice hall. chopra, s., meindl, p. . (2013) supply chain management: strategy, planning, and operation. 5th ed. england: pearson education. filament2print (2020). como escolher a tecnologia mais apropriada: fdm, sla e sls. filament2print, 3 jun. 2020. disponível em https://filament2print.com/pt/blog/92_tecnologia-fdm-sla-sls-impressao-3d.html. access: 17 fev. 2022. conner, b.p., manogharan, g.p., martof, a.n., rodomsky, l.m., rodomsky, c.m., jordan, d.c., limperos, j.w. (2014). making sense of 3-d printing: creating a map of additive manufacturing products and services. additive manufacturing, united states, 1-4, 64-76. https://doi.org/10.1016/j.addma.2014.08.005 de souza dutra, m.d. (2015). roteamento de veículos na movimentação de recursos entre canteiros de obras. (master in production engineering), universidade federal de minas gerais. available at https://repositorio.ufmg.br/handle/1843/ bubd-a47mnq. access: mar. 2022. de souza dutra, m.d. (2022) avaliação financeira de um projeto de casa inteligente para uma residência no ceará. exacta, 20(1), 176-197. https://doi.org/10.5585/exactaep.2021.17173 dino. (2018). tipos diferentes de demanda exigem estratégias específicas para cada público-alvo. mundo do marketing, 23 abril. available at https://www.mundodomarketing.com.br/noticias-corporativas/conteudo/163828/tipos-diferentesde-demanda-exigem-estrategias-especificas-para-cada-publico-alvo. access: mar. 2022. gasparino, j.h. ballstaedt. (2021) impressoras 3d e seus reflexos tributários. turivius, 25 mar. available at https://www. turivius.com/portal/impressoras-3d-e-seus-reflexos-tributarios/. access: fev. 2022. impressão 3d e o impacto na indústria 4.0. rodojacto, 07 jul. 2020. available at https://transporte.rodojacto.com.br/ impressao-e-o-impacto-na-industria/. access: mar. 2022. lawrence, p.r., lorsch, j.w. (1967). organization and environment. boston: graduate school of business administration, harvard university. mintzberg, h. (2003). criando organizações eficazes. são paulo: atlas. nabais, j.c. (2012). o dever fundamental de pagar impostos: contributo para a compreensão constitucional do estado fiscal contemporâneo. coimbra: editora almedina. nogueira, p.a. (2021). mensuração de desempenho para análise da estratégia competitiva na manufatura aditiva industrial. (master in production engineering), universidade federal de santa maria. available at https://repositorio. ufsm.br/bitstream/handle/1/23727/dis_ppgep_2021_nogueira_plinio.pdf?sequence=1&isallowed=y. access: 7 de março de 2022. oliveira, m.a. (2020). análise pestel de uma empresa: entenda com exemplos! 7 abril. gen.negócios & gestão, available at https://gennegociosegestao.com.br/analise-pest-empresa-entenda-exemplos/#%3a~%3atext%3da. access: mar. 2022. porter, m. (2004). estratégia competitiva: técnicas para análise de indústrias e da concorrência. elsevier, 409 p. rocha, j. (2019). o que é inflação e como ela afeta seus negócios? voitto, 21 nov. available at https://www.voitto.com.br/ blog/artigo/inflacao. access: mar. 2022. ryan, m.j., eyers, d.r., potter, a.t., purvis, l., gosling. j. (2017). 3d printing the future: scenarios for supply chains reviewed. international journal of physical distribution & logistics management, 47(10), 992-1014. https://doi.org/10.1108/ijpdlm-12-2016-0359 silva, c.v.x., trindade, g.b., andrade, l.r.c. de, assis, m.g. de, de souza dutra, m.d. (2022). network design for additive manufacturing enterprise: strategic level. submitted to independent journal of management & production. preprint available at: https://doi.org/10.31224/2314 souza dutra, m.d. (2019) a framework for peak shaving through the coordination of smart homes. 2019 ieee pes innovative smart grid technologies conference-latin america (isgt latin america), p. 1-6. https://doi.org/10.1109/ isgt-la.2019.8895476 veit, d.r. (2018). impactos da manufatura aditiva nos sistemas produtivos e suas repercussões nos critérios competitivos. (phd in production engineering), universidade do vale do rio dos sinos, 2018. available at http://www.repositorio. jesuita.org.br/handle/unisinos/6927. access: mar. 2022. int. j. prod. manag. eng. (2022) 10(2), 211-223creative commons attribution-noncommercial-noderivatives 4.0 international supply chain network design: a case study of the regional facilities analysis for a 3d printing company 223 http://repositorio.ipea.gov.br/bitstream/11058/3017/1/td_1954.pdf http://repositorio.ipea.gov.br/bitstream/11058/3017/1/td_1954.pdf https://filament2print.com/pt/blog/92_tecnologia-fdm-sla-sls-impressao-3d.html https://doi.org/10.1016/j.addma.2014.08.005 https://repositorio.ufmg.br/handle/1843/bubd-a47mnq https://repositorio.ufmg.br/handle/1843/bubd-a47mnq https://doi.org/10.5585/exactaep.2021.17173 https://www.mundodomarketing.com.br/noticias-corporativas/conteudo/163828/tipos-diferentes-de-demand https://www.mundodomarketing.com.br/noticias-corporativas/conteudo/163828/tipos-diferentes-de-demand https://www.turivius.com/portal/impressoras-3d-e-seus-reflexos-tributarios https://www.turivius.com/portal/impressoras-3d-e-seus-reflexos-tributarios https://transporte.rodojacto.com.br/impressao-e-o-impacto-na-industria https://transporte.rodojacto.com.br/impressao-e-o-impacto-na-industria https://repositorio.ufsm.br/bitstream/handle/1/23727/dis_ppgep_2021_nogueira_plinio.pdf?sequence=1& https://repositorio.ufsm.br/bitstream/handle/1/23727/dis_ppgep_2021_nogueira_plinio.pdf?sequence=1& https://gennegociosegestao.com.br/analise-pest-empresa-entenda-exemplos/#%3a~%3atext%3da https://doi.org/10.1108/ijpdlm-12-2016-0359 https://doi.org/10.31224/2314 https://doi.org/10.1109/isgt-la.2019.8895476 https://doi.org/10.1109/isgt-la.2019.8895476 http://www.repositorio.jesuita.org.br/handle/unisinos/6927 http://www.repositorio.jesuita.org.br/handle/unisinos/6927 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j international journal of production management and engineering survival function of a power transformer and a switch by means of non-parametric estimators aparicio ruiz, p., rodríguez palero, m., & onieva giménez, l. dpto. de organización industrial y gestión de empresas ii, escuela superior de ingeniería avda. de los descubrimientos s/n, isla de la cartuja, 41092 sevilla spain pabloaparicio@us.es mariarodriguez@us.es onieva@us.es abstract: the aim of this work is the estimation of the survival function of a power transformer and a switch for a medium voltage substation, which provides the empirical reliability of this transformer and its switch. the statistical analyses of recurrent events are used for this estimate. in this study, have been applied several estimators: in the presence of correlation, under a model of maximum likelihood and assuming a gamma frailty model. this work is part of a project applying various techniques of maintenance. these techniques are based on the reliability of the electrical substations belonging to future smart grid. key words: survival function, non-parametric estimators, transformer. 1. introduction nowadays, the power distribution companies are in a changing and fiercely competitive scenario. the traditional energy distribution will change in a short space of time towards a new paradigm. the situation which for years has been the policy framework of the electricity distributors, based on regulated monopolies typically directed by a major producer and countless captive customers. it is expected to disappear soon. the success of the current distribution is to adapt to the new situation where there will be many producers, smaller, and also a host of customers, with much higher requirements for quality of service and performance. in this new energy framework, the traditional distribution networks start to become obsolete. the energy distribution must become more pro-active, and must develop and use new tools, new concepts and new responses, to maintain and improve service quality by developing and incorporating new technologies and techniques. one of the aspects to be included in this new situation is the new maintenance management techniques. one of the aspects to be included in this new situation is the new maintenance management techniques. the standard une-en 13306:2011 define maintenance as the combination of all technical, administrative and managerial actions during the life cycle, these are realised by means of maintenance planning, control and supervision of maintenance, improvement of methods in the organization including economics aspects. one of the keys to improving electric service is to increase the continuity of electric service, by carrying out a continuous supply in time, and even wiping out or minimizing interruptions of supply to end customers. at the same time, due to this aim, electrical substations should increase their reliability indices. given the powerful network of generation and distribution of electric utilities networks, actions must be carried out globally increase the reliability of all components of the network. this work is included in one of largest whose purpose is to evaluate the reliability of electrical substation medium voltage distribution through the application of rcm (reliability centered maintenance). this document describes the work that has been developed for estimating the reliability function of a power transformer and its switch, http://dx.doi.org/10.4995/ijpme.2013.1517 received: 2013-05-16 accepted: 2013-05-31 https://ojs.upv.es/index.php/ijpme 3int. j. prod. manag. eng. vol. 01 (2013): 3-11creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2013.1517 considered critical element in substations via a nonparametric estimation. the goal is to plan and carry out maintenance plans to improve the reliability obtained. 2. antecedent as mentioned above, this work is included within a major project. the development of such work has been performed to identify the functional blocks in which a substation may be divided. once these components are identified, an analysis of the different failure modes that can occur in each element. for the determination of elements and critical failures, has done a failure modes and effects analysis (fmea). this methodology looks set for each equipment, the failure mode, its causes and effects. in the beginning of this paper, this study was performed once identified the critical elements of the substation. once identified failure modes, the study was completed with a preliminary hazard analysis (pha). following these studies identified the critical elements in the functioning of substations: the power transformer and switch. just as the most common failure modes of these critical elements. in the reliability analysis of transformers, it is interesting to know the independence or dependence of the parameters that indicate the operation of the transformer. throughout the development of the works have been collected different control parameters of transformers under study. the parameters that were collected for the transformers are shown in table 1. in this table, it also indicates the abbreviations used to identify this parameters. as seen below, the analysis of the reliability of the transformers was performed for different levels of one of these parameters. to study the dependence between variables, were monitored and were calculated the correlation matrix for two specific computers. analyzing the values collected by two transformers, in two medium voltage outdoor electrical substations. the correlation matrix is obtained from the analysis of time series, that are obtained from the collected control signals. the data set must be uniform with respect to the time interval to which they belong. in this way, we can know the evolution of the parameters at the same operating conditions. for the realization of the matrix of each equipment, have been analized the time series in the significant parameters that have been collected in the project’s development. in table 2, is shown graphically by means a color code, the level of correlations between the four measure taken for the power transformer. the darker color in the cell represents the higher correlation between the measures adopted. in the same way, on the right-hand side of the table 2 is shown in detail the interval for each color level. for example, for set variables mt2 and mt3, the magnitude of correlation index between both, is defined in the positive range between 0.80 and 0.90, where this interval is said to be right-open. that is, the limit value range on the right does not belong to it. it is noteworthy that the highest correlation occurs between the oil temperature (mt1) with other input parameters. it can be seen that there is a high degree of correlation, positive in this case, of all actions taken and the indicated (mt1). all this suggests a high degree of relationship between them. it can also be noted that the lower correlation occurs for the air temperature (mt2) relative to the humidity (mt3) and gases in the oil in the power transformer (mt4). this indicates that an increase or decrease in this parameter has an effect of the same sign, in case of positive correlation, but to a lesser extent than produced by a proportional effect. for the election of parameters used in the reliability analysis was used the parameter greater correlation with others, because it is understood that this will be 4 int. j. prod. manag. eng. vol. 01 (2013): 3-11 aparicio ruiz, p., rodríguez palero, m., & onieva giménez, l. creative commons attribution-noncommercial 3.0 spain table 1. control parameters used in the power transformer. code description mt1 oil temperature mt2 air temperature mt3 humidity in the oil mt4 presence of gases in the oil table 2. correlation matrix. mt1 mt1 mt2 mt2 mt3 mt3 mt4 mt4 interpretation code [0.90,1.00) (-1,-0.90] [0.80,0.90) (-0.90,-0.80] [0.60,0.80) (-0.80,-0.60] [0.00,0.60) (-0.60,0.00] the one that best characterizes the state of the power transformer and therefore the most significant to estimate the reliability thereof. in the case of the switch, the parameters considered are shown in table 3. analogously, was studied the correlation matrix of the parameters used in the switch case, the results obtaining are shown in table 4. in this are shown the correlations for measures relating to the switch. the interpretation of the color code is similar to the previous. the first column, corresponding to the first measure, presents the highest correlation with other measured parameters. as can be seen, the measure is identified by the code mi1, reflected in the table with the darker color, is humidity in sf6. this suggests, that the measure is the more effect it has on the other, so choose it is convenient to characterize the state of the switch. described below the methodology used to esti-mate the reliability function of the power transformer in response to most frequently identified failure as the failure mode 1 and mode 2 failures and the reliability function of the switch, in this case for a single failure mode, identified as mode 3. the aim of this comparison is to identify the faults that have more influence on the values of system reliability. the models obtained can also be used to describe the occurrence hazard of an event of interest. 3. methodology the reliability analysis of a system includes the concepts, tools and techniques, which study the time of occurrence until an event occurs. these studies are currently widely applied in scientific research in various fields such as health (gonzález and peña, 2004) or the financial markets (fuentelsalz, 2004). in the case at issue in this work, assuming as initial time instant immediately following the first recorded failure, consider n types of modes of transformer failure, which are considered independent of each other, is called by the variable tij randomly while operating the transformer to the jth failure for the ith failure mode. this random variable is a function of probability of failure is unknown, which is defined according to equation (1). ( ) ( )f t p t tij #= (1) another way of expressing the same distribution is through the reliability function, r(t). which determines the cumulative probability that the event under study occurs after time t. this dis-tribution and the relationship with the above equations are defined by (2) and (3). ( ) ( )r t p t tij $= (2) ( ) ( )r t f t1= (3) the failure probability function, f(t), can be expressed also by their failure density function f(t) as shown in (4) ( ) ( )f t f u du t 0 = # (4) this density function can also relate to the previous equations according to the expression (5), (fuentelsaz et al., 2004). ( ) limf t ( ) ( ) ( ) t t p t t t t dt df t dt ds t 0 ij = + = = " 1# d d d+ a k (5) another feature of interest is the cumulative hazard function, h(t), given in definition of the above function as expressed in equations (6) and (7), (fuentelsaz et al., 2004). ( ) ( )h t h u du t 0 = # (6) 5int. j. prod. manag. eng. vol. 01 (2013): 3-11 survival function of a power transformer and a switch by means of non-parametric estimators creative commons attribution-noncommercial 3.0 spain table 3. the control parameters used in the switch. code description mi1 humidity in sf6 mi2 number of operations in servomotor mi3 wear poles mi4 number of operations table 4. correlation matrix in the parameters list of switch. mi1 mi1 mi2 mi2 mi3 mi3 mi4 mi4 interpretation code [0.90,1.00) (-1,-0.90] [0.80,0.90) (-0.90,-0.80] [0.60,0.80) (-0.80,-0.60] [0.00,0.60) (-0.60,0.00] where h(t) is the failure rate function, ( ) ( ( ) ( ) limr t t p t t t t s t f t t ij 0 1# d d = + + ="d c m (7) in cases in which can be seen in more than one occasion an occurrence of interest, in each individual or system; specific techniques to estimate for recurring events should be used. techniques for reliability analysis differ in the event that the behavior of the variables studied, regarding covariates, follow a known probability distribu-tion or not. in the case where the variable is not following according to a known probability distribution, is used nonparametric estimates. for this reason, the estimation of the survival function through traditional methods, such as the product limit estimator of kaplan-meier (1958) or other more modern, ceases to be useful because it operates with unique events, besides of independence in the occurrence of events. in the present paper, we perform a non-parametric estimation of the time between the occurrence of faults for a power transformer, assuming that there are recurring events and interdependence between them. the variable of interest is the random variable of time between two instants when the power transformer chang-es its state from available to unavailable associ-ated with each failure mode. it is intends to conduct an estimations of transformer reliability function. this problem has been approached by wang and chang (1999) and peña et al. (2001), which present different estimators for the reliability function for recurring events. to determine the reliability of the power trans-former and the switch have been used estimators previously discussed, which are described in more detail in the following sections. 3.1. wc estimator wang and chang (1999) propose an estimator for the case of recurrent events in the presence of correlation, for the particular case that the observations are independent and identically distributed. the authors present a model of fragility where the estimator for the reliability function is given by the expression (8), ( )r t 1, *( ) *( ) t t t r tj d t j j t j = -! # t c m% (8) where d*(t) is the sum of the proportions of the devices in which the inter-occurrence times are equal to t, when there is at least one fault, and r*(t) represents the average of devices that are in risk at instant t. 3.2. psh estimators in peña et al. (2001) are presented two possible estimators. the first is an estimator non-parametric maximum likelihood for a model with variables independent and identically dis-tributed, called iidple (independent and iden-tically product limit estimator). this estimator is an extension of the product limit estimator for recurring events through counters processes, and is expressed through two functions doubly indexed to time scales: calendar time, s, and inter-occurrence time, t. iidple expression of the estimator is presented in (9),infrastructure ( )r t 1 ( , ) ( , ) w t y s w n s w = # dt c m% (9) where n(s,t) represents the observed events number that occur in the time interval [0, s], where interocurrence times do not exceed the time unit, t, and y(s,t) represents the number of observed events in the calendar time [0, s], where interocurrencia times are at least, t, time units. the second estimator proposed by the same authors is used to determine the distribution of the times of occurrence when times are correlated according to a gamma frailty model with shape and scale parameters equal to α and unknown. the proposed estimator is presented in expression (10), \ ( , ) ( , ) s s t h s t0\ \ = + t t t t t; e (10) where \t and ( , )h s t0t are estimators of the scale parameter, α, and the marginal accumulative risk function of h0(t), respectively. 4. reliability assessment to perform the desired estimates, it is necessary to have a database that is sufficiently extensive and debugged, which affect a greater accuracy in the results. to estimate the reliability of the power transformer, were analyzed the faults in one of the substation transformers involved in the project, during the development of this activity. 6 int. j. prod. manag. eng. vol. 01 (2013): 3-11 aparicio ruiz, p., rodríguez palero, m., & onieva giménez, l. creative commons attribution-noncommercial 3.0 spain for each faults have been analyzed times of calendar and intercurrence between successive failure modes, as well as those times between failures categorized with the same failure mode. for the power transformer we have been analyzed the two more frecuently failure modes. these modes are failure mode 1 and failure mode 2. for each failure occurred in the transformer under study, we also analyzed a control parameter. the parameter collects oil temperature at the last moment just before the occurrence of the fault. for this parameter have been defined three levels of temperature level i, ii and iii, of low to high temperature. level i represents the range closest to the right operational level, but this is higher than the suitable operating range. the tables shown (table 5 and 6) the characteristic descriptive statistics of the observations collected for the two failure modes analyzed in the transformer corresponding temperature units (c.t.u.). as shown in table 5, over 70% of the observations collected for the failure mode 1 correspond to the first temperature level. for these observations, 50% of the data presented a value below than 6 c.t.u. and 75% lower than 11 c.t.u., as indicated by the corresponding values of the median and interquartile columns of the first row. the following values in the row make reference to the skewness and kurtosis of the distributions. according to the data, the distribution is skewed to the right and is more pointed than a normal distribution given the value corresponding skewness and kurtosis. regarding the temperature level ii, the population of data that is available represent 20% of the population, being 53.5 c.t.u. the value which leaves 50% of the population to the left. furthermore, 75% of the population, takes values below 70 c.t.u., in this case, the distribution is symmetrical. finally, the observations for the third temperature range are only 6% of the population. for these observations, the 50% of the observations have measured values below 182.50 c.t.u. and 75% below 207.75 c.t.u. according to observations made to the failure mode 2, 47% correspond to the first level of temperature, as is shown in the frequency columns. in this case, the median of the population corresponds to a level of 12.5 c.t.u. and third quartile to 19.5 u.t.c. a similar analysis can be done to the population values corresponding to the temerature levels ii and iii. as shown, the distribution in the three cases are asymmetric to the right and are much more pointed than a normal distribution. in the case of the switch, a similar analysis was performed. were collected failure modes, and were analyzed the most common failure mode, which was identified as failure mode 3. in this case the monitored parameter which has been used is the humidity level in sf6. to characterize this data was divided into three humidity levels. regarding the recorded values are considered correct from the operational level to the most extreme value, which was divided into three intervals. by analogy with the above treatment, is called level i, ii and iii in an increasing order. in table 7 are shown the descriptive statistics of the population of data collected for the switch. in this case, over 60% of the observations correspond to the first level of humidity. while the rest of the population was evenly divided between levels ii and iii. as in previous cases, once the data have been sorted, it was observed that the distribution proved to be skewed to the right and slightly more pointed than normal. in all cases, we examined whether the available data were censored or not. there was a censored failure 7int. j. prod. manag. eng. vol. 01 (2013): 3-11 survival function of a power transformer and a switch by means of non-parametric estimators creative commons attribution-noncommercial 3.0 spain table 5. the statistical descriptions for the failure mode 1. fault 1. type frequency % median interquartile skewness kurtosis level i 0.724138 72.41% 6.00 11.00 1.711362 2.016715 level ii 0.206897 20.69% 53.50 70.00 0.000000 -0.867089 level iii 0.068966 6.90% 182.50 207.75 nc nc table 6. the statistical descriptions for the failure mode 2. fault 2. type frequency % median interquartile skewness kurtosis level i 0.476190 47.62% 12.50 19.25 2.961455 9.059048 level ii 0.285714 28.57% 23.00 40.25 2.138618 4.658857 level iii 0.238095 23.81% 119.00 191.00 2.099625 4.477970 rate of 12 percent compared to the total available data. in addition, has been tested the goodness of fit of the failure times, which allowed to test the suitability of non-parametric techniques, where the null hypothesis of the law that follows the random variable, tj, is known. at least three distributions were used: normal, exponential and weibull. in reference to the failure occurrence times, in the failure mode 1, 29 observations were analyzed, over a range of values between 0 and 233 u.t. (units of time). from these observations, 3 were censored from the right, not existing in any case, left censorship. adjustments were made in two distributions, exponential and normal. the weibull distribution can not be used, because exists one observation with 0 value. after the adjustments, the kolmogorov-smirnov test was performed. table 8 are shown p-values obtained. p-values less than 0.05 would suggest that the analyzed data are not from the selected distributions with 95% confidence. as can be seen, these values are in the range that allows us to reject the hypothesis for adjusting the selected distributions. in reference to the failure occurrence times, in the failure mode 2, 21 observations were analyzed, over a range of values between 1 and 686 u.t. (units of time). from these observations, 3 were censored from the right, not existing in any case, left censorship adjustments were made in to three distributions, exponential, normal and weibull. as seen in table 8, these values are in the range that allows reject the adjustment hypotheses for the exponential distribution. normal and weibull distributions have a p-value greater than the rejection value. however, both distributions are so different from the point of view of inference. for this reason, one can conclude that the results do not present the quality needed to say that the general population decreases according to these distributions. for observations in the switch, identified as failure mode 3. 11 observations were analyzed, of which three were censored. adjustment was made to the same distributions as in the previous case and similarly was performed kolmogorov-smirnov test. as can be seen, the p-values are in the range that does not allow to reject the null hypothesis. we conclude that the observations can be set at three selected distributions. this is because the number of elements of the population can not discern the correct distribution for adjustment, can adjust the three distributions shown large differences between them. in this case, it is equally correct to use nonparametric methods which do not require the choice of the distribution followed by variable analyzed. therefore, the results obtained indicate the adequacy of the proposed techniques for both equipment analyzed. after completing the estimates, the empirical reliability of the transformer has been obtained. the three estimators have been applied for two most common failure modes occurred at different levels of the control parameter used; the temperature of transformer oil (failure mode 1 and 2). the results are presented through graphs for the two types of failure and the three temperature levels defined, see figures 1 to 6. first, the results obtained are presented to the power transformer, with reference to the most common failure mode, called failure mode 1, as the first two levels of temperature (figure 1 and 2). for the most extreme level of temperature, level iii, 2 observations were analyzed, and one of these was censored. therefore, the graphs obtained correspond to a single point, as shown in figure 3. in figures 3, 4, 5 and 6 are shown the results for the failure mode 2, in the three temperature levels. also shown are the results obtained by temperature level for both failure modes together, see figures 7, 8 and 9. figure 10 shows the result obtained for the failure mode analyzed, in the case of the switch. called as 8 int. j. prod. manag. eng. vol. 01 (2013): 3-11 aparicio ruiz, p., rodríguez palero, m., & onieva giménez, l. creative commons attribution-noncommercial 3.0 spain table 7. the statistical descriptions for the failure mode 3. fault 3. type frequency % median interquartile skewness kurtosis level i 0.636364 63.64% 6.00 39.00 1.572202 1.367747 level ii 0.181818 18.18% 46.00 67.50 nc nc level iii 0.181818 18.18% 202.00 213.50 nc nc table 8. p-values obtained using the kolmogorovsmirnov test. distribution exponential normal weibull failure mode i 0,0072243 0,0263075 not applicable failure mode ii 0,0128391 0,0709869 0,890452 failure mode iii 0,1532750 0,4530200 0,865070 9int. j. prod. manag. eng. vol. 01 (2013): 3-11 survival function of a power transformer and a switch by means of non-parametric estimators creative commons attribution-noncommercial 3.0 spain authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 9 population can not discern the correct distribution for adjustment, can adjust the three distributions shown large differences between them. in this case, it is equally correct to use nonparametric methods which do not require the choice of the distribution followed by variable analyzed. therefore, the results obtained indicate the adequacy of the proposed techniques for both equipment analyzed. after completing the estimates, the empirical reliability of the transformer has been obtained. the three estimators have been applied for two most common failure modes occurred at different levels of the control parameter used; the temperature of transformer oil (failure mode 1 and 2). the results are presented through graphs for the two types of failure and the three temperature levels defined, see figures 1 to 6. first, the results obtained are presented to the power transformer, with reference to the most common failure mode, called failure mode 1, as the first two levels of temperature (figure 1 and 2). for the most extreme level of temperature, level iii, 2 observations were analyzed, and one of these was censored. therefore, the graphs obtained correspond to a single point, as shown in figure 3. in figures 3, 4, 5 and 6 are shown the results for the failure mode 2, in the three temperature levels. also shown are the results obtained by temperature level for both failure modes together, see figures 7, 8 and 9. figure 10 shows the result obtained for the failure mode analyzed, in the case of the switch. called as failure mode 3, for humidity level i. for humidity levels ii and iii, the number of events uncensored, is unique in both levels, therefore, the graphs correspond to a single point in space to each estimator (figures 11 and 12). figure 1. empirical reliability of transformer failure mode 1. temperature level i figure 2. empirical reliability of transformer failure mode 1. temperature level ii figure 3. empirical reliability of transformer failure mode 1. temperature level iii figure 4. empirical reliability of transformer failure mode 2. temperature level i 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 10,00 20,00 30,00 40,00 50,00 r(t) time  (ut) failure  1  –  temperature  level  i estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 20,00 40,00 60,00 80,00 r(t) time  (ut) failure  1  -­‐  temperature  level  ii estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,0 0 100,00 150,0 0 200,00 250,0 0 r(t) time  (ut)   failure  1  -­‐  temperature  level  iii estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 10,00 20,00 30,00 r(t) time  (ut)( failure  2  -­‐  temperature  level  i estimator  1 estimator  2 estiamtor  3 figure 1. empirical reliability of transformer failure mode 1. temperature level i. authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 10 figure 5. empirical reliability of transformer failure mode 2. temperature level ii figure 6. empirical reliability of transformer failure mode 2. temperature level iii figure 7. empirical reliability of transformer. temperature level i figure 8. empirical reliability of transformer. temperature level ii figure 9. empirical reliability of transformer. temperature level iii   0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  i estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  i   estimator  1   estimator  2   estimator  3   time  (ut)   figure 10. empirical reliability of switch. humidity level i   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 50,00 100,00 150,00 200,00 250,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  ii estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  ii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 11. empirical reliability of switch. humidity level ii   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 1,00 2,00 3,00 4,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  iii estimador  1 estimador  2 estimador  3 failure  iii.  humidity  level  iii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 12. empirical reliability of switch. humidity level iii 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 50,00 100,00 150,00 200,0 0 r(t) time  (ut) failure  2  –  temperature  level  ii estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,0 0 150,00 200,00 250,00 r(t) time  (ut) failure  2.  temperature  level  iii estimator  2 estimator  3 estimator  1 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) time  (ut) temperature  level  i estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  ii estimator  1 estimator  2 estimator  3 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  iii estimator  1 estimator  2 estimator  3 figure 5. empirical reliability of transformer failure mode 2. temperature level ii authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 9 population can not discern the correct distribution for adjustment, can adjust the three distributions shown large differences between them. in this case, it is equally correct to use nonparametric methods which do not require the choice of the distribution followed by variable analyzed. therefore, the results obtained indicate the adequacy of the proposed techniques for both equipment analyzed. after completing the estimates, the empirical reliability of the transformer has been obtained. the three estimators have been applied for two most common failure modes occurred at different levels of the control parameter used; the temperature of transformer oil (failure mode 1 and 2). the results are presented through graphs for the two types of failure and the three temperature levels defined, see figures 1 to 6. first, the results obtained are presented to the power transformer, with reference to the most common failure mode, called failure mode 1, as the first two levels of temperature (figure 1 and 2). for the most extreme level of temperature, level iii, 2 observations were analyzed, and one of these was censored. therefore, the graphs obtained correspond to a single point, as shown in figure 3. in figures 3, 4, 5 and 6 are shown the results for the failure mode 2, in the three temperature levels. also shown are the results obtained by temperature level for both failure modes together, see figures 7, 8 and 9. figure 10 shows the result obtained for the failure mode analyzed, in the case of the switch. called as failure mode 3, for humidity level i. for humidity levels ii and iii, the number of events uncensored, is unique in both levels, therefore, the graphs correspond to a single point in space to each estimator (figures 11 and 12). figure 1. empirical reliability of transformer failure mode 1. temperature level i figure 2. empirical reliability of transformer failure mode 1. temperature level ii figure 3. empirical reliability of transformer failure mode 1. temperature level iii figure 4. empirical reliability of transformer failure mode 2. temperature level i 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 10,00 20,00 30,00 40,00 50,00 r(t) time  (ut) failure  1  –  temperature  level  i estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 20,00 40,00 60,00 80,00 r(t) time  (ut) failure  1  -­‐  temperature  level  ii estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,0 0 100,00 150,0 0 200,00 250,0 0 r(t) time  (ut)   failure  1  -­‐  temperature  level  iii estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 10,00 20,00 30,00 r(t) time  (ut)( failure  2  -­‐  temperature  level  i estimator  1 estimator  2 estiamtor  3 figure 2. empirical reliability of transformer failure mode 1. temperature level ii. authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 10 figure 5. empirical reliability of transformer failure mode 2. temperature level ii figure 6. empirical reliability of transformer failure mode 2. temperature level iii figure 7. empirical reliability of transformer. temperature level i figure 8. empirical reliability of transformer. temperature level ii figure 9. empirical reliability of transformer. temperature level iii   0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  i estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  i   estimator  1   estimator  2   estimator  3   time  (ut)   figure 10. empirical reliability of switch. humidity level i   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 50,00 100,00 150,00 200,00 250,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  ii estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  ii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 11. empirical reliability of switch. humidity level ii   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 1,00 2,00 3,00 4,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  iii estimador  1 estimador  2 estimador  3 failure  iii.  humidity  level  iii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 12. empirical reliability of switch. humidity level iii 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 50,00 100,00 150,00 200,0 0 r(t) time  (ut) failure  2  –  temperature  level  ii estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,0 0 150,00 200,00 250,00 r(t) time  (ut) failure  2.  temperature  level  iii estimator  2 estimator  3 estimator  1 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) time  (ut) temperature  level  i estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  ii estimator  1 estimator  2 estimator  3 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  iii estimator  1 estimator  2 estimator  3 figure 6. empirical reliability of transformer failure mode 2. temperature level iii. authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 9 population can not discern the correct distribution for adjustment, can adjust the three distributions shown large differences between them. in this case, it is equally correct to use nonparametric methods which do not require the choice of the distribution followed by variable analyzed. therefore, the results obtained indicate the adequacy of the proposed techniques for both equipment analyzed. after completing the estimates, the empirical reliability of the transformer has been obtained. the three estimators have been applied for two most common failure modes occurred at different levels of the control parameter used; the temperature of transformer oil (failure mode 1 and 2). the results are presented through graphs for the two types of failure and the three temperature levels defined, see figures 1 to 6. first, the results obtained are presented to the power transformer, with reference to the most common failure mode, called failure mode 1, as the first two levels of temperature (figure 1 and 2). for the most extreme level of temperature, level iii, 2 observations were analyzed, and one of these was censored. therefore, the graphs obtained correspond to a single point, as shown in figure 3. in figures 3, 4, 5 and 6 are shown the results for the failure mode 2, in the three temperature levels. also shown are the results obtained by temperature level for both failure modes together, see figures 7, 8 and 9. figure 10 shows the result obtained for the failure mode analyzed, in the case of the switch. called as failure mode 3, for humidity level i. for humidity levels ii and iii, the number of events uncensored, is unique in both levels, therefore, the graphs correspond to a single point in space to each estimator (figures 11 and 12). figure 1. empirical reliability of transformer failure mode 1. temperature level i figure 2. empirical reliability of transformer failure mode 1. temperature level ii figure 3. empirical reliability of transformer failure mode 1. temperature level iii figure 4. empirical reliability of transformer failure mode 2. temperature level i 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 10,00 20,00 30,00 40,00 50,00 r(t) time  (ut) failure  1  –  temperature  level  i estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 20,00 40,00 60,00 80,00 r(t) time  (ut) failure  1  -­‐  temperature  level  ii estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,0 0 100,00 150,0 0 200,00 250,0 0 r(t) time  (ut)   failure  1  -­‐  temperature  level  iii estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 10,00 20,00 30,00 r(t) time  (ut)( failure  2  -­‐  temperature  level  i estimator  1 estimator  2 estiamtor  3 figure 3. empirical reliability of transformer failure mode 1. temperature level iii. authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 10 figure 5. empirical reliability of transformer failure mode 2. temperature level ii figure 6. empirical reliability of transformer failure mode 2. temperature level iii figure 7. empirical reliability of transformer. temperature level i figure 8. empirical reliability of transformer. temperature level ii figure 9. empirical reliability of transformer. temperature level iii   0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  i estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  i   estimator  1   estimator  2   estimator  3   time  (ut)   figure 10. empirical reliability of switch. humidity level i   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 50,00 100,00 150,00 200,00 250,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  ii estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  ii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 11. empirical reliability of switch. humidity level ii   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 1,00 2,00 3,00 4,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  iii estimador  1 estimador  2 estimador  3 failure  iii.  humidity  level  iii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 12. empirical reliability of switch. humidity level iii 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 50,00 100,00 150,00 200,0 0 r(t) time  (ut) failure  2  –  temperature  level  ii estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,0 0 150,00 200,00 250,00 r(t) time  (ut) failure  2.  temperature  level  iii estimator  2 estimator  3 estimator  1 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) time  (ut) temperature  level  i estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  ii estimator  1 estimator  2 estimator  3 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  iii estimator  1 estimator  2 estimator  3 figure 7. empirical reliability of transformer. temperature level i. authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 9 population can not discern the correct distribution for adjustment, can adjust the three distributions shown large differences between them. in this case, it is equally correct to use nonparametric methods which do not require the choice of the distribution followed by variable analyzed. therefore, the results obtained indicate the adequacy of the proposed techniques for both equipment analyzed. after completing the estimates, the empirical reliability of the transformer has been obtained. the three estimators have been applied for two most common failure modes occurred at different levels of the control parameter used; the temperature of transformer oil (failure mode 1 and 2). the results are presented through graphs for the two types of failure and the three temperature levels defined, see figures 1 to 6. first, the results obtained are presented to the power transformer, with reference to the most common failure mode, called failure mode 1, as the first two levels of temperature (figure 1 and 2). for the most extreme level of temperature, level iii, 2 observations were analyzed, and one of these was censored. therefore, the graphs obtained correspond to a single point, as shown in figure 3. in figures 3, 4, 5 and 6 are shown the results for the failure mode 2, in the three temperature levels. also shown are the results obtained by temperature level for both failure modes together, see figures 7, 8 and 9. figure 10 shows the result obtained for the failure mode analyzed, in the case of the switch. called as failure mode 3, for humidity level i. for humidity levels ii and iii, the number of events uncensored, is unique in both levels, therefore, the graphs correspond to a single point in space to each estimator (figures 11 and 12). figure 1. empirical reliability of transformer failure mode 1. temperature level i figure 2. empirical reliability of transformer failure mode 1. temperature level ii figure 3. empirical reliability of transformer failure mode 1. temperature level iii figure 4. empirical reliability of transformer failure mode 2. temperature level i 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 10,00 20,00 30,00 40,00 50,00 r(t) time  (ut) failure  1  –  temperature  level  i estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 20,00 40,00 60,00 80,00 r(t) time  (ut) failure  1  -­‐  temperature  level  ii estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,0 0 100,00 150,0 0 200,00 250,0 0 r(t) time  (ut)   failure  1  -­‐  temperature  level  iii estimator  1 estimator  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 10,00 20,00 30,00 r(t) time  (ut)( failure  2  -­‐  temperature  level  i estimator  1 estimator  2 estiamtor  3 figure 4. empirical reliability of transformer failure mode 2. temperature level i. authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 10 figure 5. empirical reliability of transformer failure mode 2. temperature level ii figure 6. empirical reliability of transformer failure mode 2. temperature level iii figure 7. empirical reliability of transformer. temperature level i figure 8. empirical reliability of transformer. temperature level ii figure 9. empirical reliability of transformer. temperature level iii   0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  i estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  i   estimator  1   estimator  2   estimator  3   time  (ut)   figure 10. empirical reliability of switch. humidity level i   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 50,00 100,00 150,00 200,00 250,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  ii estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  ii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 11. empirical reliability of switch. humidity level ii   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 1,00 2,00 3,00 4,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  iii estimador  1 estimador  2 estimador  3 failure  iii.  humidity  level  iii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 12. empirical reliability of switch. humidity level iii 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 50,00 100,00 150,00 200,0 0 r(t) time  (ut) failure  2  –  temperature  level  ii estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,0 0 150,00 200,00 250,00 r(t) time  (ut) failure  2.  temperature  level  iii estimator  2 estimator  3 estimator  1 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) time  (ut) temperature  level  i estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  ii estimator  1 estimator  2 estimator  3 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  iii estimator  1 estimator  2 estimator  3 figure 8. empirical reliability of transformer. temperature level ii. failure mode 3, for humidity level i. for humidity levels ii and iii, the number of events uncensored, is unique in both levels, therefore, the graphs correspond to a single point in space to each estimator (figures 11 and 12). 5. conclusions the results obtained for the transformer, under the same duration conditions, have proved that a lower temperature levels, result in a decrease in the reliability. although a priori can be contradictory, a brief analysis explains this result. when the temperature variable is at level i, above the level of operational performance, it is usual that the failure occurs. it is unusual for the temperature to reach achieve the following levels of temperature, levels ii and iii, without causing the fault, because in that case the equipment would be operating outside the correct operating range for this parameter. in this way, the time between events occurring at higher temperatures are more distant from each other. this explanation also seems to be in line with the fact that are insignificant the failures occurred at the extreme temperature level, level iii. in general, the fault will occur before reaching the temperature extreme level, which would require that the equipment had been operating at a level far from the recommended range. this could also be due to a very sharp increase in temperature occurred for very serious faults, which of course, have a very low frequency of repetition. moreover, considering the results obtained with reference to the three estimators used, can be observed other interesting conclusions. it has been observed that at lower temperature levels, the three estimators used, have presented similar results. this coincidence suggests that events do not depend each other, because the estimator can collect dependence of the events, produces the same result that the two estimators with interdependence. in contrast to the higher temperature levels, where each estimator results are quite different from each other, this is shown clearly in the higher temperature level. this suggests that the events at higher temperature level are interdependent. finally, these results can be concluded that the failures occurred in the equipments are recurring events but interdependent, except when major failures in which there is a relationship of cause and effect, where the concatenated events have eliminated the independence. the results for the switch has been presented for the first level of humidity, level i. the results obtained for 10 int. j. prod. manag. eng. vol. 01 (2013): 3-11 aparicio ruiz, p., rodríguez palero, m., & onieva giménez, l. creative commons attribution-noncommercial 3.0 spain authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 10 figure 5. empirical reliability of transformer failure mode 2. temperature level ii figure 6. empirical reliability of transformer failure mode 2. temperature level iii figure 7. empirical reliability of transformer. temperature level i figure 8. empirical reliability of transformer. temperature level ii figure 9. empirical reliability of transformer. temperature level iii   0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  i estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  i   estimator  1   estimator  2   estimator  3   time  (ut)   figure 10. empirical reliability of switch. humidity level i   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 50,00 100,00 150,00 200,00 250,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  ii estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  ii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 11. empirical reliability of switch. humidity level ii   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 1,00 2,00 3,00 4,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  iii estimador  1 estimador  2 estimador  3 failure  iii.  humidity  level  iii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 12. empirical reliability of switch. humidity level iii 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 50,00 100,00 150,00 200,0 0 r(t) time  (ut) failure  2  –  temperature  level  ii estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,0 0 150,00 200,00 250,00 r(t) time  (ut) failure  2.  temperature  level  iii estimator  2 estimator  3 estimator  1 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) time  (ut) temperature  level  i estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  ii estimator  1 estimator  2 estimator  3 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  iii estimator  1 estimator  2 estimator  3 figure 9. empirical reliability of transformer. temperature level iii. authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 10 figure 5. empirical reliability of transformer failure mode 2. temperature level ii figure 6. empirical reliability of transformer failure mode 2. temperature level iii figure 7. empirical reliability of transformer. temperature level i figure 8. empirical reliability of transformer. temperature level ii figure 9. empirical reliability of transformer. temperature level iii   0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  i estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  i   estimator  1   estimator  2   estimator  3   time  (ut)   figure 10. empirical reliability of switch. humidity level i   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 50,00 100,00 150,00 200,00 250,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  ii estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  ii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 11. empirical reliability of switch. humidity level ii   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 1,00 2,00 3,00 4,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  iii estimador  1 estimador  2 estimador  3 failure  iii.  humidity  level  iii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 12. empirical reliability of switch. humidity level iii 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 50,00 100,00 150,00 200,0 0 r(t) time  (ut) failure  2  –  temperature  level  ii estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,0 0 150,00 200,00 250,00 r(t) time  (ut) failure  2.  temperature  level  iii estimator  2 estimator  3 estimator  1 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) time  (ut) temperature  level  i estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  ii estimator  1 estimator  2 estimator  3 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  iii estimator  1 estimator  2 estimator  3 figure 10. empirical reliability of switch. humidity level i. authors, int. j. prod. manag. eng. vol. vv (yyyy) ppp-ppp. creative commons licence 10 figure 5. empirical reliability of transformer failure mode 2. temperature level ii figure 6. empirical reliability of transformer failure mode 2. temperature level iii figure 7. empirical reliability of transformer. temperature level i figure 8. empirical reliability of transformer. temperature level ii figure 9. empirical reliability of transformer. temperature level iii   0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  i estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  i   estimator  1   estimator  2   estimator  3   time  (ut)   figure 10. empirical reliability of switch. humidity level i   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 50,00 100,00 150,00 200,00 250,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  ii estimador  1 estiamdor  2 estimador  3 failure  iii.  humidity  level  ii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 11. empirical reliability of switch. humidity level ii   0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 1,00 2,00 3,00 4,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  iii estimador  1 estimador  2 estimador  3 failure  iii.  humidity  level  iii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 12. empirical reliability of switch. humidity level iii 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 0,00 50,00 100,00 150,00 200,0 0 r(t) time  (ut) failure  2  –  temperature  level  ii estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,0 0 150,00 200,00 250,00 r(t) time  (ut) failure  2.  temperature  level  iii estimator  2 estimator  3 estimator  1 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 r(t) time  (ut) temperature  level  i estimator  1 estiamtor  2 estimator  3 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  ii estimator  1 estimator  2 estimator  3 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 0,00 50,00 100,00 150,00 200,00 250,00 r(t) time  (ut) temperature  level  iii estimator  1 estimator  2 estimator  3 figure 11. empirical reliability of switch. humidity level ii.     0,00 0,20 0,40 0,60 0,80 1,00 1,20 0,00 1,00 2,00 3,00 4,00 r(t) duración  (ut) fallo  3.  nivel    de  humedad  iii estimador  1 estimador  2 estimador  3 failure  iii.  humidity  level  iii   estimator  1   estimator  2   estimator  3   time  (ut)   figure 12. empirical reliability of switch. humidity level iii. the other two levels, as indicated previously, because it only has been a non-censored observation of each, correspond to points in space. the results obtained at the level i, where the reliability obtained with the third empirical estimator differs substantially from those obtained with the other two estimators. this difference may be caused by the dependence of time in any of the covariates was not considered in the study. examples of possible covariates would be the outside temperature at the time of the failure, the percentage of humidity or some particular feature on the switch type. it would be interesting to do a new analysis, including or discarding some of the control signals. regarding the results for levels ii and iii, note that the expected time to failure with level ii is higher than the level i. this can be explained due to faults with this level of humidity are detected and the equipment is subjected to preventive maintenance. this operation aims to recover the operating condition. consequently, the time between failures increases after this maintenance. for level iii shows a sharp decline in survival time. if the humidity levels are reached, could be occurring severe failures related to each other, occurring in an uncontrolled manner in a small interval of time, without the reaction capacity necessary to perform the maintenance operation for the equipment. it is noteworthy that the results are specific to the equipment studied, because they are directly related to their own operating conditions and environment. this allows improve the design and adaptation of maintenance policies to the operational status of each equipment. if properly combined with policies of condition based maintenance, the reliability expected of assets will increase and decrease the repair costs, replacement or unavailability of equipment. references böhmer, p. e. (1912). theorie de unabhängigen wahrscheinlichkeiten, rapports, mémoires et procès verbaux du septième congrès international d’actuaires. fuentelsaz, l., gómez, j., & polo, y. (2004). aplicaciones del análisis de supervivencia a la investigación en economía de la empresa. cuadernos de economía y dirección de la empresa, 19, 081-114. gonzález, j. r., & peña, e. a. (2004). estimación no paramétrica de la función de supervivencia para datos con eventos recurrentes. revista española de salud pública, 78(2), 189-199. doi:10.1590/s1135-57272004000200006 kaplan, a., & meier, p. (1958). non-parametric estimation for incomplete estimations. journal of the american statistical association, 53(282), 457-481. doi:10.1080/01621459.1958.10501452 peña, e., strawderman, r., & hollander, m. (2001). nonparametric estimation with recurrent event data. journal of the american statistical association, 96(456), 1299-1315. doi:10.1198/016214501753381922 wang, m. c., & chang, s. h. (1999.) nonparametric estimation of a recurrent survival function. journal of the american statistical association, 94(445), 146-153. doi:10.1080/01621459.1999.10473831 11int. j. prod. manag. eng. vol. 01 (2013): 3-11 survival function of a power transformer and a switch by means of non-parametric estimators creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.1590/s1135-57272004000200006 http://dx.doi.org/10.1080/01621459.1958.10501452 http://dx.doi.org/10.1198/016214501753381922 http://dx.doi.org/10.1080/01621459.1999.10473831 pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering received: 2021-12-30 accepted: 2022-06-7 implementation of green, lean and six sigma operations for sustainable manufacturing. a review khalid nagadi university of jeddah, jeddah, makkah, saudi arabia. knagadi@uj.edu.sa abstract: with ever increasing environmental concerns and global warming, green manufacturing has gained momentum to make the manufacturing processes sustainable and efficient. this review aims to analyse the models to integrate three different management systems that are green, six sigma and lean for the sustainability of various manufacturing processes. research gaps for such integration are identified through a literature review of various studies. the importance of the concepts of eco-friendly and sustainability in business operations and practices is growing rapidly as a result of public pressure, government regulations and social responsibilities. the first step is the identification of sustainability assessment for the current industrial processes and then to make them eco-friendly and more efficient through different green, six sigma and lean tools available. the methodology presented in this review will not only help in sustainability but also is helpful in the integration of various models for the improvement of the processes. green lean six sigma (gls) is an approach known to minimize emissions and carbon footprints while improving process efficiency. gls includes green, six sigma and lean methodologies for high performance, sustainability, social development, economic progress and environmental protection. the successful integration of this gls approach is dependent on different theoretical indicators and the model is developed based on dmaic. various tools, enablers and integration methods are employed for the gls approach. key words: green, manufacturing, lean, six sigma, sustainability. 1. introduction poverty, inequality, climate change, population increases, pollution, and the rising cost of energy and raw materials represent the most significant issues humanity faces in the contemporary global context. customers, regulators, and other stakeholders are pressuring businesses throughout the world to run their operations responsibly to enhance the performance and to alter behaviours at both environmental and social levels. due to this, increasing sustainability and operational performance, decreasing the adverse environmental impacts and social effects of these industrial operations has become an essential corporate requirement: sustainability has evolved into new competitive criteria (garza-reyes, 2015). the rising consumer awareness of sustainability and appetite for environmentally friendly products has led businesses to reconsider their value chains and production methods. in today’s competitive economy, manufacturing firms are taking the lead. its goal is to occupy a strong market position and please clients by meeting their needs to maximize profit. manufacturing companies have a significant impact on the economy because they sell goods and services on a worldwide scale. organizational greening has become a growing issue in a range of industries, owing to the competitive and strategic challenges to cite this article: nagadi, k. (2022). implementation of green, lean and six sigma operations for sustainable manufacturing. a review. international journal of production management and engineering, 10(2), 159-171. https://doi.org/10.4995/ijpme.2022.16958 https://doi.org/10.4995/ijpme.2022.16958 int. j. prod. manag. eng. (2022) 10(2), 159-171creative commons attribution-noncommercial-noderivatives 4.0 international 159 http://creativecommons.org/licenses/by-nc-nd/4.0/ that these companies face. the costs and penalties that may arise from poor environmental performance represent tangible concerns. organizations must also handle intangible factors such as image and reputation as strategic priorities (zhu et al., 2018). the ability of contemporary enterprises to evolve alongside the external environment is critical to their survival (kaswan & rathi, 2020a). to remain competitive, industrial companies should develop and adopt technologies for lower carbon emissions (duarte & cruz-machado, 2019). nowadays, businesses are investing a lot of money in developing sustainable production and consumption practices. many concepts and methodologies, such as green, lean, and six sigma, have emerged in recent decades amid part of efforts to generate high-quality goods. sustainability is frequently characterized as the means of achieving a balance between current and future generations’ environmental, economic, and social demands via a strategy that is referred to as a “triple bottom line” approach. while this triple bottom line strategy encompasses all three main areas, aspects related to environmental sustainability are the most discussed after the introduction of sustainability notion of sustainability worldwide. this is certainly the case in the manufacturing context. the phrase “green manufacturing,” which is sometimes used synonymously with “sustainability,” was created to describe manufacturing processes and tactics that are conscious of environmental implications throughout production and operations (erdil et al., 2018). the methodologies of lean and six sigma are becoming increasingly popular in the search for more efficient production practices. lean provides value by pinpointing and removing waste in the manufacturing and distribution process, which improves the flow of the process and lead time. six sigma adds value by detecting and minimizing variance in the process output. lean six sigma (lss) is a next-generation framework and model to improve quality that incorporates both methodologies. lss is associated with a myriad of advantages, including fewer errors and rework, more rapid production, reduced inventory levels, reduced space requirements, less transportation, less downtime, and higher employee engagement. lean six sigma (lss) is used to improve the process and to solve different operational problems for businesses and individuals. companies may gain a competitive edge by using lss, which has been found to reduce lead time by up to 80%, costs by up to 20%, and quality and delivery time improvements of up to 99%. the relentless pursuit of six sigma zero process variation while disregarding the needs of consumers might result in non-optimal use of resources. as a balanced implementation technique, lss meets customer expectations by producing enough worth to sustain and retain market share while also lowering necessary variety and, thereby, minimizing the associated costs. critical success factors (csfs) are some of the fundamental input variables for a successful and effective lss deployment (flor vallejo et al., 2020). as a result, there are substantial similarities in terms of the underlying goals of lss and sustainability. green, lean, and six sigma are methods of increasing profitability by reducing rework, waste, and emissions. at all levels of a company, lean pushes for the methodical reduction of waste via excellence within the value chain. green technology decreases a product’s harmful environmental effects by making it more environmentally friendly. six sigma lowers variability in the process, resulting in fewer product rejections. however, when green and lean six sigma are merged, the resultant approach can lead to the development of a cost-effective product that is both of satisfactory quality and environmentally beneficial (hussain et al., 2019). a research study examined financial data spanning 170 manufacturing organizations to determine the average delivery timeframe and the overall percentage of improvements. the findings revealed that although some businesses can successfully use lss, others fail to do so. practically half of the organizations studied had their average delivery time fall over time, signifying a drop in quality. this was due to the errors made during the shift from theory to practice, rather than a lack of lss expertise. that said, a large proportion of businesses reported an improvement rate of between 100% and 300%. despite this, because there is a lack of a standardized lss roadmap or change in plans, lss deployment is frequently doomed to fail. companies and organizations must employ a plan or roadmap as a guide to ensure a successful lss deployment. this roadmap should define the activities or requirements that must be completed to achieve the required and desired goals. lss roadmaps may be customized for different businesses based on their requirements (baker, 2003). manufacturing companies face multiple challenges in the contemporary world, including low-quality products, excessive production costs, inability to fulfil customer demand, due to demand mismatches, and long delivery times, among others, all of which are caused by a lack of an effective operational plan. int. j. prod. manag. eng. (2022) 10(2), 159-171 creative commons attribution-noncommercial-noderivatives 4.0 international nagadi 160 http://creativecommons.org/licenses/by-nc-nd/4.0/ quality, adaptability, and customer happiness have arisen as competitive factors in more recent years, adding to the traditional criteria of manufacturing efficiency and profitability. according to bergmiller & mccright (2009), organizations must address the three elements of sustainability (economic, social, and environmental) to achieve and maintain a competitive advantage. the challenge for businesses in this environment is to adequately address stakeholder requirements while achieving strong economic performance and striking the correct balance between each element that forms the triple bottom line of sustainability. the possibility of combining lean, six sigma, and sustainability is gaining popularity; numerous academics and business professionals have contributed to the field’s study and development, resulting in over 118 publications. however, despite the significant number of publications, very few literature reviews of the existing literature have been published. only six of the publications we looked at were designed to assess the current level of research into the linkages between lean, six sigma, and sustainability. however, there is no systematic assessment that examines the drivers, challenges, advantages, and critical success factors (csfs) with the underlying objective of developing a potential integrated model. the current state of information about possible synergies and contradictions among the three techniques is nascent. furthermore, the presented frameworks, patterns, and approaches for integrating lean, six sigma, and sustainability have not been investigated (cherrafi et al., 2016; johansson & sundin, 2014). 2. terminologies 2.1. sustainability one of the most commonly used keywords in the last two decades has to be “sustainability.” there appears to be nothing that can’t be defined as “sustainable”: anything can be hyphenated or coupled with it. cities, resource management, careers, businesses, and livelihoods are all discussed within the concept of sustainable development (scoones, 2007). the most commonly used definition for sustainability is a development that satisfies the requirements of current populations while also guaranteeing that future generations are provided for. the term “sustainable development” first emerged in a publication of a study by the united nations world commission on environment and development (wced) in 1987. the notion or term for sustainable development subsequently referred to as “sustainability,” received international recognition and the focus of state representatives throughout the world in the aftermath of the publication of the report (erdil et al., 2018). epa report (kidwell, 2006) contains a comparable definition that emphasizes the importance of preserving a balance between profit, the environment, and people: “sustainability creates and maintains the conditions under which humans and nature can exist in productive harmony, that permit fulfilling the social, economic and other requirements of present and future generations.” if any of the triple bottom line aspects—environment, economy, or society— deteriorate or collapse, systems and organizations will not be able to withstand the test of time. institutions are sustained by the balance of these three elements, not by the supremacy of one over the other two. in the initial tbl research (elkington, 1994), the three pillars of the triple bottom line (tbl), which are these days generally referred to as social impact, environmental concerns, and economic benefits, were characterized as the 3ps: people, planet, and profit. the environmental aspect (planet) is concerned with the activities and practises that are associated with the use of natural resources, consumption of energy, ecological health, and pollution; the social aspect (people) is centred around human needs and spans aspects such as the provision of education, availability of jobs, living standards, and health and safety; the economic aspects examines strategies that promote and enhance economic prosperity and profitability, enhance cost efficiencies, and promote health and safety. sustainability in the manufacturing sector attempts to generate manufactured goods that maximize profitability while minimizing negative environmental consequences, conserving energy and raw materials, and maximizing safety for employees, end users, and communities. in addition, production practices should guarantee that a population’s demands are addressed. as a result, attaining manufacturing sustainability necessitates a holistic approach that considers not just the product and the procedures associated with its creation, but also the end-to-end supply chain and production system (faulkner & badurdeen, 2014). since the focus on sustainability emerged, numerous systems for environmental control, health and safety, and social responsibility management have been established to achieve sustainability. many procedures and guidelines have been established for this purpose to assist businesses in implementing good corporate social responsibility. int. j. prod. manag. eng. (2022) 10(2), 159-171creative commons attribution-noncommercial-noderivatives 4.0 international implementation of green, lean and six sigma operations for sustainable manufacturing. a review 161 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2.2. lean six sigma in the late 1990s and the start of the 2000s, the phrase “lean six sigma” emerged to characterize and explain the relevance and overlapping of lean and six sigma ideas. the goal of this integration was to address the flaws inherent in each approach. businesses were able to achieve a notable increase in improvement by combining the two continuous improvement approaches (byrne et al., 2007). lean six sigma is a corporate strategy and technique that enhances quality, timeliness, and cost to improve process capability and achieve customer happiness, leadership, and bottom-line outcomes. it does this by combining lean and six sigma frameworks and methodologies. many firms worldwide have adopted lean six sigma as one of the most wellknown hybrid continuous improvement approaches to solve their operational challenges and increase their competitiveness (singh & rathi, 2019). lean strives to increase customer satisfaction by reducing waste within the value chain, be it in the form of inventory, movement, motion, rework, waiting, overprocessing, and overproduction. these seven forms of waste are referred to using relatively conventional definitions in the literature. another type of waste that has attracted attention in more recent times is skills. skill wastage arises when people’s abilities, skills, and expertise are not fully leveraged. six sigma is a data-driven technique for eliminating variance in processes. the define, measure, analyze, improve, and control (dmaic) cycle is used in most six sigma deployments. dmaic is a proven paradigm for achieving large improvements in performance by providing a disciplined approach to improvement efforts. improvements made using either strategy have an influence on long-term sustainability (singh & rathi, 2019). 3. embedding sustainability into lean manufacturing and six sigma model 3.1. lean manufacturing and the six sigma model lean six sigma is the widely utilized method for improving processes and ensuring their long-term viability. it is an effective technique that concentrates on four essential aspects: profitability, quality, productivity, and cost (evans & lindsay, 2014). the five-step define, measure, analyze, improve, and command (dmaic) method provides a systematic approach to project management and implementation that covers a wide variety of lss tools in a goalfocused way. dmaic is a tried-and-true method for achieving considerable performance gains. to integrate the model with sustainability, the dmaic framework assists in the absence of implementation strategies. this framework will give access to tools and well-known practices to achieve sustainability goals and pave the path for wider implementation of sustainability principles and goals for the success of businesses. the important goals in the process are attained with the use of lss and its widespread implementation throughout different industries. this integration begins in the define phase with the identification of sustainability prospects that can be linked to an improvement of the project and proceeds with the definition of corresponding benchmarks to allow follow-ups in the later stages to reach the required sustainability performance. table 2 summarises the steps required to apply dmaic to ensure its long-term viability and overall continuous improvements (kaswan & rathi, 2019). figure 1. integration and implementation of gls model (kaswan & rathi, 2020b) int. j. prod. manag. eng. (2022) 10(2), 159-171 creative commons attribution-noncommercial-noderivatives 4.0 international nagadi 162 http://creativecommons.org/licenses/by-nc-nd/4.0/ the most significant step in the integration of sustainability into any form of the improvement process is to establish a link between the firm’s sustainability objectives and the various aims of the concerned project. it is crucial to define the sustainability demands and goals of a company for this to be effective and for the targets to be aligned successfully. as a result, before any firm embarks on an lss project with sustainability features, it must first conduct a sustainability assessment in order to define and prioritize the company’s sustainability aims in terms of environmental, social, and economic factors. this type of evaluation includes the assessment and evaluation of the present sustainability performance of a company, as well as producing documentation about the company’s requirements and objectives to improve teams or groups to use as a guide for creating sustainability goals and kpis for their initiatives (cherrafi et al., 2016). the next section describes an importance-impactimplementation analysis that was created for sustainability evaluation. table 3 compiles a list of the common sustainability indicators described within the literature. this table offers a variety of indicators, which may be changed depending on the goals of the organization (cherrafi et al., 2016, 2017). 3.2. sustainability importance-impactimplementation analysis first, as indicated in tables 2 and 3, sustainability indicators are defined in line with importance, effect, and implementation criteria. the importance of an indicator is determined by its relevance to the firm’s operations. the items that are of direct significance to company operations are considered vital metrics. the significance of indicators is used to assess the anticipated improvements. if these metrics improve, the company will see a considerable increase in profits (miller et al., 2010). finally, to ensure the process’s long-term viability and efficiency, these elements must be implemented. table 4 lists the damic duties, which are detailed below. table 1. tasks for systematic approach using dmaic. tasks define measure analyze improve control development of project charter that includes problem statements, , and total resources establishment of performance met-rics analysis of data in identification of sources for varia-tion and discrepan-cies development and evaluation of solutions verification of implemented improvements mapping the pro-cess validation of measurement sys-tem examination of the process in identification of root caus-es implemention of selected alternatives development of standard operating procedures (sops) project aims and goals, data collection development and implementation of control plans quality needs and requirements determination of process baseline team and group members and associated roles and total available resources table 2. dmaic process for the sustainability of the process. tasks actions define identification of sustainability indicators and clarification of final sustainability goals measure development of sustainability goals matrices analyze analysis the problems in context with sustainability goals improve cost and benefit analysis of the solutions for sustainable goals control verification of the solutions and impact on sustainability goals int. j. prod. manag. eng. (2022) 10(2), 159-171creative commons attribution-noncommercial-noderivatives 4.0 international implementation of green, lean and six sigma operations for sustainable manufacturing. a review 163 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.2.1. define and measure every dmaic-based improvement project begins with the creation of a project charter, which summarizes the background, definition, scope, performance criteria, deliverables, and definition associated with the project. the project charter serves as a road map that guides the project, keeping teams geared towards achieving the project’s objectives. as a result, the define phase is critical for incorporating sustainability goals into continual improvement and project charters. table 3. common sustainability indicators. sustainability indicators economic environmental social consumption analysis control over emissions risk and safety analysis of industrial workers r&d for improving existing processes, technologies and addition of new products waste water treatment for environmentally sustainable process talent acquisition growth analysis use of carbon capturing technologies in industries to reduce emissions education for latest skills financial distribution utilization of energy efficient processes equality and diversity community development table 4. literature review of studies using lean, green and six sigma approaches. sector model contribution short comings automobiles green and lean a model was analyzed to integrate lean and sustainability factors. the model employed kaizen concept for the conversation of energy and to improve the efficiency of production line. application of the model is limited as it is implemented at pilot project. large scale implementation is needed to check the sustainability of the model. semiconductor manufacturing unit green and lean a model and framework were incorporated to indulge human and physical factors in lean operations for the sustainability of the manufacturing unit. economic and environmental factors were not taken into consideration for sustainability. green and six sigma a framework was proposed in order to combine six sigma and sustainable manufacturing. economic factors for sustainability were not included and the model is not implemented for real time manufacturing environment. metal industry green and lean a model was proposed for the analysis for the relationship between environmental factors and lean operations of different manufacturing processes in metal industry. the scope of the model is limited to only few manufacturing processes in metal industries. foundries and casting industries green and lean a framework was analyzed to study the scope of sustainability and lean. the implementation of lean and environmental factors was studied too. the scope of this model is limited to few particular manufacturing and casting industrial processes. automobiles green and lean the study analyzed the supply chain operation management for green and lean operation ability for the sustainable overall performance. the study is only limited to portugal automobile industry and necessarily does not cover the approach for other countries. metal industry green and lean the aim of the study was to implement and integrate green and lean methodology in the manufac-turing. matric system was pro-posed to integrate and implement green and lean operations. the scope of this study is lim-ited to few industries and the study does not have universal approach. int. j. prod. manag. eng. (2022) 10(2), 159-171 creative commons attribution-noncommercial-noderivatives 4.0 international nagadi 164 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. green lean six sigma model the origins of gls may be known back to the adoption of the lean concept. after wwii, japan developed the lean idea to compete with the united states’ mass manufacturing method (bhamu & singh sangwan, 2014). the toyota production system (tps), which was developed by japanese engineers taiichi ohno and shigeo shingo (kaswan et al., 2019), gave birth to the contemporary notion of lean manufacturing. although the lean strategy helps to decrease waste, it does not help to mitigate negative environmental impacts. the use of green technology can help to overcome the constraints of lean frameworks. green technology contributes to lean because it eliminates hazardous environmental consequences and waste. it is a long-term solution that minimizes global warming, acidification, and other environmental issues along the supply chain. even though green technology minimizes environmental pollutants, it is unable to reduce lean wastes (dües et al., 2013). as a result, a combined green lean (gl) strategy is needed which eliminates lean wastes and decreases carbon footprints. the two strategies that aim to reduce waste and involve waste management practices have a lot in common (sureeyatanapas et al., 2018). the combined gl technique has certain shortcomings, according to the literature. although gl minimizes wastes and emissions, it does not employ statistical methods to eliminate fluctuations in the process. even though it saves waste and pollution, the gl method is incapable of manufacturing high-quality products. as a result, there is a pressing need to devise a strategy that combines tools and strategies to overcome these constraints. six sigma is a projectbased statistical data-driven strategy that utilizes certified tools to augment gl methodology (siegel et al., 2019). six sigma is based on the principle that if flaws can be accurately pinpointed a solution can be devised to overcome them. this is a well-known statistical method that can cut faults down to 3.4 per million. although the six sigma process decreases variability, it does not eliminate waste or emissions. based on the extensive examination of green, lean, and six sigma, it may be settled that each strategy has its own set of advantages and disadvantages (sreedharan et al., 2018). each one of the green, lean, and six sigma methods complements the others, resulting in the evolution of green lean six sigma. gls is known for a sustainable development method and technique that focuses on reducing waste, pollution, and faults while delivering highquality and environment friendly goods. despite the evolution of gls, little effort has been made to implement this long-term strategy in industrial organizations. the absence of green, lean, and six sigma integration methods to implement different frameworks is one of the key causes behind this. furthermore, no paradigm exists in the literature that is appropriate regardless of the size, kind, or culture of the organization (kaswan & rathi, 2019). figure 2 depicts the gls framework as well as the specific procedures needed. to make the gls process viable, the framework is separated into six different interconnected aspects. figure 2 depicts the gls framework as well as the associated procedures. to make the gls process viable, the framework is separated into six different interconnected aspects. figure 2 shows the framework for the implementation of lean, green and six sigma. figure 2. framework for green, lean and six sigma. identification for the need of gls assement of whole project/process in the current state identification of problems and respective root causes solutions for the problems using gls approch making the solutions sustainable 4.1. identification of the need for gls the gls framework’s initial step involves choosing an acceptable project that is based on the amount of waste, faults, emissions, and customer feedback. gls is a method that is to be implemented on one project at a time, addressing each department or component separately. the implementation of gls necessitates a significant financial commitment as well as structural changes in the organization or processes. it’s critical to pick a gls project that has the most room for development in terms of sustainability. complete research of the various segments of the industry is required for this purpose. the comprehensive analysis of the whole industry reveals wastes, faults, and corresponding environmental emission levels for various industrial segments. after an appropriate project has been chosen, a framework or roadmap is created that is based on the timetable, scope and team members involved in the concerned project (bhamu & singh sangwan, 2014). int. j. prod. manag. eng. (2022) 10(2), 159-171creative commons attribution-noncommercial-noderivatives 4.0 international implementation of green, lean and six sigma operations for sustainable manufacturing. a review 165 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.2. assessment of the project/process in the current state the gls framework’s second phase is to estimate the current level of the project or system of the said project. the performance and efficiency of the chosen gls project is evaluated using multiple green, lean, and six sigma indicators. furthermore, green technology methods such as life cycle assessment are used to estimate co2 consumption, green energy coefficient, material usage, and so on (lca). value stream mapping (vsm) is a valuable lean method used to determine the present scope or level of different connected wastes. furthermore, in the measuring process, life cycle assessment (lca) is used to analyze the environmental effect of the different subprocesses in several environmental impact groups. the combination of vsm and lca results in the measurement of different lean and green wastes, providing a foundation for future development (faulkner & badurdeen, 2014). 4.3. identification of problems and root causes the gls framework’s next step is to identify the causative factors of high-level wastes, emissions, and faults in the chosen project. first, non-value-added activities and value-added are recognized from both a customer and a company standpoint. the process cycle efficiency is then calculated and compared to world-class standards to ascertain the level of improvement that is required. simultaneously, a thorough study of the project is carried out to detect bottlenecks and limits in the chosen project. after a thorough, in-depth examination of the project in question, the probable causes of waste, emissions, variances, and faults are identified. at this point, procedures like failure mode effect analysis (fmea), brainstorming, cause and effect analysis (c& e), five whys analysis, life cycle impact assessment, and others are used to determine the probable causes of the flaws that have been detected. after looking into all of the various causes, the search is narrowed down to only a few key factors for project inefficiency. this stage leads to an investigation of the primary sources of inefficiencies that must be addressed in order to improve the present project or system in question (erdil et al., 2018). 4.4. solutions for the problems using gls once the root causes of waste and inefficiency have been identified, various solutions or strategies are tested and delineated, and the best one among the solutions is implemented to eliminate the root causes. during this phase, the established and defined cause and effect relationship (from the analyzing phase) is employed to generate a wide range of viable solutions. upcycling, anaerobic digestion (ad), refuse-derived fuel (rdf), recirculation or recycling of water, and other options may be suggested. those involved are expected to be very creative during this phase. to find the best answer, various solutions (alternatives) are fleshed out, criteria are defined, and the solutions are assessed. to define the assessment criteria, all sources of data are examined, including stakeholders, consumers, project sponsors, and personnel (kaswan & rathi, 2020b). following the selection of the best available solution, the existing vsm is changed to represent the process after the modifications have been implemented. enhanced vsm is also used to estimate time savings, improved quality, and other related quality metrics. the best option is subsequently implemented as a pilot project. the tasks to be completed are documented, and the pilot participants are instructed on various components of the best solution. during the pilot project, the pilot solution is deployed in a specific industry sector (faulkner & badurdeen, 2014). 4.5. sustainable solutions if the existing system or process under evaluation records a significant improvement, this stage deals with maintaining or controlling the optimal option. to determine the degree of waste and emissions reduction, the entire process is re-evaluated using vsm and lca. numerous observations, data collecting, and control charts are utilized in this stage to revaluate the sigma level, water, electricity, and material usage, among other things. if the reevaluated performance characteristics are better than those measured in the previous stage, the chosen solution is maintained. once a viable solution for the pilot project has been established for a long time, it is replicated in other areas of the industry. through the distribution of eco-friendly products, the broad application of gls in the sector leads to better sustainability and a higher reputation on a worldwide scale (garza-reyes, 2015). 5. factors influencing lean six sigma models the proposed frameworks represent systems that describe the strong correlation between green int. j. prod. manag. eng. (2022) 10(2), 159-171 creative commons attribution-noncommercial-noderivatives 4.0 international nagadi 166 http://creativecommons.org/licenses/by-nc-nd/4.0/ lean, six sigma, and sustainability, along with the drivers and associated barriers, different conflicts, compatibility issues, and crucial success factors, that are associated with the integration’s benefits in addition to frameworks and techniques. these components operate in combination as a distinct, integrated approach to assist the company in identifying strengths and areas for development, as well as monitoring the effect and depth of change inside the business to reach conclusions and address various economic, environmental issues, and social effectiveness. these approaches enable businesses to take advantage of this combination of green lean, six sigma, and sustainability to increase efficiency and performance through understanding different drivers and obstacles, synergies, various conflicts and compatibility issues, essential success factors, and leveraging lean six sigma tools and procedures (banawi & bilec, 2014). 5.1. drivers and barriers to integrating the systems because some drivers and obstacles impact the adoption of lean/six sigma and sustainability efforts, it is vital to investigate the primary drivers and barriers to lean/six sigma and sustainability integration. both external and internal factors directly impact the combination of lean/six sigma and sustainability. cost savings, profitability, risk management, brand image enhancement, and resource management are examples of internal motivations (herrmann et al., 2008). for example, the cost of raw materials, energy, and resources is always increasing due to rising demand and associated resource limitations. furthermore, because it is challenging to predict cost trends, companies must improve their material efficiency to improve their competitiveness and performance. to improve one’s market position, the triple bottom line of sustainability must be met at the same time. many studies have linked trash reduction, emissions reduction, and increased recycling to improved financial performance (wadhwa, 2014). according to king & lenox (2001), a dedication to reducing environmental effects may help a company’s brand image, which can help it perform better in the industry. in fact, better environmental performance is regarded as a good indicator of corporate social responsibility. consumers, regulators, and stockholders are all external drivers (kadry, 2013). customers, regulators, rivals, and other stakeholders put pressure on all businesses, regardless of their size, location, or sector, to analyze and adjust their operations to enhance their social and environmental performance (wilson, 2010). furthermore, the general public is becoming more environmentally conscious, and customers are actively seeking “greener” alternatives. environmental reporting, compliance, and openness are being reshaped by regulators and politicians. environmental and social performance has become a key priority for investors, shareholders, banks, and insurance corporations (kadry, 2013). many companies have responded to these forces by introducing lean/six sigma and sustainability initiatives that enable them to enhance their operational, environmental, and social outcomes. even though lean/six sigma and sustainability have been successfully integrated within many settings, the road ahead is not without obstacles. lack of environmental knowledge (rothenberg et al., 2001), a perception of increased costs, and organizational structures that separate environmental and continuous improvement choices (dakov & novkov, 2007) are some of the barriers that impact achievement. the traditional belief that improving environmental and social performance is an impediment to economic progress (found, 2009) has persisted due to a lack of environmental responsibility. companies will only substantially integrate environmental and social features if they are confident that doing so would drastically increase revenues, according to simboli et al. (2014). furthermore, research has shown that excluding human resources from lean six sigma programmes reduces the likelihood of achieving greater long-term advantages. 5.2. benefits of the integration of lean/six sigma and sustainability business techniques and operations such as lean/ six sigma and sustainability may be beneficial to businesses. previous studies have consistently found that implementing lean/six sigma and sustainability can have a favourable impact on a company’s performance (dües et al., 2013). when lean/six sigma and sustainability are adopted in combination as opposed to individually, they can have a greater and more favourable influence on an organization’s achievements (miller et al., 2010). internal benefits are relatively more valued than external advantages because, unlike the choice to implement sustainability, the reasons for integrating the two strategies are more internal than external. furthermore, because of the broader scope involved in the integration, the benefits rendered as a result of the integration are more significant when strategies are implemented in combination as opposed to independently. int. j. prod. manag. eng. (2022) 10(2), 159-171creative commons attribution-noncommercial-noderivatives 4.0 international implementation of green, lean and six sigma operations for sustainable manufacturing. a review 167 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5.3. tools, techniques and methods lean six sigma provides a number of methods to help businesses decrease waste. according to several researchers, these technologies appear to decrease manufacturing enterprises’ environmental and social consequences (chiarini & vagnoni, 2015; epa, 2003). value stream mapping, cellular manufacturing, standard work, visual management, just in time, smed, supplier relationship, six sigma, statistical process control, analysis tools, and plant layout reconfiguration are all examples of these methodologies. many of these strategies have been modified and expanded to achieve greater environmental and social improvement (langenwalter, 2006). many factors encourage the adoption of lean/six sigma technologies and approaches to enhance sustainable development. first, the tools are already in place and have been tested comprehensively. second, employees already understand and use them (chiarini, 2014). there are several scenarios where lean/six sigma tools and techniques have the potential to enhance environmental and social risks or impacts (herrmann et al., 2008), but these risks can be reduced or eliminated if environmental and social considerations are integrated pro-actively and deliberately as part of lean/six sigma implementation. putting tools and processes in place and keeping them up to date typically involves a large amount of time and effort. as a result, it’s critical to ensure that this work pays off in all aspects of consideration over time (herrmann et al., 2008). the use of techniques and technologies in the integration of lean/six sigma and sustainability has been deemed critical (chiarini, 2014). the tools/techniques must be carefully chosen and utilized strategically, and they must be compatible with the existing structure (epa, 2003). 5.4. synergies and conflicts between lean/ six sigma and sustainability numerous researchers have looked into the relationship between lean/six sigma and long-term sustainability. others have studied the relationship between lean/six sigma and sustainability (king & lenox, 2001), while others still have examined the interconnections between the two ideas (bergmiller & mccright, 2009). according to this research, combining lean/six sigma with sustainability can have both positive and negative consequences for economic, social, and environmental performance. corporations can close the gap between lean/six sigma and sustainability by comprehending the synergies and tensions between the two strategies. 5.4.1. synergies and compatibility lean/six sigma and sustainability frequently go hand-in-hand, according to ng et al. (2015), and this link is apparent in the extant literature. many scholars (bergmiller & mccright, 2009; herrmann et al., 2008; pampanelli et al., 2014) have found that lean/six sigma and sustainability have a substantial synergy, implying that firms experienced with lean/ six sigma would quickly comprehend sustainability and vice versa. dües et al. (dües et al., 2013) recently discovered that the two worlds have a positive and powerful interaction. the united states environmental protection agency (epa) recognized this synergistic link more than 15 years ago, and they are now using lean/six sigma concepts and techniques to generate economic, social, and environmental advantages. according to larson & greenwood (2004), there are significant prospects for combining these two parallel worlds, which would result in significant competitiveness and sustainability improvements. lean/six sigma and sustainability strategies are frequently seen to be complementary. furthermore, prasad and sharma (prasad & sharma, n.d.) argued that lean/six sigma and sustainability might be combined to provide better financial and operational results. waste reduction, a continuous improvement mentality backed by supply chain relationships, performance measurement, management commitment and staff participation, customer happiness, and common tools and techniques are all accessible as a result of the implementation of lean/six sigma and sustainability strategies. 5.4.2. conflicts and potential shortcoming despite the numerous synergies mentioned in the preceding section, lean/six sigma and sustainability cannot be combined perfectly (dües et al., 2013). some conflicts exist between lean/six sigma and sustainability because of lean/six sigma’s focus on ensuring customer demands for quality and durability are met, even if this necessitates additional packing or the use of more harmful chemicals. the goal of lean/six sigma is to reduce waste by eliminating faulty goods. this strategy, however, pays little regard to the long-term value of goods, as well as the environmental risk of the materials and transformation processes utilized to make them (larson & greenwood, 2004; wilson, int. j. prod. manag. eng. (2022) 10(2), 159-171 creative commons attribution-noncommercial-noderivatives 4.0 international nagadi 168 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2010). furthermore, several researchers have found that lean/six sigma focuses their long-term efforts on transformation processes, ignoring material extraction, product usage, and final disposition (dakov & novkov, 2007; larson & greenwood, 2004; maskell & pojasek, 2008). furthermore, there are rare instances where the use of particular lean/six sigma principles is incompatible with long-term viability. several studies show that even a single-time adoption increases the frequency of deliveries in small-scale quantities and smaller vehicles, increases traffic congestion, and increases greenhouse gas emissions (carvalho et al., 2011; venkat & wakeland, 2006). the continuous improvement mindset, according to pagell & gobeli (2009), may help a company become more sustainable. however, when a company needs to drastically restructure its operations to become sustainable, the same concept may impede dramatic innovation (benner & tushman, 2003). on the social side, wilson (wilson, 2010) pointed out that the lean/six sigma methodology does not address social factors. the ideology exclusively considers customer safety and ignores the health and safety of employees. in addition, the lean/six sigma approach does not manage other social sustainability problems like as human rights and community impact. 5.5. frameworks, models and methods companies have been pushed to apply lean/six sigma and sustainability to enhance their performance because of the beneficial association between the two (dües et al., 2013; epa, 2003). to combine and apply lean/six sigma with sustainability, several scholars have presented numerous models, frameworks, and approaches. these models, frameworks, and approaches were examined in this literature review. this examination exposes several flaws that are present in the majority of business models, frameworks, and approaches. these models, frameworks, and methods emphasize the importance of leadership, employee involvement, and a mature deployment level in using and applying lean/six sigma tools, as well as a high level of environmental awareness for cultural transformation and continuous improvement that leads to a high-performing organization (ng et al., 2015; pampanelli et al., 2014). a culture of continual improvement underpins all of the models, frameworks, and methodologies. 6. conclusion this study includes a literature review for the successful integration of six sigma, lean and sustainability in different industrial manufacturing processes. the review helped in the determination of research gaps and identification of theoretical elements for efficient integration of the green, lean and six sigma models. major gaps were identified to integrate the gls model. the integration of gls involves the identification of the need for gls, assessment of current manufacturing processes, identification of problems and root causes, solutions using the gls approach and making these solutions sustainable. gls approach not only addresses the sustainability but also the economic, social and environmental aspects of any manufacturing process. moreover, barriers and enablers for six sigma and lean operations were identified using the dmaic model. gls is known to reduce the overall negative environmental implications with highly efficient manufacturing processes. there is a need to understand and identify different crucial elements for the successful implementation of the gls approach to achieve sustainability goals. the successful integration of gls is dependent on various theoretical elements, enablers and integration tools. integration tools and methods aid in the implementation of the gls approach by overcoming the barrier along the path of integration. dmaic model works well to execute the integration of the gls approach that identifies the enablers for successful integration. this review presents a complete roadmap for gls implementation and integration through assessment and identification of desired improvements in existing manufacturing processes. references baker, b. (2003). lean six sigma: combining six sigma quality with lean speed. quality progress, 36(10), 96. banawi, a., & bilec, m.m. (2014). a framework to improve construction processes: integrating lean, green and six sigma. international journal of construction management, 14(1), 45–55. https://doi.org/10.1080/15623599.2013.87 5266 benner, m.j., & tushman, m.l. (2003). exploitation, exploration, and process management: the productivity dilemma revisited. academy of management review, 28(2), 238–256. https://doi.org/10.5465/amr.2003.9416096 int. j. prod. manag. eng. (2022) 10(2), 159-171creative commons attribution-noncommercial-noderivatives 4.0 international implementation of green, lean and six sigma operations for sustainable manufacturing. a review 169 https://doi.org/10.1080/15623599.2013.875266 https://doi.org/10.1080/15623599.2013.875266 https://doi.org/10.5465/amr.2003.9416096 http://creativecommons.org/licenses/by-nc-nd/4.0/ bergmiller, g.g., & mccright, p.r. (2009). are lean and green programs synergistic. proceedings of the 2009 industrial engineering research conference, 1155–1160. bhamu, j., & singh sangwan, k. (2014). lean manufacturing: literature review and research issues. international journal of operations & production management, 34(7), 876–940. https://doi.org/10.1108/ijopm-08-2012-0315 byrne, g., lubowe, d., & blitz, a. (2007). using a lean six sigma approach to drive innovation. strategy & leadership, 35(2), 5–10. https://doi.org/10.1108/10878570710734480 carvalho, h., duarte, s., & machado, v.c. (2011). lean, agile, resilient and green: divergencies and synergies. international journal of lean six sigma, 2(2), 151-179 https://doi.org/10.1108/20401461111135037 cherrafi, a., elfezazi, s., chiarini, a., mokhlis, a., & benhida, k. (2016). the integration of lean manufacturing, six sigma and sustainability: a literature review and future research directions for developing a specific model. journal of cleaner production, 139, 828–846. https://doi.org/10.1016/j.jclepro.2016.08.101 cherrafi, a., elfezazi, s., govindan, k., garza-reyes, j.a., benhida, k., & mokhlis, a. (2017). a framework for the integration of green and lean six sigma for superior sustainability performance. international journal of production research, 55(15), 4481–4515. https://doi.org/10.1080/00207543.2016.1266406 chiarini, a. (2014). sustainable manufacturing-greening processes using specific lean production tools: an empirical observation from european motorcycle component manufacturers. journal of cleaner production, 85, 226–233. https://doi.org/10.1016/j.jclepro.2014.07.080 chiarini, a., & vagnoni, e. (2015). world-class manufacturing by fiat. comparison with toyota production system from a strategic management, management accounting, operations management and performance measurement dimension. international journal of production research, 53(2), 590–606. https://doi.org/10.1080/00207543.2014.958596 dakov, i., & novkov, s. (2007). assessment of the lean production effect on the sustainable industrial enterprise development. business: theory and practice, 8(4), 183–188. https://doi.org/10.3846/btp.2007.26 duarte, s., & cruz-machado, v. (2019). green and lean supply-chain transformation: a roadmap. production planning & control, 30(14), 1170–1183. https://doi.org/10.1080/09537287.2019.1595207 dües, c.m., tan, k.h., & lim, m. (2013). green as the new lean: how to use lean practices as a catalyst to greening your supply chain. journal of cleaner production, 40, 93–100. https://doi.org/10.1016/j.jclepro.2011.12.023 elkington, j. (1994). towards the sustainable corporation: win-win-win business strategies for sustainable development. california management review, 36(2), 90–100. https://doi.org/10.2307/41165746 epa, u. (2003). lean manufacturing and the environment: research on advanced manufacturing systems and the environment and recommendations for leveraging better environmental performance. united states environmental protection agency. erdil, n.o., aktas, c.b., & arani, o.m. (2018). embedding sustainability in lean six sigma efforts. journal of cleaner production, 198, 520–529. https://doi.org/10.1016/j.jclepro.2018.07.048 evans, j.r., & lindsay, w.m. (2014). an introduction to six sigma and process improvement. cengage learning. faulkner, w., & badurdeen, f. (2014). sustainable value stream mapping (sus-vsm): methodology to visualize and assess manufacturing sustainability performance. journal of cleaner production, 85, 8–18. https://doi.org/10.1016/j. jclepro.2014.05.042 flor vallejo, v., antony, j., douglas, j.a., alexander, p., & sony, m. (2020). development of a roadmap for lean six sigma implementation and sustainability in a scottish packing company. the tqm journal, 32(6), 1263–1284. https://doi.org/10.1108/tqm-02-2020-0036 found, p. (2009). lean and low environmental impact manufacturing. garza-reyes, j.a. (2015). green lean and the need for six sigma. international journal of lean six sigma, 6(3), 226-248. https://doi.org/10.1108/ijlss-04-2014-0010 herrmann, c., thiede, s., stehr, j., & bergmann, l. (2008). an environmental perspective on lean production. in m. mitsuishi, k. ueda, & f. kimura (eds.), manufacturing systems and technologies for the new frontier (pp. 83–88). springer. https://doi.org/10.1007/978-1-84800-267-8_16 hussain, k., he, z., ahmad, n., & iqbal, m. (2019). green, lean, six sigma barriers at a glance: a case from the construction sector of pakistan. building and environment, 161, 106225. https://doi.org/10.1016/j.buildenv.2019.106225 johansson, g., & sundin, e. (2014). lean and green product development: two sides of the same coin? journal of cleaner production, 85, 104–121. https://doi.org/10.1016/j.jclepro.2014.04.005 kadry, s. (2013). six sigma methodology for the environment sustainable development. in mechanism design for sustainability (pp. 61–76). springer. https://doi.org/10.1007/978-94-007-5995-4_4 kaswan, m.s., & rathi, r. (2019). analysis and modeling the enablers of green lean six sigma implementation using interpretive structural modeling. journal of cleaner production, 231, 1182–1191. https://doi.org/10.1016/j. jclepro.2019.05.253 int. j. prod. manag. eng. (2022) 10(2), 159-171 creative commons attribution-noncommercial-noderivatives 4.0 international nagadi 170 https://doi.org/10.1108/ijopm-08-2012-0315 https://doi.org/10.1108/10878570710734480 https://doi.org/10.1108/20401461111135037 https://doi.org/10.1016/j.jclepro.2016.08.101 https://doi.org/10.1080/00207543.2016.1266406 https://doi.org/10.1016/j.jclepro.2014.07.080 https://doi.org/10.1080/00207543.2014.958596 https://doi.org/10.3846/btp.2007.26 https://doi.org/10.1080/09537287.2019.1595207 https://doi.org/10.1016/j.jclepro.2011.12.023 https://doi.org/10.2307/41165746 https://doi.org/10.1016/j.jclepro.2018.07.048 https://doi.org/10.1016/j.jclepro.2014.05.042 https://doi.org/10.1016/j.jclepro.2014.05.042 https://doi.org/10.1108/tqm-02-2020-0036 https://doi.org/10.1108/ijlss-04-2014-0010 https://doi.org/10.1007/978-1-84800-267-8_16 https://doi.org/10.1016/j.buildenv.2019.106225 https://doi.org/10.1016/j.jclepro.2014.04.005 https://doi.org/10.1007/978-94-007-5995-4_4 https://doi.org/10.1016/j.jclepro.2019.05.253 https://doi.org/10.1016/j.jclepro.2019.05.253 http://creativecommons.org/licenses/by-nc-nd/4.0/ kaswan, m.s., & rathi, r. (2020a). investigating the enablers associated with implementation of green lean six sigma in manufacturing sector using best worst method. clean technologies and environmental policy, 22(4), 865–876. https://doi.org/10.1007/s10098-020-01827-w kaswan, m.s., & rathi, r. (2020b). green lean six sigma for sustainable development: integration and framework. environmental impact assessment review, 83, 106396. https://doi.org/10.1016/j.eiar.2020.106396 kaswan, m.s., rathi, r., & singh, m. (2019). just in time elements extraction and prioritization for health care unit using decision making approach. international journal of quality & reliability management, 36(7), 1243–1263. https://doi.org/10.1108/ijqrm-08-2018-0208 kidwell, m. (2006). lean manufacturing and the environment. target, 22(6), 13–18. king, a.a., & lenox, m.j. (2001). lean and green? an empirical examination of the relationship between lean production and environmental performance. production and operations management, 10(3), 244–256. https://doi.org/10.1111/j.1937-5956.2001.tb00373.x langenwalter, g. (2006). life” is our ultimate customer: from lean to sustainability. target, 22(1), 5–15. larson, t., & greenwood, r. (2004). perfect complements: synergies between lean production and eco-sustainability initiatives. environmental quality management, 13(4), 27–36. https://doi.org/10.1002/tqem.20013 maskell, b.h., & pojasek, r.b. (2008). for lean to be green the performance measurements must change. bma, hanover. miller, g., pawloski, j., & standridge, c.r. (2010). a case study of lean, sustainable manufacturing. journal of industrial engineering and management (jiem), 3(1), 11–32. https://doi.org/10.3926/jiem.2010.v3n1.p11-32 ng, r., low, j.s. c., & song, b. (2015). integrating and implementing lean and green practices based on proposition of carbon-value efficiency metric. journal of cleaner production, 95, 242–255. https://doi.org/10.1016/j. jclepro.2015.02.043 pagell, m., & gobeli, d. (2009). how plant managers’ experiences and attitudes toward sustainability relate to operational performance. production and operations management, 18(3), 278–299. https://doi.org/10.1111/j.19375956.2009.01050.x pampanelli, a.b., found, p., & bernardes, a.m. (2014). a lean & green model for a production cell. journal of cleaner production, 85, 19–30. https://doi.org/10.1016/j.jclepro.2013.06.014 prasad, s., & sharma, s.k. (n.d.). lean and green manufacturing: concept and its implementation in operations management. 6. rothenberg, s., pil, f.k., & maxwell, j. (2001). lean, green, and the quest for superior environmental performance. production and operations management, 10(3), 228–243. https://doi.org/10.1111/j.1937-5956.2001.tb00372.x scoones, i. (2007). sustainability. development in practice, 17(4–5), 589–596. https://doi.org/10.1080/09614520701469609 siegel, r., antony, j., garza-reyes, j.a., cherrafi, a., & lameijer, b. (2019). integrated green lean approach and sustainability for smes: from literature review to a conceptual framework. journal of cleaner production, 240, 118205. https://doi.org/10.1016/j.jclepro.2019.118205 simboli, a., taddeo, r., & morgante, a. (2014). value and wastes in manufacturing. an overview and a new perspective based on eco-efficiency. administrative sciences, 4(3), 173–191. https://doi.org/10.3390/admsci4030173 singh, m., & rathi, r. (2019). a structured review of lean six sigma in various industrial sectors. international journal of lean six sigma, 10(2), 622-664 https://doi.org/10.1108/ijlss-03-2018-0018 sreedharan v.r., sandhya, g., & raju, r. (2018). development of a green lean six sigma model for public sectors. international journal of lean six sigma, 9(2), 238–255. https://doi.org/10.1108/ijlss-02-2017-0020 sureeyatanapas, p., poophiukhok, p., & pathumnakul, s. (2018). green initiatives for logistics service providers: an investigation of antecedent factors and the contributions to corporate goals. journal of cleaner production, 191, 1–14. https://doi.org/10.1016/j.jclepro.2018.04.206 venkat, k., & wakeland, w. (2006). is lean necessarily green? proceedings of the 50th annual meeting of the isss-2006, sonoma, ca, usa. wadhwa, r.s. (2014). synergizing lean and green for continuous improvement. in b. grabot, b. vallespir, s. gomes, a. bouras, & d. kiritsis (eds.), advances in production management systems. innovative and knowledge-based production management in a global-local world (pp. 154–161). springer. https://doi.org/10.1007/978-3-662-447369_19 wilson, a. (2010). sustainable manufacturing: comparing lean, six sigma, and total quality manufacturing. usa: strategic sustainability consulting, washington, dc. zhu, q., johnson, s., & sarkis, j. (2018). lean six sigma and environmental sustainability: a hospital perspective. supply chain forum: an international journal, 19(1), 25–41. https://doi.org/10.1080/16258312.2018.1426339 int. j. prod. manag. eng. (2022) 10(2), 159-171creative commons attribution-noncommercial-noderivatives 4.0 international implementation of green, lean and six sigma operations for sustainable manufacturing. a review 171 https://doi.org/10.1007/s10098-020-01827-w https://doi.org/10.1016/j.eiar.2020.106396 https://doi.org/10.1108/ijqrm-08-2018-0208 https://doi.org/10.1111/j.1937-5956.2001.tb00373.x https://doi.org/10.1002/tqem.20013 https://doi.org/10.3926/jiem.2010.v3n1.p11-32 https://doi.org/10.1016/j.jclepro.2015.02.043 https://doi.org/10.1016/j.jclepro.2015.02.043 https://doi.org/10.1111/j.1937-5956.2009.01050.x https://doi.org/10.1111/j.1937-5956.2009.01050.x https://doi.org/10.1016/j.jclepro.2013.06.014 https://doi.org/10.1111/j.1937-5956.2001.tb00372.x https://doi.org/10.1080/09614520701469609 https://doi.org/10.1016/j.jclepro.2019.118205 https://doi.org/10.3390/admsci4030173 https://doi.org/10.1108/ijlss-03-2018-0018 https://doi.org/10.1108/ijlss-02-2017-0020 https://doi.org/10.1016/j.jclepro.2018.04.206 https://doi.org/10.1007/978-3-662-44736-9_19 https://doi.org/10.1007/978-3-662-44736-9_19 https://doi.org/10.1080/16258312.2018.1426339 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.17946 received: 2022-06-27 accepted: 2022-07-18 the factors of competitiveness management of manufacturing enterprise personnel in conditions of uncertainty kovtunenko, k.v. a1 , kovtunenko yu.v. a2 , fomina n.m. a3* , kovalchuk o.v. a4 , kovtunenko d.yu. a5 a imi department, odessа polytechnic national university, odessa, shevchenko av., 1, 65044, ukraine. a1 k.v.kovtunenko@op.edu.ua, a2 y.v.kovtunenko@op.edu.ua, a3 nfomina@epsa.upv.es, a4 kovalchuk.o.v@op.edu.ua, a5 9391661@stud.op.edu.ua abstract: personnel competitiveness has always been one of the main advantages of any manufacturing enterprise. therefore, competitiveness management of manufacturing enterprise personnel is one of the most foreground types of management. the article covers the most widespread issues and difficulties of personnel management implementation nowadays, considering current complicated conditions of uncertainty, which become increasingly sophisticated every time. moreover, our study demonstrates some possible ways for manufacturing enterprises to overcome existing obstacles and become ready for new ones in conditions of uncertainty. as covid-19 and martial law in ukraine with their restrictions have contributed to the way of personnel management realization, many manufacturing enterprises have been suffering because of that and have been forced to change their management principles to meet the new conditions. therefore, the latest theories of personnel competitiveness management are introduced in our science study to help manufacturing enterprises cope with the changing conditions of uncertainty in the modern world. key words: competitiveness, personnel competitiveness, manufacturing enterprise, uncertainty, covid-19, martial law in ukraine. 1. introduction in  the  modern  realities  of  the  development  of  industrial  enterprises  in  the  world,  a  new  type  of  society  is  formed  based  on  knowledge  and  information. it adheres to the vector of  innovative  development. usually, in the international market, the  competitive advantage is given to the manufacturing  company that has managed to adapt its development  strategy to the external environment, has succeeded  in  implementing  innovations  and  usually  applies  the  latest  theories  of  personnel  management  and  competitiveness. thus, it is worth noting that in the  process of forming the competitiveness of personnel  of modern manufacturing enterprises, many features  must be taken into account by entrepreneurs to ensure  their effective operation in conditions of uncertainty.  in  2022  conditions  of  uncertainty  are  still  caused  by  the  covid-19  pandemic  and  exacerbated  by  military actions in ukraine. in general, the conditions of uncertainty are faced by  more and more personnel of industrial enterprises in  ukraine. it certainly affects the entire global personnel  market because, due to the escalation of the military  conflict,  much  talented  personnel  are  migrating  to cite this article: kovtunenko, k.v., kovtunenko, yu.v., fomina, n.m., kovalchuk, o.v., kovtunenko, d.yu. (2022). the factors of competitiveness management of manufacturing enterprise personnel in conditions of uncertainty. international journal of production management and engineering, 10(2), 225-235. https://doi.org/10.4995/ijpme.2022.17946 int. j. prod. manag. eng. (2022) 10(2), 225-235creative commons attribution-noncommercial-noderivatives 4.0 international 225 https://orcid.org/0000-0002-3759-7950 https://orcid.org/0000-0001-8528-605x https://orcid.org/0000-0002-2334-8267 https://orcid.org/0000-0002-7324-0828 https://orcid.org/0000-0001-8567-8504 http://creativecommons.org/licenses/by-nc-nd/4.0/ across  europe,  allowing  industrial  enterprises  to  increase  their  competitiveness  staff  from  ukraine.  the labour market in current conditions is constantly  changing  and  introducing  the  latest  management  theories is the main task of every manager. in today’s  needs, to increase the competitiveness of staff, it is  necessary to implement the latest and most effective  management theories to keep up with uncertainty. however,  given  the  crucial  importance  of  this  economic  category,  a  comprehensive  theoretical  justification of  the role of staff competitiveness  in  the further effective development of  the enterprise  in  conditions  of  uncertainty  remains  relevant.  furthermore,  it  determines  the  need  for  further  research in this area. 2. methodology and literature review 2.1. methodological framework since the topic of our scientific work has not only a  three-dimensional theoretical basis but also extensive  application in practice, we used such a method of  theoretical and empirical research, including analysis  and  synthesis,  comparison  and  generalization.  furthermore,  modern  computer  technology  was  used for data processing. after defining the topic, the  following systematic work was carried out, which  consisted of several stages. the  first  stage  was  the  collection  of  data  on  the  topic,  searching  for  statistical  and  analytical  sources  of  information,  and  separating  those  data  that may be useful for the study. the second stage  was  forming  and  structuring  data  according  to  the  criteria  of  conciseness  and  logic.  the  main  stage was analysing and synthesizing  the received  information and discussing the main problems and  obstacles to their solution. then all the received data  were supplemented, clarified and edited, as well as  the elimination of  redundant  information. the  last  stage was the generalization of information, i.e. the  formation of conclusions based on the study’s results. 2.2. literature review the information base of the study consisted of works  of domestic and foreign scientists, practitioners on  the  management  of  competitiveness  of  personnel  of  the  manufacturing  enterprise  in  conditions  of  uncertainty,  current  legislation,  analytical  and  statistical  data  of  domestic  and  international  open  sources and the results of their research. as  issues  covered  in  our  article  are  standard  worldwide,  the  study  was  carried  out  using  the  results  of  previous  scientific  works.  among  the  foreign  authors  should  be  noted  the  contribution  of drucker p., porter m., strickland a., thompson  a., etc. the theoretical and methodological bases of  personnel competitiveness management are devoted  to the works of boginya (2003), buchynska (2016), kozak (2011), marchenko  (2015), dunska  (2012), filyppova (2019), kovalenko (2020), etc. 3. analysis, results and discussion 3.1. personnel competitiveness in  this context,  it  is first appropriate  to define  the  concept  of  “staff  competitiveness”  and  consider  its  components.  staff  competitiveness  is  a  multicomponent  and  complex  economic  category.  competitiveness is formed as an integral indicator  of those qualities due to which a particular employee  is  better  than  others  for  a  specific  position  (from  the enterprise’s point of view and due to which the  enterprise  gives  it  an  advantage)  (nesterenko  and  chernyakova, 2011). personnel  competitiveness  is  a  complex  economic  category,  the  essence  of  which  is  revealed  by  the  following conceptual approaches to its definition: 1.  the  ability  of  managers,  practitioners  and  professionals  to  generate  ideas  using  all  the  opportunities  that  arise  in  the  external  and  internal  environment;  timely  identify  threats  to  the  enterprise;  solve  problems;  invent  and  implement  innovations  faster  than  competitors  at all stages of the product life cycle; to ensure  the  achievement  of  goals  and  fulfilment  of  its  mission; 2.  ability  to  effectively  perform  management  functions  and  timely  make  qualified  decisions  to  develop,  manufacture  and  sell  high-quality  products  with  unique  consumer  properties  (services) with the most efficient use of resources; 3.  the  ability  to  more  fully,  compared  to  other  candidates  for  vacant  positions,  meet  the  requirements of employers in terms of knowledge,  skills and abilities; int. j. prod. manag. eng. (2022) 10(2), 225-235 creative commons attribution-noncommercial-noderivatives 4.0 international kovtunenko et al. 226 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.  ability to show their personal, professional and  business  qualities  faster  and  better  than  other  employees to realize their potential under specific  conditions  prevailing  in  the  enterprise,  which  allows them to receive appropriate remuneration,  achieve  appropriate  social  status  and  ensure  professional growth (halaz, 2008). thus, the personnel is the “driver” of any enterprise.  without the human factor, the effective functioning  of the organization is impossible. without qualified  specialists,  no  firm  can  achieve  its  goal.  the  experience of many foreign companies from ukraine,  germany, canada, etc. shows the strengthening of the  role of human resource management in the system  of factors that ensure competitiveness. to achieve a  more stable competitive position, the company must  constantly  develop  staff  development  programs  to  provide  employees  with  many  necessities,  development  and  improvement  professional  and  general education. intangible  assets  should  be  included  as  one  of  the  modern  components  of  staff  competitiveness.  successful  intangible assets are  the result of wellplanned  strategic  activities  of  the  enterprise.  an  experienced  leader,  first  of  all,  is  engaged  in  improving the knowledge base of his subordinates.  it  increases  the  intellectual  value  and  potential  of  the  enterprise  as  a  whole.  nowadays,  not  many  experienced professionals create specific resources  for the company in which they work. thanks to such  “specific personnel”,  the value of  the enterprise  is  completed,  and  the  so-called  “competitiveness  of  the  enterprise  personnel”  is  formed  (kovtunenko,  2013). the essence of  the expansion of  intangible  assets  is  to  turn  ordinary  people  into  experienced  “generators”  of  ideas,  i.e.  “specific  resources”  for  this company. as a result, such innovators can bring  additional profits to the company. the ongoing development of modern technologies  and  the  it  sector  plays  a  significant  role  in  the  process  of  intellectualization  of  personnel.  it  also  affects the process of managing the competitiveness  of  personnel  of  modern  enterprises.  most  modern  manufacturing  companies  are  increasingly  beginning to implement new it departments in their  organizational structure that specialize in optimizing  the enterprise’s business processes and increasing its  competitiveness. examples of so-called optimization  processes are the introduction of intelligent document  management  or  modern  agile  methodologies  that  allow to manage production cycles of the enterprise  with  special  software  and  achieve  the  desired  goals  faster.  the  development  of  such  advanced  technologies, of course, affects the competitiveness  of staff in general, as new professionals are involved,  and improves the skills of existing personnel. 3.2. lifelong learning concept influential  aspects  of  staff  competitiveness  also  include  the  concept  of  lifelong  learning,  which  is  becoming  more  relevant  and  aware  among  young  people  and  professionals  in  various  fields.  in  addition, the desire to develop and be one step ahead  of its competitors is a great driving force and “spark”  that motivates staff to build their personality further  and work tirelessly on themselves. the concept of lifelong learning is one of the most  popular trends, which more and more professionals  seek  from  around  the  world  at  the  present  stage.  awareness  and  constantly  learning  new  things  are  becoming  a  fashion  trend  in  today’s  society.  the  concept  of  continuing  education  means  conscious  continuous  improvement  and  holistic  development  of  the  specialist  throughout  life,  increasing  opportunities  for  work  and  social  adaptation  in  a  changing world. in general,  this concept first appeared in  the early  20th  century.  the  reason  for  its  emergence  was  the rapid development of society and the need for  a  more  skilled  workforce.  however,  this  concept  did  not  receive  approval  for  a  long  time  until,  in  1975, it was adopted by unesco. the concept of  lifelong learning allows us to act in a new situation,  using  all  our  experiences  together,  and  to  update  and  transform  it  in  terms  of  our  own  beliefs  and  values,  as  well  as  feelings,  feelings  and  emotions  that  are  manifested  concerning  this  situation  and  its solution. therefore, adherence to this concept in  managing personnel  competitiveness,  especially  in  conditions of uncertainty -  is essential for modern  manufacturing enterprises. 3.3. personnel evaluation of  course,  the  timely  assessment  also  plays  an  essential  role  in  managing  staff  competitiveness.  personnel  appraisal  is  a  procedure  of  checking  employees’ work for compliance with specific criteria  for the effectiveness of certain actions within the job  responsibilities within the company. most companies  have their employee appraisal system, according to  which  employees  are  appraised  regularly.  it  is  an  int. j. prod. manag. eng. (2022) 10(2), 225-235creative commons attribution-noncommercial-noderivatives 4.0 international the factors of competitiveness management of manufacturing enterprise personnel in conditions of uncertainty 227 http://creativecommons.org/licenses/by-nc-nd/4.0/ opportunity to assess the employee’s progress, praise  his/her achievements and work together to achieve  goals  to  increase  staff  competitiveness  and  help  achieve the company’s goals. evaluation results are  often critical in raising career ladders, bonuses, pay  raises,  and  so  on.  in  addition,  regular  evaluations  help employees better understand what is expected of  them, improve communication between management  and employees, giving employees proper recognition  of  their  work  because  the  assessment  of  staff  allows  concluding  the  professional  development  of a particular employee and promptly offering the  necessary training or retraining. employee  evaluation  has  several  goals  aimed  at  increasing  the  competitiveness  of  staff.  the  advantages of professional assessment of employees  include (martynenko et al., 2013): helping  employees  better  understand  what  is  expected of them; an  opportunity  for  the  manager  to  better  understand  the  strengths  and  motivation  of  employees; providing employees with valuable feedback on  how they can improve their work in the future; assistance to the employee and the manager in  planning the future career of the employee; objective  reviews  of  people  based  on  standard  indicators can be helpful for a fair assessment of  promotion, promotion and bonuses. currently,  many  methods  are  used  for  business  evaluation of personnel  in personnel management,  each of which is relevant at a particular enterprise  stage.  however,  the  whole  evaluation  system  is  effective  only  in  a  comprehensive  assessment,  the  structure  of  which  can  be  represented  as  a  model  (figure 1), combining three groups of characteristics,  including  employee  quality,  work  behaviour  and  performance. from  the  presented  model,  it  is  seen  that  the  characteristics  of  the  three  groups  are  closely  related.  it  is  not  surprising  because  the  quality  of  the employee is the basis for the formation of labour  activity, which at the same time serves as a process  for the construction of performance. there  are  many  methods  of  personnel  evaluation  used  by  enterprises  and  organizations.  however,  they are conventionally divided into traditional and  modern.  traditional  methods  are  focused  on  the  individual employee and based only on the subjective  assessment of the head as a result of the analysis of  the achievements of the employee. the disadvantage  of this method is to provide an evaluation, ignoring  the company’s goals and prospects and colleagues’  opinions,  which  makes  it  relevant  only  for  large  enterprises  with  a  stable  external  environment.  traditional methods include: biographical method; ranking; the method of a given score; graphic profile method; method of pairwise comparisons; evaluation by results. figure 1. model of comprehensive employee evaluation. source: authors’ own development. int. j. prod. manag. eng. (2022) 10(2), 225-235 creative commons attribution-noncommercial-noderivatives 4.0 international kovtunenko et al. 228 http://creativecommons.org/licenses/by-nc-nd/4.0/ modern methods are focused on the organization’s  future and assess the employees’ effectiveness in the  group and their ability to develop and master new  knowledge and processes. they include: situation modelling method; committee method; method of 360 degrees; method of business games; goal management method; method of solving situations; method of assessment centre; analysis of human resources. 3.4. factors of personnel competitiveness thus, the components influencing the formation of  staff competitiveness should include local (internal to  the manufacturing enterprise) factors (organization  of the day, organization of the workplace, working  conditions,  wages,  labour  evaluation,  training,  incentives  and  motivation)  and  personal  elements  (development  of  unique  qualities,  self-education,  physical  and  spiritual  growth,  self-esteem,  selfimprovement).  the  main  factors  influencing  staff  competitiveness  should  be  shown  graphically  in  figure 2. 3.5. conditions of uncertainty. covid-19. martial law in ukraine since  the  main  features  of  the  formation  of  staff  competitiveness  have  already  been  mentioned,  it  is  advisable  to  shift  the  focus  to  managing  the  competitiveness  of  personnel  of  manufacturing  enterprises under conditions of uncertainty. however,  before considering the conditions of uncertainty, we  propose to dwell on the definition of uncertainty and  separate  its  significance  in  the  modern  world  for  manufacturing enterprises. uncertainty  is  the  incompleteness or  inaccuracy  of  information,  which  is  the  main  obstacle  to  developing  the  manufacturing  enterprise.  unfortunately, uncertainty is gaining momentum  with  each  passing  day.  uncertainty  haunts  everyone now, and all because of  the factors of  uncertainty, which are increasing with each passing  day.  these  factors  destabilize  confidence  in  the  economy, business, and the international market.  but every business needs this confidence because  the  company’s  stability  guarantees  its  existence  and is the foundation for the future development  of the manufacturing enterprise. the  conditions  of  uncertainty  include  any  changes,  usually  in  the  external  environment  of  the manufacturing enterprise, to which it is almost  urgent to adapt. in the world, there are many factors  figure 2. factors affecting staff competitiveness. source: authors’ own development. competitiveness of staff sociopsychological climate of the enterprise formation and development of a strong and healthy corporate culture stimulation of innovations, active innovation policy efficiency of the enterprise as a whole increase productivity improving the system of motivation and remuneration, evaluation and development of employees providing an appropriate social package for employees int. j. prod. manag. eng. (2022) 10(2), 225-235creative commons attribution-noncommercial-noderivatives 4.0 international the factors of competitiveness management of manufacturing enterprise personnel in conditions of uncertainty 229 http://creativecommons.org/licenses/by-nc-nd/4.0/ and  conditions  of  uncertainty,  and  it  is  suggested  to note  the main ones, figure 3. one of  the most  severe  and  urgent  factors  of  uncertainty  today  is  the covid-19 coronavirus pandemic and  the war  in ukraine, leaving a fatal and inevitable mark on  history and causing a critical situation in the world  economy. in this context, it  is advisable to consider in more  detail covid-19, which forced the world to become  more digitalized. modern manufacturing companies  survived the crisis and could only survive those who  could effectively manage the competitiveness of their  staff and transfer business processes online. during  covid-19,  managers  were  forced  to  encourage  employees to expand their capabilities and do any  variety of work if necessary - and employees adapted  to these changes, understanding the critical situation  and the needs of enterprises. the staff used their capabilities in a new way. this  crisis  during  the  pandemic  showed  that  by  giving  employees a chance in a difficult economic situation  to support the needs of the enterprise, employees use  their potential and show themselves to the best of their  ability. employees began to use their opportunities to  the maximum, and managers would never know the  employee’s full potential without giving him such an  opportunity (social report, 2021). as  a  result  of  covid-19,  many  manufacturing  companies have been challenged to rapidly change  working  conditions  for  staff  in  the  context  of  digitalization. as a result, managers were forced to  make a large number of new implementations in a  brief period, namely: review  and  change  the  load  on  different  departments of the enterprise in connection with  changing needs of the population; dismiss  staff  from  some  positions  that  are  not  relevant during the pandemic; hire more people in the delivery department due  to the heavy workload; monitor the health of staff and prevent the spread  of the disease; organize safety in the workplace by conducting  periodic  disinfection  and  providing  staff  with  masks, gloves, antiseptics, etc.; ensure the implementation of established plans,  even if most employees are ill; to  train  staff  to  work  in  new  conditions  of  digitalization by establishing round-the-clock it  support; organize  online  courses  for  those  who  find  it  challenging to master online work; provide  staff  with  appropriate  conditions  for  online work (create online databases and online  services and provide staff with proper gadgets); transfer staff to work online; ensure  stable  and  preferably  accessible  communication of staff with managers; organize periodic meetings and rallies to discuss  complaints and suggestions for new theories of  personnel management; monitor the efficiency of the personnel at home; make many reports to monitor the effectiveness  of  the  latest  personnel  management  theories  constantly. figure 3. the main current factors of uncertainty. source: authors’ own development. economic crisis migration of population inflation covid-19 pandemic martial law in ukraine natural disasters opaque tax system digitalization int. j. prod. manag. eng. (2022) 10(2), 225-235 creative commons attribution-noncommercial-noderivatives 4.0 international kovtunenko et al. 230 http://creativecommons.org/licenses/by-nc-nd/4.0/ we  can  conclude  that  managers  had  to  reconsider  their views on outdated management methods and  act  quickly  using  the  latest  management  theories.  currently, managers face a new challenge – the need  to restore staff competitiveness following the effects  of  covid-19  and  adapt  to  the  new  conditions  of  uncertainty  that  are  presently  suffering  from  hostilities in ukraine. the  martial  law  in  ukraine,  which  began  on  february 24, 2022, due to the invasion of the russian  federation,  destroyed  the  economic  component  of  ukraine and had a tremendous negative impact on the  world economy. due to a large number of occupied  cities,  destroyed  infrastructure,  and  refugees  to  neighbouring  countries,  many  small  and  mediumsized  businesses,  especially  in  the  manufacturing  sector,  were  forced  to  close  their  businesses.  the  martial law in ukraine creates new problems for the  global economic environment and risks hampering  the  restoration  of  confidence  in  investment  in  ukraine’s manufacturing enterprises. the war has also affected international manufacturing  companies  with  the  imposition  of  large-scale  economic sanctions, which have restricted access to  energy resources and thus raised prices for materials,  manufacture and, as a  result,  the  transportation of  products around the world. due to the global increase  in  prices,  managers  of  manufacturing  enterprises  were forced to reduce the staff or increase the amount  of work while maintaining pre-war wages or even  reducing them. due  to  the  hostilities  in  ukraine,  a  large  number  of people (approximately 4.4 million) were forced  to leave the country and leave their jobs. it means  that  ukraine’s  financial  security  has  significantly  decreased  due  to  the  migration  of  the  working  population abroad. because all the money received  by  ukrainian  migrants  remains  in  the  country  of  migration and does not raise the economy of ukraine,  the cabinet of ministers of ukraine adopted several  legislative acts to stabilize the situation and reduce  the consequences of the war. given the above, we offer the following ways and  methods of organizing personnel management of a  manufacturing enterprise in a state of war: if possible, provide housing for displaced persons  as one of the types of assistance and employment  bonuses; hire a psychotherapist or apply to the appropriate  company to provide psychological assistance to  employees and do so regularly; offer  health  insurance,  which  must  cover  the  services of a psychotherapist; constantly organize webinars and psychological  training; transfer to remote work all employees who have  the  opportunity  to  work  from  home. after  all,  it will save jobs even when a person has been  forced to go abroad and reduce travel, energy and  safety of workers; provide  employees  with  the  constant  communication so that employees have access to  contact in case of emergency; reduce the working day. speaking of international manufacturing companies,  we offer the following recommendations: simplify obtaining the right to work for migrants  from ukraine; facilitate  the  bureaucracy  process  and  reduce  time and procedures. society  must  understand  that  the  consequences  of  hostilities  in  ukraine  have  already  reached  the  whole  world.  in  case  of  prolongation,  the  conflict  will  only  intensify  in  the  territory  of  ukraine  and  all  over  the  world.  therefore,  following  the  proposed recommendations, personnel managers at  manufacturing  plants  will  be  able  to  stabilize  the  situation in the uncertain environment. managers  are  changing  organizational  thinking,  overcoming  uncertainty,  and  investing  in  building  trust with staff to develop an action plan to restore  staff  competitiveness  that  will  provide  a  solid  foundation for the future in today’s uncertainty. the  primary purpose of developing the latest theories of  personnel management should be – a shift in thinking:  from  today  to  tomorrow.  the  effectiveness  of  managers’ use of the basic principles of development  of  the  latest  theories  of  personnel  management  in  conditions  of  uncertainty  will  depend  on  the  competitiveness of personnel and the growth of the  manufacturing enterprise in general (figure 4). all the principles in the diagram are indicated in the  same percentage (25% out of 100%). that is, all the  principles  are  equally  important  and  not  applying  only  one  of  them  reduces  the  effectiveness  of  int. j. prod. manag. eng. (2022) 10(2), 225-235creative commons attribution-noncommercial-noderivatives 4.0 international the factors of competitiveness management of manufacturing enterprise personnel in conditions of uncertainty 231 http://creativecommons.org/licenses/by-nc-nd/4.0/ personnel management in conditions of uncertainty  by as much as 25%. that is why modern managers  should  review  management  policies  and  make  necessary changes based on these principles. in  the  modern  sense,  staff  competitiveness  is  a  concept  that  reveals  the  nature  of  the  processes  of  competition  from  the  standpoint  of  advantage  and  leadership.  the  latest  theories  of  personnel  competitiveness management have long been studied  and developed by various authors. first, however, it  is necessary to pay attention to competitiveness as  a  purely  practical  phenomenon,  as  long-term  and  sustainable development of the enterprise will be a  reward for ensuring the competitiveness of staff in  the context of digitalization (dobrianska, 2017). on the way to adapting to uncertainties, managers  who  know  their  business  must  inspire  their  staff  to  overcome  all  obstacles  during  this  period.  but  excellent personnel management requires even more  followers - and trust nurtures devotion. as a result,  many  managers  have  amassed  a  significant  bank  of  faith,  skillfully  overcoming  the  first  crazy  and  unpredictable stages of the crisis in the context of  digitalization (dobrianska, 2015). while some may  think of  trust as an abstract, unearthly concept,  in  reality, it is a concrete foundation needed to validate  strong  relationships  with  stakeholders  in  effective  governance. two attributes of trust are fundamental  in this regard. first,  trust  is  an  actual  exchange  of  values.  trust  is valuable only in interaction with others when its  meaning is revealed, for example, with customers,  suppliers, employees, investors and team members.  similarly, trust is built only in a relationship where  actual compromises provide mutual value. trust also  promotes growth: if invested wisely and prudently,  it  grows  through  the  confirmation  of  repetitive  experiences;  poor  investment  in  trust  quickly  depreciates.  in  addition,  research  shows  that  trust  also yields economic growth and shareholder value,  increased innovation, increased social stability and  even improved health. second, trust is most effective  in many ways. trust between stakeholders is formed  in  four  areas:  physical,  emotional,  financial  and  digital (figure 5). it  is  worth  noting  that  trust  begins  to  form  on  a  human,  interpersonal  level.  covid-19  has  increased  the  sensitivity  of  stakeholders  to  these  four parameters and opened up more opportunities  for  action  to  strengthen  or  lose  staff  confidence  in  their  manager  in  digitalization.  for  example,  employees’  trust  in  management  may  arise  when  managers  think  carefully  about  reorganizing  a  manufacturing  enterprise’s  manufacture  process  through  digitalization  (to  provide  individual  staff  with appropriate devices and transfer them to remote  work, so it will be possible to reconfigure the space  on-premises  for  another  team,  taking  into  account  social distancing). another  example  of  trust  management,  where  managers  do  their  best  to  save  as  many  jobs  as  possible,  valuing  their  staff  and  their  contribution  to the company, rather than laying off most workers  figure 4. the share of the basic principles of development of the latest theories of personnel management in conditions of uncertainty. source: authors’ own development. int. j. prod. manag. eng. (2022) 10(2), 225-235 creative commons attribution-noncommercial-noderivatives 4.0 international kovtunenko et al. 232 http://creativecommons.org/licenses/by-nc-nd/4.0/ to save profits in the short term, is not aware of the  dire consequences. actions in the future. similarly,  customers  can  build  trust  when  businesses  add  additional  security  measures  to  protect  customer  data from cyber threats. after all, in the context of  global digitalization, cybersecurity is also becoming  a  significant  issue  for  managers,  which  requires  finance and constant monitoring. after analysing figure 4, we can safely say that staff  trust is a multidimensional concept. if the manager  wants  to  unleash  the  potential  of  his  staff  to  the  maximum, he must ensure the staff’s trust in him,  taking into account all four points of faith. after all,  to  effectively  manage  a  team’s  competitiveness  in  conditions of uncertainty, which affects the overall  management  of  the  enterprise,  adapting  to  these  conditions, a professional manager must understand  that staff need support and ensure the four dimensions  of human trust. to do this, effective managers need  to consider the following issues: what parameters of the company’s activities in  conditions  of  uncertainty  are  most  important  for  each  of  the  stakeholders,  and  what  will  be  essential for them when we move to the result of  the prosperity of our business? do we communicate our intentions and actions to  staff clearly and transparently, even if we do not  have answers? can we competently fulfil what we promise our  staff? how do we track and measure our progress in  meeting stakeholder needs in the four dimensions  of trust? managers of the enterprise, who will be able to show  loyalty to their staff even in difficult times of crisis  and thus confirm that human resources are significant  to  them, will  leave the company in high positions  in  the  market.  manufacturing  companies  that  are  focused  on  people  (on  their  staff)  and  improving  working  conditions  can  count  on  the  appropriate  loyalty  and  trust  not  only  of  their  employees  but  also customers. for this purpose, the manager needs  to  develop  the  newest  theories  of  management  of  competitiveness  of  the  personnel  based  on  trust,  provide  the  personnel  with  constant  professional  support and organize uninterrupted work, even in the  conditions of uncertainty (prodius, 2009). in  current  conditions,  the  latest  management  theories are becoming the foundation of leadership,  characterized by increasing staff competitiveness due  to the spread of digitalization. also, in these conditions,  the volume of  intellectualization of work increases,  which is why the presence of highly qualified staff in  the team is considered the most  important strategic  resource.  it  is  also  facilitated  by  uniting  european  states into the european union and its globalization  policy. as a result, there are changes in the workload  distribution from physical to intellectual. therefore,  more critical  in personnel management  is assessing  its  intellectual  potential.  today,  more  than  ever,  managers  are  interested  in  forming  a  reliable  and  professional staff motivated  to grow personal skills  and qualifications and improve their contribution to  achieving  the  ultimate  goals  of  the  manufacturing  enterprise (prodius and alekseev, 2015). the covid-19 pandemic, as a factor of uncertainty,  has  radically  changed  our  way  of  life.  it  forced  figure 5. the main components of staff confidence. source: authors’ own development. int. j. prod. manag. eng. (2022) 10(2), 225-235creative commons attribution-noncommercial-noderivatives 4.0 international the factors of competitiveness management of manufacturing enterprise personnel in conditions of uncertainty 233 http://creativecommons.org/licenses/by-nc-nd/4.0/ us  to  take  a  step  towards  the  digitalization  of  the  personnel of modern enterprises. nevertheless,  the  opportunity hypothesis in a crisis remains as accurate  as ever. history has repeatedly shown that problems  stimulate  innovation  and  the  development  of  the  latest  management  theories.  whether  technical,  scientific or product related  to business models or  public  institutions,  all  such  innovations  have  one  thing in common: they solve problems. the most striking example of the use of innovation  in conditions of uncertainty is  the development of  digitalization during a pandemic as a major factor  in  staff  competitiveness.  covid-19  accelerates  digital transformation in “seven-mile increments”. in  response to restrictions imposed to slow the spread of  the virus, businesses have switched to digitalization  at  a  rate  far  above  that  achieved  by  all  previous  corporate  investments  and  government  quarantine  programs.  although  the  logic  of  digitalization  remains the same as in the past, three reasons behind  this active perception come to mind: as always, demand is determined by customer  needs.  when  there  is  no  alternative  to  online  services,  customers  often  focus  on  solutions  that support some semblance of normal business  conduct and are willing to accept them, even if  they are still developing. opportunities  for  supply  are  increasing,  so  companies  are  focused  on  the  future  and  act  quickly. as a result, minor issues are deprived of  priority, allowing companies to work much more  efficiently as bureaucracy is eliminated suddenly. regulators  create  encouraging  conditions  for  transparent  business.  governments  intervene  through laws, regulations, and support programs,  prioritizing crisis response through digitalization.  everyone adopts these regulations because now  the  public  benefit  to  society  is  placed  above  individual considerations. 4. conclusion after  analysing  the  impact  of  uncertainty  on  the  management  of  competitiveness  of  manufacturing  personnel, we can say that managers today need to be  more decisive than ever and make quick decisions.  due to the constantly changing economic situation  in the world, manufacturing companies must meet  the requirements of today to continue to conduct and  develop their business. it is appropriate to note that manufacturing companies  in  all  countries  must  understand  the  causal  links  between conditions of uncertainty (such as martial  law in ukraine or the covid-19 pandemic) and the  world economic situation. as well as must realize  that joining forces will help to cope with most factors  of uncertainty to establish the financial process in the  world  and  the  organization  of  stable  management  of  staff  competitiveness,  based  on  international  standards and principles. managers  must  always  consider  human  resource  management’s  peculiarities,  namely  their  psychological component. up to it depends on the  personnel  and  the  enterprise’s  competitiveness.  therefore,  enterprises  must  take  appropriate  measures  to  maintain  the  psychologically  healthy  condition of  the staff. examples are psychological  pieces  of  training,  webinars,  seminars,  etc.  furthermore,  enterprises  should  organize  constant  psychological assistance to employees. as a result,  the  potential  of  the  staff  will  be  revealed,  the  competitiveness of the team will increase, and as a  result, the profit of the manufacturing enterprise will  increase.  managers  should  always  keep  humanity  and  friendliness  when  organizing  the  process  of  managing  the  competitiveness  of  manufacturing  staff.  the  trust  of  staff,  and  as  a  consequence  of  customer  trust,  is  based  on  the  reputation  of  the  manufacturing enterprise, i.e., its methods and ways  of personnel management. in  addition,  speaking  of  digitalization,  it  should  be noted that the capabilities of man and machine  are  most  productively  used  in  the  development  of  systems  in  which  people  and  machines  work  together, complementing each other’s strengths and  balancing  each  other’s  limitations.  therefore,  the  personnel  management  scheme  will  be  the  most  effective  because  the  importance  of  information  technology and digitalization should not be seen as  an obstacle but as an opportunity for improvement. references boginya,  d.p.  (2003).  competitiveness of the labour force in the system of social and labour relations.  institute  of  economics. 213 p. [in ukrainian]. int. j. prod. manag. eng. (2022) 10(2), 225-235 creative commons attribution-noncommercial-noderivatives 4.0 international kovtunenko et al. 234 http://creativecommons.org/licenses/by-nc-nd/4.0/ buchynska, t.v. (2016). scientific bulletin of uzhhorod national university. series: international economic relations and  the world economy, 10(1). 74-77. [in ukrainian]. dobrianska, n.a. (2015), theoretical foundations of enterprise competitiveness management. investments: practice and experience. pp. 84-87. [in ukrainian]. dobrianska,  n.a.  (2017).  scientific  -  theoretical  foundations  of  the  competitiveness  of  an  enterprise.  theoretical, methodological and practical aspects of enterprise competitiveness, monograph. pp. 29-37. [in ukrainian]. dunska, a.r. (2012). international competitiveness management at national enterprises on innovative grounds, econom. enterp. managm., 7(133), 104-109 [in ukrainian] filyppova, s.v., kochevoi, m.m., androsenko, v.d. (2019). strategic priorities for reinforcing the competitiveness of  ukraine in the context of the european integration. economy: the realities of time. scientific journal, 1(41), 57-65.  available at https://economics.opu.ua/files/archive/2019/no1/57.pdf halaz, l.v. (2008). the role of competitive labour potential in the innovative development of the enterprise. bulletin of the national university “lviv polytechnic”. 162-169. available at http://vlp.com.ua/files/25_24.pdf kovalenko, o.m., stanislavyk, o.v.  (2020). preconditions of ensuring  the competitiveness of modern manufacturing  enterprise. (p. 113-118). the world economy and international economic relations. international scientific collection.  volume iii. scientific editors: y. kozak, t. shengelia, a.gribincea. kyiv: cul, 126p. available at https://ibn.idsi.md/ sites/default/files/imag_file/world_economy_valume_3_2020_kozak_%25d0%2592%25d0%25a1%25d0%2595_ %25d0%25905.pdf kovtunenko, k.v. (2013). the laws and the main dominants of machine-building enterprises development in conditions  of  strategic  changes.  economic annals-xxi.  scopus, 5-6(1),  pp.  75-78 available  at  http://soskin.info/ea/2013/56/201326.html kozak, k.b. (2011). research of personnel management problems at modern enterprises. edition 4 pp. 52-55. [in ukrainian] marchenko, v.m. (2015). scientific support of enterprise competitiveness. national aviation university. economic sciences.  “young scientist”. edition 1 (16), 15-17. [in ukrainian]. martynenko, v.m., dreval, yu. d., konotoptseva, yu. v. et al. (2013). modern technology for assessing the staff and human  resources of the organization and its socio-psychological aspect: science. development: k.: napa. 52 p. available at  https://mmgh.kname.edu.ua/images/gayduchenko/gaydu_1.pdf [in ukrainian]. nesterenko l.o., chernyakova a.i. (2011) the role of staff in increasing the competitiveness of the enterprise. url: http:  //www.rusnauka.com/8_nnd_2011/economics/10_81252.doc.htm [in ukrainian]. prodius, i.p. (2009). the main mechanisms of personnel management  in a crisis.  proceedings of odessa polytechnic university, 1(31), 181-184. [in ukrainian]. prodius, o., alekseev, m. (2015). motivation of staff as the main factor in improving the efficiency of the enterprise.  scientific bulletin of odessa national economic university, 12, 188-200. [in ukrainian]. the social enterprise in a world disrupted. leading the shift from survive to thrive (2021). deloitte global human capital  trends:  special  report.  available  at  https://www2.deloitte.com/content/dam/insights/us/articles/6935_2021-hctrends/di_human-capital-trends.pdf int. j. prod. manag. eng. (2022) 10(2), 225-235creative commons attribution-noncommercial-noderivatives 4.0 international the factors of competitiveness management of manufacturing enterprise personnel in conditions of uncertainty 235 https://economics.opu.ua/files/archive/2019/no1/57.pdf http://vlp.com.ua/files/25_24.pdf https://ibn.idsi.md/sites/default/files/imag_file/world_economy_valume_3_2020_kozak_%25d0%2592%25d0%25a1%25d0%2595_%25d0%25905.pdf https://ibn.idsi.md/sites/default/files/imag_file/world_economy_valume_3_2020_kozak_%25d0%2592%25d0%25a1%25d0%2595_%25d0%25905.pdf https://ibn.idsi.md/sites/default/files/imag_file/world_economy_valume_3_2020_kozak_%25d0%2592%25d0%25a1%25d0%2595_%25d0%25905.pdf http://soskin.info/ea/2013/5-6/201326.html http://soskin.info/ea/2013/5-6/201326.html https://mmgh.kname.edu.ua/images/gayduchenko/gaydu_1.pdf https://www2.deloitte.com/content/dam/insights/us/articles/6935_2021-hc-trends/di_human-capital-trends.pdf https://www2.deloitte.com/content/dam/insights/us/articles/6935_2021-hc-trends/di_human-capital-trends.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.3031 received 2014-05-29 accepted 2014-06-05 collaborative networks in industry and the role of pro-ve camarinha-matos, l. m. new university of lisbon, faculty of sciences and technology – campus de caparica 2829-516 monte caparica – portugal. cam@uninova.pt editorial most industrial enterprises are going under a big pressure, having to cope with continuously and rapidly changing market conditions and related business environments (alberts, 2011; wef, 2013). the accumulated effects of a number of factors such as the acceleration of the globalization, changes in regulations for environmental protection and working conditions, more demanding quality standards, economic crisis in some regions, demographic shifts, and fast technological evolution, among others, led to what is often called market turbulence. under these conditions, the threats to business sustainability lead to higher levels of risk; furthermore, trends show that unexpected disruptive events are increasing in frequency and in their effects (camarinha-matos, 2014). collaboration is often pointed out as a mechanism to facilitate agility and resilience, and thus a way to mitigate the effects of disruptions (camarinhamatos et al., 2009; peters, 2010). by dynamically combining the best fitting set of competencies and resources, communities of enterprises can be reshaped in different organizational forms, in order to cope with unexpected changes and disruptions, while also pursuing new business opportunities. for instance, in supply chains, an increased level of visibility along the chain, which is inherent to collaboration, can help enterprises to quickly adjust to demand fluctuations and disruptions (peters, 2010). on the other hand, advances in ict, and more specifically on internetrelated technologies, have induced or enabled new organizational forms such as the extended enterprise, virtual enterprise, virtual organization, business ecosystem, and many others, materializing different cases of collaborative networks and constituting highly interconnected and dynamic value chains. in connection with these organizational forms, and also led, to some extent, by a technology push, new business models emerged. in this context, the pro-ve (www.pro-ve.org) series of working conferences on virtual enterprises have been playing a major role along the last 15 years in terms of knowledge sharing, identification of research and development needs, and setting the trends. knowledge sharing and multi-disciplinary convergence. although topics related to collaborative networks appear in many recent conferences, pro-ve established its long-term position as a well-focused conference in the area. one of its most relevant characteristics, since the beginning, has been the effort to welcome contributions from a multidisciplinary nature and promote fruitful interactions among researchers and practitioners from different backgrounds. as such, pro-ve is designed and organized to offer an effective opportunity to synergistically combine contributions from computer science, manufacturing, engineering, economics, management, and socio-human communities. such approach contributes to the creation of a more holistic understanding of the challenges, as a basis to reach more comprehensive solutions. identifying needs and setting the trends. an analysis of the focus of each annual edition of the conference (figure 1) and corresponding proceedings shows the role of pro-ve in promoting the identification of gaps, trends, and research challenges in this scientific area. in fact, several relevant international research roadmaps have been published and discussed in 53int. j. prod. manag. eng. (2014) 2(2), 53-56creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2014.3031 http://www.pro-ve.org http://creativecommons.org/licenses/by-nc-nd/4.0/ these events. as such, the conference has been contributing to help the research community defining and promoting its research agenda. as briefly illustrated in figure 1, various phases can be identified along the time line: virtual enterprise support infrastructures. in the early phases of pro-ve, most emphasis was put on ict infrastructures to support single networks such as virtual enterprises / virtual organizations. information sharing and federation, secure communications, and distributed business processes management / inter-organizational workflows were among the most addressed topics. business ecosystem and virtual organizations breeding environments (vbe). the need to support rapid formation of agile virtual enterprises in response to business opportunities, coping with the challenges of trust building and achieving preparedness for collaboration, led to the emergence of long-term strategic networks. these organizational forms constitute proper breeding environments for the creation of dynamic goal-oriented networks. understanding the organizational and business models, supporting the various stages of the life cycle of different types of networks, and developing models and systems for the management of vbes and dynamic consortia formation in such context, became relevant research topics. emergence of collaborative networks as a discipline. as the number and variety of forms of networked organizations increased, and empiric knowledge on those networks accumulated, the research community started to feel the need to better organize related knowledge and activities as a new scientific discipline (camarinha-matos and afsarmanesh, 2005; kühnle and dekkers, 2012). the first manifesto for launching the discipline of collaborative networks was presented in the 2004 edition. theoretical foundation and science base. as part of the consolidation of the area, and boosted by the aim of establishing the new discipline, the following years encompassed several dedicated efforts to create a sounder scientific and engineering basis for collaborative networks. attention was directed to modelling the various aspects of collaborative networks, with particular relevance to the establishment of reference models such as arcon. a closer look at other relevant disciplines and an effort to integrate and synthetize contributions from those disciplines was carried out and discussed in various pro-ve editions, with more emphasis since 2007. a particularly relevant stream of research started by looking into the “soft computing” area in order to find suitable approaches for modelling aspects related to human behaviour in collaborative organizations and to handle the issues of decision making and behaviour management in the contexts of incomplete and imprecise knowledge. pervasiveness of collaborative networks. the consolidation of the area and the fast progress on ict support led to an expansion of the collaborative networks paradigm to multiple application areas. besides the “traditional” sectors represented in advanced supply chains, virtual enterprises, virtual organizations, virtual teams, and their breading environments, new forms of collaborative structures started to emerge in all sectors of the society. examples can be found in e-government, intelligent transportation systems, collaborative virtual laboratories, agribusiness, elderly care, health care, education, collaborative 2012 bournemouth uk 2000 florianopolis brazil 2002 sesimbra portugal 2003 lugano switzerland 2004 toulouse france 2005 valencia spain 2006 helsinki finland 2007 guimarães portugal 2008 poznan poland 2009 thessaloniki greece 2010 st. etienne france 2011 são paulo brazil 1999 porto portugal 2013 dresden germany 2014 amsterdam netherlands infrastructures for virtual enterprises e-business and virtual enterprises collaborative business ecosystems and virtual enterprises processes and foundations for virtual organizations virtual enterprises and collaborative networks collaborative networks and their breeding environments network-centric collaboration and supporting frameworks establishing the foundation of collaborative networks pervasive collaborative networks leveraging knowledge for innovation in collaborative networks collaborative networks for a sustainable world adaption and value creating collaborative networks collaborative networks in the internet of services collaborative systems for reindustrialization collaborative systems for smart networked environments figure 1. setting the trends on collaborative networks 54 int. j. prod. manag. eng. (2014) 2(2), 53-56 creative commons attribution-noncommercial-noderivatives 4.0 international camarinha-matos, l. m. http://creativecommons.org/licenses/by-nc-nd/4.0/ logistics networks, etc. by promoting the sharing of experiences from all these sectors, pro-ve has been contributing not only to enlarge the scope of the area, but also to consolidate concepts and terminology, which is fundamental to facilitate cross-sector and open innovation. sustainability and societal challenges. more recently, and in line with major worldwide trends and challenges raised by the economic crisis, pro-ve editions have been motivating the research community to more pro-actively apply the collaborative networks paradigm to critical societal challenges. aiming to reach a sustainable world calls for a wider collaboration among multiple stakeholders from different origins, as the changes needed for sustainability exceed the capacity and capability of any individual entity. collaboration thus plays an important role. examples of addressed issues include carbon-efficient value networks, active ageing, rescue and humanitarian organizations, collaborative health networks, and environment monitoring and management. another relevant discussion focus is re-industrialization, which appears as the economic, social, and political process of organizing resources for the purpose of re-establishing / revitalizing industries in order to reinvigorate the economy. collaboration is essential here, especially to small and medium enterprises in order to acquire critical mass, reach new markets, and leverage skills. but the reindustrialization cannot simply follow the steps of past century. new perspectives of industry are needed, as exemplified by issues widely discussed in recent editions of pro-ve: (i) focusing on service-enhanced products; (ii) addressing the full life cycle of products, including refurbishing / retrofitting and recycling; (iii) taking on board the concerns of energy saving and reduction of ecological footprint; and (iv) having a glocal perspective, while relying on co-innovation and co-evolution. extended networked environments. fast progress in pervasive computing and areas such as internet of services, internet of things, cyber-physical systems, and smart environments, create new business opportunities and open new possibilities for expanding collaborative networks. collaborative systems provide a promising basis for smart networked environments wherein humans, organizations, intelligent agents, and devices collaborate. modelling, design, and development of collaborative systems of systems is likely to support a large number of novel applications in areas including security, transportation, construction, sustainability and energy management, education, government, and manufacturing. building a community. an important outcome of the pro-ve series of conferences was the gradual building of a research community on collaborative networks. the focused scope of the events facilitates the interactions among participants, even if they are coming from different backgrounds. complemented with a deliberate effort to promote mutual respect among the contributing disciplines, the focus and aims of the conference facilitated the creation of collaborative bonds among many participants. this is reflected in the fact that a good percentage of attendants participate in consecutive editions and that many collaborative projects started among them. in 2007 the informal community gathered around pro-ve launched a more formal organizational structure – socolnet, the international society of collaborative networks (www.socolnet.org). this society, which currently counts members in 48 countries, is a not for profit research association, that aims at promoting and stimulating scientific research, education, technological development, and scientific and technical interactions among researchers in the area of collaborative networks. complementarily, pro-ve has been contributing to education. a number of curricula proposals and teaching experiences on collaborative networks have been discussed in its various editions. the conference proceedings, published by springer under the titles of the main themes shown in figure 1, and included in the ifip advances in information and communication technologies series, provide important background material for the area. being indexed in major databases such as web of science, scopus, and dblp, increased the visibility of the knowledge generated in this area. its impact is also reflected in the fact that many pro-ve papers have reached a very good number of citations. the increasing range of applications of collaborative networks, combined with the possibilities offered by new technologies (e.g. cloud computing, smart mobile devices, natural user interfaces, etc.), induce new organizational forms and new business models, all pointing to a strong and more dynamic interconnectivity, which in turn raises new research challenges. some examples include: behavioural aspects – the success and sustainability of collaboration requires better 55int. j. prod. manag. eng. (2014) 2(2), 53-56creative commons attribution-noncommercial-noderivatives 4.0 international collaborative networks in industry and the role of pro-ve http://www.socolnet.org http://creativecommons.org/licenses/by-nc-nd/4.0/ understanding of the involved behavioural aspects, which will provide a basis for the development of sounder governance principles and support tools. multiplex networks – more and more, complex applications require the involvement and interplay of multiple networks. for instance, in the area of service-enhanced products (or product-service systems), various collaborative networks need to be involved, namely for product manufacturing, creation or co-creation of business services that enhance the product, service provision along the life cycle of the product, involvement of the customer and other local stakeholders close to the customer in the process of co-creation / coinnovation, etc. furthermore, enterprises can be involved in multiple business communities, with different degrees of membership. it is also necessary to consider the co-existence of formal and informal networks. risks and complexity – particularly in turbulent environments, it is necessary to deal not only with endogenous risks (due to misalignments of values and strategies of network members), but also with exogenous ones (terrorism, natural disasters and occurrences, acceleration of globalization, demographic shift, cyber risks, etc.). interconnected worlds fast progress towards smart environments, i.e. context sensitive systems in which the physical and the cyber worlds are interwoven through seamless integration of sensors, actuators and other everyday objects, progressively enriched with computational and decision making power, and interconnected through networks. a challenge for the pro-ve / socolnet community is to keep up with these new requirements and work together to devise adequate solutions in the near future. references alberts, d. s. (2011). the agility advantage – a survival guide for complex enterprises and endeavors. dod, isbn 978-1-893723-23-8, http://www.dodccrp.org/files/agility_advantage/agility_advantage_book.pdf (access 28 may 2014). camarinha-matos, l. m. (2014). collaborative networks: a mechanism for enterprise agility and resilience. in enterprise interoperability vi (k. maertins et al. eds.). springer, pp. 3-11. doi:10.1007/978-3-319-04948-9_1 camarinha-matos, l. m., afsarmanesh, h. (2005). collaborative networks: a new scientific discipline. j. intelligent manufacturing, 16(4-5): 439-452. doi:10.1007/s10845-005-1656-3 camarinha-matos, l. m., afsarmanesh, h., galeano, n., molina, a. (2009). collaborative networked organizations concepts and practice in manufacturing enterprises. computers & industrial engineering, 57(1); 46-60. doi:10.1016/j.cie.2008.11.024 kühnle, h., dekkers, r. (2012). some thoughts on interdisciplinarity in collaborative networks’ research and manufacturing sciences. j. manufacturing technology management, 23(8): 961-975. doi:10.1108/17410381211276826 peters, c. (2010). improving supply chain resilience with network centric manufacturing, dsn innovations white paper. http:// thelucrumgroup.com/documents/improvingsupplychainresliencewithncm_final_cpeters.pdf (access 25 nov 2013) wef. (2013). building resilience in supply chains. world economic forum, jan 2013. http://www3.weforum.org/docs/wef_rrn_mo_ buildingresiliencesupplychains_report_2013.pdf (access 28 may 2014) 56 int. j. prod. manag. eng. (2014) 2(2), 53-56 creative commons attribution-noncommercial-noderivatives 4.0 international camarinha-matos, l. m. http://www.dodccrp.org/files/agility_advantage/agility_advantage_book.pdf http://dx.doi.org/10.1007/978-3-319-04948-9_1 http://dx.doi.org/10.1007/s10845-005-1656-3 http://dx.doi.org/10.1016/j.cie.2008.11.024 http://dx.doi.org/10.1108/17410381211276826 http://thelucrumgroup.com/documents/improvingsupplychainresliencewithncm_final_cpeters.pdf http://thelucrumgroup.com/documents/improvingsupplychainresliencewithncm_final_cpeters.pdf http://www3.weforum.org/docs/wef_rrn_mo_buildingresiliencesupplychains_report_2013.pdf http://www3.weforum.org/docs/wef_rrn_mo_buildingresiliencesupplychains_report_2013.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2022.17169 received: 2022-02-15 accepted: 2022-06-27 evaluation of supply chain risks by fuzzy dematel method: a case study of iron and steel industry in turkey asuman üstündağ a1 , sinan çıkmak b* , merve çankaya eyiol c , mustafa cahit ungan a2 a school of business, sakarya university, esentepe campus 54187, sakarya/turkey. b social sciences vocational school, duzce university, uzunmustafa st, 81600, duzce/turkey. c graduate school of business, sakarya university, esentepe campus 54187, sakarya/turkey. a1 austundag@sakarya.edu.tr, b sinancikmak@duzce.edu.tr, c cankayamerve@gmail.com, a2 ungan@sakarya.edu.tr abstract: business practices to strengthen competitiveness increase the vulnerability of supply chains to risks. risks that can adversely affect the effectiveness and efficiency of supply chain activities are events that disrupt the flow of information, materials, money, and products. therefore, supply chain risk management is vital for companies. it is necessary to identify the risks that threaten the supply chain and prioritize them. in addition, examining the effects of risks on each other will determine the success of supply chain risk management. this study evaluates turkey’s leading iron and steel company’s supply chain risk groups and sub-risks. the fuzzy dematel method was used to determine the relative importance of the risks and the effects of the risks on each other. results show that the most critical risk group is business risks. business risk is followed by customer risks, supplier risks, transportation risks, environmental risks, and, finally, security risks. this study provides originality by evaluating the supply chain risks from a broader perspective. key words: fuzzy dematel, iron and steel industry, risk assessment, supply chain risk management. 1. introduction almost every industry is exposed to increasing competitive pressure with the globalization of its business environment and markets. companies tend to follow practices such as outsourcing, overseas manufacturing, lean manufacturing, inventory reduction, and supply chain collaboration. although these practices strengthen companies’ competitiveness, they increase the vulnerability of their supply chains to risks. in other words, companies are increasingly exposed to unexpected disruptions that affect the entire supply chain (munir et al., 2020). a disruption at one stage of a supply chain will affect the entire chain and negatively affect firms’ different levels (parast & subramanian, 2021). the earthquake and the tsunami in japan in 2011 caused an interruption in supply and demand and therefore slowed production in other countries (tukamuhabwa et al., 2015). recent crises such as natural disasters and epidemics have significantly interrupted supply chain activities. the covid-19 pandemic has severely disrupted supply chains globally and locally (pujawan & bah, 2022). due to the global epidemic, companies operating in different geographies but with the same supply chain have been interrupted in their production capabilities. not only risks at to cite this article: üstündağ, a., çıkmak, s., çankaya eyiol, m., ungan, m.c. (2022). evaluation of supply chain risks by fuzzy dematel method: a case study of iron and steel industry in turkey. international journal of production management and engineering, 10(2), 195-209. https://doi.org/10.4995/ijpme.2022.17169 int. j. prod. manag. eng. (2022) 10(2), 195-209creative commons attribution-noncommercial-noderivatives 4.0 international 195 https://orcid.org/0000-0003-3529-6027 https://orcid.org/0000-0002-4704-3409 https://orcid.org/0000-0003-1255-613x https://orcid.org/0000-0003-2041-1344 http://creativecommons.org/licenses/by-nc-nd/4.0/ the worldwide level, but companies may also face machine breakdowns, exchange rate fluctuation, low supplier integration, inaccurate shipment from suppliers, inaccurate shipment to customers, and order fluctuation in their daily operational processes (dong & cooper, 2016). supply chain risk management, which includes the identification, evaluation, and management of risks in supply chain processes that are critical to business performance, has become an important area in supply chain management research (ceryno et al., 2015; dong & cooper, 2016; ho et al., 2015). while previous research concentrated on supply chain risk management (hermoso-orzáez, & garzón-moreno, 2021; ho et al., 2015; mital et al., 2018; zimmer et al., 2017), classification of risks (alora & barua, 2022; duong et al., 2022; kumar et al., 2020; oke & gopalakrishnan, 2009; rangel et al., 2015) and assessments of risks (ali et al., 2021; alora & barua, 2022; mital et al., 2018; zimmer et al., 2017), only a few studies (khan et al., 2021a; pfohl et al., 2011; sharma & routroy, 2016) have examined the relationships between risks. in practice, supply chain risks are generally related, but this situation is ignored in traditional risk management. it should be stressed that revealing the relationships between risks will help decision-makers determine appropriate mitigation strategies and achieve more successful risk management outcomes. to manage various operations of an industry, there is a need to understand the links between risks (lahane & kant, 2021). although the supply chain structures are generally similar, some sectoral differences exist. therefore, it is possible to mention that not all supply chains have the same types of risks (gurtu & johny, 2021; hermoso-orzáez, & garzón-moreno, 2021; srivastava & rogers, 2021). for this reason, it would be more beneficial to focus on a particular industry to determine the causal relationships between risks more accurately. in this context, the study focuses on the supply chain risks of turkey’s iron and steel industry. the iron and steel industry is of great importance to the overall performance of the manufacturing industry in turkey due to its high production and export potential and inputs to other sectors (kabak et al., 2016). with this background, this study seeks answers to the following research questions: rq 1. what are the main risk groups in the supply chain and the sub-risks? rq 2. what is the relative importance of the risk groups and sub-risks? rq 3. what is the interrelationship between supply chain risks? the rest of this paper is organized as follows. section 2 discusses the main concepts of the study. the fuzzy dematel method is discussed in section 3. section 4 introduces the results of the research. finally, conclusions and discussions are given in section 5. 2. literature review 2.1. supply chain risk management risk is a phenomenon that can affect the efficiency of an organization’s key processes (hopkin, 2018). a supply chain risk is damage or loss resulting from supply chain disruption. supply chain disruption is an undesirable, abnormal triggering event that occurs somewhere in or out of the supply chain (wagner & bode, 2008). ho et al. (2015) define supply chain risk as: “the probability and impact of unexpected macro and/or micro-level events that adversely affect any part of the supply chain, leading to operational, tactical or strategic level failures or irregularities.” supply chain risk management has emerged to implement various strategies and practices to manage supply chain networks. assessing the risks and reducing vulnerabilities (gurtu & johny, 2021; oturakçı & yıldırım, 2022; rajesh & ravi, 2017) will help to improve supply chain safety and performance. finding ways to mitigate the effects of supply chain risks is critical for successfully managing supply chains in a volatile environment (hachicha & elmsalmi, 2014). in recent years, supply chain risk management has received more attention to overcome threats and challenges (can saglam et al., 2020). generally, supply chain risk management is considered a phased process and consists of four steps: risk identification, risk assessment, risk mitigation, and risk monitoring (ho et al., 2015). this study discusses the first two steps of supply chain risk management. 2.2. supply chain risks identification the first step of the supply chain risk management is identifying risk types. it is essential to understand the firm’s internal and external dynamics to evaluate the potential for supply chain disruptions (sreedevi et al., 2021; srivastava & rogers, 2021). at this step, int. j. prod. manag. eng. (2022) 10(2), 195-209 creative commons attribution-noncommercial-noderivatives 4.0 international üstündağ et al. 196 http://creativecommons.org/licenses/by-nc-nd/4.0/ managers focus on recognizing and clearly defining all risks. thus, decision-makers become conscious of the events that cause uncertainty (hallikas et al., 2004). resources such as literature review, brainstorming, expert opinions, and examination of past events can be used to identify risks. in addition to qualitative methods, quantitative methods are also used to identify potential supply chain risks (ho et al., 2015). with the literature review and expert opinions, it is possible to talk about the existence of more than a hundred risks affecting businesses (venkatesh et al., 2015). to provide an overview, the supply chain risk classifications that different researchers included in their studies are shown in table 1. in some studies, risks are classified into two basic categories: internal risks and external risks. while internal risks are related to the activities in the supply chain processes, external risks consist of macro risks outside the supply chain and not under the control of the enterprise. some other studies classify supply chain risks into three categories: internal risks, external risks within the supply chain, and risks outside the supply chain. table 1. supply chain risks. authors risk types (jüttner et al., 2003) environmental, network-related, organizational risks (chopra & sodhi, 2004) disruptions, delays, information systems, forecast, intellectual property, procurement, receivables, inventory, and capacity risks (christopher & peck, 2004) internal risks: process and control risk, risks outside the firm but within the supply chain network: demand and supply risks, risks outside the supply chain: environmental risks (manuj & mentzer, 2008) supply, operational, demand, security, macro, political, competitive, and resource risks (wagner & bode, 2008) demand-side, supply-side, legal and regulatory, infrastructure, and catastrophic risks (trkman & mccormack, 2009) endogenous (internal) risks: market and technology turbulence, exogenous (external) risks; discrete events (e.g., terrorist attacks, contagious diseases, workers’ strikes, and continuous events (e.g., inflation rate, consumer price index changes) (kumar et al., 2010) internal operational risks: demand, production and distribution, supply risks, external operational risks: terrorist attacks, natural disasters, exchange rate fluctuation (tummala & schoenherr, 2011) demand, delay, disruption, inventory, manufacturing (process) interruption, capacity, supply, system, sovereign, and transportation risks (samvedi et al., 2013) supply, process, demand, and environmental risks (punniyamoorthy et al., 2013) supply, manufacturing, demand, logistics, and environmental risks (ho et al., 2015) macro risks, micro risks: demand, manufacturing, supply, and infrastructure risks (information systems, transportation, and financial risks) (rangel et al., 2015) planning: strategic, inertia, informational, capacity, and demand risks, source: supply, financial and relational risks, make: operational and disruption risks, delivery: customer risk, returns: legal risk, other: environmental and cultural risks (prakash et al., 2017) supply, demand, control, process, and environmental risks (shahbaz et al., 2019) supply, process, demand, logistics, collaboration, financial, and environmental risks (chu et al., 2020) political, environmental, financial, supply and demand, logistics, system, and operational risks (ali et al., 2021) natural, human-made, system accidents, and financial (hermoso-orzáez, & garzónmoreno, 2021) operational, direct process to the product/service, suppliers, security, and labor rights (lahane & kant, 2021) operational and technological risks, product recovery risks, supply risks, demand risks, environmental risks, economic risks, social risks (parast & subramanian, 2021) supply risks, demand risks, process risks, and environmental risks (srivastava & rogers, 2021) operational, infrastructure, legal, economic, supplier, forecasting, transportation, and labor (duong et al., 2022) external risk, time risk, supply risk, operational risk, and demand risk (oturakçı & yıldırım, 2022) supply, manufacturing, demand, financial, macro, transportation, and information risks int. j. prod. manag. eng. (2022) 10(2), 195-209creative commons attribution-noncommercial-noderivatives 4.0 international evaluation of supply chain risks by fuzzy dematel method: a case study of iron and steel industry in turkey 197 http://creativecommons.org/licenses/by-nc-nd/4.0/ when the research on supply chain risks is reviewed, it has been observed that there is no universal classification. classifications differ by industry or scope. as the content of this research is enormous, the number of risks was kept as wide as possible. accordingly, environmental, safety, customer, supplier, transportation, and business risk groups were used in this article. 2.3. supply chain risks assessment risk assessment is the second step of supply chain risk management. it is a critical step as the risk assessment affects the managerial decisions such as risk prioritization and resource allocation to mitigate risks (sodhi & tang, 2012; hallikas et al., 2004). there is a need to assess and prioritize risks to identify appropriate management actions for the risks at the focal business and supply chain levels. since it will not be possible to take measures at the same level for all supply chain risks, it is necessary to determine each risk’s impact and rank them. due to the reasons like budget, resources, time, labor constraints, etc., it is not possible to attach the same importance to all risks. hence, it is a need to handle the risks with a proper approach (oturakçı & yıldırım, 2022). ranking the risks helps managers focus on the risks that need immediate attention and choose the appropriate mitigation strategies. various methods used in risk assessment are shown in table 2. a literature review for this study showed that the dematel method is employed in some risk studies. however, to our best knowledge, there is no study using the dematel to assess supply chain risks in the iron and steel industry. the dematel method is accepted as one of the best tools for evaluating the importance and causal relationships among evaluation criteria (hsu et al., 2013). 3. methodology the fuzzy dematel method and its methodological steps are explained in this section. 3.1. fuzzy dematel method dematel (decision making trial and evaluation laboratory) is one of the multi-criteria decisionmaking methods that helps to evaluate human judgments numerically. a better understanding of the causal relationship allows for planning and solving the problems by dividing the relevant factors into cause-effect groups. also, there are no sample size limitations with the dematel method (govindan & chaudhuri, 2016). the main handicap in risk analysis is the subjectivity of the inputs obtained from the experts. therefore, it is possible to use fuzzy or gray theories to minimize this subjectivity (samvedi et al., 2013). using fuzzy logic, experts can make inferences about the problems they encounter under uncertainty and quantitatively define the bilateral relations they evaluate with verbal expressions (lin & wu, 2008). therefore, the dematel method combining it with fuzzy theory is used in this study. the steps of the fuzzy dematel method are given below (baykasoğlu et al., 2013): step 1: constructing the fuzzy direct relationship matrix. the relationships between the criteria are determined using the pairwise comparisons in table 3. here, a matrix is created with the answers obtained from each participant. the symbol z͂ij indicates the degree to which the i factor affects the j factor. in the formula z͂k=[z͂kij] k refers to each participant and should be 1≤k≤p z͂1, z͂2, z͂3…z͂p, participants show separate table 2. risk assessment methods. methods authors ahp (hermoso-orzáez, & garzónmoreno, 2021) ahpfuzzy topsis (alora & barua, 2022) anpimproved grey relational analysis (hashemi et al., 2015) ahpfuzzy dematel (mzougui et al., 2020) bayesian networks (lockamy iii & mccormack, 2009) fuzzy set theorymultiobjective mathematical programming (ji & zhu, 2012; kumar et al., 2004) fuzzy-ahp (samvedi et al., 2013) fuzzy-dematel (khan et al., 2021b) fuzzy-bwm (khan et al., 2021a) fuzzy inference system bow-tie analysis (aqlan & lam, 2015) dematel -anp (tarei et al., 2018) newsvendor model (cheong & song, 2013) hybrid petri-nets (khilwani et al., 2011) pf-ahp&pf-vikor (lahane & kant, 2021) sem-fuzzy ahp (oturakçı & yıldırım, 2022) simulation (durowoju et al., 2012) int. j. prod. manag. eng. (2022) 10(2), 195-209 creative commons attribution-noncommercial-noderivatives 4.0 international üstündağ et al. 198 http://creativecommons.org/licenses/by-nc-nd/4.0/ matrices of answers. if the number of participants is more than one, z͂=[z͂ij] matrix is formed by calculating the arithmetic averages over the answers given to calculate the direct relationship matrix. table 3. the fuzzy linguistic scale. linguistic variables triangular fuzzy numbers no influence (no) (0, 0, 0.25) very low influence (vl) (0, 0.25, 0.50) low influence (l) (0.25, 0.50, 0.75) high influence (h) (0.50, 0.75, 1) very high influence (vh) (0.75, 1, 1) 𝑍𝑍" = �̃�𝑧! + �̃�𝑧" + ⋯+ �̃�𝑧# 𝑝𝑝 (1) 𝑍𝑍" = $ 0 �̃�𝑧!" ⋯ �̃�𝑧!# �̃�𝑧"! 0 ⋯ �̃�𝑧"# ⋮ ⋮ ⋱ ⋮ �̃�𝑧#! �̃�𝑧#" ⋯ 0 + (2) step 2: calculating the normalized fuzzy direct relationship matrix the ‘u’ value, the last of the triangular numbers, is used to create this matrix. the ‘r’ value obtained with the help of equation (3) is divided by each value in the fuzzy direct relationship matrix. 𝑋𝑋" = $ 𝑥𝑥&!! 𝑥𝑥&!" ⋯ 𝑥𝑥&!# 𝑥𝑥&"! 𝑥𝑥&"" ⋯ 𝑥𝑥&"# ⋮ ⋮ ⋱ ⋮ 𝑥𝑥&#! 𝑥𝑥&#" ⋯ 𝑥𝑥&## * = $%!" & = , ℓ!" & , (!" & , )!" & .𝑟𝑟 = 𝑚𝑚𝑚𝑚𝑥𝑥!*+*#2∑ 𝑢𝑢+, # ,-! 5 ijx! (3) step 3: calculating the fuzzy total relationship matrix. in the normalized fuzzy direct relationship matrix, each triangular number group is divided into separate matrices, and using equation (5) are combined into a single matrix to form a fuzzy total relationship matrix. 𝑋𝑋ℓ = # 0 ℓ"#$ ⋯ ℓ"%$ ℓ#"$ 0 ⋯ ℓ#%$ ⋮ ⋮ ⋱ ⋮ ℓ%"$ ℓ%#$ ⋯ 0 ) , 𝑋𝑋& = # 0 𝑚𝑚"#$ ⋯ 𝑚𝑚"%$ 𝑚𝑚#"$ 0 ⋯ 𝑚𝑚#%$ ⋮ ⋮ ⋱ ⋮ 𝑚𝑚%"$ 𝑚𝑚%#$ ⋯ 0 ), 𝑋𝑋' = # 0 𝑢𝑢"#$ ⋯ 𝑢𝑢"%$ 𝑢𝑢#"$ 0 ⋯ 𝑢𝑢#%$ ⋮ ⋮ ⋱ ⋮ 𝑢𝑢%"$ 𝑢𝑢%#$ ⋯ 0 ) (4) !ℓ!" ″ " = 𝑋𝑋ℓ(𝐼𝐼 − 𝑋𝑋ℓ)#$ !𝑚𝑚!" ″ " = 𝑋𝑋%(𝐼𝐼 − 𝑋𝑋%)#$ !𝑢𝑢!" ″ " = 𝑋𝑋&(𝐼𝐼 − 𝑋𝑋&)#$ 𝑇𝑇= . �̃�𝑡$$ �̃�𝑡$' ⋯ �̃�𝑡$( �̃�𝑡'$ �̃�𝑡'' ⋯ �̃�𝑡'( ⋮ ⋮ ⋱ ⋮ �̃�𝑡($ �̃�𝑡(' ⋯ �̃�𝑡(( 4 �̃�𝑡!" = (ℓ!" ) ,𝑚𝑚!" ) ,𝑢𝑢!" ) ) (5) step 4: defuzzification. different defuzzification methods are mentioned in the literature. in this study, the defuzzification method of cfcs (converting data into crisp values) proposed by opricovic & tzeng (2003) is used. thus, triangular fuzzy numbers are converted to more net numbers. r=maxj uij, l=minj ℓij and δ=r−l (6) xℓj=(ℓij−l)/∆ (7) xmj=(mij−l)/∆ (8) xuj=(uij−l)/∆ (9) xjls= xmj /(1+xmj−xℓj) (10) xjrs= xuj /(1+xuj−xmj) (11) xjcrisp =[xjls(1−xjls)+xjrsxjrs]/[1−xjls+xjrs] (12) f͂ijcrisp =l+ xjcrisp∆ (13) 𝑇𝑇"!"# = ⎣ ⎢ ⎢ ⎢ ⎡�̃�𝑡$$ !"# �̃�𝑡$% !"# ⋯ �̃�𝑡$& !"# �̃�𝑡%$ !"# �̃�𝑡%% !"# ⋯ �̃�𝑡%& !"# ⋮ ⋮ ⋱ ⋮ �̃�𝑡&$ !"# �̃�𝑡&% !"# ⋯ �̃�𝑡&& !"#⎦ ⎥ ⎥ ⎥ ⎤ �̃�𝑡'( !"# = (ℓ'( ) ,𝑚𝑚'( ) ,𝑢𝑢'( ) )!"# (14) step 5: identifying cause and effect groups. the “d͂def ” matrix is obtained by taking the sum of rows of defuzzified t͂def total relation matrix. the matrix “r͂def ” is obtained by transposing the matrix formed by the sum of the total relationship matrix columns, d͂idef shows the effects i. factor on other factors, r͂idef shows the sum of the direct and indirect effects on factor i. d͂idef+r͂idef shows the importance of int. j. prod. manag. eng. (2022) 10(2), 195-209creative commons attribution-noncommercial-noderivatives 4.0 international evaluation of supply chain risks by fuzzy dematel method: a case study of iron and steel industry in turkey 199 http://creativecommons.org/licenses/by-nc-nd/4.0/ the element of i in the whole system, and d͂idef−r͂idef shows the net effect of i criteria. if d͂idef−r͂idef is positive, it means it is a cause, and if negative, it is an effect. step 6: calculating the threshold value and obtaining a cause and effect diagram. in the t͂def total relationship matrix, a threshold value is determined to eliminate criteria with a relatively low degree of influence. an expert can determine this threshold value, or it can be obtained by summing the values in the total relationship matrix and averaging them. to easily understand the relations between the criteria and their positions relative to each other, an impact-relationship graph can be created that d͂idef+r͂idef is placed on the horizontal axis and d͂idef−r͂idef on the vertical axis. 4. a case study on the iron and steel industry in turkey supply chain risks and their impact levels may differ depending on the industry characteristics of businesses. the iron and steel industry is of great importance for the overall performance of the manufacturing industry in turkey due to its high rate of production and export potential and provision of inputs to other sectors (kabak et al., 2016). in 2019, the turkish steel industry ranked 8th in the world and 2nd among steel producers in europe after germany (iron steel sector report, 2020). due to its connection with many industries, disruptions in the iron and steel industry, directly and indirectly, affect other sectors. the iron and steel company chosen for this study is among the top 30 companies in turkey’s top 500 industrial enterprises. it has a corporate risk policy to identify and manage risks to increase its competitive advantage by reducing losses. the company’s senior management has adopted the vision of developing a practical risk management approach throughout the company and its suppliers. a comprehensive literature review was made to identify main supply chain risks and sub-risks in this case study. this review resulted in a large number of supply chain risks. some of these risks were eliminated based on the opinions of two academicians in the field of operations and supply chain management. experts’ views at the executive level in the production, purchasing, and marketing departments were taken to determine the risks specific to the company’s supply chain. this process resulted in six risk groups and 58 risk types (table 4). table 4. identified supply chain risk group and sub-risks. risk groups risks environmental risks (r1) e1: p political uncertainty e2: pexchange rate e3: praw material price fluctuation e4: pbureaucracy e5: peconomic crisis e6: pcompetition changes e7: penergy supply e8: pregulatory security risks (r2) sc1: pnatural disaster sc2: pterrorism sc3: poccupational disease sc4: pwar sc5: pcyberattack sc7: poccupational accident customer risks (r3) c1: pinsufficient information about customer orders or demand c2: porder cancellation c3: pcustomers’ desire to expedite orders c4: pcustomers unable to make payments on time c5: pdemand uncertainty c6: pincorrect information about customer orders or demand supplier risks (r4) s1: psupplier inability to deliver materials on time s2: psupplier inability to provide material in desired quantity s3: psupplier inability to provide materials of desired quality s4: psupplier bankruptcy s5: psupplier inability to respond to different types of material demand s6: pinability to select the right supplier s7: pfailure to share order information correctly with the supplier s8: pinability to fully share order information with the supplier s9: pinsufficient storage and handling s10: plack of supplier capacity transportation risks (r5) t1: pincrease in transportation costs t2: pdelays due to railway capacity t3: pport strikes t4: pdelays due to port capacity t5: phigh transportation costs business risks (r6) b1: pinsufficient or excess capacity b2: pinformation infrastructure breakdown b3: pinability to find qualified employees b4: pinsufficient or excess inventory (table 4 continues on next page) int. j. prod. manag. eng. (2022) 10(2), 195-209 creative commons attribution-noncommercial-noderivatives 4.0 international üstündağ et al. 200 http://creativecommons.org/licenses/by-nc-nd/4.0/ risk groups risks business risks (r6) b5: pcompany’s inability to meet demand changes b6: pcompany’s inability to respond to different types of material demand b7: pmachine failure /production disruption b8: plabor absenteeism b9: phigh unit production cost b10: pcapacity cost b11: pmanagement of labor strikes and union processes b12: pinability to deliver orders on time b13: phigh labor turnover b14: ppoorly designed process b15: pinsufficient process improvements b16: pinability to retain qualified employee b17: psecurity of critical information b18: pbusiness continuity disruption b19: penvironmental pollution b20: pinternal transportation and semifinished product/ finished product stocking b21: pcorporate communications b22: pfailure to make appropriate investments then, the experts made pairwise comparisons with linguistic expressions to determine the relationships between the six risk groups. the same linguistic comparison procedure was also performed for the sub-risks. finally, risk groups and sub-risks were analyzed with the fuzzy dematel method, and the results were evaluated. after obtaining the pairwise comparison data on the risk groups, the following fuzzy dematel methodology steps were followed. step 1: establish the fuzzy direct relation matrix. in table 5, the binary linguistic comparisons of expert 1 about the risk groups are shown as an example. for this purpose, expert 1 used the linguistic expressions given in table 3. table 5. linguistic assessment of risk groups of the expert 1. r1 r2 r3 r4 r5 r6 r1 l h l h vl r2 l l l h vl r3 no vl l no h r4 vl no l l h r5 vl no h l h r6 l l vh h h the linguistic expressions obtained from the experts were converted into fuzzy triangular numbers. the fuzzy direct relationship matrix in table 6 was obtained by taking the arithmetic average of the experts’ assessments. step 2: calculate the normalized fuzzy direct relation matrix (table 7). step 3: calculate the total relation matrix (table 8). step 4: defuzzification. cfcs (converting data into crisp values) defuzzification method proposed by opricovic & tzeng (6), (7), (8), (9), (10), (11), (12), (13), and (14) triangular numbers were converted into crisp values, and defuzzified total relation matrix was created (table 9). step 5: identify cause and effect groups. d͂def and r͂def values, which give the row and column sums of the defuzzified total relation matrix, are calculated, and the importance level of risk groups, the cause-and-effect risk clusters are formed. step 6: calculate the threshold value and draw the cause and effect diagram. this study calculates the threshold value by taking the average values in the defuzzified fuzzy total relationship matrix, 0.440. in table 9, threshold values and higher values are marked in bold. 4.1. findings considering the analysis results of the risk groups in the study, the risk groups were divided into two clusters, as seen in figure 1 and table 10. these are the cause cluster with the positive d͂idef−r͂idef value and effect cluster with the negative d͂idef−r͂idef value. considering the d͂idef+r͂idef values, which show the degree of prominence of the criteria, it is seen that the highest valued risk group is business risks (r6) (5.945). this risk is followed by customer risks (r3) (5.529), supplier risks (r4) (5.211), transportation risks (r5) (5.206), environmental risks (r1) (5.077), and finally, security risks (r2) (4.716). looking at the d͂idef−r͂idef values, the cause-and-effect clusters of risks emerge. r1 with the highest positive d͂idef−r͂idef value (0.830) possesses the most substantial effect on others, while r3 with the smallest negative value (−0.996) is the most influenced risk group. (table 4 continues on next page) int. j. prod. manag. eng. (2022) 10(2), 195-209creative commons attribution-noncommercial-noderivatives 4.0 international evaluation of supply chain risks by fuzzy dematel method: a case study of iron and steel industry in turkey 201 http://creativecommons.org/licenses/by-nc-nd/4.0/ the fuzzy dematel steps applied for the risk groups were also carried out for the sub-risks, and a clustered values table was prepared to show d͂idef+r͂idef, d͂idef−r͂idef of the criteria and sorted according to these values. when the environmental risks in table 11 are examined, it can be seen that the most important environmental risk with the highest d͂idef+r͂idef value is the risks related to the economy (e5) (4.667). this risk is followed by e2 (4.172), e1 (4.135), e6 (3.833). however, when the d͂idef−r͂idef values are examined, it is seen that e1 is the risk that most affects other risks with a value of 0.833. this risk was followed by e2 (0.383). among the risks most influenced by the risks in this class, e6 (−0.845) took first place, while e3 (−0.429) took second place. looking at all interactions in the security risks in table 12, it can be seen that war is the highest risk with d͂idef+r͂idef value 4.443 (sc4). then, sc2 (4.383), sc5 (4.229), sc6 (4.048), sc7 (3.588), sc1 (2.791), and sc3 (2.622) follow as well. considering the d͂idef−r͂idef values, sc1 (1.030) is the risk most affects other risks, and sc6 (−0.076) is the most influenced risk. among the customer risks in table 13, the risk of c1 (2.792) was determined as the most important risk. this risk was followed by c6 (2.537) and c2 (2.446). again, while c1 (0.505) had the strongest effect on other risks, c4 (−0.511) was the most influenced risk. table 9. the total relation matrix defuzzified with cfcs. r1 r2 r3 r4 r5 c6 d͂idef d͂idef+r͂idef d͂idef–r͂idef r1 0.313 0.415 0.595 0.537 0.527 0.566 2.953 5.077 0.830 r2 0.390 0.270 0.569 0.492 0.471 0.542 2.734 4.716 0.752 r3 0.327 0.296 0.369 0.398 0.399 0.477 2.267 5.529 –0.996 r4 0.330 0.312 0.528 0.336 0.435 0.504 2.445 5.211 –0.321 r5 0.340 0.291 0.577 0.480 0.333 0.539 2.560 5.206 –0.085 r6 0.423 0.398 0.625 0.523 0.480 0.434 2.883 5.945 –0.179 r͂idef 2.123 1.982 3.262 2.766 2.645 3.062 table 6. fuzzy direct relation matrix. r1 r2 r3 r4 r5 r6 r1 0.000 0.000 0.000 0.333 0.583 0.833 0.375 0.625 0.875 0.417 0.667 0.917 0.458 0.708 0.958 0.375 0.625 0.792 r2 0.250 0.458 0.708 0.000 0.000 0.000 0.417 0.667 0.875 0.333 0.583 0.792 0.292 0.542 0.792 0.417 0.667 0.833 r3 0.208 0.375 0.625 0.125 0.292 0.542 0.000 0.000 0.000 0.208 0.375 0.625 0.292 0.458 0.708 0.417 0.667 0.875 r4 0.125 0.292 0.542 0.125 0.292 0.542 0.417 0.667 0.875 0.000 0.000 0.000 0.333 0.542 0.792 0.417 0.667 0.917 r5 0.125 0.292 0.542 0.042 0.125 0.375 0.625 0.875 1.000 0.375 0.667 0.833 0.000 0.000 0.000 0.542 0.792 0.958 r6 0.375 0.583 0.750 0.333 0.542 0.750 0.667 0.917 1.000 0.417 0.667 0.917 0.292 0.500 0.708 0.000 0.000 0.000 table 7. normalized fuzzy direct relation matrix. r1 r2 r3 r4 r5 r6 r1 0.000 0.000 0.000 0.076 0.133 0.190 0.086 0.143 0.200 0.095 0.152 0.210 0.105 0.162 0.219 0.086 0.143 0.181 r2 0.057 0.105 0.162 0.000 0.000 0.000 0.095 0.152 0.200 0.076 0.133 0.181 0.067 0.124 0.181 0.095 0.152 0.190 r3 0.048 0.086 0.143 0.029 0.067 0.124 0.000 0.000 0.000 0.048 0.086 0.143 0.067 0.105 0.162 0.095 0.152 0.200 r4 0.029 0.067 0.124 0.029 0.067 0.124 0.095 0.152 0.200 0.000 0.000 0.000 0.076 0.124 0.181 0.095 0.152 0.210 r5 0.029 0.067 0.124 0.010 0.029 0.086 0.143 0.200 0.229 0.086 0.152 0.190 0.000 0.000 0.000 0.124 0.181 0.219 r6 0.086 0.133 0.171 0.076 0.124 0.171 0.152 0.210 0.229 0.095 0.152 0.210 0.067 0.114 0.162 0.000 0.000 0.000 table 8. total relation fuzzy matrix. r1 r2 r3 r4 r5 r6 r1 0.034 0.163 1.028 0.099 0.265 1.145 0.154 0.414 1.595 0.140 0.361 1.448 0.145 0.353 1.418 0.144 0.388 1.518 r2 0.085 0.245 1.093 0.026 0.136 0.913 0.153 0.397 1.491 0.116 0.326 1.334 0.106 0.305 1.299 0.144 0.373 1.424 r3 0.069 0.196 0.953 0.048 0.169 0.904 0.052 0.207 1.149 0.081 0.242 1.151 0.095 0.246 1.133 0.130 0.319 1.264 r4 0.054 0.191 0.993 0.049 0.178 0.955 0.144 0.360 1.392 0.038 0.177 1.092 0.107 0.275 1.212 0.135 0.338 1.343 r5 0.059 0.203 0.998 0.036 0.157 0.931 0.195 0.418 1.419 0.123 0.326 1.258 0.042 0.180 1.064 0.168 0.380 1.357 r6 0.115 0.279 1.122 0.101 0.256 1.082 0.212 0.460 1.542 0.140 0.356 1.382 0.115 0.313 1.313 0.067 0.260 1.295 int. j. prod. manag. eng. (2022) 10(2), 195-209 creative commons attribution-noncommercial-noderivatives 4.0 international üstündağ et al. 202 http://creativecommons.org/licenses/by-nc-nd/4.0/ as seen in table 14, s1 with 𝐷𝑖𝑑𝑒𝑓+𝑅𝑖𝑑𝑒𝑓 value of 7.940 is the first, and s2 (7.864) is the second most important risk in the supplier risks group. s8, with a value of 0.898, is the most affected, and s3 (−0.864) has emerged as the most influenced risk. among the transportation risks, t5 (8.840) was the most critical risk, while t2 took the last place in the order of importance. while t3 was the most affected transportation risk, t5 (−1.701) was the most influenced risk (table 15). figure 1. the casual diagram of supply chain risks. r1 r2 r3 r4 r5 r6 -1 -0,8 -0,6 -0,4 -0,2 0 0,2 0,4 0,6 0,8 1 0 1 2 3 4 5 6 7d -r d+r table 10. the rank of main risk groups. rank risks d͂idef+r͂idef rank risks d͂idef–r͂idef cluster 1 r6 5.945 1 r1 0.830 cause cluster 2 r3 5.529 2 r2 0.752 3 r4 5.211 3 r5 –0.085 effect cluster 4 r5 5.206 4 r6 –0.179 5 r1 5.077 5 r4 –0.321 6 r2 4.716 6 r3 –0.996 table 11. environmental risks. rank risks d͂idef+r͂idef rank risks d͂idef–r͂idef cluster 1 e5 4.667 1 e1 0.833 cause cluster 2 e2 4.172 2 e2 0.383 3 e1 4.135 3 e8 0.230 4 e6 3.833 4 e4 0.025 5 e3 3.745 5 e5 –0.012 effect cluster 6 e4 3.425 6 e7 –0.185 7 e7 3.391 7 e3 –0.429 8 e8 3.186 8 e6 –0.845 table 12. the rank of security risks. rank risks d͂idef+r͂idef rank risks d͂idef–r͂idef cluster 1 sc4 4.443 1 sc1 1.030 cause cluster2 sc2 4.383 2 sc2 0.356 3 sc5 4.229 3 sc4 0.147 4 sc6 4.048 4 sc5 –0.076 effect cluster 5 sc7 3.588 5 sc3 –0.212 6 sc1 2.791 6 sc7 –0.585 7 sc3 2.622 7 sc6 –0.658 int. j. prod. manag. eng. (2022) 10(2), 195-209creative commons attribution-noncommercial-noderivatives 4.0 international evaluation of supply chain risks by fuzzy dematel method: a case study of iron and steel industry in turkey 203 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 13. the rank of customer risks. rank risks d͂idef+r͂idef rank risks d͂idef–r͂idef cluster 1 c1 2.792 1 c1 0.505 cause cluster2 c6 2.537 2 c6 0.368 3 c2 2.446 3 c3 0.170 4 c5 2.036 4 c2 –0.220 effect cluster5 c3 1.979 5 c5 –0.312 6 c4 1.748 6 c4 –0.511 table 15. the rank of transportations risks. rank risks d͂idef+r͂idef rank risks d͂idef–r͂idef cluster 1 t5 8.840 1 t3 0.960 cause cluster2 t1 8.699 2 t4 0.913 3 t3 7.478 3 t2 0.864 4 t4 7.428 4 t1 –1.037 effect cluster 5 t2 6.803 5 t5 –1.701 table 16. the rank of business risks. rank risks d͂idef+r͂idef rank risks d͂idef–r͂idef cluster 1 b6 5.094 1 b11 0.995 cause cluster 2 b15 5.068 2 b16 0.960 3 b14 5.008 3 b21 0.762 4 b5 4.923 4 b8 0.683 5 b22 4.618 5 b17 0.555 6 b10 4.568 6 b2 0.487 7 b1 4.567 7 b7 0.339 8 b4 4.507 8 b3 0.278 9 b12 4.503 9 b15 0.256 10 b18 4.464 10 b13 0.222 11 b3 4.360 11 b14 0.081 12 b9 4.325 12 b22 –0.005 effect cluster 13 b16 4.197 13 b19 –0.189 14 b8 4.160 14 b10 –0.262 15 b7 4.151 15 b1 –0.317 16 b20 4.140 16 b18 –0.415 17 b13 4.075 17 b6 –0.466 18 b21 3.903 18 b20 –0.567 19 b11 3.854 19 b4 –0.692 20 b2 3.606 20 b5 –0.698 21 b17 3.024 21 b9 –0.861 22 b19 2.864 22 b12 –1.145 table 14. the rank of supplier risks. rank risks d͂idef+r͂idef rank risks d͂idef–r͂idef cluster 1 s1 7.940 1 s8 0.898 cause cluster 2 s2 7.864 2 s7 0.805 3 s5 7.344 3 s10 0.641 4 s6 7.339 4 s6 0.384 5 s3 7.098 5 s4 0.035 6 s4 6.601 6 s5 –0.433 effect cluster 7 s8 6.443 7 s1 –0.445 8 s7 6.397 8 s2 –0.494 9 s10 6.380 9 s9 –0.526 10 s9 5.758 10 s3 –0.864 int. j. prod. manag. eng. (2022) 10(2), 195-209 creative commons attribution-noncommercial-noderivatives 4.0 international üstündağ et al. 204 http://creativecommons.org/licenses/by-nc-nd/4.0/ finally, when we look at the business risks in table 15, according to the value, b6 (5.094) is in the first place, while b15 (5.068) is the second. and b14 (5.008) is the third most significant risk. according to values, b11 (0.995), one of the most affected business risks, is first, while b16 (0.960) is second. b12 (−1.145) ranks first in the influenced business risk, and b9 (−0.861) ranks second. 5. conclusions and discussions identifying and assessing risks in supply chains is very important to eliminate or reduce the consequences of risks. it would be insufficient to address the risks in a supply chain one by one and assess their individual effects. evaluating each risk individually and with other risks enables managers to obtain more accurate and meaningful results. this study aims to determine the supply chain risk groups and sub-risks of a leading company operating in turkey’s iron and steel industry and examine the causal relationships between the risks. for this purpose, the pairwise comparison data about the risks obtained from the experts in the company’s purchasing, production, and marketing departments were analyzed with the fuzzy dematel method. results show that the most crucial risk group is the business risks. this is followed by customer risks, supplier risks, transportation risks, environmental risks, and, finally, security risks. business risks, also called operational risks, arise in a company’s internal product development, manufacturing, and distribution operations (deloitte, 2012). companies invest in programs such as total quality management (tqm), lean manufacturing, and six sigma to improve their quality and capabilities. however, these programs reduce the tolerance of faults and increase the negative effect of any problems (punniyamoorthy et al., 2013). therefore, it is not surprising that business risk is first in the study. also, it can be said that business operations take the most part for a greater supply chain performance (duong et al., 2022). the results showed that the environmental risk is the most influential of the other risks. this risk group has a low probability of occurring but a high impact. therefore, risks in this category drastically cause many disruptions in the supply chain (alora & barua, 2022). kumar et al. (2010) evaluated environmental risks as interaction risks and stated that the supply chain environment emerges from interaction with physical, social, legal, operational, economic, and political factors. similarly, samvedi et al. (2013) stated that environmental risks could affect a single level or organization or the entire supply chain. in our study, customer risk is most affected by the other risk groups. one of the supply chain management purposes is to achieve customer satisfaction by meeting customer needs and expectations. therefore, any disruption at any stage of the supply chain will directly or indirectly affect customer satisfaction. when the results of environmental risks are examined, it is observed that economic risk is the most critical risk. similarly, kumar et al. (2020) assessed various risks affecting demand for the indian automotive sector. they found that companies are most affected by economic risks than other risks. political uncertainty affects the other environmental risks the most. this could be because those political uncertainties in turkey harm the economy. political tensions can also affect companies within the country and, therefore, supply chain partners in other countries (mostafa et al., 2021). environmental risk analysis also indicated that competition risk is the most affected environmental risk type. regarding the security risks, it has been seen that the two most crucial security risks are terrorism and war risk. a current example of war risk is the ukrainerussia war. ukraine exports components such as iron ores, ferro-silico manganese, and pig iron that are input to the european steel industry. due to the war, there is a potential for supply chain disruption in the european iron and steel industry (world bank, 2022). a natural disaster is a risk that triggers other environmental risks. chopra & sodhi (2004) stated that natural disasters, which they consider one of the unpredictable and rare disruptions, interrupt the physical flow in the supply chain. when the importance of customer risks is assessed, insufficient information about customer orders/ demand and incorrect information about customer orders/ demand come first and second, respectively. the analysis emphasizes the importance of demandside risks since the lack of demand information is the most influential on other types of customer risks. to reduce the demand-side supply chain risks, decisionmakers should identify the factors that increase the demand risks and make appropriate process improvements to reduce them. regarding the supplier risks, the most important is that the company’s suppliers cannot deliver the materials in the desired time and quantity. in the int. j. prod. manag. eng. (2022) 10(2), 195-209creative commons attribution-noncommercial-noderivatives 4.0 international evaluation of supply chain risks by fuzzy dematel method: a case study of iron and steel industry in turkey 205 http://creativecommons.org/licenses/by-nc-nd/4.0/ literature, some authors (e.g., kumar et al., 2010; punniyamoorthy et al., 2013) indicated that supplierrelated risks would negatively affect the ability of the focal firm to meet customer demand (both in terms of quantity and quality) at the anticipated costs and at the desired time. therefore, it would be beneficial for company managers to develop reactive and proactive action plans against supplier-related risks. among the transportation risks, high transportation costs and an increase in transportation costs take the first and second places in the company. experts consider transportation costs as an essential risk that threatens the company. as disruptions in transportation operations prevent the timely supply of materials, it can disrupt the company’s production activities (paul et al., 2020). so, transportation is seen as one of the critical risks (schoen et al., 2018). transport mode, which depends on the final products’ characteristics, is a strategic variable that increases supply chain performance (oliveira et al., 2017). maritime transportation is very important in this company. therefore, port strikes were found to have the most effect on other transportation risks. on the other hand, it was determined that high transportation costs were influenced by the other transportation risks the most. according to this finding, it can be said that the company should reduce other transportation risks to reduce high transportation costs. regarding the business risks, the first risk is the inability to respond to different types of material demand. the second risk is insufficient process improvements. and the third risk is poorly designed processes. these results indicate that there are deficiencies in the company’s own internal operational processes. for this reason, the managers should identify inefficient business processes and make necessary process improvements. the most affected business risk is that orders cannot be delivered on time, proving that internal problems are reflected on the customer. this study is comprehensive as it includes many supply chain risks. the case study presented sets an example for practitioners and researchers to identify supply chain risks and assess the impact on each other. however, the lack of evaluation of experts from other companies in the iron and steel industry limits the generalization of the research findings. similar studies in different sectors can be conducted to enrich the field of supply chain risk management. in addition, studies that include risk mitigation strategies, which is the third stage of supply chain risk management, will provide significant benefits to academicians and managers working in supply chain risk management. finally, this study only used the fuzzy dematel method. in the future, hybrid techniques may be used for model creation. references ali, s.m., paul, s.k., chowdhury, p., agarwal, r., fathollahi-fard, a.m., jabbour, c.j.c., & luthra, s. (2021). modelling of supply chain disruption analytics using an integrated approach: an emerging economy example. expert systems with applications, 173, 114690. https://doi.org/10.1016/j.eswa.2021.114690 alora, a., & barua, m.k. (2022). development of a supply chain risk index for manufacturing supply chains. international journal of productivity and performance management. 71(2), 477-503. https://doi.org/10.1108/ijppm-11-2018-0422 aqlan, f., & lam, s.s. (2015). a fuzzy-based integrated framework for supply chain risk assessment. international journal of production economics, 161, 54-63. https://doi.org/10.1016/j.ijpe.2014.11.013 baykasoğlu, a., kaplanoğlu, v., durmuşoğlu, z.d.u., & şahin, c. (2013). integrating fuzzy dematel and fuzzy hierarchical topsis methods for truck selection. expert systems with applications, 40(3), 899-907. https://doi.org/10.1016/j.eswa.2012.05.046 can saglam, y., yildiz çankaya, s., & sezen, b. (2020). proactive risk mitigation strategies and supply chain risk management performance: an empirical analysis for manufacturing firms in turkey. journal of manufacturing technology management. 32(6), 1234-1244. https://doi.org/10.1108/jmtm-08-2019-0299 ceryno, p.s., scavarda, l.f., & klingebiel, k. (2015). supply chain risk: empirical research in the automotive industry. journal of risk research, 18(9), 1145-1164. https://doi.org/10.1080/13669877.2014.913662 cheong, t., & song, s.h. (2013). the value of information on supply risk under random yields. transportation research part e: logistics and transportation review, 60, 27-38. https://doi.org/10.1016/j.tre.2013.09.006 chopra, s., & sodhi, m.m.s. (2004). managing risk to avoid supply-chain breakdown. mit sloan management review, 46(1), 53-61. christopher, m., & peck, h. (2004). building the resilient supply chain. the international journal of logistics management, 15(2), 1-14. https://doi.org/10.1108/09574090410700275 int. j. prod. manag. eng. (2022) 10(2), 195-209 creative commons attribution-noncommercial-noderivatives 4.0 international üstündağ et al. 206 https://doi.org/10.1016/j.eswa.2021.114690 https://doi.org/10.1108/ijppm-11-2018-0422 https://doi.org/10.1016/j.ijpe.2014.11.013 https://doi.org/10.1016/j.eswa.2012.05.046 https://doi.org/10.1108/jmtm-08-2019-0299 https://doi.org/10.1080/13669877.2014.913662 https://doi.org/10.1016/j.tre.2013.09.006 https://doi.org/10.1108/09574090410700275 http://creativecommons.org/licenses/by-nc-nd/4.0/ chu, c.y., park, k., & kremer, g.e. (2020). a global supply chain risk management framework: an application of text-mining to identify region-specific supply chain risks. advanced engineering informatics, 45, 101053. https://doi.org/10.1016/j.aei.2020.101053 deloitte. (2012). supply chain resilience: a risk intelligent approach to managing global supply chains, https://www2. deloitte.com/global/en/pages/governance-risk-and-compliance/articles/risk-intelligent-approach-managing-supplychains.html dong, q., & cooper, o. (2016). an orders-of-magnitude ahp supply chain risk assessment framework. international journal of production economics, 182, 144-156. https://doi.org/10.1016/j.ijpe.2016.08.021 duong, a.t.b., vo, v.x., carvalho, m.d.s., sampaio, p., & truong, h.q. (2022). risks and supply chain performance: globalization and covid-19 perspectives. international journal of productivity and performance management. https://doi.org/10.1108/ijppm-03-2021-0179 durowoju, o.a., chan, h.k., & wang, x. (2012). entropy assessment of supply chain disruption. journal of manufacturing technology management, 23(8), 998-1014. https://doi.org/10.1108/17410381211276844 govindan, k., & chaudhuri, a. (2016). interrelationships of risks faced by third party logistics service providers: a dematel based approach. transportation research part e: logistics and transportation review, 90, 177-195. https://doi.org/10.1016/j.tre.2015.11.010 gurtu, a., & johny, j. (2021). supply chain risk management: literature review. risks, 9(1), 16. https://doi.org/10.3390/ risks9010016 hachicha, w., & elmsalmi, m. (2014). an integrated approach based-structural modeling for risk prioritization in supply network management. journal of risk research, 17(10), 1301-1324. https://doi.org/10.1080/13669877.2013.841734 hallikas, j., karvonen, i., pulkkinen, u., virolainen, v.m., & tuominen, m. (2004). risk management processes in supplier networks. international journal of production economics, 90(1), 47-58. https://doi.org/10.1016/j.ijpe.2004.02.007 hashemi, s.h., karimi, a., & tavana, m. (2015). an integrated green supplier selection approach with analytic network process and improved grey relational analysis. international journal of production economics, 159, 178-191. https:// doi.org/10.1016/j.ijpe.2014.09.027 hermoso-orzáez, m.j., & garzón-moreno, j. (2021). risk management methodology in the supply chain: a case study applied. annals of operations research, 1-25. https://doi.org/10.1007/s10479-022-04583-w ho, w., zheng, t., yildiz, h., & talluri, s. (2015). supply chain risk management: a literature review. international journal of production research, 53(16), 5031-5069. https://doi.org/10.1080/00207543.2015.1030467 hopkin, p. (2018). fundamentals of risk management: understanding, evaluating and implementing effective risk management (5th edition). kogan page. hsu, c.-w., kuo, t.-c., chen, s.-h., & hu, a.h. (2013). using dematel to develop a carbon management model of supplier selection in green supply chain management. journal of cleaner production, 56, 164-172. https://doi.org/10.1016/j.jclepro.2011.09.012 iron steel sector report. (2020). https://www.sanayi.gov.tr/assets/pdf/plan-program/demirçeliksektörraporu2020.pdf (in turkish) ji, g., & zhu, c. (2012). a study on emergency supply chain and risk based on urgent relief service in disasters. systems engineering procedia, 5, 313-325. https://doi.org/10.1016/j.sepro.2012.04.049 jüttner, u., peck, h., & christopher, m. (2003). supply chain risk management: outlining an agenda for future research. international journal of logistics research and applications, 6(4), 197-210. https://doi.org/10.1080/13675560310001627016 kabak, ö., ülengin, f., çekyay, b., önsel, ş., & özaydın, ö. (2016). critical success factors for the iron and steel industry in turkey: a fuzzy dematel approach. international journal of fuzzy systems, 18(3), 523-536. https://doi.org/10.1007/s40815-015-0067-7 khan, s., haleem, a., & khan, m.i. (2021a). assessment of risk in the management of halal supply chain using fuzzy bwm method. supply chain forum: an international journal, 22(1), 57-73. https://doi.org/10.1080/16258312.2020. 1788905 khan, s., haleem, a., & khan, m.i. (2021b). risk management in halal supply chain: an integrated fuzzy delphi and dematel approach. journal of modelling in management, 16(1), 172-214. https://doi.org/10.1108/jm2-09-20190228 khilwani, n., tiwari, m.k., & sabuncuoglu, i. (2011). hybrid petri-nets for modelling and performance evaluation of supply chains. international journal of production research, 49(15), 4627-4656. https://doi.org/10.1080/00207543. 2010.497173 int. j. prod. manag. eng. (2022) 10(2), 195-209creative commons attribution-noncommercial-noderivatives 4.0 international evaluation of supply chain risks by fuzzy dematel method: a case study of iron and steel industry in turkey 207 https://doi.org/10.1016/j.aei.2020.101053 https://www2.deloitte.com/global/en/pages/governance-risk-and-compliance/articles/risk-intelligent-approach-managing-supply-chains.html https://www2.deloitte.com/global/en/pages/governance-risk-and-compliance/articles/risk-intelligent-approach-managing-supply-chains.html https://www2.deloitte.com/global/en/pages/governance-risk-and-compliance/articles/risk-intelligent-approach-managing-supply-chains.html https://doi.org/10.1016/j.ijpe.2016.08.021 https://doi.org/10.1108/ijppm-03-2021-0179 https://doi.org/10.1108/17410381211276844 https://doi.org/10.1016/j.tre.2015.11.010 https://doi.org/10.3390/risks9010016 https://doi.org/10.3390/risks9010016 https://doi.org/10.1080/13669877.2013.841734 https://doi.org/10.1016/j.ijpe.2004.02.007 https://doi.org/10.1016/j.ijpe.2014.09.027 https://doi.org/10.1016/j.ijpe.2014.09.027 https://doi.org/10.1007/s10479-022-04583-w https://doi.org/10.1080/00207543.2015.1030467 https://doi.org/10.1016/j.jclepro.2011.09.012 https://www.sanayi.gov.tr/assets/pdf/plan-program/demirçeliksektörraporu2020.pdf https://doi.org/10.1016/j.sepro.2012.04.049 https://doi.org/10.1080/13675560310001627016 https://doi.org/10.1007/s40815-015-0067-7 https://doi.org/10.1080/16258312.2020.1788905 https://doi.org/10.1080/16258312.2020.1788905 https://doi.org/10.1108/jm2-09-2019-0228 https://doi.org/10.1108/jm2-09-2019-0228 https://doi.org/10.1080/00207543.2010.497173 https://doi.org/10.1080/00207543.2010.497173 http://creativecommons.org/licenses/by-nc-nd/4.0/ kumar, g., singh, r.k., jain, r., & kain, r. (2020). analysis of demand risks for the indian automotive sector in globally competitive environment. international journal of organizational analysis, 30(4), 836-863. https://doi.org/10.1108/ ijoa-03-2020-2076 kumar, m., vrat, p., & shankar, r. (2004). a fuzzy goal programming approach for vendor selection problem in a supply chain. computers and industrial engineering, 46(1), 69-85. https://doi.org/10.1016/j.cie.2003.09.010 kumar, s.k., tiwari, m.k., & babiceanu, r.f. (2010). minimisation of supply chain cost with embedded risk using computational intelligence approaches. international journal of production research, 48(13), 3717-3739. https://doi.org/10.1080/00207540902893425 lahane, s., & kant, r. (2021). evaluation and ranking of solutions to mitigate circular supply chain risks. sustainable production and consumption, 27, 753-773. https://doi.org/10.1016/j.spc.2021.01.034 lin, c.j., & wu, w.w. (2008). a causal analytical method for group decision-making under fuzzy environment. expert systems with applications, 34(1), 205-213. https://doi.org/10.1016/j.eswa.2006.08.012 lockamy iii, a., & mccormack, k. (2009). examining operational risks in supply chains. supply chain forum: an international journal, 10(1), 2-14. https://doi.org/10.1080/16258312.2009.11517204 manuj, i., & mentzer, j.t. (2008). global supply chain risk management. journal of business logistics, 29(1), 133-155. https://doi.org/10.1002/j.2158-1592.2008.tb00072.x mital, m., del giudice, m., & papa, a. (2018). comparing supply chain risks for multiple product categories with cognitive mapping and analytic hierarchy process. technological forecasting and social change, 131, 159-170. https://doi.org/10.1016/j.techfore.2017.05.036 mostafa, a.i., rashed, a.m., alsherif, y.a., enien, y.n., kaoud, m., & mohib, a. (2021, october). supply chain risk assessment using fuzzy logic. in 2021 3rd novel intelligent and leading emerging sciences conference (niles) (pp. 246-251). ieee. https://doi.org/10.1109/niles53778.2021.9600100 munir, m., jajja, m.s.s., chatha, k.a., & farooq, s. (2020). supply chain risk management and operational performance: the enabling role of supply chain integration. international journal of production economics, 227, 107667. https://doi.org/10.1016/j.ijpe.2020.107667 mzougui, i., carpitella, s., certa, a., el felsoufi, z., & izquierdo, j. (2020). assessing supply chain risks in the automotive industry through a modified mcdm-based fmeca. processes, 8(5), 579. https://doi.org/10.3390/pr8050579 oke, a., & gopalakrishnan, m. (2009). managing disruptions in supply chains: a case study of a retail supply chain. international journal of production economics, 118(1), 168-174. https://doi.org/10.1016/j.ijpe.2008.08.045 oliveira, f.l., junior, a.d.r.o., & rebelo, l.m.b. (2017). adapting transport modes to supply chains classified by the uncertainty supply chain model: a case study at manaus industrial pole. international journal of production management and engineering, 5(1), 39-43. https://doi.org/10.4995/ijpme.2017.5775 opricovic, s., & tzeng, g.h. (2003). defuzzification within a multi-criteria decision model. international journal of uncertainty, fuzziness and knowlege-based systems, 11(5), 635-652. https://doi.org/10.1142/s0218488503002387 oturakçı, m., & yıldırım, r.s. (2022). analysis of supply chain risks by structural equation model and fuzzy analytical hierarchy process. pamukkale university journal of engineering sciences, 28(1), 117-127. https://doi.org/10.5505/ pajes.2021.34119 parast, m.m., & subramanian, n. (2021). an examination of the effect of supply chain disruption risk drivers on organizational performance: evidence from chinese supply chains. supply chain management: an international journal, 26(4), 548-562. https://doi.org/10.1108/scm-07-2020-0313 paul, s., kabir, g., ali, s.m., & zhang, g. (2020). examining transportation disruption risk in supply chains: a case study from bangladeshi pharmaceutical industry. research in transportation business & management, 37, 100485. https://doi.org/10.1016/j.rtbm.2020.100485 pfohl, h.c., gallus, p., & thomas, d. (2011). interpretive structural modeling of supply chain risks. international journal of physical distribution and logistics management, 41(9), 839-859. https://doi.org/10.1108/09600031111175816 prakash, s., soni, g., & rathore, a.p.s. (2017). a critical analysis of supply chain risk management content: a structured literature review. journal of advances in management research, 14(1), 69-90. https://doi.org/10.1108/jamr-10-20150073 pujawan, i.n., & bah, a.u. (2022). supply chains under covid-19 disruptions: literature review and research agenda. supply chain forum: an international journal, 23(1), 81-95. https://doi.org/10.1080/16258312.2021.1932568 punniyamoorthy, m., thamaraiselvan, n., & manikandan, l. (2013). assessment of supply chain risk: scale development and validation. benchmarking, 20(1), 79-105. https://doi.org/10.1108/14635771311299506 rajesh, r., & ravi, v. (2017). analyzing drivers of risks in electronic supply chains: a grey–dematel approach. international journal of advanced manufacturing technology, 92(1-4), 1127-1145. https://doi.org/10.1007/s00170017-0118-3 int. j. prod. manag. eng. (2022) 10(2), 195-209 creative commons attribution-noncommercial-noderivatives 4.0 international üstündağ et al. 208 https://doi.org/10.1108/ijoa-03-2020-2076 https://doi.org/10.1108/ijoa-03-2020-2076 https://doi.org/10.1016/j.cie.2003.09.010 https://doi.org/10.1080/00207540902893425 https://doi.org/10.1016/j.spc.2021.01.034 https://doi.org/10.1016/j.eswa.2006.08.012 https://doi.org/10.1080/16258312.2009.11517204 https://doi.org/10.1002/j.2158-1592.2008.tb00072.x https://doi.org/10.1016/j.techfore.2017.05.036 https://doi.org/10.1109/niles53778.2021.9600100 https://doi.org/10.1016/j.ijpe.2020.107667 https://doi.org/10.3390/pr8050579 https://doi.org/10.1016/j.ijpe.2008.08.045 https://doi.org/10.4995/ijpme.2017.5775 https://doi.org/10.1142/s0218488503002387 https://doi.org/10.5505/pajes.2021.34119 https://doi.org/10.5505/pajes.2021.34119 https://doi.org/10.1108/scm-07-2020-0313 https://doi.org/10.1016/j.rtbm.2020.100485 https://doi.org/10.1108/09600031111175816 https://doi.org/10.1108/jamr-10-2015-0073 https://doi.org/10.1108/jamr-10-2015-0073 https://doi.org/10.1080/16258312.2021.1932568 https://doi.org/10.1108/14635771311299506 https://doi.org/10.1007/s00170-017-0118-3 https://doi.org/10.1007/s00170-017-0118-3 http://creativecommons.org/licenses/by-nc-nd/4.0/ rangel, d.a., de oliveira, t.k., & leite, m.s.a. (2015). supply chain risk classification: discussion and proposal. international journal of production research, 53(22), 6868-6887. https://doi.org/10.1080/00207543.2014.910620 samvedi, a., jain, v., & chan, f.t.s. (2013). quantifying risks in a supply chain through integration of fuzzy ahp and fuzzy topsis. international journal of production research, 51(8), 2433-2442. https://doi.org/10.1080/00207543.2 012.741330 schoen, q., sanchis, r., poler, r., lauras, m., fontanili, f., & truptil, s. (2018). categorisation of the main disruptive events in the sensitive products transportation supply chains. international journal of production management and engineering, 6(2), 79-89. https://doi.org/10.4995/ijpme.2018.10369 shahbaz, m.s., rm rasi, r.z., & bin ahmad, m.f. (2019). a novel classification of supply chain risks: scale development and validation. journal of industrial engineering and management, 12(1), 201. https://doi.org/10.3926/jiem.2792 sharma, s., & routroy, s. (2016). modeling information risk in supply chain using bayesian networks. journal of enterprise information management, 29(2), 238-254. https://doi.org/10.1108/jeim-03-2014-0031 sodhi, m.s., & tang, c.s. (2012). managing supply chain risk. in springer science & business media. https://doi. org/10.1007/978-1-4614-3238-8. sreedevi, r., saranga, h., & gouda, s.k. (2021). impact of a country’s logistical capabilities on supply chain risk. supply chain management: an international journal, https://doi.org/10.1108/scm-09-2020-0504. srivastava, m., & rogers, h. (2021). managing global supply chain risks: effects of the industry sector. international journal of logistics research and applications, 1-24. tarei, p.k., thakkar, j.j., & nag, b. (2018). a hybrid approach for quantifying supply chain risk and prioritizing the risk drivers: a case of indian petroleum supply chain. journal of manufacturing technology management, 29(3), 533-569. https://doi.org/10.1108/jmtm-10-2017-0218 trkman, p., & mccormack, k. (2009). supply chain risk in turbulent environments-a conceptual model for managing supply chain network risk. international journal of production economics, 119(2), 247-258. https://doi.org/10.1016/j. ijpe.2009.03.002 tukamuhabwa, b.r., stevenson, m., busby, j., & zorzini, m. (2015). supply chain resilience: definition, review and theoretical foundations for further study. international journal of production research, 53(18), 5592-5623. https://doi.org/10.1080/00207543.2015.1037934 tummala, r., & schoenherr, t. (2011). assessing and managing risks using the supply chain risk management process (scrmp). supply chain management, 16(6), 474-483. https://doi.org/10.1108/13598541111171165 venkatesh, v.g., rathi, s., & patwa, s. (2015). analysis on supply chain risks in indian apparel retail chains and proposal of risk prioritization model using interpretive structural modeling. journal of retailing and consumer services, 26, 153-167. https://doi.org/10.1016/j.jretconser.2015.06.001 wagner, s.m., & bode, c. (2008). an empirical examination of supply chain performance along several dimensions of risk. journal of business logistics, 29(1), 307-325. https://doi.org/10.1002/j.2158-1592.2008.tb00081.x world bank. 2022. the impact of the war in ukraine on global trade and investment. washington, dc. world bank. https://openknowledge.worldbank.org/handle/10986/37359, license: cc by 3.0 igo zimmer, k., fröhling, m., breun, p., & schultmann, f. (2017). assessing social risks of global supply chains: a quantitative analytical approach and its application to supplier selection in the german automotive industry. journal of cleaner production, 149, 96-109. https://doi.org/10.1016/j.jclepro.2017.02.041 int. j. prod. manag. eng. (2022) 10(2), 195-209creative commons attribution-noncommercial-noderivatives 4.0 international evaluation of supply chain risks by fuzzy dematel method: a case study of iron and steel industry in turkey 209 https://doi.org/10.1080/00207543.2014.910620 https://doi.org/10.1080/00207543.2012.741330 https://doi.org/10.1080/00207543.2012.741330 https://doi.org/10.4995/ijpme.2018.10369 https://doi.org/10.3926/jiem.2792 https://doi.org/10.1108/jeim-03-2014-0031 https://doi.org/10.1007/978-1-4614-3238-8 https://doi.org/10.1007/978-1-4614-3238-8 https://doi.org/10.1108/scm-09-2020-0504 https://doi.org/10.1108/jmtm-10-2017-0218 https://doi.org/10.1016/j.ijpe.2009.03.002 https://doi.org/10.1016/j.ijpe.2009.03.002 https://doi.org/10.1080/00207543.2015.1037934 https://doi.org/10.1108/13598541111171165 https://doi.org/10.1016/j.jretconser.2015.06.001 https://doi.org/10.1002/j.2158-1592.2008.tb00081.x https://openknowledge.worldbank.org/handle/10986/37359 https://doi.org/10.1016/j.jclepro.2017.02.041 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.1866 received 2013-11-12 accepted 2013-12-16 knownet: exploring interactive knowledge networking across insurance supply chains. dr. susan grant* school of engineering and design brunel university, kingston lane, uxbridge, middlesex ub8 3ph, united kingdom. susan.grant@brunel.ac.uk abstract: social media has become an extremely powerful phenomenon with millions of users who post status updates, blog, links and pictures on social networking sites such as facebook, linkedin, and twitter. however, social networking has so far spread mainly among consumers. businesses are only now beginning to acknowledge the benefits of using social media to enhance employee and supplier collaboration to support new ideas and innovation through knowledge sharing across functions and organizational boundaries. many businesses are still trying to understand the various implications of integrating internal communication systems with social media tools and private collaboration and networking platforms. indeed, a current issue in organizations today is to explore the value of social media mechanisms across a range of functions within their organizations and across their supply chains. the knownet project (an ec funded marie curie iapp) seeks to assess the value of social networking for knowledge exchange across insurance supply chains. a key objective of the project being to develop and build a web based interactive environment a supplier social network or ssn, to support and facilitate exchange of good ideas, insights, knowledge, innovations etc across a diverse group of suppliers within a multi level supply chain within the insurance sector. key words: supply chains, knowledge networking, insurance, social media. 1. packaging and supply chain the astronomical growth and evolution of platforms such as linked in, facebook and twitter reflect the success of social media technologies; more importantly it illustrates how businesses and consumers will expect to interact with and use digital media in the future, for all sorts of reasons, including driving innovation from knowledge sharing and generation. the knownet project sets out to explore the potential and value of current social networking technology to support sustained knowledge sharing and generation across a multilevel supply chain in the insurance sector both in the uk and spain. the aims and objectives of knownet are to develop, build and test an interactive supplier social network (ssn) framework, designed to support local innovation and leaning where explicit, implicit and tacit knowledge and experience of suppliers and their employees can be shared. the ssn will consist of a set of web based tools, such as forums, blogs, wikis, faqs, public recommendations/suggestion pools and exercises and applications specially designed to utilise a range of learning processes (e.g. learning by doing, learning from others, and applications supporting the formation of communities of inquiry and promoting learning through social interaction. specifically, the knownet ssn platform will bring together supply chain members in a highly interactive real-time 3d environment, who be able to communicate quickly and effectively with sound and image and will promote the sharing and adoption of good ideas, practices etc. specifically, the combination of exercises, applications and social interaction tools, ensures a holistic knowledge sharing platform which *dr susan grant is the principal investigator on the knownet project. an fp7 marie curie iapp funded project. 7int. j. prod. manag. eng. (2014) 2(1), 7-14creative commons attribution-noncommercial 3.0 spain http://dx.doi.org/10.4995/ijpme.2014.1866 encourages contact between supply chain partners encourages knowledge transfer across and between supply chain partners develops reciprocity and cooperation among supply chain partners uses active learning techniques gives prompt feedback reduces ‘misinterpretation’ of information communicates high expectations respects diverse talents and ways of learning and knowledge given that the idea of web based interactive ssn’s is relatively novel, and a comparatively unexplored area in the field of supply chain management, the project also explores participants attitudes and behaviours surrounding the concept of virtual supplier communities. crucial to the research is the use of social networking analysis techniques to map and measure the relationships and flows of information/knowledge between individuals and groups within the various supply chains, and gain insight into the roles they play within the network. the concept of collaborative networking is particularly timely in the insurance industry as it looks to strengthen the inter-organizational ties between its suppliers and external agencies, for improving processes, accelerating innovation, fostering creativity, and sharing experiences and local knowledge amongst its supplier networks. the research outputs will enable the industrial participants to assess the inherent value and efficacy of social networking as a knowledge sharing tool which can impact on a range of kpi’s within a large supplier base, as well as provide opportunities for rethinking core processes across a breadth of insurance categories. knownet will be achieved by the knowledge exchange programme between two academic partners (brunel university and universitat politècnica de valència), and one private sector partner (royal & sun alliance insurance plc uk). 2. knowledge networking knowledge is the foundation of a firm’s competitive advantage and ultimately the driver of a firm’s value (teece 2000). organizations therefore need to recognise it as being a valuable asset and develop a mechanism for tapping into the collective intelligence and skills of employees and supplier partners in order to create a greater organizational knowledge base (bollinger and smith 2001). much of the information that companies share — data on inventory levels, sales, production schedules and prices — is easy to codify and transmit. but other types of knowledge are just as important to exchange and more difficult to codify: knowhow, managerial and communication skills and organizational memory. intercompany knowledge sharing should be a joint activity between supply chain partners; the parties share knowledge and then jointly interpret and integrate it into a relationship-domain-specific memory that influences relationship-specific behavior (selnes and sallis 2003). myers and cheung (2008) found typically three types of knowledge sharing within the supply chain, each offering distinct benefits to buyers and suppliers: information sharing, joint sense making and knowledge integration. information sharing takes place when companies exchange important data about sales, customer needs, market structures and demand levels. joint sense making occurs when supply chain partners work together to solve operational problems, analyze and discuss strategic issues and facilitate communication about the relationship. since individual partners often interpret the same information differently, intercompany teams can help create a common understanding. knowledge integration occurs when supply chain partners develop relationship-specific memories, providing everyone with a common understanding of idiosyncratic routines and procedures governing the relationship. this often results in collective problem solving that benefits both the companies and the relationship as a whole. these knowledge-sharing activities constitute mechanisms that can make or break supply chain relationships. in organizations, typically however vital corporate knowledge often gets trapped in information silos like email inboxes, functional silos, structured information systems like erp, crm and srm systems, and more importantly within the minds of employees and supply partners who create, recognize, archive, access and apply knowledge in carrying out their daily tasks (nonaka and konno 1998). 8 int. j. prod. manag. eng. (2014) 2(1), 7-14 creative commons attribution-noncommercial 3.0 spain dr. susan grant indeed, many companies’ today, regardless of location, size or industry sector are struggling with interconnecting knowledge, talent, ideas and relationships both within their own organisational environment and across their supply chains. with the development and evolution of social networking sites such as facebook, linkedin, twitter etc, whereby people connect and collaborate, share personal experiences, and subjective insights, the appeal of social networking for companies to achieve close communities with employees, customers, and suppliers is increasing (khan & khan 2012; mangold & faulds, 2009; mayer, 2009; yang & chen, 2008). such virtual communities can provide similar benefits to traditional social networking methods that enhance innovation and collaborative activity, but with the added advantage of speed, and free from boundaries of time or space (ganley & lampe 2009). indeed, recent evidence shows companies’ are beginning to consider web based ‘social networking’ as a community-building platform to sharing knowledge, (bredl et al., 2012; annabi et al., 2012; álvarez et al., 2009; tsai, 2009. a recent special report in the economist stated that social-networking technologies are creating considerable benefits for the businesses that embrace them. the openness and richness of social networks can foster a fertile environment for the creation of entirely new knowledge, while also accelerating the innovation rate (majewski et al., 2012; seufert et al., 1999; tsai and ghoshal, 1998). the social aspect of learning and acquiring knowledge (know how, know why, know who and know what), is recognised as significant in these innovations. as such knowledge networking and community building to leverage, create sustain and share knowledge in a collaborative way is strongly emphasised through tools that support dialogue, discussion, observation and participation (chatti et al., 2007). whilst the literature on social networking as a collaborative tool (for learning and generating new knowledge) across and within commercial enterprises and their supply chains is relatively new, there are some good examples of existing commercial social networking software that has been applied successfully, that specifically encourage effective collaboration e.g. yammer, a cloudbased enterprise social networking (esn) system; tsb’s_connect portal, and podio from citrix. asda’s recent launch of ‘sustain and save exchange’ (procurement leaders staff 2012) and caterpillar inc’s ‘knowledge network’ are good examples of systems where opportunities for information, knowledge and learning’s can be shared, questions raised, key documents posted, and focused activities attended, to spur new ideas and solve problems amongst members of a supply network. research into supply chain networking is a particularly salient area expected to deliver a significant contribution to the knowledge transfer (and productivity) debate, and indeed there is increasingly recognition that supply chains are beginning to prioritise knowledge creation and exchange (wu 2008) as in the case of asda above. successful management of a supplier network in particular can potentially enhance the productivity of the supply chain through diffusion of knowledge. there is however, a generally adopted view that the potential of scm synergies for the creation and transfer of useful knowledge has not yet been materialised (giannakis 2008), and extensive knowledge sharing across supply chain still appears to be the exception rather than the rule (lin 2005). indeed, the findings of a recent study for the creation of value in organizations for example suggest that although firms in the uk for example, assign great importance to their suppliers as sources of new knowledge creation, their involvement in the generation of knowledge is low (edwards et al., 2004). there are a number of reasons and challenges associated with this. a key challenge concerns motivating supply chain members to engage in knowledge sharing and generating activities in the first place (grant 2013, ardichvili et al., 2003), and a second challenge is the difficulty in generating and transforming knowledge into organizational action, both internally as well as across supply chain partners (capo vicedo et al., 2011). a further key issue concerns the reluctance of companies to share information and knowledge beyond their own internal boundaries. this has implications for generating systems based supply chains innovations, which can impact greatly on customer focus as well as on operational efficiencies. 3. rationale for study conducting business in the financial services sector, requires collaboration across multiple parties within a supply chain. indeed, for agile industries such as insurance and banking, which depend on complex processes of multiple individuals exchanging information, knowledge, ideas, and insights, interaction, via social networks for example, could potentially deliver a huge set of efficiencies and 9int. j. prod. manag. eng. (2014) 2(1), 7-14creative commons attribution-noncommercial 3.0 spain knownet: exploring interactive knowledge networking across insurance supply chains. opportunities for rethinking core supply chain and internal processes. business in the financial services industry traditionally requires the input, participation and decisions of many stakeholders. for example, risk managers, actuaries, it and marking/distribution staff often collaborate in product development. lloyds of london uses collaborative technologies to cut claims costs for all the claims in the entire london insurance market (kontzer 2002). in motor vehicle claims processing, repairers, assessors claims staff, policy holders and legal representatives need to provide inputs and make decisions at different stages of the claims process. despite this need, and some minor developments in collaborative knowledge sharing, up to now, firms in the financial services industry are not seen as conducive to fostering knowledge sharing and generating collaborations across their supply chains in a proactive way (dawson 2004). insurers are beginning to look to incorporate collaboration technologies into their operating models, to improve process efficiency and knowledge sharing (josefowicz 2011, kontzer 2002), and the use of social media to assist in the coordination of knowledge sharing and other business activities is only starting to be explored. this can allow companies to stay close to the changing desires of their customers and the changing trends in the market. however, the use of such approaches and technologies presents a new set of challenges to these organizations, who are not used to managing knowledge transfer in this way. included in these challenges are monitoring appropriate content for sharing or archiving issues, measuring the benefits of these new tools, integrating these new tools into existing workflow, communication and archiving systems and understanding the motivations prompting people to share knowledge or participate in virtual communities, in an industry that has typically always used private communication channels. the knownet project seeks to build on these challenges by identifying and measuring the value of social networking across multiple groups and stakeholders in two insurance companies and their suppliers. specifically, the project addresses the organizational contexts, and commitments, motivations of multiple groups and stakeholders prior to developing, building and trialling a bottom up, user designed web based interactive environment a supplier social network (ssn), to support and facilitate exchange of good ideas, insights, tacit and explicit knowledge, innovations etc. the project will also develop a tool for measuring accurate and effective knowledge transfer, as well as measuring participant engagement and motivation to sharing new ideas, insights and knowledge in a conservative sector such as insurance. social network analysis in addition to building a socially interactive ssn framework, the project also uses social network analysis (sna) techniques as a modelling tool to better understand knowledge management in a multilevel sc. the sna perspective views any system as a set of interrelated actors or nodes. actors represent entities at various levels of collectivity, such as persons, companies, countries, and so on (borgatti and li, 2009). sna is essentially the mapping and measuring of relationships and flows between people, groups, organizations, computers, or other information and knowledge processing entities (hanenman 2002). a key output of sna is the knowledge map which provides insight for improving business and organisational processes (liebowictz 2005). knowledge maps may help identify intellectual capital (liebowitz 2003, socialise new members and enhance organizational learning (wexler 2001). several authors propose sna techniques (boschma and ter wal, 2007; borgatti et al., 2009) as appropriate to model business networks. in fact, there have been many previous works from supply chain management using these techniques (carter et al., 2007; mueller et al., 2007; ozkul and barut, 2009; borgatti and li, 2009; choi and wu, 2009; bernardes, 2010). the use of sna techniques in this project is expected to provide useful insights into how rsa’s ssn can reinforce their collaborative behaviours and activities to not only enhance their relationships, but to also achieve competitive advantages for the ssn as a whole. 4. methodology the knownet project will be implemented in 3 phases. in phase one, the consortium partners will engage in exchanging knowledge to initially develop, build and test an interactive supplier social network, prior to conducting parallel trials in the uk and spain to assess its knowledge transfer capability. in phase two, the consortium will identify ‘optimal knowledge exchange and transfer tools and applications within a digital social networking environment, subsequent to evaluating user 10 int. j. prod. manag. eng. (2014) 2(1), 7-14 creative commons attribution-noncommercial 3.0 spain dr. susan grant engagement and knowledge transfer capability of the provisional integrated system. phase three will measure knowledge adoption and transfer capability within the revised framework, prior to finalising the platform. the figure below outlines the 3 phases of the programme as shown in figure 1, the research programme will be implemented in 3 phases. in phase 1, the key activities are to develop, build and test a provisional ssn platform. the development of a conceptual ssn model which will inform the build of the platform, will be based on a range of sources including a literature review, results from a recent pilot study (britac 2011) examining attitudes to ssn engagement across a motor insurance supply chain in the uk; social networking analysis theory and of course, user centric discussions on requirements ( via user stories) during the period of a secondment unit with the insurance companies. the major output of phase 1 is to establish the parameters for a ssn platform via continuous discussion and negotiation with the partners and users, establish an architecture and specification for the ssn, and build a provisional ssn framework (incorporating web based tools,) that successfully enable diverse supply chain partners from selected insurance supply chains in the uk and spain, to interact for the purposes of knowledge transfer. consultation on requirements: this initial step involved an intense period of consultation with key stakeholders in the company. this took the form of discussions and presentations during two figure 1. methodology. 11int. j. prod. manag. eng. (2014) 2(1), 7-14creative commons attribution-noncommercial 3.0 spain knownet: exploring interactive knowledge networking across insurance supply chains. secondments over a period of months. the process began by identifying requirements with key stakeholders to reveal key areas where the adoption of social networking tools could provide a benefit. six potential areas were identified and ranked using swot analysis. following this, suitable participants from across business processes and externally (suppliers) were identified to take part in the trial. part of the process of identifying requirements of users involved gathering a catalogue of user stories, that would help drive the selection of tools/functions and serve as benchmarks to assess whether the platform solutions were delivering to requirements. trial: november 2013. the trial is due to launch towards the end of the first year, and will serve a dual purpose. firstly as an iterative process with user centred feedback over a period of 12 months, to drive changes to the look, feel and functionality of the platform in order to optimise knowledge sharing/ learning the second purpose of the trial will be to log data on supplier engagement and usage, as well as survey supplier satisfaction and attitudes to interaction within an social networking environment. the project will also carry out social networks analysis (sna) techniques to model knowledge flows across the multi levels of the supply chain. much of the data will be gained from the logging data as well as via online interviews. the sna modelling should reveal the structural properties of the network and the implications of these structural properties for the design of social network based systems. in phase two, engagement with the platform over the trials will provide an opportunity to study the characteristics of online social networking and an understanding on how to improve/modify a later version of the ssn. the major task of this phase will be to identify the optimal social media tools and applications (e.g. which provide wiki function, blog function etc,) capable of promoting the sharing of good ideas, and knowledge transfer through social interaction. this process will be iterative, until the end of the 12month period, when the trials conclude. the platform modification process will be process driven and user centric in nature. the key activities in phase 3 will be to measure knowledge transfer and knowledge adoption within the revised ssn framework. participants will be surveyed/interviewed to ascertain ease of use and satisfaction with the tools, the acquisition of new content (via leaning by doing), the acquisition of new insights (via learning from others), and followon exercises later in the trial to measure ability of participants to implement the new content gained from the interaction with other participants and tools/exercises. a further activity will involve identifying structural/ cultural inhibitors and enablers’ to engagement and interaction using interviews, surveys and continued sna modelling during the field trials. the findings from the trials in phase 3 will define the generic constructs of the ‘knowledge sharing ssn’ determine its’ usefulness across other domestic domains, and refine the evaluation tool. this tool will continue to monitor and measure engagement and usage, supplier feedback and the impact of the different learning processes and tools on the depth and breadth of knowledge transfer, motivation to share ideas, etc, until the end of the trial and highlight patterns of behaviour and adoption of new learning. the ssn framework will be applied to a number of participants (primarily sme’s) within a multilevel supply insurance chain in spain and the uk. phase 3 aims to refine and validate the framework developed in phase 1 and evaluate this interactive medium for transferring ideas, insights, experiences and learning from others. the trial also allows researchers to explore structural (using sna analysis) and cultural differences across the 2 groups. the ssn needs to be verified in different countries having different cultures. as a result the knownet consortium consists of partners from different cultures to execute the field trials in the uk and spain. if the ssn is shown to promote knowledge networking (knowledge exchange and generation and learning) across the partner’s cultures, then it is felt that a similar framework using web based tools and applications, would be accepted by a number of other countries with similar cultures, within the eu. the knownet project is underpinned by the transfer of knowledge between the commercial and academic partners. all phases of the research will include transfer of knowledge via secondments. the secondments to date have taken the form of workshop discussions, presentations and software demonstrations, and have been both intersectoral and international in nature. 5. conclusion the project is currently in the first year of operation and partners are working towards refining user 12 int. j. prod. manag. eng. (2014) 2(1), 7-14 creative commons attribution-noncommercial 3.0 spain dr. susan grant requirements for the platform. a third secondment unit is scheduled for the summer of 2014 where all partners will collaborate on midterm findings and lessons learned from the uk and spanish trials. the knownet project is a 4 year project funded by the fp7 ec industry-academia partnerships and pathways (iapp) marie curie actions. references álvarez, i., marín, r., fonfría, a (2009). the role of networking in the competitiveness of firms. technological forecasting & social change, 76(3), 410-421. doi:10.1016/j.techfore.2008.10.002 annabi, h., mcgann, s.t., pels, s., arnold, p., rivinus, c. (2012). guidelines to align communities of practice with business objectives: an application of social media (hicss), 45th hawaii international conference on system science. 4-7th january, 2012. maui, hawai, usa. ardichvili, a., page, v., wentling, t. (2003). motivation and barriers to participation in virtual knowledge-sharing communities of practice. journal of knowledge management, 7(1), 64-77. doi:10.1108/13673270310463626 bernardes, e.s. (2010). the effect of supply management on aspects of social capital and the impact in performance: a social network perspective. journal of supply chain management, 46(1), 45-55. doi:10.1111/j.1745-493x.2009.03185.x bollinger, a.s., smith, r.d. (2001). managing organizational knowledge as a strategic asset. journal of knowledge management, 5(1), 8-18. doi:10.1108/13673270110384365 borgatti, s.p., mehra, a., brass, d.j., labianca, g. (2009). network analysis in the social sciences. science, 323(5916), 892-895. doi:10.1126/science.1165821 boschma, r.a., ter wal, a.l.j. (2007). knowledge networks and innovative performance in an industrial district: the case of a footwear district in the south of italy. industry & innovation, 14(2), 177-199. doi:10.1080/13662710701253441 bredl, k., grob, a., hunniger, j., fleischer, j. (2012). the avatar as a knowledge worker? how immersive 3d virtual environments may foster knowledge acquisition. electronic journal of knowledge management,10(1), 15-25. bulmer, d. (2009). social networks and b2p. capó-vicedo, j., mula, j., capó, j. (2011). a social network-based organizational model for improving knowledge management in supply chains. supply chain management: an international journal, 16(4), 284-293. doi:10.1108/13598541111139099 carter, c.r., ellram, l.m., tate, w. (2007). the use of social network analysis in logistics research. journal of business logistics, 28(1), 137-168. doi:10.1002/j.2158-1592.2007.tb00235.x chatti, m., jarke, m. (2007). the future of e-learning: a shift to knowledge networking and social software. international journal of knowledge and learning, 3(4-5) 404-420. chesborough, h. (2003). open innovation: the new imperative for creating and profiting from technology. harvard business school press, cambridge, ma, usa. chiu, c., hsu, m., wang, e. (2006). understanding knowledge sharing in virtual communities: an integration of social capital and social cognitive theories. decision support systems, 42(3), 1872-1888. choi, t.y., wu, z. (2009). triads in supply networks: theorizing buyer-supplier-supplier relationships. journal of supply chain management, 45(1): 8-25. doi:10.1111/j.1745-493x.2009.03151.x chungsuk, r., kim, y.j., abhijit, c., raghav, r. (2005). knowledge acquisition via three learning process in enterprise information portals: learning by investment, learning by doing and learning from others. mis quarterly, 29(2), 245-278 ciborra, c., patriota, g. (1998). groupware and teamwork in r&d: limits to learning and innovation. r&d management 28(1), 43-52. doi:10.1111/1467-9310.00080 cummings, j.l., teng, b.s. (2003). transferring r&d knowledge : the key factors affecting knowledge transfer success. journal of engineering and technology management, 20(1-2), 39-68. doi:10.1016/s0923-4748(03)00004-3 dawson, r. (2004). review of the collaborations in the financial services conferences, 29th september 2004. available at: http://rossdawsonblog.com/weblog/archives/2005/07/review_of_colla.html dede, c. (2009). immersive interfaces for engagement and learning. science, 323(5910), 66-69. doi:10.1126/science.1167311 delong, d., fahey, l. (2000). diagnosing cultural barriers to knowledge management. emerald management reviews, 14(4), 113-127. duan, w. (2009). special issue on online communities and social network: an editorial introduction. decision support systems, 47(3), 167168. doi:10.1016/j.dss.2009.02.010 edwards, t., battisti, g., neely, a. (2004). value creation and the uk economy: a review of strategic options. international journal of management reviews,5(3-4), 191-213. doi:10.1111/j.1460-8545.2004.00103.x ganley, d., lampe, c. (2009). the ties that bind: social network principles in online communities. decision support systems, 47(3), 266274. doi:10.1016/j.dss.2009.02.013 giannakis, m. (2008). facilitating learning and knowledge transfer through supplier development. supply chain management: an international journal, 13(1), 62-72. doi:10.1108/13598540810850328 13int. j. prod. manag. eng. (2014) 2(1), 7-14creative commons attribution-noncommercial 3.0 spain knownet: exploring interactive knowledge networking across insurance supply chains. http://dx.doi.org/10.1016/j.techfore.2008.10.002 http://dx.doi.org/10.1108/13673270310463626 http://dx.doi.org/10.1111/j.1745-493x.2009.03185.x http://dx.doi.org/10.1108/13673270110384365 http://dx.doi.org/10.1126/science.1165821 http://dx.doi.org/10.1080/13662710701253441 http://dx.doi.org/10.1108/13598541111139099 http://dx.doi.org/10.1002/j.2158-1592.2007.tb00235.x http://dx.doi.org/10.1111/j.1745-493x.2009.03151.x http://dx.doi.org/10.1111/1467-9310.00080 http://dx.doi.org/10.1016/s0923-4748(03)00004-3 http://rossdawsonblog.com/weblog/archives/2005/07/review_of_colla.html http://dx.doi.org/10.1126/science.1167311 http://dx.doi.org/10.1016/j.dss.2009.02.010 http://dx.doi.org/10.1111/j.1460-8545.2004.00103.x http://dx.doi.org/10.1016/j.dss.2009.02.013 http://dx.doi.org/10.1108/13598540810850328 gilbert, m., cordey-hayes, m. (1996). understanding the process of knowledge transfer to achieve successful technological innovation. technovation, 16(6), 301-312. doi:10.1016/0166-4972(96)00012-0 grant, s.b. (2013). exploring attitudes to knowledge networking in uk insurance supply chains. unpubl. josefowicz, m. (2011). sharing alike: insurers need to figure out how to use new collaboration tools. inn sight, 14(7), 28. khan, a., khan, r. (2012). embracing new media in fiji: the way forward for social network marketing and communication strategies. strategic direction, 28(4), 3-5. doi:10.1108/02580541211212754 kontzer, t. (2002). come togethercompanies are collaborating to save money, save time, and drive new business. information week, october 7th 2002. kotabe, m., martin, x., domoto, h. (2003). gaining from vertical partnerships: knowledge transfer, relationship duration and supplier performance improvement in the us and japanese automotive industries. strategic management journal, 24(4), 293-316. doi:10.1002/smj.297 leug, c. (2003). knowledge sharing in online communities and its relevance to knowledge management in the e-business era. international journal of electronic business, 1(2), 140. doi:10.1504/ijeb.2003.002170 lin, c., wu, c. (2005). managing knowledge contributed by iso 9001:2000. international journal of quality and reliability management, 22(9), 968-985. doi:10.1108/02656710510625239 majewski, g., usoro, a., khan, i. (2012). knowledge sharing in immersive virtual communities of practice. vine, 41(1), 41-62. mangold, w.g., faulds, d.j. (2009). social media: the new hybrid element of the promotion mix. business horizons, 52(4), 357-365. doi:10.1016/j.bushor.2009.03.002 martz, w.b., shepherd, m. (2007). managing distance education for success. international journal of web-based learning and teaching technologies, 2(2), 50-58. doi:10.4018/jwltt.2007040104 mason, c., castleman, t., parker, c. (2006). virtual communities of practice for regional small businesses: bringing informal networks online, bled, slovenia june 5-7 2006. mayer, a. (2009) online social networks in economics. decision support systems, 47(3), 169-184. doi:10.1016/j.dss.2009.02.009 mclure wasko, m., faraj, s. (2000). "it is what one does..": why people participate and help others in electronic communities of practice. journal of strategic information systems, 9(2-3), 155-173. doi:10.1016/s0963-8687(00)00045-7 mueller, rae., buergelt, d., seidel-lass, l. (2007). supply chains and social network analysis. 1st international european forum on innovation and system dynamics in food networks, 1-12. myers, m.b., cheung, m.s. (2008). sharing global supply chain knowledge. mit sloan management review, 1st july. nevo, s., nevo d., carmel, e. (2011). unlocking the business potential of virtual worlds. mit sloan magazine: spring 2011. newell, s., swan, j. (2000). trust and inter-organizational networking. human relations. 53(10), 1287-1328. nonaka, i., konno, n. (1998). the concept of ‘ba’: building a foundation for knowledge creation, california management review, 40(3), 4054. doi:10.2307/41165942 ozkul, a., barut, m. (2009). measuring supply chain relationships: a social network approach. international journal of integrated supply management, 5(1), 38-61. doi:10.1504/ijism.2009.026204 pl staff. (2012). asda iniative aims to improve supply chain sustainability. procurement leaders, 14th june 2012 scarbrough, h., swan, j., preston, j. (1999). knowledge management and the learning organization. ipd, london, united kingdom. selnes, f., sallis, j. (2003). promoting relationship learning. journal of marketing, 67(3), 80-95. doi:10.1509/jmkg.67.3.80.18656 seufert, a., krogh, g.v., bach, a. (1999).towards knowledge networking. journal of knowledge management, 3(3), 180-190. doi:10.1108/13673279910288608 stuart, i., deckert, p., mc cutcheon, d., kunst, r., (1998). case study: a leveraged learning network. sloan management review, 39(4), 81-93 swan, j., newell, s., scarbrough, h., hislop, d. (1999). knowledge management and innovation: networks and networking. journal of knowledge management, 3(4), 262-275. doi:10.1108/13673279910304014 szulanski, g. (1996). exploring internal stickiness: impediments to the practice of best practice within the firm. strategic management journal 17, 27-44 special issue teece, d.j. (2000). strategies for managing knowledge assets: the role of firm structure and industrial context. long range planning, 33(1), 35-54. doi:10.1016/s0024-6301(99)00117-x tsai, k. (2009). collaborative networks and product innovation performance: toward a contingency perspective. research policy, 38(5), 765-778. doi:10.1016/j.respol.2008.12.012 walter, j., lechner, c., kellermanns, f.w. (2007). knowledge transfer between and within alliance partners: private versus collective benefits of social capital. journal of business research, 60(7), 698-710. doi:10.1016/j.jbusres.2007.01.026 wheeler, s. (2007). in moore, m.g. (1972) learner autonomy: the second dimension of independent learning. convergence, 5(2), 76-88 wu, c. (2008). knowledge creation in a supply chain. supply chain management: an international journal,13(3), 241-250. doi:10.1108/13598540810871280 yang, s.j.h., chen, i.y.l. (2008). a social network-based system for supporting interactive collaboration in knowledge sharing over peer-topeer network. international journal of human-computer studies, 66(1), 36-50. doi:10.1016/j.ijhcs.2007.08.005 14 int. j. prod. manag. eng. (2014) 2(1), 7-14 creative commons attribution-noncommercial 3.0 spain dr. susan grant http://dx.doi.org/10.1016/0166-4972(96)00012-0 http://dx.doi.org/10.1108/02580541211212754 http://dx.doi.org/10.1002/smj.297 http://dx.doi.org/10.1504/ijeb.2003.002170 http://dx.doi.org/10.1108/02656710510625239 http://dx.doi.org/10.1016/j.bushor.2009.03.002 http://dx.doi.org/10.4018/jwltt.2007040104 http://dx.doi.org/10.1016/j.dss.2009.02.009 http://dx.doi.org/10.1016/s0963-8687(00)00045-7 http://dx.doi.org/10.2307/41165942 http://dx.doi.org/10.1504/ijism.2009.026204 http://dx.doi.org/10.1509/jmkg.67.3.80.18656 http://dx.doi.org/10.1108/13673279910288608 http://dx.doi.org/10.1108/13673279910304014 http://dx.doi.org/10.1016/s0024-6301(99)00117-x http://dx.doi.org/10.1016/j.respol.2008.12.012 http://dx.doi.org/10.1016/j.jbusres.2007.01.026 http://dx.doi.org/10.1108/13598540810871280 http://dx.doi.org/10.1016/j.ijhcs.2007.08.005 pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2015.3320 received 2014-10-17 accepted: 2015-05-07 quantitative assessment of sustainable city logistics rafael grosso-delavegaa,i and jesús muñuzuria,ii a dpto. de organización industrial y gestión de empresas ii. etsi. universidad de sevilla. camino de los descubrimientos, s/n. isla de la cartuja, 41092, seville, spain a,i rgrosso@us.es a,ii munuzuri@etsi.us.es abstract: the aim of this paper is to seek an answer to an specific question: how to make city logistics sustainable? this question in principle has no specific answer. by contrast, it could be answered in many and varied ways. behind the search for some of these answers lies the development of a roadmap which this work aims to present. the research lines, the theoretical framework and methodology of the roadmap will be explained. although the current status of the roadmap, its duration and timing still need to be completed, the main facts, as well as the results obtained to date and the expected results are here presented. key words: city logistic, sustainable policies, access time windows, waste collection, optimization. 1. introduction 1.1. the paradigmatic framework a multiplicity of different kinds of goods are constantly entering, transiting and leaving urban areas: consumer goods, building materials, waste, packaging and mailings, etc. (dablanc, 2007). it is well known that the urban freight transport includes heterogeneous goods and different types of vehicles of different sizes. in addition, the movement urban goods has a direct and fundamental influence on economics (muñuzuri et al., 2005) and is vital to industry performance. that is why urban freight management is a necessary challenge which nevertheless implies a high complexity. but goods are not only transported in urban environments, so the first problem which efficient management of urban freight transport finds is the very notion of transport itself and its variants (figliozzi, 2012). traffic congestion has become a daily phenomenon due to the increasing amount of traffic and the limited capacity of the road network. and these growing delays are very costly for both private road users and logistics and distribution providers. this causes high economic costs to these providers, in an attempt to avoid possible delays in deliveries or collections to customers, by additional vehicles and its own drivers. in addition, violations of driving and traffic rules need close attention. furthermore, it increases externalities related to the environment, such as emissions of co2. another important problem that urban freight transport has to face is the urban morphology. european cities have several common characteristics that influence directly their mobility and their businesses. likewise it imposes some restrictions on the flow associated to the supply of goods. first, most of its city centers have a radial structure with a high concentration of shopping areas, restaurants and other centers of social attraction. this structure, which is inherited from the middle ages, generates asymmetric flows of people (going to work, to shop, to eat or to visit tourist attractions) with those flows associated with goods. parking problems, which virtually exist at the center of all urban areas, increase in europe due to its peculiar morphology consisting of alleys and narrow streets (ligocki & zonn, 1984; muñuzuri et al., 2012b). in addition, the road transport sector in spain has not been considering city logistics as an industrial subsector. therefore, there are no databases showing the importance of this activity. 97int. j. prod. manag. eng. (2015) 3(2), 97-101creative commons attribution-noncommercial-noderivatives 4.0 international on the other hand, from the 90´s the concept of sustainable development has been attracting worldwide attention. sustainable development has proven to be an enduring and compelling concept because it points towards a clear management policy. also, it is also flexible enough to adapt to new challenges, technological and economic conditions and social aspirations. it appeals to the general public and the scientific community in particular, as it involves a systemic view of economy and ecology, and requires solutions that protect the interests of future generations (goldman & gorham, 2006). sustainable development meets the needs of the present without compromising the ability of future generations to meet their own needs. it is widely accepted that these needs include economic, social and environmental developments (see figure 1). this “triple” point of view understands that development should be bearable (socially and environmentally), fair (socially and economically) and viable (environmentally and economically) and, therefore, sustainable and durable. the representation of the “three pillars of sustainability” implies the fact that the concept of sustainability itself is the result of interactions between these three dimensions. that is the reason why they cannot, or rather should not, be analysed separately from each other (rossi et al., 2012). in response to the intersection of urban freight transport and the concept of sustainability, a holistic approach to globalize planning and urban management needs to be adopted (robusté et al., 2000). such a challenge needs to consider together all operations and services present in the city; special attention to the sustainability of the system should be paid. this new discipline, which aims at systemic or holistic optimization of city services, could be called sustainable city logistics. figure 1. schematic view of the “three pillars of sustainability”. 1.2. answering the questions therefore the question to answer in this roadmap is how to make city logistics sustainable. this question is very broad and covers many answers. that is why there has been an attempt at setting more specific objectives for our roadmap. consquently, and for this purpose this work aims to solve two sub questions related to this capital aspect. 1.2.1. are sustainability policies really sustainable? the first sub question which is being addressed is about road freight transport. it is well known that road freight transport is causing a number of social, environmental and economic negative impacts in many cities around the world. therefore sustainable city logistics must be the solution to the problems of urban centers, and researchers must have as their main objective to reduce these impacts without penalizing cities needs (chang & yen, 2012). moreover, policy makers and decision makers aim at decreasing the above mentioned variety of negative social, environmental and economic impacts of urban freight transport. because of this several initiatives and policies have been implemented to try reduce them (temporal regulation of access, promotion of cooperation between public and private sector, etc.). some of the objectives of these policies are to improve the environment (air and noise quality), securing pedestrian’s space and the prevention of accidents. they all have sustainability as the ultimate goal (dalkmann & brannigan, 2007). in this situation, city logistics researches, reflect upon the impact of these policies on the different areas and upon the interests of the different stakeholders involved in urban areas and its centers (citizens, residents, merchants, transporters, local authorities, etc.). this is a field that has been investigated in recent years (quak & de koster, 2009; gonzalez-feliu et al., 2012; stathopoulos et al., 2012). given the heterogeneity of the interests of these stakeholders, coordination becomes somewhat cumbersome, so they generally act independently and without any centralized control. but this paper seeks to answer a less particular issue; a question which captures the overall interests of all stakeholders involved (general interests should be above individuals): are sustainability policies really sustainable? therefore, the first purpose of this work is to evaluate one of these policies in a quantitative way to answer the question (muñuzur et al., 2013). 98 int. j. prod. manag. eng. (2015) 3(1), 97-101 creative commons attribution-noncommercial-noderivatives 4.0 international grosso-delavega, r. and muñuzuri, j. 1.2.2. how to make urban fleets more sustainable? the second sub-question that arises from the main one is about the urban fleets, more specifically about the fleet in charge of recyclable waste. trying to solve the problem of waste collection in cities is not a new problem. back in the 70’s, authors already addressed the problem, either from a mathematical point of view (marks & liebman, 1970), or by modeling and solving a vehicle routing problem (vrp) (beltrami & bodin, 1974; turner & hougland, 1975). this problem is not easy to solve since it falls under the classification of np-hard. the increased levels of consumption and the waste generation associated with it, the environmental considerations and the sustainability of cities have led to the emergence of new european and national policies regarding the management of municipal waste. an example of this is the national integrated waste plan implemented in spain in 2009, which is to continue the previous national urban waste plan (pnru). among other things, it enforces municipalities over 5000 inhabitants to ensure proper separation for a selective collection of waste. such measures imply the consideration of new challenges to municipalities, even more so in the economic recession framework in which we live. different types of dumpsters, different types of waste, the location of dumpsters, pollution, energy consumption, cost reduction and the like, are some of these challenges. in this sense, authors address the problem from such perspectives as the consumption of fuel (sonesson, 2000), or having in mind environmental and economic goals. nowadays the emergence of new technologies and the drop in their price allow researchers to find new tools to solve this problem. examples of these new technologies are, among others, the geographic information system (gis), volumetric sensors, or radio frequency identification (rfid). by using this technology, some issues may be addressed. these include eliminating unnecessary stops, fleet reduction and balancing according to demand, pollution impact reduction, operating costs reduction, etc. these issues are actually the basis of some research projects undertaken in recent years (chang, lu, & wei, 1997; nuortio et al., 2006) . needlles to say that all these new lines of reseach offer a great potential for future work. it is in this direction that this work moves. this part of the project addresses the problem of waste disposal in urban areas with the real-time level data of the dumpsters. in particular, the work focuses on the collection of glass containers. a more sustainable collect policy is present and compared with other classical optimization algorithms (grosso-delavega et al., 2014). 2. proposed solutions the objectives of the roadmap will focus on: characterize and analyse the situation of city logistics and characterize and analyze the situation of recyclable waste collection in the european union and spain. study the existing scientific literature on city logistics and recyclable waste collection, especially in the field of sustainability and city centers. study of the determining factors for sustainable development of city logistics in centers in european and the particularly factors in spanish cities. design optimization models for sustainable city logistics improvement and for better understanding and analysis. development of a simulation environment, using heuristics and metaheuristics, specifically designed for city logistic problems in city centers. validation of the models proposed in the simulation environment. as already mentioned above, the proposed methodology focuses on optimization algorithms. also, also solutions need to found in a relatively short time; in this way fast optimization mechanisms such as metaheuristics, heuristics and techniques are implemented. these will be compared with existing techniques in order to be able to verify the hypothesis. the work has been divided into four stages, which be conducted sequentially: 1. study of the history of freight transport in europe and the state of the art in terms of optimization of urban transport routes and its sustainability. 2. development of a simulation environment in which to test the heuristics and metaheuristics. 3. development and codification of the different heuristics and metaheuristics are considered to solve the said problems. 99int. j. prod. manag. eng. (2015) 3(2), 97-101creative commons attribution-noncommercial-noderivatives 4.0 international quantitative assessment of sustainable city logistics 4. analysis of heuristics raised. study and comparison of the results obtained. analysis of the improvements that the system would provide in a real environment. 3. expected and existent contributions this roadmap project was initiated in september 2012. since then there have been many experiments and some intermediate results have been obtained. there are still results to be complete, however, although some of the responses to the issues raised have been published in the following papers derived from the roadmap: grosso-delavega et al. (2014) and muñuzuri et al. (2013). other published works related to the main theme of the roadmap are: muñuzuri et al. (2012a) and muñuzuri et al. (2011). in their paper, muñuzuri et al., (2013) developed a model based on vrp logic called vehicle routing problem with access time windows (vrpatw). this model was solved using genetic algorithms. they provided conclusive results, about the sustainability of the policies adopted in the city centers. following the line of research initiated earlier, the autors are currently working on the development and resolution of the model. it is intended to solve as a mathematical model and using a greedy heuristic. in this way the model would be solved by mathematical programming, using a metaheuristic and also a heuristic. the aims are: to be able to answer the questions raised in a more precise way. perform a comparison of the different techniques used in terms of methodology. this comparison is intended be accomplished in terms of: o proximity to the optimal solution o size of the problem that can be solved with each technique o solving times. at present, this research project is at the design of the experiment phase stage. these experiments must be designed in order to be solved by means of the three techniques. it must be said that the greedy heuristic is being tested so that it solves the problem satisfactorily. with respect to the line of garbage collection, this research project is currently trying to improve the resolution algorithm in order to to improve the results. at the time that this work was written had another year and a half to the end of the stipulated period of time for the finalization of the roadmap. given the published results and the results that could be obtained, it is expected that two publications can be submitted in the period of time left. aditionally, potential contributions of the roadmap might include the following: a move from the theoretical level to the practical level and transfer the results of this roadmap to local authorities. continue developing as a scientist. references beltrami, e. j., bodin, l. d. (1974). networks and vehicle routing for municipal waste collection. networks, 4(1): 65-94. doi:10.1002/ net.3230040106 chang, n.-b., lu, h., wei, y. (1997). gis technology for vehicle routing and scheduling in solid waste collection systems. journal of environmental engineering, 123(9): 901-910. doi:10.1061/(asce)0733-9372(1997)123:9(901) chang, t.-s., yen, h.-m. (2012). city-courier routing and scheduling problems. european journal of operational research, 223(2): 489-498. doi: 10.1016/j.ejor.2012.06.007 dablanc, l. (2007). goods transport in large european cities: difficult to organize, difficult to modernize. transportation research part a: policy and practice, 41(3): 280-285. doi: 10.1016/j.tra.2006.05.005 dalkmann, h., brannigan, c. (2007). transport and climate change. module 5e. sustainable transport: a sourcebook for policy-makers in developing cities. deutsche gesellschaft fuer technische zusammenarbeit (gtz). available online at http://www.gtkp.com/assets/ uploads/20091123-095443-1692-5e_tcc.pdf. figliozzi, m. a. (2012). the time dependent vehicle routing problem with time windows: benchmark problems, an efficient solution algorithm, and solution characteristics. transportation research part e: logistics and transportation review, 48(3): 616-636. doi: 10.1016/j. tre.2011.11.006 100 int. j. prod. manag. eng. (2015) 3(1), 97-101 creative commons attribution-noncommercial-noderivatives 4.0 international grosso-delavega, r. and muñuzuri, j. goldman, t., gorham, r. (2006). sustainable urban transport: four innovative directions. technology in society, 28(1-2): 261-273. doi: 10.1016/j.techsoc.2005.10.007 gonzalez-feliu, j., ambrosini, c., pluvinet, p., toilier, f., routhier, j.-l. (2012). a simulation framework for evaluating the impacts of urban goods transport in terms of road occupancy. journal of computational science, 3(4): 206-215. doi:10.1016/j.jocs.2012.04.003 grosso-delavega, r., muñuzuri sanz, j., rodriguez palero, m., teba fernandez, j. (2014). optimization of recyclable waste collection using real-time information. annals of industrial engineering 2012 (pp. 171-177): springer. ligocki, c., zonn, l. e. (1984). parking problems in central business districts. cities, 1(4): 350-355. doi:10.1016/0264-2751(84)90006-4 marks, d. h., liebman, j. (1970). mathematical analysis of solid waste collection. public health service publication (vol. 2104): department of health, education and welfare. may, a. d. (2013). urban transport and sustainability: the key challenges. international journal of sustainable transportation, 7(3): 170-185. doi:15568318.2013.710136 muñuzuri, j., cortés, p., grosso, r., guadix, j. (2012a). selecting the location of minihubs for freight delivery in congested downtown areas. journal of computational science, 3(4): 228-237. doi: 10.1016/j.jocs.2011.12.002 muñuzuri, j., cortés, p., guadix, j., onieva, l. (2012b). city logistics in spain: why it might never work. cities, 29(2): 133-141. doi: 10.1016/j.cities.2011.03.004 muñuzuri, j., grosso, r., cortés, p., guadix, j. (2011). development of a cost model for intermodal transport in spain. in s. renko (ed.), supply chain management new perspectives. doi:10.5772/22740 muñuzuri, j., grosso, r., cortés, p., guadix, j. (2013). estimating the extra costs imposed on delivery vehicles using access time windows in a city. computers, environment and urban systems, 41: 262-275. doi: 10.1016/j.compenvurbsys.2012.05.005 muñuzuri, j., larrañeta, j., onieva, l., cortés, p. (2005). solutions applicable by local administrations for urban logistics improvement. cities, 22(1): 15-28. doi: 10.1016/j.cities.2004.10.003 nuortio, t., kytöjoki, j., niska, h., bräysy, o. (2006). improved route planning and scheduling of waste collection and transport. expert systems with applications, 30(2): 223-232. doi: 10.1016/j.eswa.2005.07.009 quak, h. j., de koster, m. b. m. (2009). delivering goods in urban areas: how to deal with urban policy restrictions and the environment. transportation science, 43(2): 211-227. doi: 10.1287/trsc.1080.0235 robusté, f., campos, j. m., galván, d. (2000). nace la logística urbana. paper presented at the actas del iv congreso de ingeniería del transporte. editado por jv colomer y a. garcía. schleske, e., lozano, a., antún, jp (2001). location of a logistic platform for improving the shoe distribution in mexico city. proceedings of the xxxii annual conference of the operational research society of italy. rossi, r., gastaldi, m., gecchele, g. (2012). comparison of fuzzy-based and ahp methods in sustainability evaluation: a case of traffic pollution-reducing policies. european transport research review, 5(1): 11-26. doi: 10.1007/s12544-012-0086-5 sonesson, u. (2000). modelling of waste collection–a general approach to calculate fuel consumption and time. waste management and research, 18(2): 115-123. doi:10.1177/0734242x0001800203 stathopoulos, a., valeri, e., marcucci, e. (2012). stakeholder reactions to urban freight policy innovation. journal of transport geography, 22: 34-45. doi: 10.1016/j.jtrangeo.2011.11.017 turner, w. c., hougland, e. s. (1975). the optimal routing of solid waste collection vehicles. aiie transactions, 7(4): 427-431. doi:10.1080/05695557508975027 101int. j. prod. manag. eng. (2015) 3(2), 97-101creative commons attribution-noncommercial-noderivatives 4.0 international quantitative assessment of sustainable city logistics pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.2078 received 2014-01-12 accepted 2014-05-26 cell formation heuristic procedure considering production data shailendra kumari,* and dr rajiv kumar sharmaii mechanical engineering department, national institute of technology, hamirpur, himachal pradesh, india-177005. i tyagi_sk@yahoo.com ii rksnithmr@gmail.com abstract: manufacturing cell formation is one of foremost, and critical aspect of any manufacturing cell design problem. a large number of cell formation methods are developed and research in this area is still in progress. in this paper an attempt has been made by authors to develop a simple, easy to understand and implement cell formation heuristic, having the capability to handle production data viz. operation sequence, production volume, and inter-cell movement cost simultaneously. the results obtained from proposed method are in tune with some highly complex methods, which validates the performance of proposed procedure. to demonstrate its ability to handle other production parameters with little modifications, a modification for consideration to part processing cost in addition to above mentioned production data is developed and explained. towards the end the procedure to handle alternate process plans in conjugation with production data by the proposed cell formation procedure is also discussed. key words: alternate process plan, cell formation heuristic, inter-cell movement cost, operation sequence, part processing cost, production volume. 1. introduction the present era, of increasing global competition, complexity, and high levels of customisation, turn the attention of the industry leaders to critical issues of productivity, quality, efficiency, and manufacturing cost. cell design is critical to any cellular manufacturing problem. cell formation (cf) is the first step of cell design. cf is to form the machine cells for the families of parts needed similar manufacturing requirements (sarker, 1996, boutsinas, 2013, and mukattash et al., 2002). identification of machine cells is most important and basic need in the design of a cellular manufacturing system (mukattash et al., 2002, seifoddini, 1998, venugopal and narendran, 1994). the purpose of any cell formation technique is to form a set of mutually independent set of machines each capable of fully processing the part families assigned to it (venugopal and narendran, 1994). it makes cf a complex and tedious task. the cf techniques developed so far can be categorized into number of categories (boutsinas, 2013, yin and yasuda, 2006, papaioannou and wilson, 2010, and yasuda et al., 2005) (i) similarity coefficient based methods, (ii) mathematical programming based methods, (iii) artificial intelligence based approaches, (iv) heuristics / meta-heuristics, and any combination of these. among cf techniques similarity coefficient based methods are more flexible and easy to implement (yin and yasuda, 2006). numerous cell formation methods are developed so far and the counting is still raising. no single algorithm can provide all the desired benefits (mukattash et al., 2002). work by yin and yasuda, (2005), chu and tsai, (1990), shafer and meredith, (1990), and miltenburg and zhang, (1991) may be referred for the comparative study of various cf techniques. the thrust of most cf techniques is to arrange binary part machine incidence (pmi) matrix in such a fashion that maximum possible 1s and minimum possible 0s are arranged inside diagonal blocks (lokesh and jain, 2010). table 1 presents the literature studies related to cell formation. susanto et al. (2009) revealed that about 80% of cf techniques do not consider important production factors like production volume, machine capacity, operation sequence, inter-cell / intra-cell transportation cost, part processing cost, processing time, machine capacity, etc whereas considerations to such production data would make them more 75int. j. prod. manag. eng. (2014) 2(2), 75-84attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2014.2078 http://creativecommons.org/licenses/by-nc-nd/4.0/ realistic and effective, but rarely any such similarity coefficient based simple cf technique has been developed so far. the literature reflects the need of efforts to incorporate production and manufacturing flexibility related data (realistic data) in cf procedures in a simple manner. thus, to abridge this gap, authors in present study developed similarity coefficient / commonality score based simple cell formation heuristic, which possess the capability to handle the production data i.e. operation sequence, production volume, and inter-cell movement cost simultaneously. to demonstrate its ability to handle other production parameters with little modifications, a modification for consideration to part processing cost in addition to above mentioned production data is developed and explained. towards the end the procedure to handle alternate process plans in conjuction with production data by the proposed cell formation procedure is also discussed. the outline of the paper is as follows: section 2 describes development of a commonality score based manufacturing cf procedure. detailed procedure is explained in section 2.1. a cf problem with production data is solved with proposed cf procedure in section 2.2, followed by comparison of results in section 3. further, the proposed approach is modified to consider part processing cost along with other production data viz. operation sequence, production volume, and inter-cell movement cost in section 4. for better understanding of procedure again, one more cf problem is solved in section 4.1. a discussion on the procedure to handle alternate process plans along with above production data is given in section 5. at last the conclusions from the study are drawn and scope for future work is given in section 6. 2. development of cf procedure an exceptional element indicates that part needs to be processed on a machine located outside the manufacturing cell, hence adding towards the intercell movement, which in-turn adds to inter-cell movement cost. total inter-cell movement cost will depend on production volume, per part per move inter cell movement cost, and number of inter-cell moves generated due to exceptional element(s). from a little observation of operation sequence we can infer that a machine could add maximum one inter-cell move per part if it is either at starting or at ending position of the operation sequence of a particular part, otherwise it could add maximum two inter-cell moves (won and lee, 1991). further, while clustering, if two or more machines lie outside the manufacturing cell and they are in consecutive order in operation sequence of a particular part, in this case total inter-cell moves generated by them will be much lesser than the simple sum, for all such cases. it must be taken into consideration while estimating required inter-cell moves or cost accordingly. though all similarity coefficients are intuition based and there is no strict reasoning why one of them is better than others (krushinsky and goldengorin, 2012). jaccard similarity coefficient is found the efficient and most stable one among the twenty table 1. summary of work observed on cell formation with consideration to flexibility and production data author & year data considered lian et al., 2013 multiple identical machines, processing time, set-up time, machine capacity, production volume, cell size, alternative routes gupta et al., 2012 operation sequence ahi et al., 2009 operational time, operation sequence pandian and mahapatra, 2009 operation sequence, operation time paydar and sahebjamnia, 2009 operation sequence susanto et al. 2009 sequence of operations, part-volume, alternative routes kumar and jain, 2008 operation sequence, time, production volume masmoudi et al. 2008 alternative routes kim et al., 2004 machine sequence, alternative routes mahesh and srinivasan, 2002 processing time, alternative routes mukattash et al., 2002 multiple parallel machines, processing time, alternative routes won and lee, 2001 operation sequence, production volume nair and narendran, 1998 operation sequence beaulieu et al., 1997 intra-cell movement cost, material handling cost, machine cost (alternate machines), alternative routes beaulieu et al., 1993 production cost, work load, machine flexibility, routing flexibility 76 int. j. prod. manag. eng. (2014) 2(2), 75-84 attribution-noncommercial-noderivatives 4.0 international kumar & kumar sharma http://creativecommons.org/licenses/by-nc-nd/4.0/ compared in a comparative study of similarity coefficients made by yin and yasuda (2005). for efficiency and stability of proposed cf procedure a variant of jaccard similarity coefficient is used. chow and hawaleshka (1992), claimed that the common source of machine chaining problem seems to be in the implementation of each step of grouping procedure in a disjoint manner. to avoid chaining in the proposed cf procedure, the input from a grouping step is used in next grouping step similar to chow and hawaleshka (1992). 2.1. proposed cf procedure number of minimum exceptional elements does not guarantee the minimum inter-cell moves / movement cost. the prime motive of any cf procedure is to minimize the number of inter-cell moves / inter-cell movement cost (sivraj and sharma, 2012, and arkat et al., 2012) this is also the basis of proposed cf procedure. the steps of proposed cf procedure is explained by a flow chart in figure 1 and elaborated below: step 1: it is dedicated to convert the pertaining data into maximum possible inter-cell movement cost matrix. the elements of this matrix could be obtained as product of ‘production volume of part’, ‘sum of maximum possible moves could be generated by concerned machine for a concerned part’, and ‘per unit per move inter-cell movement cost’. the procedure of conversion is explained in four sub steps from ‘1a’ to ‘1d’ detailed below. any step out of ‘step 1b’, ‘step 1c’, and ‘step 1d’ may be skipped if concerned parameter is not considered. step 1a. make pmi matrix of size m x n, for ‘m’ parts and ‘n’ machines under consideration step 1b. convert it into a matrix indicating maximum possible inter-cell moves for unit production of each part, by considering their respective operation sequence. by using the logic figure 1. flow chart for proposed cf procedure 77int. j. prod. manag. eng. (2014) 2(2), 75-84attribution-noncommercial-noderivatives 4.0 international cell formation heuristic procedure considering production data http://creativecommons.org/licenses/by-nc-nd/4.0/ that a machine could add maximum one inter-cell move per part if it is either at starting or at ending position of the operation sequence of a particular part, otherwise it could add maximum two intercell moves (won and lee, 1991). step 1c. convert this maximum possible intercell moves matrix into a matrix containing maximum possible inter-cell movement cost for unit production of each part, by multiplying each element by their respective per part per move inter-cell movement cost. thus obtained matrix may be referred as the maximum possible inter-cell movement cost matrix for single unit production volume. step 1d. to consider production volume, multiply each element of this matrix by respective production volume. after multiplication with respective production volume this matrix is converted to the maximum possible inter-cell movement cost matrix for given production volume. this resultant matrix is deduced from the information about operation sequence, production volume, and inter-cell movement cost in addition to the machine(s) required for processing of a particular part. step 2: compute the similarity coefficient among all possible machine pairs from the matrix obtained from ‘step 1’ by using a variant of jaccard similarity coefficient. the proposed variant of jaccard similarity coefficient used here is represented by eq. 1. commonality score = a / (a+b+c) (1) where, a → sum of elements common to both machines in concerned machine pair (in this case, maximum possible inter-cell movement cost for parts visiting both machines) a a ijk k n k1 = = = ^ h/ , a(ij)k → elements common to both machines mi and mj, for k = 1 to k = n parts b → sum of values of elements concerned to only first machine in pair (in this case, maximum possible inter-cell movement cost for parts visiting only first machine) b b i kk k n 1 = = = ^ h/ , b(i)k → elements concerned to machine mi but not machine mj, for k = 1 to k = n parts c → sum of values of elements concerned to only second machine in pair (in other terms, maximum possible inter-cell movement cost parts visiting only second machine) c c kk k n j1 = = = ^ h/ , c(j)k → elements concerned to machine mj but not machine mi, for k = 1 to k = n parts step 3: group machine pair having highest value of commonality score, and transform this machine pair into a machine unit mr having elements m(i,j)r , , a if a a a if a a m ,i j i r i r j r j r r j r r i < = $ ^ ^ ^ ^ ^ ^ ^ h h h h h h h * where, m(i, j)r → corresponding elements of machine unit mr, obtained after transformation of machine mi and machine mj into a single machine unit ai(r) & aj(r) → corresponding elements of machine mi and mj respectively. step 4: formulate the tree/ dandogram accordingly (optional) step 5: replace machine mi and mj with machine unit mr in the incidence matrix step 6: stop and assign parts to machine cells so as to maximise the work load inside these cells, if the number of machine cells in the new incidence matrix is either only one or desired number of machine cells or , otherwise proceed to step 2. 2.2. numerical example solved by proposed cf procedure in this section for a good understanding of proposed procedure, the procedure is implemented on the problem of five machines and five parts adapted from won and lee (1991), and given in table 2. ‘p1’, ‘p2’, ‘p3’, ‘p4’, ‘p5’, indicates from part number 1, to part number 5 whereas ‘m1’, ‘m2’, ‘m3’, ‘m4’, ‘m5’ indicates from machine number 1 to machine number 5 respectively in order. table 2. initial data for cell formation problem. part no. operation sequence production volume p1 m2→m4→m2→m4→m5 20 p2 m1→m3 10 p3 m1→m3→m1→m5 50 p4 m4→m2→m4 40 p5 m2→m1→m5→m1→ m2→m1→m5→m1 30 step 1: step 1a: deduce pmi matrix from the data of problem presented by table 2. step 1b: matrix for maximum possible intercell moves for unit production of each part is formulated by considering the potential of every 78 int. j. prod. manag. eng. (2014) 2(2), 75-84 attribution-noncommercial-noderivatives 4.0 international kumar & kumar sharma http://creativecommons.org/licenses/by-nc-nd/4.0/ machine to generate maximum possible intercell moves, by taking care of their respective operation sequence. the resulting matrix is shown in table 3. table 3. maximum possible inter-cell moves matrix for unit production of each part. p1 p2 p3 p4 p5 m1 0 1 3 0 7 m2 3 0 0 2 3 m3 0 1 2 0 0 m4 4 0 0 2 0 m5 1 0 1 0 4 step 1c: inter-cell movement cost is not considered, this step may be skipped. step 1d: deduce matrix in table 3 into the matrix indicating total maximum possible inter-cell moves for given production volume. for this conversion multiply each element of table 3 by their respective production volume. for example entry corresponding to machine ‘m1’ and part ‘p3’ is calculated as 50(3) = 150 i.e. product of ‘production volume of part’ and ‘sum of maximum possible moves could be generated by concerned machine’. matrix developed by this process is tabulated in table 4. table 4. matrix for maximum possible inter-cell moves for given production volume. p1 p2 p3 p4 p5 m1 0 10 150 0 210 m2 60 0 0 80 90 m3 0 10 100 0 0 m4 80 0 0 80 0 m5 20 0 50 0 120 step 2: construct the commonality score matrix on the basis of data available in the resultant matrix from step 1. the similarity coefficient matrix is represented in table 5. table 5. commonality score matrix based on data in table 4. m1 m2 m3 m4 m5 m1 1 {90/(280+140+90} = 0.176 0.297 0 0.436 m2 1 0 0.56 0.355 m3 1 0 0.2 m4 1 0.061 m5 1 step 3: machines ‘m2’ and ‘m4’ have the highest value of commonality score, therefore they must be clubbed to form a machine unit. step 4: it is simply the construction of dendogram or tree. the step may be skipped. step 5: a new data matrix by clubbing machine ‘m2’ and ‘m4’ in a single machine unit is developed and represented in table 6. table 6. new data matrix considering machine 2 and machine 4 as a single machine unit. p1 p2 p3 p4 p5 m1 0 10 150 0 210 m2,4 80 0 0 80 90 m3 0 10 100 0 0 m5 20 0 50 0 120 at this stage (table 6) the number of machine cells are neither optimum nor one, hence, proceed to step 2. final clustered maximum possible movement based machine part incidence matrix is represented in table 7 and machine cells are encircled by bold lines. table 7. final clustered maximum possible movement based machine part incidence matrix. p3 p5 p2 p1 p4 m1 150 210 10 0 0 m5 50 120 0 20 0 m3 100 0 10 0 0 m4 0 0 0 80 80 m2 0 90 0 60 80 after clustering machines as per the scheme tabulated in table 7, we can find the maximum number of possible inter-cell moves is 110 (i.e. 90+20). machine ‘m5’ is only at the end of operation sequence of part ‘p1’, in this case maximum possible inter-cell moves are same as total inter-cell moves required. machine ‘m2’ is neither at the start nor at the end of operation sequence of part ‘p5’, and machine ‘m5’ is used only once in the operation sequence of part ‘p5’, in this case also maximum possible inter-cell moves are same as total inter-cell moves required. hence, total inter-cell moves required for given production volume will be 110. this is also an optimum solution and in the tune of won and lee (1991). 3. comparison of results the results from proposed procedure is compared with the results of some well-known methods. these well-known methods were compared and found better than several other methods in the studies made by their respective authors. basically, the clustering of these cf problems are same as those found by their respective authors. the comparison of results is summarized in table 8. 79int. j. prod. manag. eng. (2014) 2(2), 75-84attribution-noncommercial-noderivatives 4.0 international cell formation heuristic procedure considering production data http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. proposed cf procedure with consideration to part processing cost the proposed procedure is corrected for consideration to part processing cost along with operation sequence, production volume, and inter-cell material handling cost by a little modifications in deduction of cost matrix for computation of commonality score. the desired modification is limited only to step 1 of proposed procedure, and all other steps remain same. the procedural step of deduction of cost matrix (step 1) is explained with the help of a flow chart in figure 2. 4.1. numerical example with modified cf procedure for illustration part processing, and inter-cell movement costs are introduced arbitratrly to problem adopted from won and lee (1991), discussed in section 2.2. the revised problem is given in table 9. part processing cost per operation is 10, 40, 30, table 8. comparison of results. source of problem size of problem (part×machine) exceptional elements required inter-cell moves from both methods remarks proposed procedure source author’s method elbenani and ferland (2012) 8 × 6 6 6 ------same groups as elbenani and ferland (2012), 0/1 matrix only pandian and mahapatra, (2009) 7 × 5 6 6 5 for unit production volume same clustering as of source author ahi, et al., (2009) 20 × 8 10 14 16 for unit production volume from proposed method machine cells are same as ahi et al. (2009) won and lee (2001) 5 × 5 2 2 110 for given production volume same clustering as won and lee (2001) nair and narendran (1998) 20 × 6 9 9 16 for unit production volume same clustering as nair and narendran (1998), and paydar and sahebjamnia (2009)   make  pmi  matrix production  volume  based  pmi   matrix multiply  each  element  of  pmi  with  production  volume  of  corresponding  part   matrix  for  processing  cost  for  given   production  volume multiply  corresponding  entries  by  per   part  processing  cost  on  related  machines   matrix  for  maximum  possible  inter-­‐cell  moves  for   given  production  volume consider  operation  sequence   maximum  possible  inter-­‐cell  movement  cost  matrix multiply  by  inter-­‐cell  movement  cost  per  part total  production  cost  matrix add  the  corresponding  elements  of  both  matrix figure 2. flow chart for modifications required (step 1) to accommodate processing cost. 80 int. j. prod. manag. eng. (2014) 2(2), 75-84 attribution-noncommercial-noderivatives 4.0 international kumar & kumar sharma http://creativecommons.org/licenses/by-nc-nd/4.0/ 25, 20 units on machine ‘m1’, ‘m2’, ‘m3’, ‘m4’, ‘m5’ respectively. in this illustration part processing cost is same for all parts processed on a particular machine but this cost may be different for different part for a particular machine. matrix obtained at various stages of proposed procedure are given from table 10 to table 14. after completing the procedure the final clustered matrix is obtained as detailed in table 14. from final clustered matrix the followings inferences can be deduced: exceptional elements = 2, required intercell moves = 110, required inter-cell movement cost = 310 unit, processing cost for parts processed outside manufacturing cells = 1490 units, and total production cost of exceptional elements for entire lot =1910 units. 5. handling of multiple process plans along with production data by proposed cf procedure in the case, when multiple process plans are to be considered along with operation sequence, production volume, inter-cell material handling cost, and part processing cost. the cf procedure will remain similar as explained above, the only difference is first consider all process plans as process plans for different parts, form manufacturing cells by considering all process plans in the line of mukattash et al. (2002). after cell formation, accept only one process plan for each part, which needs minimum inter-cell movement cost / processing cost for part outside manufacturing cell, and reject all other process plans for that part. 5.1. numerical example solved by proposed modified cf procedure for considerations of alternate process plans, the same are added to the above discussed numerical problem arbitrarily, rest of other data remains same, as detailed in table 15. it becomes a 5 machines, 5 parts, and 10 process plan problem. final clustered matrix obtained from proposed cf table 10. matrix for processing cost for given production volume. p1 p2 p3 p4 p5 m1 0 100 500 0 300 m2 800 0 0 1600 1200 m3 0 300 1500 0 0 m4 500 0 0 1000 0 m5 400 0 1000 0 600 table 11. maximum possible inter-cell movement cost matrix. p1 p2 p3 p4 p5 m1 0 10 300 0 630 m2 120 0 0 80 270 m3 0 10 200 0 0 m4 160 0 0 120 0 m5 40 0 100 0 360 table 12. combined (total production) cost matrix. p1 p2 p3 p4 p5 m1 0 110 800 0 930 m2 920 0 0 1680 1470 m3 0 310 1700 0 0 m4 660 0 0 1120 0 m5 440 0 1100 0 960 table 9. modified data for cell formation problem. part operation sequence production volume inter-cell movement cost / part / move p1 m2→m4→m2→m4→m5 20 2 unit p2 m1→m3 10 1 unit p3 m1→m3→m1→m5 50 2 unit p4 m4→m2→m4 40 1 unit p5 m2→m1→m5→m1→m2→m1→m5→m1 30 3 unit table 14. final clustered combined cost matrix. part p3 p5 p2 p1 p1 m1 800 930 110 0 0 m3 1700 0 310 0 0 m5 1100 960 0 0 440 m2 0 1470 0 1680 920 m4 0 0 0 1120 680 table 13. commonality score matrix (first iteration). m1 m2 m3 m4 m5 m1 1 0.187 0.31 0 0.663 m2 1 0 0.437 0.271 m3 1 0 0.323 m4 1 0.115 m5 1 81int. j. prod. manag. eng. (2014) 2(2), 75-84attribution-noncommercial-noderivatives 4.0 international cell formation heuristic procedure considering production data http://creativecommons.org/licenses/by-nc-nd/4.0/ references ahi, a., aryanezhad, m.b., ashtiani, b., makui, a. (2009). a novel approach to determine cell formation, intracellular machine layout and cell layout in the cms problem based on topsis method. computers & operations research, 36(5): 1478-1496. http://dx.doi. org/10.1016/j.cor.2008.02.012 arkat, j., farahani, m.h., hosseini, l. (2012). integrating cell formation with cellular layout and operations scheduling. int. j. adv. manuf. tech., 61: 637-647. http://dx.doi.org/10.1007/s00170-011-3733-4 beaulieu, a., ait-kadi, d., gharbi, a. (1993). heuristic for flexible machine selection problems. journal of decision systems, 2: 241-253. doi:10.1080/12460125.1993.10511583 procedure is given in table 16. results deduced from final clustered matrix are as follows: exceptional elements = 2, required total intercell moves = 110, required inter-cell movement cost = 220 unit, processing cost for parts processed outside manufacturing cells = 1600 units, and total production cost of exceptional elements for entire production = 1820 units. a little consideration of results obtained without alternate process plans (table 14), and with alternate process plans (table 16), it is clear that procedure selects the process plan which ultimately reduces the total production cost. table 16. final clustered combined cost matrix with alternate process plan. part p1 p2 p3 p4 p5 process plan 6 10 5 1 8 m1 800 720 0 0 0 m5 1100 840 210 440 0 m3 1800 0 310 0 0 m4 0 0 0 660 1080 m2 0 1380 0 920 1680 6. conclusion the proposed cf heuristic procedure is simple, easy to understand, and implement. it has the ability to use the production data such as production volume, operation sequence, and inter-cell movement cost simultaneously. it produces the results which commensurate with some highly advanced and complex cell formation methods requiring very high computational power. the modifications for part processing cost, and alternate process plans demonstrates its ability to handle other production parameters too. the proposed procedure could also be implemented if part processing, and inter-cell movement costs are replaced by part processing, and inter-cell movement time respectively. the procedures would be highly beneficial for a low to mid-size flexible manufacturing system. future work may be carried out in the development of more realistic and efficient formulations with considerations to more realistic parameters such as setup cost, machine capacity, multiple identical machines, decisions on number of manufacturing cells & size, reliability, work imbalance, and various manufacturing flexibility related parameters, etc for large-size cf problems. table 15. data for cell formation problem. part no. process plan operation sequence prod. vol. inter-cell movement cost /part / move p1 1 m2→m4 →m2→m4→m5 20 2 2 m2→m1 →m2 m3→m5 3 m3→m1 m5→m3→m4 p2 4 m1→m4 10 1 5 m3→m5 p3 6 m1→m3 m1→m5 50 2 7 m1→m5→m1→m4 p4 8 m4→m2→m4 40 1 9 m4→m2 →m3 p5 10 m2 →m1→m5 →m1m2→m1 →m5→m1 30 2 82 int. j. prod. manag. eng. (2014) 2(2), 75-84 attribution-noncommercial-noderivatives 4.0 international kumar & kumar sharma http://dx.doi.org/10.1016/j.cor.2008.02.012 http://dx.doi.org/10.1016/j.cor.2008.02.012 http://dx.doi.org/10.1007/s00170-011-3733-4 http://dx.doi.org/10.1080/12460125.1993.10511583 http://creativecommons.org/licenses/by-nc-nd/4.0/ beaulieu, a. gharbi, a., ait-kadi. (1997). an algorithm for the cell formation and the machine selection problems in the design of a cellular manufacturing system. international journal of production research, 35(7): 1857-1874. doi:10.1080/002075497194958 boutsinas, b. (2013). machine-part cell formation using biclustering. european journal of operational research, 230(3): 563-572. http:// dx.doi.org/10.1016/j.ejor.2013.05.007 chow, w.s., hawaleshka, o. (1992). an efficient algorithm for solving the machine chaining problem in cellular manufacturing. computers and industrial engineering., 22(1): 95-100. http://dx.doi.org/10.1016/0360-8352(92)90036-j chu, c.h., tsai, m. (1990). a comparison of three array-based clustering techniques for manufacturing cell formation. international journal of production research, 28(8): 1417-1433. http://dx.doi.org/10.1080/00207549008942802 elbenani, b., ferland, j.a. (2012). cell formation problem solved exactly with the dinkelbach algorithm. https://www.cirrelt.ca/ documentstravail/cirrelt-2012-07.pdf accessed on 25.11.2013. gupta, a., jain, p.k., kumar, d. (2012). formation of part family in reconfigurable manufacturing system using principle component analysis and k-means algorithm. in: annals of daaam for 2012 & proceedings of the 23rd international daaam symposium, volume 23, no.1, ed. b. katalinic, published by daaam international, vienna, austria. jayakrishnan nair, g.j., narendran, t.t. (1998). case: a algorithm for cell formation with sequence data. international journal of production research, 36(1): 157-180. http://dx.doi.org/10.1080/002075498193985 kim, c.o.,baek, j.g., baek, j.k. (2004). a two-phase heuristic algorithm for cell formation problems considering alternative part routes and machine sequences. international journal of production research, 42(18): 3911-3927. doi:10.1080/00207540410001704078 krushinsky, d., goldengorin, b. (2012). an exact model for cell formation in group technology. comput. manag. sci., 9: 323-338. doi:10.1007/s10287-012-0146-2. kumar, l., jain, p.k. (2008). part-machine group formation with operation sequence, time, and production volume. int. j. simul. model, 7(4): 198-209. http://dx.doi.org/10.2507/ijsimm07(4)4.113 lian, j., liu, c.g., li, w.j., evans, s., yin, y. (2013). formation of independent manufacturing cells with the consideration of multiple identical machines. international journal of production research, 52(5):1363-1400. doi:10.1080/00207543.2013.843797 lokesh, k., jain, p.k. (2010). concurrently part-machine groups formation with important production data. int. j. simul. model, 9(1): 5-6. http://dx.doi.org/10.2507/ijsimm09(1)1.133 masmoudi, f., hachicha, w., haddar, m. (2008). a new combined framework for the cellular manufacturing systems design. in: proceedings of the 2008 international conference of manufacturing engineering and engineering management, london: united kingdom. miltenburg, j., zhang, w. (1991). a comparative evaluation of nine well-known algorithms for solving the cell formation problem in group technology. journal of operations management, 10(1): 44-72. http://dx.doi.org/10.1016/0272-6963(91)90035-v mukattash, a.m., adil, m.b., tahboub, k.k. (2002). heuristic approaches for part assignment in cell formation. computers & industrial engineering, 42: 329-341. http://dx.doi.org/10.1016/s0360-8352(02)00020-7 mahesh, o., srinivasan, g. (2002). incremental cell formation considering alternative machines. international journal of production research, 40(14): 3291-3310. doi:10.1080/00207540210146189 pandian, r.s., mahapatra, s.s. (2009). manufacturing cell formation with production data using neural networks. computers & industrial engineering, 56(4): 1340-1347. http://dx.doi.org/10.1016/j.cie.2008.08.003 papaioannou, g., wilson, j.m. (2010). the evolution of cell formation problem methodologies based on recent studies (1997–2008): review and directions for future research”. european journal of operational research, 206(3): 509-521. http://dx.doi.org/10.1016/j. ejor.2009.10.020 paydar, m.m., sahebjamnia, n. (2009). designing a mathematical model for cell formation problem using operation sequence. journal of applied operational research, 1(1): 30-38. sarker, b.r. (1996). the resemblance coefficients in group technology: a survey and comparative study of relational metrics. computers ind. engng., 30(1): 103-116. doi:10.1016/0360-8352(95)00024-0 seifoddini, h. (1998). machine grouping expert systems: comparison between single linkage and average linkage clustering techniques in forming machine cells. computers & industrial engineering 15(14): 210-216. doi:10.1016/0360-8352(88)90088-5 shafer, s.m., meredith, j.r. (1990). a comparison of selected manufacturing cell formation techniques. international journal of production research, 28(4): 661-673. http://dx.doi.org/10.1080/00207549008942747 sivaraj, a., sharma, r.k. (2012). cluster analysis in cellular manufacturing by using proposed algorithm, in: proceedings of 3rd international conference on production and industrial engineering cpie-2013, dr b.r. ambedkar nit jalandhar, india, 192-202. susanto, s., al-dabass, d., bhattacharya, a. (2009). optimised cell formation algorithm considering sequence of operations, alternative routing and part-volume, in: proceedings of third asia international conference on modelling & simulation, 25-29 may, 2009, bali, india. 79-84. doi:10.1109/ams.2009.145 venugopal, v., narendran, t.t. (1994). machine-cell formation through neural models. international journal of production research, 32(9): 2105-2116. http://dx.doi.org/10.1080/00207549408957061 83int. j. prod. manag. eng. (2014) 2(2), 75-84attribution-noncommercial-noderivatives 4.0 international cell formation heuristic procedure considering production data http://dx.doi.org/10.1080/002075497194958 http://dx.doi.org/10.1016/j.ejor.2013.05.007 http://dx.doi.org/10.1016/j.ejor.2013.05.007 http://dx.doi.org/10.1016/0360-8352(92)90036-j http://dx.doi.org/10.1080/00207549008942802 https://www.cirrelt.ca/documentstravail/cirrelt-2012-07.pdf https://www.cirrelt.ca/documentstravail/cirrelt-2012-07.pdf http://dx.doi.org/10.1080/002075498193985 http://dx.doi.org/10.1080/00207540410001704078 http://dx.doi.org/10.1007/s10287-012-0146-2 http://dx.doi.org/10.2507/ijsimm07(4)4.113 http://dx.doi.org/10.1080/00207543.2013.843797 http://dx.doi.org/10.2507/ijsimm09(1)1.133 http://dx.doi.org/10.1016/0272-6963(91)90035-v http://dx.doi.org/10.1016/s0360-8352(02)00020-7 http://dx.doi.org/10.1080/00207540210146189 http://dx.doi.org/10.1016/j.cie.2008.08.003 http://dx.doi.org/10.1016/j.ejor.2009.10.020 http://dx.doi.org/10.1016/j.ejor.2009.10.020 http://dx.doi.org/10.1016/0360-8352(95)00024-0 http://dx.doi.org/10.1016/0360-8352(88)90088-5 http://dx.doi.org/10.1080/00207549008942747 http://dx.doi.org/10.1109/ams.2009.145 http://dx.doi.org/10.1080/00207549408957061 http://creativecommons.org/licenses/by-nc-nd/4.0/ won, y., lee, k.c. (1991). group technology cell formation considering operation sequences and production volumes. international journal of production research, 39(13): 2755-2768. doi:10.1080/00207540010005060 yasuda, k., hu, l., yin, y. (2005). a grouping genetic algorithm for multi-objective cell formation problem. international journal of production research, 43(4): 829-853. http://dx.doi.org/10.1080/00207540512331311859 yin, y., yasuda k. (2005). similarity coefficient methods applied to the cell formation problem: a comparative investigation. computers & industrial engineering 48(3): 471-489. http://dx.doi.org/10.1016/j.cie.2003.01.001 yin, y., yasuda, k. (2006). similarity coefficient methods applied to the cell formation problem: a taxonomy and review. int. j. production economics, 101(2): 329-352. http://dx.doi.org/10.1016/j.ijpe.2005.01.014 84 int. j. prod. manag. eng. (2014) 2(2), 75-84 attribution-noncommercial-noderivatives 4.0 international kumar & kumar sharma http://dx.doi.org/10.1080/00207540010005060 http://dx.doi.org/10.1080/00207540512331311859 http://dx.doi.org/10.1016/j.cie.2003.01.001 http://dx.doi.org/10.1016/j.ijpe.2005.01.014 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2015.3314 received 2014-10-13 accepted: 2014-12-12 integrating forward and reverse logistics network for commercial goods management. an integer linear programming model proposal eva ponce-cuetoi and melisa molenat muelasii industrial engineering and logistics research group. department of industrial engineering, business administration and statistics. escuela técnica superior de ingenieros industriales de madrid, universidad politécnica de madrid josé gutiérrez abascal, 2, 28006 madrid, spain i eva.ponce@upm.es ii melisa.molenat@gmail.com abstract: in this paper, an optimization model is formulated for designing an integrated forward and reverse logistics network in the consumer goods industry. the resultant model is a mixed-integer linear programming model (milp). the objective is to minimize the total costs of the closed-loop supply chain network. it is important to note that the design of the logistics network may involve a trade-off between the total costs and the optimality in commercial goods management. the model comprises a discrete set as potential locations of unlimited capacity warehouses and fixed locations of customers’ zones. it provides decisions related to the facility location and customers’ requirements satisfaction, all of this related with the inventory and shipment decisions of the supply chain. finally, an application of this model is illustrated by a real-life case in the food and drinks industry. we can conclude that this model can significantly help companies to make decisions about problems associated with logistics network design. key words: forward/reverse logistics network, network design, mixed-integer linear programming, facility location, commercial goods. 1. introduction different authors examine the interdependence between facility location, transportation and inventory issues in a distribution network design problem (jayaraman, 1998). in this network design problem, warehouses location plays an important role. teo and shu (2004) study the distribution network design problem focusing on warehouse location by integrating transportation and inventory cost function, but considering only forward logistics. to determine how many warehouses to set up, where to locate them and how to serve many points of sales using these warehouses, reducing costs and maintaining a good service level are not trivial problems. in addition, and particularly if we focus on commercial goods management (such as modular counters, display cabinets, frozen food fridges or retail displays), reverse logistics activities need to be considered in the problem, creating a closed loop system, where those activities are included encompassing the transportation and recollecting of commercial products. few authors have however addressed such problem (jayaraman et al., 2003; lu and bostel, 2007; srivastava, 2008; yongsheng and shouyang, 2008) and further research on the topic is still required (cardoso et al., 2013). the aim of this paper is to develop a logistics network model for helping companies to configure their warehouse network of commercial goods in the consumer goods industry (food and beverage). a mixed integer linear programming model in which reverse logistics activities are considered simultaneously with forward supply chain activities, will be proposed. to determine the optimal number of warehouses that satisfy customers’ requirements with minimum costs is the main contribution of this paper. the paper is structured as follows: in section 2, the main problem characteristics are detailed. in section 3, the mathematical formulation is presented. in section 4, an application and analysis to a real life case study of consumers’ goods company is presented. finally, in section 5, conclusions and future works are discussed. 25int. j. prod. manag. eng. (2015) 3(1), 25-32creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2015.3314 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. method 2.1. problem definition in this paper, a supply chain network design problem for commercial goods is considered, following the mathematical model for the reverse distribution problem proposed in the article written in 2003 by jayaraman et al. it is important to point out that their model can be adapted for different type of products such as end of life and commercial returns, considering reverse functions like recycling, remanufacturing, reuse... the supply chain of the present paper is formed by several warehouses, which store products and deliver them directly to markets. the points of sale (poss) are the last echelon in the considered supply chain. each pos demands commercial goods (such as modular counters, display cabinets, frozen food fridges or retail displays) at the beginning of each summer season. warehouses integrate the received orders and transform them into an assembly service. at the end of the summer season, the product flow is reverse: products are collected by warehouses from pos, in order to store and reuse them in the next summer season. in the reverse product flow, it is assumed that disassembly services consist of the sum of: (1) non-recovered products that are a fraction of the products supply to markets at each time period, which will never be collected at the end of the time period, (2) returned products to be sent to warehouses, to be stored and reuse in a next time period and (3) end of life products to be sent to disposal sites, to be disassembled and recycled. the amount of each type of these products is calculated as a percentage of the total number of disassembly services in each time period. to resume, in this problem there are four services: assembly, disassembly, maintenance (from a technical or a commercial point of view) and storage. the total amount of each service is associated with a percentage of the installed-base of products in the entire market, based on the number of services done in previous years. the fact that the commercial goods are recovered from the forward supply chain and sent again to the warehouses for a new use configures a closed-loop supply chain. figure 1. a closed-loop supply chain is configured. the time horizon is a month, in order to consider the evolution of the consumer’s demand over a year. the problem is modelled through a multi-product mixed integer linear programming (milp) model. the warehouses of the supply chain will be located in a number of pre-determined candidate locations. as stated before, the exact location of pos is known. the aim of this problem is to determine the optimal network structure of the supply chain that minimises the total logistics cost. the problem formulation is detailed in section 2.2. 2.2. problem formulation having in consideration the problem characteristics described previously and following the example proposed by jayaraman et al. (2003), a milp model is developed and formulated. the nomenclature, indices and parameters used are shown in the table 1.1, table 1.2, table 2.1 and table 2.2. 2.2.1. variables a continuous variable is used to define the total cost related to the assembly services of the commercial goods in the points of sale: assemblyc. a continuous variable is used to define the total cost related to the disassembly services of the installed commercial goods in the points of sale: disassemblyc. a continuous variable is used to define the total cost related to the maintenance services to repair goods installed in the points of sale: tecchangec. a continuous variable is used to define the total cost related to the commercial changes: commercialexchangec. as shown in the table 2.1, in the continous variables defined previously, the number of goods that have to be assembled, disassembled or changed for 26 int. j. prod. manag. eng. (2015) 3(1), 25-32 creative commons attribution-noncommercial-noderivatives 4.0 international ponce-cueto, e. and molenat muelas, m. http://creativecommons.org/licenses/by-nc-nd/4.0/ a commercial or technical reason in a pos by a warehouse w ∈ w is converted in a kilometric distance multiplying by the distance between the warehouse w ∈ w and the gravity center of the costumers’ zone z ∈ z. in theory, each assembly, disassembly, technical or commercial change means a round trip, and that is why the kilometric distance is multiplying by 2. but in reality, in some cases the technical service cannot do the activity in the predicted moment (because the pos is closed or because the responsible of the pos is not agreed). in those cases, the technical service has to come back in another moment. that is why the kilometric distance is multiplying in these particular situations by 4. in addition synergies are considered transporting as much goods as it is possible, depending on the size of the box truck and on the dimensions of the items. finally, the total kilometrics are converted in a cost multiplying by a unit kilometric cost. a continuous variable is used to define the total cost related to the storage of commercial goods in the warehouses: storagec. a continuous variable is used to define the total cost related to the opening of the warehouses: openingc. a binary variable is used to define the location of warehouses (equal to 1 if the warehouse w ∈ w table 1.1 sets and parameters used in the model. type identifier definition set w ∈ w set of potential locations for locating warehouses. z ∈ z set of customers' zones. p ∈ p set of products families in function of the dimensional characteristics of each commercial good. parameters percenrepresentp percentage of each products family p ∈ p. kmw,z kilometric distance between the warehouses w and the centre of gravity of the customers' zone z ∈ z. numwarehouses pre-determined number of warehouses considered to design the supply chain network. installedbasez quantity of installed commercial goods in a customers’ zone z ∈ z. installedbasetohibernatez quantity of installed commercial goods to hibernate associated to a customers’ zone z ∈ z. percenassembly percentage of assembled commercial goods of the installed goods. percenassembliesfailed percentage of the assemblies which cannot be done in the predicted moment. percendisassembly percentage of disassembled commercial goods of the installed goods. percendisassembliesfailed percentage of the disassemblies which cannot be done in the predicted moment. percendisassemblieslost percentage of the disassemblies which cannot be done because the good has disappeared (robbery...). percendisassembliesdisposal percentage of the disassemblies which are going to be disposed. percentecchange percentage of technical changes in the installed commercial goods. percentecchangesfailed percentage of the technical changes which cannot be repaired. table 1.2 sets and parameters used in the model. type identifier definition parameters percencomchange percentage of commercial changes of the installed commercial goods. percencomchangesfailed percentage of commercial changes which cannot be done in the predicted moment. percensecuritystock percentage of the security stock considered. percentohibernate percentage of the installed commercial goods to hibernate. truckboxsize surface of the truck box which can be used to transport commercial goods. prodfamilysizep size of the products family p ∈ p. palletsoccupiedp number of pallets occupied by the products family p ∈ p. transportuc variable cost per unit of kilometric distance. warehouseufc fixed cost of opening a warehouse. warehouseuvc variable cost per unit of pallets occupied by the stored products family p ∈ p. 27int. j. prod. manag. eng. (2015) 3(1), 25-32creative commons attribution-noncommercial-noderivatives 4.0 international integrating forward and reverse logistics network for commercial goods management. an integer linear programming model proposal http://creativecommons.org/licenses/by-nc-nd/4.0/ is located in potential location w and equal to 0 otherwise): openingw. a binary variable is used to define the customers’ zone service (equal to 1 if the customers’ zone is served by the warehouse w ∈ w and equal to 0 otherwise): allocationzww,z the objective function minimizes the total logistics cost: min(c)= assemblyc+disassemblyc+tecchangec+ comchangec+storagec+openingc (1) the total cost of the supply chain is the sum of the transportation costs and the facilities costs. the transportation costs are those related to assembly services, disassembly services and technical and commercial maintenance services. the facilities costs are those related to warehouses location and operation. (includes in equation 1 as storagec and openingc. see table 2.2 for variables definition). subject to: sw openingw = numwarehouses (2) sw allocationzww,z = 1, ∀w∈w (3) openingw ≥ allocationzww,z, ∀w∈w, ∀z∈z (4) table 2.1 variables used in the model. identifier definition assemblyc transportuc·σwσz ((installedbasez·percenassembly·allocationzww,z· kmwzw,z· (2· (1-percenassembliesfailed) + 4·percenassembliesfailed))/(truckboxsize/ σp(percenrepresentp·prodfamilysizep)) disassemblyc transportuc·σwσz ((installedbasez·percendisassembly·allocationzww,z· kmwzw,z· (2·(1-percendisassembliesfailed) + 4·percendisassembliesfailed))/(truckboxsize/ σp(percenrepresentp·prodfamilysizep)) tecchangec transportuc·σwσz ((installedbasez·percentecchange·allocationzww,z· kmwzw,z· (2·(1-percentecchangesfailed) + 4·percentecchangesfailed))/(truckboxsize/ σp(percenrepresentp·prodfamilysizep)) commercialexchangec transportuc·σwσz ((installedbasez·percencomchange·allocationzww,z· kmwzw,z· (2·(1-percencomchangesfailed) + 4·percencomchangesfailed))/(truckboxsize/ σp(percenrepresentp·prodfamilysizep)) table 2.1 variables used in the model. identifier definition openingc σw warehouseufcw·openingw openingw binary variable equal to 1 if the warehouse w ∈ w is located in potential location w and equal to 0 otherwise allocationzww,z binary variable equal to 1 if the customers’ zone is served by the warehouse w ∈ w and equal to 0 otherwise. storagec months of the summer season: σw warehouseuvcw·σz ((installedbasez·percenassembly·allocationzww,z) + (installedbasez·percentecchange·allocationzww,z)+(installedbasez·percencomchange· allocationzww,z) + (installedbasez·percenassembly·allocationzww,z)·percensecuritystock – (installedbasez·percendisassembly·allocationzww,z)·(1-percendissembliesdisposal – percendisassemblieslost) · (σp(percenrepresentp· palletsoccupiedp)) months of the winter season: σw warehouseuvcw·σz ((installedbasez·percenassembly·allocationzww,z) +(installedbasez· percentecchange·allocationzww,z)+ (installedbasez·percencomchange·allocationzww,z)+ (installedbasetohibernatez·percentohibernate·allocationzww,z)–((installedbasez· percendisassembly·allocationzww,z)·(1-percendisassembliesdisposal–percendisassemblieslost)· (σp(percenrepresentp·palletsoccupiedp)) 28 int. j. prod. manag. eng. (2015) 3(1), 25-32 creative commons attribution-noncommercial-noderivatives 4.0 international ponce-cueto, e. and molenat muelas, m. http://creativecommons.org/licenses/by-nc-nd/4.0/ openingw ∈ {0,1}, allocationzww,z ∈ {0,1} ∀w∈w, ∀z∈z (5) assemblyc, disassemblyc, tecchangec, comchangec, storagec, openingc ≥ 0 (6) the equation (2) is a logical constraint ensuring that the number of open warehouses is exactly equal to the number of warehouses defined. the equation (3) ensures that all customers’ zones are taken into account and that each one is served by a unique warehouse. inequality (4) ensures that before allocating a customers’ zone to a warehouse, the warehouse has to be opened. finally, constraint (5) is for binary variables and constraint (6) the nonnegative variables constraint. 3. case study application in this section, the formulated model is applied to a real-life industrial case study (based on a consumer goods company, which operates in the food and drink industry). in particular, the case study includes the supply chain network of fridges for ice creams.   figure 2. example of commercial goods considered in the case study. specifically, the company manages the supply chain of goods, controlling the storage, the delivery, the maintenance and repair, the recovery and, if needed, the scrapping of those items. the total of warehouses considered varies between 1 (centralized supply chain network) and 47 (decentralized supply chain network). 47 is the maximum number of warehouses considered since there is only one candidate location for locating warehouses in each spanish province. in the supply chain network more than 140,000 points of sale are considered and aggregated per spanish provinces. the exact location of each point of sale cannot be described due to confidentiality company reasons. the average amount of commercial goods installed in spain in 2013 was 173,461. in this case study, 8 products families are considered. table 3. characteristics of the products family. products family % of each products family number of pallets occupied by the products family 1 22,03% 1 2 5,08% 1 3 6,78% 1 4 1,69% 1 5 38,98% 1 6 3,39% 1 7 3,39% 2 8 18,64% 1 the percentages introduced in the model are the same as the month percentages of the last year. it is assumed that these percentages are the same in all the provinces, without geographical distinction. table 4.1 month percentages considered with respect to all goods installed in pos. month %assemblies %disassemblies january 6,6% 9,5% february 6,5% 10,7% march 13,7% 9,9% april 15,4% 8,4% may 14,1% 7,6% june 13,6% 7,1% july 9,0% 6,7% august 4,7% 4,8% september 3,6% 12,0% october 7,4% 12,6% november 2,1% 6,7% december 3,2% 4,0% 29int. j. prod. manag. eng. (2015) 3(1), 25-32creative commons attribution-noncommercial-noderivatives 4.0 international integrating forward and reverse logistics network for commercial goods management. an integer linear programming model proposal http://creativecommons.org/licenses/by-nc-nd/4.0/ as shown in table 4.1, the end of the summer season is in august because the disassemblies percentage increases and the assemblies percentage decreases in september. so, to calculate the amount of commercial goods to hibernate, the information related to the amount of goods installed in spain in august has to be considered. specifically, 8.3% of goods installed in pos in august have to be storage during the hibernation period (from september to february). table 4.2 month percentages considered with respect to all goods installed in pos. month %commercial changes %technical changes january 4% 5% february 10% 11% march 14% 11% april 17% 11% may 15% 7% june 12% 10% july 10% 10% august 6% 10% september 5% 9% october 3% 6% november 2% 5% december 2% 5% table 4.3 month percentages considered with respect to all goods installed in pos. %assemblies failed 10% %disassemblies failed 10% %technical changes failed 15% %commercial changes failed 12% %disassemblies lost 5% %disassemblies disposal 8% the kilometric distances considered are those defined among capitals of provinces. in order to estimate the transportation cost, the database of the infraestructure ministry of spain, acotram 2.4.0 was consulted. table 5. cost parameters settings in example (source: acotram 2.4.0). description value transportuc 0.93 €/km warehouseufc 18000 €/warehouse warehouseuvc 3.0 €/palet·month in this study, the surface of the truck box which is used to transport commercial goods is equal to 16.17 square meters. 4. results considering all numerical parameters defined in the previous section, the model formulated was solved. the tool used to solve the model is aimms v. 3.13 and the solver cplex. the reported computational results were obtained on a 1.65 ghz amd e series processor, 3 gb ddr3sdram ram, 1333 mhz memory speed and 320 gb sata hard drive, running windows xp professional. the computational time needed to find the results was lower than 1 minute. the results of the model shows the influence of the warehouses number in the annual total cost: figure 3. relationship between the annual total cost and the number of warehouses considered in the model. the minimum point of the annual total cost curve defines the optimal number of warehouses that has to be considered in the developed model: 7 warehouses. in addition, the result of the model shows the location of these 7 warehouses. figure 4 shows, in a schematic way, the locations of all warehouses and the consumers’ zone allocation.   figure 4. model results: warehouse’s locations and consumers’ zones allocation (aimms 3.13). 30 int. j. prod. manag. eng. (2015) 3(1), 25-32 creative commons attribution-noncommercial-noderivatives 4.0 international ponce-cueto, e. and molenat muelas, m. http://creativecommons.org/licenses/by-nc-nd/4.0/ finally, the results of the model specifies the amount of the total logistics costs for this particular case study: --costs related with the transport: o assemblies: 462,437.0€ o disassemblies: 586,741.0€ o technical changes: 1,456,208.0€ o commercial changes: 591,765.0€ --costs related with warehouses: o fixed costs for opening warehouses: 337,645.0€ o variable costs for storage products: 1,412,744.0€ --annual total cost: 4,847,540.0€. as shown, in the developed model, the annual total transport cost represents the 63.9% and the annual total cost related with warehouses represents the 36.1%. figure 5 shows the current situation of the company. 97 warehouses are spread out throughout the spanish geography. each warehouse serves in its zone. so there is in total 97 consumers’ zones defined in this situation. figure 5. current situation in the company. the current total logistics costs for the current situation is: --costs related with the transport: o assemblies: 217,617.9€ o disassemblies: 275,727.8€ o technical changes: 669,812.5€ o commercial changes: 672,451.9€ --costs related with warehouses: o fixed costs for opening warehouses: 4,678,795.0€ o variable costs for storage products: 2,132,487.1€ --annual total cost: 8,646,892.2€ in that model, the annual total transportation cost represents 21.9% and the annual total cost related with warehouses supposes 78.1%. comparing the results obtained with the model developed with the current situation implemented by the company nowadays, the annual total transportation cost has increased 1.69 times, but the installation cost of warehouses has been considerably reduced (from 6,811,282€ to 1,750,389€), so the annual total costs have descreased in a 44%. 5. discussion and conclusions in the present work, it has been proposed an optimization model for the design of a supply chain that integrates forward and reverse flows with application in the consumers’ goods industry. the model developed was applied to the optimization of the supply chain of a spanish company, where cabinets for ice creams were studied. the model application shows that the situation proposed results in a decrease of the total costs (including transportation, fixed costs for opening warehouses and variable costs for storage products in the warehouses). more specifically, in comparison with the current situation, the total costs have decreased in a 44%. considering all its results, it has been shown that the situation described as the solution tends to create a more profitable, simple and efficient supply chain network. as future work, an improvement of model formulation will be explored. having in mind that, in the present work, the level defined by plants and the one defined by infrastructures for used and recovered products which life is ended, are not considered. multi-objective approaches should be then addressed as an extension of the present study. 31int. j. prod. manag. eng. (2015) 3(1), 25-32creative commons attribution-noncommercial-noderivatives 4.0 international integrating forward and reverse logistics network for commercial goods management. an integer linear programming model proposal http://creativecommons.org/licenses/by-nc-nd/4.0/ references acotram. (2014). asistente para el cálculo de costes del transporte de mercancías por carretera. published by the infrastructure ministry of spain. cardoso, s.r., barbosa-povoa, a.p., relvas, s. (2013). design and planning of supply chains with integration of reverse logistics activities under demand uncertainty. european journal of operational research, 226(3): 436-451. doi:10.1016/j.ejor.2012.11.035 jayaraman, v. (1998). transportation, facility location and inventory issues in distribution network design: an investigation. international journal of operations & production management, 18(5): 471-494. doi:10.1108/01443579810206299 jayaraman, v., patterson, r.a., rolland, e. (2003). the design of reverse distribution networks: models and solution procedures. european journal of operational research, 150(1): 128-149. doi:10.1016/s0377-2217(02)00497-6 lu, z., bostel, n. (2007). a facility location model for logistics systems including reverse flows: the case of remanufacturing activities. computers & operations research, 34(2): 299-323. doi:10.1016/j.cor.2005.03.002 srivastava, s.k. (2008). network design for reverse logistics. omega, 36(4): 535-548. doi:10.1016/j.omega.2006.11.012 yongsheng, z., shouyang, w. (2008). generic model of reverse logistics network design. journal of transportation systems engineering and information technology, 8(3): 71-78. doi:10.1016/s1570-6672(08)60025-2 teo, c.p., shu, j. (2004). warehouse-retailer network design problem, operations research, 52(3): 396-408. doi:10.1287/opre.1030.0096 32 int. j. prod. manag. eng. (2015) 3(1), 25-32 creative commons attribution-noncommercial-noderivatives 4.0 international ponce-cueto, e. and molenat muelas, m. http://dx.doi.org/10.1016/j.ejor.2012.11.035 http://dx.doi.org/10.1108/01443579810206299 http://dx.doi.org/10.1016/s0377-2217(02)00497-6 http://dx.doi.org/10.1016/j.cor.2005.03.002 http://dx.doi.org/10.1016/j.omega.2006.11.012 http://dx.doi.org/10.1016/s1570-6672(08)60025-2 http://dx.doi.org/10.1287/opre.1030.0096 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2016.4534 received 2016-01-12 accepted: 2016-01-13 understanding organisational engineering angel ortiz centro de investigación de gestión e ingeniería de la producción (cigip). universitat politècnica de valència. camino de vera s/n, edificio 8b acceso l nivel 2. 46022 valencia (spain). aortiz@cigip.upv.es abstract: this paper analyses the concept of organisational engineering by stressing the main functions or phases to be developed and linking them with enperprise management but also emphasizing the differences. the paper also states that organisational engineering is well-founded and robust discipline. key words: organisational engineering, industrial engineering, enterprise management, design, build, operate, improve. 1. introduction this paper is about organizational engineering (oe). the objectives of the paper are twofold: firstly, to introduce and clarify the concepts of organisational engineering and analysing the differences with enterprise management (em); secondly, to demonstrate that oe is powerful engineering. this paper attempts to avoid confusion and clarify concepts. section 2 is devoted to some definitions, section 3 examines the differences between oe and em, section 4 demonstrates that oe is a discipline, and the paper finishes with a conclusions section. 2. definitions this section introduces the definitions of the main terms used in the paper. according to merriamwebster (2015): engineering: the work of designing and creating large structures (such as roads and bridges) or new products or systems by using scientific methods management: the act or skill of controlling and making decisions about a business, department, sports team, etc. organisation: a company, business, club, etc., that is formed for a particular purpose enterprise: a business organisation these four words have lots of combinations used in natural language and in research (i.e. enterprise engineering, enterprise management, engineering management, organisational engineering, organisational management). the next few paragraphs include some definitions, and some additional ones are found in section 3: enterprise engineering (vernadat, 1996): the art of understanding, defining, specifying, analysing, and implementing business processes for the entire enterprise life cycle, so that the enterprise can achieve its objectives, be cost-effective, and be more competitive in market environment. enterprise management (de miguel, 2005): the functions of plan, organise, lead and control and enterprise. figure 1. enterprise management. engineering management (ieee, 1990): the discipline that addresses making and implementing decisions for strategic and operational leadership in current and emerging technologies and their impact on interrelated systems. int. j. prod. manag. eng. (2016) 4(1), 1-3creative commons attribution-noncommercial-noderivatives 4.0 international 1 http://dx.doi.org/10.4995/ijpme.2016.4534 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3. organisational engineering: definition and interactions with enterprise management in this paper organisational engineering is used as being synonymous of the term industrial engineering. industrial engineering according to merriam-webster (2015), is the engineering that deals with the design, improvement, and installation of integrated systems (as of people, materials, and energy) in industry. therefore, why do we use organisational engineering instead of industrial engineering? because in some countries (including spain), industrial engineering has a broader sense (it includes subjects which, in other countries, are organised as several branches of engineering, such as mechanical engineering, electrical engineering, and many more. a broad analysis of this topic can be found in (companys et al., 2016). now the use of the term organisational engineering (oe) is clarified, oe is defined herein as: organisational engineering is engineering that deals with designing, building, operating and improving organisations (like processes, people, materials, information and money). figure 2. organisational engineering. it is interesting to note that enterprise management is developed in the operating phase of oe (figure 3). figure 3. organisational engineering vs. enterprise management. figure 3 presents some ideas to discuss. the first one is that an organisational engineer addresses every stage (design, build, operate and improve) of oe. this is not obvious in other engineering disciplines. let me translate the schema to “traditional” engineering, for example, mechanical engineering, and how this life cycle is implemented in a university programme. normally, a major part of the program focuses on the design phase and how to facilitate the phases below to other actors: build; good designs and specific tasks, such as drawings, constructive instructions, and so on, will facilitate skilled workers the construction of a, e.g., a car operate; a good design, thinking about the human-machine interface and appropriate operative instructions will facilitate the car’s operation for users improve; sensors and indicators will provide engineers with information to improve the car. sometimes the mechanical engineer can be involved in other phases (i.e. it is usual in formula 1 that engineers build the car), but it is not usual that such engineers generally participate in the operate phase (i.e. a formula 1 driver is not usually a mechanical engineer). this situation differs slightly when analysing this aspect in oe. normally, oe curriculy strongly emphase knowledge about the operate part of the organisation. this is important added value for organisational engineers. they are able to execute all the life-cycle phases of oe, but there is also a risk. the risk is to confuse engineering with management in the oe discipline, and to train engineers as managers and not as engineers with the capacity to design, build, operate (manage) and improve organisations. oe universities programmes must balance the time spent on different phases to train good organisational engineers, and it must be accepted that oe is an engineering discipline. 4. organisational engineering as a discipline according to (liles et al. 1995) a discipline has six basic characteristics: (1) a study focal point; (2) a world view or paradigm; (3) a set of reference disciplines used to establish the discipline; (4) principles and practices associated with the discipline; (5) an active research or theory development agenda; (6) the deployment of education and the promotion of professionalism. int. j. prod. manag. eng. (2016) 4(1), 1-3 creative commons attribution-noncommercial-noderivatives 4.0 international ortiz, a. 2 http://creativecommons.org/licenses/by-nc-nd/4.0/ the oe definition is found in section 3. according to the definitions, the study focal point can be the development of architectures, methods and tools for oe (ortiz et al., 1999). extensive information about points (2), (3), (4) and (5) on oe can be found in carrasco et al. (2015). oe (also know as industrial engineering in many countries) has a long-standing track record in education and professionalism. therefore, following the proposal of liles, oe is also demonstrated as a discipline. 5. conclusions. this paper clarifies what oe is. the definition and main phases of the oe life cycle are presented, as well as the link with enterprise management. additionally, oe has also been demonstrated as a founded discipline. references carrasco, j., mataix c., carrasco-gallego r. (2015). organisational engineering: the emerging state of industrial engineering. 9th international conference on industrial engineering and industrial management, aveiro, portugal. july 6-8, 2015. companys, r., lario f.c., vicens, e., poler, r., ortiz, a. (2016). an approach to the industrial organisation engineering background in spain. lecture notes in management and industrial engineering. springer. de miguel fernandez, e. (2005). introducción a la gestión (management). editorial universitat politècnica de valència. liles, d.h., johnson, m.e., meade, l.m., underdown, d.r. (1995). enterprise engineering: a discipline?, society for enterprise engineering conference proceedings. ortiz, a., lario, f., ros, l. (1999). enterprise integration business processes integrated management: a proposal for methodology to develop enterprise integration programs. computers in industry, 40(2): 155-171. vernadat, f. (1996). enterprise modeling and integration. priciples and applications. chapman&hall. int. j. prod. manag. eng. (2016) 4(1), 1-3creative commons attribution-noncommercial-noderivatives 4.0 international understanding organisational engineering 3 file:///volumes/lacie/%5b%2b%5d%20trabajos%20pendientes/upv/*ojs/ijpme/vol%204-1/4534/orig/javascript:opengatewaylink('http://gateway.webofknowledge.com/gateway/gateway.cgi?gwversion=2&srcauth=rid&srcapp=rid&destlinktype=fullrecord&destapp=all_wos&keyut=000083470400007') file:///volumes/lacie/%5b%2b%5d%20trabajos%20pendientes/upv/*ojs/ijpme/vol%204-1/4534/orig/javascript:opengatewaylink('http://gateway.webofknowledge.com/gateway/gateway.cgi?gwversion=2&srcauth=rid&srcapp=rid&destlinktype=fullrecord&destapp=all_wos&keyut=000083470400007') http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2015.3318 received 2014-10-16 accepted: 2015-02-28 fuzzy maintenance costs of a wind turbine pitch control device mariana carvalhoa,b, eusébio nunesb,i and josé telhadab,ii a polytechnic institute of cávado and ave, barcelos, portugal a mcarvalho@ipca.pt b centro algoritmi, university of minho, braga, portugal b,i enunes@dps.uminho.pt b,ii telhada@dps.uminho.pt abstract: this paper deals with the problem of estimation maintenance costs for the case of the pitch controls system of wind farms turbines. previous investigations have estimated these costs as (traditional) “crisp” values, simply ignoring the uncertainty nature of data and information available. this paper purposes an extended version of the estimation model by making use of the fuzzy set theory. the results alert decision-makers to consequent uncertainty of the estimations along with their overall level, thus improving the information given to the maintenance support system. key words: wind turbine, pitch control, maintenance cost, fuzzy sets. 1. introduction wind power technology is one of the major growing areas in the energy sector. in a few years’ time wind power has gone from a minor energy source to a large-scale industry. proper and well-planned service and maintenance strategies are very important to ensure an efficient energy production and required to effectively reduce the costs associated with wind turbine (wt) support. maintenance management approaches aim to find a sound balance between costs and benefits of performing maintenance. some experiments and studies show that there is a large potential to reduce overall costs in the maintenance of wts (e.g. bertling et al., 2006). according to morthorst (2003), operation and maintenance (o&m) costs constitute a sizeable share of the total annual costs of a wt. for a new machine, o&m costs might easily have an average share over the lifetime of the wt of approximately 20% to 25% of the total cost per kwh produced. in an attempt to shed some light on this problem, other works (e.g. carvalho et al., 2013a) has been focused on studying the active power control system or pitch control system of wts. this system assumes primordial importance in the wind turbine, because: i) it is crucial in the optimization of the turbine efficiency; ii) it is very important with regard to the safety of the turbine (naranjo et al., 2011); and iii) reveals frequent failures and large residence time in failure state compared to other systems of the machine (nilsson & bertling, 2007; carvalho et al., 2013b). consequently, to guarantee a normal operation, they are usually needs maintenance actions, which are only provided by the manufacturer (teresa, 2007). moreover, information related to failure modes, (un) availability and maintenance costs of these systems remain confidential and only the manufacturer has knowledge about them. this situation does not facilitate, for example, the work of company managers who search for better warranty and maintenance contracts. in complex systems, such as pitch control systems, the maintenance management function is commonly supported by analyses of collected data as well as on the quality and experience of maintenance engineers and others experts in this field. in this context, this function is often very difficult, and unrealistic decisions come out from the process with undesirable frequency. so, it is expected that the fuzzy set theory, applied in the maintenance management, will lead to more realistic decisions. 103int. j. prod. manag. eng. (2015) 3(2), 103-109creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2015.3318 http://creativecommons.org/licenses/by-nc-nd/4.0/ the study presented in this paper is based on an analysis of two years data collected from 21 identical wts installed in a wind farm in portugal. the data were provided confidentially by the company that manages the wind farm. for this reason, the name of the company and the wt brand are not revealed. each wt under analysis has a nominal power of 2 mw, three rotor blades and an active power control (pitch). the main objective of the study consists on reporting the gathering process of information about wt functioning and its failures and costs, and conducting some reliability analyses, providing an estimate of the associated maintenance costs of the pitch system. the remainder part of the paper is organized as follows. section 2 introduces the fundamentals of the fuzzy set theory. section 3 describes the system under study, the pitch control of the wts, and its fault and error states. in section 4, it is proposed a model for the fuzzy maintenance costs of the pitch control system. section 5 reports the results of the application of the proposed model and discusses its practical relevance. finally, the main conclusions of this study are discussed in section 6. 2. fuzzy set theory 2.1. introduction the fuzzy set theory has been extensively studied in the past 30 years, largely motivated by the need for a more expressive mathematical structure to deal with human factors. this theory has a major impact on industrial engineering and maintenance management systems. during the last decade, several models for maintenance management problems have been incorporating uncertainty of their parameters by using fuzzy sets (e.g. yuniarto and labib, 2006; khanlari et al., 2008; sharma et al., 2008; shen et al., 2009). nevertheless, most of the current literature on maintenance simply omits the uncertainty that is inherent to real processes. fuzzy sets are adequate, for instance, to estimate the lifetime or the failure rate of a given equipment that operates in different environments. in most cases, statements in plain language may be the best form to express the knowledge about a system. however, this information is naturally very inaccurate and any realistic estimate inferred from that is always an approximation. 2.2. basic concepts a fuzzy set a, in the universe of discourse x, is defined by a membership function, μa(x): x→[0,1], which assigns, for each element of x, a membership degree to a. definition 1: given a fuzzy set a defined on x and any number αî(0, 1], the α-cut set, aα, is the crisp set expressed by eq. (1). aα ={x: a(x) ≥ α} (1) the α-cut set concept allows us to manipulate fuzzy sets by using the interval arithmetic. alternatively, such manipulation can be performed by the extension principle introduced by zadeh (1975). this is an important tool by which classical mathematical theories can be fuzzified. on the other side, defuzzification is the conversion of a fuzzy quantity to a crisp quantity. despite the fact that the bulk of the information emerging every day is fuzzy, most of the actions or decisions implemented by humans or machines are crisp or binary. a detailed application of defuzzification methods can be found in klir and yuan (1995). among the innumerous types of fuzzy sets, those that are defined in the set of the real numbers assume a particular importance. these sets have a quantitative meaning and under certain conditions they can be treated as fuzzy numbers, e.g. when intuitively and linguistically they represent approximate numbers, such as “the preventive maintenance duration is around τ hours” (ross 1995). in reliability and maintenance studies, the triangular and trapezoidal numbers are the most used number patterns because their simplicity and adequacy on representing uncertainty, vagueness and subjectivity of data and human judgment. without loss of generality, this paper only deals with triangular fuzzy numbers. a triangular fuzzy number, a, can be defined by a triplet (a1, a2, a3), where μa(a1) = μa(a3) = 0 and μa(a2) = 1. each α-cut, a α, is a closed interval represented as aα = [a1 α, a3 α], where: a1 α = (a2 – a1) α + a1 a3 α = a3 – (a3 – a2) α (2) the family of cut sets {aα: αî(0, 1]} is a representation of the fuzzy number a. an illustrative graphic representation of a is shown in figure 1. 104 int. j. prod. manag. eng. (2015) 3(1), 103-109 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho, m., nunes, e. and telhada, j. http://creativecommons.org/licenses/by-nc-nd/4.0/ x1 αa 1 α aµ a a1 a3a2 3 αa figure 1. representation of a triangular fuzzy number. like functions, f, are important in mathematical modeling, fuzzy functions, f̃, are important in fuzzy modeling. the usual way of obtaining a fuzzy function is to extend a function to map fuzzy sets to fuzzy sets. there are two common methods to accomplish this extension: the extension principle procedure; and the α-cut and interval arithmetic procedure. this paper uses the extension principle (zadeh 1975). the extension principle may be generalized to functions of many independent variables x1, x2,…, xn (buckley & eslami, 2002). let a1, a2,…, an be triangular fuzzy numbers of x1, x2,…, xn respectively, represented by the α-cuts: a1 α=[a11 α, a13 α], a2 α=[a21 α, a23 α],…., an α=[an1 α, an3 α]. using the extension principle, it is possible to extend f to f ̃, where b=f(a1, …, an). if b α=[b1 α, b3 α], then b1 α=min{f(x1, …, xn): x1∈a1 α, …, xn∈an α} b3 α=max{f(x1, …, xn): x1∈a1 α, …, xn∈an α} note that min and max can be used in these equations, for the reason that a continuous function in a closed interval takes its maximum and minimum. therefore, if there are two triangular fuzzy numbers, a1 and a2, and supposing that: ∂f ⁄ ∂x1>0 and ∂f ⁄ ∂x2<0 that is, f is an increasing (decreasing) function in x1(x2), hence, for all α: b1 α = f(a11 α, a21 α ) and b3 α=f(a13 α,a23 α ) (3) 3. pitch control device the main purpose of a pitch system is to prevent the power of the electric generator from being exceeded in the case of high wind speeds, as well as preventing relieve strain on the structure and components of the wind turbine. this system acts on the aerodynamic forces by controlling the loads and power. in the active control, blades may undergo rotation about its longitudinal axis, which makes it changes the angle of attack of the blades with respect to the relative wind speed. this process takes place through hydraulic or electric systems, which respond to an electronic control which checks the power output. whenever the power is too high, the control triggers the mechanism. the main advantages of this system are related to its capability to limit the power for high wind speeds, facilitating the start-up operation, to diminish the efforts and to optimize power when the turbine is operating at partial load. the pitch system also assumes primordial importance with regard to the safety of the turbine. a flaw in this system, combined with adverse climatic situation (e.g. a storm) may lead to an uncontrolled rotation speed of the blades and catastrophic consequences, including, in the limit, a total destruction of the turbine. the states that actively influence the reliability of the active control power are: fault blade load control (state s1): the control effort in the turbine is constantly monitored. this state means that an undue effort has been exercised in the blade. the wind turbine is still operating, but it is under reduced power. the maintenance service has to rectify stress effects. this state actively influences the state s2. pitch control error (state s2): the angles of the three blades are continuously monitored. when there is a difference in the angles of the blades (can be erroneously due to a measurement error), state s2 arises, which leads to a shutdown of the turbine and the engine restarts automatically. if the problem persists for a predefined number of times, the maintenance service will have to repair the fault. 4. fuzzy maintenance costs of the pitch control device recent studies have been emphasizing the importance of the pitch system for the functioning, cost optimization and security of the wind turbine (carvalho et al., 2013a). in this study it was very difficult to estimate the maintenance costs related to these two states, s1 and s2. the wind farm company only knows the information that can be observed from the data made available to this study. from that, with relevance to the analysis of 105int. j. prod. manag. eng. (2015) 3(2), 103-109creative commons attribution-noncommercial-noderivatives 4.0 international fuzzy maintenance costs of a wind turbine pitch control device http://creativecommons.org/licenses/by-nc-nd/4.0/ maintenance costs, one can highlight the record of the exact time of occurrence of each state in each turbine and the wind speed at the time of occurrence. such information allows estimating the downtime cost, for states s1 and s2, as well as the number of preventive and corrective maintenances carried out in two years. however, the costs of corrective and preventive maintenance, and the number of replacements made, were not revealed by the wt manufacturer (who performs maintenances as well). one can only know approximate values from the experience of managing experts from this and from others wind farms who were consulted in the context of this work. therefore, part of the information collected does not follow statistical analysis, but rather statements of experts, based on their knowledge and experience and, consequently, they are subject to an increased level of uncertainty. 4.1. maintenance costs knowing the costs of maintenance, albeit mere approximations, allows the wind farm company to make better decisions, particularly with regard to contracts for the maintenance established with the manufacturer. however, this information either does not exist or is not public. in the context of this study, the contract that the wind farm company has with the manufacturer assumes the execution of four interventions per year in each turbine, conducted at quarterly intervals. specifically, the manufacturer performs an electrical preventive maintenance, a mechanical preventive maintenance, a visual inspection and a lubrication operation. the manufacturer is also responsible for any corrective maintenance that is necessary, as well as some improvement maintenance he may consider as fundamental. this maintenance provided by the manufacturer is a necessary condition to offer warranty to the wind farm company. associated costs are 38000€ per year per turbine by 15 years. in reality, it is not possible to know the exact cost for each preventive and corrective maintenance, since the maintenance contract does not explicit these costs. 4.1.1. costs of unavailability table 1 summarizes the frequency, duration and cost of the resulting unavailability of states s1 and s2, for the 21 turbines, in the two years. table 1. resume of the effects of states s1 and s2 for the 21 machines in two years. state s1 s2 n.º of occurrences 196 431 unavailable time (hh:mm:ss) 949:55:08 1609:32:52 unavailable cost (€) 108276.40 129733.20 average unavailable time (h) 5 4 average unavailable cost (€) 552 301 the cost of downtime shown in the last row of table 1 was estimated as a function of wind speed records and the ratio of power with wind speed, displayed in figure 2. figure 2. power curve as a function of wind speed. the data of average wind speed and the respective wasted power by the turbine, and other details, resulting from the appearance of these two states can be found in qiu et al. (2012). the energy wasted by the occurrence of each state is given by: energy [mwh]= (power[kw]×mdt[h])/1000 where the mean downtime (mdt) is the average time that a system is non-operational for being either in state s1 or state s2. the cost of down-time was estimated supposing that the energy produced is sold at 90€ per mwh. more details about this estimation can be found in carvalho et al. (2013b). 4.1.2. preventive maintenance costs for the preventive maintenance of the pitch system, experts mentioned that a preventive maintenance of the active power control system costs at least 580€ and requires about 4 hours, which additionally 106 int. j. prod. manag. eng. (2015) 3(1), 103-109 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho, m., nunes, e. and telhada, j. http://creativecommons.org/licenses/by-nc-nd/4.0/ represents an approximate unavailability cost of 293€ due to preventive maintenance, assuming an average wind speed of 8 m/sec. 4.1.3. corrective maintenance costs analyzing data from the 431 recorded instances of state s2, it resulted in 53 repairs, i.e. 12.3% of cases requiring maintenance. the maintenance usually consists in replacing the engine of blades. to ensure machine availability, repairs are never made on site. the engine is replaced and thereafter is repaired in the manufacturer. the experts indicate that the cost of each engine is around 2000€. among the 196 occurrences of state s1, 46 triggered a corrective maintenance, which corresponds to 23.5% of all cases. when the maintenance team makes an intervention in the failure load control (s1), usually they replace the sensors that somehow quantify the load exerted on the blade. a new sensor costs around 50€. 4.2. total maintenance costs the total maintenance cost of the active power control system for the 21 machines in the two years can be given by eq. (4). c=ccms1×ncms1+ccms2×ncms2+2×4×21× ×(cpm+cupm)+nos1×cus1+nos2×cus2 (4) where: c: total cost of maintenance of the active power control system for the 21 machines in the two years; ccms1: corrective maintenance cost of state s1; ccms2: corrective maintenance cost of state s2; cpm: preventive maintenance cost of pitch sytem; cupm: unavailability cost, due to preventive maintenance; cus1: average unavailability cost, due to state s1; cus2: average unavailability cost, due to state s2; ncms1: number of corrective maintenance of the state s1; ncms2: number of corrective maintenance of the state s2; nos1: number of occurrences of the state s1; nos2: number of occurrences of the state s2. table 2 presents estimates for the values of these parameters. these estimates were calculated from data provided by the management of the wind farm and information obtained from interviews with managers of the park. applying these values in eq. (4), it was estimated the amount of 492,887€ to the cost spent on pitch system maintenance of the 21 turbines of the wind farm, in the two years under review. thus, on average, the annual maintenance cost of each active power control system was around 11735€. note that the only available information for the maintenance cost is that which prevails at the contract between the company and the manufacturer, i.e. 38000€ per year per turbine. as mentioned above, the true costs of corrective and preventive maintenance, and the number of replacements made, were not revealed by the wt manufacturer. thus the total maintenance cost obtained by eq. (4) contain a certain level of uncertainty which depend of their parameters uncertainty. some issues may arise at this point, such as: what is the confidence level for the total value of the maintenance cost obtained by eq. (4)? how to represent non-probabilistic uncertainty present in some of the cost components? how the uncertainty in the cost components affect the uncertainty in the total cost of maintenance? in the following section, these issues will be addressed using the theory of fuzzy sets introduced above, in section 2. 5. fuzzy total maintenance cost analysis consider the same parameters of eq. (4), but admit now that the uncertainty inherent to the following parameters must be not neglected: preventive maintenance cost, cpm, and unavailable costs due to preventive maintenance, cupm. estimates for these costs (table 2) have a very fragile analytical basis due to limited access to the data (these are not provided by the companies providing maintenance services to the park), so it is assumed that the uncertainty associated with these costs is high. the preventive maintenance cost, cpm, for example, is exclusively known by the experts’ opinion. the unavailable costs due to preventive maintenance, cupm, are also very uncertainty, because it is assumed an average wind speed of 8 m/s. the uncertainty associated with these parameters was appraised from the great experience and indispensable collaboration of two managers of the park. by consensus, the managers presented, for each of these cost parameters, the value they considered most plausible, and the values below and 107int. j. prod. manag. eng. (2015) 3(2), 103-109creative commons attribution-noncommercial-noderivatives 4.0 international fuzzy maintenance costs of a wind turbine pitch control device http://creativecommons.org/licenses/by-nc-nd/4.0/ above from which they consider as impossible to occur. based on this information, is was set up the fuzzy triangular numbers for cpm and cupm. table 3 shows the parameters of the eq. (4), assuming those as triangular fuzzy numbers. table 3. crisp and fuzzy parameters estimates of the maintenance cost function. parameter estimative (€) ccms1 50 ccms2 2000 c̃pm (450, 580, 800) c̃upm (180, 293, 350) cus1 552 cus2 301 eq. (4) can now be rewritten as: c̃= ccms1×ncms1+ccms2×ncms2+2×4×21× ×(c̃pm+c̃upm)+nos1×cus1+nos2×cus2 (5) using the extension principle, c extends to c̃, where c̃=c(cp̃m, c̃upm, ccms1, … ). if cα = [c1 α, c3 α], by eq. (2) and eq. (3) results: c1 α =c[(580–450)α+450, (293–180)α+180, 50,…] and c3 α=c[800–(800–580)α, 350–(350–293)α, 50,…] then, by eq. (5), the maintenance cost will be between 452,063€ and 539,423€. these values determine the confidence interval of the total maintenance cost c (universe of discourse c). the higher the magnitude of this interval, the greater is the uncertainty present in the cost. figure 3 represents this result graphically, as a triangular fuzzy number. 452063 492887 539423 cost € 0.2 0.4 0.6 0.8 1 a figure 3. fuzzy maintenance cost. by determining the mean total cost for each turbine, cw̃t, it is obtained the fuzzy number: cw̃t =c̃/21×2=(452063, 492 887, 539423)/42 = (10763, 11735, 12843) this approach seems to give rise to more realistic solutions, allowing for better decisions in decision making processes. on the other hand, the difficulty of interpreting the results increases. these difficulties are due to the large quantity of possible outcomes for cwt, represented by the universe of discourse of c̃wt. the way to reduce uncertainty in the value of cwt involves obtaining more information about cpm and cupm, thus reducing the universe of discourse of the fuzzy parameters c̃pm and c̃upm. figure 4 shows the fuzzy maintenance cost, c̃, when the universe of discourse of c̃pm and c̃upm is reduced by 30%. in this case, we had set c̃pm=(490, 580, 735) and c̃upm=(210, 293, 330). it is thus noted that the uncertainty reduction of about 30% of c̃pm and c̃upm leads to the same level of reduction in the uncertainty of c̃. 463823 492887 525143 cost € 0.2 0.4 0.6 0.8 1 a figure 4. fuzzy maintenance cost (reducing uncertainty). frequently, the membership function is defuzzificated to obtain a crisp number. however, a lot of information that can be relevant to the decision process is lost in the defuzzification operation. that is, the fuzzy result is richer than the crisp result, and the former should be preferred whenever possible. 6. conclusions in complex systems it is impossible to has a perfect knowledge about the involved parameters (failure rates, unavailability times, etc.) and about their interdependency relationships. considering these results as “crisp” values is equivalent to assume that there is no uncertainty in these data. but, in fact, the uncertainty of data is intrinsic to the system and it is not probabilistic. to overcome these limitations, the application of the fuzzy set theory proves to be 108 int. j. prod. manag. eng. (2015) 3(1), 103-109 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho, m., nunes, e. and telhada, j. http://creativecommons.org/licenses/by-nc-nd/4.0/ an interesting approach for capturing the vagueness and fuzziness of the cost parameters. the application of this theory allows to propagate the uncertainty from parameters to results in the modelling process. the study reported in this paper has demonstrated the validity of these conclusions in the case of a particular maintenance cost problem. moreover, the proposed fuzzy modelling approach will allow managers to make their decisions based on a reacher set of information than that they would have by the application of tradicional crisp valued parameters approach. acknowledgements this work was financed with feder funds by programa operacional fatores de competitividade – compete and by national funds by fct – fundação para a ciência e tecnologia, project: fcomp-01-0124-feder-022674. references bertling, l., ackermann, t., nilsson, j., ribrant j. (2006). pre-study on reliability-centered maintenance for wind power systems with focus on condition monitoring systems. elforsk report 06:39, may 2006. buckley, j. j., eslami, e. (2002). an introduction to fuzzy logic and fuzzy sets. physica-verlag, heidelberg, new york. doi:10.1007/9783-7908-1799-7 carvalho, m., nunes, e., telhada, j. (2013a). maintenance costs of a pitch control device of a wind turbine. proceedings of the world congress on engineering 2013, july 3-5, london, 569-574. carvalho, m., nunes, e., telhada, j. (2013b). state-space characterization and estimation of unavailability costs of a wind turbine. proceedings of the international conference on industrial engineering and operation management 2013, july 10-12, valladolid, spain. hong, d. h. (2006). renewal process with t-related fuzzy inter-arrival times and fuzzy rewards. information sciences, 176(16): 2386-2395. doi:10.1016/j.ins.2005.06.008 khanlari, a., mohammadi, k., sohrabi, b. (2008). prioritizing equipments for preventive maintenance (pm) activities using fuzzy rules. computers & industrial engineering, 54(2): 169-184. doi:10.1016/j.cie.2007.07.002 klir, g. j., yuan, b. (1995). fuzzy sets and fuzzy logic: theory and applications. englewood cliffs, nj, prentice-hall. morthorst, p. e. (2003). wind energy, the facts, costs and prices. risø national laboratory, denmark, vol. 2, 94-110. naranjo, e., sumper, a., bellmunt, o., ferre, a., rojas, m. (2011). pitch control system design to improve frequency response capability of fixed-speed wind turbine systems. european transactions on electrical power, 21(7): 1984-2006. doi:10.1002/etep.535 nilsson, j., bertling, l. (2007). maintenance management of wind power systems using condition monitoring systems-life cycle cost analysis for two case studies. ieee transactions on energy conversion, 22(1): 223-229. doi:10.1109/tec.2006.889623 nunes, e., faria, j., matos, m. (2006). using fuzzy sets to evaluate the performance of complex systems when parameters are uncertain. safety and reliability for managing risk. in c. guedes soares and e. zio (eds.). proceedings of the esrel 2006, 3: 2351-2359. estoril, portugal, taylor & francis group. qiu, y., feng, y., tavner, p., richardson, p., erdos, g., chen, b. (2012). wind turbine scada alarm analysis for improving reliability. wind energy, 15(8): 951-966. doi:10.1002/we.513 ross, t. j. (1995). fuzzy logic with engineering applications. mcgraw-hill, inc. sharma, r.k., kumar, d., kumar, p. (2008). fuzzy modelling of system behavior for risk and reliability analysis. international journal of systems science, 39(6): 563-581. doi:10.1080/00207720701717708 shen, q., zhao, r., tang, w. (2009). random fuzzy alternating renewal processes. soft computing, 13(2): 139-147. doi:10.1007/s00500008-0307-y teresa, h. (2007). wind turbines: designing with maintenance in mind. power engineering, 111(5): 36(3). yuniarto, m. n., labib, a. w. (2006). fuzzy adaptive preventive maintenance in a manufacturing control system: a step towards selfmaintenance. international journal of production research, 44(1): 159-180. doi:10.1080/13528160500245723 zadeh, l. a. (1975). the concept of a linguistic variable and its application to approximate reasoning (i, ii, iii). information sciences: i, 8(3): 199-249; ii,8(4): 301-357; iii, 9(1): 43-80. doi:10.1016/0020-0255(75)90017-1 109int. j. prod. manag. eng. (2015) 3(2), 103-109creative commons attribution-noncommercial-noderivatives 4.0 international fuzzy maintenance costs of a wind turbine pitch control device http://dx.doi.org/10.1007/978-3-7908-1799-7 http://dx.doi.org/10.1007/978-3-7908-1799-7 http://dx.doi.org/10.1016/j.ins.2005.06.008 http://dx.doi.org/10.1016/j.cie.2007.07.002 http://dx.doi.org/10.1002/etep.535 http://dx.doi.org/10.1109/tec.2006.889623 http://dx.doi.org/10.1002/we.513 http://dx.doi.org/10.1080/00207720701717708 http://dx.doi.org/10.1007/s00500-008-0307-y http://dx.doi.org/10.1007/s00500-008-0307-y http://dx.doi.org/10.1080/13528160500245723 http://dx.doi.org/10.1016/0020-0255(75)90017-1 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2016.4106 received 2015-09-04 accepted: 2015-12-17 sociological diagnostics in staff competency assessments: evidence from russian museums asiia usiaevaa,i, mariia rubtcovaa,ii*, irina pavlenkovaa,i and svetlana petropavlovskayab a department of social management and planning, faculty of sociology, saint petersburg state university, 7-9, universitetskaya nab., saint petersburg, 199034, russian federation a,i dkpi@mail.ru a,ii m.rubtsova@spbu.ru b training programs department laboratory of pedagogical cultural studies at spb gbuk gmp “st. isaac’s cathedral”, 4, st isaac’s square, st. petersburg, 190000 russian federation abstract: this paper presents the sociological diagnostics analysis that aims to the assessment of staff’s competencies in the russian museum complex. the current research was carried out in 2014 in russia (st. petersburg). the methodology is a structured observation and “mystery shopper” method. monitoring of staff competencies was conducted according to the competence model that specified criteria and indicators of observation in assessment cards. results indicated the level of professional and communicative competence and revealed that the communication with museum visitors is the least developed competence. key words: sociological diagnostics, staff competency assessment, museum management. 1. introduction organization is a complex system. its parts are working together to provide the stable and effective operating. however, the most popular methods of staff competency assessment in russia are hardly applying to analyze some staff’s traits and to make a competence model. moreover, the efficiency of the company depends on skills of employees; therefore, the policy of human resource management (hrm) should be based on the results of staff competency research. the research is based on view on an organization as a system, composed of tasks, technology, structure and people (leavitt, 1965:1146). people are one of the basic elements of organization. the condition of staff, the level of its effectiveness and problem areas have a significant impact on the whole system. according to parsons (1951, 1968), there are four subsystems in each social system. these subsystems have to perform four functions to provide survival of the system: adaptation to the environment (a), goal attainment (g), integration of parts (i) and maintenance its latent pattern (l). moreover, there are parts in each subsystem that also should function in agil paradigm. each competence of people could be focused on the agil functions. staff’s skills and behaviour attitudes should provide stability to an organization. the construction of competency models allows to find out which staff’s skills are required to provide adaptation, goal attainment, integration or maintenance of latent pattern. therefore, each competence in profile should maintain stability and survival of the system. the competence-based approach is a key way to harmonize hrm policy with the strategy and the main tasks of companies. the model of competencies is widely used in various directions of hrm: hiring, training, job rotation, employee motivation, organization culture etc. (see e.g. cameron, 1999). it allows not only to identify the most skilful, professional, psychologically mature and the most competent staff, but also to predict the productivity of their work in this particular organization with a certain structure, the type of leadership, norms and organizational culture (see e.g. caldas and brandao soares de carvalho, 2008; pavenkov, 2015). this approach gives the opportunity to apply common int. j. prod. manag. eng. (2016) 4(1), 29-33creative commons attribution-noncommercial-noderivatives 4.0 international 29 http://dx.doi.org/10.4995/ijpme.2016.4106 mailto:dkpi@mail.ru mailto:m.rubtsova@spbu.ru http://creativecommons.org/licenses/by-nc-nd/4.0/ schemes and to develop unique models based on features of the environment and organizational structure in museums (kimchang, 2011; haldma and laats, 2012; dedova, 2013). sociological diagnosis as a method is described in sufficient detail (deviatko, 1993; prigogin, 2003; glazov, 2004; pavlenkova, 2014; pakhlova, 2015; rubtcova et al., 2015). usually sociological diagnostics used as a social technology that aims to the developing decisions about social issues and the elements of the organization. it consists of three stages: 1. description of the current situation; 2. assignment of the standard norm, the required state of the organization; 3. comparative analysis of the actual state of the object with the required. the studied museum complex (“the cathedral” and “the savior”) in 2010 formulated new strategic goals. they are concentrated on the development of the competitive capacity of the museums through the expansion of the scope of services, the increased flow of visitors, the improvement of image and they led to the need to amend the human resources of the museums. in 2010 the new system evaluation and subsequent development of the personnel began to operate. our research explores features and possible application of sociological diagnostics in analyzing of employees’ competencies. research question is the following: is there the accordance between requirements to the staff competency and actual competencies of museum employees? 2. data and methodology assessment of competencies in the studied museum made in accordance with the follow scheme. first of all, it was specified a competency model that set a standard (required) level of staff’s skills. secondly, using structured observation the researchers collected information about the current state of employee’s competencies. finally, deviations in staff’s behavior were analyzed. the method is a structured observation of the staff daily work in two specific museums. monitoring of staff competencies was conducted according to the competence model that specified criteria and indicators of observation in assessment cards. in the current research communicative and professional competencies of museum employees were evaluated. the study was performed in several phases. firstly, the researchers examined the literature and recent surveys of organizations, its structure, staff assessment, competencies and hrm. secondly, the museum complex was chosen as the object of study, where it would be possible to apply methods of sociological diagnostics of competencies. then it was conducted the observation of staff in two divisions of this museum in “the cathedral” and in “the savior”. as a conclusion, we analyzed the data and developed recommendations to the hr department of this museum. one of the researchers was involved in the service process as a mystery visitor. the staff daily work of whole technological chain of the museum was observed: cashiers, ticket inspectors, administrators, tour guides and sellers. this method allowed creating natural conditions of everyday service, to regard the actual skills and behavior patterns of employees and to make valid conclusions about their competencies. all employees who worked in the museums in the research dates were involved. the study was conducted according to the professional ethical code of sociologists by the russian society of sociologists. this means that according to requirements of anonymity, signed participant consent agreement cannot be asked for. the researchers informed the head of hr department about purpose of the survey and guaranteed saving anonymity and using data in a generalized form. the researchers have changed the names of museums. 3. results quality of service and competence of the employees is measured using assessment cards. as a result, 26% of discrepancies to the dress code (accepted in the code of corporate ethics) were revealed in 2010. despite the fact that employees adhere to a classical (business) official style, these deviations from the standard have been associated primarily with non-compliance with the corporate symbolic include corporate symbols and badges. the number of inconsistencies appearance standards decreased to 23% after the adoption the new code of corporate ethics and carrying out systematic monitoring of the quality of service. int. j. prod. manag. eng. (2016) 4(1), 29-33 creative commons attribution-noncommercial-noderivatives 4.0 international usiaeva, a., rubtcova, m., pavlenkova, i. and petropavlovskaya, s. 30 http://creativecommons.org/licenses/by-nc-nd/4.0/ implementation of the service quality monitoring has reduced the number of deviations, however, some problems still remain. for the more detailed study of the emerging organizational pathologies in the period from 17 february to march 4, 2014 year staff competency assessment has been made in two sites of the museum complex – the «savior» and the «cathedral». one of the researchers as a «mystery visitors» has analyzed the communicative skills, professional competence and actual behavior of cashiers, ticket inspectors, administrators, tour guides and sellers. during the observation the researcher completed 127 assessment cards: 63 in “the savior” and 64 in “the cathedral” (see table 1.) table 1. the number of the completed assessment cards. position the cathedral the savior cashiers 15 14 administrators 10 11 ticket inspectors 13 12 tour guides 13 12 sellers 13 14 total 64 63 as a result, 49 deviations of professional competence were detected. they are following: 5 ticket cashiers (10% of the total); 16 administrators (33% of the total); 5 ticket inspectors(10% of the total); 17 tour guides (35% of the total); 6 sellers (14%) (see figure 1) 10% 33% 10% 35% 12% cashiers administrators tickets inspectors tour guides sellers figure 1. proportion of deviations of museum employees, 2014. it is important to specify the observations, in which there were no deviations, and observations that recorded at least one. cashiers have the ratio 86.2% and 13.8% respectively; administrators – 57.1% and 42.9%; ticket inspectors – 84% and 16%; tour guides – 48% and 52%; sellers – 85.2% and 14.8%. as we can see, administrators and tour guides have the greatest number of deviations and the smallest gap between them. the researchers found out that in both departments of the museum complex tour guides and administrators more frequently deviated standards than other museum employees (see figure 1). the greatest number of administrators` deviations linked to their absence from the workplace, while their job responsibilities include the work with visitors that pass through the turnstiles. in addition, they should regulate the flow of visitors and help them in a case of difficulties. tour guides have deviations connected with refusal the methodology and the plan of excursion (shortening trips, ignoring the key stop, etc.). cashiers show their professional incompetence in 10% of cases. despite the correctness of the execution of monetary transactions policy during the ticket sales, they do not clarify the ticket types and possibilities of discount. professional deviations of ticket inspectors dealt with the provision of information on travel services (10.0%). finally, group, demonstrating a completely professionally competent, are sellers: there was no case of deviation from instructions or standards during the observation period. then there was the assessment of communicative competence of the personnel of two museums. as a result, it was found that administrators more frequently deviated communicative standards than other museum employees did (75% deviation of communications standards). a major disruption in communication is the following: 67% cases-non-final communication phrases; 25% – no greeting visitors; 8% – absence of polite forms during interactions (authoritative speech, without the polite words “thank you”, “please”, “сan i be of any assistance”, “can i help you”, etc.) however, during our observation, we found that when there is a large flow of visitors the museum staff could not perform the communication acts according to corporative speech standards. corporative speech standards constitute a common set of rules for all divisions of the museums and the prescribed phrase to complete the interaction can be considered as “illogical” for some divisions. the smallest number of communicative deviations is fixed with guides – 25%. during the diagnosis, only two violations were observed. communicative competence of tour guides is most developed. if we take into account that these employees, as a rule, have the longest contact with visitors, such indicators are a good achievement for the effective functioning of museums. int. j. prod. manag. eng. (2016) 4(1), 29-33creative commons attribution-noncommercial-noderivatives 4.0 international sociological diagnostics in staff competency assessments: evidence from russian museums 31 http://creativecommons.org/licenses/by-nc-nd/4.0/ however, tour guides have 56.3% deviation of dress code, as a rule they don't use the corporate symbols, namely uniform scarf. such deviations from the prescribed standards were reported in almost every third card. visual self-presentation employee, his/her association with the museum is important to create a certain visitor's impression. the level of communicative deviations in 20102014 decreased. in 2010 the biggest problem was communication with visitors – 52% of deviations (see figure 2). in 2014 the number of deviations reduced to 13,5%. however deviations in professional behavior decreased only from 17% in 2014 to 15, 5% in 2014. 52,0% 17,0% 13,5% 15,5% 0,0% 10,0% 20,0% 30,0% 40,0% 50,0% 60,0% 70,0% 80,0% 90,0% 100,0% communicative competence professional competence 2010 2014 figure 2. amount of communicative and professional deviations, 2010-2014. despite the fact that communicative competencies were developed, they still remained as a weakness of this organization (see figure 3). 80,0% 20,0% 75,0% 25,0% 80,0% 20,0% 23,5% 24,0% 83,3% 0,0% 0,0% 10,0% 20,0% 30,0% 40,0% 50,0% 60,0% 70,0% 80,0% 90,0% communicative competence professional competence cashiers administrators ticket inspectors tour guides sellers figure 3. communicative and professional deviations of museum staff, 2014. the most common deviations were lack of politeness and noncompliance of dress-code. however, it should be said that the analysis revealed that speech standards were not adapted to the specifics of the ticket inspectors and administrators that is why there were registered a lot of false deviations and poor development of communicative competence. it was also discovered an incomplete awareness among employees about the evaluation criteria, which reduced their loyalty to the organization and confidence in the monitoring. the criteria for assessment of professional competence were job descriptions, so the deviations were related to the fact that employees do not perform their duties (for example, the administrators were absent, guides reduced the duration of the tour etc.). 4. discussion and conclusion to sum up it should be said that relevant training system and regular staff assessment (diagnostics of employee’s competencies) in the state museums increased competitiveness of this organization and improved staff competence (pavlenkova & petropavlovskaya, 2013). nevertheless, the researcher made the following recommendations to this museum that were based on the survey. it is necessary to upgrade speech standards, according to specifics of service in each stage of technological chain. employees should enlighten about the criteria of competence assessment. organizational identity and commitment to the corporate culture should be developed in order to ensure that staff is loyal to the museum and strives to improve their business efficiency. instructions and training of professional duties should be made to reduce the deviations in this field. in conclusion, it should be emphasized that our results are consistent with ideas expressed by v. sherbina (1993, 2004). according to his opinion, sociological diagnostics is a very effective way to analyze and assessment of problems in the organization, therefore it should be used to solve practical problems and in order to make management decisions. acknowledgements the authors would like to thank the anonymous reviewers for their insightful and helpful comments on our work. declaration of conflicting interests the authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. funding the authors received no financial support for the research, authorship, and/or publication of this article. int. j. prod. manag. eng. (2016) 4(1), 29-33 creative commons attribution-noncommercial-noderivatives 4.0 international usiaeva, a., rubtcova, m., pavlenkova, i. and petropavlovskaya, s. 32 http://creativecommons.org/licenses/by-nc-nd/4.0/ references caldas, r. f. and brandao soares de carvalho, j. a. (2008). approach for the human resources management analysis in libraries and museums: case study of knowledge cities. k. s. soliman, ed. cameron, k. (1999). diagnosing and changing organizational culture. upper saddle river, nj: prentice hall series in organizational development. dedova, m. (2013). development of public cultural services management: study of the night of museums. proceedings of the 17th international conference current trends in public sector research, 21-28. deviatko, i. (1993). the diagnostic procedure in sociology. essay on history and theory. moscow: nauka. glazov, m. (2004). functional diagnostics of industrial company. st. petersburg: rggmu. haldma, t. and laats, k. (2012). the balanced scorecard as a performance management tool for museums. best practices in management accounting, 232-252. http://dx.doi.org/10.1057/9780230361553.0028 kimchang b, (2011). management innovation of non-profit organizations an empirical study on factors affecting museum performance in korean museums. the e-business studies, 12(3): 461-485. http://dx.doi.org/10.15719/geba.12.3.201109.461 leavitt, h. (1965). applied organizational change in industry: structural, technological and humanistic approaches, in: handbook of organizations /edited by j.g. march. chicago: rand mcnally. pakhlova, i. (2015). competence model as a tool of hr management in organization. moscow: gramota parsons, т. (1951). the social system. london: routledge & kegan paul. parsons, t. (1968). the structure of social action: a study in social theory with special reference to a group of recent european writers. ny.: the free press pavenkov, o., pavenkov, v. and rubtcova, m.v. (2015). the influence of social norms for creativity in an organization of the media industry: theoretical framework. international scientific-practical conference “actual problems of development of the media industry at the present stage”, st. petersburg. st. petersburg state university of cinema and television. 28-29 november 2014. http://dx.doi. org/10.2139/ssrn.2573754 pavlenkova, i. (2014).the management of organizational conflicts in the museum field: social and communicative aspects. unpublished phd thesis. st.petersburg: spbu pavlenkova, i. and petropavlovskaya, s. (2013). social and cultural aspects of the service of the northern capital of museum services (the case of the state museum «the cathedral»). st. petersburg: the state museum «the cathedral» prigogin, a. (2003). methods of development organizations. moscow: mcfer. rubtcova, m., pavenkov, o., pavenkov, v., martianova, n. and martyanov, d. (2015). deprofessionalisation as a performance management dysfunction: the case of inclusive education teachers in russia. asian social science, 11(18): 339-349. http://dx.doi.org/10.5539/ ass.v11n18p339 sherbina, v. (1993). methods of sociological diagnostics in management. moscow: socis sherbina, v. (2004). human resource management. moscow: msu int. j. prod. manag. eng. (2016) 4(1), 29-33creative commons attribution-noncommercial-noderivatives 4.0 international sociological diagnostics in staff competency assessments: evidence from russian museums 33 http://dx.doi.org/10.1057/9780230361553.0028 http://dx.doi.org/10.15719/geba.12.3.201109.461 http://dx.doi.org/10.2139/ssrn.2573754 http://dx.doi.org/10.2139/ssrn.2573754 http://dx.doi.org/10.5539/ass.v11n18p339 http://dx.doi.org/10.5539/ass.v11n18p339 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2015.3245 received 2014-08-26 accepted: 2015-05-25 total quality management implementation in greek businesses: comparative assessment 2009-2013 martha vranakia,i, stergios vranakisa,ii and lazaros sarigiannidisb a democritus university of thrace, school of engineering, department of production and management engineering. xanthi, greece a,i martha@vranakis.gr a,i stergios@vranakis.gr b technological educational institute of kavala, school of business and economics, department of business administration. kavala, greece b lsarigia@pme.duth.gr abstract: the competition in the greek manufacturing sector has become very intense and the need for businesses to survive, under these very difficult conditions, forces them to find new ways to increase their profits, but also to attract new customers and to retain old. a necessary condition for long-term business survival is to maintain a high product quality level. the implementation of total quality management (tqm) approach is a key factor to achieve this goal.the main objective of this research is to identify the current situation as far as the implementation of tqm by greek manufacturing firms, and finally to compare the results between the current research and the previous research of 2009 (vranaki et al., 2010). the research model that has been developed incorporates nine factors which are found in literature to influence total quality management. a structured questionnaire has been developed and distributed to executives of 61 companies. descriptive statistics as well as structure equation modeling (sem) techniques were used to analyze the data. key words: quality, total quality management, business performance, management leadership, supplier management, customer focus. 1. introduction in the first decades after the second world war, the competitiveness of products and services in international trade was defined by two related features, quality and production cost. a more recent important dimension of competitiveness is the ability to develop innovations in products and production processes. the ability to develop innovations frequently combined with quality and productivity, determine the time our chances to survive a business in a complex and uncertain environment in a context of rapid globalization, technological developments. moreover, development in recent decades led many companies to consider the quality as the basic and most effective condition for success. this explains the ease of penetration in foreign countries many products in japan and germany, software packages and various technical and financial services in the u.s., and their example followed by south korea, singapore, taiwan etc. . what ultimately establish and differentiate the products of the countries is the high quality that offer the purchaser in relation to their cost, in other words a great value compared to cost to the customer. 2. literature review in recent years, increasing attention has been paid to improving the overall quality. many companies have taken initiatives to implement various techniques of quality management. an important strategy for achieving high quality is tqm (total quality management). the total quality management (tqm) was defined as a management system to improve efficiency within a business to maximize customer satisfaction, conduct continuous improvements and great support to the involvement of employees. 87int. j. prod. manag. eng. (2015) 3(2), 87-95creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2015.3245 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2.1. factors that affect operational results improving business results through quality, assumes that many factors are quite important for enterprises. a comprehensive review and a classification of the relevant empirical literature revealed that, in general, the nine factors discussed below were the most important parameters in the application of tqm. 2.2. leadership as documented by various researchers (deming, 1986; juran, 1986), the administrative leadership is an important factor in the implementation of tqm because it improves performance by influencing other practices of tqm (anderson et al., 1995; flynn et al., 1995; ahire and o’shaughnessy, 1998; wilson and collier, 2000). the successful implementation of tqm requires effective change in the culture of a company. it is almost impossible to have changes in a company without any a concentrated effort by the administration, which aims at a continuous improvement in an open communication and cooperation throughout the enterprise (bell and burnham, 1989; ettkin et al., 1990; goodstein and burke, 1991; handfield and ghosh, 1994; choi, 1995; hamlin et al., 1997; zeitz et al., 1997; daft, 1998; abraham et al., 1999; adebanjo and kehoe, 1999; ho et al., 1999). 2.3. human resource management the administration has a complex role in the implementation of tqm. it is impossible to improve the procedures of any business without a well-trained workforce. the management of human resources, previously known as personnel management, has been upgraded to the science that studies the staff not as a factor that causes costs, but as an asset in which each company must invest. the administration should provide the necessary resources for the training of staff in the use of new concepts and tools and creates a work environment that encourages employee participation in the process of change (bell and burnham, 1989; schroeder et al., 1989; burack et al., 1994; anderson et al., 1995; flynn et al., 1995; hamlin et al., 1997; ahire and o’shaughnessy, 1998; daft, 1998; handfield et al., 1998; ho et al., 1999; wilson and collier, 2000). top management should also ensure that the necessary resources for the relevant quality training is available (ahire and o’shaughnessy, 1998; anderson et al., 1995; flynn et al., 1995; handfield et al., 1998; ho et al., 1999). it takes more than education to be effective and successful change. employees should be involved at this stage. a crucial factor in accordance with the adebanjo and kehoe (1999), is that the participation of workers, because affected by the creation of a new working environment that encourages and facilitates open communication. in such an environment, it is possible for workers to commit themselves to work and contribute their ideas in that it facilitates and enhances the process of change (burack et al., 1994; anderson et al., 1995; flynn et al., 1995; das et al., 2000). 2.4. information and data analysis the information and analysis of data related to quality, including the unnecessary actions of a “poor” quality, such as repetitive labor costs, waste and control charts to identify quality problems and provide information on the areas of potential improvement (choi, 1995; lockamy, 1998; ho et al., 1999). the data relating to quality have a positive effect on firm performance through three business practices of tqm. specifically, through the quality management of suppliers, to design new products / services and through management processes. 2.5. supplier management since all businesses (especially large) have their suppliers from whom they buy either materials or products, the quality that they provide them is able to affect the overall quality of the finished products. so the complete identification of products needed by their suppliers a company is a hub avoiding production of defective products and, therefore, increase business performance. the quality management of suppliers requires regular monitoring of suppliers by creating a database that measures this performance, a critical tool for improving material and raw materials costs required to develop, market prices and responsiveness of suppliers (krause et al., 1998). with this database, companies can pursue qualitative measures such defective parts-per-million (parts-per-million defective), the reliability and the rate of discarded products (forza and flippini, 1998; krause et al., 1998; trent and monczka, 1999), as well as timely delivery and performance in the percentage of acceptable incoming materials (tan et al., 1998). 2.6. product design each product has specific characteristics. for the design of the process or production processes, products are categorized into groups depending on 88 int. j. prod. manag. eng. (2015) 3(1), 87-95 creative commons attribution-noncommercial-noderivatives 4.0 international vranaki, m., vranakis, s. and sarigiannidis, l. http://creativecommons.org/licenses/by-nc-nd/4.0/ their type, their production volume, complexity, and on the basis of characteristics of the contact point with these firm-customer. regarding the model can distinguish the specific products that can be manufactured in many different styles, standard products and products of mass production (adamidis, 2002). under the igi, efforts design new products have two objectives: planning the construction part of the products, and the design quality of the products (flynn et al., 1995; handfield et al., 1999). top management of a company is responsible for the design of products for the market and meet consumer needs (deming, 1986; garvin, 1987; shetty, 1988; flynn et al., 1995). this focus is critical for the development of products, especially when they meet customer needs (juran, 1981; leonard and sasser, 1982; flynn et al., 1995; hackman and wageman, 1995). to simplify the design of products, top management uses interoperable groups to reduce the number of parts that make up the product and standardize these parts (chase et al., 2001). by doing so achieves a more efficient management processes by reducing process complexity and differences between the procedures (flynn et al., 1995; ahire and dreyfus, 2000). 2.7. process management another factor that affects the operational results through management procedures. management processes in an enterprise implies a proactive method to improve the quality, such as the design of processes that provide stable production schedules and distribution work (saraph et al., 1989; flynn et al., 1995) to reduce the complexity of processes (flynn et al., 1995) with the build quality of the product during the production phase (handfield et al., 1999). reducing the complexity of the process increases the uniformity of production, while reducing duplication and defective (anderson et al., 1994; forza and flippini, 1998) because the quality problems are identified and corrected immediately (ahire and dreyfus, 2000). the process used to produce a product directly affects the quality. the market, for example, a machine that will facilitate the production and thus improve the quality of a solution where the money will be invested in the market will be amortized from the best production, the easiest and best price sale. 2.8. customer focus one element of tqm is the focus on customers. the establishment and maintenance of an open relationship between the firm and its clients facilitate the design of new products. this is achieved because there is immediate clarification of needs and wants of customers. the key to nurturing strong relationships with customers is to establish communication between the firm and its clients (tillery, 1985). these practices include frequent contact with customers. the wright and snell (2002) argue that simply focus and customer acquisition is not always good for business. since customers can easily be lost in case they have a bad experience with the product or even if a new product does not attract them. businesses should target customer trust to have improved operational results. 2.9. strategic planning strategic planning is the process of development and analysis of the mission and the vision, objectives, strategies and defining the sources of business. strategic planning has a long time horizon, considering the external environment and determines the general direction of the business. this programming will be made by the highest levels of administration (jackson and ferguson, 1952). 3. proposed conceptual framework and research hypotheses through this research aims to study o role of tqm in greek businesses, and the comparison to the applications of the principles of tqm in the years 2009 and 2013. research model is a synthesis of research findings from the literature. the opinions are varied and numerous, so an attempt was made to include as much as the model to be an integrated presentation that takes into account all factors affecting the igi. the 9 factors of tqm presented will serve as part of the model. all are interrelated and the proper functioning of one affects the proper functioning of the other. all are considered particularly important for an enterprise to improve its results, should take them seriously. even the improvement of some of these factors will lead to greater earnings. since the model will create some initial assumptions that depending on the findings of the investigation or will be verified or disproved. the research model is framed by an external agent is the economic crisis. the processing in our country now is now at breaking point. the economic crisis and the number of bureaucratic barriers that are in any healthy business initiative have created uncertainty and insecurity in the market. 89int. j. prod. manag. eng. (2015) 3(2), 87-95creative commons attribution-noncommercial-noderivatives 4.0 international total quality management implementation in greek businesses: comparative assessment 2009-2013 http://creativecommons.org/licenses/by-nc-nd/4.0/ meanwhile, the business of the country, especially manufacturing companies based in the greek region, seeking development measures that the state promised one part, and the other was obliged to take to remove disincentives to entrepreneurship and the establishment of structural economic reforms. given the negative sentiment in the market, it is obvious that a new approach to the development effort, especially in the field of policies to improve the external business environment in greece. as shown in figure 1 above, created the following assumptions: hypothesis 1: the administrative leadership positively affects: a) strategic planning, b) customer focus, c) the information and data analysis, d) human resource management, e) management procedures, f) and supplier management. hypothesis 2: the strategic planning positively affects: a) customer focus, and b) operational results. hypothesis 3: the focus of customer positively affects business results. hypothesis 4: the information and data analysis positively affects: a) strategic planning, b) customer focus, c) the design of products, d) human resource management, e) management procedures, f) managing suppliers. hypothesis 5: the management of human resources positively affects: a) the management of suppliers, b) customer focus, and c) operational results. hypothesis 6: the process management positively affects business results. hypothesis 7: managing suppliers is positive: a) designing products, and b) operational results. figure 1. proposed conceptual framework. figure  1:  proposed  conceptual  framework       human  resource   management strategic  planning customer  focus leadership supplier   management   product   design business  results process   management data  analysis h1a h1b h1c h1d h1e h1f h2a h2b h3 h4a h4b h4c h4d h4e h4f h5a h5b h5c h6 h8 h7a h7b 90 int. j. prod. manag. eng. (2015) 3(1), 87-95 creative commons attribution-noncommercial-noderivatives 4.0 international vranaki, m., vranakis, s. and sarigiannidis, l. http://creativecommons.org/licenses/by-nc-nd/4.0/ hypothesis 8: the product design positively affects management procedures. 4. research methodology field survey of the research are greek companies that belong to the manufacturing sector of greek economy, and employ more than 20 employees. the final sample consisted of 61 correctly completed questionnaires from the secondary sector. 67 of the 95 companies responded, returning 61 completed questionnaires. however, questionnaires six of them, were deemed unsuitable because responses completed poorly. therefore 61 questionnaires (from 67 firms) were assessed as suitable for statistical analysis with a response rate of approximately 64.21% of the total population (95). 5. exploratory factor analysis one measure of sample adequacy is the ratio of the kaiser-meyer-olkin (kmo), and must take values greater than 0.5 (malhotra, 1999). in this study, the kmo values are satisfactory and acceptable. an additional check of the correlations of our data is testing sphericity of bartlett (1950). note that variables removed from the tables because of low loadings (see appendix). the results of the checks carried out, allow to assert, that the deterministic variables are compact and reliable structures, able to contribute to the measurement of the agent to which they belong. to assess the goodness of fit of deterministic variables applied confirmatory factor analysis. initially, took control of the overall model, and then testing the structural model. in the model below, the encodings are as follows: a. leadership, b. strategic planning, c. customer focus, d. information & data analysis, e. human resource management, f. process management, g. supplier management, h. product design, i. business results. the overall model was estimated using four indicators. acceptable values of the indicators are: cmin/df<3, gfi>0.9, cfi>0.9, rmr<0.05 (smith & mcmillan, 2001). the levels of these markers suitability is acceptable, so the model is valid. in summary, it should be noted that at first glance, observed that the main core of the model is the administrative leadership and information and analysis. the first factor directly influences the figure 2. fitness model. cmin/df cfi gfi rmr 1,946 0,878 0,847 0,041 figure  2:  fitness  model       cmin/df   cfi   gfi   rmr   1,946   ,878   ,847   ,041       table  1:  results  of  hypothesis  testing   hypotheses investigated relationships regression result 1a α  b 0.31*** accepted 1b α  c rejected 1c α  d 0.60*** accept 1d α  e 0.21*** accept 1e α  f rejection 1f α  g rejection 2a β  c rejection 2b β  i rejection 3 c  i rejection 4a d  b 0.19*** accept 4b d  c 0.24*** accept 4c d  h rejection 4d d  e 0.19*** accept 4e d  f 0.25*** accept 4g d  g rejection 5a e  g rejection 5b e  c 0.25*** accept 5c e  i rejection 6 f  i rejection 7a g  h 0.46*** accept 7b h  i 0.31*** accept 8 i  f rejection ***p<0.001 level, **p<0.05 level   ,00 a ,15 b ,29 c ,36 d ,21 e ,40 f g ,42 h ,58 i e1 e2 e3 e4 e5 e6 e7 e8 e9 ,25 ,60 ,24 ,19 ,31 ,25 ,31 ,31 ,46 ,44 ,46 ,70   ,23   ,19 ,00 91int. j. prod. manag. eng. (2015) 3(2), 87-95creative commons attribution-noncommercial-noderivatives 4.0 international total quality management implementation in greek businesses: comparative assessment 2009-2013 http://creativecommons.org/licenses/by-nc-nd/4.0/ strategic planning, the factor relating to information and data analysis and management of human resources, which verifies three of the initial assumptions. while indirectly affects the other actors and the operational results. the second factor is that the core is the information and data analysis, which directly-affected in human resource management, the operational results and customer focus. 6. conclusions the aim of this study was to analyze the factors affecting the igi operating results, the impact of the economic crisis and to compare the results of this research with the corresponding 2009. comparing the results of this research with the research conducted in 2009 (vranaki et al., 2010) resulted in the following conclusions: 1. to improve operational results, emphasis should be placed on all factors of tqm. 2. focusing on customers is a key objective of greek firms. 3. changes in customer preferences significantly affect the management of suppliers. 4. factor information is a “station” of administrative leadership. the first and very impressive conclusion drawn from this research are the indirect effects that accept business results, which verified in earlier research (vranaki et al., 2010). it was expected that these factors will directly affect business performance to some extent. the significance of this finding is the indirect influence of these factors on business outcomes. the interpretation of the above can be a very useful tool in the hands of greek firms. more specifically, from the above we understand that companies need to pay attention to many parameters to achieve their purpose. it is not enough to be consumed in a particular agent and others to fail. the focus of the customer no effect. unlike the earlier survey where the customer focus impacted upon four factors, and this in turn is impacted upon the management of suppliers. thus, we conclude that the customer satisfaction and knowledge on the requirements of customers is the second most important goal you want to achieve the greek companies, but also that most businesses do not make changes in supplier management with the slightest change in customer needs. the administrative leadership does not act directly to target customers, but indirectly through other factors, in contrast to the 2009 survey. administrations business to achieve its approach and establishment of good relations with clients through the collection of information, training workers but also through product design. at this point it should be noted that research verifies the wright and snell (2002), the who argue that simply focus and customer acquisition is not always good for business. table 1. results of hypothesis testing. hypotheses investigated relationships regression result 1a α → b 0.31*** accepted 1b α → c rejected 1c α → d 0.60*** accept 1d α → e 0.21*** accept 1e α → f rejection 1f α → g rejection 2a β → c rejection 2b β → i rejection 3 c → i rejection 4a d → b 0.19*** accept 4b d → c 0.24*** accept 4c d → h rejection 4d d → e 0.19*** accept 4e d → f 0.25*** accept 4g d → g rejection 5a e → g rejection 5b e → c 0.25*** accept 5c e → i rejection 6 f → i rejection 7a g → h 0.46*** accept 7b h → i 0.31*** accept 8 i → f rejection ***p<0.001 level, **p<0.05 level 92 int. j. prod. manag. eng. (2015) 3(1), 87-95 creative commons attribution-noncommercial-noderivatives 4.0 international vranaki, m., vranakis, s. and sarigiannidis, l. http://creativecommons.org/licenses/by-nc-nd/4.0/ since customers can easily be lost in case they have a bad experience with the product or even if a new product does not attract them. the greek companies surveyed are showing great interest in customer retention. besides, studies have shown that attracting new customers is much more expensive strategy than keeping existing customers (kotler, 1982). the greek firms, given the large and increasing competition are trying to focus on customer satisfaction rather than on improving operational results, but for obtaining “good reputation.” the next conclusion we reached was that the administrative leadership through the management of human resources affects product design product design. in the 2009 survey design products in turn impacted upon on business outcomes. however, it is very encouraging that most companies place great emphasis on training their employees. course, must be included with the necessary resources for the training of staff in total business expenses. on the other hand, when a company has fully trained staff on quality issues as avoiding possible mistakes and defective products and therefore achieves customer focus. it should be noted that the management of human resources including the health and safety of workers. as we can conclude, businesses protecting employees from any accidents aimed at improving their emoluments as well as to improve the image of the company. furthermore, observed that the level of training of governing and management procedures, which is repeated from 2009. process management involves reducing the complexity of the processes in the production stage. officials, however, the companies to be able to respond to change a process must first have the proper training. in any other case, the “change” in business processes will have no positive benefit to business results. vendors directly affect the design of new products. any change in production processes or customer habits involves the review of suppliers. as mentioned above, the quality of raw materials of products is the basis for good quality of finished products. finally, reference should be made to study the economic crisis as an external factor. the economic crisis, according to the frequency analysis, seems to have the most negative effects on human resource management and management of suppliers. this was expected, considering the increase in the unemployment rate in the country the last two years, but also the need for companies to increasingly seeking “best prices” for their raw materials. 6.1. research limitations observing the results of the investigation, it is useful also to refer to some restrictions. the survey was conducted with a sample of 61 greek and craft industries in the manufacturing sector. a larger sample would likely give different results. all companies operate in manufacturing sector, but 45 of the 61 belong to the food industry, so they subject to each agent from a different perspective, than if it were operating in different manufacturing activity. questions contain elements of subjectivity. thus, some of the respondents may be overestimated to a question by scoring 1 in likert scale that can be “worth” 2 or underestimated some grading at 7 in the likert scale that can ‘deserved’ 6. references abraham, m., crawford, j., fisher, t. (1999). key factors predicting effectiveness of cultural change and improved productivity in implementing total quality management, international journal of quality and reliability management, 16(2): 112–132. doi:10.1108/02656719910239910 adebanjo, d., kehoe, d. (1999). an investigation of quality culturedevelopment in uk industry, international journal of operations and production management, 19(7): 633–649. doi:10.1108/01443579910271656 ahire, s.l., o’shaughnessy, k.c. (1998). the role of top management commitment in quality management: an empirical analysis of the auto parts industry, international journal of quality science, 3(1): 5–37. doi:10.1108/13598539810196868 ahire, s.l., dreyfus, p. (2000). the impact of design management and process management on quality: an empirical examination, journal of operations management, 18(5): 549–575. doi:10.1016/s0272-6963(00)00029-2 anderson, j.c., rungtusanatham, m., schroeder, r.g. (1994). a theory of quality management underlying the deming management method, academy of management review, 19(3): 472-509. anderson, j.c., rungtusanatham, m., schroeder, r.g., devaraj, s. (1995). a path analytic model of a theory of quality management underlying the deming management method: preliminary empirical findings, decision sciences, 26(5): 637–658. doi:10.1111/j.1540-5915.1995. tb01444.x bartlett, g. r. (1950). american journal of physiology, 163: 614. 93int. j. prod. manag. eng. (2015) 3(2), 87-95creative commons attribution-noncommercial-noderivatives 4.0 international total quality management implementation in greek businesses: comparative assessment 2009-2013 http://dx.doi.org/10.1108/02656719910239910 http://dx.doi.org/10.1108/01443579910271656 http://dx.doi.org/10.1108/13598539810196868 http://dx.doi.org/10.1016/s0272-6963(00)00029-2 http://dx.doi.org/10.1111/j.1540-5915.1995.tb01444.x http://dx.doi.org/10.1111/j.1540-5915.1995.tb01444.x http://creativecommons.org/licenses/by-nc-nd/4.0/ bell, r.r., burnham, j.m. (1989). the paradox of manufacturing productivity and innovation, business horizons, 32(5): 58–64. doi:10.1016/0007-6813(89)90084-0 burack, e.h., burack, m.d., miller, d.m., morgan, k. (1994). new paradigm approaches in strategic human resource management, group and organization management, 19(2): 141–159. doi:10.1177/1059601194192002 burt, d.n. (1989). managing suppliers up to speed, harvard business review, 67(4): 127–135. chase, r.b., aquilano, n.j., jacobs, f.r. (2001). operations management for competitive advantage, 9th ed. mcgraw-hill, boston, ma. choi, t.y. (1995). conceptualizing continuous improvement: implications for organizational change, omega, 23(6): 607–624. doi:10.1016/0305-0483(95)00041-0 cooper, m.c., ellram, l.m. (1993). characteristics of supply chain management and the implications for purchasing and logistics strategy, the international journal of logistics management, 4(2): 13–24. doi:10.1108/09574099310804957 daft, r.l. (1998). organization theory and design. south-western college publishing, cincinnati, oh. das, a., handfield, r.b., calantone, r.j., ghosh, s. (2000). a contingent view of quality management—the impact of international competition on quality, decision sciences, 31(3): 649–690. doi:10.1111/j.1540-5915.2000.tb00938.x deming, w.e. (1986). out of the crisis. mit press, cambridge, ma. ellram, l.m. (1991). a managerial guideline for the development and implementation of purchasing partnerships, international journal of purchasing and materials management, 27(3): 2–8. ettkin, l.p., helms, m.m., haynes, p.j. (1990). people: a critical element of new technology implementation, industrial management, 32(5): 27–29. evans, j., lindsay, w. (2008). the management and control of quality, 7th edition. usa: thomson south western. flynn, b.b., schroeder, r.g., sakakibara, s. (1995). the impact of quality management practices on performance and competitive advantage, decision sciences, 26(5): 659–691. doi:10.1111/j.1540-5915.1995.tb01445.x forza, c., filippini, r. (1998). tqm impact on quality conformance and customer satisfaction: a causal model, international journal of production economics, 55(1): 1-20. doi:10.1016/s0925-5273(98)00007-3 garvin, d.a. (1987). competing on the eight dimensions of quality, harvard business review, 65(6): 101–109. goodstein, l., burke, w. (1991). creating successful organization change, organizational dynamics, 19(4): 5–17. doi:10.1016/00902616(91)90050-j hackman, j.r., wageman, r. (1995). total quality management: empirical, conceptual, and practical issues, administrative science quarterly, 40(2): 309–342. doi:10.2307/2393640 hamlin, b., reidy, m., stewart, j. (1997). changing the management culture in one part of the british civil service through visionary leadership and strategically led researchbased od interventions, journal of applied management studies, 6(2): 233–251. handfield, r., ghosh, s. (1994). creating a quality culture through organizational change: a case analysis, journal of international marketing, 2(3): 7–36. handfield, r., ghosh, s., fawcett, s. (1998). quality-driven change and its effects on financial performance, quality management journal, 5(3): 13–30. handfield r.b., jayaram j., ghosh s. (1999). an empirical examination of quality tool deployment patterns and their impact on performance, international journal of production research, 37(6): 1403-1426. doi:10.1080/002075499191328 ho, d.c.k., duffy, v.g., shih, h.m. (1999). an empirical analysis of effective tqm implementation in the hong kong electronics manufacturing industry, human factors and ergonomics in manufacturing, 9(1): 1–25. doi:10.1002/(sici)1520-6564(199924)9:1<1::aidhfm1>3.0.co;2-l jackson, r.w.b., ferguson, g.a. (1952). manual of education statistics, university of toronto, department of educational research, toronto. juran, j.m. (1981). product quality – a prescription for the west, part i. management review, 70(6): 8–14. juran, j.m. (1988). on planning for quality. collier macmillan, london. kotler, p. (1982). marketing for nonprofit organization, 2nd edition. prentice-hall, englewood cliffs, nj. krause, d.r., handfield, r.b., scannell, t.v. (1998). an empirical investigation of supplier development: reactive and strategic processes, journal of operations management, 17(1): 39–58. doi:10.1016/s0272-6963(98)00030-8 leonard, f.s., sasser, w.e. (1982). the incline of quality. harvard business review, 60(5): 163–171. lockamy iii, a. (1998). quality-focused performance measurement systems: a normative model, international journal of operations and production management, 18(8):740–766. doi:10.1108/01443579810217440 malhotra, n.k. (1999). marketing research. an applied orientation. international edition. 3rd edition prentice hall, london. saraph, j.v., benson, g.p., schroeder, r.g. (1989). an instrument for measuring the critical factors of quality management, decision sciences, 20(4): 810–829. doi:10.1111/j.1540-5915.1989.tb01421.x schroeder, r.g., scudder, g.d., elm, d.r. (1989). innovation in manufacturing, journal of operations management, 8(1): 1–15. doi:10.1016/ s0272-6963(89)80002-6 94 int. j. prod. manag. eng. (2015) 3(1), 87-95 creative commons attribution-noncommercial-noderivatives 4.0 international vranaki, m., vranakis, s. and sarigiannidis, l. http://dx.doi.org/10.1016/0007-6813(89)90084-0 http://dx.doi.org/10.1177/1059601194192002 http://dx.doi.org/10.1016/0305-0483(95)00041-0 http://dx.doi.org/10.1108/09574099310804957 http://dx.doi.org/10.1111/j.1540-5915.2000.tb00938.x http://dx.doi.org/10.1111/j.1540-5915.1995.tb01445.x http://dx.doi.org/10.1016/s0925-5273(98)00007-3 http://dx.doi.org/10.1016/0090-2616(91)90050-j http://dx.doi.org/10.1016/0090-2616(91)90050-j http://dx.doi.org/10.2307/2393640 http://dx.doi.org/10.1080/002075499191328 http://dx.doi.org/10.1002/(sici)1520-6564(199924)9:1<1::aid-hfm1>3.0.co;2-l http://dx.doi.org/10.1002/(sici)1520-6564(199924)9:1<1::aid-hfm1>3.0.co;2-l http://dx.doi.org/10.1016/s0272-6963(98)00030-8 http://dx.doi.org/10.1108/01443579810217440 http://dx.doi.org/10.1111/j.1540-5915.1989.tb01421.x http://dx.doi.org/10.1016/s0272-6963(89)80002-6 http://dx.doi.org/10.1016/s0272-6963(89)80002-6 http://creativecommons.org/licenses/by-nc-nd/4.0/ shetty, y.k. (1988). managing product quality for profitability, sam advanced management journal, 53(4): 33–38. smith,t.d., mcmillan,b.f. (2001). a primer of model fit indices in structural equation model, annual meeting of the southwest educational research association. new orleans, la, february 1-3. tillery, k.r. (1985). an exploratory study of quality function with five manufacturing organizations. phd thesis, georgia state university. trent, r.j., monczka, r.m. (1999). achieving world-class supplier quality, total quality management, 10: 927–938. doi:10.1080/0954412997334 vranaki, m., vranakis, s., chatzoglou, p. (2010). the impact of total quality management in the greek business performance, portuguese journal of quantitative methods, 1(2): 39-51 wilson, d.d., collier, d.a. (2000). an empirical investigation of the malcolm baldrige national quality award causal model, decision sciences, 31: 361–390. doi:10.1111/j.1540-5915.2000.tb01627.x wisner, j.d., eakins, s.g. (1994). a performance assessment of the us baldrige quality award winner, international journal of quality and reliability management, 11(2): 8–25. doi:10.1108/02656719410051643 wright, p.m. and snell, s.a. (2002). research update, human resource planning, vol. 25 no. 2, pp. 45-54. zeitz, g., johannesson, r., ritchie jr., j.e. (1997). an employee survey measuring total quality management practices and culture, group and organization management, 22(4): 414–444. doi:10.1177/1059601197224002 appendix 1. check the one-dimensional nature and reliability factors variables loading kmo tve bartlett’s sig cronbach alpha leadership aa1 aa2 aa3 aa4 ab1 ab2 0,745 0,710 0,690 0,668 0,710 0,750 0,750 60,386 0,00 0,675 strategic planning b2 b3 b4 0,730 0,907 0,809 0,584 67,025 0,00 0,747 customer focus c1 c3 c4 c5 0,671 0,522 0,814 0,692 0,625 52,623 0,00 0,596 information and data analysis da1 da2 db2 db3 db4 db5 0,677 0,792 0,877 0,856 0,850 0,614 0,748 62,575 0,00 0,705 human resource management ε1 ε2 ε3 ε4 0,734 0,810 0,727 0,513 0,685 51,647 0,00 0,658 process management f1 f2 f3 0,674 0,746 0,731 0,611 51,514 0,00 0,525 supplier management g1 g3a g3b 0,702 0,650 0,673 0,801 52,070 0,00 0,674 product design h1 h2 h3 h4 0,780 0,710 0,752 0,561 0,664 50,846 0,00 0,664 business results i1 i2 i3 i4 i5 0,826 0,776 0,540 0,667 0,652 0,704 56,941 0,00 0,704 95int. j. prod. manag. eng. (2015) 3(2), 87-95creative commons attribution-noncommercial-noderivatives 4.0 international total quality management implementation in greek businesses: comparative assessment 2009-2013 http://dx.doi.org/10.1080/0954412997334 http://dx.doi.org/10.1111/j.1540-5915.2000.tb01627.x http://dx.doi.org/10.1108/02656719410051643 http://dx.doi.org/10.1177/1059601197224002 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2015.3528 received 2015-02-05 accepted: 2015-05-28 which of dea or ahp can best be employed to measure efficiency of projects? marisa a. sánchez departamento de ciencias de la administración, universidad nacional del sur campus palihue, bahía blanca, argentina mas@uns.edu.ar abstract: this paper compares analytic hierarchy process (ahp) and data envelopment analysis (dea) approaches for monitoring projects, in order to determine their performance in terms of economic, environmental and social organizational goals. this work is founded on an existing methodology to select and monitor projects based on dea, and discusses modifications and additions arising from using ahp. the proposal is applied to a real case. the results indicate that ahp constitutes an insightful approach in situations requiring a modelling of managerial preferences regarding the relative importance of organizational goals. key words: analytic hierarchy process, project management, sustainability, data envelopment analysis. 1. introduction consumers and regulators exert continuous pressure on firms to innovate in ways that will reduce their impact on the natural environment (yalabik and fairchild, 2011). porter argues that ‘for profit’ companies are well suited to solve social problems while at the same time serving their shareholder’s interest to maximize investor returns (porter and kramer, 2011). executives face the challenge of balancing sustainability related to the whole strategic priorities. this requires an effective project portfolio management that is supportive of sustainability driven strategies. in this paper we consider the problem of assessing projects so that they provide maximum value and minimize environmental and social impacts. the main challenges are evaluating projects that support different goals, some of them provide benefits that cannot be measured in monetary terms, and prioritize them together with the existing company´s portfolio. in a previous work (sánchez, 2014) an approach that integrates sustainability into project management is proposed. data envelopment analysis is used for selection and monitoring of projects. dea is widely recognized as an effective means of evaluating the relative efficiency of a group of homogeneous decision making units which produce multiple outputs by using multiple inputs. however, there are some drawbacks. managers need to assess projects considering different scenarios due to uncertainty. joro and viitala (2004) note that all inputs and outputs may not be equally relevant to the organizations analyzed and their stakeholders. and hence, it is useful to assign preferences to organizational goals or costs. however, by using dea goals are modeled as having the same preference. another issue with dea is the homogeneity assumption where all dmus are required to undertake the same processes, they should use the same inputs to produce the same outputs, and it is required that they operate within the same environment (mar, 2009). then, a very large unit is deemed efficient because there are no other units with similar production levels (madlener et al., 2006). as a consequence, dea would prevent direct efficiency comparisons between small and large projects. in particular, this research addresses the project monitoring problem through the integration of ahp into the project management framework presented in the work of sánchez (2014). the method is based 111int. j. prod. manag. eng. (2015) 3(2), 111-122creative commons attribution-noncommercial-noderivatives 4.0 international on a multi-criteria decision making technique that allows incorporating preferences among criteria. hence, the research question becomes ‘which of dea or ahp can best be employed to measure efficiency of projects and are decisions derived from dea and ahp consistent?’ in order to answer this research question we describe how to develop an ahp-based model to monitor projects. the proposal is applied to a case study and results are compared with rankings produced by dea. this work is organized in the following sections. section 2 introduces background concepts such as dea and ahp. section 3 describes the ahp model and discusses issues such as how to model costs and benefits. section 4 describes the results obtained when using the different methods. section 5 concludes with general findings on the applicability and consistency of the methodologies. 2. theoretical background 2.1. data envelopment analysis dea, first proposed by charnes (charnes et al., 1978), is a non-parametric technique used to measure the efficiency of decision making units (dmus). each dmu is seen as being engaged in a transformation process, in which, some inputs (resources) are used to try to produce some outputs (goods or services). in management contexts, dea serves as a tool for control and evaluation of past accomplishments as well as a tool to aid in planning future activities (banker et al., 1984; cook and seiford, 2009). dea models return an efficient projection point of operation on the frontier for each inefficient dmu, thus identifying the dmus that can be used as performance benchmarks for the dmus that are operating inefficiently. there are some limitations regarding the application of dea as described in section 1. another issue of using dea as a benchmarking tool is that it may provide inappropriate benchmark dmus for inefficient dmus when the inputs (or outputs) are derived from two distinct objectives (camanho and dyson, 1999; shimshak et al., 2009; chang and yang, 2010). for instance, when quality outputs and operating outputs are directly mixed together for executing dea. dmus with low quality outputs (or low operating outputs) can be recognized as benchmark dmus. this happens because dea is linear-programming-based technique for evaluating efficiency of each dmu by selecting the most beneficial weights for inputs and outputs; once the outputs are sufficiently high, the low outputs of another one may be ignored due to zero weights. 2.2. analytic hierarchy process the ahp allows decision makers to model a complex problem in a hierarchical structure showing the relationships of the goal, objectives (criteria), sub-objectives, and alternatives (saaty, 1997). after arranging the problem in a hierarchical fashion, the decision-maker makes subjective assessments with respect to the relative importance of each of the criteria, and indicates the preference of each alternative with respect to each of the criteria. comparison matrices are used for pairwise comparisons between the sub-criteria with respect to its parent node, and each pair of alternatives with respect to each sub-criterion. these comparisons may be taken from actual measurements or from a scale that reflects the relative strength of preference, relevance or probability. given n criteria and m alternatives, n matrices of order m×m and order n×n should be built, which makes that ahp is a non-scalable method. once judgments have been entered for each part of the model, the information is synthesized to rank the alternatives in relation to the overall goal. 2.3. ahp and project management kumar (2004) developed a model based on ahp for project selection. the criteria are structured using a pre-defined list of organizational, technical, strategic and financial factors. pairwise comparisons of factors reflect the importance of each of them. candidate projects are evaluated using a grading scale with five elements. in kendrick and saaty (2007) the authors propose a four-step process to define a portfolio. a hierarchy of business drivers is defined. projects are rated against criteria and the priority of each project is represented as a measure of its relative value toward the stated goals and objectives of the organization. the ratings are derived through pairwise comparisons. finally, optimization techniques are use to define the portfolio that maximizes value for cost based on business rules. the portfolio value measures the overall aggregated priority of all the projects that are funded. in bible and bivins (2011) the authors provide a project selection method based on the ahp. the objectives hierarchy in the evaluation model is a representation of the strategic plan. an alignment matrix shows which candidate projects support 112 int. j. prod. manag. eng. (2015) 3(1), 111-122 creative commons attribution-noncommercial-noderivatives 4.0 international sánchez, m. a. which objectives from the strategic plan (but not how much or how well). to produce accurate priorities for the alternatives (candidate projects) with respect to the objectives they support, alternatives can be evaluated using pairwise comparisons, or by using absolute measurement scales. pairwise comparisons can be cumbersome and time consuming. absolute measurement scales are based on a point scale with assigned intensity levels for each point. synthesized results of the evaluation provide the relative priorities of the candidate projects with respect to achieving a goal. the list of prioritized projects is used as the input to derive a portfolio using an optimization algorithm (the combination of projects that provides the greatest benefit for the budget level). 2.4. project management framework the methodology comprises four steps: (1) cover stakeholders’ concerns by means of stakeholder analysis; (2) define a strategy map and a balanced scorecard; (3) conduct sustainability analysis; and (4) perform a global optimization of projects (sánchez, 2014). the full description of the framework exceeds the goals and scope of this paper. in what follows, a brief description of the main steps is provided. the tasks involved in stakeholder analysis include identifying stakeholders and their interests. as a result of the analysis, stakeholders’ interests are translated into goals, and a balanced scorecard (bsc) (kaplan and norton, 2004) is drawn. the bsc is structured using four perspectives: triple bottom line, stakeholders, the internal process and learning and growth. the triple bottom line perspective includes economic, environmental and social value goals. the stakeholders perspective balances the interests of all stakeholders. to meet stakeholders’ expectations, the internal process perspective defines goals for processes. finally, the learning and growth perspective includes goals related with the skills, culture and technology necessary for its employees to do the required work. for each goal, key performance indicators (kpis) are described. then, actions plans and projects are defined. the strategy map links together several domains and elements of the strategy in the four key perspectives. the technique used to assess the environmental impact of projects depends on the characteristics of the project (e.g. production or service project). the portfolio selection is formulated as a dea problem where dmus represent portfolios; inputs represent initial investments, development, operational and disposal costs, and socio-environmental impacts derived from sustainability analysis; outputs represent the estimated contribution of portfolios to each goal. in this way, dea results provide a ranking of portfolios based on the simultaneous analysis of eco-impacts and contribution to organizational goals. similarly, project monitoring is represented as a dea problem where each project defines two dmus: one dmu represents the ongoing projects and input and output data are given by incurred costs and by realized value, or updated cost forecasts and value if the project is not closed. the other dmu represents the planned project and input and output data are given by initial estimated expenditures and expected value contribution. ideally, dmus representing planned projects would define the efficient frontier and would be the reference set for ongoing projects. 3. portfolio selection and project monitoring based on ahp the aim of this section is to describe how to perform the portfolio selection and project tracking task using ahp. the proposal is aimed to be used instead of dea. the multi-criteria analysis should include criteria to represent goals defined in the bsc, and economic, environmental, and social costs that arise of project implementation. ahp has been adapted to perform a benefit/cost analysis. in its more general form, two ahp hierarchies are used for the same set of alternatives, one for benefits and the other for costs (azis, 1990; saaty and ozdemir, 2003). after synthesis of information, the benefits’ priorities are then compared to the costs’ priorities to see which option has the highest ratio. a final ranking is calculated using the following expression (saaty and ozdemir, 2003): a b rmaxi i c 1 i $= ^ h (1) where ai represents alternative i, bi are the benefits of alternative ai , ci are the costs of alternative ai , and rmax represents the maximum of 1/ci, 1≤ i ≤ n, where n is the number of alternatives. 3.1. portfolio selection 3.1.1. benefit hierarchy criteria representing benefits are derived from goals defined in the bsc. these criteria are structured in a benefit hierarchy. each perspective in the bsc defines a branch in the hierarchy, and goals and sub113int. j. prod. manag. eng. (2015) 3(2), 111-122creative commons attribution-noncommercial-noderivatives 4.0 international which of dea or ahp can best be employed to measure efficiency of projects? goals provide criteria and sub-criteria. finally, kpis are represented at the preliminary final level of the hierarchy. alternatives are given by candidate portfolios (see figure 1) and are presented at the final level of the hierarchy. figure 1. benefit hierarchy template for portfolio selection. once the hierarchy has been stated, the relative preference of each criterion is defined by performing pairwise comparisons. for the higher part of the hierarchy, the evaluation of the importance of the criteria and sub-criteria refers to management concerns. judgments will be expressed by managers of different areas (e.g. project management, financial, marketing) according to their requirements. in the case of alternatives, assessment does not consider preferences but the evaluation is based on the forecasted contribution toward each kpi if the portfolio is funded. a portfolio is composed of projects and the portfolio contribution to a kpi is given by the maximum project contribution. assume there are z candidate portfolios. let p={pi ,1 ≤ i ≤z} be the set of portfolios. let pi={pk i ,1 ≤ k ≤ ni} denote the projects in portfolio pi where ni is the number of projects and 1 ≤ i ≤ z. the contribution of project pk i to each criterion is denoted by bk i ={bk i j, 1 ≤ j ≤ w} (2) where w is the number of criteria. hence, the contribution of portfolio pi to each criterion is described by bi ={b i max,j , 1 ≤ j ≤ w} (3) where bimax,j represents the maximum of b i k,j, 1 ≤ k ≤ ni. if a portfolio does not contribute to improve a particular kpi, then the current value of the kpi is used in the analysis. then, alternatives assessment is performed using raw data. priorities can be derived from data as well as from pairwise comparisons (forman, 2001) assuming a linear or inverse linear relationship is deemed to be reasonable. simple arithmetic is adequate to derive the priorities by adding up each alternative data value, and dividing by the total to normalize such that the priorities add up to one. similarly, inverse relationships can be calculated when a higher data value is less desirable. 3.1.2. cost hierarchy criteria representing costs are derived from economic and financial analysis developed for each candidate project in the portfolio. additionally, environmental costs defined during the sustainability analysis step provide criteria and data. the relative preference of each cost is performed using pairwise comparisons. alternatives assessment is performed using raw data (see figure 2). the portfolio cost measures the overall aggregated cost of all projects. more formally, the costs of project pk i is denoted by ck i ={ck i j , 1 ≤ j ≤ q} (4) where q is the number of cost categories. hence, the cost of portfolio pi is described by ,c c c j q1i ji ik v 1 # #= = = kj i" ,/ (5) figure 2. cost hierarchy template for portfolio selection. 3.2. project monitoring 3.2.1. benefit hierarchy the benefit hierarchy defined during portfolio selection is updated according to changes in the 114 int. j. prod. manag. eng. (2015) 3(1), 111-122 creative commons attribution-noncommercial-noderivatives 4.0 international sánchez, m. a. bsc. hence, some goals and kpis may be added and others may be eliminated. alternatives are given by new proposals and ongoing projects (see figure 3) and are presented at the final level of the hierarchy. if there is a change in the hierarchy defined during portfolio selection, then it is necessary to perform pairwise comparisons to calculate the relative preference of each criterion. in the case of alternatives, assessment is based on each alternative (project) contribution to each criterion. alternatives assessment is performed by specifying the forecasted value for each key measure if the project is implemented. if a project does not contribute to improve a particular kpi, then the current value of the kpi is used in the analysis. during project development, measures will remain unchanged, but once a project is completed, projects will deliver benefits and measures will be updated accordingly. figure 3. benefit hierarchy template for project monitoring. for the lower level of the hierarchy (the level of the alternatives) the evaluation considers numerical information updated at the control point of interest. in this way, kpis’ values recorded in the bsc can be directly used and time-consuming pairwise comparisons are avoided. 3.2.2. cost hierarchy the approach to define the cost hierarchy is similar to the one used during portfolio selection. criteria representing costs are derived from economic, environmental and social analysis developed for each project in the funded portfolio. the relative preference of each cost is performed using pairwise comparisons. alternatives assessment is performed using raw data (see figure 4). 4. application case alas ingenieria is a small information technology company located in bahía blanca (argentina) since 1991. the company provides advanced solutions for engineering and information management for industrial plants. they also provide support to develop, implement and integrate applications. currently the company is organized under two segments –software products and services. the owner announced intent to explore options to promote growth, efficiency and improve the company´s social responsibility image. after performing a stakeholders’ analysis (whose description is out of the scope of this paper), a strategy map (see figure 5) and a bsc are defined. in what follows, the results of project monitoring are described. the funded portfolio is composed of projects described in table 1. figure 4. cost hierarchy template for project monitoring. 4.1. criteria and sub-criteria definition the benefit hierarchy reflects the information provided by the bsc, i.e. perspectives; goals and kpis (see figure 6 and figure 7). the cost impact categories which are particularly significant for this study are energy consumption, paper use and economic (initial costs and total cost of ownership). the cost hierarchy is structured using these categories (figure 8). 4.2. alternatives definition alternatives are given by projects included in table 1. for each project, two alternatives are defined: a) alternative pi, 1 ≤ i ≤ 18, represents a project as planned. b) alternative ri, 1 ≤ i ≤ 18, represents the project´s status at a control point. 115int. j. prod. manag. eng. (2015) 3(2), 111-122creative commons attribution-noncommercial-noderivatives 4.0 international which of dea or ahp can best be employed to measure efficiency of projects? figure 5. strategy map (partial view). table 1. projects and supported goals in the strategy map. projects goals in the strategy map process map definition 9 10 15 18 equipment repair and donation 13 23 train employees 12 16 17 25 process control 17 iso 9001:2008 certification 18 account information processing for financial analysis 1 18 cost analysis 1 18 train customers about responsible use of products 19 upgrade appliances and electronics 20 train employees about energy efficiency 20 paper less initiative 21 sustainable acquisition of products 22 responsibly disposal of batteries 23 paper recycling 23 develop employee discount programs 12 26 conduct employee performance evaluation 16 financial software module deployment 2 crm software module deployment 3 4 6 8 116 int. j. prod. manag. eng. (2015) 3(1), 111-122 creative commons attribution-noncommercial-noderivatives 4.0 international sánchez, m. a. figure 6. benefit hierarchy. figure 7. benefit hierarchy (continuation). figure 8. cost hierarchy. 4.3. criteria, sub-criteria and alternatives definition in this case, the same relative preference was given to all criteria. in this way, ahp and dea results are derived from the same scenario. the benefits and costs hierarchy show global and local priorities for each criteria and sub-criteria. for example, criteria triple bottom line, stakeholders, internal process and learning and growth have the same priority with respect to the global goal (25%). alternatives assessment at control t point is performed using data updated to that instant of time. 4.4. synthesis results once that all judgments had been defined, numerical evaluations are computed using expert choice® software. table 3 (appendix i) includes results of benefits, costs, and an integrated score. the final score that integrates benefits and costs is calculated using expression (1) (see section 3). in order to give an interpretation to ahp priorities, recall that data used to assess alternatives pi belong from project plans; while data used to assess alternative ri at control point t (1 ≤ i ≤ 18), belong from data updated at this control point. hence, the aim of ahp step is analyzing each pair pi , ri, and finding out if priorities are different. for example, if pi score is greater than ri score, then it may be that benefits have not been realized yet, or that ri spending has been more than planned. table 4 (appendix ii) includes an interpretation of ahp synthesis. it can be seen that resulting priorities reflect the status of projects. the worst score is for project 5 because its costs are much higher than the rest of the projects. alternative p1 has a low priority since forecasted cost is quite high. projects 17 and 18 show a low priority because the score based on cost information is relatively bigger than others (see table 3 in appendix i). 4.5. dea versus ahp this section is devoted to compare and discuss ahp and dea results. since score numbers are not comparable in absolute terms, the aim is to find out if decisions derived from ahp results are consistent with decisions derived from ahp scores. in other words, for each pair pi , ri it is discussed if ahp scores are consistent with dea ones. table 2 summarizes results and table 5 (appendix iii) includes dea scores. in general, they support the 117int. j. prod. manag. eng. (2015) 3(2), 111-122creative commons attribution-noncommercial-noderivatives 4.0 international which of dea or ahp can best be employed to measure efficiency of projects? same decisions based in ahp. it can be concluded that dea scores are more precise with respect to the current progress of projects. dea scores may not be realistic when the reference set of a project ri is pj, i ≠ j (see remarks about project 1 in table 2). discrepancies in ahp and dea rankings are not surprising since techniques are different. for example, dea gives an outstanding rank for r5 (number 5). the score of p5 is smaller than r5´s score, and r5 is p5´s reference set. in other words, p5 is punished because it is compared to r5. in addition, there are no other units with similar production levels so r5 is deemed as efficient. this result arises because of the specialization problem that is a known drawback of dea. by using ahp, the ranking of p5 (18) is also worse than r5´s rank (17), but the difference is not so large. to summarize, while the comparison of ahp and dea scores are consistent for pairs pi ,ri, rankings do not provide useful information because they compare all projects and management decisions should be based on the score analysis of pairs pi ,ri. 5. conclusions this work describes how to use ahp as an aid in project management. the proposal is grounded on an existing project management framework to select and monitor projects based on dea, and discusses modifications and additions arising from using ahp. table 2. summary of ahp and dea scores comparison for each pair pi , ri. projects ahp and dea scores for pi , ri process map definition even though dea finds both p1 and r1 efficient, p1 score is better than r1. on the other hand, ahp computes a better score for r1. the current status is that almost all benefits have been realized and the spending is much less than planned. ahp results are consistent with this. dea finds p7 as a reference set of r1, then r1 is not compared with p1. equipment repair and donation both dea and ahp give a better score to p2. however, dea provides a bigger difference between p2 and r2. p2 is the reference set of r2. in fact, the total budget was spent and full benefits are expected in the future. dea better highlights the situation. train employees similar remarks as for project 2. process control similar remarks as for project 2. iso 9001:2008 certification consistent. account information processing for financial analysis consistent. cost analysis consistent. train customers about responsible use of products consistent. upgrade appliances and electronics consistent. train employees about energy efficiency consistent. paper less initiative there is a slight difference between p11 and r11 ahp scores while dea scores are equal. however, it is doubtful that the decimal points are relevant. so it may be concluded that scores are consistent. sustainable acquisition of products consistent. responsibly disposal of batteries consistent. paper recycling consistent. develop employee discount programs dea finds p15 inefficient and r15 efficient. ahp provides the same score. the current scenario is that r15 provided more benefits than planned with the estimated budget. then, dea reflects the situation while ahp does not. conduct employee performance evaluation consistent. financial software module deployment consistent. crm software module deployment consistent. 118 int. j. prod. manag. eng. (2015) 3(1), 111-122 creative commons attribution-noncommercial-noderivatives 4.0 international sánchez, m. a. the main components of the ahp model is a benefit hierarchy that structures goals represented in a bsc, and a cost hierarchy that includes costs derived from economic, environmental and social impacts derived from project development. criteria and sub-criteria assessment is realized doing pairwise comparisons. for the case of alternatives, the evaluation considers raw data. the use of ahp allows overcoming some limitations of dea. the first is the introduction of preferences. the possibility of assigning preferences to criteria allows considering different scenarios and performing what-if analysis. scenario analysis is often a requirement when selecting projects since uncertainty in many factors such market development, cost variance, among many others. in particular, when project selection takes into account economic, environmental and social dimensions, reasoning about the impact of each dimension enhances the analysis. the second limitation of dea is the homogeneity assumption where all dmus are required to have comparable production levels. since ahp does not require alternatives to be similar, projects of different size may be compared. in addition, ahp does not make assumptions about the number of alternatives. in dea, it is desirable that the number of dmus exceeds the number of inputs and outputs several times. finally, ahp allows multiple decision makers to give judgments. while this option was not used in the work, it may be relevant since it favors inclusive approaches that allow the participation of multiple actors. on the other hand, ahp has some disadvantages. sometimes, it is not advisable to derive priorities directly from hard data because preferences are often not linearly related to data. for instance, if with respect to initial cost, alternative a is two times more preferable than b, then a may not be twice as preferable as b. how to systematically derive rating scales from raw data is a potential direction for research. references azis, i. (1990). analytic hierarchy process in the benefit-cost framework: a post-evaluation of the trans-sumatra highway project. european journal of operational research, 48(1): 38-48. doi:10.1016/0377-2217(90)90059-k banker, r., charnes, a., cooper, w. (1984). some models for estimating technical and scale inefficiencies in data envelopment analysis. management science, 30(9): 1078-1092. doi:10.1287/mnsc.30.9.1078 bible, m., bivins, s. (2011). mastering project portfolio management. fort lauderdale: j. ross publishing, inc. camanho, a., dyson, r. (1999). efficiency, size, benchmarks and targets for bank branches: an application of data envelopment analysis. journal of the operational research society, 50: 903-915. doi:10.1057/palgrave.jors.2600792 chang, d., yang, f. (2010). data envelopment analysis with two distinct objectives of inputs or outputs. in proc. the 6th international symposium on management, engineering and informatics. international institute of informatics and systemics, florida, usa. charnes, a., cooper, w., rhodes, e. (1978). measuring the efficiency of decision making units. european journal of operational research, 2(6): 429-444. doi:10.1016/0377-2217(78)90138-8 cook, w., seiford, l. (2009). data envelopment analysis (dea) -thirty years on. european journal of operational research, 192(1): 1-17. doi:10.1016/j.ejor.2008.01.032 forman, e. s. (2001). decision by objectives how to convince others that you are right. river edge, new jersey: world scientific publishing. joro, t., viitala, e. (2004). weight-restricted dea in action: from expert opinions to mathematical models. journal of the operational research society, 55: 814-821. doi:10.1057/palgrave.jors.2601752 kaplan, r., norton, r. (2004). strategy maps. converting intangible assets into tangible outcomes. boston: harvard business school press. kendrick, j., saaty, d. (2007). use analytic hierarchy process for project selection. six sigma forum magazine, pp. 22-29. kumar, s. (2004). ahpbased formal system for r&d project evaluation. journal of scientific & industrial research, 63: 888-896. madlener, r., antunes, c., dias, l. (2009). assessing the performance of biogas plants with multi-criteria and data envelopment analysis. european journal of operational research, 197(3):1084-1094. doi:10.1016/j.ejor.2007.12.051 mar, c. (2009). specialization versus diversification: non-homogeneity in data envelopment analysis. proc. 3rd international conference on industrial engineering and industrial management, barcelona, spain, pp. 1125-1133. porter, m., kramer, m. (2011). creating shared value. how to reinvent capitalism -and unleash a wave of innovation and growth. harvard business review, 1-17. saaty, t. (1997). toma de decisiones para líderes: el proceso analítico jerárquico. la toma de decisiones en un mundo complejo. pittsburgh: rws publications. 119int. j. prod. manag. eng. (2015) 3(2), 111-122creative commons attribution-noncommercial-noderivatives 4.0 international which of dea or ahp can best be employed to measure efficiency of projects? saaty, t., ozdemir, m. (2003). negative priorities in the analytic hierarchy process. mathematical and computer modelling, 37(9-10): 10631075. doi:10.1016/s0895-7177(03)00118-3 sánchez, m. (2014). integrating sustainability issues into project management. journal of cleaner production, 96: 319-330. doi:10.1016/j. jclepro.2013.12.087. shimshak, d., lenard, m., klimberg, r. (2009). incorporating quality into data envelopment analysis of nursing home performance: a case study. omega, 37(3): 672-685. doi:10.1016/j.omega.2008.05.004 yalabik, b., fairchild, r. (2011). customer, regulatory, and competitive pressure as drivers of environmental innovation. international journal of production economics, 131(2): 519-527. doi:10.1016/j.ijpe.2011.01.020 appendix i. ahp synthesis results table 3. detail of ahp results. projects (alternatives) benefits costs integral score p1: process map definition 0.026 0.054 0.00048148 p2: equipment repair and donation 0.029 0.001 0.029 p3: train employees 0.028 0.013 0.00215385 p4: process control 0.028 0.001 0.028 p5: iso 9001:2008 certification 0.029 0.455 6.3736×10–5 p6: account information processing for financial analysis 0.027 0.001 0.027 p7: cost analysis 0.028 0.001 0.028 p8: train customers about responsible use of products 0.028 0.053 0.0005283 p9: upgrade appliances and electronics 0.027 0.001 0.027 p10: train employees about energy efficiency 0.027 0.001 0.027 p11: paper less initiative 0.027 0.015 0.0018 p12: sustainable acquisition of products 0.029 0.001 0.029 p13: responsibly disposal of batteries 0.028 0.001 0.028 p14: paper recycling 0.027 0.001 0.027 p15: develop employee discount programs 0.029 0.018 0.00161111 p16: conduct employee performance evaluation 0.03 0.008 0.00375 p17: financial software module deployment 0.025 0.033 0.00075758 p18: crm software module deployment 0.03 0.03 0.001 r1: process map definition 0.027 0.023 0.00117391 r2: equipment repair and donation 0.028 0.001 0.028 r3: train employees 0.027 0.001 0.027 r4: process control 0.028 0.001 0.028 r5: iso 9001:2008 certification 0.028 0.18 0.00015556 r6: account information processing for financial analysis 0.027 0.001 0.027 r7: cost analysis 0.028 0.001 0.028 r8: train customers about responsible use of products 0.028 0.018 0.00155556 r9: upgrade appliances and electronics 0.027 0.001 0.027 r10: train employees about energy efficiency 0.027 0.001 0.027 r11: paper less initiative 0.027 0.013 0.00207692 r12: sustainable acquisition of products 0.028 0.001 0.028 r13: responsible disposal of batteries 0.028 0.001 0.028 r14: paper recycling 0.028 0.001 0.028 r15: develop employee discount programs 0.029 0.018 0.00161111 r16: conduct employee performance evaluation 0.028 0.008 0.0035 r17: financial software module deployment 0.025 0.033 0.00075758 r18: crm software module deployment 0.028 0.03 0.00093333 120 int. j. prod. manag. eng. (2015) 3(1), 111-122 creative commons attribution-noncommercial-noderivatives 4.0 international sánchez, m. a. appendix ii. project monitoring based on ahp table 4. project monitoring results based on ahp. projects decision alternative score comments process map definition continue p1 0.00048 in progress. achievement of some benefits. costs less than planned.r1 0.0018 equipment repair and donation continue p2 0.029 full benefits expected on the final period. total budget spent.r2 0.028 train employees continue p3 0.002154 in progress. some benefits expected on the final period, others benefits expected in the longterm. costs less than planned.r3 0.027 process control continue p4 0.028 in progress. some benefits expected on the final period, others benefits expected in the longterm. costs less than planned.r4 0.028 iso 9001:2008 certification continue p5 6.37363×10–5 in progress. achievement of some benefits. costs less than planned.r5 0.00016 account information processing for financial analysis completed p6 0.027 deliverables fully accomplished. on budget. r6 0.027 cost analysis completed p7 0.028 deliverables fully accomplished. on budget. r7 0.028 train customers about responsible use of products continue p8 0.00053 in progress. achievement of some benefits. costs less than planned.r8 0.0016 upgrade appliances and electronics completed p9 0.027 deliverables fully accomplished. on budget. r9 0.027 train employees about energy efficiency completed p10 0.027 deliverables fully accomplished. on budget. r10 0.027 paper less initiative continue p11 0.0018 in progress. achievement of some benefits. costs less than planned.r11 0.0021 sustainable acquisition of products continue p12 0.029 in progress. achievement of some benefits. costs less than planned.r12 0.028 responsible disposal of batteries continue p13 0.028 in progress. achievement of some benefits. costs less than planned.r13 0.028 paper recycling continue p14 0.027 in progress. achievement of all benefits. costs less than planned.r14 0.028 develop employee discount programs continue p15 0.0016 in progress. achievement of some benefits. on budget.r15 0.0016 conduct employee performance evaluation continue p16 0.00375 in progress. achievement of some benefits. on budget.r16 0.0035 financial software module deployment completed p17 0.00076 deliverables fully accomplished. on budget. r17 0.00076 crm software module deployment continue p18 0.001 in progress. achievement of some benefits. on budget.r18 0.00093 121int. j. prod. manag. eng. (2015) 3(2), 111-122creative commons attribution-noncommercial-noderivatives 4.0 international which of dea or ahp can best be employed to measure efficiency of projects? appendix iii. project monitoring based on dea table 5. project monitoring scores based on dea. dmu score dmu score dmu score p1 1.03 p7 1.00 p13 1.02 r1 1.00 r7 1.00 r13 1.03 p2 1.01 p8 1.00 p14 1.00 r2 0.38 r8 1.02 r14 1.03 p3 1.06 p9 0.77 p15 0.76 r3 1.00 r9 0.77 r15 1.02 p4 1.02 p10 1.00 p16 1.02 r4 0.35 r10 1.00 r16 0.82 p5 1.01 p11 0.76 p17 0.17 r5 1.03 r11 0.76 r17 0.17 p6 1.00 p12 1.01 p18 1.05 r6 1.00 r12 1.02 r18 0.31 122 int. j. prod. manag. eng. (2015) 3(1), 111-122 creative commons attribution-noncommercial-noderivatives 4.0 international sánchez, m. a. pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2018.8617 received 2017-09-21 accepted: 2018-04-13 does cmmi implementation affect the performance of the firm? an evaluation from a dynamic capabilities approach eduardo ajenjoa1, natalia martín-cruza2, cristina ruiz-martina3 and adolfo lópez-paredesa4 grupo insisoc. escuela de ingenierías industriales. universidad de valladolid, pº. del cauce 59, 47011-valladolid, spain a1 eduardo.ajenjo@gmail.com, a2 ambiela@eco.uva.es, a3 cruiz@eii.uva.es, a4 aparedes@eii.uva.es abstract: in this paper, we study the impact of the capability maturity model integration (cmmi) on firm performance both during and after its implementation. the literature pointed out that cmmi is theoretically related to the generation of dynamic capabilities. to give an empirical view of these theories, we built a database of economic and financial data from spanish firms involved in programming, consultancy or another computer-related sector. this data allowed us to study the relationship between the use of cmmi and the firm economic and financial performance in an empirical way. the main finding of the analysis is a negative relationship between the use of cmmi and profitability in the firms during the analyzed period and sector. key words: dynamic capabilities, cmmi, firm performance. 1. introduction in the last decades, several maturity models have been developed with the focus on improving different tasks in organizations such as the processes, the project management or the knowledge management. one of these models is the capability maturity model integration (cmmi). cmmi is a set of good practices developed by the software engineering institute. it is focused on improving the company processes with special application in the software industry. this maturity model has been theoretically related to the development of several capabilities in the organizations that helps in their performance (lee & wu, 2007). our goal is to test whether there is or not an empirical relation between the implementation of maturity models and the performance of the organizations. to achieve our goal, we performed an empirical analysis based on spanish firms involved in programming, consultancy or another computer-related sector to test two hypotheses. we will examine if the performance of a firm is likely to increase after the process of cmmi implementation and if it is likely to decline during the process of cmmi implementation in the following sections, we discuss the results we obtained. in section 2, we present the background and our research hypothesis. in section 3, we explain the methodology we have followed to drive our analysis. in section 4, we describe the results of our study. finally, in section 5 we discuss our results, we present the main conclusions and future lines of this research. to cite this article: ajenjo. e., martín-cruz, n., ruiz-martin, c. and lópez-paredes, a. (2018). does cmmi implementation affect the performance of the firm? an  evaluation from a dynamic capabilities approach. international journal of production management and engineering, 6(2), 57-64. https://doi.org/10.4995/ijpme.2018.8617 int. j. prod. manag. eng. (2018) 6(2), 57-64creative commons attribution-noncommercial-noderivatives 4.0 international 57 mailto:eduardo.ajenjo@gmail.com mailto:ambiela@eco.uva.es mailto:cruiz@eii.uva.es mailto:aparedes@eii.uva.es http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. background and hypothesis nowadays, more and more organizations adopt maturity models, such as cmmi or other project management maturity models (pmmm), to improve their performance. despite cmmi has been criticized for the lack of developed procedures (reifer, 2000; sun & liu, 2010) or for the difficulty in finding the areas where the improvement efforts have to be focused (huang & han, 2006), it is a good tool to improve schedule, costs and even the return of investment in the organizations (gibson, goldenson, & kost, 2006; goldenson & gibson, 2003). other maturity models, such as pmmm are also related to a better organizational performance (de oliveira moraes & barbin laurindo, 2013; nenni, arnone, boccardelli, & napolitano, 2014), although they have been criticized for being irrelevant and for the need for greater flexibility and adaptability to the organization (mullaly, 2014). young et al. (2014) state that higher levels of maturity will help to increase the organization performance. however, other authors conclude that a higher level of maturity does not necessarily imply greater success; each organization needs to find its appropriate level (albrecht & spang, 2014; de oliveira moraes & barbin laurindo, 2013). staples and niazi (2008) affirm that the most common reasons to adopt capability maturity models are to improve quality and project performance, and to enhance process management. lee and wu (2007) study cmmi as a source of strategic capabilities, which are related to the developing of dynamic capabilities. the dynamic capability approach (teece, pisano, & shuen, 1997) tries to explain the internal processes that a firm uses to be competitive. dynamic capabilities are best practices and organizational processes that allow creating competitive advantages (eisenhardt and martin, 2000). organizations need them for long-term enterprise success (augier & teece, 2009; wu, he, & duan, 2013), but they also require a good strategy and own vrin resources (valuable, rare, in-imitable and non-substitutable resources) (teece, 2014). furthermore, dynamic capabilities affect positively to organizational (chien & tsai, 2012; tseng & lee, 2014) and innovation performance (zheng, zhang, & du, 2011). dynamic capabilities can be viewed as the result of knowledge management activities (nielsen, paarup, & paarup, 2006; zheng et al., 2011). tseng and lee (2014) give evidence to support that development of knowledge management promotes the growth of dynamic capabilities. besides, knowledge management and cmmi can be used together, allowing the organization to be more efficient (dayan & evans, 2006). despite the lack of consensus on the terminology used in knowledge management area (klimko, 2001), several authors state the possibility of implementing knowledge management maturity models to enable the organization to pursue a path of improvement in this field (kuriakose & raj, 2010). several scholars have found a positive correlation between better knowledge management and financial performance of the firm (andreeva & kianto, 2012; tanriverdi, 2005; zack, mckeen, & singh, 2009), innovation capacity (darroch, 2005) and organizational performance (kruger & johnson, 2011; tseng & lee, 2014). based on previous literature, we present the first hypothesis: h1. the performance of a firm is likely to increase after the process of cmmi implementation. the implementation of cmmi requires a deep organizational change, for which is need to invest several resources and time. these kinds of changes can mean financial losses at least in the short term (mellert, scherbaum, oliveira, & wilke, 2015) caused by different reasons such as the resistance to change or a bad planning (kotter, 2007). several scholars have pointed the lack of standardized procedures or a clear structure to implement cmmi as some of the main disadvantages of this model (niazi, wilson, & zowghi, 2005; reifer, 2000), which could hinder achieving the desired results. in a practical way, shih et al., 2013 study the implementation of cmmi in an organization, noting that in a first step this change has negative effects on the firm performance. int. j. prod. manag. eng. (2018) 6(2), 57-64 creative commons attribution-noncommercial-noderivatives 4.0 international ajenjo. e., martín-cruz, n., ruiz-martin, c. and lópez-paredes, a. 58 http://creativecommons.org/licenses/by-nc-nd/4.0/ based on previous literature, we present the second hypothesis: h2. the performance of a firm is likely to decline during the process of cmmi implementation. 3. methods to study whether the implementation of cmmi creates or not a competitive advantage that is reflected in the economic or financial performance of the firm, we carried a statistical analysis to compare the results of different firms which have or do not have cmmi implemented. for our analysis, we have chosen spanish firms in the information and technology (it) sector. this election is based on the fact that despite cmmi is now used in almost every sector, it was initially developed for it companies. the firm’s economic and financial data were obtained from the database “amadeus”. we selected the spanish firms belonging to the nace code 62 (statistical classification of economic activities in the european community). all these firms belong to the it sector. we obtained 2130 firms. we refined this search because, although amadeus usually contains data of the main economic and financial indicators for the last 10 years, it had incomplete data for some firms. we selected a period of time that gives us a balance between the length of the time interval to study and the number of firms we are able to analyze. we selected the period 20082013 for the following reasons: it is long enough (6 years) to see a tendency it belongs to the same economic cycle: the economic crisis the number of firms in this period is quite large: 899 firms, giving us 5394 observations. if we extend the period one year more, the number of firms will drop to 779. we have selected the following financial and economic indicators for our analysis as they are non-dimensional and they are not prejudiced by the size of the firm: return on equity (roe), return on assets (roa) and return on capital (roc). the data regarding cmmi implementation was acquired analyzing the results of the cmmi appraisal published on the cmmi website (https://sas.cmmiinstitute.com/pars/pars.aspx). despite this is the official site for cmmi, it has a drawback for getting the data: it only shows the firms with an active cmmi certification. it means, that we cannot get when the organization first got the certification or if a firm has abandoned it. to deal with this issue, we also looked at javier garzas’ blog (http://www.javiergarzas.com/2011/12/empresasespanolas-evaluadas-con-cmmi.html). in this blog, we checked which companies had the certification in 2011. the caelum consulting website showed which firms had the certification upon february 2007. as these last two sources of information were not the official ones, we validated the information directly with the firms. we contacted them either by phone or email. we found 40 firms that have or have had a cmmi certification during the time period 20082013. the annual distribution is shown in table 1. we can easily see that some firms may have abandoned the certification, and others have got it. table 1. cmmi annual distribution for spanish firms in the it sector (nace code 62). year 2008 2009 2010 2011 2012 2013 # of firms with cmmi certification 17 23 26 31 36 38 4. results we have first performed a descriptive analysis of the data used in this study for a better understanding of the influence of cmmi on the firm performance. we cannot forget that the period under study (20082013) belongs to the economic crisis. the initial number of observations was 5394. however, we see that during the analysis this number has decreased due to the closure of some firms. 4.1. descriptive analysis of the data we analyze the number of employees in the firms, foundation year, sales revenue, roa, roc and roe to understand the evolution during the studied time period. 4.1.1. number of employees in the firm in table 2, we depict the evolution of the number of employees of the firms during the time period. we compare firms with and without cmmi. we can see that the number of employees globally increased despite the economic crisis. conversely, int. j. prod. manag. eng. (2018) 6(2), 57-64creative commons attribution-noncommercial-noderivatives 4.0 international does cmmi implementation affect the performance of the firm? an evaluation from a dynamic capabilities approach 59 https://sas.cmmiinstitute.com/pars/pars.aspx http://www.javiergarzas.com/2011/12/empresas-espanolas-evaluadas-con-cmmi.html http://www.javiergarzas.com/2011/12/empresas-espanolas-evaluadas-con-cmmi.html http://creativecommons.org/licenses/by-nc-nd/4.0/ the number of employees within the firms which implement cmmi decreases over this time period. one explanation may be that the firms are more efficient and they need fewer employees. however, a better one may be that at the beginning only big corporations implemented cmmi, but over time more and more medium and small firms adopt this maturity model. 4.1.2. foundation year analysing the firms’ foundation year, we found that most firms were founded in the 90s. we need to take into account that there are much older firms (founded in the 30s) and much younger firms (founded after 2000). firms with cmmi implemented are on average a bit older (1993) than firms without cmmi (1997). 4.1.3. sales revenue analysing the sales revenue we observe a great variability. there are firms without incomes (probably representing closed companies) and companies such as “indra sistemas” with almost 3.000 million €. distinguishing firms with and without cmmi, we found that on average, firms with cmmi implemented have a greater average (151.70 million €) than the ones without cmmi (8.97 million €). 4.1.4. roa analysing the evolution of the roa, we find a decreasing tendency during the studied period. we found that the roa was higher in the firms with cmmi during 2008 and 2011. it was the opposite during 2009, 2010, 2012 and 2013. for a clear analysis, we discard the 50 more extreme observations, which is less than 1% of our data (see table 3 for the trend). we observed the same tendency and we reached the same conclusion when we analysis firm with cmmi and without it separately. table 3. evolution of the roa annual mean without the 50 more extreme observations. year 2008 2009 2010 2011 2012 2013 roa. all firms 9.01 5.93 6.16 5.10 5.17 4.33 roa. firms with cmmi 9.05 5.39 6.07 6.25 4.17 1.07 roa. firms without cmmi 9.00 5.95 6.17 5.06 5.21 4.48 4.1.5. roc analysing the evolution of the roc, we find a decreasing tendency during the studied period. we found that the roc was higher in the firms with cmmi during 2008 and 2010. it was the opposite during 2009, 2011, 2012 and 2013. for a clear analysis, we discard the 50 more extreme observations, which is less than 1% of our data (see table 4 for the trend). we observed the same tendency and we reached the same conclusion when we analysis firm with cmmi and without it separately. however, when we discard the extreme observations we find that the distance between the firms with and without cmmi is smaller. table 4. evolution of roc annual mean without the 50 more extreme observations. year 2008 2009 2010 2011 2012 2013 roc. all firms 22.47 14.99 15.87 13.35 12.74 10.93 roc. firms with cmmi 26.81 12.62 17.19 10.76 9.41 1.83 roc. firms without cmmi 22.37 15.06 15.83 13.45 12.9 11.39 4.1.6. roe analysing the evolution of the roe, we find a decreasing tendency during the studied period. we found that the roe was higher in the firms with cmmi only during 2008. for a clear analysis, we discard the 50 more extreme observations, which is less than 1% of our data (see table 5 for the trend). table 2. evolution of the number of employees. the data represent the annual mean of employees. year 2008 2009 2010 2011 2012 2013 # of employees. all firms 121.88 120.02 124.84 134.13 143.00 145.13 # of employees. firms with cmmi 2706.80 2058.10 2047.20 1911.40 1838.10 1767.80 # of employees. firms without cmmi 68.08 66.57 67.76 72.62 71.60 72.71 int. j. prod. manag. eng. (2018) 6(2), 57-64 creative commons attribution-noncommercial-noderivatives 4.0 international ajenjo. e., martín-cruz, n., ruiz-martin, c. and lópez-paredes, a. 60 http://creativecommons.org/licenses/by-nc-nd/4.0/ we observed the same tendency and we reached the same conclusion when we analysis firm with cmmi and without it separately. table 5. evolution of roe annual mean without the 50 more extreme observations. year 2008 2009 2010 2011 2012 2013 roe. all firms 26.55 15.45 16.11 13.91 11.62 9.42 roe. firms with cmmi 28.17 14.97 16.04 8.24 9.27 -11.08 roe. firms without cmmi 26.52 15.47 16.11 14.11 11.72 10.37 4.2. panel analysis after describing our sample, we focus on the influence of cmmi on the firm performance. for this purpose, we do a regression analysis during the period of the panel data. the dependent variables are roa, roc, roe and their interannual differences (dif roa, dif roc, dif roe). the independent variables are: cmmi: it takes the value 1 if the company has cmmi in the studied year and 0 otherwise. this variable allows us to study the effect of cmmi certification on the firm performance. cmmi level: if the company has cmmi in the studied year it takes the value of the cmmi certification level and 0 otherwise. this variable allows us to study the effect of cmmi certification level on the firm performance. lag cmmi: if the company has cmmi in the previous year it takes the value 1 and 0 otherwise. it may be possible that the effect of cmmi in the firm performance is not immediate. this variable allows studying the effect of cmmi on the firm performance 1 year after its implementation. lag2 cmmi: if the company has cmmi in the two previous years it takes the value 1 and 0 otherwise. this variable allows studying the effect of cmmi on the firm performance 2 years after its implementation. we use the number of employees and the foundation year as control variables. we introduce lag cmmi and lag2 cmmi independent variables to distinguish the implementation period and the post-implementation identified in our research hypothesis. we identify when the company adopted cmmi since cmmi will take the value 1 and lag cmmi and lag2 cmmi the value 0. we analyze if cmmi has influence in the interannual variation of the economic and financial performance, and on the absolute economic and financial performance of the firm. as a robustness analysis, we repeat the analysis dividing the sample into two periods. we choose 2010/2011 as breaking point since, in the descriptive analysis, we found that the firms with and without cmmi had a similar performance during the first three years. however, during the second period, the performance of the firms with cmmi was quite worse than the ones without it. we consider that the regressions are significant if prob > f is less than 0.05 and that the variable is significant if p > | t | is less 0.05. we have to notice that when the regression has only one variable prob > f is equal to p > | t |. 4.2.1. study of the interannual variation of the firm performance we study the impact of cmmi on the variation of the firm performance. we analyze 12 statistical regressions to relate dif roa, dif roc and dif roe to our independent variables (cmmi, cmmi level, lag cmmi and lag2 cmmi) in an individual manner. we use the number of employees as a control variable. we find that none of these regressions is significant. table 6. variation on the firm performance vs the implementation of cmmi. cmmi cmmi level lag cmmi lag2 cmmi coef. p>|t| coef. p>|t| coef. p>|t| coef. p>|t| dif. roa -3.26 0.407 -1.25 0.375 -2.41 0.540 -4.13 0.390 dif. roc -16.77 0.308 -6.41 0.311 14.87 0.362 -5.02 0.800 dif. roe -21.13 0.274 -8.67 0.231 12.91 0.507 -10.95 0.650 int. j. prod. manag. eng. (2018) 6(2), 57-64creative commons attribution-noncommercial-noderivatives 4.0 international does cmmi implementation affect the performance of the firm? an evaluation from a dynamic capabilities approach 61 http://creativecommons.org/licenses/by-nc-nd/4.0/ we repeat the study without the employee number to check if this variable is distorting our results. however, we also find that the regressions are not significant as shown in table 6. 4.2.2. study of the firm performance we repeat the study to analyze the effect on the firm performance. we related roa, roc, and roe to the independent variables (cmmi, cmmi level, lag cmmi and lag2 cmmi) individually. we use the number of employees and the foundation year as control variables. we find that both the adoption and level of cmmi affects negatively to the roe and roc as shown in table 7. the other regressions are not significant, but all of them pointed out the same negative relation. we repeat the analysis removing the foundation year and the foundation year and a number of employees. we get the same conclusion. 4.2.3. study of the firm performance. sample divided in two periods we repeat the study to analyze the effect on the firm performance dividing the time period into two slots as explained above. we related roa, roc, and roe to the independent variables (cmmi, cmmi level, lag cmmi and lag2 cmmi) individually. in this case, we do not use control variables, as the effect of this variables was tested in the previous regression without dividing the time period. the regression results are shown in table 8. during the first period, we do not find any significant regression. lag2 cmmi column is empty because we do not have enough data to calculate the regression. during the second period, we find that both the adoption and level of cmmi affects negatively to the roe and roc 5. discussion and conclusions based on the results presented in section 4, we only found two significant variables to explain the performance of the firm based on the implementation or not of cmmi. these two variables are cmmi (i.e. the existence or absence of cmmi in the firm) and cmmi level (i.e. the level of cmmi the company has implemented). both variables are negatively related to the performance of the firm. when we divided the sample into two time periods, we found that these variables only are significant between 2011 and 2013. cmmi is a maturity model that provides a set of best practices to improve the firm processes, especially in software companies. looking at cmmi website table 7. firm performance vs. the implementation of cmmi. cmmi cmmi level coef. p>|t| prob > f coef. p>|t| prob > f roa -5.37 0.036 0.1700 -1.99 0.049 0.2100 roc -36.30 0.001 0.0039 -16.76 0.000 0.0007 roe -31.64 0.019 0.0320 -13.85 0.015 0.0260 lag cmmi lag2 cmmi coef. p>|t| prob > f coef. p>|t| prob > f roa -4.82 0.104 0.364 -3.20 0.353 0.8000 roc -24.40 0.058 0.160 -14.87 0.303 0.6500 roe -14.73 0.358 0.310 -12.73 0.474 0.7093 table 8. firm performance vs. the implementation of cmmi. period 2008-2010 and 2011-2013. cmmi cmmi level lag cmmi lag2 cmmi 2008 2010 coef. p>|t| coef. p>|t| coef. p>|t| coef. p>|t| roa 1.330 0.798 0.508 0.804 1.16 0.865 * * roc -5.000 0.815 -2.260 0.788 1.20 0.969 * * roe -2.625 0.930 -1.390 0.907 2.50 0.949 * * 2011 2013 coef. p>|t| coef. p>|t| coef. p>|t| coef. p>|t| roa -7.40 0.065 -2.41 0.112 -6.40 0.129 -3.56 0.450 roc -53.94 0.001 -25.70 0.0001 -17.37 0.316 -13.78 0.456 roe -48.72 0.017 -20.02 0.011 -11.55 0.589 -11.78 0.645 int. j. prod. manag. eng. (2018) 6(2), 57-64 creative commons attribution-noncommercial-noderivatives 4.0 international ajenjo. e., martín-cruz, n., ruiz-martin, c. and lópez-paredes, a. 62 http://creativecommons.org/licenses/by-nc-nd/4.0/ we find which companies have adopted this maturity model and when. based on the literature, we established a positive relation of the application of cmmi and the firm performance (goldenson & gibson 2003; gibson et al., 2006). however, cmmi has also been criticized for its difficulty to put the model in practice (reifer 2000; sun & liu 2010). dynamic capabilities are defined as the capacity to organize, integrate and develop new capabilities (teece et al., 1997). teece (2014) considers that dynamic capabilities are inimitable; they should be built and learned. on the other hand, eisenhardt & martin, (2000) consider them as replicable processes. following this view, cmmi may help to develop dynamic capabilities (lee & wu 2007). cmmi implementation triggers a complex change in the organization. it requires an investment of time and other resources. this kind of changes may carry economic losses in the company in the short term (mellert et al., 2015). this is also the case when implementing cmmi (shih et al., 2013). this statement has been validated by our empirical study; however, we do not consider six year is a short time period. we were not able to validate our research hypothesis 1 “the performance of a firm is likely to increase after the process of cmmi implementation”. we found a negative relationship between the use of cmmi and the performance of the firm. we neither could validate research hypothesis 2 “the performance of a firm is likely to decline during the process of cmmi implementation” since lag cmmi and lag2 cmmi variables were not significant. to be able to validate this second hypothesis lag cmmi and lag 2 cmmi should have been positively correlated to firm performance since these variables show if the firm has just implemented cmmi, it has been implemented in the company for one year (lag cmmi) or for two years or more (lag2 cmmi). these results should be understood as a preliminary study and they must be put in the right context. the time period under study belongs to the economic crisis, which had a special impact in spain. moreover, the number of companies that use cmmi is relatively small compared to the total sample. we know the results of the cmmi appraisal; however, we do not know the commitment of the company with cmmi. we also need to take into account that the firms with cmmi grew significantly during the time period under study. so, this growth may also be a cause for its economic and financial recession. to drive general conclusions, this study should be replicated in other countries and during longer time periods. moreover, other variables, such as the commitment of the company to cmmi or the reason for implemented cmmi should be included. acknowledgements the authors want to acknowledge the financial support of bpmsat, banco santander and university of valladolid grant and the following projects: (1) computational models for industrial management (cm4im) project, funded by the valladolid university general foundation and (2) eco201678128-p project, funded by mineco. references albrecht, j. c., pang, k. (2014). linking the benefits of project management maturity to project complexity. international journal of managing projects in business, 7(2), 285-301. https://doi.org/10.1108/ijmpb-08-2013-0040 andreeva, t., kianto, a. (2012). does knowledge management really matter? linking knowledge management practices, competitiveness, and economic performance. journal of knowledge management, 16, 617-636. https://doi.org/10.1108/13673271211246185 augier, m., teece, d. j. (2009). dynamic capabilities and the role of managers in business strategy and economic performance. organization science, 20(2), 410-421. https://doi.org/10.1287/orsc.1090.0424 chien, s., tsai, c. (2012). dynamic capability, knowledge, learning, and firm performance. journal of organizational change management, 25(3), 434-444. https://doi.org/10.1108/09534811211228148 darroch, j. (2005). knowledge management, innovation, and firm performance. journal of knowledge management, 9(3), 101-115. https://doi.org/10.1108/13673270510602809 dayan, r., evans, s. (2006). km your way to cmmi. journal of knowledge management, 10(1), 69-80. https://doi.org/10.1108/13673270610650111 int. j. prod. manag. eng. (2018) 6(2), 57-64creative commons attribution-noncommercial-noderivatives 4.0 international does cmmi implementation affect the performance of the firm? an evaluation from a dynamic capabilities approach 63 https://doi.org/10.1108/ijmpb-08-2013-0040 https://doi.org/10.1108/13673271211246185 https://doi.org/10.1287/orsc.1090.0424 https://doi.org/10.1108/09534811211228148 https://doi.org/10.1108/13673270510602809 https://doi.org/10.1108/13673270610650111 http://creativecommons.org/licenses/by-nc-nd/4.0/ de oliveira moraes, r., barbin laurindo, f. j. (2013). maturity and performance in information technology project management. journal of technology management & innovation, 8, 5-6. https://doi.org/10.4067/s0718-27242013000300003 eisenhardt, k. m., martin, j. a. (2000). dynamic capabilities: what are they? strategic management journal, 21, 1105-1121. https://doi.org/10.1002/1097-0266(200010/11)21:10/11%3c1105::aid-smj133%3e3.0.co;2-e gibson, d., goldenson, d., kost, k. (2006). performance results of cmmi-based process improvement. goldenson, d., gibson, d. (2003). demonstrating the impact and benefits of cmmi: an update and preliminary results. software engineering institute. https://doi.org/10.21236/ada418481 huang, s. j., han, w. m. (2006). selection priority of process areas based on cmmi continuous representation. information and management, 43, 297-307. https://doi.org/10.1016/j.im.2005.08.003 klimko, g. (2001). knowledge management and maturity models: building common understanding. in proceedings of the 2nd european conference on knowledge management (pp. 269-278). bled, slovenia. kotter, j. p. (2007). leading change: why transformation efforts fail. harvard business school publishing hbr. kruger, c. j. (neels), johnson, r. d. (2011). is there a correlation between knowledge management maturity and organizational performance? vine, 41(3), 265-295. https://doi.org/10.1108/03055721111171618 kuriakose, k., raj, b. (2010). knowledge management maturity models-a morphological analysis. journal of knowledge management practice, 11(3). lee, j.-f., wu, m.-j. (2007). organizational capabilities building through cmmi: the case of taiwan software industry. journal of the chinese institute of industrial engineers, 24(4), 327-339. https://doi.org/10.1080/10170660709509048 mellert, l. d., scherbaum, c., oliveira, j., wilke, b. (2015). examining the relationship between organizational change and financial loss. journal of organizational change management, 28(1), 59-71. https://doi.org/10.1108/jocm-11-2013-0236 mullaly, m. (2014). if maturity is the answer, then exactly what was the question? international journal of managing projects in business, 7(2), 169-185. https://doi.org/10.1108/ijmpb-09-2013-0047 nenni, m. e., arnone, v., boccardelli, p., napolitano, i. (2014). how to increase the value of the project management maturity model as a business-oriented framework. international journal of engineering business management, 6(1), 1. https://doi.org/10.5772/58292 niazi, m., wilson, d., zowghi, d. (2005). a maturity model for the implementation of software process improvement: an empirical study. journal of systems and software, 74(2005), 155-172. https://doi.org/10.1016/j.jss.2003.10.017 nielsen, paarup, a., paarup, a. (2006). understanding dynamic capabilities through knowledge management. journal of knowledge management, 10, 59-71. https://doi.org/10.1108/13673270610679363 reifer, d. j. (2000). the cmmi: it’s formidable. journal of systems and software, 50, 97-98. https://doi.org/10.1016/s0164-1212(99)00119-3 shih, s.-p., shaw, r.-s., fu, t.-y., cheng, c.-p. (2013). a systematic study of change management during cmmi implementation: a modified activity theory perspective. project management journal, 44(4), 84-100. https://doi.org/10.1002/pmj.21358 staples, m., niazi, m. (2008). systematic review of organizational motivations for adopting cmm-based spi. information and software technology, 50, 605-620. https://doi.org/10.1016/j.infsof.2007.07.003 sun, y., liu, x. (2010). business-oriented software process improvement based on cmmi using qfd. information and software technology, 52(1), 79-91. https://doi.org/10.1016/j.infsof.2009.08.003 tanriverdi, h. (2005). information technology relatedness knowledge management capability, and performance of multibusiness firms. mis quarterly, 29, 311-334. https://doi.org/10.2307/25148681 teece, d. (2014). the foundations of enterprise performance: dynamic and ordinary capabilities in an (economic) theory of firms. academy of management perspectives, 28(4), 328-352. https://doi.org/10.5465/amp.2013.0116 teece, d., pisano, g., shuen, a. (1997). dynamic capabilities and strategic management. strategic management journal, 18(7), 509-533. https://doi.org/10.1002/(sici)1097-0266(199708)18:7%3c509::aid-smj882%3e3.0.co;2-z tseng, s.-m., lee, p.-s. (2014). the effect of knowledge management capability and dynamic capability on organizational performance. journal of enterprise information management, 27(2), 158-179. https://doi.org/10.1108/jeim-05-2012-0025 wu, q., he, q., duan, y. (2013). explicating dynamic capabilities for corporate sustainability. euromed journal of business, 8(3), 255-272. https://doi.org/http://dx.doi.org/10.1108/emjb-05-2013-0025 young, m., young, r., romero zapata, j. (2014). project, programme and portfolio maturity: a case study of australian federal government. international journal of managing projects in business, 7(2), 215-230. https://doi.org/10.1108/ijmpb-08-2013-0034 zack, m., mckeen, j., singh, s. (2009). knowledge management and organizational performance: an exploratory analysis. journal of knowledge management, 13(6), 392-409. https://doi.org/10.1108/13673270910997088 zheng, s., zhang, w., du, j. (2011). knowledge-based dynamic capabilities and innovation in networked environments. journal of knowledge management, 15, 1035-1051. https://doi.org/10.1108/13673271111179352 int. j. prod. manag. eng. (2018) 6(2), 57-64 creative commons attribution-noncommercial-noderivatives 4.0 international ajenjo. e., martín-cruz, n., ruiz-martin, c. and lópez-paredes, a. 64 https://doi.org/10.4067/s0718-27242013000300003 https://doi.org/10.1002/1097-0266(200010/11)21:10/11%3c1105::aid-smj133%3e3.0.co;2-e https://doi.org/10.21236/ada418481 https://doi.org/10.1016/j.im.2005.08.003 https://doi.org/10.1108/03055721111171618 https://doi.org/10.1080/10170660709509048 https://doi.org/10.1108/jocm-11-2013-0236 https://doi.org/10.1108/ijmpb-09-2013-0047 https://doi.org/10.5772/58292 https://doi.org/10.1016/j.jss.2003.10.017 https://doi.org/10.1108/13673270610679363 https://doi.org/10.1016/s0164-1212(99)00119-3 https://doi.org/10.1002/pmj.21358 https://doi.org/10.1016/j.infsof.2007.07.003 https://doi.org/10.1016/j.infsof.2009.08.003 https://doi.org/10.2307/25148681 https://doi.org/10.5465/amp.2013.0116 https://doi.org/10.1002/(sici)1097-0266(199708)18:7%3c509::aid-smj882%3e3.0.co;2-z https://doi.org/10.1108/jeim-05-2012-0025 https://doi.org/http://dx.doi.org/10.1108/emjb-05-2013-0025 https://doi.org/10.1108/ijmpb-08-2013-0034 https://doi.org/10.1108/13673270910997088 https://doi.org/10.1108/13673271111179352 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.2096 received 2014-01-21 accepted 2014-06-12 ism approach to model offshore outsourcing risks sunand kumara,i, rajiv kumar sharmaa,ii and prashant chauhanb,iii a national institute of technology, department of mechanical enginering, hamirpur, himachal pradesh, india. i susand@nith.ac.in ii rksnithmr@gmail.com b jss academy of technical education, noida, uttar pradesh, india. iii pchauhan33@rediffmail.com abstract: in an effort to achieve a competitive advantage via cost reductions and improved market responsiveness, organizations are increasingly employing offshore outsourcing as a major component of their supply chain strategies. but as evident from literature number of risks such as political risk, risk due to cultural differences, compliance and regulatory risk, opportunistic risk and organization structural risk, which adversely affect the performance of offshore outsourcing in a supply chain network. this also leads to dissatisfaction among different stake holders. the main objective of this paper is to identify and understand the mutual interaction among various risks which affect the performance of offshore outsourcing. to this effect, authors have identified various risks through extant review of literature. from this information, an integrated model using interpretive structural modelling (ism) for risks affecting offshore outsourcing is developed and the structural relationships between these risks are modeled. further, micmac analysis is done to analyze the driving power and dependency of risks which shall be helpful to managers to identify and classify important criterions and to reveal the direct and indirect effects of each criterion on offshore outsourcing. results show that political risk and risk due to cultural differences act as strong drivers. key words: offshore outsourcing, service, risk, interpretive structural modelling, supply chain. 1. introduction in an effort to achieve a competitive advantage via cost reductions and improved market responsiveness, organizations are increasingly employing outsourcing as a major component of their supply chain strategies (lockamy & mccormack, 2010). outsourcing refers to the practice of a firm entrusting to an external entity the performance of an activity that was earlier performed in-house. the outsourced activity could either be the manufacturing of a good or the performance of a service, outsourcing to third party firms based in other countries is commonly referred to as offshore outsourcing (varadarajan, 2009). offshore-outsourcing entails that the service being conducted by sub-contractors in other countries, who are not employees of the organization (honeycutt et al., 2012). this phenomenon has gained increased importance and attention in both theory as well as practice and has been coined ‘‘the next wave of globalization’’ (dossani and kenney, 2007). in actual terms, offshore outsourcing is more risky than domestic outsourcing, given the lack of vendor’s information, managerial difficulties, political or economical uncertainty, and the cost of knowledge transfer in a culturally different environment, and further adding the costs of stolen intellectual property, the challenge become greater (jiang et al., 2007). offshoring is, after all, an inherently risky business due to the complexity of achieving ‘‘suitable management oversight’’ and control from a distance (wright, 2005). due to increasing globalization and technological discontinuities, firms strive to develop new product capabilities and flexibilities by engaging in outsourcing activities and adopting modular systems. however, these strategies contain risks of opportunistic expropriation of tacit knowledge and costs related to monitoring sourcing partners who are geographically and culturally distant (harmancioglu, 2009). according to fel & griette, (2012) many risks present in offshoring outsourcing including natural and political risks of disruptive events, as well as intellectual-property risk and environment, including government support, business environment, local culture and accessibility. 101int. j. prod. manag. eng. (2014) 2(2), 101-111creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2014.2096 http://creativecommons.org/licenses/by-nc-nd/4.0/ during the last decade, enormous research has been done in the offshore outsourcing specifically for risk mitigation e.g. aron et al. (2005) and ellram et al. (2008) in their papers, utilized the framework of transaction cost economics to develop an understanding of how firms manage the costs and risks of offshore outsourcing of professional services. hertah & kishore, (2009) discussed the applicability and potential of balanced score card method to effectively implement an outsourcing strategy and reduce the risks. stringfellow et al. (2008) combined existing service operations theory with insights from the literature on communications and culture to present a new conceptual framework to find out cost drivers related to risk due to cultural differences. youngdahl & ramaswamy, (2008) presented two complementary conceptual models that help to shed light on the complexities of offshoring service and knowledge work related to operational risk. jensen, (2012) used activity-based approach to the study of the offshore outsourcing of high-value, advanced services and presented the theoretical framework to integrate resource-based view for analyzing the risks. youngdahl et al., (2010) developed conceptual model that links economic development, national cultural predispositions, and the future of offshoring service and knowledge functions. tjader et al., (2013) combined the analytic network process and the balanced score card approach to build a cohesive decision model for determining firm level it outsourcing strategy and further examined the robustness of the model through sensitivity analysis. chou & chou, (2009) identified an information systems outsourcing life cycle through three project related periods: precontract phase, contract phase, and post-contract phase. also, various risk factors associated with each phase of the information system outsourcing practice have been identified and examined. mathew & chen, (2013) focused on three major modes of relational norms: norm of flexibility, norm of solidarity and norm of information exchange for achieving offshore software development success, thus mitigating the risks involved. stratman, (2008) used transaction cost theory and operations management models of service process to identify challenges to the effective offshoring of service processes. doh, (2005) suggested that international labour and environmental standards and corporate codes of conduct could mitigate some of the most intense concerns raised about offshoring but conclude that offshoring is likely to present challenges to societies, corporations, and stakeholders for many decades. cai et al., (2011) presented a theoretic method to control outsourcing risks by designing the incentive and monitoring mechanism of the producer services outsourcing contract. hahn & bunyaratavej, (2010) empirically examined theoretical development of service cultural alignment and investigated the impact of cultural dimensions on the location of service offshoring projects. as evident from above literature studies, number of approaches, models, empirical as well as conceptual has been developed by researchers to study or model the impact of various risks on offshore outsourcing. but very limited research, which examines the relationship between various types offshoring risks is found (aron et al., 2005; wright, 2005; goo & huang, 2008; harmancioglu, 2009; chou & chou, 2009, 2011; youngdahl et al., 2010; datta & roy, 2012). owing to the complex nature of offshore outsourcing because of its interface between cultures, organizations, disciplines, technologies and tacit knowledge of employees, it is very difficult to analyze the inter-relationship among the various risks. in literature, various methods such as ahp, ism and anp are used by authors to examine the interrelationships (see, e.g., shang et al., 2004; wei et al., 2005; raj et al., 2008; subramanian & ramanathan, 2012). authors, in the present study make use of interpretive structural modelling (ism), a well established methodology for identifying relationships among specific items, which defines risk. the main objectives of this paper are: 1. to identify risks involved in offshore outsourcing of professional services 2. to establish the relationship between these identified risks using interpretive structural modeling 3. to propose a structural model for risks of offshore outsourcing 4. to classify the identified risks into various categories using micmac analysis the remainder of this paper is organized as follows. after introduction in section 1, section 2 presents ism methodology. section 3 presents the literature review with respect to nine types of potential risks. section 4 presents the details of ism approach to model offshore outsourcing risks section 5 presents the discussions. conclusion and further research directions are presented in section 6. 102 int. j. prod. manag. eng. (2014) 2(2), 101-111 creative commons attribution-noncommercial-noderivatives 4.0 international kumar, s., kumar sharmab, r. & chauhanc, p. http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. an overview of ism approach interpretive structural modelling (ism) is an interactive learning process whereby a set of different indirectly and directly related elements are structured into a comprehensive systemic model. the presence of indirectly or directly related elements complicates the structure of the system which may or may not be articulated in a clear fashion. it becomes difficult to deal with such a system where structure is not clearly defined. hence, a methodology needs to be developed which aids in the identification of a structure within a system, interpretive structural modelling is such a methodology. several examples of the use of ism have appeared in the literature. there are two basic concepts which are essential to understand the ism methodology. one is the concept of reachability and the other is that of transitivity. common terminology used to represent relationship between elements is discussed as under: four symbols used to denote the direction of relationship between the elements are given below (i and j) v: → element i will reaches element j a: → element j will reaches element i x: → elements i and j will help to alleviate each other o: → elements i and j will not related to each other this information is represented in the form of binary matrix and it is called initial reachability matrix. if an element i reaches element j, then the entry in the cell (i, j) of the reachability matrix is 1 and if element i does not reach element j, then entry in the cell (i, j) of the reachability matrix is 0. if element i reaches to element j and element j reaches to element k, then transitivity implies element i reaches to element k. the steps involved in ism approach are shown in figure 1. 3. identification of risks related to offshore outsourcing recent concerns over the intellectual property protection in software production are an early indication of what could be a growing phenomenon (doh, 2005). as with all outsourcing contracts, the threat of intellectual property (ip) infringement is very serious and needs to be taken into account prior to signing a contract (currie et al., 2008). rao (2004) explored issues about doing business overseas, and discussed factors from the availability of telecommunications infrastructure to cultural differences and language barriers, as well as legal and regulatory challenges of conducting business elsewhere. regulatory uncertainty in developing countries is generally considered a risk that raises transaction costs (stratman, 2008). hahn & bunyaratavej, 2010 considered the possibility that firms may be averse to countries that have higher levels of political risk as firms may prefer to do business in more stable environments. outsourcing may be subject to political risk both from home country protectionist pressures and the traditional risk of operating in a foreign country (sambharya & rasheed, 2012). figure 1. flow diagram for ism approach. 103int. j. prod. manag. eng. (2014) 2(2), 101-111creative commons attribution-noncommercial-noderivatives 4.0 international ism approach to model offshore outsourcing risks http://creativecommons.org/licenses/by-nc-nd/4.0/ when the service process is non-standardized, requires complex judgment and has reciprocal interdependence among steps and sequences, the reliability and assurance of service quality are at risk (stringfellow et al., 2008). it happens with knowledge process work or r & d related work of outsourcing. mitigation of operational risk is very critical in the process because the cost of failure can be high (youngdahl & ramaswamy, 2008). language, cultural and time zone differences are obvious problems that complicate offshore process management (stratman, 2008). when outsourcing is carried out with an offshore vendor it poses additional risks such as cultural differences, language barriers, and geographical and time zone related barriers (hertah & kishore, 2009). risks that result from opportunistic behavior of one or both parties i.e buyer and supplier (aron et al., 2005). this opportunistic behavior may result in failure of cooperative innovation and it includes shirking, cheating, and distorting information (li et al., 2008). a strong commitment to and identification with the project at all levels of the organization, involving forbearance and avoidance of opportunistic behavior would be essential (søderberg et al., 2013). some structural risk arises because vendors can stop investing in training or employ people who aren’t as qualified as the agents they presented during negotiations (aron & singh, 2005). yet another risk is the supplier making changes to processes, technologies, and procedures without properly informing the buying firm (ellram et al., 2008). the mid-contract sag occurs after the supplier has dispensed all their transformational levers (consolidation, standardization, reduced headcount, better technology, and better processes) (lacity et al., 2008). few cost related risks, such as unexpected transition and management costs, switching costs, costly contractual amendments, disputes and litigation (chou & chou, 2009). there is a potential risk of incurring transition costs, and project and vendor management costs, which can more than offset the savings from outsourcing, resulting in a net loss (tjader, et al., 2013). according to carmel & agarwal (2002), the use of offshore resources creates uncertainty and turmoil among internal staff.the loss of critical knowledge is seen as the greatest source of workforce-related risk around outsourcing (pfannenstein & tsai, 2004). after the first few years of a large outsourcing contract, the client’s knowledge retention can dramatically erode through attrition (lacity et al., 2008). based upon the extant review of literature, authors grouped them under nine categories presented in table 1. 4. ism approach to modelling the various steps involved in ism technique used to model the structural relationship among identified risks are discussed in the following paragraphs: 4.1. establishing the contextual relationship among variables (risks) after identifying and enlisting the 9 risks, the next step is to analyze the risks. for this purpose, a contextual relationship of ‘reaches to’ type is selected. this means that one risk reaches to another risk. based on this principle, a contextual relationship is developed. some experts, from various organizations related to outsourcing were consulted to assist in developing the contextual relationships between the risks. keeping in mind the contextual relationship for each risk, the existence of a relation between any two risks (i and j) and the associated direction of this relation is decided. to analyze the risk for the development of the structural self-interaction matrix (ssim), the following four symbols are used to denote the direction of the relationship between the risks (i and j). 1. v is used for the relation from risk i to risk j (i.e. if risk i reaches risk j). 2. a is used for the relation from risk j to risk i (i.e. if risk j reaches risk i). 3. x is used for both direction relations (i.e. if risks i and j influence each other). 4. o is used for no relation between two risks (i.e. if risks i and j are unrelated). 4.2. development of a structural self interaction matrix (ssim) based on the contextual relationship between the risks, the ssim was developed. to achieve consensus, the ssim was discussed in a group of experts. based on their responses, the ssim was finalized and is presented in table. 2. 104 int. j. prod. manag. eng. (2014) 2(2), 101-111 creative commons attribution-noncommercial-noderivatives 4.0 international kumar, s., kumar sharmab, r. & chauhanc, p. http://creativecommons.org/licenses/by-nc-nd/4.0/ table 1. identification of risks related to offshore outsourcing. s. no. name of risk definition references 1 intellectual property risk intellectual property (ip) risk is where the vendors or their staff will misappropriate ip even where the contract stipulates that ip rights solely belong to the client. (fleming & sorenson, 2001), (doh, 2005), (frank, 2005), (jiang et al., 2007), (li et al., 2008), (currie et al., 2008), (lacity et al., 2008), (raiborn et al., 2009), (tjader et al., 2010), (chou & chou, 2011), (fel & griette, 2012), (nassimbeni et al., 2012), (mathew & chen, 2013) 2 compliance and regulatory risk compliance and regulatory risk is where an outsourcing contract inhibits the client from complying with a compliance or regulatory framework. (rao, 2004), (graham, 2004), (currie et al., 2008), (stratman, 2008), (aron et al., 2008), (forte, 2009), (luo et al., 2010), (chou & chou, 2011), (benlian & hess, 2011), (bachlechner et al., 2013) 3 political risk these are risks associated with different regions with their different socio-political systems and different historical contexts. (prasad & babbar, 2000), (aron et al., 2005), (wright, 2005), (jiang et al., 2007), (ellram et al., 2008), (stratman, 2008), (currie et al., 2008), (nakatsu & iacovou, 2009), (tjader et al., 2010), (hahn & bunyaratavej, 2010), (cappelli, 2011), (fel & griette, 2012), (sambharya & rasheed, 2012) 4 operational risk operational risk is where services will not be delivered as expected or that there will be failure in infrastructure or technology that will impede continuity of service to customers. (quélin & duhamel, 2003), , (aron & singh, 2005), (aron et al., 2005), (currie et al., 2008), (aron et al., 2008), (youngdahl & ramaswamy, 2008), (ellram et al. 2008), (goo & huang, 2008) (stringfellow et al., 2008), (chou & chou, 2009), (hertah & kishore, 2009), (raiborn et al., 2009), (krishnamurthy et al., 2009), (bachlechner et al., 2013) 5 risk due to cultural differences cultural differences relate to deep-seated values and are also often more difficult to observe than language differences, so they may go unnoticed. (prasad & babbar, 2000), (rao, 2004), (ellram et al. 2008), (stringfellow et al., 2008),(youngdahl & ramaswamy, 2008), (stratman, 2008), (nakatsu & iacovou, 2009), (hertah & kishore, 2009), (tjader et al., 2010), (youngdahl et al., 2010), (honeycutt et al., 2012), (fel & griette, 2012) 6 opportunistic risk opportunistic risk is related with behavior of service provider which may result in failure of cooperative innovation and it includes shirking, cheating, and distorting information. (aron et al., 2005), (li et al., 2008), (stratman, 2008), (aron et al., 2008), (goo & huang, 2008), (mao et al., 2008), (chou & chou, 2009), (harmancioglu, 2009), (raiborn et al., 2009), (tjader et al., 2010), (lacity, et al., 2011), (cai et al., 2011), (datta & roy, 2012), (nassimbeni et al., 2012), (mathew & chen, 2013), (søderberg et al., 2013) 7 organization structural risk the offshore service provider can stop investing in training or employ people who aren’t as qualified as the agents they presented during negotiations. (henderson & clark, 1900), (quélin & duhamel, 2003), (aron & singh, 2005), (ellram et al. 2008), (shekhar, 2008), (lacity et al., 2008) 8 financial risk risks, such as unexpected transition and management costs, switching costs, costly contractual amendments, disputes and litigation. (overby, 2003), (pfannenstein & tsai, 2004), (ellram et al. 2008), (lacity et al., 2008), (chou & chou, 2009), (hertah & kishore, 2009), (fel & griette, 2012), (sambharya & rasheed, 2012), (tjader, et al., 2013) 9 loss of core professionals the loss of critical knowledge is seen as the greatest source of workforce-related offshore outsourcing risk. (carmel & agarwal, 2002), (quélin & duhamel, 2003), (pfannenstein & tsai, 2004), (aron et al., 2005), ( ellram et al. 2008), (lacity et al., 2008), (chou & chou, 2009), (hertah & kishore, 2009), (jensen, 2012), (tayauova, 2012) 105int. j. prod. manag. eng. (2014) 2(2), 101-111creative commons attribution-noncommercial-noderivatives 4.0 international ism approach to model offshore outsourcing risks http://creativecommons.org/licenses/by-nc-nd/4.0/ table 2. ssim (structural self interaction matrix). s. no.variables(risks) 2 3 4 5 6 7 8 9 1 intellectual property risk o o x o a o x x 2 compliance and regulatory risk a v a x x v o 3 political risk v v o o v o 4 operational risk a a a v o 5 risk due to cultural differences o v v o 6 opportunistic risk v v o 7 organisation structural risk v o 8 financial risk o 9 loss of core professionals 4.3. development of the initial reachability matrix (irm) the ssim was converted into a binary matrix, called the initial reachability matrix by substituting v, a, x and o with 1 and 0 as per the case and is presented in table 3. the substitution of 1s and 0s are as per the following rules: 1. if the (i, j) entry in the ssim is v, the (i, j) entry in the initial reachability matrix becomes 1 and the (j, i) entry becomes 0. 2. if the (i, j) entry in the ssim is a, the (i, j) entry in the initial reachability matrix becomes 0 and the (j, i) entry becomes 1. 3. if the (i, j) entry in the ssim is x, the (i, j) entry in the initial reachability matrix becomes 1 and the (j, i) entry also becomes 1. 4. if the (i, j) entry in the ssim is o, the (i, j) entry in the initial reachability matrix becomes 0 and the (j, i) entry also becomes 0. table 3. irm (initial reachability matrix). s. no. variables(risks) 1 2 3 4 5 6 7 8 9 1 intellectual property risk 1 0 0 1 0 0 0 1 1 2 compliance and regulatory risk 0 1 0 1 0 1 1 1 0 3 political risk 0 1 1 1 1 0 0 1 0 4 operational risk 1 0 0 1 0 0 0 1 0 5 risk due to cultural differences 0 1 0 1 1 0 1 1 0 6 opportunistic risk 1 1 0 1 0 1 1 1 0 7 organization structural risk 0 1 0 1 0 0 1 1 0 8 financial risk 1 0 0 0 0 0 0 1 0 9 loss of core professionals 1 0 0 0 0 0 0 0 1 4.4. development of the final reachability matrix (frm) the initial reachability matrix was converted into a final reachability matrix (frm) and is presented in table 4. it considers transitivity concept of ism methodology. table 5. shows final reachability matrix with driving power and dependence. table 4. frm (final reachability matrix). s. no. variables(risks) 1 2 3 4 5 6 7 8 9 1 intellectual property risk 1 0 0 1 0 0 0 1 1 2 compliance and regulatory risk 1 1 0 1 0 1 1 1 1 3 political risk 1 1 1 1 1 1 1 1 1 4 operational risk 1 0 0 1 0 0 0 1 1 5 risk due to cultural differences 1 1 0 1 1 1 1 1 1 6 opportunistic risk 1 1 0 1 0 1 1 1 1 7 organization structural risk 1 1 0 1 0 1 1 1 1 8 financial risk 1 0 0 1 0 0 0 1 1 9 loss of core professionals 1 0 0 1 0 0 0 1 1 table 5. frm (final reachability matrix with driving power and dependence). s. no. variables(risks) 1 2 3 4 5 6 7 8 9 driving power 1 intellectual property risk 1 0 0 1 0 0 0 1 1 4 2 compliance and regulatory risk 1 1 0 1 0 1 1 1 1 7 3 political risk 1 1 1 1 1 1 1 1 1 9 4 operational risk 1 0 0 1 0 0 0 1 1 4 5 risk due to cultural differences 1 1 0 1 1 1 1 1 1 8 6 opportunistic risk 1 1 0 1 0 1 1 1 1 7 7 organization structural risk 1 1 0 1 0 1 1 1 1 7 8 financial risk 1 0 0 1 0 0 0 1 1 4 9 loss of core professionals 1 0 0 1 0 0 0 1 1 4 dependence 9 5 1 9 2 5 5 9 9 4.5. partitioning the final reachability matrix once the reachability matrix has been created, it must be processed to extract the structural model. the reachability set consists of the risk (i) itself and the other risks which are reachable from that particular risk (i). for every column which contains 1 in the row of the considered risk (i), the risk that column represents is included in the reachability set. 106 int. j. prod. manag. eng. (2014) 2(2), 101-111 creative commons attribution-noncommercial-noderivatives 4.0 international kumar, s., kumar sharmab, r. & chauhanc, p. http://creativecommons.org/licenses/by-nc-nd/4.0/ similarly, the antecedent set consists of the risk (i) itself and the other risks which may reach the risk (i). for every row which contains 1 in the column of considered risk (i), the risk that row represents is included in the antecedent set. after finding the reachability and antecedent sets for each risk, the intersection of these sets is derived for all the risks and levels. the variables for which the reachability and the intersection are the same are given the top level in the ism hierarchy. this procedure is continued till all levels of the structure are identified. these identified levels help in the development of the model. in the present case the level identification process for the 9 risks was completed in four iterations and is shown in tables 6 -9. further in table 10, ism based levels of variables or risks are shown. table 6. first iteration. variablesreachability antecedent intersection level 1 1,4,8,9 1,2,3,4,5,7, 8,9 1,4,8,9 i 2 1,2,4,6,7,8,9 2,3,5,6,7 2,6,7 3 1,2,3,4,5,6, 7,8,9 3 3 4 1,4,8,9 1,2,3,4,5,6,7, 8,9 1,4,8,9 i 5 1,4,5,6,7,8,9 3,5, 5 6 1,2,4,6,7,8,9 2,3,5,6,7 2,6,7 7 1,2,4,6,7,8,9 2,3,5,6,7 2,6,7 8 1,4,8,9 1,2,3,4,5,6,7, 8,9 1,4,8,9 i 9 1,4,8,9 1,2,3,4,5,6,7, 8,9 1,4,8,9 i table 7. second iteration. variables reachability antecedent intersection level 2 2,6,7 2,3,5,6,7 2,6,7 ii 3 2,3,5,6,7 3 3 5 2,5,6,7 3,5 5 6 2,6,7 2,3,5,6,7 2,6,7 ii 7 2,6,7 2,3,5,6,7 2,6,7 ii table 8. third iteration. variables reachability antecedent intersection level 3 3,5 3 3 5 5 3,5 5 iii table 9. fourth iteration. variable reachability antecedent intersection level 3 3 3 3 iv table 10. ism based levels of variables. s. no. variables(risks) levels 1 intellectual property risk i 2 compliance and regulatory risk ii 3 political risk iv 4 operational risk i 5 risk due to cultural differences iii 6 opportunistic risk ii 7 organization structural risk ii 8 financial risk i 9 loss of core professionals i 4.6. development of conical matrix a conical matrix is developed by clubbing together risks in the same level, across the rows and columns of the final reachability matrix is presented in table 11. the driving power of a risk is derived by adding the number of ones in the rows, and the dependency is derived by adding up numbers ones in the columns. table 11. conical matrix. risks ( number of risk) 1 4 8 9 2 6 7 5 3 driving power intellectual property risk (1) 1 1 1 1 0 0 0 0 0 4 operational risk (4) 1 1 1 1 0 0 0 0 0 4 financial risk (8) 1 1 1 1 0 0 0 0 0 4 loss of core professionals (9) 1 1 1 1 0 0 0 0 0 4 compliance and regulatory risk (2) 1 1 1 1 1 1 1 0 0 7 opportunistic risk (6) 1 1 1 1 1 1 1 0 0 7 organization structural risk (7) 1 1 1 1 1 1 1 0 0 7 risk due to cultural differences (5) 1 1 1 1 1 1 1 1 0 8 political risk (3) 1 1 1 1 1 1 1 1 1 9 dependence 9 9 9 9 5 5 5 2 1 4.7. development of diagraph based on the conical matrix an initial diagraph including transitivity links is drawn. this is drawn by the nodes and the lines of edges. after removing the transitivity, a final diagraph is drawn (figure 2). if there is a relationship between the risks j and i this is shown by an arrow which points from risk i to risk j. 107int. j. prod. manag. eng. (2014) 2(2), 101-111creative commons attribution-noncommercial-noderivatives 4.0 international ism approach to model offshore outsourcing risks http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.8. development of ism model the diagraph is converted into an ism model by replacing the nodes with name of risks as shown in figure 3. 4.9. micmac analysis matrice d’impacts croises-multipication appliqu´ean classment (cross-impact matrix multiplication applied to classification) is abbreviated as micmac. the micmac principle is based on the multiplication properties of matrices. the purpose of a micmac analysis is to analyze the driver power and dependency of the variables (raj et al., 2008; govindan et al., 2012). this is done to identify the key risks that drive the system. based on their driver power and dependency, the risks, in this present case, have been classified into four categories as follow: 1 2 6 5 3 7 4 8 9 figure 2. diagraph showing the level of offshore outsourcing risks figure 3. ism model showing the level of offshore outsourcing risks. (1) autonomous risks: these risks have weak driver power and weak dependence. they are relatively disconnected from the system, with which they have few strong links. (2) linkage risks: these have strong driver power as well as strong dependence. they are also unstable. any action on them has an effect on others and also a feedback effect on themselves. (3) dependent risks: this category includes those risks which have strong dependence power but weak driver power. (4) independent risks: these have strong driver power but weak dependence power. it is generally observed that a risk with a very strong driver power, called a ‘key risk’ falls into the category of independent or linkage risks. figure 4, presents the results of micmac analysis. figure 4. driving power and dependence diagram. 5. findings and discussion the objective of this research was to identify and analyze the risks that significantly affect success of offshore outsourcing so that managers may effectively deal with these risks. in this research, an ism-based model was developed to analyze the relationship among different risks of offshore outsourcing so that management can get an insight into these risks and understand their relative importance and interactions. some of the valuable findings from the study are as under: (a.) from the driving power and dependence diagram (figure 4), it is observed that two risks, 108 int. j. prod. manag. eng. (2014) 2(2), 101-111 creative commons attribution-noncommercial-noderivatives 4.0 international kumar, s., kumar sharmab, r. & chauhanc, p. http://creativecommons.org/licenses/by-nc-nd/4.0/ namely political risk (3) and risk due to cultural differences (5) have strong driving power and are less dependent on other risks. therefore, these all independent variables are strong drivers and may be treated as the root causes for all risks, so managers need to address these risks as a priority for success of offshore outsourcing. (b.) from the driving power and dependence diagram (figure 4) it is observed that intellectual property risk (1), operational risk (4), financial risk (8) and loss of core professionals (9) are weak drivers but strongly dependent on the other risks. these four risks are at the top of the ism hierarchy, therefore are considered as the most important risks. decision taking authorities should, therefore, accord high priority in resolving these risks for achieving success of offshore outsourcing and should understand the dependence of these risks on other risks. operational risks are caused by the breakdown in operations at the vendor location. these risks are not caused by deliberate actions by the vendor or by unethical behavior of the vendor. rather, they are a by-product of the complexity of operations, the geographic separation between client and vendor, the cultural gap between the environments of the client and the vendor, or the limitations of the communications and transmission systems between the two (aron et al., 2005; krishnamurthy et al., 2009). 6. conclusion based upon the extant review of literature, authors identified 9 key risks that could affect performance of offshore outsourcing. further to examine the complex relationship between them, an ism model and micmac approach was used. the findings provide important classification of risks under four categories i.e. independent (risk due to cultural differences and political risk), linkage (compliance and regulatory risk, opportunistic risk and organization structural risk), dependent (intellectual property risk, operational risk, financial risk and loss of core professionals) and autonomous (no risk in this case). the results obtained with the help of ism are being used to gain insights into the driver and dependence power of risks related to offshore outsourcing. future research may be directed towards confirmatory approach to data analysis supported by structural equation modelling (sem) and inclusion of more risks which affect the process of offshore outsourcing. references ambroggi, m. de., & trucco, p. (2011). modelling and assessment of dependent performance shaping factors through analytic network process. reliability engineering and system safety, 96(7): 849-860. doi:10.1016/j.ress.2011.03.004 aron, r., bandyopadhyay, s., jayanty, s., & pathak, p. (2008). monitoring process quality in off-shore outsourcing: a model and findings from multi-country survey. journal of operations management, 26(2): 303-321. doi:10.1016/j.jom.2007.02.014 aron, r., clemons, e. k., & reddi, s. (2005). just right outsourcing: understanding and managing risk. journal of management information systems, 22(2): 37-55. aron, r., & singh, j. v. (2005). getting offshoring right. harvard business review, dec 2005. bachlechner, d., thalmann, s., & maier, r. (2013). security and compliance challenges in complex it outsourcing arrangements: a multistakeholder perspective. computers & security, 40: 38-59. doi:10.1016/j.cose.2013.11.002 benlian, a., & hess, t. (2011). opportunities and risks of software-as-a-service: findings from a survey of it executives. decision support systems, 52(1): 232-246. doi:10.1016/j.dss.2011.07.007 cai, s., ci, k., & zou, b. (2011). producer services outsourcing risk control based on outsourcing contract design: industrial engineering perspective. systems engineering procedia, 2: 308-315. doi:10.1016/j.sepro.2011.10.043 cappelli, p. (2011). hr sourcing decisions and risk management. organizational dynamics, 40(4): 310-316. doi:10.1016/j. orgdyn.2011.07.008 carmel, e., agarwal, r., (2002). the maturation of offshore sourcing of information technology work. mis quarterly executive vol. 1 no. 2. chou, d. c., & chou, a. y. (2009). information systems outsourcing life cycle and risks analysis. computer standards & interfaces, 31(5): 1036-1043. doi:10.1016/j.csi.2008.09.032 chou, d. c., & chou, a. y. (2011). innovation outsourcing: risks and quality issues. computer standards & interfaces, 33(3): 350-356. doi:10.1016/j.csi.2010.10.001 currie, w. l., michell, v., & abanishe, o. (2008). knowledge process outsourcing in financial services: the vendor perspective. european management journal, 26(2): 94-104. doi:10.1016/j.emj.2007.11.002 109int. j. prod. manag. eng. (2014) 2(2), 101-111creative commons attribution-noncommercial-noderivatives 4.0 international ism approach to model offshore outsourcing risks http://dx.doi.org/10.1016/j.ress.2011.03.004 http://dx.doi.org/10.1016/j.jom.2007.02.014 http://dx.doi.org/10.1016/j.cose.2013.11.002 http://dx.doi.org/10.1016/j.dss.2011.07.007 http://dx.doi.org/10.1016/j.sepro.2011.10.043 http://dx.doi.org/10.1016/j.orgdyn.2011.07.008 http://dx.doi.org/10.1016/j.orgdyn.2011.07.008 http://dx.doi.org/10.1016/j.csi.2008.09.032 http://dx.doi.org/10.1016/j.csi.2010.10.001 http://dx.doi.org/10.1016/j.emj.2007.11.002 http://creativecommons.org/licenses/by-nc-nd/4.0/ datta, p. p. & roy, r. (2011). incentive issues in performance-based outsourcing contracts in the uk defence industry: a simulation study. production planning & control, 1-16, ifirst. doh, j. p. (2005). offshore outsourcing: implications for international business and strategic management theory and practice. journal of management studies, 42(3): 695-704. doi:10.1111/j.1467-6486.2005.00515.x dossani, r., & kenney, m. (2007).the next wave of globalization: relocating service provision to india. world development, 35(5): 772-791. doi:10.1016/j.worlddev.2006.09.014 ellram, l. m., tate, w. l., & billington, c. (2008). offshore outsourcing of professional services: a transaction cost economics perspective. journal of operation management, 26(2): 148-163. doi:10.1016/j.jom.2007.02.008 fel, f., & griette, e. (2012). an analysis of the offshoring decision process: the influence of the company’s size. procedia social and behavioral sciences, 58, 596-605. doi:10.1016/j.sbspro.2012.09.1037 fleming, l., sorenson, o., (2001). technology as a complex adaptive system: evidence from patent data. research policy, 30(7): 10191039. doi:10.1016/s0048-7333(00)00135-9 forte, d. (2009). state of the art security management. computer fraud & security, 10: 17-18. doi:10.1016/s1361-3723(09)70130-7 frank, s. j. (2005). source out, risk in. ieee spectrum, april 2005. goo, j., & huang, c. d. (2008). facilitating relational governance through service level agreements in it outsourcing: an application of the commitment–trust theory. decision support systems, 46(1): 216-232. doi:10.1016/j.dss.2008.06.005 govindan, k., palaniappan, m., zhu, q., & kannan, d. (2012). analysis of third party reverse logistics provider using interpretive structural modelling. int. j. production economics, 140(1): 204-211. doi:10.1016/j.ijpe.2012.01.043 graham, n. (2004). data protection and business sales – risks and solutions?. computer law & security report, 20(4): 306-309. doi:10.1016/s0267-3649(04)00053-6 hahn, e. d., & bunyaratavej, k. (2010). services cultural alignment in offshoring: the impact of cultural dimensions on offshoring location choices. journal of operations management, 28(3): 186-193. doi:10.1016/j.jom.2009.10.005 harmancioglu, n. (2009). portfolio of controls in outsourcing relationships for global new product development. industrial marketing management, 38(4): 394-403. doi:10.1016/j.indmarman.2009.04.001 henderson, r.m., & clark, k. b. (1990). architectural innovation: the reconfiguration of existing product technologies and the failure of established firms. administrative science quarterly, 35(1), special issue: technology, organizations, and innovation, pp. 9-30. herath, t., & kishore, r. (2009). offshore outsourcing: risks, challenges, and potential solutions. information systems management, 26(4): 312326. doi:10.1080/10580530903245549 honeycutt, e. d. jr., magnini, v. p., & thelen, s. t. (2012). solutions for customer complaints about offshoring and outsourcing services. business horizons, 55(1): 33-42. doi:10.1016/j.bushor.2011.09.001 jensen, p. d. ø. (2012). a passage to india: a dual case study of activities, processes and resources in offshore outsourcing of advanced services. journal of world business, 47(2): 311-326. doi:10.1016/j.jwb.2011.04.018 jiang, b., belohlav, j. a., & young, s. t. (2007). outsourcing impact on manufacturing firms’ value: evidence from japan. journal of operations management, 25(4): 885-900. doi:10.1016/j.jom.2006.12.002 krishnamurthy, k., jegen, d., & brownell, b. (2009). strategic out-tasking: creating ‘‘win–win’’ outsourcing partnerships. information & management, 46(1): 42-51. doi:10.1016/j.im.2008.11.004 lacity, m.c., willcocks, l.p. & rottman, j.w. (2008). global outsourcing of back office services: lessons, trends, and enduring challenges. strategic outsourcing: an international journal, 1(1): 13-34. doi:10.1108/17538290810857457 lacity, m., willcocks, l., & khan, s. (2011). beyond transaction cost economics: towards an endogenous theory of information technology outsourcing. journal of strategic information systems, 20(2): 139-157. doi:10.1016/j.jsis.2011.04.002 li, y., liu, y., li, m., & wu, h. (2008). transformational offshore outsourcing: empirical evidence from alliances in china. journal of operations management, 26(2): 257-274. doi:10.1016/j.jom.2007.02.011 lo, chi-c., chen, & wan-j. (2012). a hybrid information security risk assessment procedure considering interdependences between controls. expert systems with applications, 39(1): 247-257. doi:10.1016/j.eswa.2011.07.015 lockamy iii, a., & mccormack, k. (2010). analyzing risks in supply networks to facilitate outsourcing decisions. international journal of production research, 48(2): 593-611. doi:10.1080/00207540903175152 luo, y., zheng, q., & jayaraman, v. (2010). managing business process outsourcing. organizational dynamics, 39(3): 205-217. doi:10.1016/j.orgdyn.2010.03.005 mao, ji-y., lee, jae-n., & deng, chun-p. (2008). vendors’ perspectives on trust and control in offshore information systems outsourcing. information & management, 45(7): 482-492. doi:10.1016/j.im.2008.07.003 mathew, s. k., & chen, y. (2013). achieving offshore software development success: an empirical analysis of risk mitigation through relational norms. journal of strategic information systems, 22(4): 298-314. doi:10.1016/j.jsis.2013.03.001 110 int. j. prod. manag. eng. (2014) 2(2), 101-111 creative commons attribution-noncommercial-noderivatives 4.0 international kumar, s., kumar sharmab, r. & chauhanc, p. http://dx.doi.org/10.1111/j.1467-6486.2005.00515.x http://dx.doi.org/10.1016/j.worlddev.2006.09.014 http://dx.doi.org/10.1016/j.jom.2007.02.008 http://dx.doi.org/10.1016/j.sbspro.2012.09.1037 http://dx.doi.org/10.1016/s0048-7333(00)00135-9 http://dx.doi.org/10.1016/s1361-3723(09)70130-7 http://dx.doi.org/10.1016/j.dss.2008.06.005 http://dx.doi.org/10.1016/j.ijpe.2012.01.043 http://dx.doi.org/10.1016/s0267-3649(04)00053-6 http://dx.doi.org/10.1016/j.jom.2009.10.005 http://dx.doi.org/10.1016/j.indmarman.2009.04.001 http://dx.doi.org/10.1080/10580530903245549 http://dx.doi.org/10.1016/j.bushor.2011.09.001 http://dx.doi.org/10.1016/j.jwb.2011.04.018 http://dx.doi.org/10.1016/j.jom.2006.12.002 http://dx.doi.org/10.1016/j.im.2008.11.004 http://dx.doi.org/10.1108/17538290810857457 http://dx.doi.org/10.1016/j.jsis.2011.04.002 http://dx.doi.org/10.1016/j.jom.2007.02.011 http://dx.doi.org/10.1016/j.eswa.2011.07.015 http://dx.doi.org/10.1080/00207540903175152 http://dx.doi.org/10.1016/j.orgdyn.2010.03.005 http://dx.doi.org/10.1016/j.im.2008.07.003 http://dx.doi.org/10.1016/j.jsis.2013.03.001 http://creativecommons.org/licenses/by-nc-nd/4.0/ nakatsu, r. t., & iacovou, c. l. (2009). a comparative study of important risk factors involved in offshore and domestic outsourcing of software development projects: a two-panel delphi study. information & management, 46(1): 57-68. doi:10.1016/j.im.2008.11.005 nassimbeni, g., sartor, m., & dus, d. (2012). security risks in service offshoring and outsourcing. industrial management & data systems, 112(3): 405-440. doi:10.1108/02635571211210059 overby, s. (2003). the hidden costs of offshore outsourcing, cio magazine, september 1. pfannenstein, l.l. & tsai, r.j. (2004): offshore outsourcing: current and future effects on american it industry. information systems management, 21(4): 72-80. doi:10.1201/1078/44705.21.4.20040901/84190.9 prasad, s., & babbar, s. (2000): international operations management research. journal of operations management, 18(2): 209-247. doi:10.1016/s0272-6963(99)00015-7 quélin, b., & duhamel, f. (2003). bringing together strategic outsourcing and corporate strategy: outsourcing motives and risks. european management journal, 21(5): 647-661. doi:10.1016/s0263-2373(03)00113-0 raiborn, c. a., butler, j. b., & massoud, m. f. (2009). outsourcing support functions: identifying and managing the good, the bad, and the ugly. business horizons, 52(4): 347-356. doi:10.1016/j.bushor.2009.02.005 raj, t., shankar, r., & suhaib, m. (2008). an ism approach for modelling the enablers of flexible manufacturing system: the case for india. international journal of production research, 46(24): 6883-6912. doi:10.1080/00207540701429926 rao, m.t. (2004). key issues for global it sourcing: country and individual factors. information systems management, 21(3): 16-21. doi:10 .1201/1078/44432.21.3.20040601/82472.3 sambharya, r. b., & rasheed, a. a. (2012). global risk in a changing world: new paradigms and practice. organizational dynamics, 41(4): 308-317. doi:10.1016/j.orgdyn.2012.08.006 shang, j. s., tjader, y., & ding, y. (2004). a unified framework for multi criteria evaluation of transportation projects. ieee transactions on engineering management, 51(3): 300-313. doi:10.1109/tem.2004.830848 shekhar, s. (2008), benchmarking knowledge gaps through role simulations for assessing outsourcing viability. benchmarking: an international journal, 15(3): 225-241. doi:10.1108/14635770810876575 søderberg, a. m., krishna, s., & bjørn, p. (2013). global software development: commitment, trust and cultural sensitivity in strategic partnerships. journal of international management, 19(4): 347-361. doi:10.1016/j.intman.2013.04.004 stratman, j. k. (2008). facilitating offshoring with enterprise technologies: reducing operational friction in the governance and production of services. journal of operations management, 26(2): 275-287. doi:10.1016/j.jom.2007.02.006 stringfellow, a., teagarden, m. b., & nie, w. (2008). invisible costs in offshoring services work. journal of operations management, 26(2): 164-179. doi:10.1016/j.jom.2007.02.009 subramanian, n., & ramanathan, r. (2012). a review of applications of analytic hierarchy process in operations management. int. j. production economics, 138(2): 215-241. doi:10.1016/j.ijpe.2012.03.036 tayauova g. (2012). advantages and disadvantages of outsourcing: analysis of outsourcing practices of kazakhstan banks. procedia social and behavioral sciences, 41:188-195. doi:10.1016/j.sbspro.2012.04.023 tjader, y. c., shang, j. s., & vargas, l. g. (2010). offshore outsourcing decision making: a policy-makers perspective. european journal of operational research, 207(1): 434-444. doi:10.1016/j.ejor.2010.03.042 tjader, y., may, j.h., shang, j., vargas, l. g. & gao, n. (2013). firm-level outsourcing decision making: a balanced scorecard-based analytic network process model. international journal of production economics. 147: 614-623 varadarajan, r. (2009). outsourcing: think more expansively. journal of business research, 62(11): 1165-1172. doi:10.1016/j. jbusres.2008.09.006 wei, chun-c., chien, chen-f., & wang, mao-jiun j. (2005). an ahp-based approach to erp system selection. int. j. production economics, 96(1): 47-62. doi:10.1016/j.ijpe.2004.03.004 wright, t. (2005). outsourcing financial services authority report on offshoring. computer law & security report, 21: 500-504. doi:10.1016/j.clsr.2005.08.008 youngdahl, w., & ramaswamy, k. (2008). offshoring knowledge and service work: a conceptual model and research agenda. journal of operations management, 26(2): 212-221. doi:10.1016/j.jom.2007.02.007 youngdahl, w.e., ramaswamy, k., & dash, k. c. (2010). service offshoring: the evolution of offshore operations. international journal of operations & production management, 30(8): 798-820. doi:10.1108/01443571011068171 111int. j. prod. manag. eng. (2014) 2(2), 101-111creative commons attribution-noncommercial-noderivatives 4.0 international ism approach to model offshore outsourcing risks http://dx.doi.org/10.1016/j.im.2008.11.005 http://dx.doi.org/10.1108/02635571211210059 http://dx.doi.org/10.1201/1078/44705.21.4.20040901/84190.9 http://dx.doi.org/10.1016/s0272-6963(99)00015-7 http://dx.doi.org/10.1016/s0263-2373(03)00113-0 http://dx.doi.org/10.1016/j.bushor.2009.02.005 http://dx.doi.org/10.1080/00207540701429926 http://dx.doi.org/10.1201/1078/44432.21.3.20040601/82472.3 http://dx.doi.org/10.1201/1078/44432.21.3.20040601/82472.3 http://dx.doi.org/10.1016/j.orgdyn.2012.08.006 http://dx.doi.org/10.1109/tem.2004.830848 http://dx.doi.org/10.1108/14635770810876575 http://dx.doi.org/10.1016/j.intman.2013.04.004 http://dx.doi.org/10.1016/j.jom.2007.02.006 http://dx.doi.org/10.1016/j.jom.2007.02.009 http://dx.doi.org/10.1016/j.ijpe.2012.03.036 http://dx.doi.org/10.1016/j.sbspro.2012.04.023 http://dx.doi.org/10.1016/j.ejor.2010.03.042 http://dx.doi.org/10.1016/j.jbusres.2008.09.006 http://dx.doi.org/10.1016/j.jbusres.2008.09.006 http://dx.doi.org/10.1016/j.ijpe.2004.03.004 http://dx.doi.org/10.1016/j.clsr.2005.08.008 http://dx.doi.org/10.1016/j.jom.2007.02.007 http://dx.doi.org/10.1108/01443571011068171 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2019.8757 received 2017-10-15 accepted: 2018-10-12 an oil package study aiming the logistics optimization on the palletizing capacity samed, m.m.a.a1 and umemoto, a.l.t.a2 astate university of maringá a1 mmasamed@uem.br, a2 amanda.tanji@hotmail.com abstract: the palletizing layout affects directly the company’s costs and logistics efficiency, because with an optimal layout is possible to reduce the number of truck loading to meet the same demand. therefore, this paper is a case study, showing the actual scenario of soybean oil manufacturing company, to propose a packaging study which will define the optimal dimensions for the package and palletizing. the reached results show a 25% pallet capacity increase, as well as a 25% reduction of the number of truck loading necessary to meet the demand. key words: logistics, optimization, package. 1. introduction adding value to the products is something that could be done in several perspectives, as price, consumer behavior and/or market strategy (vilckas and nantes, 2007). the logistics develops the role to improve the product quality focusing on the transportation and making the products meet the demand, being on the right place at the right time, and the transportation is characterized as the most representative element of its chain (ballou, 2010). according to the national confederation of transport (cnt, 2016), the road transportation in brazil represents more than 60%, of all types of modals. in fact, the road transportation is considered an inefficient modal, because it offers to its products a negative impact, due to the occurrence of damages, breakdowns, delays and other problems (monaro et al., 2015). in order to reduce the handling time to transportation of the products or to minimize this kind of problem, it has being studied the unitization through the palletizing, which reduces the number of trips to deliver the same quantity of products and the number of volumes handled, offering less risks to damage the load (ballou, 2010). furthermore, the main objective of this work is to maximize the capacity of the oil package palletizing on the studied company, consequently, improve the transportation by reducing the costs implied to it, showing all the results obtained on the proposed study and premises adopted to achieve the desired goal. to cite this article: samed, m.m.a. and umemoto, a.l.t. (2019). an oil package study aiming the logistics optimization on the palletizing capacity. international journal of production management and engineering, 7(1), 13-21. https://doi.org/10.4995/ijpme.2019.8757 int. j. prod. manag. eng. (2019) 7(1), 13-21creative commons attribution-noncommercial-noderivatives 4.0 international 13 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. literature review according to ballou (2010), the logistics involve a coordinated management of interrelated activities of several areas that compose an enterprise, such as finances, marketing, production/operations, in addition of the basic activities as transportation and storage. the author still indicates that the logistics resources when well applied and used may add more value to the product or service, attending better the final customers needs, and consequently, increasing the chances to improve the sales. to novaes (2007), the logistics add value of space/place, providing the product on ideal place to the client, time, related to the delivery time of the product or service, quality, including the delivery conditions of the product or service, and information, additional data about the product. thus, the logistics is directly influenced by the type of product transported, handled or storaged. with this, knowing the context of the product development process is really relevant, as well as your package. so the product development process is a set of activities that considers the technological restrictions, competitive and product strategies of the enterprise, to define product specifications to meet a certain demand (rozenfeld and forcellini, 2006). and is crucial to the enterprises survival, because it is a tool to identify the costumer’s needs and determine market’s opportunities (cobra, 2012). by that, the product development process is guided by goals that try to attend the market demand, satisfying the customers requirements and needs, at the same time that add value to the final product, hence achieving the business target (baxter, 2003). in this context, to elaborate this work, it was necessary understanding the conception of the product packages, as an important factor on the product development process, as described bellow. an important step during the product development process is the creation of the packing, which has the basic functions as containing, protecting and facilitating the product transportation. however, over the years it also has been used to product conservation, exposition, selling and to consumer’s communication and attraction (mestriner, 2002a,b). besides that, the packing has impacts on the operation productivity and efficiency, because it affects the loading, storage and transportation of the products (bowersox et al., 2014). finally, the package is defined as a set of project and production activities of the container or wrap of a product, that creates a value to it, when well developed and implemented (kotle and keller, 2006). 3. case study the enterprise studied is a manufacturer of soybean oil, which manufactures the product fully on your own installations. so, basically the raw materials are treated and after the moisture is ready, it goes to the oil packaging area. the preform for the plastic bottles are blown, and then, filled with the oil on the same production line. the preform and plastic bottles after being blowed are shown in the figure 1. figure 1. preform and actual package. table 1. actual dimensions. 1 unit 5 units 4 units package width 79 mm 395 mm 316 mm the actual scenario is based on the actual package which has 249 mm high, 79 mm diameter and 900 mm volume. the secondary package contains 20 plastic bottles, arranged in a 5×4 layout, showed in figure 2 and table 1. int. j. prod. manag. eng. (2019) 7(1), 13-21 creative commons attribution-noncommercial-noderivatives 4.0 international samed, m.m.a. and umemoto, a.l.t. 14 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 2. 5×4 layout arrange (source: author, 2017). the palletizing layout is shown in figure 3, according to the pallet standard brazilian dimensions, as 120 cm width and 100 cm length. figure 3. actual palletizing layout (source: author, 2017). it is important to emphasize that the layers need to be braided, that is, the layout for each layer can’t be the same, where the direction for the empty spaces alternates in the layers. although it looks more simple, it was tested and showed bad results to resistance on the shacking. so, the actual palletizing layout has capacity of 48 boxes, divided in six layers. also it is known that the transportation to the distribution centers is done by three types of trucks: two, three or four axles. considering the pallet capacity in pallets and the weight for each pallet with the 48 boxes, indicated on the table 2, it is possible to note that there is a considerable loading idleness, due the openings on the layout mentioned before. it means that part of the truck is empty during the transportation, with 13%, 20% and 19% for the types of trucks, respectively. table 2. actual dimensions. 2 axles 3 axles 4 axles capacity (kg) 14,000 26,000 30,000 boxes per pallet 48 48 48 capacity (in pallets) 14 24 28 load gross weight (kg) 12,125.7 20,768.9 24,251.4 idleness (%) 13.39 20.05 19.16 analysing the annual numbers, the company is capable of produces more than 8 millions of bottles per year. and approximately 30% of this is transported on the 2 axles trucks, around 65% on 3 axles trucks, and only 2% on the 4 axles type. so, comparing with the idleness for each type of truck, on the table 3 is possible to see the amount of bottles that could be distributed in additional if the idleness were reduced. table 3. idleness per truck in amount of bottles. axles idleness (%) % per year annual production quantity 2 13.39 31.52 8,400,000 354,547 3 20.05 65.72 1,106,832 4 19.16 2.76 44,412 therefore, due the empty spaces on the actual palletizing layout, more than 1.5 millions of bottles could be transported using the same amount of trucks, and with no additional costs, mainly with freight. so the main problems found in the analyzed company are: i. storage, moving and transportation idleness, ii. stacking low resistance, iii. product loss due to the damages in the boxes, iv. damaged loading returns and loss of products for selling. 3.1. restrictions to propose a different arrange and layout, it is necessary analyze the company’s restrictions. in relation with the product, the shape is one of the factors that can not be changed, due the blowing int. j. prod. manag. eng. (2019) 7(1), 13-21creative commons attribution-noncommercial-noderivatives 4.0 international an oil package study aiming the logistics optimization on the palletizing capacity 15 http://creativecommons.org/licenses/by-nc-nd/4.0/ process line available. a change to a squared shape, for example, requires different and better raw materials and machines, then, the proposal may considers a cylinder shape for the new package. besides, looking to the logistics and trade aspects, it is important to verify the storage, transportation and distribution conditions, mainly due to the height of the bottle/package. on the logistics, it may affect the stock height of the customers and distribution centers, and for trade marketing, affects on the gondola heights. 4. methodology after that, it was verified the viability of the project implementation, the risks involved, and mechanical and financial factors affected by the package modification. besides, due to the information from the company is calculated a representative payback. the actual palletizing layout has several risks, because of the idleness there is low stacking resistance favoring the loads collapse, consequently is more likely that damages happen. then the mechanicals aspects are related with load transportation and movement. the main aim of the implementation is having a more resistant load stacking, which implies in an easier movement and a bigger stacking stability, optimizing the mechanical process. then, to check the project viability is necessary to verify the mechanical and financial factors affected. the mechanical aspects are related with load transportation and movement. the main aim of the implementation is having a more resistant load stacking, which implies in a easier movement and a bigger stacking stability, optimizing the mechanical process. due the company’s privacy, the financial information as cost, revenue and profit, could not be revealed. so it was calculated a simple representative payback, getting that the bottles amount lost due to damages is approximately one million. after that, it was possible to construct the problem modelling considering all the relevant information to the packaging. the first new information is about the height of the market racks equal to 26.2 cm, determined by the trade marketing department. besides, as a market requirement the packaging arrangement contains 20 bottles, organized as 5×4 units. the secondary package is a cardboard box which has 2 mm thickness for each side of the box, summing 4 mm per box. furthermore, to have a reasonable bottle diameter, each side of pallet can fit a maximum of 15 bottles. this, creates limitations of minimum and maximum diameter dimensions for the bottles, as 0.066 m and 0.079 m, respectively, due the standard measurements for the pallets in brazil. after collecting all the data, it is possible to make a packaging study to verify which are the optimal dimensions for this product. basically, different palletizing layouts were proposed using the cad software, trying to fit the largest amount of boxes for each layer on the pallet, considering the maximum and minimum limits and an interval of 3.25 mm, originating the five new scenarios. with the 5×4 units arrange, five units organized side by side indicate the width, and four units indicate the length of the package, as shown in table 4. table 4. possible layouts. axles idleness (%) % per year annual production quantity 1 0.06600 0.334 0.268 0.090 2 0.06925 0.350 0.281 0.100 3 0.07250 0.367 0.294 0.108 4 0.07575 0.383 0.307 0.118 5 0.07900 0.399 0.320 0.128 according to the pallet size (1.2×1.0 m), it is known that its area is equal to 1.2 m². so, knowing the five scenarios dimensions, it is possible to calculate the optimal number of boxes that one palletizing layer could fit, that is, the biggest quantity of boxes that can be arranged in a layer, considering the area occupied by each scenario, as is presented on table 5. table 5. optimal quantities per scenario. scenario package area (m²) pallet area (m²) boxes per pallet layer 1 0.090 1.2 13 2 0.100 12 3 0.108 11 4 0.118 10 5 0.128 9 int. j. prod. manag. eng. (2019) 7(1), 13-21 creative commons attribution-noncommercial-noderivatives 4.0 international samed, m.m.a. and umemoto, a.l.t. 16 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 4. the five scenarios proposed. however, since the arrange must be grouped in the 5×4 layout, it may infer a complication to fit this amount of boxes, existing the possibility do not reach it, due to the empty spaces among the boxes. another consideration is about the bottle volume, which must fit 900 mm. thus, to find the volume, according to the height determined before as 26.2 cm, is considered the bottle as a cylinder. and to have an acceptable limit, it was defined that the cylinder has 1-liter volume, since part of the packaging goes empty, then it is found the minimum diameter that the bottle must have to fit this volume, which is 0.0697 m, discarding the scenarios 1 and 2. besides, the scenario 5 has the same dimensions that the actual bottle, so it won’t be analyzed. 5. results to have a considerable number of possibilities, it was proposed three arranges for each one of the five scenarios. although it could be proposed more arranges, it was verified that most of then had the same quantity of boxes that the actual lay-out and bigger empty spaces, so it is shown only the three best arranges found. starting from the scenario 3, as illustrated on figures 5, 6 and 7. as discussed before, the optimal quantities do not imply in the real number of boxes fit on the pallet, as it possible to see in the arranges 1 and 2, for scenario 3, as the expectation was to have eleven boxes for each layer, fitting only ten. and the arrange 2 has the big right-side space, which may compromise the stacking, while the number 3 present measurements out of the pallet, remaining only the first one. similarly, the scenario 4 was arranged, as shown in the figure 8, 9 and 10. figure 5. scenario 3, arrange 1 (source: author, 2017). figure 6. scenario 3, arrange 2 (source: author, 2017). figure 7. scenario 3, arrange 3 (source: author, 2017). int. j. prod. manag. eng. (2019) 7(1), 13-21creative commons attribution-noncommercial-noderivatives 4.0 international an oil package study aiming the logistics optimization on the palletizing capacity 17 http://creativecommons.org/licenses/by-nc-nd/4.0/ as well as the last scenario, the arranges 1 and 3 fit only nine boxes, which is the same quantity of the actual palletizing layout, not presenting any efficiency improvement. besides, the arrange 2 has bigger dimensions than the pallet. so from the six arranges, only one was left. to have more satisfactory results, it became necessary to analyze more scenarios. then, due to the elimination of scenario 4, 1 and 2 were discarded because they could not fit the volume required, it was defined more scenarios between 3 and 4, as shown in the following table 6. table 6. new scenarios. scenario diameter (m) area boxer per pallet layer 3 0.0725 0.108 11 6 0.0736 0.111 10 7 0.0747 0.114 10 4 0.07575 0.118 10 similarly to the previous scenarios, it was analyzed the possible arranges for the two new scenarios, with three possible layouts tested, showed on figures 11, 12 and 13. it is possible to verify the big empty spaces on the arrange 2, implying in serious problems to stack the layers, discarding it, and remaining the arranges 1 and 3. the last one, the scenario 7 is also shown in the figures 14, 15 and 16. once again, the arrange 1 has remaining dimensions, so it is discarded. the others arrange fit only nine boxes, the same quantity for the actual palletizing layout, implying in the elimination of all arranges in the scenario 7. the three remaining possibilities are originated from the scenario 3, 6 and 6 again, with the arranges 1, figure 8. scenario 4, arrange 1 (source: author, 2017). figure 9. scenario 4, arrange 2 (source: author, 2017). figure 10. scenario 4, arrange 1 (source: author, 2017). figure 11. scenario 6, arrange 1 (source: author, 2017). figure 12. scenario 6, arrange 2 (source: author, 2017). figure 13. scenario 6, arrange 3 (source: author, 2017). int. j. prod. manag. eng. (2019) 7(1), 13-21 creative commons attribution-noncommercial-noderivatives 4.0 international samed, m.m.a. and umemoto, a.l.t. 18 http://creativecommons.org/licenses/by-nc-nd/4.0/ 1 and 3, respectively, and all of them has the capacity of 200 bottles per layer, or 10 boxes, stacking 60 boxes in the layout. the proposed scenarios are shown on table 7 and figures 17, 18 and 19. as the goal of this work as minimize the idleness, it was verified the available area on each layer, by calculating the available area on the pallet subtracting the area occupied by the packages, with 0.12 m² for the first option and 0.09 m² for the others two, eliminating the first arrange from scenario 3, and remaining the others two. from the two left options, both has the same occupied area and contain the same number of boxes per layer on the palletizing layout, however, the arrange 3, from scenario 6, presents a layout that could not be braided, which is necessary, meaning figure 14. scenario 7, arrange 1 (source: author, 2017). figure 15. scenario 7, arrange 2 (source: author, 2017). figure 16. scenario 7, arrange 3 (source: author, 2017). figure 17. scenario 3, arrange 1 (source: author, 2017). figure 18. scenario 6, arrange 1 (source: author, 2017). figure 19. scenario 6, arrange 3 (source: author, 2017). table 7. analyzed arranges follow-up. diameter (m) status reason 1 0.06600 discarded smaller dimensions than the minimum radius to fit 1 liter volume 2 0.06925 discarded 3 0.07250 1 arrange 4 0.07575 discarded the arranges have dimensions out of the pallet or fit the same quantity of boxes per layer of the actual layout 5 0.07900 discarded actual bottle dimensions 6 0.07360 2 arranges 7 0.07470 discarded the arranges have dimensions out of the pallet or fit the same quantity of boxes per layer of the actual layout int. j. prod. manag. eng. (2019) 7(1), 13-21creative commons attribution-noncommercial-noderivatives 4.0 international an oil package study aiming the logistics optimization on the palletizing capacity 19 http://creativecommons.org/licenses/by-nc-nd/4.0/ that all the layers have the same layout, not inverting the direction of the boxes, and it highly compromises the resistance of the stacking, as already mentioned before. therefore, the arrange chosen is the option 2, from scenario 6 on the proposed arrange 1, showed in the figure 20. figure 20. new pallet layout. the optimal size or diameter for the bottle is equal 0.0735 m. and considering the occupied area on the palletizing layout, according with the dimensions of pallet layer, the idleness is less than 8.5%. then, the new pallet layout fits 60 boxes, or 10 boxes in each one of the six layers. compared with the actual scenario, it represents an increase of 25% on the capacity of only one pallet. it means that on the transportation of one single pallet, there is a gain of 12 boxes or 240 bottles if used the new pallet layout proposed. the information is represented on table 8. table 8. proposed arrange information for each type of truck. scenario 2 axles 3 axles 4 axles bottles per box 20 20 20 boxes per pallet 60 60 60 pallet gross weight (kg) 1,073.9 1,073.9 1,073.9 capacity (in pallets) 13 24 27 load gross weight (kg) 13,960.7 25,772.6 28,995.3 idleness (%) 0.29 0.87 3.35 finally, the improvements of the palletizing optimization are analyzed, showing the positive results, as well as the numbers for the idleness to the proposed layout. the data shown in table 2 is now updated in table 8. it is possible to notice that several factors were optimized due the modification on the bottle dimensions. table 9. comparison of the actual layout and the proposed using the demand met for 2015. scenario 2 axles 3 axles 4 axles boxes transported 672 780 1,152 1,440 1,344 1,620 gain (%) 16.07 25 20.53 loads 1,485 1,280 1,806 1,445 65 54 reduction (%) 13.8 19.99 16.92 using the demand numbers met in 2015 for the production of the oil, it is calculated how many trips would be necessary to meet the same demand with the proposed layout. since the numbers of boxes per pallet is higher, it is expected that the numbers of trips are smaller, obtaining a positive result for all of the three types of transportation, as shown in table 9. the first improvement is about the boxes per pallet which had an increase of 25%, having the capacity of 60 boxes compared to the 48 in the actual layout. with this, the loading capacity of the trucks is also expanded, reaching 13, 24 and 27 pallets per load for the two, three and four axles, respectively. consequently, the loading idleness is reduced too. in the old layout, the idleness was 13% for the two axles trucks, 20% for the three axles and 19% for the four axles. implementing the proposal, it could change drastically the idleness, getting 0.29% for two axles, 0.87% for three axles and 3.35% for four axles. the numbers show that the proposal has a gain of 16% for the two axles trucks, 25% for the three axles and 20% for the four axles. at the same time, it reduces the number of loadings necessary, reaching 13%, 20% and 16% respectively. it means that is necessary less trips to meet the same demand met in 2015, since it is possible to transport more boxes per pallet, when compared with the actual layout. 6. conclusion the proposed study resulted in an increase of 25% on the pallet capacity, where the old layout fitd 48 boxes, or 960 bottles, and with the implementation of the proposal, the new palletizing layout would fit 60 boxes, or 1,200 bottles. besides, the logistic idleness when using the actual palletizing layout represents 13.39% of the truck with 2 axles, 20.05% for trucks with 3 axles and, int. j. prod. manag. eng. (2019) 7(1), 13-21 creative commons attribution-noncommercial-noderivatives 4.0 international samed, m.m.a. and umemoto, a.l.t. 20 http://creativecommons.org/licenses/by-nc-nd/4.0/ finally 19.16% for those trucks with 4 axles. then, with the proposed layout, the idleness values became 0.29%, 0.87% and 3.35%, respectively for the types of trucks mentioned. consequently, considering the demand met on 2015, there was a gain of 16.07% of the quantity fit on the pallets loading for the trucks with 2 axles, 25% for those with 3 axles, and 20.53% for the 4 axles. it indicates that the implementation would bring to the company an increase from 672 to 780 boxes for 2 axles truck, 1,152 to 1,440 boxes to 3 axles, and 1,344 to 1,620 boxes for 4 axles. thus, to met the demand were necessary 1,485, 1,806 and 65 loads for the 2 axles, 3 axles and 4 axles trucks, respectively. so, after the implementation, the number of loads to met the same demand would be reduced to 1,280, 1,445 and 54, for the same types of trucks, respectively, showing a significant reduction on the quantity of trips that the trucks should do to met the costumers needs in 2015. besides, it is is possible to reduce the idleness on the palletizing layout, which currently is equal to 14.88%. with the improvement it would be increased to 7.5%, which means that, previously, analyzing the occupied volume for the six layers on pallet, approximately 15% represented empty spaces, while with the proposed layout would turn out to be only 7.5%, because the layout can not fir all the space available on the truck, as explained before. with a smaller area empty or idle on the proposed palletizing layout, there is a better stacking during the transportation and storage too, because there is less idle spaces among the boxes on the layers, it results on less damaged loads or products during the transportation or storage. besides, it was shown a considerable gain in the logistics efficiency, this is because it will be necessary a smaller number of shipments of loads to meet the requested demand, and the company will be able to deliver that in a shorter period, compared with the actual palletizing layout and transportation characteristics. in summary, for all types of trucks, the proposal had really positive numbers in idleness, number of trips to meet the demand, quantity of boxes per pallet, amount of boxes transported, and several another benefits could be noticed. finally, the results show that the study does not works only for the oil bottles, but for all similarshape packages, being useful for any kind of product, since it was presented all the information necessary to a package study. references ballou, r.h. (2010). gerenciamento da cadeia de suprimentos/logística empresarial. porto alegre, brazil: bookman. baxter, m.r. (2003). projeto de produto: guia prático para o desing de novos productos. são paulo, brazil: blücher. bowersox, d.j., closs, d.j., cooper, m.b., bowersox, j.c. (2014). gestão logística da cadeia de suprimentos. porto alegre, brazil: amgh. cobra, m. (2012). marketing básico: uma perspectiva brasileira. são paulo, brazil: atlas. kotler, p., keller, k.l. (2006). administração de marketing. são paulo, brazil: pearson. mestriner, f. (2002a). design de embalagem: curso básico. são paulo, brazil: pearson makron books. mestriner, f. (2002b). gestão estratégica de embalagem: uma ferramenta de competitividade para sua empresa. são paulo, brazil: prentice hall brasil. monaro, r.l.g., pinton, c.g.s., monaro, d.l.g. (2015). a influência da paletização na qualidade das cargas durante o transporte físico. xxxv encontro nacional de engenharia de producao, perspectivas globais para a engenharia de produção. fortaleza, ce, brasil, 13 a 16 de outubro de 2015. available via http://www.abepro.org.br/biblioteca/tn_sto_206_222_27431.pdf. accessed 15 may 2016. national confederation of the transportation – cnt (2016). boletim estatístico. available via http://cms.cnt.org.br/imagens%20cnt/ boletim%20estat%c3%8dstico/boletim%20estat%c3%8dstico%202016/boletim%20estat%c3%adstico%20-%2001%20 -%202016.pdf. accessed 15 may 2016. novaes, a.g. (2007). logística e gerenciamento da cadeia de distribuição: estratégia, operação e avaliação. rio de janeiro, brazil: campus. rozenfeld, h., forcellini, f.a. (2006). gestão de desenvolvimento de produtos: uma referência para a melhoria do processo. são paulo, brazil: saraiva. vilckas, m., nantes, j.f.d. (2007). agregação de valor: uma alternativa para expansão do mercado de alimentos orgânicos. organizações rurais & agroindustriais, 9(1), 26-37 available via http://www.redalyc.org/articulo.oa?id=87890102>. accessed 15 may 2016. int. j. prod. manag. eng. (2019) 7(1), 13-21creative commons attribution-noncommercial-noderivatives 4.0 international an oil package study aiming the logistics optimization on the palletizing capacity 21 http://www.abepro.org.br/biblioteca/tn_sto_206_222_27431.pdf http://www.redalyc.org/articulo.oa?id=87890102 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2016.4425 received 2015-12-10 accepted: 2015-12-14 social network analysis and supply chain management raúl rodriguez-rodriguez* and ramona d. leon research centre on production management and engigeering, universitat politècnica de valència, cno. de vera s/n, 46022 valencia (españa). * raurodro@upvnet.upv.es abstract: this paper deals with social network analysis and how it could be integrated within supply chain management from a decision-making point of view. even though the benefits of using social analysis have are widely accepted at both academic and industry/services context, there is still a lack of solid frameworks that allow decision-makers to connect the usage and obtained results of social network analysis – mainly both information and knowledge flows and derived resultswith supply chain management objectives and goals. this paper gives an overview of social network analysis, the main social network analysis metrics, supply chain performance and, finally, it identifies how future frameworks could close the gap and link the results of social network analysis with the supply chain management decision-making processes. key words: social network analysis, supply chain management, metrics. 1. introduction it is currently widely accepted that the global economy has drifted and it is more and more ictsupported than ever. many sectors are dominated by organisations that represent this digital breakdown and fully rely on ict advances to make a difference: skype with no telecommunication infrastructure, facebook with no content creation, airbnb (world’s accommodation leader) with no property owned etc. this is possible due to the fact that linking people, services and products together as well as interconnect them via information and knowledge exchange is the differentiate point nowadays. the advent of ict shapes the organisations from inside out; it affects employees’ productivity and firm’s capacity of innovating (molina-castillo et al., 2012) and at the same time, it fosters the relationship between the company and its customers (gunawan and huarng, 2015). in this way, organisations have available different tools in order to support and carry out these information and knowledge exchange processes. one of the most famous one is through online social networks e.g. facebook or twitter. however, these are open networks and organisations usually use them to communicate with their final customers instead of with the actors upstream the supply chain, namely with their suppliers. an alternative to this are the called private online social networks, which are accessed only by invitation and represent an optimal opportunity for organisations to build up a social network with their supply chain partners to foster the information and knowledge interchange. once they have set up the social network and it has been working for a while, the application of social network analysis will outcome important additional information to make decisions. however, it is possible to affirm that decision-makers do not have yet tools, mechanisms or frameworks that help them out to make decisions regarding how social network analysis is affecting to supply chain management and, extensively, how supply chain both as a whole and from its individual members point of view, can be linked together under an integrative approach. this paper tackles this line of research, and it is structured as follows: the next section presents a brief literature review on i) social network analysis; ii) social network analysis metrics; iii) supply chain measurement. emerging from this point, the main advantages of using social network analysis at the supply chain level as well as some specific lines where both social network analysis and supply chain management could be brought together are then presented. finally, the main conclusions are highlighted. int. j. prod. manag. eng. (2016) 4(1), 35-40creative commons attribution-noncommercial-noderivatives 4.0 international 35 http://dx.doi.org/10.4995/ijpme.2016.4425 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. brief literature review 2.1. social network analysis social network analysis has its roots in the work of kurt lewin (1936) who combined the abstract character of mathematics with the subjective/ interpretative character of sociology, and put the bases of the mathematical based field theory, the predecessor of the graph theory. both findings contribute to the development of what is called today “sociometry” and provide the necessary context for analysing, measuring and understanding relationship’s value. against this framework, a social network is defined as set of relationships developed among the members of a group. from an abstract perspective, a social network can be a representation of series of nodes and lines in which the nodes describe an individual, a team, an organization, a community or a country and the lines emphasize a relation that has been established between two nodes, based on preferences or necessity (sandru, 2012). from a subjective approach, a social network describes the information, tacit and explicit knowledge that flows within a group of people; each node represents an individual who can act as a knowledge holder (transmitting emotions, values, ideas, experiences, advice, stories, best practices etc. to the others) or as a knowledge receiver (interacting with some of the others members in order to receive answers to his/her current or potential problems). nevertheless, both approaches highlight a mathematical graph based on a binary interaction matrix; no matter the actors’ nature, the relationships are usually encoded based on a binary code where 1 symbolizes the existence of interaction between two members and 0 reflects its absence. within the social network theories, social network analysis appears as a branch of mathematical sociology, capable of evaluating structural positions and characteristics at both node and network level. however, it can be used for both exploratory and confirmatory issues. on the one hand, in line with a confirmatory approach, social network analysis allows testing hypothesis by converting the subjective nature of relationships into abstract parameters and probabilities. it provides the necessary framework for testing hypothesis regarding networks’ and groups’ means, densities, correlations and regression and it also facilitates the prediction of future relationships (hanneman and riddle, 2005). on the other hand, in line with the exploratory approach, it facilitates the visualization and exploration of nodes’ and networks’ characteristics. at the node level, various analysis may be conducted in order to emphasize individuals’ position and characteristics (zhu et al., 2010; hu, 2013); their importance within the network is reflected based on degree, betweenness and closeness centrality, reachability and connectivity while their preferences are highlighted through the homophily analysis. at the network level, several analysis can be developed in order to bring forward networks characteristics (gunawan and huarng, 2015; lin and lo, 2015); its potential is emphasized based on networks’ density, its diversity is reflected based on clustering analysis, structural and automorphic equivalence, while its efficiency is brought forward krackhardt gdt analysis. the sna measures that are usually applied are presented further. 2.2. social network analysis metrics through the application of social network analysis techniques, a graphical approach of the main relationships among and between the members of such a social network is first taken into account to make decisions. however, there is much more information when looking at the metrics that these analysis techniques output. (turetken and sharda, 2007) state that the three most used social network analysis metrics are: degree centrality, degree of intermediation (betweenness centrality) and proximity (closeness). these three metrics are further developed next. the degree centrality is the number of direct ties that an actor (or node) possesses; for instance, how many other nodes are directly connected. this metric indicates who is the most connected member in a group. then, a member with a high degree has got many connections to other network members, carrying out hub tasks within the network. this implies that a member with a high degree has got a high degree to influence other members. then, organisations that precisely identify who the hub members are possess important additional information when trying to disseminate both information and knowledge within the network. additionally, when looking at interorganisational contexts such a hub identification process will provide, in return, a key and valuable contact point with external supply chain companies. figure 1 shows an example of degree centrality where big squares mean a high degree and the small int. j. prod. manag. eng. (2016) 4(1), 35-40 creative commons attribution-noncommercial-noderivatives 4.0 international rodriguez-rodriguez, r. and leon, r.d. 36 http://creativecommons.org/licenses/by-nc-nd/4.0/ squares a low degree. then, it can be seen that the user u126 has got the highest degree centrality value whereas that other users located at the margins of the network such as u89, u98 or u127 has got the lowest level of degree centrality. figure 1. example of degree centrality. the degree of intermediation shows how often a node appears in the shortest section that connects two others. in other words, it indicates when a member plays the role of intermediary between two other members that do not keep any relationship within the network. then, it is necessary to differentiate when these two actors that do not keep any relationships with each other: a) keep relationships with other members of the network; b) does not keep any other relationship with other members of the network, only the only kept through the intermediary. in the former case, the removal of the intermediary member is not as dramatic as in the latter one. removal of intermediary members, carried out by organisations by any reason, when they are the only link for two isolated members will lead to a knowledge sharing loss. these intermediary members of isolated members should be the last ones to leave the network, therefore actions to foster their presence and participation should be carried out and monitored. figure 2 shows an example of degree of intermediation, where the big squares mean a high degree and the small squares a low degree. then, it can be seen that the user u126 has got the highest degree of intermediation value whereas that other users located at the margins of the network such as u113 or u104 has got the lowest level of degree of intermediation. figure 2. example of betweenness centrality. finally, the closeness or proximity metric indicates how close is a node from the rest of the network. it represents the ability of a member to reach others within the network. figure 3 shows an example of closeness, where the circles mean a high degree of closeness and the squares show a low level. looking at both the incoming and outcoming arrows, it is possible to observe that users u126, u108 and u128 are the ones with a highest degree of closeness within the network. figure 3. example of closeness. in any case, it is necessary to take into account that other metrics can be obtained when applying int. j. prod. manag. eng. (2016) 4(1), 35-40creative commons attribution-noncommercial-noderivatives 4.0 international social network analysis and supply chain management 37 http://creativecommons.org/licenses/by-nc-nd/4.0/ social network analysis. these metrics are mainly related to contingency, hily, strength, heterogeneity and network density. additionally, it is possible to develop knowledge flows forecasting by applying the holland and leinhardt’s p1 model (1981). these additional metrics have not been so popular within the literature and further attention should be put on their application to situations as the one presented in this paper. 2.3. supply chain measurement when looking at the supply chain management context aiming to establish how to measure its results and performance, many frameworks have been developed in the last years. starting with the plane application of the well-known balanced scorecard (kaplan and norton, 1992) to other most sophisticated approaches. to mention some recent balanced-scorecard works, chang et al. (2013) developed a supply chain performance measurement system that integrates r&d activities as well as marketing policies; both tajbakhsh and hassini (2015) and shafiee et al. (2014) developed a supply chain performance evaluation system based on the combined use of both data envelopment analysis and balanced scorecard. regarding supplier evaluation and selection, heidarzade et al. (2016) developed a clustering method based on fuzzy logic, and bruno et al. (2016) present a framework where they several combined supplier selection methodologies in a multistakeholder environment. on the other hand, there have been also some supply chain performance measurement frameworks to measure the impact of collaboration over performance. then, verdecho et al. (2012) applied the analytic network process to identify the main collaborative factors and to evaluate how collaboration practices affect to performance; keehung et al. (2015) presented a work regarding how environmental information sharing affects to supply chain partners and performance. however, when it comes to link together supply chain management and social network analysis there is a gap from a researching point of view. in other words, if members of a supply chain share and develop both information and knowledge for a certain time period, they will not have available any solid framework to help them out to interpret and project their findings, at the social network context, towards the supply chain one. the latter will ask to quantify how certain actions of, for instance, knowledge sharing within the social network will help to achieve shorter cycle times. these two-sided intrinsically linked decisionmaking contexts should be further studied, and solid and effective frameworks developed. in the next point, we provide some insights that could be taken when trying to accomplish this task. 3. social network analysis and supply chain management derived from the previous point, it is possible to highlight the main advantages of maintaining a social network and develop associated analysis through its metrics. such advantages are mainly: fostering innovation capabilities, increment of the task orientation level and facilitating of both internal and external communication. these three are next further developed. the implementation and use of suppliers’ social network enhances the innovation capability of each member and the supply chain as a whole. first of all, it ensures the creation of a shared vision and it increases members’ cohesion. based on this, networks leaders act as distributors, disseminating the image of what they are as a supply chain and what do they aim to be. as a consequence, the members understand what happens with their products once they leave their factory, what is the role they play into the big system (the supply chain) and with which organisations they should cooperate more in order to improve their strengths and diminish their vulnerabilities. second of all, it increases the level of task orientation. once each firm knows where does it stands within the supply chain, it can adapt its internal processes and structures so that it increases its performance and also the supply chain performance. at this level process innovation may appear in order to improve decision-making, collaboration between several departments or teams etc. third of all, it facilitates internal and external communication and knowledge sharing. at the internal level, it supports organizational learning and it facilitates the process of intergenerational learning since knowledge is shared among the social network. besides strengthening the relationships int. j. prod. manag. eng. (2016) 4(1), 35-40 creative commons attribution-noncommercial-noderivatives 4.0 international rodriguez-rodriguez, r. and leon, r.d. 38 http://creativecommons.org/licenses/by-nc-nd/4.0/ between employees, it also supports the processes of human resources evaluation (especially, those related to skills and abilities evaluation), and it offers the necessary framework for designing internal knowledge maps (which shows who knows what). therefore, it is widely accepted that social network analysis brings organizational competitive advantages. the remaining question is: to what extent? the creation, implementation and maintenance of a social network within a supply chain will lead, if properly managed to, among others, the advantages above highlighted. but the question is: how is paying-off to create and maintain a social network? further, to what extend is such a social network contributing to reach specific supply chain objectives? up to now there are not any solid/integral framework available to respond these questions. some possibilities are: to subjectively quantify the impact of social network analysis over the supply chain’s results. this could be done applying subjective techniques such as questionnaires, surveys, multi-criteria decision-aid techniques, etc. the main advantage of this approach is that it does not need to wait until historical data is available. however, its main disadvantage lies in its intrinsic subjectively, as decision-makers will rely on subjective judgements instead on real data evolution. to objectively quantify the impact of social network analysis over the supply chain result’s. this approach will need from real data from different metrics: the ones from the social network analysis –mainly the previously presented– and from the supply chain –metrics regarding operations, finance, customer, etc.–, applying then statistical techniques to find out whether there is a relationships between these set of metrics. in other words, to identify whether the social network via its metrics lead to a change – either positive or negativeon the supply chain metrics. the main advantage of this approach is that it does not rely on subjective judgement but on real data. the main disadvantage is that it requires more time for collecting the data and that the statistical techniques to be applied will be, surely, more complicated than the ones applied in the subjective approach. any of these two approaches are valid and could be considered as a first step for presenting a solid framework for analysing whether and to what extent social network analysis is affecting to supply chain management. 4. conclusions in today’s world, it is widely accepted that ict support greatly organizational decision-making processes, specially the interchange of both information and knowledge at both intra and interorganisational contexts. one of the most recent tools to carry out such a interchange is social networks, being specially indicated for organizations closed social networks. by applying social network analysis, decision-makers has got available a set of metrics from which they can get to know who the key users, from a knowledge interchange, creation and absorption point of view, are. however, there is a lack of research when coming to connect the results obtained from social network analysis with the supply chain side. works that either subjectively or objectively carry out such a connection are needed, as they will allow to organisations to identify how and to what extent social network analysis and practices are affecting to supply chain management. acknowledgements “the research reported in this paper is supported by the european commission for the project “engaging in knowledge networking via an interactive 3d social supplier network (knownet)” (fp7people-2013-iapp 324408)”. references bruno, g., esposito, e., genovese, a., simpson, m. (2016). applying supplier selection methodologies in a multi-stakeholder environment: a case study and a critical assessment. expert systems with applications, 43: 271-285. doi:10.1016/j.eswa.2015.07.016 chan, f.t., nayak, a., raj, r., chong, a.y.l., manoj, t. (2013). an innovative supply chain performance measurement system incorporating research and development (r&d) and marketing policy. computers & industrial engineering, 69: 64-70. doi:10.1016/j.cie.2013.12.015 gunawan, d.d., huarng, k.h. (2015). viral effects of social network and media on consumers´ purchase intention. journal of business research, 68(11): 2237-2241. doi:10.1016/j.jbusres.2015.06.004 int. j. prod. manag. eng. (2016) 4(1), 35-40creative commons attribution-noncommercial-noderivatives 4.0 international social network analysis and supply chain management 39 http://dx.doi.org/10.1016/j.eswa.2015.07.016 http://dx.doi.org/10.1016/j.cie.2013.12.015 http://dx.doi.org/10.1016/j.jbusres.2015.06.004 http://creativecommons.org/licenses/by-nc-nd/4.0/ hanneman, r.a., riddle, m. (2005). introduction to social network methods. riverside, ca: university of california, riverside. heidarzade, a., mahdavi, i., mahdavi-amiri, n. (2016). supplier selection using a clustering method based on a new distance for interval type-2 fuzzy sets: a case study. applied soft computing, 38: 213-231. doi:10.1016/j.asoc.2015.09.029 holland, p., leinhardt, j. (1981). an exponential family of probability distributions for directed graphs. journal of the american statistical association, 76(373): 33-50. doi:10.1080/01621459.1981.10477598 hu, y. (2013). hyperlinked actors in the global knowledge communities and diffusion of innovation tools in nascent industrial field. technovation, 33(2-3): 38-49. doi:10.1016/j.technovation.2012.10.001 kaplan, r.s., norton, d.p. (1992). the balanced scorecard. measures that drive performance. harvard business review, 70(1): 71-9. kee-hung, l., wong, w.y., lee lam, s. (2015). sharing environmental management information with supply chain partners and the performance contingencies on environmental munificence. international journal of production economics, 164: 445-453. doi:10.1016/j. ijpe.2014.12.009 lewin, k. (1936) principles of topological psychology. martino fine books. doi:10.1037/10019-000 sandru, c. (2012). epistemic and methodological aspects of network analysis. bulletin of the transilvania university of braşov, 3: 63-74. lin, s.w., lo, l.y.s. (2015). mechanisms to motivate knowledge sharing: integrating the reward systems and social network perspectives. journal of knowledge management, 19(2): 212-235. doi:10.1108/jkm-05-2014-0209 molina-castillo, f.j., lopez-nicolas, c., soto-acosta, p. (2012). interaction effects of media and message on perceived complexity, risk and trust of innovative products. european management journal, 30(6): 577–587. doi:10.1016/j.emj.2012.07.005 shafiee, m., lotfi, f.h., saleh, h. (2014). supply chain performance evaluation with data envelopment analysis and balanced scorecard approach. applied mathematical modelling, 38(21-22): 5092-5112. doi:10.1016/j.apm.2014.03.023 tajbakhsh, a., hassini, e. (2015). data envelopment analysis approach to evaluate sustainability in supply chain networks. journal of cleaner production, 105: 74-85. doi:10.1016/j.jclepro.2014.07.054 turetken, o., sharda, r. (2007). visualization of web spaces: state of the art and future directions. data base, 38(3): 51-81. doi:10.1145/1278253.1278260 verdecho, m.j., alfaro-saiz, j.j., rodriguez-rodriguez, r., ortiz-bas, a. (2012). a multi-criteria approach for managing inter-enterprise collaborative relationships. omega, 40(3): 249-263. doi:10.1016/j.omega.2011.07.004 zhu, b., watts, s., chen, h. (2010). visualizing social network concepts. decision support systems, 49(2): 151-161. doi:10.1016/j. dss.2010.02.001 int. j. prod. manag. eng. (2016) 4(1), 35-40 creative commons attribution-noncommercial-noderivatives 4.0 international rodriguez-rodriguez, r. and leon, r.d. 40 http://dx.doi.org/10.1016/j.asoc.2015.09.029 http://dx.doi.org/10.1080/01621459.1981.10477598 http://dx.doi.org/10.1016/j.technovation.2012.10.001 http://dx.doi.org/10.1016/j.ijpe.2014.12.009 http://dx.doi.org/10.1016/j.ijpe.2014.12.009 http://dx.doi.org/10.1037/10019-000 http://dx.doi.org/10.1108/jkm-05-2014-0209 http://dx.doi.org/10.1016/j.emj.2012.07.005 http://dx.doi.org/10.1016/j.apm.2014.03.023 http://dx.doi.org/10.1016/j.jclepro.2014.07.054 http://dx.doi.org/10.1145/1278253.1278260 http://dx.doi.org/10.1016/j.omega.2011.07.004 http://dx.doi.org/10.1016/j.dss.2010.02.001 http://dx.doi.org/10.1016/j.dss.2010.02.001 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2015.3150 received 2014-06-29 accepted: 2014-12-15 integration of six sigma and ism to improve supply chain coordination – a conceptual framework pratima mishrai and rajiv kumar sharmaii department of management and social sciences, national institute of technology (nit) hamirpur, himachal pradesh-177005, india i pratima_mishra1@rediffmail.com ii rksnithmr@gmail.com abstract: to achieve competitiveness and to improve supply chain performance, supply chain coordination (scc) is considered as a key challenge by the companies to satisfy their customers. in today’s turbulent economic environment, scc is a topic of great significance among business houses because scc creates understanding, molds human behavior and improve competitiveness. as observed from literature that the dynamics of global market has resulted in serious pressure and distraction to activities of various supply chain entities i.e. suppliers, manufacturer, distributors, wholesaler, retailers and customers, which ultimately affects the sc performance. thus, supply chains are exposed to risks due to uncertain and turbulent economic environment. to overcome these challenges, authors in this study developed a conceptual framework based on six sigma and ism which can be used to study various supply chain dimensions be it human, process or quality dimensions. the main advantage of this framework is that it not only helps to understand information regarding the strength and weaknesses of various supply chain entities in a supply chain but also helps to determine the structural relationship among key dimensions of interest. the proposed framework can be applied by industries to model and analyze their processes effectively, compare their performance both within and outside their industry segment and thus improve competitiveness by following various supply chain management practices. key words: globalization, supply chain, six sigma, ism. 1. introduction in today’s turbulent environment supply chain coordination (scc) is a topic of much significance in organizations because scc creates understanding, molds behavior and improve competitiveness. intensive competition in the market place has enforced companies to focus on process performance and coordination between the supply chain network for the customer value of products and services in recent times. to create a customer value in the market, companies has to concentrate on supply chains coordination. in the modern global manufacturing environment, change has become a constant phenomenon. changes occur due to fluctuation in technology, environmental requirements, regulatory policies, social needs and the economy is the most important things. since, today’s world competition is no longer between individual organizations but between the supply chains. supply chain includes all the activities related to the processing of materials and the conversion of goods from the stage of raw materials to the stage of delivery to final customer, as well as the informational and financial processes related to them, along with coordinated and integrated management (shafia et al., 2008; harrison and van hoek, 2008). in other words, supply chain defines as, the network of entities through which material flows. those entities include suppliers, manufacturers, wholesaler, retailers and customers (lummus and alber, 1997). different entities in a supply chain function subject to different sets of constraints and objectives. however, these entities are highly interdependent when it comes to improving performance of the supply chain in terms of objectives such as on-time delivery, quality assurance, and cost minimization. as a result, performance of any entity in a supply chain depends on the performance of others, and their willingness and ability to coordinate activities within the supply chain. 75int. j. prod. manag. eng. (2015) 3(1), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2015.3150 http://creativecommons.org/licenses/by-nc-nd/4.0/ a supply chain consists of all stages involved, directly or indirectly, in fulfilling a customer request. the supply chain not only includes the manufacturer and suppliers, but also warehouses, distributors, wholesaler, retailers and customers themselves (chopra and meindl, 2007). to improve the overall performance of supply chain, the members of supply chain may act as a part of an integrated system and coordinate with each other. supply chain members cannot compete as independent members. the product used by the end customer passes all the way through a number of entities contributed in the value addition of the product before its consumption. thus coordination comes into focus. 1.1. supply chain coordination according to lau et al., (2004) and hervani et al., (2005), supply chain is coordination of independent enterprises in order to improve the performance of the whole supply chain by considering their individual needs. supply chain activities transform natural resources, raw materials and components into a finished product that is delivered to the end customer. supply chain coordination is a strategic response to the challenges that arise from the dependencies supply chain members (xu and beamon, 2006). according to arshinder (2008), supply chain coordination can be defined as identifying interdependent supply chain activities between supply chain members and devise mechanisms for manage those interdependencies. it is the measure of extent of implementation of such aggregated coordination mechanism, which helps in improving the performance of supply chain in the best interest of participating members. lee (2000) proposed supply chain coordination as a vehicle to redesign decision rights, work flow, and resources between chain members to influence better performance such as higher profit margins, improved customer service performance, and faster response time. simatupang et al., (2002) classified the coordination effort in supply chain into four categories i.e. logistics coordination, information sharing, incentive alignment and collective learning which needs to be further investigated particularly with respect to human dimensions such as information sharing and collective learning. kim and oh (2005) presented systems dynamics approach to coordinate supplier and manufacturer decisions regarding improvement in quality and the new product development. according to cao et al., (2008), supply chain coordination includes every effort of information exchange and integration during the courses of developing, producing and delivering a product or services to end market places. singh (2011) developed a framework for improving the coordination in supply chain and stressed that coordination cannot be achieved by focusing on only one or two factors of the supply chain. there are many other factors involved in achieving coordination such as human, technology, strategies, relationship, rewards, profits, and risk. most of the literatures focus on coordination issues in supply chain from an operational or tactics sense. how to integrate the mechanism and methods of coordination from a comprehensive or holistic sense have been rarely discussed. the static operation reference based on best practice cannot support the new intra and inter-enterprise collaboration. it is necessary to develop a systematic methodology to help enterprise to develop their requirement, to analyze the target system, to structure the coordination system and to monitor its operational performance. in this paper, authors have tried to develop a conceptual model based on six sigma and ism to improve the coordination between the dimensions so as to have better understanding to interorganizational coordination within a holistic supply chain network from the view of an industry in turbulent environment. paper is organized as follows: a brief introduction of paper presented in section 1, section 2 presents a six sigma approach. section 3 presents the ism approach. in the section 4 benefits of six sigma and ism are presented. research methodology is presented in section 5. integrated framework of six sigma and ism for improving scc performance is presented in section 6. finally conclusions are presented in section 7. 2. six sigma six sigma is a wellstructured knowledge management methodology that focuses on reducing variation, measuring defects, and improving the quality of products, processes and services. six sigma is a statistical quality goal that equals no more than 3.4 defects per million opportunities. six sigma is also a business improvement program that targets process variation. sixsigma is not only for manufacturing, but any process where an opportunity exists for error. six sigma projects are based on a problem solving methodology called dmaic (define, measure, analyze, improve, and control). motorola created the six sigma methodology in 1986, in order to increase its competitiveness against japanese companies in the electronics industry by improving quality levels. the name of the six 76 int. j. prod. manag. eng. (2015) 3(1), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international mishra, p. and kumar sharma, r. http://creativecommons.org/licenses/by-nc-nd/4.0/ sigma methodology is derived from the greek alphabet symbol utilized in statistics for standard deviation, a measurement to quantify variation and process inconsistency (pande et al., 2000). the dmaic is the classic six sigma problem solving process. dmaic resolve issues of defects or failure, deviation from a target, excess cost or time, and deterioration. six sigma reduces variation within and across the value-adding steps in a process. dmaic identifies key requirements, deliverables, task, and standard tools for a project team to utilize when tackling problem. for solving any problem in the six sigma methodology is done by formulating a team of people associated with the process. the six sigma organizational structure includes such as leadership is provided by a team of champions: senior champion, deployment champion and projects champion at corporate, unit and department level, respectively, supported by a team of experts. the experts are referred as master black belts (who provides mentoring, training and expert support to the black belts. table 1. literature related to ism. s. no. use of ism in literature authors 1 knowledge management in engineering industry singh et al., (2003) 2 ism for the barriers in it enablement of supply chain jharkharia and shankar (2005) 3 ism for improving competitiveness of smes singh et al., (2007) 4 enablers of six sigma soti et al., (2010) 5 ism methodology for tqm practices talib et al., (2011) 6 developing the framework for coordination in sc of smes singh (2011) 7 framework for vendor managed inventory adoption in indian industries borade and bansood (2012) black belts who usually work full time on projects at process level to solve critical problems and achieve bottomline results and green belts are the employees who take up six sigma implementation along with their job responsibilities, operating under the guidance of black belts. yellow belts employees that have basic training in six sigma tools. several studies have investigated how six sigma methodology can effectively be employed in scm to measure, monitor and improve the performance of the whole supply chain network. dasgupta (2003) make out the application of six sigma metrics as a complete and flexible framework for evaluating and benchmarking the performance of supply chain and its entities beside world class standards. wang et al., (2004) investigated how quality management can be employed in scm to improve the performance of various issues in the whole supply chain network using six sigma methodology. knowles et al., (2005) concluded that six sigma does have something novel to offer organizations over the contribution of existing approaches to supply chain improvement. they proposed a conceptual model that integrates the balanced scorecard, scor model (supply chain reference model) and six sigma dmaic methodology in a strategic and operational level cycle. liu (2006) presented an application of six sigma to reduce cycle time and defects in clinical report entry. yousef et al., (2008) stated the importance of introducing the concept of six sigma as an effective methodology for monitoring and controlling supply chain variables. nabhani and shokri (2009) presented a case study to reduce the delivery lead time in food distribution with the implementation of dmaic procedures based on the six sigma methodology. chang et al. (2012) applied the six sigma (dmaic) to manage and improve the performance of the production planning procedures. wei and yi-zhong (2013) proposed a framework using six sigma metrics which is comprehensive, flexible and easy to measure and improve supply chain performance. 3. ism methodology ism was first proposed by warfield (1974). it is an interactive learning qualitative tool in which a set of different and directly related elements are structured into a comprehensive systematic model (warfield, 1974 and sage, 1977). it helps to improve orders and direction to the complexity of relationship among various element of the system (sage, 1977). the ism process transform unclear, poorly articulated mental models of system into visible and well defined models useful for many purposes (ahuja et al., 2009). according to the sharma et al. (2011), it is a method for developing the hierarchy of system enablers to represent the system structure. thus it is a models technique as the specific relationship and orders structure are portrayed in a graphical model (singh and kant, 2008, borade and bansod, 2012). hence, ism is a powerful technique, which can be applied in various fields. the application of ism has been applied by many researchers in various literatures. mandal and deshmukh (1994) used the 77int. j. prod. manag. eng. (2015) 3(1), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international integration of six sigma and ism to improve supply chain coordination – a conceptual framework http://creativecommons.org/licenses/by-nc-nd/4.0/ ism methodology to analyze some of the important vendor selection criteria and have shown the interrelationship between criteria and their levels. singh et al. (2003) applied ism for knowledge management in indian industry. singh et al. (2007) applied ism for improving competitiveness of smes. soti et al. (2010) used the ism to study the enablers of six-sigma and to establish the relationship among them. talib et al. (2011) utilize the ism methodology to understand the mutual relationship among the tqm practices and presented a hierarchy-based model of the practices. singh (2011) employed ism approach to develop the structural relationship among different factors of coordination and responsiveness in supply chain to take strategic decision. borade and bansod (2012) applied ism based framework for vendor managed inventory adoption in indian industry. few of the literature related to ism are shown in table 1. the various steps involved in the ism technique are shown in figure 1: step 1: literature review of scm dimensions step 2: identification of dimensions step 3: prioritization of dimensions on the basis of six-sigma value step 4: establish contextual relationship between dimensions step 5: develop a structural self-interaction matrix step 6: develop a reachability matrix step 7: partition the reachability matrix into different levels step 8: prepare the ism step 9: calculate driver and driven power of dimensions step 10: micmac analysis figure 1. flow diagram for research methodology of ism. 4. benefits of six sigma six sigma benefits are related to various areas such as reduction in process variability, reduction in inprocess defect levels, reduction in maintenance inspection time, improving capacity cycle time, improving inventory on time delivery, increasing savings in capital expenditures, increase in profitability, reduction of operational costs, reduction in the cost of poor quality, increase in productivity, reduction cycle time, reduction of customer complaints, improved sales and reduce inspection (antony et al., 2005, 2007, knowles et al. 2005). it is also a customer and data driven problem solving methodology and provide training and development to the organizations widely. although six sigma has been established highly successful in many industries and functional applications, one of the critical weaknesses of six sigma is the lack of a fundamental methodology for leveraging strategic and operational opportunities to drive the selection and execution of high priority projects. the six sigma approach also relies upon the existence of fundamental process capabilities and some levels of organizational maturity around the process. six sigma focuses on reducing the variability in order to improve the process, but it does not believe an early possibility to replenish them. six sigma is data dependent tools and techniques difficult to use in situations where data is not available or readily collected. so, if we come across at the strength and weakness of each of these methodologies it is apparent that they need each other. the strong fact-based, data-driven problem solving approaches that six sigma has helps to discover root causes. so it needs to develop integrated framework in turbulent environment to improve the supply chain performance and satisfy the customers. 5. research methodology in this paper authors has considered two methodologies from a management point of you, by investigating snag and opportunity. the approach to enable this study has been a qualitative one. the secondary data ranging from 2000 to 2014 has been studied to work upon the gaps and subsequently to identify the dimensions. on the basis of the identified dimensions, a conceptual framework has been proposed based on six sigma and ism by considering various supply chain dimensions i.e. human, process and output dimensions such as quality, cost etc. the section presents summary of some significant studies with reference to supply chain coordination. the various supply chain coordination issues ranging from administrative to technical i.e. management commitment, supplier selection, information processing, information technology, order fulfillment, edi, vmi etc. which help in coordination among various supply chain entities and improve supply chain performance as shown in table 2. 78 int. j. prod. manag. eng. (2015) 3(1), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international mishra, p. and kumar sharma, r. http://creativecommons.org/licenses/by-nc-nd/4.0/ no. authors year dimensions methodology m an ag er ia l s up po rt c oo rd in at io n in fo rm at io n te ch no lo gy su pp lie r s el ec tio n d is tr ib ut io n sy st em in fo rm at io n sh ar in g l ea de rs hi p o rd er fu lfi llm en t c us to m er s at is fa ct io n fe ed ba ck a nd r ew ar d v en do r m an ag em en t i nv en to ry tr ai ni ng a nd e du ca tio n c om m un ic at io n & c oo pe ra tio n fo re ca st in g q ua lit y m an ag em en t/i m pr ov em en t l ea d tim e se lf -m an ag em en t in no va tio n pe rf or m an ce te am fl ex ib ili ty e m pl oy ee p ar tic ip at io n pr oc es s im pr ov em en t o ri en ta tio n l ite ra tu re re vi ew c on ce pt ua l m od el / fr am ew or k c as e st ud y e m pi ri ca l a na ly si s 1 kanji and wallace 2000 √ √ √ √ √ √ √ √ 2 brah et al. 2000 √ √ √ √ √ √ √ √ √ 3 romano and vinelli 2001 √ √ √ √ √ √ 4 tracy et al. 2001 √ √ √ 5 chen and paulraj 2004 √ √ √ √ √ √ √ √ √ √ 6 lin et al. 2005 √ √ √ √ √ √ √ √ 7 sila and ebrahimpour 2005 √ √ √ √ √ √ √ √ √ √ √ √ √ 8 robinson and malhotra 2005 √ √ √ √ √ √ √ √ √ 9 burgess et al. 2006 √ √ √ √ √ √ 10 li et al. 2006 √ √ √ √ √ 11 li and lin 2006 √ √ √ √ √ √ √ √ 12 samat et al. 2006 √ √ √ √ √ √ √ 13 hsu et al. 2007 √ √ √ √ √ 14 koh et al. 2007 √ √ √ 15 ooi et al. 2007 √ √ √ √ √ √ √ √ 16 yousuf et al. 2007 √ √ √ √ √ √ √ √ √ 17 arshinder & deshmukh 2008 √ √ √ √ √ 18 kaynak and hartley 2008 √ √ √ √ √ √ √ √ √ √ 19 tews and tracey 2008 √ √ √ √ 20 chang 2009 √ √ √ √ √ √ √ √ 21 sit et al. 2009 √ √ √ √ √ √ √ 22 khang et al. 2010 √ √ √ √ √ 23 li et al. 2010 √ √ √ √ √ 24 wu and weng 2010 √ √ √ √ √ 25 srikantha et al. 2010 √ √ √ √ √ 26 chong et al. 2011 √ √ √ √ √ √ √ 27 mishra & sharma 2011 √ √ √ √ √ √ 28 singh 2011 √ √ √ √ √ √ √ √ √ 29 talib et al. 2011 √ √ √ √ √ √ √ √ √ √ √ 30 agus and hajinoor 2012 √ √ √ √ 31 borade and bansood 2012 √ √ √ √ √ 32 vanichchinchai 2012 √ √ √ √ 33 bala 2013 √ √ √ √ √ √ 34 luai and sawalha 2013 √ √ 35 karimi and rafiee 2014 √ √ √ √ √ √ table 2. dimensions of supply chain coordination. 79int. j. prod. manag. eng. (2015) 3(1), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international integration of six sigma and ism to improve supply chain coordination – a conceptual framework http://creativecommons.org/licenses/by-nc-nd/4.0/ the technical dimensions includes use of information technology, six sigma, e-business technology, jit and lean practices, use of edi, vmi for inventory management. the implementation of it technologies, such as edi, enterprise resource planning and crm systems can improve supply chain performance. it helps supply chain members to share information in real time (chong et al., 2011). 6. integrated framework of six sigma and ism for improving scc performance figure 2 presents an integrated framework of six sigma and ism for improving supply chain coordination (scc) performance. the framework consists of three phases i.e. (i) identification, (ii) prioritization, (iii) modeling. in order to achieve customer satisfaction, companies need coordination, both in internal process/functions and between different players in the supply chain network. as a consequence, the organization of the company must be considered in an integrated way through staff involvement and the sharing of objectives between the different dimensions. six sigma and ism focus on these fundamentals. thus the integrated framework is ideal for discovering a new convergent approach. it helps in achieving coordination between different dimensions such as human dimensions, process dimensions and output dimensions as it is described in literature as the most promising model for supply chain strategic decisions making. scm practices are defined as the set of activities undertaken by an organization to promote effective management of its supply chain and are recognized as one of the most http://dx.doi.org/10.4995/ijpme.2014.3150 received: 2014-06-29 accepted: 2014-12-15 creative commons attribution-noncommercial noderivatives 4.0 international int. j. prod. manag. eng. (2015) 3(1), ppp-ppp | 1 https://ojs.upv.es/index.php/ijpme supplier manufacturer distributors wholesaler retailer customer s c m d i m e n s i o n s f l o w o f i n f o r m a t i o n f l o w o f g o o d s s c m n e t w o r k sixsigma methodology interpretive structural modeling phase-i identification phase-ii prioritization phase-iii modeling i ii iii ism model micmac analysis prioritization of dimensions human dimensions process dimensions output dimensions structural equation modeling (sem) validation contextual relationship figure 2. integrated framework for effective scc with focus on dimensions. 80 int. j. prod. manag. eng. (2015) 3(1), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international mishra, p. and kumar sharma, r. http://creativecommons.org/licenses/by-nc-nd/4.0/ important areas for competitiveness and growth of the industries. in literature authors (gunasekaran, 2001; wang et al., 2004; knowles et al., 2005; li et al., 2011) developed various frameworks based on balanced scorecard (bsc), supply chain operations and reference model (scor), analytic hierarchy process (ahp), and maturity models to integrate and analyze scm dimensions. at the same time they stressed upon the need for development of integrated type of framework for measuring scm performance. due to growing importance of scm in global market, large business houses are heavily dependent on small to medium sized enterprises for good quality of products at low costs. thus, forcing the enterprises to imbibe & practice quality culture in their business activities which may help to minimize waste & reduce variability, improve consistency, and increase efficiency and productivity (gunasekaran, 2001; wang et al., 2004; li et al., 2011). the different steps of the framework are presented below in following paragraphs. phase-i: identification: phase–i consists of four sub steps i.e. scm network, scm process, scm dimensions. sc network: supply chain network is a coordinate system of organizations, people, activities, information and resources involved in moving a product or service in physical or virtual manner from supplier to customer. in the study a scn model is developed (figure 1) to show various entities such as suppliers, manufacturer, distributors, wholesaler, retailers and customers through which raw materials are acquired, transformed and are distributed to the customers. the objective of each entity is to make easy the schedule of materials from upstream to downstream and, in turn, deliver products to customers. in order to improve supply to the end customer, it is important to develop strong coordination and partnership within the scn. scm process: scm process depicts flow of goods/ flow of information in a network consisting of supplier, manufacturer, distributor, wholesaler, retailer and customer linked together via the feed forward flow of materials and the feedback flow of information. these entities are involved in providing and delivering final products from supplier’s to the customer’s in the whole process. scm dimensions: scm practices are a set of activities undertaken in an organization to promote effective management of various supply chain dimensions may be classified as (i) human, (ii) process and (iii) output dimensions. based upon critical review of literature, authors have identified various human dimensions such as (i) managerial support, (ii) information sharing, (iii) leadership, (iv) feedback & reward, (v) communication & co-operation, (vi) training & education, (vii) self-management, (viii) team flexibility, (ix) employee participation, (x) process improvement. process dimension of paint industry has been considered for improving the scm performance. output dimensions such as perfect order fulfillment, quality level and cost will be considered for measuring supply chain performance. phase-ii: prioritization: it consists of application of six sigma methodology to prioritization of dimensions. six sigma: six sigma methodology be required to be applied to reduce defect and improve job successful. by their nature, they appear to be systematically at predetermined processes to improve. hence, once the supply chain objectives are accomplished, they provide solutions at every stage of detail by eliminating the nonvalue added activities and reducing the process variability. the dynamic effects like changes in production rate, poor quality in raw materials and other effects related to the bullwhip behavior of a supply chain. six sigma completes a process, by aligning the strategic opportunities with the capability to execute them. phase -iii: modeling: it makes use of interpretive structural modeling (ism) framework to establish contextual relationships among scm dimensions and sem is used to validate the model. 6.1. interpretive structural modeling (ism) after that authors tried to measure and investigate the structural relationships between key dimensions for effective scc. it makes use of interpretive structural modeling (ism) framework to establish contextual relationships among scm dimensions. ism model developed after removing the transitivity as described in ism methodology. in order to determine driving power and dependence of each dimension micmac analysis has been done. micmac is “matrice d’ impacts croises multiplication appliquee a un classement”. the objective of cross-impact matrix multiplication applied to the classification (micmac) analysis is to analyze driving and dependence power of each dimension. according to singh et al., 2007; talib et al., 2011; and borade and 81int. j. prod. manag. eng. (2015) 3(1), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international integration of six sigma and ism to improve supply chain coordination – a conceptual framework http://creativecommons.org/licenses/by-nc-nd/4.0/ bansod 2012, micmac analysis is a classification based on driving power and dependence of each dimensions. although various dimensions are critical to organization competitiveness, research so far has tended to focus on scor, bsc, total quality management (tqm), activity based costing (abc), just in time (jit), etc., but in literature hardly any description of an integrated framework which prioritize and establish structural relationship between various dimensions using ism methodology is presented. as none of the previous research has grouped or analyzed sc dimensions critical to scc which can be structured into any framework or model for analyzing the structural relationship among them. however, the ism model needs to be statistically validated. the model can be validated with some empirical data and case studies with the help of structural equation modeling (sem). 7. conclusion the purpose of this paper is to propose a conceptual supply chain integrated framework based on six sigma and ism methodology to improve an entire supply chain performance. a three level framework for achieving the integration has been proposed which is in the form of (i) identification, (ii) prioritization, (iii) modeling. the focus of the six sigma and ism is to develop the concept of an integrated deployment models, refine the six sigma deployment methodology, add metrics, tools, technologies and capabilities through ism to address the relationship and integrate a highly structured six sigma an implementation methodology with the ism approach. the integration of six sigma is based on the ism supply chain strategic approach, simplified by the tools of the dmaic problem solving method. six sigma beliefs recommend the potential to refine current approaches to supply chain improvement. it offers likely benefits in delivering reduced variations over and above the elimination of waste and non-value added activity delivering by existing approaches. the application of six sigma methodology in the network model helps us to understand information regarding the strength and weaknesses of various supply chain entities in a supply chain. the model can be applied in practice by organizations in near future by following case study approach. hence, this framework allow the companies to evaluate their own processes effectively, compare their performance with others i.e. companies both within and outside their industry segment. main contribution of this study is that framework developed will be very useful for the management in selecting suitable dimensions for cultivating coordination in a supply chain and getting competitive advantage. the present model can be statistically validated with use of structural equation modeling (sem) which has the ability to test the validity of such models. the model also can be validated with some empirical data and case studies. acknowledgements the authors would like to acknowledge the valuable support received from editor-in-chief and anonymous reviewers who have helped us with their comments to improve the manuscript in present form. references ahuja,v., yang, j., shankar, r. (2009). benefits of collaborative ict adoption for building project management, construction innovation, 9(3): 323-340. http://dx.doi.org/10.1108/14714170910973529 agus, a., hajinoor m. s. (2012). lean production supply chain management as driver towards enhancing product quality and business performance, case study of manufacturing companies in malaysia, international journal of quality & reliability management, 29(1): 92-121. http://dx.doi.org/10.1108/02656711211190891 antony, j., kumar, m., tiiwari, m.k. (2005). an application of six sigma methodology to reduce the engine-overheating problem in an automotive company, proceedings of the institution of mechanical engineers, part b. journal of engineering manufacturer, 219(8): 633-46. http://dx.doi.org/10.1243/095440505x32418 antony, j., antony, f., kumar, m. (2007). six sigma in service organizations: benefits, challenges and difficulties, common myths, empirical observations and success factors, international journal of quality & reliability management, 24(3): 294-311. http://dx.doi. org/10.1108/02656710710730889 arshinder, k. a., deshmukh, s. g. (2008). supply chain coordination: perspective, empirical studies and research directions, international journal of production economics, 115(2): 316-335. http://dx.doi.org/10.1016/j.ijpe.2008.05.011 bala, h. (2013). the effects of it-enabled supply chain process change on job and process outcomes: a longitudinal investigation, journal of operations management, 31: 450-473. http://dx.doi.org/10.1016/j.jom.2013.07.014 82 int. j. prod. manag. eng. (2015) 3(1), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international mishra, p. and kumar sharma, r. http://dx.doi.org/10.1108/14714170910973529 http://dx.doi.org/10.1108/02656711211190891 http://dx.doi.org/10.1243/095440505x32418 http://dx.doi.org/10.1108/02656710710730889 http://dx.doi.org/10.1108/02656710710730889 http://dx.doi.org/10.1016/j.ijpe.2008.05.011 http://dx.doi.org/10.1016/j.jom.2013.07.014 http://creativecommons.org/licenses/by-nc-nd/4.0/ borade, a. b., bansod, s. v. (2012). interpretive structural modelingbased framework for vmi adotion in indian industries, international journal of advances manufacturing technology, 58: 1227-1242. http://dx.doi.org/10.1007/s00170-011-3458-4 brah, s. a., wong, j. l., rao, b. m. (2000). tqm and business performance in the service sector: a singapore study, international journal of operations and production management, 20(11): 1293-1312. http://dx.doi.org/10.1108/01443570010348262 burgess, k., singh, p.j., koroglu, r. (2006). supply chain management: a structured literature review and implications for future research, international journal of operations & production management, 26(7): 703-29. http://dx.doi.org/10.1108/01443570610672202 cao, n., zhang, z., to, k.m., ng, k.p. (2008). how are supply chains coordinated? an empirical observation in textile apparel business, journal of fashion marketing & management, 12(3): 384-97. http://dx.doi.org/10.1108/13612020810889326 chang g. (2009). total quality management in supply chain, international business research, 2(4): 81-85. chang s.i., yen, d.c., chou, c.c., wu, h.c., lee, h.p. (2012). applying six sigma to the management and improvement of production planning procedure’s performance, total quality management & business excellence, 23(3-4): 291-308. http://dx.doi.org/10.1080/1 4783363.2012.657387 chen, i.j., paulraj, a. (2004). towards a theory of supply chain management: the constructs and measurements, journal of operations management, 22(2): 119-50. http://dx.doi.org/10.1016/j.jom.2003.12.007 chong, a.y.l., chan, f.t.s., ooi, k.b., sim, j.j. (2011). can malaysian firms improve organizational/innovations performance via scm, industrial management and data systems, 111(3): 410-431. http://dx.doi.org/10.1108/02635571111118288 chopra, s., meindl, p. (2007). supply chain management: strategy, planning, and operation, prentice-hall. dasgupta, t. (2003). using the six –sigma metric to measure and improve the performance of a supply chain, total quality management & business excellence, 14(3): 355-366. http://dx.doi.org/10.1080/1478336032000046652 gunasekaran, a., patel, c., tirtiroglu, e. (2001). performance measures and metrics in a supply chain environment, international journal of operations & production management, 21(1-2): 71-87. http://dx.doi.org/10.1108/01443570110358468 harrison, a., van hoek, r. (2008). logistics management and strategy: competing through the supply chain, london: prentice-hall. hervani, a., helms, m.m., sarkis, j. (2005). performance measurement for green supply chain management, benchmarking: an international journal, 12(4): 330-353. http://dx.doi.org/10.1108/14635770510609015 hsu, m.h., ju, t. l., yen, c.h., chang, c.m. (2007). knowledge sharing behavior in virtual communities: the relationship between trust, self-efficacy, and outcome expectations, international journal of human computer studies, 65: 153-169. http://dx.doi.org/10.1016/j. ijhcs.2006.09.003 jharkharia, s., shankar, r. (2005). it-enablement of supply chains: understanding the barriesrs, the journal of enterprise information management, 18(1): 11-27. http://dx.doi.org/10.1108/17410390510571466 karimi, e., rafiee, m. (2014). analyzing the impact of supply chain management practices on organizational performance through competitive priorities (case study: iran pumps company), international journal of academic research in accounting, finance and management sciences, 4(1): 1–15. http://dx.doi.org/10.6007/ijarafms/v4-i1/503 kanji, g.k., wallace, w. (2000). business excellence through customer satisfaction, total quality management, 11(7): 979–998. http:// dx.doi.org/10.1080/09544120050135515 kaynak, h., hartley, j.l. (2008). a replication and extension of quality management intothe supply chain, journal of operations management, 26(4): 468-489. http://dx.doi.org/10.1016/j.jom.2007.06.002 khang, t.s., arumugam, v., chong, a.y-l., chan f. t.s. (2010). relationship between supply chain management practices and organizations performance: a case study in the malaysian service industry, international journal of modeling in operation management, vol.1(1): 84106. http://dx.doi.org/10.1504/ijmom.2010.035256 kim, b., oh, h. (2005). the impact of decisions making sharing between supplier and manufacturer on their collaboration performance, supply chain management: an international journal, 10(3-4): 223-236. http://dx.doi.org/10.1108/13598540510606287 koh, s.c. demirbag, m., bayraktar, e., tatoglu, e., zaim, s. (2007). the impact of supply chain management practices on performance of smes, industrial management & data systems, 107(1): 103-124. http://dx.doi.org/10.1108/02635570710719089 knowles, g., whicker, l., femat, j. h., canales, f. d. c. (2005). a conceptual model for the application of six sigma methodologies to supply chain improvement, international journal of logistics: research and applications, 8(1): 51-65. http://dx.doi. org/10.1080/13675560500067459 lau, j.s.k., huang, g.q., mak, k.l. (2004). impact of information sharing on inventory replenishment in divergent supply chains, international journal of production research, 42(5): 919– 941. http://dx.doi.org/10.1080/00207540310001628911 lee, h.l. (2000). creating value through supply chain integration, supply chain managemen review, 4(4): 30-36. li, s., ragu-nathan, b., ragu-nathan, t. s., rao, s. s. (2006). the impact of supply chain management practices on competitive advantage and organizational performance, international journal of management science, omega, 32): 107-124. http://dx.doi.org/10.1016/j. omega.2004.08.002 li, y., chang, k.c., chen h.g., jiang, j.j. (2010). software development team flexibility antecedents, journal system software, 83(10): 1726-1734. http://dx.doi.org/10.1016/j.jss.2010.04.077 83int. j. prod. manag. eng. (2015) 3(1), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international integration of six sigma and ism to improve supply chain coordination – a conceptual framework http://dx.doi.org/10.1007/s00170-011-3458-4 http://dx.doi.org/10.1108/01443570010348262 http://dx.doi.org/10.1108/01443570610672202 http://dx.doi.org/10.1108/13612020810889326 http://dx.doi.org/10.1080/14783363.2012.657387 http://dx.doi.org/10.1080/14783363.2012.657387 http://dx.doi.org/10.1016/j.jom.2003.12.007 http://dx.doi.org/10.1108/02635571111118288 http://dx.doi.org/10.1080/1478336032000046652 http://dx.doi.org/10.1108/01443570110358468 http://www.husdal.com/2007/09/01/book-review-logistics-management-and-strategy/ http://dx.doi.org/10.1108/14635770510609015 http://dx.doi.org/10.1016/j.ijhcs.2006.09.003 http://dx.doi.org/10.1016/j.ijhcs.2006.09.003 http://dx.doi.org/10.1108/17410390510571466 http://dx.doi.org/10.6007/ijarafms/v4-i1/503 http://dx.doi.org/10.1080/09544120050135515 http://dx.doi.org/10.1080/09544120050135515 http://dx.doi.org/10.1016/j.jom.2007.06.002 http://dx.doi.org/10.1504/ijmom.2010.035256 http://dx.doi.org/10.1108/13598540510606287 http://dx.doi.org/10.1108/02635570710719089 http://dx.doi.org/10.1080/13675560500067459 http://dx.doi.org/10.1080/13675560500067459 http://dx.doi.org/10.1080/00207540310001628911 http://dx.doi.org/10.1016/j.omega.2004.08.002 http://dx.doi.org/10.1016/j.omega.2004.08.002 http://dx.doi.org/10.1016/j.jss.2010.04.077 http://creativecommons.org/licenses/by-nc-nd/4.0/ li, l., su, q., chen x., (2011). ensuring supply chain quality performance through applying the scor model, international journal of production research, 49(1): 33-57. http://dx.doi.org/10.1080/00207543.2010.508934 lin, c., chow, w. s., madu, c. n., kuei, c.h., yu, p. p. (2005). a structural equation model of supply chain quality management and organizational performance, international journal of production economics, 96(3): 355-365. http://dx.doi.org/10.1016/j.ijpe.2004.05.009 li, s., lin, b. (2006). accessing information sharing and information quality in supply chain management, decision support systems, 42(3): 1641–1656. http://dx.doi.org/10.1016/j.dss.2006.02.011 luai e. jraisat, ihab h. sawalha (2013). quality control and supply chain management: a contextual perspective and a case study, supply chain management: an international journal, 18(2): 194–207. http://dx.doi.org/10.1108/13598541311318827 lummus, r.r., alber, k.l., (1997). supply chain management: balancing the supply chain with customer demand, the educational and resource foundation of apics, falls church, va. mandal, a., deshmukh, s.g. (1994). vendor selection using interpretive structural modelling, international journal of operations & production management, 14(6): 52-9. http://dx.doi.org/10.1108/01443579410062086 mishra p., sharma, r. k. (2011). reengineering the supply chain human dimensions using six-sigma framework, international journal for quality research, 5(3): 193-203. nabhani, f., shokri, a. (2009). reducing the delivery lead time in a food distribution smes through the implementation of six sigma methodology, journal of manufacturing technology management, 20(7): 957-974. http://dx.doi.org/10.1108/17410380910984221 ooi, k.b., bakar, n.a., arumugam, v., vellapan, l., loke, a.k.y. (2007). does tqm influence employee’s job satisfaction? an empirical case analysis, international journal of quality and reliability management, 24(1): 62-77. http://dx.doi.org/10.1108/02656710710720330 pande, peter s., neuman, robert p., cavanach, roland r., (2000), the six sigma way, mcgraw-hill, new york. robinson, c., malhotra, m. (2005). defining the concept of supply chain quality management and its relevance to academic and industrial practice, international journal of production economics, 96(3): 315–325. http://dx.doi.org/10.1016/j.ijpe.2004.06.055 romano, p., vinelli, a. (2001). quality management in supply chain perspectives, strategies and operative choices in a textile-apparel network. international journal of operations & production management, 21(4): 446-60. http://dx.doi.org/10.1108/01443570110381363 sage, a.p. (1977). interpretive structural modeling: methodology for large-scale systems, mcgraw-hill, new york: 91-164. samat, n., ramayah, t., saad, n.h. (2006). tqm practices, service quality, and market orientation – some empirical evidence from a developing country, management research news, 29(11): 713–728. http://dx.doi.org/10.1108/01409170610716025 sharma s.k., panda, b.n., mahapatra, s.s., sahu, s. (2011). analysis of barriers for reverse logistics: an indian perspective, international journal model optima, 1(2): 101-106. http://dx.doi.org/10.7763/ijmo.2011.v1.18 simatupang, t.m., wright. a.c., sridharan, r. (2002). the knowledge of coordination for supply chain integration, business process management journal, 8(3): 289-308. http://dx.doi.org/10.1108/14637150210428989 singh m.d., shankar, r., narain, r., agrawal, a. (2003). an interpretive structural modeling of knowledge management in engineering industries, journal of advances in management research, 1(1): 28-40. http://dx.doi.org/10.1108/97279810380000356 singh, r.k., garg, s.k., deshmukh, s.g. (2007). interpretive structural modeling of factors for improving competitiveness of smes, international journal of productivity and quality management, 2(4): 423-40. http://dx.doi.org/10.1504/ijpqm.2007.013336 singh, m., kant, r. (2008). knowledge management barriers: an interpretive structural modeling approach, international journal of management science engineering management, 3(2): 141-150. singh, rajesh, k. (2011). developing the framework for co-ordination in supply chain of sme’s, business process management journal, 17(4): 619-638. http://dx.doi.org/10.1108/14637151111149456 sila, i., ebrahimpour, m. (2005). an investigation of the total quality management survey based research published between 1989 and 2000: a literature review, international journal of quality and reliability management, 19(7): 902-970. http://dx.doi. org/10.1108/02656710210434801 sit, w., ooi, k.b, lin, b., chong, a.y.l. (2009). tqm and customer satisfaction in malaysia’s service industry, industrial management & data systems, 109(7): 957-975. http://dx.doi.org/10.1108/02635570910982300 soti, a., shankar, r., kaushal, o.p. (2010). modeling the enablers of six sigma using interpretive structural modeling, journal of modelling in management, 5(2): 124-141. http://dx.doi.org/10.1108/17465661011060989 swartwood, d. (2001). how to scor with six sigma, (supply chain council). talib, f., rahman, z., qureshi, m.n. (2011). an interpretive structural modeling approach for modeling the practices of total quality management in service sector, international journal of modeling in operation management, 1(3): 223-250. http://dx.doi.org/10.1504/ ijmom.2011.039528 tanco, m., mateo, r., santos, j., jaca, c., viles, e. (2012). on the relationship between continuous improvement programs and their effect on quality defects: an automotive case study, total quality management & business excellence, 23(3-4): 277-290. http://dx.doi.org/1 0.1080/14783363.2011.637779 84 int. j. prod. manag. eng. (2015) 3(1), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international mishra, p. and kumar sharma, r. http://dx.doi.org/10.1080/00207543.2010.508934 http://dx.doi.org/10.1016/j.ijpe.2004.05.009 http://dx.doi.org/10.1016/j.dss.2006.02.011 http://dx.doi.org/10.1108/13598541311318827 http://dx.doi.org/10.1108/01443579410062086 http://dx.doi.org/10.1108/17410380910984221 http://dx.doi.org/10.1108/02656710710720330 http://dx.doi.org/10.1016/j.ijpe.2004.06.055 http://dx.doi.org/10.1108/01443570110381363 http://dx.doi.org/10.1108/01409170610716025 http://dx.doi.org/10.7763/ijmo.2011.v1.18 http://dx.doi.org/10.1108/14637150210428989 http://dx.doi.org/10.1108/97279810380000356 http://dx.doi.org/10.1504/ijpqm.2007.013336 http://dx.doi.org/10.1108/14637151111149456 http://dx.doi.org/10.1108/02656710210434801 http://dx.doi.org/10.1108/02656710210434801 http://dx.doi.org/10.1108/02635570910982300 http://dx.doi.org/10.1108/17465661011060989 http://dx.doi.org/10.1504/ijmom.2011.039528 http://dx.doi.org/10.1504/ijmom.2011.039528 http://dx.doi.org/10.1080/14783363.2011.637779 http://dx.doi.org/10.1080/14783363.2011.637779 http://creativecommons.org/licenses/by-nc-nd/4.0/ tracey, j.b., hinkin, t.r., tannenbaum, s., mathieu, j.e. (2001). the influence of individual characteristics and the work environment on varying levels of training outcomes, human resource development quarterly, 12 no 1): 5-23. http://dx.doi.org/10.1002/15321096(200101/02)12:1<5::aid-hrdq2>3.0.co;2-j tews, m. j., tracey, j. b. (2008). an empirical examination of post training on-the-job supplements for enhancing the effectiveness of interpersonal skills training, personnel psychology, 61(2): 375–401. http://dx.doi.org/10.1111/j.1744-6570.2008.00117.x vanichchinchai, a. (2012). the relationship between employee involvement, partnership management and supply performance: findings from a developing country, international journal of productivity and performance management, 61(2):157-172. http://dx.doi. org/10.1108/17410401211194662 warfield, j.w. (1974). developing interconnected matrices in structural modeling, ieee transactions on systems, man and cybernetics, 4(1): 51-81. wang, g., huang, s. h., dismukes, j. p. (2004). product –driven supply chainselection using integrated multi-criteria decision-making methodology, international journal of production economics, 91(1): 1-15. http://dx.doi.org/10.1016/s0925-5273(03)00221-4 wei, l., yi-zhong, m.a. (2013). a framework for using six-sigma metrics to evaluate the performance of supply chain, international conference on management science & engineering (20th) july 17-19. wu m.y., weng, y.c. (2010). a study of supplier selection factors for high-tech industries in the supply chain, total quality management & business excellence, 21(4): 391-413. http://dx.doi.org/10.1080/14783361003606662 yusuf, y., gunasekaran, a., dan, g. (2007). implementation of tqm in china and organizational performance: an empirical investigation, total quality management, 18(5): 509-530. http://dx.doi.org/10.1080/14783360701239982 yousef, a., lee, l., sang-heon, l., ashraf, m. a. (2008). optimizing order fulfillment using design for six sigma and fuzzy logic, international journal of management science and engineering management, 3(2): 83-99. xu, l., beamon, b. (2006). supply chain coordination and cooperation mechanisms: an attribute based approach, the journal of supply chain management, 42(1): 4-12. http://dx.doi.org/10.1111/j.1745-493x.2006.04201002.x 85int. j. prod. manag. eng. (2015) 3(1), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international integration of six sigma and ism to improve supply chain coordination – a conceptual framework http://dx.doi.org/10.1002/1532-1096(200101/02)12:1<5::aid-hrdq2>3.0.co;2-j http://dx.doi.org/10.1002/1532-1096(200101/02)12:1<5::aid-hrdq2>3.0.co;2-j http://onlinelibrary.wiley.com/doi/10.1111/peps.2008.61.issue-2/issuetoc http://dx.doi.org/10.1111/j.1744-6570.2008.00117.x http://dx.doi.org/10.1108/17410401211194662 http://dx.doi.org/10.1108/17410401211194662 http://dx.doi.org/10.1016/s0925-5273(03)00221-4 http://dx.doi.org/10.1080/14783361003606662 http://dx.doi.org/10.1080/14783360701239982 http://dx.doi.org/10.1111/j.1745-493x.2006.04201002.x http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2018.8758 received 2017-10-15 accepted: 2018-01-17 derailed locomotive? petrobras investments and economic growth in brazil raíssa fernandes yabiko*and rosemarie bröker bone industrial engineering department (dei) – polytechnic school of the federal university of rio de janeiro (ufrj). athos da silveira ramos avenue, 149/room f106-6 cidade universitária, rio de janeiro/rj, zip code 21941-909, brazil * rayabiko@poli.ufrj.br abstract: petrobras is the largest firm in brazil and one of the largest in the world. its investment plans are among the biggest in the oil and gas industry, focused in brazil and on e&p. petrobras is responsible for a large share of gross capital formation and gross domestic product (gdp) growth in the country. the correlation between its investments and the country investment and gdp growth is above 0.8 and shows the dependency of the economy to petrobras activity. at the same time, as a state enterprise it has been a tool of macroeconomic policy. in the 2010´s its gasoline and diesel prices were frozen to keep inflation down. the recent crisis in the company, including corruption scandals and oil price slump increased debt levels and reduced its capital expenditures. the sale of assets directive since 2016 is required to reduce its net debt. while a medium to long term survival strategy, the change in petrobras’ investment profile may decrease the prospects of gdp growth in the brazilian economy. key words: brazil, oil sector, petrobras, investment, gross domestic product, gross capital formation. 1. introduction created in 1953 as sole oil producing firm in brazil, petrobras has been responsible for the development of the oil and gas industry along the vertical chain, including exploration and production (e&p), refining, transportation, retail and all links of natural gas chain. it is one of the largest firms in the world oil and gas industry and part of the ‘new seven sisters’ – domestic oil producers that currently challenge the industry role of the international conglomerates of the ‘seven sisters’ (hoyos, 2007). even the oil sector deregulation in the 1990´s did not reduce petrobras expansion. by 2010, with the prospect of the newly found pre-salt reserves, it presented the largest investment plan, with over usd 220 billion capital expenditures (petrobras, 2011). this investment plan was backed by a regulatory change (law no. 12351/2010 that settled the terms of a shared oil partition regime) that provided petrobras a leading (if not exclusive) role in developing the pre-salt area. in addition, petrobras controlling bondholder, the brazilian government, pushed for one of the largest equity issue in history doubling petrobras capital to usd 223 billion, making it the second largest oil firm at the time (mf, 2010). the raised capital would finance the required investments for pre-salt exploration and production (e&p) and keep financial solvency indicators within investment grade level. nevertheless, by 2015 the firm and the country situation was in stark contrast. posting a loss of usd 8 billion including a significant write-off due to cite this article: yabiko, r.f., bone, r.b. (2018). derailed locomotive? petrobras investments and economic growth in brazil. international journal of production management and engineering, 6(1), 47-55. https://doi.org/10.4995/ijpme.2018.8758 int. j. prod. manag. eng. (2018) 6(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international 47 mailto:rayabiko@poli.ufrj.br http://creativecommons.org/licenses/by-nc-nd/4.0/ to corruption overcharges from suppliers, petrobras faced a second year of losses in a row. the corruption scandal and investigation known as car-wash (“lava jato”) involved directors and suppliers of petrobras. its effects were deeply felt in the administration and put great investments as the petrochemical complex of rio de janeiro in check. not only brazil itself faced an unprecedented recession with a gross domestic product per capita fall of 4.6% (ibge, 2017). given the size and role of petrobras in a key strategic sector for a country economy and development, it is not surprising that its woes may influence the country and may as well help it leave recession. this paper explores the relationship between petrobras investment outlays and the country economic welfare through indicator such as gross capital formation and gross domestic product (gdp), highlighting the company impact in brazil development, positive or negative. the study investigates the profile of petrobras investments dividing them by areas. the change in investments across activities over the last decade was not homogeneous or in line with the goals of an oil and gas (o&g) corporation. the lack of cohesion between what should be the company strategy and the effect of its role in economic growth in the country further understanding. the article is organized as follows. the next section presents the evolution of petrobras investments in total volume and broken down by activities over the oil and gas vertical chain, revealing the strategic choices made over time and the monetary impact of these actions in the company welfare. the third section highlights the role of petrobras investment on economic growth and evaluates the effect of the strategic investment choices current and future economic growth. the analysis course was to use the economy indicators to correlate then with the company investments growth. the last section collects concluding comments and possible forecasts to both brazil and petrobras. 2. petrobras investments the goal of this section is to the analyise petrobras investment levels and trends and its relationship with oil refined goods prices and their consequences to firm indebtness. investment is taken here as gross fixed capital formation, net of disinvestments. investment itself is broken down in business segments, namely exploration and production (e&p); refining, transportation and marketing (supply); gas and energy (gas & energy); distribution; biofuel; corporate; and international. the first aspect to be analyzed is the volume of petrobras total investment in the last eleven years, from 2005 to 2016. this period comprehends several critical changes in the scenario, since the discovery of pre-salt reserves, the global financial crisis, the brazilian recession, up to the launch of lava-jato investigation. petrobras investment information were collect-ed from the f-20 forms presented to the securities and exchange commission (sec) in the united states of america. $0 $10 $20 $30 $40 $50 $60 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 in b ill io ns u sd petrobras historical investment series 2005-2016 figure 1. petrobras historical investment series (total volume) 2005-2016 (source: sec form-20f, 2006-2016 and petrobras, 2017). the information displayed in figure 1 shows that petrobras overall investment did not decrease in the face of the 2008-2009 global financial crisis. investment reached a first peak in 2010 with an amount of usd 45 billion and this magnitude order were maintained up to 2012. looking closely, the year of 2011 saw a 6% decrease in total investment and 2012 a modest 1% increase in relation to the previous year, keeping it virtually frozen. in 2013 a new high was reached with over usd 50 billion spent on investment. this huge increase happened in the wake of a management change that replaced the presidency and directors. the decay began in 2014 and were prolonged until last year. in 2015 there is a decrease in total investment when the level was close to the 2007 level, reaching only usd 23 billion. the investment decrease between 2014 and 2015 reflected the fall in international oil prices, the increasing burden of debt at petrobras. the loss of its international rating agencies (fitch, moody´s and standard and poor´s) investment grade in 2015 and the brazilian real int. j. prod. manag. eng. (2018) 6(1), 47-55 creative commons attribution-noncommercial-noderivatives 4.0 international yabiko, r. f. and bone, r. b. 48 http://creativecommons.org/licenses/by-nc-nd/4.0/ (r$) devaluation from 2014 that made investment financing harder and the company more selective in its capital expenditures (capex). for many years the brazilian government imposed on petrobras a populist agenda of not passing through international oil price increases to gasoline and diesel. this generated losses for petrobras, as the sale price of gasoline was not able to cover the refining, extraction and production costs for most of the time between 2010-2014 when oil prices increased after the 2009 drop. to illustrate the contrasting scenario in the country figure 2 compares the trend of brent oil prices with the percentage variation in the conventional gasoline and diesel distribution price in brazil and usa (ny market). brent oil price is an international oil price benchmark for internationally traded oil. it is measured at the north sea production and used here given the increasing domestic influence of shale oil and gas on the other international reference price, the western texas intermediate (wti). refined goods gasoline and diesel new york prices are taken as reference for international competitive prices, while brazil prices are national averages registered by the oil regulatory agency (anp). the trends determined by the commodity international prices (brent oil prices) clearly determine refined products such as gasoline and diesel in international markets, as seen by the new york benchmark. the trend seen in internal markets in brazil is markedly different. the gasoline and diesel pricing followed petrobras main controller – the government of brazil– macroeconomic policies, such as inflation control. international oil price changes pass through to refined goods prices are smoothed over a very long term and show often different trends. given the inflation surge in 20092012 gasoline and diesel prices were frozen as an anti-inflationary policy, placing a heavy burden on petrobras cash flow. it was laid in petrobras shoulders to carry these police of inflation control. additionally, another important event to petrobras woes is the launch of the car wash corruption and embezzlement investigation in 2014. this put board decisions to a near halt from 2014 and drained the company forces to keep up investments levels. figure 3 is a clear reflection of the chaotic scenario faced by petrobras since 2010. it presents the evolution of the debt in the firm by showing the ratio of the net debt to ebtida. net debt is the amount of money the company needs to clear the liability that generates financial expense. ebitda stands for earnings before taxes, interest depreciation and amortization –a measure of operational profits and cash generation (damodaran, 1997). this indicator gives a clear view of petrobras financial health. -60,0% -50,0% -40,0% -30,0% -20,0% -10,0% 0,0% 10,0% 20,0% 30,0% 40,0% 50,0% 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 brent, conventional gasoline and diesel: brazil vs new york percentage variation 2007-2016 conventional gasoline brazil diesel brazil conventional gasoline new york diesel new york brent oil prices figure 2. percentage variation between brent oil prices, conventional gasoline and diesel in brazil and usa (new york market) 2007-2016 (source: eia, 2017 and anp, 2017). int. j. prod. manag. eng. (2018) 6(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international derailed locomotive? petrobras investments and economic growth in brazil 49 http://creativecommons.org/licenses/by-nc-nd/4.0/ the sharp rise in its indebtedness, from 1.5 of ebitda to 5 times its earnings in five years raised an urgent change in policy within the company to avoid insolvency. the petrobras latest announced business set a goal to reducing its net debt to ebitda indicator to 2.5, leading to important changes in its investment policy as will be seen below. 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 evolution of petrobras leverage 2006 2016 net debt/ebitda 2018 goal figure 3. evolution of petrobras leverage: net debt/ ebitda (source: petrobras, 2017). figure 3, when compared with the level of investments, allows us to say whether the company indebtedness is due to capital expenditures and future growth perspectives or whether it is a symptom of say poor administration. compared to figure 1, we see that indebtness rose in the wake of largest investment levels. nevertheless the difference between oil cost growth and refined goods prices hurt ebitda and pushed indebtness up. aware of the need to overcome this situation of high levels of leverage, petrobras established in its last business plan (2017-2021) a goal to reduce its indebtedness to a 2.5 reason of net debt/ebitda. besides the company started to prioritize its expenditures to maximize the cash flow and to make a series of disinvestments to promote profit in short term (petrobras, 2017). now it is possible to tread the path to understand how petrobras found itself in the midst of such crisis. also, it is time to begin the questioning of how deep it goes the role of petrobras in the brazilian economy. 2.1. petrobras investments profile in this section the investment is broken down in business segments, namely, exploration and production (e&p), that includes the oil and gas exploration and production activities onshore and offshore; refining, transportation and marketing (supply), that includes the next activities in the oil and gas vertical chain once oil is extracted, namely refining and transportation and sale of crude oil; gas and energy (gas & energy), that includes all activities related to natural gas (ng), as ng sale, and thermal power generation and its commercialization; distribution, that involves oil refined products wholesale and retail; biofuel, that covers biodiesel and co-products; as well as two additional classification: corporate, which comprehend financial management and human resources activities; and international, that consolidates activities abroad. the overall trend that was analyzed first may hinder a better view of the investment profile at petrobras. figure 4 shows the amount of cash that was focused in each of those areas and their evolution over eleven years. $0 $5 $10 $15 $20 $25 $30 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 in b ill io ns u sd petrobras historical investment series: sectoral development 2005-2016 e&p supply gas & energy biofuels distribution corporate internacional figure 4. petrobras historical investment series: sectoral development 2005-2016 (source: sec form-20f 2006-2016 and petrobras, 2017). int. j. prod. manag. eng. (2018) 6(1), 47-55 creative commons attribution-noncommercial-noderivatives 4.0 international yabiko, r. f. and bone, r. b. 50 http://creativecommons.org/licenses/by-nc-nd/4.0/ a relative ranking of business segments was nevertheless maintained over time: e&p taking the largest share of investment outlays, followed by supply, gas & energy and international investments. biofuels, distribution and corporate were always very small compared to the others. this trend is expected from an oil and gas company that draws profits from the commercialization of oil (crude or refined products). it can also be explained by the fact that e&p is the most expansive link from the petroleum chain. a contributor factor to the peak in e&p in 2013 was the realization of the first bidding round of pre-salt fields, when the consortium led by petrobras was the winner. the most relevant change visible from figure 4 is the role of supply segment. from 2008 onwards, it followed the increase in e&p investments up to 2013. this year e&p investments were made priority and supply investments started decreasing. while in 2012 supply investment reached 37% of total investment, by 2015 it accounted for just 10% of total investment. figure 4 is net of disinvestments. the 2015 figures reflect the 2015-2019 petrobras business plan revision to the new the proposed growth for the next five years, as seen in the 2017-2021 petrobras business plan that included the sale of transportation units, gas retail business and even oil fields to manage the increasing debt that rose sharply from 2011 (petrobras, 2016 and 2017). in 2014 after the beginning of car-wash the investments in the international segment were not accounted for separately. since that year onwards, the investments abroad have been divided by company activity. the decrease in supply investments had important effects on the current refining capacity and the prospects for internalizing the refining of the pre-salt oil in brazil, as shown in yabiko, medeiros and bone (2016). even though e&p is the responsible for the principal cash flow in an oil company, the largest increase in value added is in refining the oil produced from these blocks. notwithstanding the level of total investment was reduced by nearly half from 2013 to 2016, e&p investment decreased by only a third and is at the 2008 level. supply investment itself decreased more than 50% from 2013 to 2016. this strategic choice of protecting e&p investment from such large cuts makes sense given that petrobras is an integrated oil and gas company and guaranteeing a level of proven and commercial oil reserves over time has a positive impact on stock prices (ribeiro, almeida and bone, 2017). in the first semester of 2017 the brazilian government announced the realization of two more bidding round in the pre-salt area (anp, 2017). however, due to petrobras economic situation is questionable if the company is going to be able to maintain its role in petroleum exploration in brazil and contribution to the national development. 3. economic growth and the role of petrobras since its creation in the 1950´s petrobras investment decisions have a significant impact in the brazilian economy. here we compare petrobras activities with brazil´s gross fixed capital formation (aggregate investment) and gross domestic product (gdp), allowing us to show petrobras role in the economy. 3.1. gross capital formation and petrobras investments gross fixed capital formation, or gross capital formation measures the aquisition of machinery and equipment as well as vehicles and the addition of building and construction in the economy. according to the brazilian institute of geography and statistics (ibge, 2017), this national accounts measure does not substract depreciation . figure 5 presents the time series analysis of petrobras total investment and gross capital formation for brazil, data collected from the (ibge) national accounts. -60,0% -40,0% -20,0% 0,0% 20,0% 40,0% 60,0% 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 petrobras investment and gross capital formation 2006-2016 petrobras investment gross capital formation figure 5. petrobras investment and gross capital formation in brazil 2006-2016 (source: ibge, 2017 and petrobras, 2017). int. j. prod. manag. eng. (2018) 6(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international derailed locomotive? petrobras investments and economic growth in brazil 51 http://creativecommons.org/licenses/by-nc-nd/4.0/ we see that petrobras investment and gross capital formation (or aggregate investment) have a similar trend from 2006 to 2016. at the same time, petrobras investment is more volatile than gross capital formation. this is expected since aggregate investment includes residential home building and infrastructure investment, taking in consideration the whole economy and not just one sector (oecd, 2015). looking closely, from figure 4, from 2006 to 2008 petrobras investment follows an equal trend of aggregate investment. justified by the fact that petrobras is an entity responsible for the largest investments in the country economy and its strategic role is to be a mechanism that encourages economy growth in brazil. in 2009, the year affected by global financial issues, the slowdown (but not decrease) in investments generated an echo in the gross capital formation, with a rebound in 2010. then in 2011-2013 we see a different trend between petrobras and aggregate investment. this found difference is a consequence of two major events that were held in brazil: the world cup and olympic games. these occasions shielded gross capital formation as strong investments in infrastructure were made and several foreign companies saw brazil as a country to invest. however, in 2014-2015 both investment indicators showing negative growth, more pronounced in petrobras line. in this year is when the effects of an economy crisis in brazil begin conjointly with the newly found but deep-rooted deficit in petrobras cash flow. in 2016 in both indicators it is possible to see a small recovery, the early signs of an economy rebirth. 3.2. gross domestic product and gross capital formation the role of investment on economic growth is well known (blanchard, 2011). investment multipliers, that is, the effect of a 1 growth in investment (with respect to gdp) on the growth of gdp, can be larger than 1 (oecd, 2015), so that a 1% increase in investment may lead to a more than 1% increase in gdp. this effect is relevant, in the face that investment is only part of total gdp (that includes consumption, government spending and net exports). the leading effect of investment on economic growth can be seen in figure 6. upward trends of investment growth are followed by upward trends in gdp growth and negative trends of investment are followed by downward trends in gdp. -20,0% -15,0% -10,0% -5,0% 0,0% 5,0% 10,0% 15,0% 20,0% 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 gross domestic product and gross capital formation 2006-2016 gross domestic product gross capital formation figure 6. gross domestic product and gross capital formation in brazil 2006-2016 (source: ibge 2017). the trajectory of the brazilian economy from 20052016 can be divided between before and after the 2009 financial crisis. up to 2008, the country benefited from the commodity price boom and experienced increasing gdp growth, only to meet a gdp decrease in 2009. countercyclical policies, including petrobras investment policy, and investment projects in 2010 proved effective but short-lived, as growth faltered, albeit in positive territory from 2011 to 2013. the exhausted public finance could no longer sustain the attempts at economic growth and gdp contracted. in 2016 gdp also felt the remerge of the national economy, although still in contraction. according to the latest forecast published by the international monetary fund (imf) it is expected that brazil will leave this recession to grow 0.7% in 2017 and 1.5% in 2018 in terms of gdp (imf, 2017). 3.3. correlation between petrobras investments and economic indicators this section aims to learn if petrobras investment decision, as the ‘locomotive’ of the brazilian economy, affected national growth, looking at the correlation matrix of petrobras investment and national gdp and gross fixed capital formation indicators. the correlation figures show how strong int. j. prod. manag. eng. (2018) 6(1), 47-55 creative commons attribution-noncommercial-noderivatives 4.0 international yabiko, r. f. and bone, r. b. 52 http://creativecommons.org/licenses/by-nc-nd/4.0/ is the relationship between the figures and the direct of ingluende whenther inverse or direct. for computing the data it was used used in the study was the pearson correlation coefficient (ρ). it is a measure of the linear correlation between two variables and varies between +1 (perfect positive correlation) and −1 (perfect negative correlation), when the value is close to zero, the variables analyzed are linear independents. gathering all data in the table 1, it was possible to use pearson formula and calculate the correlation between petrobras investments, gdp and gross capital formation, stated in table 2. table 1. annual rate of growth gross capital formation, gdp and petrobras investment. (source: ibge, 2017 and petrobras, 2017). gross domestic product gross capital formation petrobras investment 2006 4.0% 6.7% 39.8% 2007 6.1% 12.0% 46.2% 2008 5.1% 12.3% 46.8% 2009 -0.1% -2.1% 17.6% 2010 7.5% 17.9% 30.1% 2011 3.9% 6.8% -5.6% 2012 1.9% 0.8% 1.2% 2013 3.0% 5.8% 17.3% 2014 0.1% -4.2% -26.3% 2015 -3.8% -13.9% -38.1% 2016 -3.6% -10.2% -31.9% table 2. correlation coefficient: gross capital formation, gdp and petrobras investment. (source: ibge, 2017 and petrobras, 2016). gross domestic product gross capital formation petrobras investment gross domestic product 1.00 0.99 0.84 gross capital formation 0.99 1.00 0.86 petrobras investment 0.84 0.86 1.00 from table 1 it can be verified that brazilian gdp had it best year in 2010 with 7.5% per year. this positive result can be justified by advances in gross capital formation and petrobras investments. looking through petrobras point of view, the year of 2010 was not the most expressive in terms of growing, but the previous years with remarkable performance (2008 with 46.8% per year) made it investments echoed years later. from table 2, the correlation coefficient between investment growth and gdp growth is at 0.99, highlighting the relevance of gross capital formation for the economy. such high correlation is impressive given the short sample and the use of variables in growth rates, to avoid the spurious correlation problem (gujarati, 2001). petrobras investment and gross capital formation correlation is at 0.86. and the correlation coefficient between petrobras investment and economic growth is at 0.84, showing the relevance of the firm to the brazilian economy, since these values (between 0.7 and 0.9) are characteristics of a strong link between the variables. the role of petrobras in economic growth cannot be undermined. nevertheless, the changes in the profile of petrobras investment may reduce its role as ‘locomotive’ of the brazilian economy. while distribution and supply investment are strongly associated with local productive capacity and the use of domestic inputs in construction and equipment use, e&p investments use more international supplied services and relatively less local technology, including engineering services. even under strict local content requirements on e&p, petrobras has had difficulties meeting these requirements for lack of adequate suppliers in brazil. the reduction in the relative share of investment in supply may suggest a lesser role of petrobras in leading the economy. 4. concluding comments since its creation, petrobras has been considered a firm with a development public policy role, as the main shareholder used as a policy tool. the firm directed its efforts just for the oil and gas sector, but to the domestic manufacturing industry as a whole. the impact of the firm in the economy is visible and the technological advances and r&d expenditures influenced other sectors in the economy. the degree of association of the gross capital formation in the economy and petrobras investments were visible in the decade since 2005. investment is a key variable for economic growth. as part of aggregate demand, it increases an economy productive capacity and its multiplier effect has a significant impact in income (gdp) growth. the correlation coefficient between investment and gdp growth is higher than 90% for brazil. int. j. prod. manag. eng. (2018) 6(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international derailed locomotive? petrobras investments and economic growth in brazil 53 http://creativecommons.org/licenses/by-nc-nd/4.0/ at the same time the petrobras investment was more volatile than gross capital formation. petrobras investment is more exposed than the aggregate of the economy to sector effects such as international crude oil price volatility and is more exposed to factors that led to the decline in the observed investment level and the pro-posed growth for the next five years, as seen in the 2017-2021 petrobras´ business plan (petrobras, 2017). the firm is reacting to the effect of government deficits and recovering from the initially paralyzing effect from the car wash investigation. in this crisis scenario, the profile of petrobras investment changed. e&p investments increased its share in total investment significantly, while supply (such as refining and petrochemicals) investments decreased sharply. the e&p emphasis is in line with the needs of an integrated oil and gas company and its requirements to maintain and explore reserves and produce crude oil for domestic consumption at refineries or export. but at the same time, the e&p investments may have relatively lower impact in the economy than supply in-vestments. a longterm view would suggest a more balanced approach as the extracted oil will need refining and the more balanced in-vestment profile may lead to a more effective contribution to growth and the reduction of the unemployment in the economy. acknowledgements we acknowledge the support from labecopet/poli/ ufrj and comments and suggestions from eduardo pontual ribeiro (ie/ufrj). references anp. (2017). brazil rounds. rio de janeiro: agência nacional do petróleo, gás natural e biocombusíveis/coordenadoria de defesa da concorrência. available from: http://www.brasil-rounds.gov.br/round_p2/portugues_rp2/cronograma.asp [accessed: 09 sep 2017]. anp. (2017). relatório de defesa da concorrência. rio de janeiro: agência nacional do petróleo, gás natural e biocombusíveis/coordenadoria de defesa da concorrência. available from: http://www.anp.gov.br/wwwanp/precos-e-defesa/234-precos/levantamento-de-precos/868serie-historica-do-levantamento-de-precos-e-de-margens-de-comercializacao-de-combustiveis [accessed: 25 sep 2017]. blanchard, o. (2011). macroeconomia, 5th edition. são paulo: ed. pearson. côrrea, a. (2009) ‘crise muda perfil das exportações brasileiras’, brasil. bbc brasil, são paulo. available from: http://www.bbc.com/ portuguese/noticias/2009/09/090908_crise_exportacoes_ac_np [accessed: 01 sep 2017]. damodaran, a. (1997). avaliação de investimentos: ferramentas e técnicas para a determinação do valor de qualquer ativo, 1st edition. rio de janeiro: qualitymark. eia. (2017). petroleum & other liquids. washington: u.s. energy information administration. avalaible from: https://www.eia.gov/dnav/pet/ pet_pri_spt_s1_a.htm [accessed: 10 oct 2017] gujarati, d. n. (2001). basic econometrics, 4th edition. são paulo: mcgraw-hill company. hoyos, c. (2007). ‘the new seven sisters: oil and gas giants dwarf western rivals’, oil & gas. the financial times limited, são paulo. available from: https://www.ft.com/content/471ae1b8-d001-11db-94cb-000b5df10621 [accessed: 07 jan 2017]. ibge. (2017). formação bruta de capital fixo variação em volume: taxa acumulada ao longo do ano. instituto brasileiro de geografia e estatística (ibge), rio de janeiro. available from: http://seriesestatisticas.ibge.gov.br/series [accessed: 05 oct 2017]. ibge. (2017). produto interno bruto variação em volume: taxa trimestral. instituto brasileiro de geografia e estatística (ibge), rio de janeiro. available from: http://seriesestatisticas.ibge.gov.br/series [accessed: 01 oct 2017]. imf. (2017). brazil: at a glance. united states: international monetary fund. available from: http://www.imf.org/en/countries/bra [accessed: 10 oct 2017] ipea. (2017). taxa de câmbio nominal. rio de janeiro: instituto de pesquisa econômica aplicada (ipea). available from: http://ipeadata.gov. br/exibeserie.aspx?serid=38389 [accessed: 05 mar 2017]. mf. (2010). petrobras arrecada cerca de r$ 120 bilhões em maior capitação do mundo. ministério da fazenda (mf). available from: http://www.fazenda.gov.br/noticias/2010/setembro/petrobras-arrecada-cerca-de-r-120-bilhoes-em-maior-operacao-de-captacaodo-mundo oecd. (2015). oecd chapter 3, lifting investment for higher sustainable growth. economic outlook. volume 2015/1. paduan, roberta (2016) petrobras: uma história de orgulho e vergonha. rio de janeiro: ed. objetiva. presidência da república, casa civil, subchefia para assuntos jurídicos. (2010). lei nº 12.351/2010. available from: https://www.planalto. gov.br/ccivil_03/_ato2007-2010/2010/lei/l12351.htm [accessed: 12 dec 2016]. petrobras. (2017). plano estratégico e plano de negócios e gestão 2017-2021. rio de janeiro: petrobras. available from: goo.gl/nhx5dn [accessed: 10 nov 2017]. petrobras. (2016). plano de negócios e gestão 2015-2019. rio de janeiro: petrobras. available from: goo.gl/5vfils [accessed: 20 jul 2016]. int. j. prod. manag. eng. (2018) 6(1), 47-55 creative commons attribution-noncommercial-noderivatives 4.0 international yabiko, r. f. and bone, r. b. 54 http://www.brasil-rounds.gov.br/round_p2/portugues_rp2/cronograma.asp http://www.anp.gov.br/wwwanp/precos-e-defesa/234-precos/levantamento-de-precos/868-serie-historica-d http://www.anp.gov.br/wwwanp/precos-e-defesa/234-precos/levantamento-de-precos/868-serie-historica-d http://www.bbc.com/portuguese/noticias/2009/09/090908_crise_exportacoes_ac_np http://www.bbc.com/portuguese/noticias/2009/09/090908_crise_exportacoes_ac_np https://www.eia.gov/dnav/pet/pet_pri_spt_s1_a.htm https://www.eia.gov/dnav/pet/pet_pri_spt_s1_a.htm https://www.ft.com/content/471ae1b8-d001-11db-94cb-000b5df10621 http://seriesestatisticas.ibge.gov.br/series http://seriesestatisticas.ibge.gov.br/series http://www.imf.org/en/countries/bra http://ipeadata.gov.br/exibeserie.aspx?serid=38389 http://ipeadata.gov.br/exibeserie.aspx?serid=38389 http://www.fazenda.gov.br/noticias/2010/setembro/petrobras-arrecada-cerca-de-r-120-bilhoes-em-maior-operacao-de-captacao-do-mundo http://www.fazenda.gov.br/noticias/2010/setembro/petrobras-arrecada-cerca-de-r-120-bilhoes-em-maior-operacao-de-captacao-do-mundo https://www.planalto.gov.br/ccivil_03/_ato2007-2010/2010/lei/l12351.htm https://www.planalto.gov.br/ccivil_03/_ato2007-2010/2010/lei/l12351.htm http://goo.gl/nhx5dn http://goo.gl/5vfils http://creativecommons.org/licenses/by-nc-nd/4.0/ petrobras. (2011). plano de negócios 2011-2015. rio de janeiro: petrobras. available from: http://investidorpetrobras.com.br/pt/ comunicado-e-fatos-relevantes/plano-de-negocios-2011-2015 [accessed: 13 jan 2017] petrobras. (2017). resultados financeiros: ebitda, available from: http://www.investidorpetrobras.com.br/pt/resultados-financeiros/holding [accessed: 10 oct 2017] petrobras. (2017). alavancagem: dívida líquida. available from: http://www.investidorpetrobras.com.br/pt/divida/endividamento-ealavancagem [accessed: 05 oct 2017] ribeiro e.p., de almeida w.f., bone r.b. (2017) stock market firm value effects of research and development expenditures in the oil and gas industry. in: amorim m., ferreira c., vieira junior m., prado c. (eds) engineering systems and networks. lecture notes in management and industrial engineering. springer, cham. https://doi.org/10.1007/978-3-319-45748-2_7 sec. (2016). form 20-f annual report. washington: u.s. securities and exchange commission (sec). available from: https://www.sec. gov/archives/edgar/data/ [accessed: 20 jan 2017]. yabiko, r., medeiros, g., bone, r. b. (2016). petrobras’ investment projects in brazil under check-mate: the ghost refineries case. international joint conference cio-icieom-iie-aim (ijc 2016). int. j. prod. manag. eng. (2018) 6(1), 47-55creative commons attribution-noncommercial-noderivatives 4.0 international derailed locomotive? petrobras investments and economic growth in brazil 55 http://investidorpetrobras.com.br/pt/comunicado-e-fatos-relevantes/plano-de-negocios-2011-2015 http://investidorpetrobras.com.br/pt/comunicado-e-fatos-relevantes/plano-de-negocios-2011-2015 http://www.investidorpetrobras.com.br/pt/resultados-financeiros/holding http://www.investidorpetrobras.com.br/pt/divida/endividamento-e-alavancagem http://www.investidorpetrobras.com.br/pt/divida/endividamento-e-alavancagem https://doi.org/10.1007/978-3-319-45748-2_7 https://www.sec.gov/archives/edgar/data https://www.sec.gov/archives/edgar/data http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2019.8726 received: 2017-08-10 accepted: 2019-03-01 effect of routing flexibility on the performance of manufacturing system khan, w.u.a, ali, m.b a department of mechanical engineering, amu, aligarh, india. b mechanical engineering section, university polytechnic, amu, aligarh, india. a mohdali234@rediffmail.com, b wasifuk@rediffmail.com abstract: this work presented in this paper is based on the simulation of the routing flexibility enabled manufacturing system. in this study four levels of each factor (i.e. routing flexibility, system load conditions, system capacity and four part sequencing rules) are considered for the investigation. the performance of the routing flexibility enabled manufacturing system (rfems) is evaluated using three performance measures like make-span time, resource utilization and work-in-process. the analysis of results shows that the performance of the manufacturing system may be improved by adding in routing flexibility at the initial level along with other factors. however, the benefit of this flexibility diminishes at higher levels of routing flexibilities. key words: flexibility, flexible manufacturing system, routing flexibility, makespan, work-in-process, resource utilization. 1. introduction in the present global market, manufacturers are facing vastly competitive, complex and dynamic industrial environment. the manufacturing performance is not only governed by the price of the product but other factors such as flexibility, quality, and delivery also have grate importance. thus the researchers are focused on the additional advantages of the advance technologies like computer integration in manufacturing systems, automated material handling system, robotic arms and flexible manufacturing system (fms). the most important advantage of these systems that are taken by the manufacturers is inherent flexibilities in these systems gustavsson (1984). flexibility in any manufacturing system described as the ability of a system to react in an economic way for volume change, mix requirement, status of the machine and processing capabilities. there are several types of flexibilities mentioned in literature. sethi and sethi (1990) recognize flexibility as a multi-dimensional notion within the manufacturing domain. flexibility may be reactive or proactive in nature (gerwin, 1993). joseph and sridharan (2012) studied the effect of routing flexibility, sequencing flexibility and sequencing rules of a perfect flexible manufacturing system on different performance measures. the flexibility is broadly classified as hardware flexibility and software flexibility blackburn and millen (1986). the later type of flexibility refers to the routing flexibility where as the software flexibility refers as sequencing flexibility. in routing flexibility there are options for the parts to move to one machine or other. it exists when the machines are capable to perform different type of operations without major change in the machine setup. therefore in this paper we consider routing flexibility in place of sequencing flexibility. much of the work has been done on routing flexibility in the deterministic environment. the work focused the impact of routing flexibility with different performance measures in a stochastic environment of a routing flexibility enabled manufacturing system (rfems). various measures are used to evaluate to cite this article: khan, w.u., ali, m. (2019). effect of routing flexibility on the performance of manufacturing system. international journal of production management and engineering, 7(2), 133-143. https://doi.org/10.4995/ijpme.2019.8726 int. j. prod. manag. eng. (2019) 7(2), 133-143creative commons attribution-noncommercial-noderivatives 4.0 international 133 mailto:mohdali234@rediffmail.com mailto:mohdali234@rediffmail.com http://creativecommons.org/licenses/by-nc-nd/4.0/ the performance of rfems like make-span time, average resource utilization and work-in-process of parts. taguchi principle is used to design the simulation experiments and results are statistical analyzed. the objectives of this paper are: to examine the interaction among different factors as routing flexibility, system capacity, system load condition, and part sequencing rules in a perfect rfems in a stochastic environment. to find out the effect of various factors and their levels on the performance of routing flexibility enabled manufacturing system. this paper is organized as; section 2 represents the work background. description of proposed rfems is presented in section 3. a brief explanation of the operational logic of the rfems model is presented in section 4. the section 5 describes the experiment design and methodology of the work. results and discussion are presented in section 6 and 7 respectively. finally, conclusions are given in section 8. 2. background and motivation in the past, much of the work has been done on routing problems with deterministic manufacturing environment and different solutions are proposed for the effectual control of system. but, very few researchers addressed the impact of routing flexibility on fms with stochastic environment. some of the researches and their findings are discussed here. browne et al. (1984) defined routing flexibility is exposed when there is a breakdown of machine. they provide a good discussion on the routing flexibility and their impact on the manufacturing system. this reduces the lead-time and fractional decrease in the total job make-span using alternative routes. pankaj et al. (1991) incorporates the reliability of machines to study routing flexibility. zhao et al. (2001) considers genetic algorithm to the scheduling of flexible manufacturing systems with multiple routes. barad et al. (2003) stated that routing flexibility is the means of processing parts through different routes in the system. but the setup time is a significant part of the lead time. therefore routing flexibility shows significant impact on manufacturing lead time wahab (2005). there are various manufacturing flexibilities mentioned in the literature but chen et al. (2006) stated that all the manufacturing-related flexibilities are derived by the routing flexibility in the fms. wadhwa et al. (2008) worked over the impact of routing flexibility on system performance with various planning and control strategies in an fms. ali and wadhwa (2010) revels that, the increase in the routing flexibility level may not be treated as a key role in system performance enhancement. mehdi et al. (2013) presents how the meta-heuristics (i.e. ant colony optimization, genetic algorithms, simulated annealing tabu search etc.) are adopted to solve the alternative route selection problem with real time so as to decrease the congestion in the manufacturing system. hence we observe from literature review that introduction of flexibility, mainly routing flexibility, has been found to have helped firms in the reduction of lead time, bottlenecks and uncertainties. a stochastic model was developed by savsar and aldaihani (2012) which expresses the system by the use of study state differential equations that are solved by maple software and analyze performance measures of fms under various operational conditions. this model is useful for researchers to analyze a manufacturing system. rohit et al. (2016) were discussed the unforeseen situations in manufacturing systems like deadlock, machine breakdowns etc. and strived to overcome the impact of uncertainties. they also studied the scheduling of parts and their affect on the performance of manufacturing system. and developed an extrapolative schedule that takes care of the interruption in the system and maintain the performance of the manufacturing system. our study here explicitly find outs the impact of routing flexibility, system capacity, system load conditions and sequencing rules on the performance of a rfems with respect to make-span time, average resource utilization and work-in-process under stochastic manufacturing environment. 3. description of rfems this study is carried out on routing flexibility enabled manufacturing system under the stochastic environment. this system comprises of six flexible machines i.e. m1, m2, m3, m4, m5 and m6 with a dedicated input buffer with each machine and a load/ unload station (khan and ali, 2015). 3.1. part type six types of part are considered for processing in the system i.e. p1, p2, p3, p4, p5 and p6. the details for each part type are generated as described below: int. j. prod. manag. eng. (2019) 7(2), 133-143 creative commons attribution-noncommercial-noderivatives 4.0 international khan and ali 134 http://creativecommons.org/licenses/by-nc-nd/4.0/ each parst have five different operations. for different load conditions their mean and standard deviation as taken as follows: level load condition mean standard deviation 1 lfb 27.2 9.75 2 lub 27.2 15.15 3 lumbpt 27.2 8.62 4 lbmupt 27.2 9.23 3.2. modeling routing flexibility the five different operations were considered for processing each part. the flexible system is considered four routing flexibility levels i.e. rf0, rf1, rf2 and rf3 under stochastic environment. rf=0, means that there is one to one relationship between machine and the part i.e. there is no alternative route for the parts. at rf=1, one operation can be done on two machines i.e. there is 1 more alternative machine for the same job (in addition to the machine available at rf=0). at rf=2, for one operation there are three alternative machines i.e. there is 2 more machines are available for processing the same operation in addition to the first one. similarly for rf=3, 3 alternative machines are available any part or operation as shown in the figure 1. the makespan, resource utilization and work-in-process were considered as performance measures for processing 600 parts of 6 part types. 3.3. system capacity the size of the input buffer of the machines represents the capacity of the manufacturing system. four levels of the system capacity are considered in a way that, at an instant 30, 60, 90 and 120 parts present in system for processing. 3.4. system load conditions the four system load conditions are taken in the proposed manufacturing model for the simulation i.e. load unbalanced (lub), load full balanced (lfb), load balanced machine and unbalanced processing time (lbmupt) and load unbalanced machine and balanced processing time (lumbpt). in stochastic modeling the processing time may vary from one model to another with the influence of many factors, but in this paper it is assumed as figure 1. flow of part at different routing flexibility. int. j. prod. manag. eng. (2019) 7(2), 133-143creative commons attribution-noncommercial-noderivatives 4.0 international effect of routing flexibility on the performance of manufacturing system 135 http://creativecommons.org/licenses/by-nc-nd/4.0/ normally distributed. ozcan et al. (2010) ignored the travel time in the calculation of total task time. they considered task times as normal distribution in stochastic environment. the operation times with the given load conditions are given in tables 1–4. the mean and standard deviation of each load conditions are given along with the respected tables. table 1. system operating with lub. lub m1 m2 m3 m4 m5 m6 p1 73 60 30 20 25 208 p2 40 15 22 20 15 142 p3 45 19 26 10 17 117 p4 45 26 16 12 20 119 p5 15 50 19 26 12 122 p6 26 22 15 25 20 108 113 171 192 98 113 109 816 normal distribution: mean = 27.2, stand. deviation = 15.15 table 2. system operating with lfb. lfb m1 m2 m3 m4 m5 m6 p1 33 30 30 20 23 136 p2 40 20 32 25 19 136 p3 45 19 36 16 20 136 p4 40 25 30 25 16 136 p5 58 20 18 17 23 136 p6 26 22 20 33 35 136 136 136 136 136 136 136 816 normal distribution: mean = 27.2, stand. deviation = 9.75 table 3. system operating with lbmupt. lbmupt m1 m2 m3 m4 m5 m6 p1 38 30 30 22 22 142 p2 35 20 32 23 20 130 p3 40 24 36 20 18 138 p4 37 20 30 21 18 126 p5 58 18 18 18 23 135 p6 34 24 20 32 35 145 136 136 136 136 136 136 816 normal distribution: mean = 27.2, stand. deviation = 9.23 table 4. system operating with lumbpt. lumbpt m1 m2 m3 m4 m5 m6 p1 33 25 35 18 25 136 p2 35 25 32 20 24 136 p3 40 24 30 22 20 136 p4 38 27 25 31 15 136 p5 55 23 16 17 25 136 p6 25 20 20 35 36 136 128 123 154 123 143 145 816 normal distribution: mean = 27.2, stand. deviation = 8.62 3.5. sequencing rules the sequencing rules are applied over the machine input buffer queue so that the parts are selected on account of the applicable sequencing rule (sr). the opted sequencing rules are: first-come-first-served (fcfs): part that comes first in the machine buffer will be selected first for processing. shortest processing time (spt): part having shortest processing time among parts present in the machine buffer will process first. highest processing time (hpt): part having highest processing time among parts present in the machine buffer will process first. last-come-first-served (lcfs): part that comes last in the machine buffer will be selected first for processing. 3.6. performance measures the model is evaluated by considering make-span time, resource utilization and work-in-process as the performance measures (khan and ali, 2015) are. where: makespan time cmax = max ij ctij resource utilization /ru siui sii n i n 1 1 = = = | | work-in-process (wip) is referred to all parts and partly finished parts that are at different stages of the manufacturing process. 4. explanation of simulation model this illustrated model shows the material flow and information flow in the proposed routing flexibility enabled flexible manufacturing system (figure 2). primarily the parts were created and moved in the simulation model in controlled manner. the loading station gets the information from the unloading station to release a new part of same part type to the system. as the part releases from the loading station the attributes are attached in respect to the part type. on the basis of the part type, parts are moved to the potential machine for performing respective operation. then the information is obtained to know the machine buffer condition. if the buffer is capable of holding the part, the part in turn goes to the buffer of int. j. prod. manag. eng. (2019) 7(2), 133-143 creative commons attribution-noncommercial-noderivatives 4.0 international khan and ali 136 http://creativecommons.org/licenses/by-nc-nd/4.0/ that machine and the required attributes are assigned. if number of parts in the machine buffer are equal to the capacity of the buffer then the part finds another route. once all the buffers of the selected routes are full, the systems becomes blocked. the parts have to wait in machine buffer till machine completed its task. the next part moved into the machine for processing only when the in process part moved out from the machine. the next part enters on the bases of queue sequencing rules. parts after being processed on a particular information is obtained to know whether all the operation of that part has been completed or not. if all the operations are completed, then the part is moved to the store. if any operation is left to perform then the part goes to the particular machine, where attributes are assigned. once all operations are completed for a part then it is sent for the storage. the information is sent in the form of a signal to loading station that releases same part type from the controlled input system provided in model. by this way, a constant volume is maintained in the manufacturing system. this process will go on till all the parts turned over through the system. 5. experiment design and methodology arena simulation software is used for the experimentation of proposed manufacturing model. a number of experiments have been performed to find out the effects of routing flexibility, system figure 2. schematic diagram depicting the modeled rfms. int. j. prod. manag. eng. (2019) 7(2), 133-143creative commons attribution-noncommercial-noderivatives 4.0 international effect of routing flexibility on the performance of manufacturing system 137 http://creativecommons.org/licenses/by-nc-nd/4.0/ capacity, system load condition and sequencing rule on the performance of the system. make-span time, resource utilization and work-in-process are taken as the performance measure. 5.1. assumptions of model the aim of this work is to determine the effect of routing flexibility, system capacity, system load condition and sequencing rule on the performance of rfems under stochastic environment and to highlight the actual impact of these factors under different conditions. it is assumed that the processing time of the parts is considered as normally distributed with four different load conditions that are mentioned in the above section. the system capacity is controlled by maintaining the input buffer size of each machine. sequencing rules are employed over each queue of the machine individually. the makespan time, resource utilization and work-in-process are considered as the performance measure. each machine must process one part at an instance of time. total processing time also include set-up times. table 5 shows the four factors and their levels. table 5. details of factor and their level. s.no. factor factor level level id 1 routing flexibility (rf) 0 1 2 3 1 2 3 4 2 system capacity (sc) 30 60 90 120 1 2 3 4 3 system load (sl) lub lfb lumbpt lbmupt 1 2 3 4 4 sequencing rules (sr) fcfs spt hpt lcfs 1 2 3 4 in the above designed conditions the total number of experiment required to perform the interactive study of the given factors and their levels are 256. therefore, taguchi’s concept of design of experiment is used to establish the best possible combinations of the given factors and their levels that drop the number of experiments to 16 shown in table 6. table 6. combination details of factors and levels. exp. no. rf sc sl sr 1 0 30 lfb fcfs 2 0 60 lub spt 3 0 90 lumbpt hpt 4 0 120 lbmupt lcfs 5 1 30 lub hpt 6 1 60 lfb lcfs 7 1 90 lbmupt fcfs 8 1 120 lumbpt spt 9 2 30 lumbpt lcfs 10 2 60 lbmupt hpt 11 2 90 lfb spt 12 2 120 lub fcfs 13 3 30 lbmupt spt 14 3 60 lumbpt fcfs 15 3 90 lub lcfs 16 3 120 lfb hpt phadke (1989) stated that for make-span time the natural scale is not appropriate because it gives negative calculation, which is meaningless. to avoid the negative prediction we may use well known decibel scale. so as to minimize the sensitivity of noise factor, we have to maximize α, as: α= −10 log (make span)2 in regard to signal-to-noise (s/n) ratio, three types are described by taguchi, i.e., smaller thebest, larger the-best, and nominal-the-best. for the analysis of results smaller-the-best is considered in case of make-span time and work-in-process, while for resource utilization, the larger-the-best was considered. 6. experimental results the proposed rfems model is simulated with the above consideration. the model is simulated for 600 parts of 6 part types. the results are presented in table 7. optimal factor combination is also identified, which generates the best system performance. in this study, for obtaining best optimal combination of the factors and their levels we uses analysis of means (anom). int. j. prod. manag. eng. (2019) 7(2), 133-143 creative commons attribution-noncommercial-noderivatives 4.0 international khan and ali 138 http://creativecommons.org/licenses/by-nc-nd/4.0/ the anom is used to determine the optimal factor combinations for the designed rfems model, it may be defined as follows (phadke 1989): mjk = main factor effect for the kth level of factor j, i.e.: l jki t l 1 a = | where: j = the factor (i.e., routing flexibility, system capacity, system load condition, sequencing rules); k = factor level (i.e., 1, 2, 3, or 4); αjki = the s/n ratio of the factor j with level k; l = the time that factor j with level k appears in the simulation model (i.e., 4). 6.1. optimal factor combinations according to the results getting from anom, the mjk values for rfems with the three given performance measuring i.e. mst, wip and ru are presented in table 8. the s/n ratio for each of the optimal factor combination is signified by the maximum point on the graph as shown in the figure 3, 4 and 5. table 7. orthogonal array l 16 (4) with experimental results and calculated s/n ratios. exp. no. rf sc sl sr mst/min. s/n ratio (db) wip (%) s/n ratio (db) ru (%) ru ratio (db) 1 1 1 1 1 31397 -89.93 45.59 -33.17 0.43 7.25 2 1 2 2 2 19761 -85.91 42.58 -32.58 0.69 3.16 3 1 3 3 3 18347 -85.27 43.26 -32.72 0.74 2.59 4 1 4 4 4 17596 -84.90 42.88 -32.64 0.77 2.22 5 2 1 2 3 31897 -90.07 44.52 -32.97 0.42 7.35 6 2 2 1 4 18606 -85.39 42.10 -32.48 0.73 2.70 7 2 3 4 1 15161 -83.61 40.78 -32.20 0.89 0.95 8 2 4 3 2 14142 -83.01 41.42 -32.34 0.96 0.33 9 3 1 3 4 32797 -90.31 43.46 -32.76 0.41 7.63 10 3 2 4 3 18910 -85.53 40.82 -32.21 0.71 2.87 11 3 3 1 2 14836 -83.42 40.44 -32.13 0.91 0.77 12 3 4 2 1 13903 -82.86 43.77 -32.82 0.98 0.14 13 4 1 4 2 32534 -90.24 42.96 -32.66 0.41 7.55 14 4 2 3 1 14795 -83.40 41.06 -32.26 0.91 0.74 15 4 3 2 4 14266 -83.08 42.41 -32.54 0.95 0.36 16 4 4 1 3 14258 -83.08 42.30 -32.52 0.95 0.41 table 8. factor mean effects of matrix experiment factor level main effect applicable formula mst s/n (α) ratio (db) wip s/n (α) ratio (db) ru s/n (α) ratio (db) msf0 (α1+ α2+ α3+ α4)/4 -86.51 -32.78 3.80 msf1 (α5+ α6+ α7+ α8)/4 -85.52 -32.50 2.83 msf2 (α9+ α10+ α11+ α12)/4 -85.53 --32.48 2.85 msf3 (α13+ α14+ α15+ α16)/4 -84.95 -32.50. 2.27 msc1 (α1+ α5+ α9+ α13)/4 -90.14 -32.89 7.44 msc2 (α2+ α6+ α10+ α14)/4 -85.06 -32.38 2.37 msc3 (α3+ α7+ α11+ α15)/4 -83.84 -32.40 1.17 msc4 (α4+ α8+ α12+ α16)/4 -83.46 -32.58 0.77 msl1 (α1+ α6+ α11+ α16)/4 -85.45 -32.58 2.78 msl2 (α2+ α5+ α12+ α15)/4 -85.48 -32.73 2.75 msl3 (α3+ α8+ α9+ α14)/4 -85.50 -32.52 2.82 msl4 (α4+ α7+ α10+ α13)/4 -86.07 -32.43 3.40 msr1 (α1+ α7+ α12+ α14)/4 -84.95 -32.62 2.27 msr2 (α2+ α8+ α11+ α13)/4 -85.65 -32.43 2.95 msr3 (α3+ α5+ α10+ α16)/4 -85.99 -32.61 3.30 msr4 (α4+ α6+ α9+ α15)/4 -85.92 -32.61 3.23 int. j. prod. manag. eng. (2019) 7(2), 133-143creative commons attribution-noncommercial-noderivatives 4.0 international effect of routing flexibility on the performance of manufacturing system 139 http://creativecommons.org/licenses/by-nc-nd/4.0/ it is found in figure 3 that the best factor level combination with mst is rf3, sc4, sl1 and sr1. which shall be understood as the routing flexibility level 4, the system capacity 120, load fully balanced (lfb) and sequencing rule as fcfs. it is also evident from the figure 4 that the best factor level combination with wip as performance measure is rf2, sc2, sl4 and sr2. this shall be easily read as the routing flexibility level 3, the system capacity 60, load balanced on machine and unbalanced processing time (lbmupt) and sequencing rule as spt. it is shown in figure 5 that the best factor level combination with ru is rf0, sc1, sl4 and sr3. this may be read as the routing flexibility level 1, the system capacity 30, load balanced on machine and unbalanced processing time (lumbpt), and the sequencing rule is hpt. it is also very important to discuss the relative significance of different factors on the system. for this analysis of variance (anova) is implemented. main effect of each factor level (mst) -92 -90 -88 -86 -84 -82 -80 m rf 0 m rf 1 m rf 2 m rf 3 m sc 1 m sc 2 m sc 3 m sc 4 m sl 1 m sl 2 m sl 3 m sl 4 m sr 1 m sr 2 m sr 3 m sr 4 factor level s /n r at io figure 3. main effects of each factor level (mst). main effect of each factor level (wip) -33 -32.9 -32.8 -32.7 -32.6 -32.5 -32.4 -32.3 -32.2 -32.1 m rf 0 m rf 1 m rf 2 m rf 3 m sc 1 m sc 2 m sc 3 m sc 4 m sl 1 m sl 2 m sl 3 m sl 4 m sr 1 m sr 2 m sr 3 m sr 4 factor level s /n r at io figure 4. main effects of each factor level (wip). main effect of each factor level (ru) 0 2 4 6 8 m rf0 m rf1 m rf2 m rf3 m sc 1 m sc 2 m sc 3 m sc 4 m sl 1 m sl 2 m sl 3 m sl 4 m sr 1 m sr 2 m sr 3 m sr 4 factor level s /n r at io figure 5. main effects of each factor level (ru). int. j. prod. manag. eng. (2019) 7(2), 133-143 creative commons attribution-noncommercial-noderivatives 4.0 international khan and ali 140 http://creativecommons.org/licenses/by-nc-nd/4.0/ 6.2. analysis of variance the significant factors can be found out by implementing analysis of variance (anova). where, the relative importances of factors are exposed by the error variance. the higher f-value means greater importance. minitab statistical software is used to calculate the anova at confidence level of 95%. the simulated results of the mst, wip and ru of the system are taken from table 8 for preparation of the anova table (table 9). the importance of the factor is determined by the f-value phadke (1989). from the table 9 f-value of system capacity is maximum at mst and ru at the same time effect of routing flexibility is highest at wip. but the sequencing rule is having the least effect at all performance measures. 7. results and discussions taguchi’s design of experiment (doe) gives us a quick means to find out the behavior of different factors in a manufacturing system. it is established from the above analysis the best factors and their levels combination with make-span time is rf3, sc4, sl1 and sr1, that may read as the routing flexibility level 4, system capacity 120, load unbalanced (lub), and the sequencing rule as fcfs where as for resource utilization the best combination is rf0, sc1, sl4 and sr3, whereas rf2, sc2, sl4 and sr2 is the best combination for work-in-process measurement. in light of results obtained each of factor is discussed briefly in the following sections. 7.1. routing flexibility it is seen in figure 3 that there is a major effect of routing flexibility with the make-span time on the performance of rfems. it decreases with the increase in the level of routing flexibility. at rf=0 the parts moved through fixed route. routing flexibility is exploited as rf level increases, in result there is decrease in mst. figure 4 shows that as the routing flexibility increases the wip reduce. this phenomenon happens up to rf2 and then there is a slight increase at rf3. wip is maximum at rf=0, as the parts wait in machine buffers for processing, in so doing increasing the wip. it is also view from figure 5 that the ru reduces with the increase in rf level. this is because at lower levels of rf the parts are processed through a fixed route resulting decrease in ru. it is also observed from tablex9 that table 9. anova results showing at different outputs (rf). anova for means (mst) source df seq ss adj ss adj ms f p rf 3 16320178 16320178 5440059 4.28 0.132 sc 3 782739183 782739183 260913061 205.32 0.001 sl 3 3983854 3983854 1327951 1.05 0.486 sr 3 10964563 10964563 3654854 2.88 0.204 residual error 3 3812234 3812234 1270745 total 15 817820013 anova for means (wip) source df seq ss adj ss adj ms f p rf 3 5.9762 5.9762 1.9921 3.74 0.154 sc 3 16.0854 16.0854 5.3618 10.06 0.045 sl 3 4.5263 4.5263 1.5088 2.83 0.208 sr 3 2.4233 2.4233 0.8078 1.52 0.370 residual error 3 1.5983 1.5983 0.5328 total 15 30.6095 anova for means (ru) source df seq ss adj ss adj ms f p rf 3 0.046771 0.046771 0.015590 8.19 0.059 sc 3 0.602995 0.602995 0.200998 105.56 0.002 sl 3 0.010807 0.010807 0.003602 1.89 0.307 sr 3 0.023346 0.023346 0.007782 4.09 0.139 residual error 3 0.005712 0.005712 0.001904 total 15 0.689632 int. j. prod. manag. eng. (2019) 7(2), 133-143creative commons attribution-noncommercial-noderivatives 4.0 international effect of routing flexibility on the performance of manufacturing system 141 http://creativecommons.org/licenses/by-nc-nd/4.0/ impact of routing flexibility is highest when ru is taken as performance measure and followed by mst and wip. 7.2. system capacity it is evident from the figure 3 that there is a remarkable improvement in the mst of the proposed rfems. mst decreases with increase in system capacity. increase in the system capacity means more parts are permissible to move in the system resulting better load sharing so as mst is reduced. figure 4, shows that the wip is significantly reduced from sc1 to sc2 and then there is a marginal increase with the increase in system capacity level. it is also found from figure 5 that the average resource utilization increases with the increase in the capacity of the system. it is because at higher levels of sc resources constantly in working therefore ru increases. from the results table 9 shows system capacity effect most at mst and followed by ru and wip. 7.3. system load condition figure 3 show that, the rfems perform in a different way with different system load conditions with make-span time. it is maximum at lbmupt and minimum at lfb. from figure 4, it is clear that wip is maximum with lub and minimum with lbmupt system load condition. it is evident from figure 5, that ru is highest when lbmupt is taken and it is lowest when lub is takes as system load condition. this is so because when the model is simulated with lub then there is maximum load sharing on the machines therefore average resource utilization increases. it is also observed from table 9 that the effect of sl is significant by considering wip and then followed by ru and mst as performance measure. 7.4. sequencing rules sequencing rules have a little impact on the mst and ru performance of rfems as shown in figure 3 and 5. they are minimum when fcfs is taken into consideration. while figure 4 shows the minimum wip at spt. it is also observed from table 9 that effect of sequencing rule is most dominating by considering ru as performance measure and then mst and wip. 8. conclusions in this paper, discrete-event simulation model is used to analyze the impact of some factors i.e. routing flexibility, system capacity, system load condition and sequencing rule on the performance of a rfems. for experiment design taguchi’s doe framework is used and the results are analyzed statically. it is found that increase in the routing flexibility level is not the only means for the system performance improvement. by using the proposed model, the best possible levels of some other factors are also considered, i.e. system capacity, system load condition, and sequencing rule. in spite of whole of this study, there are some limitations exist that may be explored further. one of the key limitation of this study is the model cannot be used for all manufacturing domain. an extensive data set may be modeled for getting better results. the focus of this work is for the development of a demonstrative platform to show major areas of concern and key directions. keeping these limitations in mind future work can be undertaken by considering other flexibility types, number of machines, parts, operations, etc. references ali, m., wadhwa, s. (2010). the effect of routing flexibility on a flexible system of integrated manufacturing. international journal of production research, 48(19), 5691-5709. https://doi.org/10.1080/00207540903100044 barad, m., sapir, d. e. (2003). flexibility in logistic systems-modeling and performance evaluation. international journal of production economics, 85, 155-170. https://doi.org/10.1016/s0925-5273(03)00107-5 blackburn, j., millen r. (1986). perspectives on flexibility in manufacturing: hardware versus software, in andrew r. rachamadugu and t.j. schriber kusiak modeling and design of flexible manufacturing systems, 2, 2, , ed. elsevier science publishers, amsterdam, the netherlands, 116-117. browne j., dubois, d., rathmill, k., seth, s. p., stecke, k. e. (1984). classification of flexible manufacturing systems. the fms magazine 114. chan, f. t. s., bhagwat, r., wadhwa. s. (2006). increase in flexibility: productive or counterproductive? a study on the physical and operating characteristics of a flexible manufacturing system. international journal of production research, 44(7), 1431-1445. https://doi.org/10.1080/00207540500398959 int. j. prod. manag. eng. (2019) 7(2), 133-143 creative commons attribution-noncommercial-noderivatives 4.0 international khan and ali 142 https://doi.org/10.1080/00207540903100044 https://doi.org/10.1016/s0925-5273(03)00107-5 https://doi.org/10.1080/00207540500398959 http://creativecommons.org/licenses/by-nc-nd/4.0/ gerwin, d. (1993). manufacturing flexibility: a strategic perspective. international journal of management science, 39(4), 395-410. https://doi.org/10.1287/mnsc.39.4.395 gustavsson, s. o. (1984). flexibility and productivity in complex production processes. international journal of production research, 22, 801-808. https://doi.org/10.1080/00207548408942500 joseph, o. a., sridharan, r. (2012). effect of flexibility and scheduling decisions the performance of an fms: simulation modeling and analysis. international journal of production research, 50(7), 2058-2078. https://doi.org/10.1080/00207543.2011.575091 khan w. u., ali. m. (2015), effect of sequencing flexibility on the performance of flexibility enabled manufacturing system. international journal of industrial and systems engineering, 21(4), 474-498. https://doi.org/10.1504/ijise.2015.072731 mehdi s., zaki s., ahmed, h. (2013). real-time rescheduling metaheuristic algorithms applied to fms with routing flexibility. international journal of advanced manufacturing technology, 64(1-4), 145. https://doi.org/10.1007/s00170-012-4001-y ozcan u., kellegoz, t., toklu, b. (2010). a genetic algorithm for the stochastic mixed-model u-line balancing and sequencing problem. international journal of production research, 49(6), 1605-1626, https://doi.org/10.1080/00207541003690090 pankaj, c., mihkel, m. t. (1991). models for the evaluation of routing and machine flexibility. european journal of operational research, 60(2), 156-165. https://doi.org/10.1016/0377-2217(92)90090-v phadke, m. s. (1989). quality engineering robust design. englewood cliffs, nj: prentice-hall. rohit p., nishant s., arvind s. t. (2016). performance evaluation of flexible manufacturing system (fms) in manufacturing industries. imperial journal of interdisciplinary research, 2(3), 2454-1362 savsar, m., aldaihani, m. (2012). a stochastic model for analysis of manufacturing modules. applied mathematics and information science 6(3), 587-600. sethi, a. k., sethi, s. p. (1990). flexibility in manufacturing: a survey. international journal of flexible manufacturing systems, 2, 289-328. https://doi.org/10.1007/bf00186471 wadhwa, s., bhagwat, r. (1998). judicious increase in flexibility and decision automation in semi-computerized flexible manufacturing (scfm) systems. international journal, studies in informatics and control, 2(8), 329-342. wadhwa, s., ducq, y., mohammed, a., prakash, a. (2008). performance analysis of a flexible manufacturing system under planning and control strategies. international journal, studies in informatics and control, 17(3), 273-284. wahab, m. i. m. (2005). measuring machine and product mix flexibilities of a manufacturing system. international journal of production research, 43(18), 3773-3786. https://doi.org/10.1080/00207540500147091 zhao, c., wu, z. (2001). a genetic algorithm approach to the scheduling of fmss with multiple routes. international journal of flexible manufacturing systems, 13(1), 71-88. https://doi.org/10.1023/a:1008148313360 int. j. prod. manag. eng. (2019) 7(2), 133-143creative commons attribution-noncommercial-noderivatives 4.0 international effect of routing flexibility on the performance of manufacturing system 143 https://doi.org/10.1287/mnsc.39.4.395 https://doi.org/10.1080/00207548408942500 https://doi.org/10.1080/00207543.2011.575091 https://doi.org/10.1504/ijise.2015.072731 https://doi.org/10.1007/s00170-012-4001-y https://doi.org/10.1080/00207541003690090 https://doi.org/10.1016/0377-2217(92)90090-v https://doi.org/10.1007/bf00186471 https://doi.org/10.1080/00207540500147091 https://doi.org/10.1023/a:1008148313360 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2018.8762 received 2017-10-15 accepted: 2018-01-17 clustering product innovators: a comparison between conventional and green product innovators marc ponsa1, andrea bikfalvia2 and josep llacha3 department of business administration and product design. university of girona, campus montilivi, 17071 girona, spain a1 marc.pons@udg.edu, a2 andrea.bikfalvi@udg.edu, a3 josep.llach@udg.edu abstract: this paper aims at analysing firms implementing new products. based on a cluster analysis, three types of manufacturers have been identified representing different types of product innovators according to the competitiveness factors important for their business, environmentally sensitive new products, and a performance indicator, such as the share of turnover from new products. key words: ems, manufacturing companies, product innovation, environmental impact, spain. 1. introduction according to the united nations’ approach on sustainable development goals (un, 2016) a sustainable consumption and production helps to achieve overall development plans, reduces future economic-, environmentaland social costs, strengthens economic competitiveness and reduces poverty. innovation appears as one possible action in this direction. new products, in general, and new products sensitive towards improving environmental impact, in particular, can make a considerable contribution to the society. some examples of improved environmental impact refer to: reduction of health risks when in use, extended product lifetime, reduction of energy consumption when in use, reduction of environmental pollution when in use, easier to maintain or to retrofit, and improved recycling, redemption or disposal properties. recently, a review specifically focusing on green product innovation published by dangelico (2016) makes an important contribution by analysing 63 studies in the field. it is affirmed that “with regard to capabilities in common with conventional new product development, it would be interesting for future research to investigate whether there is a difference between gpi development and conventional new product development in terms of relative importance of these capabilities and in terms of their extent of use” dangelico (2016:574). the analysis responds not just to an academic goal and a broader scientific call verbalised by dangelico (2016), but also to a global institutional priority as the europe 2020 strategy targeting improved environmental impacts and boosted innovation. moreover, using three countries’ data we contribute to other –few– data-driven approaches that combine environmental and innovation policy, translated to companies’ daily operations. 2. objectives the objective of this exploratory work is to characterise patterns of product innovative manufacturing companies distinguishing between green product innovators (gpi) and conventional product innovators (cpi). for this purpose, we proceed with a cluster classification process. more concretely, we focus our to cite this article: pons, m., bikfalvi, a., llach, j. (2018). clustering product innovators: a comparison between conventional and green product innovators. international journal of production management and engineering, 6(1), 37-46. https://doi.org/10.4995/ijpme.2018.8762 int. j. prod. manag. eng. (2018) 6(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international 37 mailto:marc.pons@udg.edu http://creativecommons.org/licenses/by-nc-nd/4.0/ analysis on firms that affirm having implemented product innovations in the last three years. we complement this aspect with a further detail, namely product innovators whose new products contemplate an improvement of the environmental impact by either using or disposing of them. 3. literature review 3.1. conceptual delimitation and definition of green product innovation a product innovative firm has been defined as the one that has implemented a new or significantly improved product during the period under review according to the oslo manual (oecd & eurostat, 2005). complementing this definition, and for the purpose of this study, green product innovation is defines as the design, production and implementation of new or significantly improved products that have a positive impact on the environment when in use or when disposing of them. different authors use a variety of terms to describe new products with environmental implications that are synonyms and combinations of eco, eco-friendly, ecological, green, sustainable, environmental and environmental-friendly with innovation, product innovation, new product (dangelico, 2016) (gerstlberger et al., 2014). “green product” and “environmental product” are used commonly to describe those that strive to protect or enhance the natural environment by conserving energy and/or resources and reducing or eliminating the use of toxic agents, pollution, and waste (ottman et al., 2006). pujari refers to the action to develop and market new products that address environmental issues. most of the sustainable innovation in npd relates to incremental or evolutionary innovation (pujari, 2006). product innovations with environmental implications should fulfil two goals simultaneously, namely improvement of environmental impact and obtaining commercial performance (gerstlberger et al., 2014). holistic approaches to model design should prevail, as the ones advocated by the 6rs (redesigning, reusing, remanufacturing, recovering, recycling, and reducing) and products with multiple life cycles (thomé et al., 2016). 3.2. determinants of green product innovation some authors tried to identify if and up to what degree, determinants of product innovation apply to green new product manufacturers. in the case of specific drivers, they also measured their effect (edison et al., 2013; keupp et al., 2012). other, grouped the factors in internal/external or by nature as technological capabilities, internal integrative capabilities, external integrative capabilities or marketing capabilities (dangelico, 2016). 4. methods our research is based on data from the european manufacturing survey (ems), 2015 edition. ems is coordinated by the fraunhofer institute for systems and in-novation research (isi, 2017) and it is the largest european survey in manufacturing activities to date. it aims to collect data relative to the modernisation of manufacturing processes and practices. it complements existing innovation surveys by including latest trends among the topics of interest. further elaborating in this direction, environmental aspects (energy and material saving technologies and practices, energy consumption, their sources and use) have been considered and updated since 2009 and on-going. our study includes data from ems spain, france and portugal, formed by 194 firms’ responses. the survey was performed on manufacturing firms having at least 20 employees. developed jointly by columbia university and yale university, the environ-mental performance index (epi) ranks 180 countries on 20 performance indicators, which track performance and progress on two broad objectives: protection of human health and protection of ecosystem (hsu and zomer, 2016). according to the latest edition all three countries are part of the top 10 of the 2016 epi rankings, spain ranks 6th with an epi score of 88.91, portugal is at position 7 with a score of 88.63, while france situates at the 10th position scoring 88.2 in the ranking where finland has taken the top spot with the maximum possible score of 90.68. all countries included in the present analysis have high epi performance int. j. prod. manag. eng. (2018) 6(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international pons, m., bikfalvi, a. and llach, j. 38 http://creativecommons.org/licenses/by-nc-nd/4.0/ indicators with a better performance than countries in their region (europe), globally. technical details of the utilized subsamples are shown in table 1. from the existing distances in a set of variables, groups of cases have been created by a k-means cluster analysis. variables were the ones in our sample representing the firm’s competitive factors significance ranked from 1 (most important) to 6 (less important): ‘product price’, ‘product quality’, ‘innovative products’ and ‘customization to customers’ demands’. other two competitive factors variables, ‘adherence to delivery/short delivery times’ and ‘service’, were not considered to obtain the clusters because they are not strictly linked with product innovation. according with the obtained clusters of product innovators, a frequencies analysis for variables representing the technological level of firms and product development and manufacturing aspects and innovation drivers are performed. other descriptive analysis have been elaborated from variables such as companies’ personnel distribution and qualification, company size (normalized with logarithms), exportation index and strategic costs as percentage of the turnover in 2014, like energy, payroll or r&d costs. the group of companies affirming their new products lead to an improvement of their environmental impact during their use or disposal differentiating them from the rest of conventional product innovatorsis called ‘green product innovators’. both groups are analysed separately to compare results and detect differences among clusters, being this the main objective of the present study. 5. results the cluster analysis results in three coherent groups of manufacturing establishments distinguishing between product innovators competing by i) customization, ii) price, and iii) innovation, as shown in figure 1 and table 2. in these three cases, companies also compete by quality as well, but we label groups with the most relevant competitiveness factor that differentiates among the groups. the differences are also reported according the presence of firms whose new products improve their environmental impact. table 2. product innovators inside each obtained cluster. cluster 1 customization cluster 2 price cluster 3 innovation n n n conventional 22 19 13 green 23 15 20 total 45 34 33 table 1. technical details for the spanish, french and portuguese subsamples of the european manufacturing survey 2015 edition. universe: spanish, french and portuguese manufacturing firms with at least 20 employees cnae 2009; codes from 10 to 33. unit of analysis: establishment sample: 194 firms: (es) 100; (fr) 61; (pt) 33 confidence margin: 95% variance: maximum indetermination p=q=50% documentation paper (8 pages questionnaire) + return envelope + presentation letter channel postal fieldwork: may to september 2015 reference period: 2012-2014; 2014 institution: dept. of business administration and product design, university of girona – girona (spain) university of lyon, iae lyon, lyon (france) dept. of mechanical and industrial engineering, universidade nova de lisboa, caparica (portugal) data base recording and creation: es: outsourced to dap gmbh – passau (germany) fr, pt: institution sample distribution: by size and sector of activity by ‘green product innovators’: ‘conventional product innovators’: 55 (es) 34; (fr) 15; (pt) 6 ‘green product innovators’: 60 (es) 23; (fr) 25; (pt) 12 int. j. prod. manag. eng. (2018) 6(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international clustering product innovators: a comparison between conventional and green product innovators 39 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5.1. technological level, product development and product manufacturing characteristics table 3 presents a frequency analysis of companies’ technological level and product development and manufacturing characteristics for the three clusters, differentiating also between green and non-green/ conventional product innovators. technological level refers to the eurostat aggregation of the manufacturing industry according to technological intensity based on firm’s nace code rev.2. it is observable that the majority of product innovative firms competing by innovation are gpi. likewise, inside the cluster competing by customization, companies are notably more gpi than cpi except in the case of the low technological intensity ones that are clearly more cpi. low technological intensity firms competing by price, are 87.5% cpi, and more equilibrated for both groups in the case of low-med, med-high and high technological intensity ones. regarding the analysed manufacturing characteristics, we obtain the results for product development customization level, manufacturing customization level, batch or lot sizes, and product complexity level. a summary of the most interesting highlights regarding gpi is presented below. 5.1.1. manufacturing characteristics for product innovators in “customization” cluster inside the cluster of innovators competing by customization, companies with a high product development customization level are more gpi than cpi. no gpi can be found among manufactures that “make to order”, that is the highest manufacturing customization level. the share of gpi increases as the lot/batch sizes decrease being a 78% of the firms in the case of manufacturing unit by unit. mainly in high but also in low product complexity level the percentage of gpi is higher. 5.1.2. manufacturing characteristics for product innovators in “price” cluster in high product development customization level, the percentage of gpi competing by price is lower than the cpi one (40% vs. 60%). regarding to the manufacturing customization degree, gpi represent a higher percentage in the group of companies that figure 1. clustering methodology for product innovators (source: own elaboration). int. j. prod. manag. eng. (2018) 6(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international pons, m., bikfalvi, a. and llach, j. 40 http://creativecommons.org/licenses/by-nc-nd/4.0/ produce with a “make to stock” system. innovators producing in high lot/batch sizes are, mostly, cpi (66%). the percentage of gpi increases as it increases the product complexity level, being a 55% in the case of companies that produce highly complex products. 5.1.3. manufacturing characteristics for product innovators in “innovation” cluster product innovative firms competing by innovation that offer a medium or high product development customization level are mostly gpi in a 69% and a 55% respectively. in product manufacturing customization level, cpi represent only a 33% of the companies that assemble to order and no one of them make to stock or make to order. gpi represents the majority of innovators producing big size and unitary lot/batch sizes with a 70% and 88% respectively. in all product complexity degrees, gpi represent the majority of firms inside this cluster with a very similar percentages: 60% for high, 61% for medium and 60% for low complexity. 5.2. main origins of impulses/ideas for innovation as it is observed in table 4 and more easily in figure 2, some differences between conventional and green product innovators appear regarding the table 3. frequency analysis for firms’ technological level, development and manufacturing customization, lot size and product complexity. cluster 1 customization cluster 2 price cluster 3 innovation n [%]column n [%]column n [%]column tech_level (from nace rev2) low conventional 9 64.3 % 7 87.5 % 6 42.9 % green 5 35.7 % 1 12.5 % 8 57.1 % med-low conventional 7 41.2 % 6 50.0 % 4 36.4 % green 10 58.8 % 6 50.0 % 7 63.6 % med-high and high conventional 6 42.9 % 6 42.9 % 3 37.5 % green 8 57.1 % 8 57.1 % 5 62.5 % product development customization level low conventional 1 100.0 % 1 50.0 % 1 50.0 % green 0 0.0 % 1 50.0 % 1 50.0 % med conventional 8 57.1 % 9 52.9 % 4 30.8 % green 6 42.9 % 8 47.1 % 9 69.2 % high conventional 13 46.4 % 9 60.0 % 8 44.4 % green 15 53.6 % 6 40.0 % 10 55.6 % manufacturing customization level make to order conventional 1 100.0 % 0 0.0 % 0 0.0 % green 0 0.0 % 0 0.0 % 1 100.0 % assemble to order conventional 5 45.5 % 3 50.0 % 3 33.3 % green 6 54.5% 3 50.0 % 6 66.7 % make to stock conventional 2 40.0 % 3 42.9 % 0 0.0 % green 3 60.0 % 4 57.1 % 3 100.0 % no production conventional 12 46.2 % 13 61.9 % 10 52.6 % green 14 53.8 % 8 38.1 % 9 47.4 % batch or lot sizes unit conventional 2 22.2 % 2 50.0 % 1 12.5 % green 7 77.8% 2 50.0 % 7 87.5 % med size conventional 13 52.0 % 9 50.0 % 9 60.0 % green 12 48.0 % 9 50.0 % 6 40.0 % big size conventional 7 63.6 % 8 66.7 % 3 30.0 % green 4 36.4 % 4 33.3 % 7 70.0 % product complexity level low conventional 4 44.4 % 2 100.0 % 2 40.0 % green 5 55.6 % 0 0.0 % 3 60.0 % medium conventional 14 56.0 % 11 57.9 % 7 38.9 % green 11 44.0 % 8 42.1 % 11 61.1 % high conventional 4 36.4 % 5 45.5 % 4 40.0 % green 7 63.6 % 6 54.5 % 6 60.0 % int. j. prod. manag. eng. (2018) 6(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international clustering product innovators: a comparison between conventional and green product innovators 41 http://creativecommons.org/licenses/by-nc-nd/4.0/ origin of impulses/ideas they declared to use for their innovations. these differences are also particular for every cluster and they could not be appreciated in a general, non-clustered analysis. green product innovators competing by customisation find inspiration for new product development in the r&d/engineering department and the customer service section. complementary, ideas for npd also come from the customer/user. the pattern is partially similar for low cost product innovators who find their main sources of inspiration in the r&d/engineering department, the customer/ user and ceo/management (in decreasing order). green product innovators competing by innovation are mainly inspired by the customer/user followed by the ceo/management and third, in the r&d/ engineering department. 5.3. companies’ characteristics since companies’ characteristics are important determinants of innovation, it is interesting to observe the results showed in table 5, table 6 and table 7. 5.3.1. personnel closely related to the previous section qualification level of employees is often related to companies’ capacity to innovate. the results in table 5 show that the highest level of qualification is characteristic to cluster of firms following a strategy based on price, followed by those firms differentiating from competitors through innovation, and last the ones focusing on customization. when comparing traditional product innovators to green product innovators, major differences in favour of gpi showing higher or equal values to the other ones, can be observed in the “low cost” category. differences are minor and do not exceed 0.2 points. as observed in the previous section both internal to the firm and external sources of ideas/ impulses for innovation can be detected. focusing the attention on the distribution of employees in the different key functional areas of the firm the analysis shows the following: i) indifferently of the cluster, research & development employees are more numerous in gpis, ii) cluster 2 shows differentiated characteristics in the sense that gpis that belong to this have higher concentration of employees in manufacturing, assembly and other areas, iii) the major difference table 4. frequency analysis for main origins of ideas/impulses for innovation by cluster. cluster 1 customization cluster 2 price cluster 3 innovation n [% ] s ha re o f c om pa ni es b y cl us te r [% ] c om pa ni es b y or ig in o f i de as n [% ] s ha re o f c om pa ni es b y cl us te r [% ] c om pa ni es b y or ig in o f i de as n [% ] s ha re o f c om pa ni es b y cl us te r [% ] c om pa ni es b y or ig in o f i de as r&d / engineering conventional 11 32.4 % 20.0 % 15 44.1 % 27.3 % 8 23.5 % 14.5 % green 17 47.2 % 28.3 % 12 33.3 % 20.0 % 7 19.4 % 11.7 % production conventional 6 37.5 % 10.9 % 7 43.8 % 12.7 % 3 18.8 % 5.5 % green 7 63.6 % 11.7 % 1 9.1 % 1.7 % 3 27.3 % 5.0 % customer service conventional 5 35.7 % 9.1 % 2 14.3 % 3.6 % 7 50.0 % 12.7 % green 12 50.0 % 20.0 % 5 20.8 % 8.3 % 7 29.2 % 11.7 % ceo/ management conventional 10 43.5 % 18.2 % 8 34.8 % 14.5 % 5 21.7 % 9.1 % green 11 39.3 % 18.3 % 8 28.6 % 13.3 % 9 32.1 % 15.0 % customer or user conventional 17 48.6 % 30.9 % 8 22.9 % 14.5 % 10 28.6 % 18.2 % green 12 34.3 % 20.0 % 9 25.7 % 15.0 % 14 40.0 % 23.3 % supplier conventional 2 40.0 % 3.6 % 1 20.0 % 1.8 % 2 40.0 % 3.6 % green 2 50.0 % 3.3 % 0 0.0 % 0.0 % 2 50.0 % 3.3 % research institutions, universities conventional 0 0.0 % 0.0 % 3 100.0 % 5.5 % 0 0.0 % 0.0 % green 3 50.0 % 5.0 % 2 33.3 % 3.3 % 1 16.7 % 1.7 % consultancy conventional 1 33.3 % 1.8 % 1 33.3 % 1.8 % 1 33.3 % 1.8 % green 0 0.0 % 0.0 % 1 50.0 % 1.7 % 1 50.0 % 1.7 % int. j. prod. manag. eng. (2018) 6(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international pons, m., bikfalvi, a. and llach, j. 42 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 2. manufacturing firm’s main origin of ideas/impulses for innovation by clusters. int. j. prod. manag. eng. (2018) 6(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international clustering product innovators: a comparison between conventional and green product innovators 43 http://creativecommons.org/licenses/by-nc-nd/4.0/ in percentage points can be observed in cluster 3, the results showing customer service as the function concentrating more employees in conventional product innovators than in gpis (22.3 vs. 10.3). 5.3.2. size, costs and economic parameters the results for gpi and cpi regarding different variables representing company size, costs and economic performance are showed in table 7. the differential of turnover as a basic financial performance indicator does not show any significant difference between gpi and cpi. the same similarities between green and conventional innovators appear in variables as number of employees, payroll costs or relative percentage of energy costs for all the clusters. the most outstanding results appear in the cluster of firms following a strategy based on price regarding table 5. descriptive analysis for the personnel qualification [1-5] by cluster. cluster1 customization cluster 2 price cluster 3 innovation µ σ µ σ µ σ global personnel qualification [1-5] 1 for lowest and 5 for highest (phd and master) conventional 2.6 0.5 3.0 0.6 2.9 0.5 green 2.8 0.4 3.1 0.5 2.9 0.6 table 6. descriptive analysis for the personnel distribution inside each company areas in % by cluster. cluster1 customization cluster 2 price cluster 3 innovation µ σ µ σ µ σ research & development conventional 3.7 3.9 6.9 6.6 3.8 5.7 green 6.3 4.6 7.4 5.3 5.4 7.7 configuration, design conventional 3.8 5.9 7.3 8.1 3.0 3.4 green 8.5 8.9 4.2 7.0 4.6 4.9 manufacturing and assembly conventional 69.8 16.9 57.9 22.3 61.6 28.6 green 61.0 14.9 65.3 12.0 58.7 19.5 customer service conventional 3.3 4.3 7.8 8.1 22.2 32.2 green 6.0 5.0 6.6 4.6 10.3 9.8 other conventional 20.7 14.1 22.3 14.7 15.5 11.5 green 19.3 11.7 22.7 23.3 20.6 12.4 main differences table 7. descriptive analysis for company size, strategic costs and exportation by cluster. cluster1 customization cluster 2 price cluster 3 innovation µ σ µ σ µ σ ln (annual turnover 2014) conventional 3.4 1.4 3.3 1.5 3.1 1.6 green 3.1 1.3 3.7 2.2 2.9 1.2 ln (number of employees 2014) conventional 4.9 1.1 4.4 1.0 4.2 1.8 green 4.3 1.0 5.2 1.5 4.5 1.1 % of r&d relative to incomes 2014 conventional 3.0 4.0 8.8 8.2 3.3 4.9 green 2.9 2.6 4.9 5.4 4.0 3.7 payroll costs as % of turnover 2014 conventional 20.8 7.7 23.0 12.6 22.9 17.1 green 23.1 11.1 25.3 16.0 24.5 12.2 % products sold abroad conventional 44.1 30.7 48.1 37.0 45.2 29.1 green 41.4 30.0 63.0 23.2 40.4 32.2 total energy costs as % of turnover 2014 conventional 2.9 4.2 3.2 2.5 6.6 10.8 green 4.5 7.8 4.0 3.5 3.8 4.0 main differences int. j. prod. manag. eng. (2018) 6(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international pons, m., bikfalvi, a. and llach, j. 44 http://creativecommons.org/licenses/by-nc-nd/4.0/ variables representing relative r&d expenditures and exportation. in this cluster, we can observe gpi declare, in average, less percentage of r&d expenditures relative to incomes than cpi. on the other hand, gpi declare they sell more percentage of products abroad than cpi. 6. conclusions the paper provides recent objective data regarding product innovation and sustainability in southwestern european manufacturing firms. introducing clusters, hidden aspects that differentiate green product innovators from conventional ones it can be observed. these differences can not be seen in an overall analysis. describing and differentiating both groups of gpi and cpi, our findings could be insights for policy makers to identify drivers and factors that impulse this type of desirable innovations or barriers that difficult their emergence. it could be informative for manufacturing practitioners in terms of characteristics and opportunities of green new product innovation. 7. contribution the present work aims to complement previous descriptive analysis on product innovation and sustainability in manufacturing firms using the same methodology (pons et al.,, 2013; palčič et al.,, 2013; pons et al.,, 2017), but adding a layer of complexity achieved by the cluster analysis as well as presenting recent data on a topic situated at the intersection of two crucial societal issues, namely environment and innovation. while manufacturers can find greening opportunities in both process and products, the product option remains one of the most perceived and visible alternative for stakeholders, being that the backbone of the present contribution. 8. future research the study could be expanded to 10 countries evaluating country effects. it would be interesting to observe if different environmental policies, regulations or green cultures affects to the results. a more sophisticated analysis of performance (environmental and economic) in relation to these gpi should be made in the future. in the framework of a wider sample, it could be possible to compute a variable capturing different degrees of greenness considering, for example, the extent of implementation of green product innovations. models testing relationships between drivers/ barriers, company characteristics and green product innovation and/or performance, have to be further studied. references dangelico, r.m. 2016. green product innovation: where we are and where we are going. business strategy and the environment, 25(8), 560–576. https://doi.org/10.1002/bse.1886 edison, h., bin ali, n., torkar, r. (2013). towards innovation measurement in the software industry. journal of systems and software, 86(5), 1390–1407. https://doi.org/10.1016/j.jss.2013.01.013 gerstlberger, w., præst knudsen, m., stampe, i. (2014). sustainable development strategies for product innovation and energy efficiency. business strategy and the environment, 23(2), 131–144. https://doi.org/10.1002/bse.1777 hsu, a., zomer, a. (2016). 2016 environmental performance index (epi). wiley statsref: statistics reference online. 1–5. https://doi.org/10.1002/9781118445112.stat03789.pub2 isi. 2016. the european manufacturing survey. available at: http://www.isi.fraunhofer.de/isi-en/i/projekte/fems.php [accessed: february 20, 2013]. keupp, m.m., palmié, m., gassmann, o. (2012). the strategic management of innovation: a systematic review and paths for future research. international journal of management reviews, 14(4), 367–390. https://doi.org/10.1111/j.1468-2370.2011.00321.x oecd & eurostat. (2005). oslo manual guidelines for collecting and interpreting innovation data. https://doi.org/10.1787/9789264013100-en ottman, j.a., stafford, e.r., hartman, c.l. (2006). avoiding green marketing myopia. environment, 48(5), 22. https://doi.org/10.3200/ envt.48.5.22-36 int. j. prod. manag. eng. (2018) 6(1), 37-46creative commons attribution-noncommercial-noderivatives 4.0 international clustering product innovators: a comparison between conventional and green product innovators 45 https://doi.org/10.1002/bse.1886 https://doi.org/10.1016/j.jss.2013.01.013 https://doi.org/10.1002/bse.1777 https://doi.org/10.1002/9781118445112.stat03789.pub2 http://www.isi.fraunhofer.de/isi-en/i/projekte/fems.php https://doi.org/10.1111/j.1468-2370.2011.00321.x https://doi.org/10.1787/9789264013100-en https://doi.org/10.3200/envt.48.5.22-36 https://doi.org/10.3200/envt.48.5.22-36 http://creativecommons.org/licenses/by-nc-nd/4.0/ palčič, i., pons, m., bikfalvi, a., llach, j., buchmeister, b. (2013). analysing energy and material saving technologies’ adoption and adopters. strojniški vestnik journal of mechanical engineering, 59, 409–417. https://doi.org/10.5545/sv-jme.2012.830 pons, m., bikfalvi, a., llach, j., palčič, i. (2013). exploring the impact of energy efficiency technologies on manufacturing firm performance. journal of cleaner production, 52, 134-144. https://doi.org/10.1016/j.jclepro.2013.03.011 pons, m., llach, j., bikfalvi, a. (2017). analysing innovators according to the environmental impact of new products. in ipdmc reykjavik. pujari, d. (2006). eco-innovation and new product development: understanding the influences on market performance. technovation, 26(1), 76–85. https://doi.org/10.1016/j.technovation.2004.07.006 thomé, a.m.t., scavarda, a., ceryno, p.s., remmen, a. (2016). sustainable new product development: a longitudinal review. clean technologies and environmental policy, 18(7), 2195–2208. https://doi.org/10.1007/s10098-016-1166-3 un. (2016). sustainable development goals 17 goals to transform our world. sustainable development goals united nations, 1–2. available at: https://www.mendeley.com/research-papers/sustainable-development-goals-17-goals-transform-world/?utm_ source=desktop&utm_medium=1.17.10&utm_campaign=open_catalog&userdocumentid=%7ba2e1fd7d-29e9-3f6d-bac8f243523eab66%7d [accessed: august 31, 2017]. int. j. prod. manag. eng. (2018) 6(1), 37-46 creative commons attribution-noncommercial-noderivatives 4.0 international pons, m., bikfalvi, a. and llach, j. 46 https://doi.org/10.5545/sv-jme.2012.830 https://doi.org/10.1016/j.jclepro.2013.03.011 https://doi.org/10.1016/j.technovation.2004.07.006 https://doi.org/10.1007/s10098-016-1166-3 https://www.mendeley.com/research-papers/sustainable-development-goals-17-goals-transform-world/?utm https://www.mendeley.com/research-papers/sustainable-development-goals-17-goals-transform-world/?utm https://www.mendeley.com/research-papers/sustainable-development-goals-17-goals-transform-world/?utm http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2015.3345 received 2014-10-29 accepted: 2014-01-07 the distributed parallel machine and assembly scheduling problem with eligibility constraints sara hatamia,i, rubén ruizb and carlos andrés-romanoa,ii a departamento de organización de empresas, universitat politècnica de valència. camino de vera s/n, 46021, valència, spain. a,i sara_sodi@yahoo.com a,ii candres@omp.upv.es b grupo de sistemas de optimización aplicada, instituto tecnológico de informática, ciudad politécnica de la innovación, edifico 8g, acc. b. universitat politècnica de valència, camino de vera s/n, 46021, valència, spain. b rruiz@eio.upv.es abstract: in this paper we jointly consider several realistic scheduling extensions: first we study the distributed unrelated parallel machines problem where there is a set of identical factories with parallel machines in the production stage. jobs have to be assigned to factories and to machines. additionally, there is an assembly stage with a single assembly machine. finished jobs at the manufacturing stage are assembled into final products in this second assembly stage. these two joint features are referred to as the distributed parallel machine and assembly scheduling problem or dpmasp. the objective is to minimize the makespan in the assembly stage. due to technological constraints, machines cannot be idle and some jobs can be processed only in certain factories. we propose a mathematical model and two high-performing heuristics. the model is tested with two state-of-the-art solvers and, together with the heuristics, 2220 instances are solved in a comprehensive computational experience. results show that the proposed model is able to solve moderately-sized instances, and that one of the heuristics is fast, giving optimal solutions close to optimum in less than half a second in the worst case. key words: distributed parallel machines, assembly stage, heuristics, model. 1. introduction nowadays, the manufacturing industry faces many challenges, namely globalization, increasing product variety, complexity and customer demands, shorter product life cycles, higher demand of customized goods instead of mass production, uncertain and dynamic global market, etc. of course, the strong competition from emerging and established economies has to be considered as well. one of the many tools to face these challenges and to meet customer’s demands is to increase the product variety that companies offer. a wide product portfolio and diversified offer is a key asset to stay competitive in such an unpredictable and ever evolving market. product variety has been defined by many authors as a number or collection of different things of a particular class of the same general kind (elmaraghy et al., 2013). in recent years, assembly systems are such as techniques that are mostly used mass production. they have been also employed in various manufacturing systems so as to increase flexibility and the capability to increase product variety. these types of manufacturing settings are referred to as assembly scheduling problems (asp). in an assembly system, different operations are performed independently, and potentially in parallel, to produce different components which are later assembled into finished products in assembly lines. a high variety of finished products, made from different combinations of produced components, can be produced in assembly systems. existence of more than one manufacturing facility in different geographical places may decrease some costs related to the production. to offset these costs, companies must operate different and specialized factories in what is known as distributed 13int. j. prod. manag. eng. (2015) 3(1), 13-23creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2015.3459 http://creativecommons.org/licenses/by-nc-nd/4.0/ manufacturing systems (dms). in a dms environment, several independent production centers or factories are run in parallel at potentially different geographical places. furthermore, distributed manufacturing allows for greater flexibility and resiliency (sluga et al., 1998). other benefits of dms are: higher product quality, lower production costs, reduced risks (kahn et al., 2004; chan et al., 2005; mahdavi et al., 2008). however, scheduling in dms is more complicated than in a single production factory. in single production centers a job schedule for each set of machines has to be defined, while in dmss, there are two interrelated decisions to be made: factory selection for each job and then scheduling at each factory. as a conclusion, and in order to reap the benefits of both assembly systems (asp) and distributed manufacturing (dms), both aspects must be jointly considered. in this paper we consider two manufacturing stages: production and assembly. for production we have a set of distributed factories and for assembly there is a single assembly facility. each one of the f distributed production centers (factories) has unrelated parallel machines as a shop configuration whereas the assembly stage consists of a single machine. transportation time for transferring jobs from production centers to assembly stage is assumed negligible. by considering the above model we define the studied problem in this paper as the distributed parallel machine and assembly scheduling problem (dpmasp). more in specifically, in the dpmasp there is a set n of n jobs that has to be processed on a set f of f identical factories. note that all factories are identical and have the same number of machines. each factory has a set m of m unrelated parallel machines. each job has to be processed at exactly one machine at one factory. furthermore, there are eligibility constraints. lfj⊆f is the subset of factories where job j can be assigned, where f ≥|lfj|≥ 1, j=1…n job j can only be assigned to an eligible factory. there is a set t of t independent products. each product is assembled at the single assembly machine ma. for the assembly of product h, h=1…t a subset nh⊆n of jobs must have been produced at the distributed factories beforehand. each job can only belong to an assembly program of a product, i.e., n nhh t 1 = = / . the assembly of product h can only start when all jobs in nh have been completed at the distributed factories. for the processing at the distributed manufacturing stage, pjk denotes the processing time of job j at machine k of any factory. note that all factories are identical and have the same number of machines. for the assembly stage, ph denotes the assembly time of product h. all processing times are positive, deterministic and known integer quantities. the objective in the proposed dpmasp is to assign jobs to machines at factories in the distributed manufacturing stage, to schedule all assigned jobs to each machine at each factory and to schedule products at the single machine assembly stage while minimizing the makespan at this assembly stage. as regards the computational complexity of the dpmasp we can conclude that it is an np-hard problem if n≫f since the regular parallel machines problem (even in the case where there are two identical machines, i.e., the p2//cmax problem) is already np-hard according to the results of lenstra et al. (1977). as we will later show, the dpmasp is an important generalization of existing problems that has not been studied before to the best of our knowledge. in this paper we propose a mathematical model to solve the problem. the model is solved with two state-of-the-art commercial solvers and results are compared. two high performing heuristics are proposed and are shown to give results that are, in many cases, close to the optimal ones. the rest of the paper is organized as follows: in the next section we present a short literature review on related problems. in section 3 we present a mixed integer linear programming (milp) model to solve the considered problem. section 4 describes two simple constructive heuristics. section 5 presents a comprehensive computational evaluation of the proposed milp and simple constructive heuristics. finally, some concluding remarks and future research directions are provided in section 6. 2. literature review as mentioned, the dpmasp contains parts from distributed manufacturing, assembly and parallel machines. as such, a complete literature review on each one of these three topics is clearly outside the scope of this paper. some of the closely related research will be reviewed instead. regarding the assembly part of the proposed dpmasp, lee et al. (1993) considered a threemachine assembly-type flowshop problem (nondistributed). the problem comprises two stages; in the first stage there are two production machines that produce two components for each single product. the second stage is a single assembly machine that assembles the two produced components to make each final product. they present a branch and bound algorithm and also an approximate 14 int. j. prod. manag. eng. (2015) 3(1), 13-23 creative commons attribution-noncommercial-noderivatives 4.0 international hatami, s., ruiz, r. and andrés-romano, c. http://creativecommons.org/licenses/by-nc-nd/4.0/ procedure. makespan minimization is considered as an objective function. later, potts et al. (1995) considered m parallel machines instead of the two production machines in the first stage. they produced approximated solutions with worse-case absolute performance guarantees. for the same problem of lee et al. (1993), hariri and potts (1997) proposed a branch-and-bound algorithm, and sun et al. (2003) presented different powerful heuristic algorithms. also, sung and kim (2008) tried to expand the model presented by lee et al. (1993) by adding multiple-assembly machines in the second stage. the objective is to minimize the sum of completion times. they proposed a lower bound and employed it in a branch-and-bound algorithm. an efficient and simple heuristic was also proposed. as mentioned, we consider eligibility constraints for assigning jobs to factories in distributed manufacturing stage. to the best of our knowledge, lin and li (2004) have a similar job to machine eligibility constraints. in this paper, the parallel machine scheduling problem with unit processing times is studied and polynomial algorithms are presented. for the distributed part of the dpmasp we have to note that dms is a general and broad manufacturing term. focusing only on distributed scheduling problems, there are few studies about, distributed flowshops and jobshops. for example, the distributed permutation flowshop scheduling problem (dpfsp) was introduced for the first time by naderi and ruiz (2010). they proposed six different alternative milp models, two simple factory assignment rules, fourteen heuristics and variable neighborhood descent methods. later, lin et al. (2013) and wang et al. (2013) proposed an effective iterated greedy (ig) method and an estimation of distribution algorithm on dpfsp, respectively. later, naderi and ruiz (2014) presented a scatter search (ss) method for the dpfsp. this ss was shown to outperform existing methods. for an updated literature review on the dpfsp, the reader is referred to this paper of naderi and ruiz (2014). recently, fernandez-viagas and framinan (2015) have presented a modified iterated greedy algorithm for the dpfsp, which is shown to outperform the initial algorithms of naderi and ruiz (2010). however, there is no comparison between the ss of naderi and ruiz (2014) and this modified iterated greedy. the distributed jobshop problem considering two different criteria is studied first by jia et al. (2002) and jia et al. (2003) where they proposed genetic algorithm (ga) to solve the problem. later, jia et al. (2007), refined the previous ga. chan et al. (2006) studied the distributed jobshop with makespan objective, also using ga. the only papers that we are aware of that jointly consider the assembly and distributed aspects are hatami et al. (2013) which recently introduced the distributed assembly permutation flowshop scheduling problem (dapfsp). in this problem, there are f distributed flowshop production centers and a single assembly center with a single machine. a milp, several constructive heuristics and simple local search based variable neigborhood descent (vnd) methods were proposed. xiong et al. (2014) presented a distributed two-stage assembly system with setup times. the authors considered f distributed factories where each factory has the same m processing parallel machines at the first stage and the same assembly machine at the second stage. each assembled product consists of m components produced by parallel machines. they developed heuristic methods and three hybrid meta-heuristics to minimize the total completion time. the problem studied by xiong et al. (2014) is different from the studied dpmasp. first, we consider a separated assembly stage, not an assembly operation at each factory. second, we allow the different jobs composing a product to be produced in different factories. third, each product might have a number of jobs (components) different from m. as we can see, and to the best of our knowledge, there is no literature on the dpmasp. 3. mixed integer linear programming model we present a mathematical model to solve the proposed dpmasp. first we detail the indexes, parameters and variables are used: index description i, j denotes jobs, i, j=0,1...n, where 0 represents a dummy job k denotes machines, k=1...m q denotes factories, q=1...f l,s denotes products, l, s=0,1...t, where 0 represents a dummy product m a sufficiently large positive number parameter description n number of jobs m number of machines f number of factories t number of products pjk processing time of job j on machine k ps processing time of product s at the assembly stage 15int. j. prod. manag. eng. (2015) 3(1), 13-23creative commons attribution-noncommercial-noderivatives 4.0 international the distributed parallel machine and assembly scheduling problem with eligibility constraints http://creativecommons.org/licenses/by-nc-nd/4.0/ gjs binary parameter equal to 1 if job j belongs to product s, and 0 otherwise variable description xijkq binary variable equal to 1 if job i is an immediate predecessor of job j on machine k in factory q yls binary variable equal to 1 if product l is an immediate predecessor of product s at the assembly machine cj completion time of job j at the production stage cas completion time of product s on the assembly stage cmax makespan the objective function of the model is to minimize the makespan: min cmax subject to the following constraints: x j1 , ijkq q q lf q lf f k m i i j n 110 j i 6= ! ! ! === /// (1) x i1 , ijkq q q lf q lf f k m j j i n 110 j i 6= ! ! ! === /// (2) ,x k q1jkq j q lf n 0 1 j 6= ! = / (3) ,x k q1i kq i q lf n 0 1 i 6= ! = / (4) , , ,x x i k q q lf0 , ijkq jikq i j j i q lf n 1 j 6 != ! ! = ^ h/ (5) , , , x x i n j i 1 1 1 ijkq jikq q q lf q lf f k m q q lf q lf f k m 1111 i j i j 6 f 2 # ! + ! ! ! ! ==== " , //// (6) , , , , c c p m x i j k q q lf lf 1j i jk i jkq i j+6 $ ! + + -^ h (7) sy 1 , ls l l s t 0 6= != / (8) y l1 , ls s l s t 1 6# != / (9) , , ,y y l t s l1 1 1ls sl 6 f 2# !+ -" , (10) ,ca c g p j s·s j js s 6$ +^ h (11) , ,ca ca p m y l s l s1s l s ls 6 !$ + + -^ h (12) c ca smax s 6$ (13) , , , , , , ,x i j k q i j q lf q lf0 1ijkq i j6 !! ! !" , (14) , ,, l s l sy 0 1ls 6 !! " , (15) cj j0 6$ (16) ca s0s 6$ (17) note that c0=ca0=0. constraint sets (1) and (2) ensure that each job must have exactly one preceding and succeeding job, respectively. sets (3) and (4) enforce that each machine at each factory has to have a dummy job 0 as predecessor and successor, respectively. note that this is a special constraint, as we do not allow any machine at any factory to be empty due to technological or economic constraints. this also requires the total number of jobs in the shop (n) to be greater or equal than f × m. constraint set (5) ensures that if a job is sequenced on a machine, then its predecessor and successor must be processed on the same machine. constraint set (6) controls that a job cannot be both a predecessor and successor of another job at the same time. constraint set (7) determines that if job j is placed immediately after job i, its processing at machine k cannot start before the processing of job i in machine k finishes. constraints (8) and (9) force that each product should have one predecessor and at most one succeeding product in the assembly factory, respectively. constraint (10) controls that a product cannot be both a predecessor and a successor of another product at the same time in the assembly machine. constraint (11) determines that each product h cannot begin to be assembled before all its jobs are completed in the corresponding machine. constraint set (12) determines that if product s is placed immediately after product l, it cannot start to be assembled on the assembly machine before the assembling of product l in assembly machine has finished. constraints (13) and (14)-(17) define the makespan and the domain of the decision variables, respectively. note that only 16 int. j. prod. manag. eng. (2015) 3(1), 13-23 creative commons attribution-noncommercial-noderivatives 4.0 international hatami, s., ruiz, r. and andrés-romano, c. http://creativecommons.org/licenses/by-nc-nd/4.0/ the necessary variables are defined, i.e., eligibility constraints are implicitly considered in the model. 4. constructive heuristic methods let us first introduce a dpmasp example that will guide the exposition of the proposed heuristics. the example consists of fourteen jobs (n=14), three products (t=3), two factories (f=2) with two unrelated parallel machines in each factory (m=2). the assembly programs for each product are: n1={2, 7, 8}, n2={1, 3, 4, 10, 12, 13} and n3={5, 6, 9, 11, 14}, i.e., jobs 2, 7 and 8 must be finished in order to assemble product 1. table 1 contains the job processing times on each machine at the production stage and eligibility constraints. processing times for assembling products 1 to 3 are 3, 12 and 7, respectively. some additional notation is the following: a product sequence is represented by π, e.g., π = {2, 1, 3}. to assign all jobs belonging to the assembly program of product h to the unrelated parallel machines at the different factories, a job to machine-factory assignment method is needed. after the application of this assignment procedure we obtain a job to machine-factory sequence for product h, referred to πh, e.g., π1 = {0, 8; 7, 2}, π2 = {1-10, 3; 12, 4-13} and π3 = {14, 5; 69, 11} as a possible job to machinefactory sequence for products of the example. at each πh, each factory is separated by “;”, each machine by “,” and the sequence of jobs at each machine is separated by “-”. a machine that is still empty (which can only occur in a partial solution) is denoted by “0” in its sequence. following the previous example for π2 we have that jobs 1, 10 and 3 are assigned to the first factory. jobs 1 and 10 are assigned to the first machine in this factory in this order and job 3 to the second machine. since πh presents the job to machine-factory sequence of a single product h, πt, referred to as the final job sequence, is the concatenation of the different πh following the product sequence π. following the previous example, πt = {1-10-14, 3-8-5; 12-7-6-9, 4-13-2-11}. once all jobs in the assembly program of a product h are completed in the production stage, it can be assembled on the assembly stage. earliest assembling time of product h is denoted as eh. in this paper two methods are employed to construct the product sequence π. the first one uses the shortest processing time heuristic (spt). this dispatching rule is known to reduce the average number of jobs in the system, in-process inventories and average job tardiness (stafford et al., 2005). we obtain the spt order using the product assembly times and refer to this method as ps1. the second method, referred to as ps2, sorts the products in ascending order of the earliest assembling times (eh). in the method to make job to machine-factory assignments for products, we need first some additional notation. we refer to uh to the set of unscheduled jobs of product h assembly program, i.e., those jobs not yet assigned to machines at factories. skq is the set of jobs already scheduled at machine k inside factory q. with this in mind, the job to machine-factory assignment considers, for a product h, all jobs inside its assembly program, assigning first the unscheduled job with the earliest completion time at any machine in every eligible factory. more in details, we assign job j*∊uh to machine k* at factory q* satisfying: , , , , argmin j k q k m q lf j u p pj h ik jk i skq ! ! ! = + ) ) ) ! " ' , 1/ the process is applied until all jobs in the assembly program of product h are scheduled. both proposed constructive heuristics consist of three main steps: in the first step, the product sequence π is constructed. in the second step, the jobs inside the assembly program of each product are assigned following the previous job to machinefactory assignment procedure, following the order of products given in π. finally, in the third step the sequence of products for the assembly stage is obtained by sorting products according to eh in table 1. job processing times and factory eligibility constraints for the example. job j1 j2 j3 j4 j5 j6 j7 j8 j9 j10 j11 j12 j13 j14 machine m1 7 3 4 3 1 3 7 4 9 7 8 3 4 7 m2 1 6 5 4 5 9 2 1 6 8 4 9 1 3 lfj 1,2 1,2 1 2 1,2 2 2 1 2 1 2 1,2 2 1,2 17int. j. prod. manag. eng. (2015) 3(1), 13-23creative commons attribution-noncommercial-noderivatives 4.0 international the distributed parallel machine and assembly scheduling problem with eligibility constraints http://creativecommons.org/licenses/by-nc-nd/4.0/ ascending order. we propose two heuristics with identical second and third steps and with a different process to build the product sequence π in the first step. 4.1. heuristics pj1 and pj2 in heuristic pj1, ps1 is used to determine the product sequence π. after processing all jobs in the production stage, eh for each product h is calculated. the product sequence on the assembly machine is updated by sorting eh in ascending order and the final makespan is calculated. pseudocode 1 explains pj1 in detail: pseudocode 1: outline of the pj1 heuristic. obtain product sequence, π, applying ps1 use the job to machine-factory assignment procedure to assign all jobs of each product following the product order in π calculate earliest assembling time of each product h, eh determine the product sequence π on the assembly stage by sorting eh in ascending order the second heuristic pj2 needs some careful explanation. it uses method ps2 in the first step to make the product sequence π. however, ps2 requires sorting products in increasing order of eh. to calculate eh, all jobs must be assigned to factories and machines. in heuristic pj2, each product’s eh is calculated in isolation. to calculate eh of each product h, only jobs belong to product h are considered. once eh is calculated for all products, they are sorted in increasing order to form the product sequence π. this product sequence π is in turn used to apply again the job to machine-factory assignment for all products, which in the end gives us the final makespan. the difference between heuristic pj2, and the first heuristic pj1, is just on the first step. as mentioned before, heuristic pj2, uses ps2 to construct π. therefore, pseudocode of heuristic pj2 is not presented due to space constraints and because of its similarity with heuristic pj1. note that if there are ties in the eh of products, they are broken by taking the first product. also the same rule is considered for breaking ties on the spt rule which is used in heuristic pj1 to calculate π.as a final note, and to enforce the technological constraint that no machine should be left empty, if after the application of any of the two proposed heuristics, any machine is left empty, we reassign to it the job with the smallest processing time at that machine. the two proposed heuristics are applied to the previous example in the next section for further clarification. 4.2. heuristic application example the example of table 1 is used to detail heuristic pj1 first. products are first sorted according to shortest processing assembly times so π = {1, 3, 2}. in the second step, following the product order in π, first we assign jobs of product 1, to factories through the job to machine-factory assignment procedure. n1 = {2, 7, 8} so we first take job 2. the earliest completion time of this first job in all machines of all factories is 3. for job 7 is 2 (considering that it can only be assigned to factory 2) and for job 8 is 1 and can only be assigned to factory 1. the minimum is 1, which corresponds to the assignment of job 8 to the second machine of factory 1. note that if there is a tie in the minimum completion time for the jobs, it is broken by taking the first job. we now have to consider the unscheduled jobs 2 and 7. we now calculate the earliest completion times of these two jobs at all machines of all eligible factories considering that job 8 is already assigned. these minimum completion times are 3 and 2 for jobs 2 and 7, respectively. therefore job 7 is scheduled at factory 2 (the only eligible for this job) and to machine 2. lastly, job 2 is scheduled with the earliest completion time of 3 at factory 1, machine 1. note that we could have assigned this job to machine 1 of factory 2 with the same completion time, so we break ties by assigning jobs to the first table 2. instance and factors for proposed instances. instance factor symbol values ga gb gc number of jobs n 10, 12, 14, 16, 18 20, 22, 24 200, 300, 400 number of machines m 2, 3 2, 3, 4 5, 10, 15 number of factories f 2, 3 2, 3, 4 4, 6, 8 number of products t 2, 3, 4 2, 3, 4 20, 30, 40 18 int. j. prod. manag. eng. (2015) 3(1), 13-23 creative commons attribution-noncommercial-noderivatives 4.0 international hatami, s., ruiz, r. and andrés-romano, c. http://creativecommons.org/licenses/by-nc-nd/4.0/ machine and factory with equal completion time. after this procedure π1 = {2, 8; 0, 7}. following the same process, the jobs in the assembly programs of products 3 and 2 are assigned to factories one after the other, resulting in the final job sequence πt = {2-12-3, 8-14-1-10; 5-6-4-13, 7-11-9}. the completion times of all jobs at the production stage are: c1 = 5, c2 = 3, c3 = 10, c4 = 7, c5 = 1, c6 = 4, c7 = 2, c8 = 1, c9 = 12, c10 = 13, c11 = 6, c12 = 6, c13 = 11 and c14 = 4. the earliest assembling time for products 1 to 3, by considering their respective assembly programs are: e1 = 3, e2 = 13 and e3 = 12, respectively. in the third step, the product sequence π on the assembly stage is updated by sorting eh in ascending order, i.e., π = {1, 3, 2} and the cmax of the application of pj1 to this example is 31. for the second heuristic pj2 we calculate the eh values for all products one by one with the job to machine-factory assignment procedure, the obtained sequences are π1 = {2, 8; 0, 7 } with e1 = 3, π2 = {1210, 1-3; 4, 13} with e2 = 10 and π3 = {5, 14; 6, 11-9} with e3 = 10, so π = {1,2,3}. note that there is a tie in the eh of products 2 and 3 so again we break ties by taking the first product. using this π we apply again the job to machine-factory assignment procedure obtaining πt={2-12-10, 8-1-3; 4-5-6-9, 7-13-14-11} with completion times for the jobs as: c1= 2, c2= 3, c3= 7, c4= 3, c5= 4, c6= 7, c7= 2, c8= 1, c9= 16, c10= 13, c11= 10, c12= 6, c13= 3 and c14= 6. in the third step, again products are sorted in increasing order of their respective eh which are e1= 3, e2= 13 and e3= 16. therefore, the updated product sequence for the assembly stage is π ={1, 2, 3} with a makespan of 32. 5. computational evaluation to test the proposed milp model and constructive heuristics, six complete sets of instances have been generated. we consider different number of problem characteristics to comprehensively evaluate and test the proposed approaches: number of jobs (n), number of machines (m), number of factories (f) and number of products (t) are four controlled instance factors. depending on the chosen values we have small, medium and large-sized instances, referred to as ga, gb and gc, respectively. the processing times of the jobs on each machine in the production stage, are generated following a random uniform distribution in the range [1, 99], as it is common in the scheduling literature. the last instance factor we consider is the distribution of the assembly processing times which are fixed as: u[|nh|,49×|nh|] and u[|nh|,99×|nh|]. these two distributions are referred to in short as 50, and 100, respectively. the final sets of instances are then denoted as ga50, ga100,…,gc100. for each combination of instance factors we have five replications. the combinations for each instance size are given in table 2. therefore, the total number of instances is 300 for ga50 and another 300 for ga100 and 405 for every set in gb50 through gc100 resulting in a grand total of 2220 instances. 5.1. milp model evaluation the proposed milp model is tested only on sets ga and gb given the impossibility to solve large instances. two state-of-the-art commercial solvers are used, namely cplex 12.6 and gurobi 5.6.3, which are, at the time of the writing of this paper, the latest versions available. two different stopping times are tested with each solver: 900 and 3600 seconds. in total we have obtained 5640 results. all tests are performed in a high performance computing cluster with 30 blades, each one containing 16 gbytes of ram memory and two intel xeon e5420 processors running at 2.5 ghz. the 30 blade servers are used only to divide the workload since experiments are performed in virtualized windows xp machines, each one with a virtualized processor table 3. performance results for solvers and time limit for instance sets of ga50, ga100, gb50 and gb100. solver time limit 900s 3600s instance set ga50 ga100 gb50 gb100 ga50 ga100 gb50 gb100 cplex % opt 96.67 98.00 79.50 87.40 97.00 98.33 81.72 88.39 % outm 0.00 0.00 2.46 1.72 0.00 0.00 12.34 6.41 gap % 0.18 0.06 0.55 0.23 0.15 0.05 0.32 0.07 av time (sec.) 48.92 28.37 201.02 133.95 133.28 79.04 391.35 286.12 gurobi % opt 95.67 98.00 74.56 81.97 97.00 98.67 77.03 84.44 gap % 0.29 0.07 1.15 0.51 0.21 0.05 0.82 0.37 av time (sec.) 61.08 36.06 292.75 221.95 159.21 83.35 932.96 658.33 19int. j. prod. manag. eng. (2015) 3(1), 13-23creative commons attribution-noncommercial-noderivatives 4.0 international the distributed parallel machine and assembly scheduling problem with eligibility constraints http://creativecommons.org/licenses/by-nc-nd/4.0/ with two cores and 2 gb of ram memory. therefore, since both cplex and gurobi are parallel solvers, the two available threads at each virtual machine are used. after solving the models with cplex and gurobi, three possible outcomes are obtained. the first type is “optimal”, which means that an optimal solution with a given makespan value was obtained in the given maximum cpu time. the second type is “nonoptimal”, meaning that a feasible integer solution was obtained within the time limit but it was not possible to demonstrate its optimality and the gap is reported. the third and last outcome is “out of memory”, by which the solver had an error and ran out of ram memory, reporting a solution and a gap calculated with respect to the best obtained solution for that instance. in general, the solvers were able to find 294 (98.00%) and 297 (99.00%) optimal solutions in sets ga50 and ga100, respectively. for gb50 and gb100 the numbers are 338 (83.45%) and 363 (89.63%) for the 405 instances, respectively. the summarized results, according to the instance factors, type of solver and time limit, are presented in table 3 for sets ga and gb. the reported values at the tables are the percentage of optimum solutions found (% opt), the percentage of cases with out of memory error (% outm), the average gap for nonoptimal solution (gap %) and the average cpu time in seconds (av time). as we can see, the effect of the distribution of the assembly times at the assembly stage is much stronger than either the type of solver or cpu time limit. for group ga, instances with more disperse assembly times are easier to solve and also need less cpu time. as regards the comparison between cplex and gurobi, for set ga we see comparable performance with slightly shorter cpu times for cplex. for instance sets gb the differences between solvers are stronger. we see that gurobi is much slower than cplex and has higher gap values. however, cplex reports out of memory errors that in some cases average more than 12% (gb50). so it is important to conclude that there is no clear winner for this problem between these two solvers. in total, the largest tested instances in sets gb have 24 jobs and 16 machines distributed in 4 factories so we can attest that the proposed mathematical model has an adequate performance. 5.2. heuristics evaluation the two proposed heuristics, pj1 and pj2, are now tested. the response variable is the relative percentage deviation (rpd), measured as: lg rpd best a best 100 sol sol sol #= where bestsol is the best makespan obtained after all experimentation in this paper for any instance and algsol is the makespan obtained by the heuristic. the heuristics are coded in c# and are compiled under visual studio 2010. the same computing platform used for the milp evaluation is employed here. the average rpd values for the proposed heuristics are given in tables 4, 5 and 6 for instances sets ga, gb and gc, respectively. all results are grouped by n and f. the average rpd values of cplex and gurobi are reported as well for reference. as can be observed, pj2 is generally much better than pj1 in all groups of instances, although the difference is not very big in the large instances. it is important to observe how in the largest instances in set gb of 24 jobs and 4 factories, pj2, gives a very small gap of just 0.35% which indicates that pj2 is a very capable table 4. average relative percentage deviation (rpd) of cplex, gurobi and the proposed heuristics for instance sets ga50 and ga100. ga50 ga100 f×n cplex gurobi pj1 pj2 cplex gurobi pj1 pj2 2×10 0.00 0.00 9.52 3.26 0.00 0.00 2.21 0.46 2×12 0.00 0.00 8.24 4.04 0.00 0.00 3.58 1.76 2×14 0.00 0.00 8.29 3.36 0.00 0.00 4.50 0.62 2×16 0.00 0.00 9.31 4.59 0.00 0.00 4.30 1.16 2×18 0.00 0.21 7.27 3.90 0.00 0.00 3.34 1.23 3×10 0.00 0.00 5.10 2.59 0.00 0.00 1.09 1.24 3×12 0.00 0.00 4.72 1.43 0.00 0.00 1.85 0.98 3×14 0.00 0.00 4.51 1.71 0.00 0.00 1.51 0.28 3×16 0.00 0.00 4.79 1.39 0.00 0.00 2.63 0.72 3×18 0.00 0.00 4.04 1.34 0.00 0.00 2.15 0.78 average 0.00 0.02 6.58 2.76 0.00 0.00 2.72 0.92 20 int. j. prod. manag. eng. (2015) 3(1), 13-23 creative commons attribution-noncommercial-noderivatives 4.0 international hatami, s., ruiz, r. and andrés-romano, c. http://creativecommons.org/licenses/by-nc-nd/4.0/ heuristic with close to optimality performance. on average, pj2 is below 1% rpd for instance groups ga and gb. for the large instances in gc it is not possible to calculate the optimum solution so we only have an overall picture were pj2 always obtains the best solution. as a matter of fact and although not reported in detail here, among the 810 instances in gc50 and gc100, pj2 is always better or equal than pj1. we report now on the cpu times of the proposed heuristics in table 7. it has to be noted that cpu times are negligible, on the verge of being below the margin of error in measurements. as can be seen, the average cpu times are below one tenth of a second for the largest instances in group gc. on average, pj2 is relatively slower than pj1 but on absolute terms the cpu times are very small. although not shown here, the largest measured cpu time corresponds to heuristic pj2 and has been 0.41 seconds. from this final evaluation and considering the relative rpd of pj2 we can conclude that it is a capable and very fast heuristic. even though the observed differences are large in all cases for the proposed heuristics and very small for the two solvers, we carry out some statistical analyses in order to ascertain if the observed differences are indeed statistically significant. all results are examined with the analysis of variance (anova) technique. anova is a powerful parametric tool, which has been used in the last 10 years in the scheduling literature with great success. for the small instances there is no statistically significant difference in the performance of cplex and gurobi and pj2 is statistically better than pj1. the detailed data is not reported for space reasons. for the medium sized-instances in set gb we observe the interaction between the distribution of the assembly processing times and tested methods in figure 1. as can be seen, the results are similar to those of set ga. the differences between the proposed heuristics table 5. average relative percentage deviation (rpd) of cplex, gurobi and the proposed heuristics for instance sets gb50 and gb100. ga50 ga100 f×n cplex gurobi pj1 pj2 cplex gurobi pj1 pj2 2×20 0.20 0.00 7.03 2.33 0.00 0.00 3.33 1.03 2×22 0.29 0.03 5.26 2.36 0.05 0.03 2.36 0.87 2×24 0.13 0.19 4.78 1.89 0.11 0.04 3.33 1.40 3×20 0.00 0.00 3.21 1.42 0.00 0.00 2.48 1.27 3×22 0.01 0.01 2.89 1.69 0.04 0.02 1.54 0.47 3×24 0.10 0.02 3.17 1.25 0.00 0.02 1.38 0.70 4×20 0.00 0.00 2.31 1.29 0.00 0.00 1.83 0.67 4×22 0.00 0.00 2.18 0.77 0.00 0.00 1.78 0.75 4×24 0.00 0.00 2.50 1.12 0.00 0.00 1.82 0.35 average 0.08 0.03 3.70 1.57 0.02 0.01 2.21 0.84 table 6. average relative percentage deviation (rpd) of the proposed heuristics for instance sets gc50 and gc100. f×n gc50 gc100 pj1 pj2 pj1 pj2 4×200 0.13 0.00 0.07 0.00 4×300 0.12 0.00 0.05 0.00 4×400 0.11 0.00 0.06 0.00 6×200 0.09 0.00 0.04 0.00 6×300 0.08 0.00 0.05 0.00 6×400 0.08 0.00 0.04 0.00 8×200 0.06 0.00 0.03 0.00 8×300 0.08 0.00 0.03 0.00 8×400 0.05 0.00 0.04 0.00 average 0.09 0.00 0.05 0.00 -0.3 0.7 1.7 2.7 3.7 4.7 a ve ra ge r el at iv e pe rc en ta ge d ev ia tio n pj1 pj2 cplex gurobi assembly times distribution 50 100 figure 1. means plot with the interaction between the distribution of the assembly processing times and the tested methods for instances gb. all means have tukey’s honest significant difference (hsd) 95% confidence intervals. 21int. j. prod. manag. eng. (2015) 3(1), 13-23creative commons attribution-noncommercial-noderivatives 4.0 international the distributed parallel machine and assembly scheduling problem with eligibility constraints http://creativecommons.org/licenses/by-nc-nd/4.0/ are large enough so as to be statistically significant whereas the differences in the performance of the solvers are not statistically relevant. as for the large instances in group gc we can only test the significance in the observed differences in the average rpd between the two heuristics. this is given in figure 2. -0.003 0.017 0.037 0.057 0.077 a ve ra ge r el at iv e pe rc en ta ge d ev ia tio n pj1 pj2 figure 2. means plot for the two heuristics in large instances (gc). all means have tukey’s honest significant difference (hsd) 95% confidence intervals. as can be observed, pj2 is statistically better than pj1 even though the absolute difference between both proposed methods is practically small. 6. conclusions and future research in this paper we have studied an interesting combination of a distributed manufacturing problem with assembly operations. more specifically, we have presented a distributed unrelated parallel machines problem by which a number of factories, each one containing unrelated parallel machines have to manufacture jobs. all these jobs are later assembled into products in a factory with a single assembly machine. the objective is to minimize the makespan in the assembly stage. such a problem has been motivated and shown not to have been studied to date. we have presented a mathematical model and two constructive heuristics. the mathematical model has been comprehensively evaluated and tested using two state-of-the-art commercial solvers. results have shown that we are able to solve optimally problems of up to 24 jobs and 16 machines distributed in 4 factories. the two proposed heuristics are inherently simple and at the same time report solutions very close to optimal in the cases for which the optimal solution has been obtained. furthermore, for large instances, the performance is very good, obtaining solutions in less than half a second. while the studied problem has many potential applications, it is very likely for additional constraints to appear in practice. for example, sequence dependent setup times at machines are ubiquitous in real industries. more complex assembly stages with parallel assembly machines, or assembly flowshops, might be of interest. lastly, other objective functions, basically those based on due dates are worthy of additional studies. furthermore, metaheuristic techniques might improve the results of the mathematical models and proposed heuristics in a significant way. we expect dealing with some of these ideas in future work. acknowledgements the spanish ministry of economy and competitiveness supports rubén ruiz, under the project “result-realistic extended scheduling using light techniques” (no. dpi201236243-c02-01). carlos andrés is partially supported by the project “hybrid methods for horizontal cooperation in green transportation and logistics greencoop” tra2013-48180-c3-3-p from the spanish ministry of economy and competitiveness. references chan, f. t. s., chung, s. h., chan, p. l. y. (2005). an adaptive genetic algorithm with dominated genes for distributed scheduling problems. expert systems with applications, 29(2): 364–371. doi:10.1016/j.eswa.2005.04.009 chan, f. t. s., chung, s. h., chan, p. l. y., finke, g., tiwari, m. k. (2006). solving distributed fms scheduling problems subject to maintenance: genetic algorithms approach. robotics and computer-integrated manufacturing, 22(5-6): 493–504. doi:10.1016/j.rcim.2005.11.005 elmaraghy, h., schuh, g., elmaraghy, w., piller, f., schönsleben, p., tseng, m., bernard, a. (2013). product variety management. cirp annals-manufacturing technology, 62(2): 629–652. doi:10.1016/j.cirp.2013.05.007 fernandez-viagas, v., framinan, j.m. (2015). a bounded-search iterated greedy algorithm for the distributed permutation flowshop scheduling problem. international journal of production research, 53(4): 1111–1123. doi:10.1080/00207543.2014.948578 22 int. j. prod. manag. eng. (2015) 3(1), 13-23 creative commons attribution-noncommercial-noderivatives 4.0 international hatami, s., ruiz, r. and andrés-romano, c. http://dx.doi.org/10.1016/j.eswa.2005.04.009 http://dx.doi.org/10.1016/j.rcim.2005.11.005 http://dx.doi.org/10.1016/j.cirp.2013.05.007 http://dx.doi.org/10.1080/00207543.2014.948578 http://creativecommons.org/licenses/by-nc-nd/4.0/ hariri, a. m. a., potts, c. n. (1997). a branch and bound algorithm for the two-stage assembly scheduling problem. european journal of operational research, 103(3): 547–556. doi:10.1016/s0377-2217(96)00312-8 hatami, s., ruiz, r., andrés-romano, c. (2013). the distributed assembly permutation flowshop scheduling problem. international journal of production research, 51: 5292–5308. doi:10.1080/00207543.2013.807955 jia, h. z., fuh, j. y. h., nee, a. y. c., zhang, y. f. (2002). web-based multi-functional scheduling system for a distributed manufacturing environment. concurrent engineering: research and applications, 10(1): 27–39. jia, h. z., nee, a. y. c., fuh, j. y. h., zhang, y. f. (2003). a modified genetic algorithm for distributed scheduling problems. journal of intelligent manufacturing, 14(3-4): 351–362. doi:10.1023/a:1024653810491 jia, h. z., fuh, j. y. h., nee, a. y. c., zhang, y. f. (2007). integration of genetic algorithm and gantt chart for job shop scheduling in distributed manufacturing systems. computers & industrial engineering, 53(2): 313–320. doi:10.1016/j.cie.2007.06.024 lee, c. y., cheng, t. c. e., lin, b. m. t. (1993). minimizing the makespan in the 3-machine assembly-type flowshop scheduling problem. management science, 39(5): 616–625. doi:10.1287/mnsc.39.5.616 lenstra, j. k., rinnooy kan, a. h. g., brucker, p. (1977). complexity of machine scheduling problems. annals of discrete mathematics, 1:343–362. doi:10.1016/s0167-5060(08)70743-x lin, s. w., ying, k. c., huang, c. y. (2013). minimising makespan in distributed permutation flowshops using a modified iterated greedy algorithm. international journal of production research, 51(16): 5029–5038. doi:10.1080/00207543.2013.790571 lin, y., li, w. (2004). parallel machine scheduling of machine-dependent jobs with unit-length. european journal of operational research, 156(1): 261–266. doi:10.1016/s0377-2217(02)00914-1 mahdavi, i., shirazi, b., cho, n., sahebjamnia, n., ghobadi, s. (2008). modeling an e-based real-time quality control information system in distributed manufacturing shops. computers in industry, 59(8): 759–766. doi:10.1016/j.compind.2008.03.005 naderi, b., ruiz, r. (2010). the distributed permutation flowshop scheduling problem. computers & operations research, 37(4): 754–768. doi:10.1016/j.cor.2009.06.019 naderi, b., ruiz, r. (2014). a scatter search algorithm for the distributed permutation flowshop scheduling problem. european journal of operational research, 239(2): 323–334. doi:10.1016/j.ejor.2014.05.024 sluga, a., butala, p., bervar, g. (1998). a multi-agent approach to process planning and fabrication in distributed manufacturing. computers & industrial engineering, 35(3-4): 455–458. doi:10.1016/s0360-8352(98)00132-6 stafford, e. f., tseng, f. t., gupta, j. n. d. (2005). comparative evaluation of milp flowshop models. journal of the operational research society, 56(1): 88–101. doi:10.1057/palgrave.jors.2601805 sun, x., morizawa, k., nagasawa, h. (2003). powerful heuristics to minimize makespan in fixed, 3-machine, assembly-type flowshop scheduling. european journal of operational research, 146(3): 499–517. doi:10.1016/s0377-2217(02)00245-x sung, c. s., kim, h. a. (2008). a two-stage multiple-machine assembly scheduling problem for minimizing sum of completion times. international journal of production economics, 113(2): 1038–1048. doi:10.1016/j.ijpe.2007.12.007 wang, s. y., wang, l., liu, m., xu, y. (2013). an effective estimation of distribution algorithm for solving the distributed permutation flowshop scheduling problem. international journal of production economics, 145(1): 387–396. doi:10.1016/j.ijpe.2013.05.004 xiong, f., xing, k., wang, f., lei, h., han, l. (2014). minimizing the total completion time in a distributed two stage assembly system with setup times. computers & operations research, 47: 92–105. doi:10.1016/j.cor.2014.02.005 23int. j. prod. manag. eng. (2015) 3(1), 13-23creative commons attribution-noncommercial-noderivatives 4.0 international the distributed parallel machine and assembly scheduling problem with eligibility constraints http://dx.doi.org/10.1016/s0377-2217(96)00312-8 http://dx.doi.org/10.1080/00207543.2013.807955 http://dx.doi.org/10.1023/a:1024653810491 http://dx.doi.org/10.1016/j.cie.2007.06.024 http://dx.doi.org/10.1287/mnsc.39.5.616 http://dx.doi.org/10.1016/s0167-5060(08)70743-x http://dx.doi.org/10.1080/00207543.2013.790571 http://dx.doi.org/10.1016/s0377-2217(02)00914-1 http://dx.doi.org/10.1016/j.compind.2008.03.005 http://dx.doi.org/10.1016/j.cor.2009.06.019 http://dx.doi.org/10.1016/j.ejor.2014.05.024 http://dx.doi.org/10.1016/s0360-8352(98)00132-6 http://dx.doi.org/10.1057/palgrave.jors.2601805 http://dx.doi.org/10.1016/s0377-2217(02)00245-x http://dx.doi.org/10.1016/j.ijpe.2007.12.007 http://dx.doi.org/10.1016/j.ijpe.2013.05.004 http://dx.doi.org/10.1016/j.cor.2014.02.005 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2016.4102 received 2015-09-02 accepted: 2016-04-29 flow shop scheduling decisions through techniques for order preference by similarity to an ideal solution (topsis) arun gupta a* and shailendra kumar b a mechanical engineering department, college of engineering, teerthanker mahaveer university, moradabad (up) india. * engg.arun12@gmail.com b mechanical engineering department, meerut institute of engineering and technology, meerut (up) india. tyagi_sk@yahoo.com abstract: the flow-shop scheduling problem (fsp) has been widely studied in the literature and having a very active research area. over the last few decades, a number of heuristic/meta-heuristic solution techniques have been developed. some of these techniques offer excellent effectiveness and efficiency at the expense of substantial implementation efforts and being extremely complicated. this paper brings out the application of a multi-criteria decision making (mcdm) method known as techniques for order preference by similarity to an ideal solution (topsis) using different weighting schemes in flow-shop environment. the objective function is identification of a job sequence which in turn would have minimum makespan (total job completion time). the application of the proposed method to flow shop scheduling is presented and explained with a numerical example. the results of the proposed topsis based technique of fsp are also compared on the basis of some benchmark problems and found compatible with the results obtained from other standard procedures. key words: scheduling, multi-criteria decision making, topsis, flow-shop. 1. introduction in modern world, where things are changing with a pace never before and every aspect of human life is affected from this change. every industry is looking for strategies and technologies not only to retain their profit margins and market-share but to improve them, manufacturing industry is not an exception. concept of cellular manufacturing seems to be a solution in this situation. an enormous and growing body of literature is available on it. in simplest words cellular manufacturing is a manufacturing process in which every manufacturing cell acts as an independent manufacturing unit. cell formation, cell layout and scheduling are the three basic steps in design of any cellular manufacturing system (kumar and sharma, 2015). cell formation involves identification and grouping of machine cells and part families (albadawi et al., 2005; kumar and sharma, 2014, 2015). cell layout emphasizes placing of machines within a manufacturing cell. scheduling is simply sequencing of tasks on different machines in a manufacturing cell. researchers are working tirelessly on all three stages of cellular manufacturing, considering them independently as well as concurrently. out of these three stages, scheduling is one of the important and critical issue in the production planning and operation of a manufacturing system. scheduling is the allocation of a set of limited resources to a number of jobs over time, with the objective of optimizing one or more performance criterion (french, 1982). in the set of machine, scheduling finds a sequence of jobs with constraints to optimize one or more objectives such as makespan, tardiness, work-in-process inventory, number of tardy jobs, idle time, etc. (nakhaeinejad and nahavandi, 2012). two common issues that frequently appear in the scheduling literature are flow shop scheduling (dannenbring, 1977; king and spachis, 1980; taillard, 1990; fink and vob, 2003) and job shop scheduling (guinet and legrand, 1998; cheung and zhou, 2001; wang and zheng, 2001; schuster and framinan, 2003). in flow shop scheduling, it is generally assumed that the jobs must be processed on the machines in the same technological or machine order. in job shop scheduling, however, jobs are commonly processed int. j. prod. manag. eng. (2016) 4(2), 43-52creative commons attribution-noncommercial-noderivatives 4.0 international 43 http://dx.doi.org/10.4995/ijpme.2016.4102 http://creativecommons.org/licenses/by-nc-nd/4.0/ following different machine orders (laha and chakraborty, 2009). flow shop scheduling has become one of the most popular scheduling problems due to its amazingly increasing practical usage in manufacturing industries (nakhaeinejad and nahavandi, 2012). the flow shop scheduling performs a set of jobs on a set of dedicated machines, where each job follows the same processing operation order. each machine processes one job at a time and each job is processed on one machine at a time without pre-emption.the criterion that is most commonly studied in flow-shop scheduling literature is the minimization of total completion time, better to say makespan (cmax), of the production sequence. researchers have optimised makespan in flow-shop problems by using various techniques like statistical tools, artificial neural network, fuzzy logic, genetic algorithm, simulated annealing, tabu search, etc. and combination of these /such techniques (ruiz et al., 2006; zobolas et al., 2009). multi-criteria decision making (mcdm) methods found their application in scheduling problems too. techniques for order preference by similarity to the ideal solution (topsis) is such a technique which can be effectively used in scheduling decisions, however a comparatively less literature is observed on application of it in scheduling decision. in this paper, efforts to bring out the application of topsis in fsp with different weight factors in scheduling problems are made. the organisation of rest of the paper is as follows: review of literature relevant to flow shop scheduling is carried out in section 2. procedure for computation of makespan, and topsis are explained in section 3, while the implementation of the same on an illustrative problem is elaborated in section 4. results are given and compared for same problem with different weight factors also, for different problems with different methods in section 5. conclusions are drawn in section 6. 2. literature review literature pertaining to flow shop scheduling problem is studied in the following categories on the basis of problem solving techniques: i. exact/statistical methods ii. heuristic techniques iii. meta-heuristic techniques iv. hybrid heuristic/meta-heuristic techniques. the flow shop scheduling has been widely studied by the researchers because of its np-completeness in the sense when number of machines is greater than three (garey et al., 1976; gonzalez and sahani, 1978, may be referred). due to the complexity of the problem, exact methods for the general flow shop scheduling (fsp) failed to achieve high quality solutions for problems of large size in reasonable computational time. therefore, academic research focused on heuristic approaches rather than exact methods to solve scheduling problems involving a large number of jobs (gupta and chauhan, 2015). the flow-shop problem (fsp) was first studied by johnson (1954) for two machines. many researchers have generalized the johnson’s rule to ‘m’ machine flow shop problems (ruiz and stutzle, 2007). it is proved that m-machine flow shop problem with the makespan objective is np-hard (garey et al., 1976; french, 1982). the first heuristic for makespan minimization for the flow shop scheduling problem was introduced by palmer (1965). subsequent work includes the one on the cds heuristic (campbell et al., 1970) and rapid access (ra) heuristic (dannenbring, 1977). the neh heuristic by nawaz et al. (1983) was regarded as the best performing heuristic method (turner and booth, 1987; ruiz and maroto, 2005). more advanced methods are published by koulamas (1998), davoud pour (2001), and laha and chakraborty (2009). various heuristics with makespan as decision criterion have been reviewed by hejazi and saghafian (2005), and ruiz and maroto (2005). in the pursuit for solutions closer to the optimum, it has become inevitable that new solution approaches should be followed by some difficult problem instances and academic interest switched to artificial intelligence based optimization methods, and meta-heuristics (zobolas et al., 2009). simulated annealing based algorithms proposed by osman and potts (1989), and ogbu and smith (1990) are the first proposed meta-heuristics for the fsp. widmer and hertz (1989), taillard (1990), reeves (1993) and nowicki and smutnicki (1996) demonstrated different tabu search approaches. genetic algorithms for solving the fsp have also appeared in chen et al. (1995), reeves (1995), wang and zheng (2003), and aldowaisan and allahvedi (2003). other algorithms are the path-based method of werner (1993), the iterated local search of stützle (1998), two very effective ant-colony optimization algorithms by rajendran and ziegler (2004), and a fast tabu search approach of grabowski and wodecki (2004). int. j. prod. manag. eng. (2016) 4(2), 43-52 creative commons attribution-noncommercial-noderivatives 4.0 international gupta, a. and kumar, s. 44 http://creativecommons.org/licenses/by-nc-nd/4.0/ the results generated by meta-heuristics are found near to the best one/optimum, which provide the basis for researchers to develop advanced hybrid approaches, by combining different concepts or components of more than one meta-heuristic (blum and roli, 2003). hybridization, when properly applied, may further enhance the effectiveness of the solution space search, and may overcome any inherent limitations of single meta-heuristic algorithms. therefore, new opportunities emerge, which may lead to even more powerful and flexible solution methods for combinatorial optimization problems. even though, by the use of these methods the quality of solution is improved, but these methods are complex and iterative in nature, need special programming skills and require more computations to arrive a solution of flow shop problem. on the other hand, topsis is a simple multi-criteria decision making techniques that provides the solution in few steps. the philosophy and procedure of topsis could be understood easily. further it does not require any special computation technique and advance mathematics for its implementation in a flow shop scheduling problem. 3. methodology the proposed work is an effort of implementation of pioneer work of yoon and hwang (1980) (called topsis) in fsp. the methodology adopted in this paper is described in two subsections. in first subsection details of topsis and its implementation steps in fsp for generation of job sequence whilst in second subsection the computation of makespan (i.e. total completion time) for selected job sequence is explained 3.1. generation of job sequence by technique for order preference by similarity to ideal solution (topsis) topsis developed by yoon and hwang (1980), is a simple and one of the most commonly utilized multicriteria/ multi-attribute decision making (mcdm/ madm) procedure for wide range of real world problems. it helps decision makers to carryout comparisons, rankings and analysis among various available options in order to select a best one. (behzadian et al., 2012; shih et al., 2007; vega et al., 2014). it attempts to choose alternatives that simultaneously have the shortest distance from the positive ideal solution and the farthest distance from the negative-ideal solution. the positive ideal solution maximizes the benefit criteria and minimizes the cost criteria, whereas the negative ideal solution maximizes the cost criteria and minimizes the benefit criteria (behzadian et al., 2012; sarraf et al., 2013). thus, the best alternative, is the alternative having shortest relative distance from positive ideal solution and farthest distance from negative ideal solution. researchers have developed large number of variants of topsis by combining it with different distribution technique, for those (behzadian et al., 2012) may be referred. the stepwise implementation of standard topsis in flow shop scheduling problem is explained through a selfexplanatory flow chart shown in figure 1. recently, a heuristic based on the reduced weightage scheme of machines to generate different combination of sequences for a fsp has been developed by gupta and chauhan, (2015). similarly, here the weight factors used in topsis are selected as per different weighting schemes to obtain a combination of job sequences in order to minimize the makespan. 3.2. computation of makespan in a flow-shop scheduling problem, a set of n jobs (1, …, n) are processed on a set of m machines (1, …, m) in the same technological order, i.e. first in machine 1 then on machine 2 and so on until machine m. the objective is to find a sequence for the processing of the jobs in the machines so that the total completion time or makespan of the schedule (cmax) is minimized. let ti,j denote the processing time of the job in position i (i = 1, 2, …, n) on machine j (j =1, 2, …, m). let ci,j represents the completion time of the job in position i on machine j. therefore, c1,1 = t1,1 (1) ci,1 = ci-1,1 + ti,1 for i = 2, …, n (2) c1,j = c1,j-1 + t1,j for j = 2, …,m (3) ci,j = max (ci,j-1 , ci-1,j ) + ti> (4) for i = 2, …, n j =2, …, m total completion time (cmax) = cn,m int. j. prod. manag. eng. (2016) 4(2), 43-52creative commons attribution-noncommercial-noderivatives 4.0 international flow shop scheduling decisions through techniques for order preference by similarity to an ideal solution (topsis) 45 http://creativecommons.org/licenses/by-nc-nd/4.0/ gupta, a. & kumar, s. creative commons attribution-noncommercial-noderivatives 4.0 international int. j. prod. manag. eng. (2016) 4(2), ppp-ppp | 1 figure 1: steps for implementation of standard topsis in flow shop scheduling step 1: draw a matrix containing processing time tij for each job ‘i' on each machine ‘j’ step 2: construct normalized decision matrix nij = tij tij 2 i=1 n ∑ for i =1, 2,…., n ; j =1, 2,…., m where tij and nij are original and normalized processing time for decision matrix, respectively step 3: assign the weights wj for each machine ‘j’ based on selected weightage criteria step 4: construct the weighted normalized decision matrix nwv ijiij ⋅= step 5: determine the positive-ideal and negative-ideal solutions. a+ = v1 +,v2 +,...,vm +{ } = max j vij j ∈ ωb( ), min j vij j ∈ ωc( ){ } a− = v1 −,v2 −,...,vm −{ } = min j vij j ∈ ωb( ), max j vij j ∈ ωc( ){ } where, ωb is the set of benefit criteria and ωc is the set of cost criteria. step 6: calculate the separation measures for each alternative (job) (i) the separation from positive-ideal alternative – si + = vj + −vij( ) 2 j=1 m ∑ , i = 1, 2,…., n, (ii) the separation from negative-ideal alternative – si − = vj − −vij( ) 2 j=1 m ∑ , i = 1, 2,…., n, step 7: calculate the relative closeness to the ideal solution. rci = si − si + +si − , i = 1, 2,…., n, 0 ≤ rci ≤ 1 step 8: ranking the alternatives (jobs) according to the decreasing value of rci and make the job sequence. step 9: compute the value of makespan according to the job sequence. figure 1. steps for implementation of standard topsis in flow shop scheduling. int. j. prod. manag. eng. (2016) 4(2), 43-52 creative commons attribution-noncommercial-noderivatives 4.0 international gupta, a. and kumar, s. 46 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. implementation with illustration the proposed flow shop scheduling procedure is implemented on an arbitrarily designed flow shop scheduling problem illustrated below in subsection 4.1. 4.1. illustrative problem for illustration purpose, a flow shop scheduling problem of five jobs and five machines with random data has been developed and given in table 1 as processing time matrix. jobs and machines are represented in rows and columns of problem matrix respectively. order of machines are pre-fixed i.e. each job has to be processed in a pre-defined sequence of machines. each cell in the matrix represents the processing time of respective job on machine concerned. in order to minimize makespan (total job completion time), job sequence is to be determined. table 1. processing time matrix for flow shop scheduling problem. jobs/machines m1 m2 m3 m4 m5 j1 8 12 9 6 4 j2 14 10 11 2 15 j3 10 7 8 11 2 j4 7 8 14 9 2 j5 3 9 5 13 8 4.2. implementation of standard topsis in the above flow shop problem step 1: first of all, a problem matrix is introduced that consists of the processing time of each job on each machine (refer table 1). step 2: the normalized decision matrix is constructed as per the procedure given in step 2 of figure 1. it transforms the various attribute dimensions into nondimensional attributes for the purpose of comparison across the attributes. the normalized matrix for the illustrative problem is given in table 2. table 2. normalized decision matrix. jobs/machines m1 m2 m3 m4 m5 j1 0.39 0.57 0.41 0.30 0.23 j2 0.68 0.48 0.50 0.10 0.85 j3 0.49 0.33 0.36 0.54 0.11 j4 0.34 0.38 0.63 0.44 0.11 j5 0.15 0.43 0.23 0.64 0.45 step 3: weight factors ‘wj’ are assigned for each machine ‘j’ based on selected weightage scheme. here, the weight factors are selected based on the decreasing rank order of machine. step 4: now, the weighted normalized decision matrix is constructed by multiplying the elements of normalized decision matrix by the relevant weight factor. thus, obtained weighted normalized decision matrix for the illustration is given in table 3. table 3. weighted normalized decision matrix jobs/machines m1 m2 m3 m4 m5 j1 1.96 2.29 1.22 0.59 0.23 j2 3.42 1.91 1.50 0.20 0.85 j3 2.45 1.34 1.09 1.09 0.11 j4 1.71 1.53 1.90 0.89 0.11 j5 0.73 1.72 0.68 1.28 0.45 step 5: at this stage, the positive-ideal and negativeideal solutions are determined as per the details given in figure 1. since the matrix (table 3) is based on the processing times of jobs on respective machines, therefore, in order to minimize the total completion time of jobs (i.e. makespan), the minimum and maximum value in each column of the weighted normalized decision matrix is selected as positiveideal (highlighted by bold letters in table 4) and negative ideal (highlighted by bold letters in table 5) solution respectively. table 4. positive-ideal solution shown in weighted normalized decision matrix. jobs/machines m1 m2 m3 m4 m5 j1 1.96 2.29 1.22 0.59 0.23 j2 3.42 1.91 1.50 0.20 0.85 j3 2.45 1.34 1.09 1.09 0.11 j4 1.71 1.53 1.90 0.89 0.11 j5 0.73 1.72 0.68 1.28 0.45 positive ideal solution a+ = (0.73, 1.34, 0.68, 0.20, 0.11) table 5. negative -ideal solution shown in weighted normalized decision matrix. jobs/machines m1 m2 m3 m4 m5 j1 1.96 2.29 1.22 0.59 0.23 j2 3.42 1.91 1.50 0.20 0.85 j3 2.45 1.34 1.09 1.09 0.11 j4 1.71 1.53 1.90 0.89 0.11 j5 0.73 1.72 0.68 1.28 0.45 negative ideal solution a= (3.42, 2.29, 1.90, 1.28, 0.85) int. j. prod. manag. eng. (2016) 4(2), 43-52creative commons attribution-noncommercial-noderivatives 4.0 international flow shop scheduling decisions through techniques for order preference by similarity to an ideal solution (topsis) 47 http://creativecommons.org/licenses/by-nc-nd/4.0/ step 6: this stage requires the calculation of the separation measures from positive-ideal and negative-ideal solutions for each alternative (job) according to step 6 of figure 1. the separation of each alternative (job) from positive ideal and negative ideal solution is tabulated in table 6 and 7 respectively. table 6. separation from positive-ideal solution. jobs separation from positive-ideal solution j1 1.70 j2 2.97 j3 1.97 j4 1.72 j5 1.20 table 7. separation from negative-ideal solution. jobs separation from negative-ideal solution j1 1.86 j2 1.22 j3 1.76 j4 2.05 j5 3.03 step 7: now, relative closeness to the ideal solution is calculated as per the details elaborated in step 7 of figure 1. the same is presented in table 8. table 8. relative closeness to ideal solution for each alternative (job). jobs relative closeness to ideal solution (rci) j1 0.52 j2 0.29 j3 0.47 j4 0.54 j5 0.72 step 8: jobs (alternatives) are ranked according to the decreasing value of relative closeness rci and job sequence are made to be processed on the predefined order of machines. from the above table, the job sequence j5-j4-j1-j3-j2 has been obtained. step 9: finally, makespan value is computed based on the above job sequence using the equation 1-4. the makespan value of 80 units is obtained for the illustrative problem, according to the job sequence generated in step 8. 5. result comparison results are compared in two phases, as explained in section 5.1 and 5.2. 5.1. comparison on the basis of different weighting scheme: in this phase the topsis procedure is repeated for the same illustrative problem for different weightage schemes (explained below) to obtain different makespan value. i. decreasing order – under this criteria, machine having the first rank (or position), is given highest weightage which decreases with the number of rank increases. ii. increasing order – under this criteria, machine having the first rank (or position), is given lowest weightage which decreases with the number of rank increases. iii. higher machine utilization – a machine with the highest load (or utilization), is given highest weightage, which decreases with decreasing the load. iv. lower machine utilization – a machine with the higher load (or utilization), is given lowest weightage which increases with decreasing the load. v. equal weightage – all the machines is assigned with the equal weights to process jobs. job sequences are generated according to above different weighting schemes, by using standard topsis procedure. then makespan is computed for each job sequence. the same is summarized in table 9. table 9. comparison of makespan with different weightage scheme. s. no. weightage for machines job sequence obtained through topsis makespan (units)weightage scheme weight factors 1. decreasing order 5, 4, 3, 2, 1 j5-j4-j1-j3-j2 80 2. increasing order 1, 2, 3, 4, 5 j1-j3-j4-j5-j2 96 3. higher machine utilization 3, 4, 5, 2, 1 j5-j3-j1-j4-j2 84 4. lower machine utilization 3, 2, 1, 4, 5 j1-j4-j3-j5-j2 99 5. equal weightage 1, 1, 1, 1, 1 j1-j3-j4-j5-j2 96 int. j. prod. manag. eng. (2016) 4(2), 43-52 creative commons attribution-noncommercial-noderivatives 4.0 international gupta, a. and kumar, s. 48 http://creativecommons.org/licenses/by-nc-nd/4.0/ selection of weighting scheme is one of key consideration in any topsis application (olson, 2004). in this problem, a minimum value of makespan is observed with decreasing order weight scheme. makespan value for job sequence based on weighting scheme namely higher machine utilisation is close to minimum makespan value. 5.2. result comparison with some standard methods first, the above problem is solved by five well-known heuristic algorithms namely palmer (palmer, 1965), gupta (gupta, 1971), campbell, dudek and smith (cds) (campbell et al., 1970), rapid access (ra) (dannenbring, 1977) and nawaz, enscore and ham (neh) (nawaz et al., 1983) makespan values of 85, 82, 82, 79 and 79 respectively are obtained. for results from topsis based proposed method, all five weighting schemes discussed above are implemented on a particular problem, and the minimum value of those, presented here, is 80. results are found to be compatible. secondly, for better comparison, results of some benchmark problems are compared with other well-known methods. these benchmark problems prescribed by j. carlier (1978) and c.r. reeves (1995) are taken from the or-library (http://people. brunel.ac.uk/~mastjjb/jeb/orlib/files/flowshop1.txt). 11 problem instances are taken for comparison purpose, 8 instances of carlier and 3 instances of reeves (1995). these problems are designed the for the purpose of comparing the heuristic algorithms with an objective of makespan minimization. details of problems taken for comparison are shown in table 10. for above tabulated problems, minimum makespan time for the completion of the process is calculated based on the schedule derived by the respective algorithms. the results are presented in table 11. the minimum figures across the algorithms are presented by bold letters in the table. table 10. benchmark problems taken for result comparison. s. no. size of the problems instances no. of jobs no. of machines 1. carlier 01 11 5 2. carlier 02 13 4 3. carlier 03 12 5 4. carlier 04 14 4 5. carlier 05 10 6 6. carlier 06 8 9 7. carlier 07 7 7 8. carlier 08 8 8 9. rec 01 20 5 10. rec 03 20 5 11. rec 05 20 5 the results indicate that introduction of topsis method in the field of scheduling provides compatible results to some extent. for benchmark problems of 10x6 and 7×7, result of proposed method is closest to the least makespan values obtained by neh and cds respectively and also for other problems, makespan value is near about the least ones. most of the minimum results obtained from proposed methods are obtained by using the decreasing order weighting scheme. table 11. comparison of makespan time obtained by topsis based proposed method with other standard heuristics. s. no. problem instances makespan (time measurable unit) proposed method palmer gupta cds ra neh 1. carlier01-11x5 7332 7472 7348 7202 7817 7038 2. carlier02-13x4 8123 7940 7534 7410 7509 7940 3. carlier03-12x5 8567 7725 7399 7399 7399 7503 4. carlier04-14x4 9170 8423 8423 8423 8357 8003 5. carlier05-10x6 8309 8520 8773 8627 8940 8190 6. carlier06-8x9 9647 9487 9441 9553 9514 9159 7. carlier07-7x7 7563 7639 7639 6819 6923 7668 8. carlier08-8x8 9345 9023 9224 8903 9062 9032 9. rec01-20x5 1595 1391 1434 1399 1399 1334 10. rec03-20x5 1289 1223 1380 1273 1159 1136 11. rec05-20x5 1479 1290 1429 1338 1434 1294 int. j. prod. manag. eng. (2016) 4(2), 43-52creative commons attribution-noncommercial-noderivatives 4.0 international flow shop scheduling decisions through techniques for order preference by similarity to an ideal solution (topsis) 49 http://people.brunel.ac.uk/~mastjjb/jeb/orlib/files/flowshop1.txt http://people.brunel.ac.uk/~mastjjb/jeb/orlib/files/flowshop1.txt http://creativecommons.org/licenses/by-nc-nd/4.0/ 6. conclusion this paper successfully presented a new approach based on a mcdm method called ‘topsis’ for minimization of makespan criterion in flow shop scheduling. the technique demonstrated is simple, easy to understand, and implement. it has been illustrated by using five different weighting schemes to generate the job sequences. the makespan value obtained from these sequences shows the compatibility of this technique in flow shop environment. the results can further be improved by incorporating a weighting scheme which could consider machine utilization and other production related parameters. the proposed procedure could also be implemented for other scheduling problems such as job shop, flexible job shops and flow shops. references albadawi, z., bashir, h.a., chen, m. (2005). a mathematical approach for the formation of manufacturing cells. computers & industrial engineering, 48(1): 3–21. http://dx.doi.org/10.1016/j.cie.2004.06.008 aldowaisan, t., allahvedi, a. (2003), new heuristics for no-wait flowshops to minimize makespan. computers & operations research, 30(8): 1219–31. http://dx.doi.org/10.1016/s0305-0548(02)00068-0 behzadian, m., otaghsara, s., yazdani, m., ignatius, j. (2012). a state-of the-art survey of topsis applications. expert systems with applications, 39(17), 13051-13069. http://dx.doi.org/10.1016/j.eswa.2012.05.056 blum, c., roli, a. (2003). metaheuristics in combinatorial optimization: overview and conceptual comparison. acm computing surveys, 35(3), 268–308. http://dx.doi.org/10.1145/937503.937505 campbell, h.g., dudek, r.a., smith, m.l. (1970). a heuristic algorithm for the n-job, m-machine problem. management science, 16(10), b630–7. http://dx.doi.org/10.1287/mnsc.16.10.b630 carlier, j. (1978). ordonnancements a contraintes disjonctives, r.a.i.r.o. recherche operationelle/operations research, 12(4), 333-351. chen, c-l., vempati, v.s., aljaber, n. (1995). an application of genetic algorithms for flow shop problems. european journal of operational research, 80(2), 389-396. http://dx.doi.org/10.1016/0377-2217(93)e0228-p cheung, w.m., zhou, h. (2001). using genetic algorithms and heuristics for job shop scheduling with sequence-dependent setup times. annals of operations research, 107(1), 65–81. http://dx.doi.org/10.1023/a:1014990729837 dannenbring, d.g. (1977). an evaluation of flow-shop sequencing heuristic. management science, 23(11), 1174–1182. http://dx.doi. org/10.1287/mnsc.23.11.1174 davoud pour, h. (2001). a new heuristic for the n-job, m-machine flow-shop problem. production planning and control, 12(7), 648–53. http://dx.doi.org/10.1080/09537280152582995 fink, a., vob, s. (2003). solving the continuous flow-shop scheduling problem by metaheuristics. european journal of operational research, 151(2), 400–414. http://dx.doi.org/10.1016/s0377-2217(02)00834-2 french, s. (1982). sequencing and scheduling: an introduction to the mathematics of the job-shop. ellis horwood limited. garey, m.r.d., johnson, d.s., sethi, r. (1976). the complexity of flow shop and job shop scheduling. mathematics of operations research, 1, 117-29. http://dx.doi.org/10.1287/moor.1.2.117 gonzalez, t., sahni, s. (1978). flow shop and job shop schedules. operations research, 26, 36–52. http://dx.doi.org/10.1287/opre.26.1.36 grabowski, j., wodecki, m. (2004). a very fast tabu search algorithm for the permutation flowshop problem with makespan criterion. computers and operations research, 31(11), 1891–909. http://dx.doi.org/10.1016/s0305-0548(03)00145-x guinet, a., legrand, m. (1998). reduction of job-shop problems to flow-shop problems with precedence constraints. european journal of operational research, 109(1), 96–110. http://dx.doi.org/10.1016/s0377-2217(97)00129-x gupta, a., chauhan, s. (2015). a heuristic algorithm for scheduling in a flow shop environment to minimize makespan. international journal of industrial engineering computations, 6(2), 173-184. http://dx.doi.org/10.5267/j.ijiec.2014.12.002 hejazi, s.r., saghafian, s. (2005). flowshop-scheduling problems with makespan criterion: a review. international journal of production research, 43(14), 2895–2929. http://dx.doi.org/10.1080/0020754050056417 johnson, s.m. (1954). optimal two and three-stage production schedules with set-up times included. naval research logistics quarterly, 1(1), 61-68. http://dx.doi.org/10.1002/nav.3800010110 king, j.r., spachis, a.s. (1980). heuristics for flowshop scheduling. international journal of production research, 18(3), 343–357. http:// dx.doi.org/10.1080/00207548008919673 koulamas, c. (1998). a new constructive heuristic for the flowshop scheduling problem. european journal of operational research; 105(1), 66–71. http://dx.doi.org/10.1016/s0377-2217(97)00027-1 kumar, s., sharma, r.k. (2014). cell formation heuristic procedure considering production data. international journal of production management and engineering, 2(2), 75–84. http://dx.doi.org/10.4995/ijpme.2014.2078 int. j. prod. manag. eng. (2016) 4(2), 43-52 creative commons attribution-noncommercial-noderivatives 4.0 international gupta, a. and kumar, s. 50 http://dx.doi.org/10.1016/j.cie.2004.06.008 http://dx.doi.org/10.1016/s0305-0548(02)00068-0 http://dx.doi.org/10.1016/j.eswa.2012.05.056 http://dx.doi.org/10.1145/937503.937505 http://dx.doi.org/10.1287/mnsc.16.10.b630 http://dx.doi.org/10.1016/0377-2217(93)e0228-p http://dx.doi.org/10.1023/a:1014990729837 http://dx.doi.org/10.1287/mnsc.23.11.1174 http://dx.doi.org/10.1287/mnsc.23.11.1174 http://dx.doi.org/10.1080/09537280152582995 http://dx.doi.org/10.1016/s0377-2217(02)00834-2 http://dx.doi.org/10.1287/moor.1.2.117 http://dx.doi.org/10.1287/opre.26.1.36 http://dx.doi.org/10.1016/s0305-0548(03)00145-x http://dx.doi.org/10.1016/s0377-2217(97)00129-x http://dx.doi.org/10.5267/j.ijiec.2014.12.002 http://dx.doi.org/10.1080/0020754050056417 http://dx.doi.org/10.1002/nav.3800010110 http://dx.doi.org/10.1080/00207548008919673 http://dx.doi.org/10.1080/00207548008919673 http://dx.doi.org/10.1016/s0377-2217(97)00027-1 http://dx.doi.org/10.4995/ijpme.2014.2078 http://creativecommons.org/licenses/by-nc-nd/4.0/ kumar, s., sharma, r.k. (2015). development of a cell formation heuristic by considering realistic data using principal component analysis and taguchi’s method. journal of industrial engineering international, 11(1), 87-100. http://dx.doi.org/10.1007/s40092-014-0093-3 laha, d., chakraborty, u.k. (2009). an efficient hybrid heuristic for makespan minimization in permutation flow shop scheduling. international journal of advanced manufacturing technology, 44(5), 559–569. http://dx.doi.org/10.1007/s00170-008-1845-2 nakhaeinejad, m., nahavandi, n. (2012). an interactive algorithm for multi-objective flow shop scheduling with fuzzy processing time through resolution method and topsis. international journal of advanced manufacturing technology, 66(5), 1047-1064. http://dx.doi. org/10.1007/s00170-012-4388-5 nawaz, m., enscore, jr. e., ham, i. (1983). a heuristic algorithm for the m-machine, n-job flow-shop sequencing problem. omega, international journal of management science, 11, 91–95. nowicki, e., smutnicki, c. (1996). a fast tabu search algorithm for the permutation flow-shop problem. european journal of operational research, 91(1), 160–75. http://dx.doi.org/10.1016/0377-2217(95)00037-2 ogbu, f.a., smith, d.k. (1990). the application of the simulated annealing algorithms to the solution of the n/m/c max flowshop problem. computers & operations research, 17(3), 243–53. http://dx.doi.org/10.1016/0305-0548(90)90001-n olson, d.l. (2004). comparison of weights in topsis models. mathematical and computer modeling, 40(7-8), 721–727. http://dx.doi. org/10.1016/j.mcm.2004.10.003 osman, i.h., potts, c.n. (1989). simulated annealing for permutation flow-shop scheduling. omega, the international journal of management science, 17(6), 551–7. palmer, d.s. (1965). sequencing jobs through a multistage process in the minimum total time: a quick method of obtaining a near optimum. operations research quarterly, 16(1), 101–7. http://dx.doi.org/10.1057/jors.1965.8 rajendran, c., ziegler, h. (2004). ant-colony algorithms for permutation flowshop scheduling to minimize makespan/total flowtime of jobs. european journal of operational research, 155(2), 426–38. http://dx.doi.org/10.1016/s0377-2217(02)00908-6 reeves, c.r. (1993). improving the efficiency of tabu search for machine scheduling problems. journal of the operational research society, 44(4), 375–82. http://dx.doi.org/10.2307/2584415 reeves, c.r., (1995). a genetic algorithm for flowshop sequencing. computers and operations research 22(1), 5–13. http://dx.doi. org/10.1016/0305-0548(93)e0014-k ruiz, r., maroto, c., alcaraz, j. (2006). two new robust genetic algorithms for the flowshop scheduling problem. omega, international journal of management science, 34, 461–76. http://dx.doi.org/10.1016/j.omega.2004.12.006 ruiz, r., stutzle, t. (2007). a simple and effective iterated greedy algorithm for the permutation flowshop scheduling problem. european journal of operational research, 177(3), 2033–2049. http://dx.doi.org/10.1016/j.ejor.2005.12.009 ruiz, r., maroto, c. (2005). a comprehensive review and evaluation of permutation flow shop heuristics. european journal of operational research, 165(2), 479–494. http://dx.doi.org/10.1016/j.ejor.2004.04.017 sarraf, a.z., mohaghar, a., bazargani, h. (2013). developing topsis method using statistical normalization for selecting knowledge management strategies. journal of industrial engineering and management, 6(4), 860-875. http://dx.doi.org/10.3926/jiem.573 schuster, c.j., framinan, j.m. (2003). approximate procedures for no wait job shop scheduling. operations research letters, 31(4), 308–318. http://dx.doi.org/10.1016/s0167-6377(03)00005-1 shih, h.s, syur, h.j., lee, e.s. (2007). an extension of topsis for group decision making. mathematical and computer modeling, 45(7-8), 801-813. http://dx.doi.org/10.1016/j.mcm.2006.03.023 stützle, t. (1998). applying iterated local search to the permutation flowshop problem. technical report, aida-98-04, tu darmstadt, fg intellektik; taillard, e. (1990). some efficient heuristic methods for the flow shop sequencing problem. european journal of operational research, 47(1), 65–74. http://dx.doi.org/10.1016/0377-2217(90)90090-x turner, s., booth, d., (1987). comparison of heuristics for flow shop sequencing. omega, the international journal of management science 15(1), 75–78. http://dx.doi.org/10.1016/0305-0483(87)90054-5 vega, a., aguaron, j., garcía-alcaraz, j., moreno-jiménez, j.m. (2014). notes on dependent attributes in topsis. procedia computer science, 31, 308 – 317. http://dx.doi.org/10.1016/j.procs.2014.05.273 wang, l., zheng, d. (2001). an effective hybrid optimization strategy for job-shop scheduling problems. computers & operations research, 28(6), 585–596. http://dx.doi.org/10.1016/s0305-0548(99)00137-9 wang, l., zheng, d.z. (2003). an effective hybrid heuristic for flow shop scheduling. the international journal of advanced manufacturing technology, 21(1), 38–44. http://dx.doi.org/10.1007/s001700300005 werner, f. (1993). on the heuristic solution of the permutation flow shop problem by path algorithms. computers and operations research 20(7), 707–722. http://dx.doi.org/10.1016/0305-0548(93)90058-q widmer m, hertz a. (1989). a new heuristic method for the flow shop sequencing problem. european journal of operational research, 41(2), 186–93. http://dx.doi.org/10.1016/0377-2217(89)90383-4 int. j. prod. manag. eng. (2016) 4(2), 43-52creative commons attribution-noncommercial-noderivatives 4.0 international flow shop scheduling decisions through techniques for order preference by similarity to an ideal solution (topsis) 51 http://dx.doi.org/10.1007/s40092-014-0093-3 http://dx.doi.org/10.1007/s00170-008-1845-2 http://dx.doi.org/10.1007/s00170-012-4388-5 http://dx.doi.org/10.1007/s00170-012-4388-5 http://dx.doi.org/10.1016/0377-2217(95)00037-2 http://dx.doi.org/10.1016/0305-0548(90)90001-n http://dx.doi.org/10.1016/j.mcm.2004.10.003 http://dx.doi.org/10.1016/j.mcm.2004.10.003 http://dx.doi.org/10.1057/jors.1965.8 http://dx.doi.org/10.1016/s0377-2217(02)00908-6 http://dx.doi.org/10.2307/2584415 http://dx.doi.org/10.1016/0305-0548(93)e0014-k http://dx.doi.org/10.1016/0305-0548(93)e0014-k http://dx.doi.org/10.1016/j.omega.2004.12.006 http://dx.doi.org/10.1016/j.ejor.2005.12.009 http://dx.doi.org/10.1016/j.ejor.2004.04.017 http://dx.doi.org/10.3926/jiem.573 http://dx.doi.org/10.1016/s0167-6377(03)00005-1 http://dx.doi.org/10.1016/j.mcm.2006.03.023 http://dx.doi.org/10.1016/0377-2217(90)90090-x http://dx.doi.org/10.1016/0305-0483(87)90054-5 http://dx.doi.org/10.1016/j.procs.2014.05.273 http://dx.doi.org/10.1016/s0305-0548(99)00137-9 http://dx.doi.org/10.1007/s001700300005 http://dx.doi.org/10.1016/0305-0548(93)90058-q http://dx.doi.org/10.1016/0377-2217(89)90383-4 http://creativecommons.org/licenses/by-nc-nd/4.0/ yoon, k., hwang, c.l. (1980). multiple attribute decision making methods and applications. a state of the art survey. berlin: springer verlag. zobolas, g.i., tarantilis, c.d., ioannou, g. (2009). minimizing makespan in permutation flow shop scheduling problems using a hybrid metaheuristic algorithm. computers & operations research 36(4), 1249-1267. http://dx.doi.org/10.1016/j.cor.2008.01.007 int. j. prod. manag. eng. (2016) 4(2), 43-52 creative commons attribution-noncommercial-noderivatives 4.0 international gupta, a. and kumar, s. 52 http://dx.doi.org/10.1016/j.cor.2008.01.007 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2017.5775 received 2016-05-14 accepted: 2017-01-16 adapting transport modes to supply chains classified by the uncertainty supply chain model: a case study at manaus industrial pole fabiana lucena oliveiraa*, aristides da rocha oliveira juniorb and luiza m. bessa rebeloc a amazonas state university, pim observatory, manaus-amazonas, brazil. b management sciences dept., federal university of amazonas, manaus-amazonas, brazil. b public safety post graduateprogram, stateuniversity of amazonas, manaus-amazonas, brazil. a flucenaoliveira@gmail.com b aristides.jr@hotmail.com b lmbrebelo@gmail.com abstract: this paper discusses transport modes supporting uncertainty supply chain model (uscm) in the case of manaus industrial pole (pim), an industrial cluster in the brazilian amazon that hosts six hundred factories with diverse logistics and supply chain managerial strategies. uscm (lee, 2002; fisher, 1997) develops a dot matrix classification of the supply chains considering several attributes (e.g., agility, cost, security, responsiveness) and argues that emergent economies industrial clusters, in the effort to keep attractiveness for technological frontier firms, need to adapt supply chain strategies according to uscm attributes. the paper takes a further step, discussing which transport modes are suitable to each supply chain classified at the uscm in pim´s case. the research´s methods covered the use of pim´s statistical official database (secondary data), interviews with the main logistical services providers of pim and phone survey with a sample of firms (primary data). findings confirm the theoretical argument that different supply chains will demand different transport modes running at the same time in the same industrial cluster (oliveira, 2009). in the case of pim, this implies investments on port and airport infrastructure and a strategic focus on air transport mode, due to (1) short life cycle of products, (2) distance from suppliers, (3) quick response to demand and (4) the fact that even pim´s standard products use, in average, forty per cent of air transport at inbound logistics. key words: uncertainty supply chain model, manaus industrial pole, transport. 1. introduction one of the most important aspects linked to the competitiveness of emergent economies and industrial clusters refer to supply chain management and its correlated logistics strategies (oliveira, 2009). to determine the conditions under which supply chains located in these less developed geographic contexts can operate in higher competitive level standards demands broad and detailed theoretical and empirical exam of the needs, limits and possibilities of each supply chain, as well of the public policies involved in their supporting. among the many supply chains theoretical perspectives (halldorsson et al., 2007) and classifying models available in the literature (e.g., marques et al., 2008), the uncertainty supply chain model (uscm) focuses as its core theoretical contribution the developing of a dot matrix classification of several supply chains, considering uncertainty as an essential parameter (oliveira, 2009; lee, 2002; fisher, 1997). this uncertainty matrix is useful to categorize the supplying (raw materials/components) and demand (consumer market processes), considering the singularities of each manufactured product, as also int. j. prod. manag. eng. (2017) 5(1), 39-43creative commons attribution-noncommercial-noderivatives 4.0 international 39 https://doi.org/10.4995/ijpme.2017.5775 mailto:aristides.jr@hotmail.com mailto:lmbrebelo@gmail.com http://creativecommons.org/licenses/by-nc-nd/4.0/ to indicate the suitable logistics strategies for the diverse domestic and global supply chains. transport modes is one of the strategic aspects to be considered in enhancing supply chain performances, but it is, at the same time, a variable strongly dependent to the characteristics of the final goods produced by the diverse existent supply chains and their respective markets. this paper explores this theoretical linkage between the uscm and transport strategies in the context of a concrete case: the manaus industrial pole (pim). operating as an industrial cluster in the very heart of the brazilian amazon (in the city of manaus) since 1967, pim hosts about six hundred companies manufacturing durable goods classified mainly in the sectors of consumer electronics, motorcycles, information technology hardware, chemicals, watches, among others. 2. the uncertainty supply chain model and transport strategy this uscm has as core framework an uncertainty matrix used to categorize the supplying (raw materials and components) and demand (consumer market) processes, considering intrinsic characteristics of each manufactured product. in general terms, the uscm classification shows that there are some goods characterized by demand and supply stability, longer cycle of life and low added technological value, that will demand a more simplified logistics strategy, and that there are other goods, characterized by demand and supply instability, very short life cycle and high added technological value, will require special logistics strategies management. figure 1 reproduces the uncertainty matrix (lee, 2002). uscm classifies products in two main categories: functional characterized by low technological added value and stable demand (consumer market) and supply (raw material/components) processes and innovativecharacterized by cutting edge technology, unstable demand (consumer market) and supply (raw material/components)processes. for each one of these product categories, a different scm strategy was theorized. this being so, the products considered to present low uncertainty of supply and low uncertainty of demand are those that aggregate low technological value in their production, in other words, the life cycles of these products are usually longer and their manufacturing depends in a low degree on technological evolution. whereas those with low uncertainty of supply and high uncertainty of demand are the audio and video, telecommunications and computer products that follow the tendencies of a market characterized by the consumption of novelties that aggregate new technologies, in the expectation of keeping up with technological evolution. these products already usually present a short life cycle and require agility in the management of their supply chains, since the tendencies in technological evolution can be very fast. those products with high degrees of uncertainty in supply and low degrees of uncertainty in demand (e.g., hydroelectric power generating equipment, cables and connections and mining equipment) and some food segments that transform specific raw materials. the sources for the supply of raw materials to manufacture these products are limited and this leads to uncertainty of supply, since demand is stable and the need for production remains constant from a source with scarce supply. goods with a high degree of uncertainty in demand and a high degree of uncertainty in supply are represented by telecommunications products, highend computers and semi-conductors. these products have sources of even scarcer supply and that are sometimes monopolized by a handful of companies. from the point of view of demand, telecom products (e.g., mobile telephony) have short life cycle, high competitiveness and a high degree of uncertainty regarding the consumer desire to buy. agility in the management of this supply chain is vital to the survival of the product´s manufacturing. industrial clusters that wishes to include companies classified in the lower quadrants of the uncertainty model, needs to consider agility as one of its pillars of development (oliveira, 2009). low (functional products) high (innovative products) low (stable process) high (devel opment process) hydroelectric apparatus, some food segments telecom, high-end computers, semi conductors uncertainty of demand u nc er ta in ty o f su pp ly candies, basics, common apparel, foodstuffs, oil and gas fashion apparel, computers, audio, vi deo figure 1. the uncertainty matrix (source: aligning supply  chain strategies with product uncertainties: lee, 2002). int. j. prod. manag. eng. (2017) 5(1), 39-43 creative commons attribution-noncommercial-noderivatives 4.0 international oliveira, f.l., da rocha oliveira, a. and bessa rebeloc, l.m. 40 http://creativecommons.org/licenses/by-nc-nd/4.0/ the strategies for the uncertainty models are classified according to four types: (1) efficient supply chains, (2) supply chains with risk coverage, (3) sensitive supply chains and (4) agile supply chains. figure 2 presents a summary of these supply chain classifications: low high (functional products) (innovative products) low (stable process) high (development process) uncertainty of demand u nc er ta in ty o f s up pl y efficient supply chains sensitive supply chains supply chains with risk coverage agile supply chains figure 2. supply chain strategies (source:  aligning  supply chain strategies with product uncertainties: lee,  2002). according to grieger (2003), the most critical variables to analyze in the uscm are: a) fast product life cycle; b) just in time production; c) cost leadership; and d) global competition. based on the behavior of these variables, it is possible to research and predict which logistics strategies will be more suitable to supply chains located on one of the four quadrants of the uncertainty matrix, as, for example, the choice of the transport modes involved either to import raw materials/components or to export final goods (oliveira, 2009). this theoretical aspect adjusts the rationality of transport decisions in the way they are currently exposed in the scm and business logistics literature (e.g., bowersox et al., 2002), which does not explicitly and formally integrates these uncertainty variable as a sc classification parameter to be considered in firms´ logistic decision-making and strategy. 3. methodology the methodology used in the empirical investigation conducted in the case of the manaus industrial pole (pim) covered three main strategies: (1) the use of pim´s official statistical database (suframa, 2015a) and pim´s companies´ official profile (suframa, 2015), as also the current customs legislation of brazil, as documental sources of secondary data (number and name of the enterprises by sector, main manufactured goods, sector revenues, imports and exports, customs processes etc.); (2) interviews conducted with the main logistical services providers (lsp´s) of pim as a source of primary data directed to the classification of pim´s supply chains in the uscm quadrants and the actual usage of transport modes by each sc; and (3) phone survey with a sample of firms (also primary data) to confirm the data collected in the previous steps. the first effort was to outline the universe of pim´s companies in such a way as to identify how this model may fit the reality of an industrial pole (geographical delimitation), and of the current customs legislation in this country, or in their respective particularities. from the pim´s sector revenues, we identified the most twenty important products of pim. based on this products list and crossing with supply chain strategies respectively, all supply chains of these products have been identified, so was their respective companies. then, the lsp´s interviews and subsequently phone survey with the companies involved to confirm their most used transport modes. this survey was done by telephone and using some personal contacts with clearance people. all the information about transport modes were filled up without previous checking just to guarantee we were capturing the real transport mode in use, even knowing some products were standard. the transport modes appointed by the theoretical model were confirmed, with increased use of air transport even in the standard product supply chains. with regard to the purposes, this survey was explanatory and applied, because it aimed not only to clear up the factors involved, but also to contribute to the making of decisions and propose concrete solutions to concrete and immediate problems. the universe for study refers to the group directly involved in the formulation of the problem, the companies in the manaus industrial pole (pim). this analysis, adapted and chose the uncertainty model in its most extreme aspect uncertainty of supply and uncertainty of demand, using an industrial unit that has its supply chain perfectly adapted to this reality as its research universe. the results obtained here, therefore, are restricted to the industrial units with extreme uncertainty regarding their supply chains, following the guidance of brazilian customs legislation, and improving the processes already identified as being promising by int. j. prod. manag. eng. (2017) 5(1), 39-43creative commons attribution-noncommercial-noderivatives 4.0 international adapting transport modes to supply chains classified by the uncertainty supply chain model: a case study at manaus industrial pole 41 http://creativecommons.org/licenses/by-nc-nd/4.0/ the case study for the brazilian customs authorities: the manaus industrial pole (pim). 4. results (case study: manaus industrial polepim) the most important products of pim represents eighty percent of the total billed by this agglomerations model within one year. a sample of these products took us twenty-seven companies whose products make up the list of the most lucrative of the pim. this research set out to identify the classification of these companies between global and multinational, and from the uncertainty model and its respective supply chains derived from them. then, to identify the modes of transportation actually used for incoming inputs, independent of type of supply chains. the first result was on the pim´s composition: seventy percent of the companies operating on it are subsidiaries of global companies. this means that logistics strategies are defined in their respective foreign dies, leaving minimal autonomy for decision and adaptation of local logistics strategies. it helps to explains why, passed almost fifty years of pim´s existence, until now there isn´t a single sc strategy formally formulated and implemented by policy-makers (e.g., suframa, amazonas state government etc.). when it checks the types of supply chains, the result was 11 agile supply chains, 05 sensitive supply chains, and 11 efficient supply chains. this means that a total of twenty seven companies surveyed, sixteen need the air mode to remain competitive because their supply chains have a level of uncertainty still present. this represents fifty nine per cent of dependent companies of air transportation and therefore potential users of the infrastructure improvements at airports for pim. this high usage rate of air mode confirms the hypothesis that to remain competitive logistics of pim, it is necessary to accept that being away from the supply base and also to final customers makes the airline the able way enable the uncertainty of supply and demand. thus, if the expectation is to attract companies whose products are technological innovation, brazil needs to be special attention to the logistical support of the air mode. table 1 presents the types of companies, their supply chains and transport modes considered for inbound. table 1. supply chains identified at pim (source: authors, 2015). company type products supply chain transport mode for inbound transport mode for final product global mobile phones agile air air global razor & toothbrush efficient sea road global pens, lighters efficient sea road & air global computers & tvs agile sea air air global computers & tvs agile sea air air global computers & tvs agile sea air air global medical equipment efficient sea air air multinational microwave oven efficient sea road multinational cameras agile sea air na multinational board assembly agile sea air na global tv agile sea road & sea global atms efficient sea road & air global mobile phones agile air air global acessories for cameras sensitive air na global batteries agile sea air na global tvs & microwave efficient/sensitive sea air road & sea global electric shaver efficient sea road & sea global battery charger for mobile phone agile sea air na global tvs & mobile phone sensitive/agile sea air road & air multinational electronic components efficient sea na global tv & audio sensitive sea air road & air multinational cd e dvd sensitive sea road multinational toner sensitive air air global air conditioner efficient sea road & sea multinational air conditioner efficient sea road & sea global air conditioner & microwave efficient sea road & sea multinational motorcycle efficient sea road & sea int. j. prod. manag. eng. (2017) 5(1), 39-43 creative commons attribution-noncommercial-noderivatives 4.0 international oliveira, f.l., da rocha oliveira, a. and bessa rebeloc, l.m. 42 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5. conclusion brazil is an emergent country which works with regional, multinational and global companies. manaus industrial pole is very important to keep around a hundred thousand employments in a city with less than two million people and which is responsible for economic activity for the north region in brazil. do not find regional or local companies listed on the most important products from pim seems to be a worrying matter to pim´s policy-makers. it means that domestic capital is not present in the manufacturing of high-tech products. in the other hands, brazil has to be able to attract and keep these international companies on different agglomeration models. if brazilian government as a representative of emergent economy, wants to keep industries´ competitiveness based in brazil, a high investment on airports and air transport infrastructure has to be done. the risk, if it is not done in a short time, is to keep in brazil only companies without innovative products and delay the consumption of innovative products by brazilian society, since it will be imported, not manufactured in the country. references bowersox, d. j., closs, d. j., cooper, m. b. (2002). supply chain logistics management (vols. 1 and 2). new york, ny: mcgraw-hill. christopher, m. (2000). the agile supply chain: competing in volatile markets. industrial marketing management, 29(1), 37-44. https://doi. org/10.1016/s0019-8501(99)00110-8 fisher, m. l. (1997). what is the right supply chain for your product? harvard business review. march-april, 1997, 105-116. grieger, m. (2003). electronic marketplaces: a literature review and a call for supply chain management research. european journal of operational research, 144(2), 280-294. https://doi.org/10.1016/s0377-2217(02)00394-6 halldorsson, a., kotzab, h., mikkola, j. h., skjøtt-larsen, t. (2007). complementary theories to supply chainmanagement. supply chain management: an international journal, 12(4), 284-296. http://dx.doi.org/10.1108/13598540710759808 lee, h. (2002). aligning supply chain strategies with product uncertainties. california management review, 44(3),167-179. https://doi. org/10.2307/41166135 marques, g. c., pereira, s. c. f., carona, n. (2008). proposta de um modelo dinâmico para classificação de cadeias de suprimentos. anais do xi simpoi, rio de janeiro. oliveira, f. l. (2009). gestão estratégica das cadeias de suprimento com base no modelo logístico de incerteza: o caso do polo industrial de manaus (pim). universidade federal do rio de janeiro (doctoralthesis). bathnagar, r., sohal, a. s. (2005). supply chain competitiveness: measuring the impact of location factors, uncertainty and manufacturing practices. technovation, 25(5), 443-456. http://dx.doi.org/10.1016/j.technovation.2003.09.012 suframa – superintendência da zona franca de manaus. (2015) indicadores de desempenho do polo industrial de manaus. suframa, manaus. available at: http://www.suframa.gov.br. last access: january 2016. suframa – superintendência da zona franca de manaus. (2015) perfil das empresas incentivadas do polo industrial de manaus. suframa, manaus. available at: http://www.suframa.gov.br. last access: january 2016. int. j. prod. manag. eng. (2017) 5(1), 39-43creative commons attribution-noncommercial-noderivatives 4.0 international adapting transport modes to supply chains classified by the uncertainty supply chain model: a case study at manaus industrial pole 43 https://doi.org/10.1016/s0019-8501(99)00110-8 https://doi.org/10.1016/s0019-8501(99)00110-8 https://doi.org/10.1016/s0377-2217(02)00394-6 http://dx.doi.org/10.1108/13598540710759808 https://doi.org/10.2307/41166135 https://doi.org/10.2307/41166135 http://dx.doi.org/10.1016/j.technovation.2003.09.012 http://www.suframa.gov.br http://www.suframa.gov.br http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2016.5964 industrial engineering: creating a network! j. carlos prado-prado university of vigo. school of industrial enginners, lagoas marcosende, 36310 vigo (pontevedra). spain jcprado@uvigo.es abstract: this paper presents a brief history of the industrial engineering conference (cio), and specially reinforces the role of the cios as a forum for building a network and creating log-term relationships. key words: industrial engineering, industrial engineering conference, network. after 30 years, no less, a group of industrial engineering teachers from different spanish universities met at la rábida (huelva, spain) for the 1th industrial organisation conference, an event organised by the department of business organisation at the seville school of industrial engineers. they were all concerned about reaching a consensus on the presence and activity of teachers in the business organisation areas in engineering schools and on defining the industrial engineering profile. later in 1995 the 2nd industrial organisation conference was held in valencia and the 3th industrial organisation conference was held in barcelona, which saw the start of adingor (the association for industrial engineering development). adingor’s first and fundamental objective was to reinforce and consolidate its own industrial engineering approach based on business organisation and engineering. moreover, attempts were made to set a meeting place, a forum, where people could exchange opinions, and share knowledge, best practices, as well as academic and research experiences, without forgetting the chance for personal contact among group members, building a network and creating long-term relationships. the meeting place has been the annual adingor conferences, known as cios. since 1999, cios have been held non-stop until this year’s 20th edition in the city of san sebastián. many things have changed since then. the environment is increasingly turbulent, and the market is ever-demanding and more globalised. competitive pressure is enormous, new technologies constantly appear, as do environmental and legal pressures, and so on. the education field is not much more promising. government cuts affect all aspects, there are no research grants, the pressure to publish is brutal, publications are the only scale to be acknowledged by, uncertainties about future syllabi, etc. all this makes our activity and day-to-day duties extraordinarily complicated. however, if we examine cios, i believe that we can see that their evolution and value have been positive for both teaching and research. in the teaching domain, i wish to stress what cios have contributed. teaching problems have always been addressed in all the cios. by way of example, in this year’s cio in san sebastián, two tracks appear: industrial engineering & operations management education and innovation in education and professional skills. i would like to point out the particular contribution made by the 2003 cio held in the city of valladolid, which coincided with university degrees being adapted to the european higher education area. int. j. prod. manag. eng. (2016) 4(2), 41-42creative commons attribution-noncommercial-noderivatives 4.0 international 41 http://dx.doi.org/10.4995/ijpme.2016.5964 http://creativecommons.org/licenses/by-nc-nd/4.0/ during the 2003 cio, the adingor general assembly passed the proposal of the industrial engineering degree title. this document was the basis on which all the titles of the different spanish schools and universities have been determined. as industrial engineers, an applied nature features in our dna. promoting relations with industry has always been present. indeed the motto of this year’s 20th edition of cio is “building bridges between researchers and practitioners”. moreover, directors of important national and international companies have always attended the plenary sessions of cios. it is a fact that cios and adingor have always known that in a global world, we could not be restricted to a national conference. after consolidating the presence of teachers and researchers at cios from all spanish universities, our internationalisation evidently became a pending matter. in 2007, the 11th cio madrid conference was also the 1th international conference on industrial engineering and industrial management, for which english and spanish were the official languages. contacts were also made with the associação brasileira de engenharia de produção, abepro, the equivalent to our industrial engineering association. this contact was subsequently reinforced and it finally shaped the joint organisation of cios with icieom. and so it was that valladolid witnessed the first joint conference in 2013. such international emphasis continued to grow as the institute of industrial engineers of the usa (now the institute of industrial & systems engineers) attended the 2014 cio in the city of málaga, the first conference to be jointly organised by three associations with english as the official language. in 2015 the conference was organised in aveiro (portugal), to which portuguese industrial engineers attended, and the european academy of industrial management will attend this year’s cio in san sebastián. this has meant that more than 20 countries have attended recent cios. in particular, cios are a meeting place and a networking opportunity for all industrial engineers from spain and other countries. i believe that this aspect is fundamental. cios are a unique opportunity to meet with colleagues from other national and international universities. it is not just a matter of attending the conference, presenting a paper and receiving feedback from other researchers. nor is it only for knowledge sharing and setting trends by means of talks given by researchers and plenary sessions. for example, in this year’s cio we will talk about new trends in manufacturing with a track in industry 4.0. it is not only to offer the chance to welcome editors from prestigious international journals, and to hear their points of view and recommendations to be able to publish in their journals and other journals. nor just to select the best papers for prestigious international journals or to be published in the springer lecture notes in management and industrial engineering. it is true that cios very much reinforce our role as trainers of a given profession, which has in recent years been extended and its image has improved, and it has been acknowledged in economic-professional and academic domains. cios are also a way of transmitting the existence of industrial engineering and its international acknowledgement to academic and professional domains. the hundreds of thousands of downloads of the conference proceedings from our website clearly evidence this. apart from all these aspects, any of which alone would justify the existence of cios, the role of “meeting point” and networking is essential. the fact that publications are seen as the only acknowledgement scale and us focusing most of our efforts there may lead us to isolation, to us being enclosed in our bubble and our shell. so above all, we need to relate with others because there is no doubt that “making friends” is vital in all walks of life, and also in what is academic. sharing with other colleagues, knowing what is happening in other universities and in the world, knowing how to do things together which cannot be done individually are what cios offer us. moreover, our cios dedicate a large part of the conference to social aspects to reinforce this interpersonal knowledge and these links. we face many challenges and cios are one of the best responses to face them. int. j. prod. manag. eng. (2016) 4(2), 41-42 creative commons attribution-noncommercial-noderivatives 4.0 international prado-prado, j. c. 42 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.3199 received 2014-07-18 accepted 2014-12-18 coupling order release methods with autonomous control methods – an assessment of potentials by literature review and discrete event simulation grundstein, s.a,i*, schukraft, s.a,ii, scholz-reiter, b.b and freitag, m.b a biba – bremer institut für produktion und logistik gmbh at the university of bremen, hochschulring 20, 28359 bremen, germany *i gru@biba.uni-bremen.de ii skf@biba.uni-bremen.de b university of bremen, bibliothekstraße 1, 28359 bremen, germany abstract: production planning and control faces increasing uncertainty, dynamics and complexity. autonomous control methods proved themselves as a promising approach for coping with these challenges. however, there is a lack of knowledge regarding the interaction between autonomous control and precedent functions of production planning and control. in particular, up to now previous research has paid no attention to the influence of order release methods on the efficiency of autonomous control methods. thereby, many researchers over the last decades provided evidence that the order release function has great influence on the logistic objective achievement in conventional production systems. therefore, this paper examines the influence of order release methods on the efficiency of autonomous control methods by both theoretic evaluation and discrete event simulation. the simulation results indicate an overall high influence. moreover, the logistic performance differs considerably depending on the implemented order release methods and the combinations of order release methods with autonomous control methods. the findings highlight demand for further research in this field. key words: autonomous control, order release, production control, reactive scheduling. 1. introduction production planning and control (ppc) has to cope with increasing complexity, dynamics and uncertainty (kim & duffie, 2004; westphal, 2001). complexity is caused by the number and variety of elements and relations in a production system (westphal, 2001). dynamics is induced by the change of characteristics of elements and relations in time (wyssusek, 1999). uncertainty is defined as difference between required and available information, both about the current and the future system state and future events (leisten, 1996). for the planning and control of production systems with a high degree of uncertainty, reactive scheduling approaches are generally proposed (byeon, wu & storer, 1998; gan & wirth, 2005; lawrence & sewell, 1997; sabuncuoglu & karabuk, 1999). in this context, autonomous control methods proved themselves as a promising approach for coping with increasing dynamics and complexity in logistic processes (scholz-reiter & freitag, 2007b). thereby, they focus on the achievement of logistic objectives. however, also the preceding order release function has significant influence on the logistic objectives achievement in conventional, i.e. non-autonomous production systems (qi, sivakumar & gershwin, 2009; wein & chevalier, 1992). generally, order release methods impose the boundaries, within control methods operate. up to now research has paid no attention to the influence of order release methods other than immediate order release on the efficiency of autonomous control methods. 43int. j. prod. manag. eng. (2015) 3(1), 43-56attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2015.3199 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 1. review of previous research. order release method: release/ creation criteria  research on autonomous control  n on e/ im m ed ia te re le as e w ith d et er m in is tic p la n da te s pe ri od ic in flu x w ith im m ed ia te re le as e (e .g . s in us ) w ip re gu la tin g ap pr oa ch w ith ou t l oa d ba la nc e a pp ro ac h w ith w or ks ta tio n sp ec ifi c lo ad b al an ce barenji, barenji & hashemipour (2014) x sudo & matsuda (2013) x owliya, saadat, anane & goharian (2012) x park & tran (2012) x pach, bekrar, zbib, sallez & trentesaux (2012) x rekersbrink (2012) x wang, tang, gu, zheng, yuan & tang (2012) x windt, becker, jeken & gelessus (2010) x scholz-reiter, görges, jagalski & naujok (2010) x pannequin, morel & thomas (2009) x wang & lin (2009) x duffie & shi (2009); duffie, roy & shi (2008) x leitão & restivo (2008) x scholz-reiter, jagalski & bendul (2008); de beer (2008); scholz-reiter, freitag, de beer & jagalski (2006) x xiang & lee (2008) x tsutsui & liu (2007) x reaidy, massotte & diep (2006) x wong, leung, mak & fung (2006) x armbruster, de beer, freitag, jagalski & ringhofer (2006) x x kornienko, kornienko & priese (2004) x siwamogsatham & saygin (2004) x cicirello & smith (2001) x x table 1 gives an overview of previous research on autonomous control in a broader sense concerning the applied order release methods according to the classification depicted later in section 2.3. the overview reveals that classes of wip regulating approaches and load balancing approaches have not yet been examined in combination with autonomous control approaches. nevertheless, wip regulating approaches and load balancing approaches are in focus of scientific consideration, as many recent publications to these topics reveal. ostgathe (2012) for instance applies the workload control approach in combination with the least slack sequencing rule in his approach examining autonomous disruption management. it is generally suggested to apply simple sequencing rules in combination with these release approaches (land, stevenson & thürer 2012). due to the proven high impact of order release methods in non-autonomous systems, this paper investigates the effect of the application of order release methods in autonomously controlled production systems. 2. fundamentals 2.1. scheduling in the context of production planning and control production planning and control provides the basis for organising and executing the production process (nyhuis & wiendahl, 2008). the basic planning tasks are the production program planning, the production requirements planning, the in-house production planning with scheduling as a subtask, and the planning and control of external production (cf. figure 1). the first task creates a production program, containing the primary demand for each product and planning period based on market and sales forecasts. this step is also known as production planning or master scheduling (pinedo, 2008). the subsequent requirement planning derives the secondary demand from the production program, i.e. the required material and resources. therefore, production orders are generated, roughly scheduled and the required production capacities are computed and adjusted. based on the results of these planning steps, the make-or-buy decision determines which parts of the production are procured externally and which are produced in-house. external production for example might be relevant if the demand exceeds available in-house production capacities. in case of in-house production, production planning determines batch sizes, detailed scheduling of production orders and the availability check of required resources. diligent data management is the basis for all these tasks. (schuh, 2006; wiendahl, 2005). 44 int. j. prod. manag. eng. (2015) 3(1), 43-56 attribution-noncommercial-noderivatives 4.0 international grundstein, s., schukraft, s., scholz-reiter, b. and freitag, m. http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 1. key tasks of ppc with focus of this paper according to luczak & eversheim (1999). this paper focuses on detailed scheduling in the context of in-house-production planning and control, which is often called just scheduling (pinedo, 2008). this step typically determines a schedule, which comprises the exact dates for the start or end of operations and thus also the sequence of production orders. the detailed allocation of resources is also part of this planning step (pinedo, 2008). scheduling strategies can be differentiated into predictive, reactive, predictive-reactive as well as proactive ones (van brackel, 2009). predictive scheduling strategies create schedules before the beginning of the production process (van brackel, 2009). they once create a schedule under the assumption of a deterministic production system. reactive scheduling strategies do not create schedules as defined in common terms (sabuncuoglu & karabuk, 1999). they control the process locally, often using priority rules (o’donovan, uzsoy & mckay, 1999). predictive-reactive scheduling strategies create an initial schedule and then adopt it in iterations to occurring disturbances (o’donovan, uzsoy & mckay, 1999). proactive scheduling strategies try to avoid rescheduling by creating robust schedules, which e.g. anticipate disturbances by inserting idle times, for example szelke & monostori (1999). autonomous control methods in this context possess characteristics of reactive scheduling. 2.2. autonomous control autonomous control methods enable coping with dynamics and complexity in the production process. they base on decentralized decision-making authority (freitag, herzog & scholz-reiter, 2004). prevailing, autonomous control approaches focus on the usage of existing flexibility potentials in the production system for generating decision alternatives (schuh, gottschalk & höhne, 2007). for that matter, parts or production orders decide autonomously on available alternative routes through the production system (scholz-reiter & freitag, 2007b). logistic objects interact with each other, exchange information, decide for themselves on this basis and execute their decisions (scholz-reiter & freitag, 2007b). simulation studies indicate that autonomous control methods are able to increase the logistic performance (scholz-reiter & freitag, 2007b). several autonomous control methods have already been developed within the collaborative research centre 637 at the university of bremen (gierth, 2009; windt, becker, jeken & gelessus, 2010). a detailed classification of these methods is given by windt et al. (2010). the following table depicts the state of the art and classifies the methods based on the descriptions of windt et al. (2010), scholzreiter et al. (2010) and schmidt et al. (2007). in general, these methods can be divided into rational methods, bounded rational (bio-analogue) methods and respective combinations of them (scholzreiter, böse, jagalski & windt, 2007a), whereas the latter have not yet been analysed quantitatively and therefore are not depicted explicitly in table 2. while rational methods decide based on the anticipation of future system states, bio-analogue methods transfer the behaviour of natural systems like ants or honey bees to production control (scholz-reiter, böse, jagalski & windt, 2007a). table 2. autonomous control methods. method type method rational methods dlrp production due date method gentelligent parts link-state internet routing protocol one logistics target per rule simple rule based 1 / queue length estimator simple rule based 2 bounded rational (bio-analogue) methods ant algorithm/ cunning ant algorithm bacterial chemotaxis bee foraging bionic manufacturing system holonic manufacturing market based control 45int. j. prod. manag. eng. (2015) 3(1), 43-56attribution-noncommercial-noderivatives 4.0 international coupling order release methods with autonomous control methods – an assessment of potentials by literature review and discrete event simulation http://creativecommons.org/licenses/by-nc-nd/4.0/ 2.3. order release and generation according to wiendahl (1997), the primary task of production control is the realisation of the production plan despite of potential disturbances. kurbel (2005) thereby defines the order release as connector between production planning and control. the task of order release determines the point of time, from which on the production is authorized to start the processing of an order. it usually triggers the material supply and the assignment of material to a specific production order (lödding, 2013). order release thus has a significant impact on the logistic performance of a production system (qi, sivakumar & gershwin, 2009; wein & chevalier, 1992). there are several frameworks for the description and classification of order release methods in literature (bergamaschi, cigolini, perona & portioli, 1997; lödding, 2013; sabuncuoglu & karapinar, 1999). according to lödding (2013), release methods are divided into the classes of immediate order release, due date based order release, wip regulating order release and approaches with workstation specific load balance as depicted in figure 2. classes of order release according to lödding (2013).. the order release function processes only orders, which are provided by the precedent order generation function. the order generation function creates production orders out of customer orders, material withdrawals or a production program. it determines the planned input into the production, the planned sequence and the planned output. in general, order generation methods can be classified as depicted in figure 3. classification criteria for order generation methods according to lödding (2013). by the generation scope, trigger logic, the method character and the primary generation criterion, which is mainly individual for every method. (lödding, 2013) 2.4. summarising basic relationships the traditional function of production control is to realise the production plan also under – potentially unavoidable – disruptions (wiendahl, von cieminski & wiendahl, 2005). a basic model of production control according to lödding (2013) is shown in figure 4. the model consists of four basic elements: the production control functions, the manipulated variables, the observed variables and the logistic objectives. the connections between the elements indicate causal relationships. the functions determine the manipulated variables in the way of the directions. the observed variables result from the deviation of two manipulated variables. the observed variables determine the degree of logistic objective achievement. hence, the four functions of production control – order generation, order release, sequencing and capacity control – directly influence the degree of logistic objective achievement. using autonomous control methods, they fulfil the sequencing task and thus, influence the logistic objectives. throughput time is the length of time between the order›s release and the end of its processing. wip level is a measure for released orders which have not yet been finished. wip can be counted either in number of orders or in figure 4. production control model according to lödding (2013). figure 3. classification criteria for order generation methods according to lödding (2013). figure 2. classes of order release according to lödding (2013). 46 int. j. prod. manag. eng. (2015) 3(1), 43-56 attribution-noncommercial-noderivatives 4.0 international grundstein, s., schukraft, s., scholz-reiter, b. and freitag, m. http://creativecommons.org/licenses/by-nc-nd/4.0/ time units. utilisation describes the ratio of the mean and maximum possible output rate of a workstation. finally, schedule reliability refers to the percentage of orders delivered within a defined delivery reliability tolerance. lödding (2013) 3. coupling order release and autonomous control methods 3.1. theoretical evaluation a theoretical evaluation is carried out in table 3. evaluation matrix of autonomous control methods and order release methods.. the evaluation considers selections of autonomous control methods and order release methods and examines the possibility of a combined application based on each method´s description in the corresponding reference. the methods are clustered within method types according to the classifications given in section 2. the character (●) represents a possible combination. however, the possibility of a combination does not coercively imply reasonability. the reasonability of a combined usage depends on each production system and its requirements. for example, the due date oriented order release can be combined with all autonomous control methods. if a company prioritises high due date reliability then a due-date oriented autonomous control method such as the dd method is more reasonable than a method such as srb1/qle which aims at minimising makespan and disregards due dates. the character (◑) indicates a possible combination under the prerequisite of an adaption of either the control or the release method. for instance, the workload control (wlc) method considers orders to be processed on a work centre for the workload calculation. however, autonomous control methods decide their route during run-time, so that adequate forecasting methods must be applied to combine these methods efficiently. therefore, the evaluation in this case is at (◑). an evaluation of (○) indicates an impossible combination. for instance, the “g_polca”-method is based on product authorisation cards to control the number of jobs in production. the “dlrp”-method is based on collaborating intelligent products. these two methods are obviously impossible to combine. the asterisks (*) mark couples which are considered in the context of the simulation in section 3.2. furthermore, compared to table 1, the evaluation only considers autonomous control methods, which are described detailed enough to be reproducible. it is also apparent, that up to now, wip regulating approaches and approaches with load balance cannot be combined with autonomous control methods without adaptions. therefore, adaptions, respectively adaption methods must be developed if a combined application of these methods is desired. table 3. evaluation matrix of autonomous control methods and order release methods. also indicates, that it depends primarily on the characteristics and requirements of each order release method, whether the combination is possible or not. this dependency reveals itself reading the table line by line: if an order release method can be combined with one autonomous control method, it can also be combined with all others (except the slrd method). the most frequent reason for incompatible combinations is that certain order release methods rely on the anticipation of future system states, for example as the wlc explained above. thereby, in most cases they presume a deterministic production plan. autonomous control methods, in contrast, decide during run-time. therefore, as long as no dynamic and unpredicted events occur (e.g. rush orders, breakdowns), all autonomous control methods can be combined with all order release methods. but as this assumption is far from reality, the evaluation is often at (◑) for these cases. besides, applying autonomous control methods is not reasonable under such static and deterministic conditions (cf. 2.1). 3.2. simulation study in order to corroborate the interdependencies of a combined application of order release and autonomous control methods, a simulation study was carried out as described in the following. the study picks up all combinations marked with a (*) in table 3. evaluation matrix of autonomous control methods and order release methods. to exemplarily show the degree of interdependency and the potential benefits of combined application. these combinations consider at least one method of each category, if the combination is possible without adaptions. the simulation study is based on a 3×3-machine model depicted in figure 5, which is often used for the evaluation of autonomous control methods (de beer, 2008; scholz-reiter, jagalski & bendul, 2008). for matters of comparability we orient the simulation study to a great extent on de beer (2008), who examines the dynamics of production systems applying autonomous control methods. 47int. j. prod. manag. eng. (2015) 3(1), 43-56attribution-noncommercial-noderivatives 4.0 international coupling order release methods with autonomous control methods – an assessment of potentials by literature review and discrete event simulation http://creativecommons.org/licenses/by-nc-nd/4.0/ table 3. evaluation matrix of autonomous control methods and order release methods. method type rational methods bounded rational methods / bio-analogue methods reference sc ho lz -r ei te r, ja ga ls ki & b en du l ( 20 08 ) sc ho lz -r ei te r, fr ei ta g, d e b ee r & j ag al sk i ( 20 06 ) w in dt , b ec ke r, je ke n & g el es su s (2 01 0) r ek er sb ri nk , ( 20 12 ) c ic ir el lo & s m ith (2 00 1) ; w in dt e t a l. (2 01 0) ts ut su i & l iu (2 00 7) ; w in dt e t a l. (2 01 0) a rm br us te r e t a l. (2 00 6) ; w in dt e t a l. (2 01 0) sc ho lz -r ei te r, ja ga ls ki & b en du l ( 20 08 ) sc ho lz -r ei te r, g ör ge s, ja ga ls ki & n au jo k (2 01 0) method si m pl e ru le b as ed 1 / q ue ue l en gt h e st im at or si m pl e ru le b as ed 2 d ue d at e m et ho d d l r p pr od uc tio n a nt a lg or ith m c un ni ng a nt a lg or ith m ph er om on e ap pr oa ch b ee fo ra gi ng b ac te ri al c he m ot ax is code srb1 / qle srb2 dd dlrp ant c-ant phe bee che immediate order release lödding (2013) immediate order release imr ●* ● ●* ● ● ● ●* ● ● due date based order release thürer, stevenson, silva, land & fredendall (2012) periodic period ● ● ● ● ● ● ● ● ● lödding (2013) due date oriented order release date ●* ● ●* ● ● ● ●* ● ● wip regulating approach without loadorientation thürer, stevenson, silva, land & fredendall (2012) constant work in process conwip ●* ● ●* ● ● ● ●* ● ● qi, sivakumar & gershwin (2009) wipload control wipload ● ● ● ● ● ● ● ● ● fernandes & carmo-silva (2011) similar set-up and latest release date slrd ● ● ● ○ ● ● ● ● ● fernandes & carmo-silva (2006) generic paired-cell overlapping loops of cards with authorization g_polca ◑ ◑ ◑ ○ ◑ ◑ ◑ ◑ ○ lödding (2001) decentral inventory oriented manufacturing control dec_boa ◑ ◑ ◑ ○ ◑ ◑ ◑ ◑ ○ lödding (2013); wein (1988) bottleneck control bot ● ● ● ● ● ● ● ● ● workstation specific load balance irastorza & deane (1974) order release with linear programming lp ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ jendralski (1978) workload control wlc ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ bechte (1980); wiendahl (1991) load-oriented order release boa ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ baykasoğlu & göçken (2011); melnyk & ragatz (1989) aggregate workload trigger, work in the next queue selection aggwnq ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ work centre workload trigger wcwt ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ thürer, stevenson, silva, land & fredendall (2012) lancaster university management school order release lmus cor ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ superfluous load avoidance release slar ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ thürer, filho & stevenson (2013) idle machine rule idlemr ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ space control order release spaceor ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ standard rule standr ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ zozom, hodgson, king, weintraub & cormier (2003) job planner jp ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ gentile & rogers (2009) workload control machine center wlcmc ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ weng, wu, qi & zheng (2008) multi-agent-based workload control for make-to-order manufacturing ma_wlc ◑ ◑ ◑ ◑ ◑ ◑ ◑ ◑ ○ 48 int. j. prod. manag. eng. (2015) 3(1), 43-56 attribution-noncommercial-noderivatives 4.0 international grundstein, s., schukraft, s., scholz-reiter, b. and freitag, m. http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 5. basic model structure (de beer, 2008, p.75). three product variants are produced (de beer, 2008, p.84). transportation times and set-up-times are included in the processing times, cf. de beer (2008, p.84). the processing times depend on station and product according to table 4 as in de beer (2008, p.84). table 4. machine and product dependent processing times (de beer, 2008, p.84). product a product b product c 1st stage s11 2h 2.5h 3h s12 3h 2h 2.5h s13 2.5h 3h 2h 2nd stage s21 2h 2.5h 3h s22 3h 2h 2.5h s23 2.5h 3h 2h 3rd stage s31 2h 2.5h 3h s32 3h 2h 2.5h s33 2.5h 3h 2h each product has to be processed on one station at each stage (de beer, 2008). de beer (2008) disregards explicit breakdowns and models them by stochastic fluctuations of processing times of maximum 10%. this paper follows scholz-reiter et al. (2008) and windt et al. (2010) by modelling explicit breakdowns, which increase both dynamics and uncertainty. to achieve a similar degree of dynamics as de beer (2008), all stations are modelled with a failure rate of 10% and a deterministic mean-time-to-repair (mttr) of 15 minutes. based on the described setup, several combinations of methods and input data are examined, as shown in figure 6. three data sets with increasing dynamics are combined with three order generation methods (cf. table 5), three order release methods and three autonomous control methods (cf. table 3). concluding from figure 6, 34=81 simulation scenarios are examined. figure 6. considered methods and combinations. each input data set (i.e. order situation) comprises 5000 orders following scholz-reiter (2011). in general, we assume customer orders with given due dates. the order situations are depicted in figure 7, figure 8 and figure 9. the stable order situation contains almost no fluctuations concerning the demand quantity. all product variants have the same demand proportion. in this order situation, the same amount of each product type is manufactured every day. figure 7. stable order situation. the second data set comprises demand fluctuations concerning the demanded quantity. all product variants still have the same demand proportion. therefore, this order situation contains higher dynamics than in the stable order situation, but the product mix is the same. 49int. j. prod. manag. eng. (2015) 3(1), 43-56attribution-noncommercial-noderivatives 4.0 international coupling order release methods with autonomous control methods – an assessment of potentials by literature review and discrete event simulation http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 8. fluctuation in quantity. the third data set contains both fluctuations in quantity and quality, i.e. product variants. the product variants show an altering, periodic recurring demand proportion. the curve progression is similar to oscillating arrival. thus, figure 9 represents a very dynamic order situation with changing order volume and a changing product mix. figure 9. fluctuation in quantity and quality. these data sets represent customer orders, which are converted into production orders by the order generation methods. three different order generation methods are considered in the simulation study in section 3. they are systematically chosen according to the classification criteria, as table 5 indicates. the order release method according to the customer orders directly converts customer orders to production orders. the cumulative production figures method divides the production into control blocks and matches produced quantities with the production plan in a regular interval (lödding, 2013). the method is implemented with the whole production as one control block and a daily matching of produced quantities. the periodic batch control method is implemented to release orders in a rolling horizon of a 6-day period. this is consistent with a weekly planning period, so that at the beginning of each week all orders for this week plus the first working day of the following week are immediately generated. table 5. selection of order generation methods, cf. lödding (2013). generation method scope of order generation trigger logic character of method primary criterion for generating customer orders single-level event oriented specific date and time cumulative production figures definable event oriented generic definable periodic batch control single-level periodic specific date and time finally, parameters for the conwip order release method and for the phe control approach have to be defined before the simulation. conwip is implemented for a maximum load of 27 orders in the shop. with uniformly distributed orders in a 3x3 machine-model, a conwip of 27 causes 3 orders per station in average with 2 orders in queue. an existing queue is necessary to efficiently make use of methods like the srb1 or dd method, so that 27 orders are an adequate parameterisation. the pheromone approach uses a pheromone length of 5 orders as suggested by armbruster et al. (2006). 3.3. results the simulation results depending on applied order release methods and data sets are summarised in table 6. the values are averaged both over the three considered autonomous control methods and order generation methods depending on the order release method. the best value per order situation and logistic objective is highlighted. table 6. averaged results dependent on order release method. immediate order release planned order release conwip st ab le o rd er si tu at io n due date reliability 0,479 0,228 0,131 wip (pcs.) 99,26 50,00 18,55 utilisation 0,929 0,902 0,898 average throughput time [h:mm:ss] 111:27:00 55:10:00 19:42 q ua nt it y flu ct ua ti on due date reliability 0,568 0,122 0,533 wip (pcs.) 238,23 247,04 20,66 utilisation 0,852 0,830 0,822 average through-put time [h:mm:ss] 381:52:00 399:06:00 26:30:00 f lu ct ua ti on in qu an ti ty a nd qu al it y due date reliability 0,011 0,112 0,110 wip (pcs.) 21,75 192,9 19,02 utilisation 0,819 0,853 0,851 average throughput time [h:mm:ss] 29:02:00 271:48:00 22:59 50 int. j. prod. manag. eng. (2015) 3(1), 43-56 attribution-noncommercial-noderivatives 4.0 international grundstein, s., schukraft, s., scholz-reiter, b. and freitag, m. http://creativecommons.org/licenses/by-nc-nd/4.0/ the table indicates that there is no dominant order release method, i.e. different order release methods achieve considerable better values than others, but every method achieves at least two best values. especially the throughput time reveals room for improvement e.g. comparing conwip to immediate order release. it is consequentially, that conwip achieves the best results both in wip and in throughput time due to the direct relationship between those logistic objectives (nyhuis & wiendahl, 2008). the immediate order release disregards the wip level and thus causes higher wip and throughput time in all examined cases. concerning the relationship between utilisation and wip, it is noticeable, that the increase of 3% utilisation from conwip to immediate order release quintuples the wip in the stable order situation and approximately decuples the wip in the order situation with fluctuations in quantity. this is consistent with the funnel model, compare nyhuis & wiendahl (2008): the utilisation increases slowly with disproportionately rising throughput time at high level of utilisation. the higher wip level also causes an excessive increase in throughput time. the immediate order release achieves the best results in the first two data sets concerning the due date reliability. this is logical, because the other two order release methods keep back known orders, while the immediate order release method releases the orders considerably earlier. nevertheless, the high due date reliability of the immediate order release method goes along with high wip and throughput time. regarding the due date reliability with fluctuation in quantity and quality, the planned order release performs best. also conwip performs considerably better than the immediate order release method in this case. these results can be explained via the load balancing of these two methods. by smoothing the fluctuations a higher overall utilisation can be achieved by a homogenous distribution of orders onto stations. thus, the due date reliability can be improved. however, it seems counterintuitive that in the most dynamic situation, the planned order release achieves the highest due date reliability. taking into account the high wip in this case, it is explainable that the plan causes high wip, so that autonomous control methods have long queues for the calculation and many decision alternatives. therefore, for the price of high wip an increase of due date reliability is possible. nevertheless, the best value of 11% is still deficient. this relationship between wip and due date reliability is not a general one. comparing these two performance indicators between conwip and immediate order release with fluctuations in quantity and quality, it is notable that conwip achieves both a better due date reliability and lower wip with also a higher utilisation. this can be explained via the specific characteristics of the autonomous control methods. table 6 comprises average values which vary depending on the applied autonomous control methods. this is exemplarily depicted in figure 10 and figure 11. they illustrate the due date reliability for different data sets and for different combinations of order release methods and autonomous control methods. figure 10. due date reliability in stable order situation. figure 11. due date reliability with fluctuation in quantity and quality. apparently, different combinations perform considerably better in different order situations. especially in dynamic situations order release methods other than immediate release achieve significant better results. as explained above, smoothing the fluctuations enables a higher overall utilisation and improves the due date reliability by a homogenous distribution of orders onto stations. thereby, especially due date oriented and queuelength oriented methods (srb 1, dd) can prove 51int. j. prod. manag. eng. (2015) 3(1), 43-56attribution-noncommercial-noderivatives 4.0 international coupling order release methods with autonomous control methods – an assessment of potentials by literature review and discrete event simulation http://creativecommons.org/licenses/by-nc-nd/4.0/ their potential. the phe method performs worst of all three considered methods because it reacts very slowly, especially in case of multiple variants. in this case, the phe method uses the throughput time of the last 5 orders of the same product type on the corresponding station to decide, on which station to be processed. thereby, deviations of processing times and queue length are considered and taken into account in decision making. however, if certain product types have not been produced for a longer time, e.g. due to demand fluctuations, throughput time information is out of date. therefore, the quality of decision making generally decreases with a higher number of product variants applying the phe method. this disadvantage could for instance be compensated by also taking into account the throughput time of other product variants and calculating the ration between the expected throughput time and measured throughput time. the last aspect to examine is the impact of order generation methods. by building the average over order release methods, table 6 implicitly assumes that the impact of order generation methods can be disregarded. therefore, figure 12 examines the absolute deviation from averaged values of performance indicators depending on the order generation method. it represents a comparison with table 6. only 3 out of 108 values show a deviation of more than 5 % with an average deviation of 0,79 % and a standard deviation of 4,23 %. these findings justify the assumption, that the effects of order generation are significantly softened by the following order release method. thus, the focus on order release is justifiable. the performance of the considered methods allows several conclusions for matching order release methods and autonomous control methods. as far as order release methods are concerned, conwip is recommended in combination with autonomous control methods in rather dynamic situations if low wip and short throughput times are focussed. high due date reliability can be achieved with planned order release even in dynamic situations. however, the plan was scheduled with inserted idle times, so that it was robust towards the occurring dynamics. the immediate order release in combination with autonomous control methods is advantageous concerning utilisation in less dynamic order situations. nevertheless, the efficiency of both order release methods and autonomous control methods strongly depends on the current order situation, logistic targets and further properties of the production system so that the overall suitability must be evaluated in each particular case. figure 12. absolute deviation from average values depending on order generation methods. 52 int. j. prod. manag. eng. (2015) 3(1), 43-56 attribution-noncommercial-noderivatives 4.0 international grundstein, s., schukraft, s., scholz-reiter, b. and freitag, m. http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. conclusions production planning and control (ppc) has to cope with increasing complexity, dynamics and uncertainty (kim & duffie, 2004; westphal, 2001). autonomous control methods are considered as a promising approach for coping with these challenges (scholz-reiter & freitag, 2007b). however, the effect of order release methods on the efficiency of autonomous control methods has not yet been examined. this paper provides a theoretic analysis of feasible method combinations. furthermore, simulations confirm the high influence of the order release function on autonomous control methods. the results further revealed that the effects of order generation methods are softened by order release methods, so that for future research the assumption is justifiable to initially focus on combining order release with autonomous control. the differences of several autonomous control methods concerning the logistic objectives also reveal that more extensive studies need to be carried out to evaluate the suitability of combinations for different production systems. there is a broad consensus, that there is no universal dominant release method, respectively control method. the efficiency of both order release methods and autonomous control methods strongly depends on the current order situation, logistic targets and further properties of the production system. users find a decision help in table 3 to preselect method combinations for detailed evaluation. future research must examine, which combinations of table 3 are generally promising for which production system, and furthermore, which adaptions can be carried out to apply combinations of release methods and autonomous control methods of table 3, which are currently not applicable. these adaptions concern especially the application of load balancing order release methods with autonomous control methods. moreover, broadening the scope of research by approaches of capacity control is also field of future research. acknowledgements this research was funded by the german research foundation (dfg) under the reference number scho 540/26-1 “methods for the interlinking of central planning and autonomous control in production”. references armbruster, d., de beer, c., freitag, m., jagalski, t., ringhofer, c. (2006). autonomous control of production networks using a pheromone approach. physica a: statistical mechanics and its applications, 363(1): 104-114. doi:10.1016/j.physa.2006.01.052 barenji, v. t., barenji, a. v., hashemipour, m. (2014). a multi-agent rfid-enabled distributed control system for a flexible manufacturing shop. the international journal of advanced manufacturing technology, 71, 1773-1791. doi:10.1007/s00170-013-5597-2 baykasoğlu, a., göçken, m. (2011). a simulation based approach to analyse the effects of job release on the performance of a multi-stage job-shop with processing flexibility. international journal of production research, 49(2): 585-610. doi:10.1080/00207540903479778 bechte, w. (1980). steuerung der durchlaufzeit durch belastungsorientierte auftragsfreigabe bei werkstattfertigung. (doctoral dissertation). universität hannover, hanover, germany. bergamaschi, d., cigolini, r., perona, a., portioli, a. (1997). order review and release strategies in a job shop environment: a review and classification. international journal of production research, 35(2): 399-420. doi:10.1080/002075497195821 byeon, e. s., wu, s. d., storer, r. h. (1998). decomposition heuristics for robust job-shop scheduling. ieee transactions on robotics and automation, 14(2): 303-313. doi:10.1109/70.681248 cicirello, v., smith, s. (2001). ant colony control for autonomous decentralized shop floor routing. proceedings of the isads-2001 fifth international symposium on autonomous decentralized systems, dallas, usa: i.e.e.e.press. de beer, c. (2008). untersuchung der dynamik von selbststeuernden prozessen in produktionslogistischen systemen anhand ereignisdiskreter simulationsmodelle. (doctoral dissertation). universität bremen, bremen, germany. duffie, n. a., roy, d., shi, l. (2008). dynamic modeling of production networks of autonomous work systems with local capacity control. cirp annals manufacturing technology, 57(1): 463-466. doi:10.1016/j.cirp.2008.03.018 duffie, n. a., shi, l. (2009). maintaining constant wip-regulation dynamics in production networks with autonomous work systems. cirp annals manufacturing technology, 58(1): 399-402. doi:10.1016/j.cirp.2009.03.012 fernandes, n. o., carmo-silva, s. (2006). generic polca—a production and materials flow control mechanism for quick response manufacturing. international journal of production economics, 104, 74-84. doi:10.1016/j.ijpe.2005.07.003 fernandes, n. o., carmo-silva, s. (2011). order release in a workload controlled flow-shop with sequence-dependent set-up times. international journal of production research, 49(8): 2443-2454. doi:10.1080/00207541003720376 53int. j. prod. manag. eng. (2015) 3(1), 43-56attribution-noncommercial-noderivatives 4.0 international coupling order release methods with autonomous control methods – an assessment of potentials by literature review and discrete event simulation http://dx.doi.org/10.1016/j.physa.2006.01.052 http://dx.doi.org/10.1007/s00170-013-5597-2 http://dx.doi.org/10.1080/00207540903479778 http://dx.doi.org/10.1080/002075497195821 http://dx.doi.org/10.1109/70.681248 http://dx.doi.org/10.1016/j.cirp.2008.03.018 http://dx.doi.org/10.1016/j.cirp.2009.03.012 http://dx.doi.org/10.1016/j.ijpe.2005.07.003 http://dx.doi.org/10.1080/00207541003720376 http://creativecommons.org/licenses/by-nc-nd/4.0/ freitag, m., herzog, o., scholz-reiter, b. (2004). selbststeuerung logistischer prozesse ein paradigmenwechsel und seine grenzen. industrie management, 20(1): 23-27. gan, h., wirth, a. (2005). comparing deterministic, robust and online scheduling using entropy. international journal of production research, 43(10): 2113-2134. doi:10.1080/00207540412331333405 gentile, f., rogers, k. j. (2009). order release and dispatching in a sequence dependent job shop. management of engineering & technology picmet 2009, 1185-1196. doi:10.1109/picmet.2009.5262014 gierth, a. (2009). beurteilung der selbststeuerung logistischer prozesse in der werkstattfertigung. (doctoral dissertation). rwth aachen, aachen, germany. irastorza, j., deane, r. (1974). a loading and balancing methodology for job-shop control. aiie transactions, 6(4): 302-307. doi:10.1080/05695557408974968 jendralski, j. (1978). kapazitätsterminierung zur bestandsregelung in der werkstattfertigung. (doctoral dissertation). technische universität hannover, hanover, germany. kim, j.-h., duffie, n. a. (2004). backlog control for a closed loop ppc system. annals of the cirp, 53(1): 357-360. doi:10.1016/s00078506(07)60716-1 kornienko, s., kornienko, o., priese, j. (2004). application of multi-agent planning to the assignment problem. computers in industry, 54(3): 273-290. doi:10.1016/j.compind.2003.11.002 kurbel, k. (2005). produktionsplanung und -steuerung im enterprise resource planning und supply chain management. munich, germany: oldenbourg. land, m., stevenson, m., thürer, m. (2012). integrating load-based order release and priority dispatching. international journal of production research, 52(4): 1059-1073. doi:10.1080/00207543.2013.836614 lawrence, s. r., sewell, s. e. (1997). heuristic, optimal, static and dynamic schedules when processing times are uncertain. journal of operations management, 15(1): 71-82. doi:10.1016/s0272-6963(96)00090-3 leisten, r. (1996). iterative aggregation und mehrstufige entscheidungsmodelle. einordnung in den planerischen kontext, analyse anhand der modelle der linearen programmierung und darstellung am anwendungsbeispiel der hierarchischen produktionsplanung. heidelberg: physica. doi:10.1007/978-3-642-57875-5 leitão, p., restivo, f. j. (2008). implementation of a holonic control system in a flexible manufacturing system. ieee transactions on systems, man, and cybernetics, 38(5): 699-709. doi:10.1109/tsmcc.2008.923881 lödding, h. (2001). dezentrale bestandsorientierte fertigungsregelung. (doctoral dissertation). universität hannover, hanover, germany. lödding, h. (2013). handbook of manufacturing control. fundamentals, description, configuration. berlin, germany: springer. doi:10.1007/978-3-642-24458-2 luczak, h., eversheim, w. (1999). produktionsplanung und -steuerung. grundlagen, gestaltung und konzepte. berlin, germany: springer. doi:10.1007/978-3-662-09472-3 melnyk, s. a., ragatz, g. l. (1989). order review/release: research issues and perspectives. international journal of production research, 27(7): 1081-1096. doi:10.1080/00207548908942609 nyhuis, p., wiendahl, h. -p. (2008). fundamentals of production logistics: theory, tools and applications. berlin, germany: springer. o’donovan, r., uzsoy, r., mckay, k. n. (1999). predictable scheduling of a single machine with breakdowns and sensitive jobs. international journal of production research, 37(18): 4217-4233. doi:10.1080/002075499189745 ostgathe, m. (2012). system zur produktbasierten steuerung von abläufen in der auftragsbezogenen fertigung und montage. (doctoral dissertation). technische universität münchen, münchen, germany. owliya m., saadat, m., anane, r., goharian, m. (2012). a new agents-based model for dynamic job allocation in manufacturing shopfloors. ieee systems journal, 6(2): 353-361. doi:10.1109/jsyst.2012.2188435 pach, c., bekrar, a., zbib, n., sallez,y., trentesaux, d. (2012). an effective potential field approach to fms holonic heterarchical control. control engineering practice, 20(12): 1293-1309. doi:10.1016/j.conengprac.2012.07.005 pannequin, r., morel, g., thomas, a. (2009). the performance of product-driven manufacturing control: an emulation-based benchmarking study. computers in industry, 60(3): 195-203. doi:10.1016/j.compind.2008.12.007 park, h.-s., tran, n.-h. (2012). an autonomous manufacturing system based on swarm of cognitive agents. journal of manufacturing systems, 31(3): 337-348. doi:10.1016/j.jmsy.2012.05.002 pinedo, m. l. (2008). scheduling. theory, algorithms and systems. new york, usa: springer. qi, c., sivakumar, a. i., gershwin, s. b. (2009). an efficient new job release control methodology. international journal of production research, 47(3): 703-731. doi:10.1080/00207540701455335 reaidy, j., massotte, p., diep, d. (2006). comparison of negotiation protocols in dynamic agent-based manufacturing systems. international journal of production economics, 99, 117-130. doi:10.1016/j.ijpe.2004.12.011 54 int. j. prod. manag. eng. (2015) 3(1), 43-56 attribution-noncommercial-noderivatives 4.0 international grundstein, s., schukraft, s., scholz-reiter, b. and freitag, m. http://dx.doi.org/10.1080/00207540412331333405 http://dx.doi.org/10.1109/picmet.2009.5262014 http://dx.doi.org/10.1080/05695557408974968 http://dx.doi.org/10.1016/s0007-8506(07)60716-1 http://dx.doi.org/10.1016/s0007-8506(07)60716-1 http://dx.doi.org/10.1016/j.compind.2003.11.002 http://dx.doi.org/10.1080/00207543.2013.836614 http://dx.doi.org/10.1016/s0272-6963(96)00090-3 http://dx.doi.org/10.1007/978-3-642-57875-5 http://dx.doi.org/10.1109/tsmcc.2008.923881 http://dx.doi.org/10.1007/978-3-642-24458-2 http://dx.doi.org/10.1007/978-3-662-09472-3 http://dx.doi.org/10.1080/00207548908942609 http://dx.doi.org/10.1080/002075499189745 http://dx.doi.org/10.1109/jsyst.2012.2188435 http://dx.doi.org/10.1016/j.conengprac.2012.07.005 http://dx.doi.org/10.1016/j.compind.2008.12.007 http://dx.doi.org/10.1016/j.jmsy.2012.05.002 http://dx.doi.org/10.1080/00207540701455335 http://dx.doi.org/10.1016/j.ijpe.2004.12.011 http://creativecommons.org/licenses/by-nc-nd/4.0/ rekersbrink, h. (2012). methoden zum selbststeuernden routing autonomer logistischer objekte. (doctoral dissertation). universität bremen, bremen, germany. sabuncuoglu, i., karabuk, s. (1999). rescheduling frequency in an fms with uncertain processing times and unreliable machines. journal of manufacturing systems, 18(4): 268-283. doi:10.1016/s0278-6125(00)86630-3 sabuncuoglu, i., karapinar, h. (1999). analysis of order review/release problems in production systems. international journal of production economics, 62(3): 259-279. doi:10.1016/s0925-5273(98)00248-5 schmidt, m., fronia, p., fisser, f., nyhuis, p. (2007). decentralized planning and control for assembly areas driven by gentelligent® parts. industrial engineering and engineering management 2007 ieee, 1088-1092. doi:10.1109/ieem.2007.4419360 scholz-reiter, b., harjes, f., mansfeld, j., stasch, o. (2011). an inventory and capacity-oriented production control concept for the shop floor based on artificial neural networks. proceedings of the 7th international conference on digital enterprise technology 2011 (det 2011), 213-220. athens, greece: n.p. scholz-reiter, b., görges, m., jagalski, t., naujok, l. (2010). modelling and analysis of an autonomous control method based on bacterial chemotaxis. proceedings of the 43rd cirp international conference on manufacturing systems 2010 (icms 2010), 699-706. wien, austria: neuer wissenschaftlicher verlag. scholz-reiter, b., jagalski, t., bendul, j. (2008). autonomous control of a shop floor based on bee’s foraging behaviour. in haasis, h.-d., kreowski, h.-j., b. scholz-reiter (eds.), dynamics in logistics (pp. 415-423). berlin, germany.: springer. doi:10.1007/978-3-54076862-3_41 scholz-reiter, b., böse, f., jagalski, t., windt, k. (2007a). selbststeuerung in der betrieblichen praxis. ein framework zur auswahl der passenden selbststeuerungsstrategie. industrie management, 23(3): 7-10. scholz-reiter, b., freitag, m. (2007b). autonomous processes in assembly systems. cirp annals, 56(2): 712-729. doi:10.1016/j. cirp.2007.10.002 scholz-reiter, b., freitag, m., de beer, c., jagalski, t. (2006). the influence of production network’s complexity on the performance of autonomous control methods. proceedings of the 5th cirp international seminar on computation in manufacturing engineering, 317320. schuh, g. (ed.). (2006). produktionsplanung und -steuerung. berlin, germany: springer. doi:10.1007/3-540-33855-1 schuh, g., gottschalk, s., höhne, t. (2007). high resolution production management. annals of the cirp, 56(1): 439-442. doi:10.1016/j. cirp.2007.05.105 siwamogsatham, t., saygin, c. (2004). auction-based distributed scheduling and control scheme for flexible manufacturing systems. international journal of production research, 42(3): 547-572. doi:10.1080/00207540310001613683 sudo, y., matsuda, m. (2013). agent based manufacturing simulation for efficient assembly operations. procedia cirp, 7: 437-442. doi:10.1016/j.procir.2013.06.012 szelke, e., monostori, l. (1999). reactive scheduling in real time production control. in brandimarte, p. & villa, a. (eds.), modeling manufacturing systems (pp. 65-113). berlin, germany: springer. doi:10.1007/978-3-662-03853-6_5 thürer, m., filho, m. g., stevenson, m. (2013). coping with finite storage space in job shops through order release control: an assessment by simulation. international journal of computer integrated manufacturing, 26(9): 830-838. doi:10.1080/0951192x.2013.799779 thürer, m., stevenson, m., silva, c., land, m. j., fredendall, l. d. (2012). workload control and order release: a lean solution for make-toorder companies. production and operations management, 21(5): 939-953. doi:10.1111/j.1937-5956.2011.01307.x tsutsui, s., liu, l. (2007). cunning ant system for quadratic assignment problem with local search and parallelization. in ghosh, a., de, r.k. & pal, s.k. (eds.), pattern recognition and machine intelligence (pp 269-278). berlin, germany: springer. doi:10.1007/978-3-54077046-6_33 van brackel, t. (2009). adaptive steuerung flexibler werkstattfertigungssysteme. (doctoral dissertation). universität paderborn, paderborn, germany. wang, l., tang, d.-b., gu, w.-b., zheng, k., yuan, w.-d., tang, d.-s. (2012). pheromone-based coordination for manufacturing system control. journal of intelligent manufacturing, 23, 747-757. doi:10.1007/s10845-010-0426-z wang, l.-c., lin, s.-k. (2009). a multi-agent based agile manufacturing planning and control system. computers & industrial engineering, 57(2): 620-640. doi:10.1016/j.cie.2009.05.015 wein, l. m. (1988). scheduling semiconductor wafer fabrication. ieee transactions on semiconductor manufacturing, 1(3): 115-130. doi:10.1109/66.4384 wein, l. m., chevalier, p. b. (1992). a broader view of the job-shop scheduling problem. management science, 38(7): 1018-1033. doi:10.1287/mnsc.38.7.1018 weng, m. x., wu, z., qi, g., zheng, l. (2008). multi-agent-based workload control for make-to-order manufacturing. international journal of production research, 46(8): 2197-2213. doi:10.1080/00207540600969758 westphal, j. r. (2001). komplexitätsmanagement in der produktionslogistik ein ansatz zur flussorientierten gestaltung und lenkung heterogener produktionssysteme. wiesbaden, germany: deutscher universitäts verlag. 55int. j. prod. manag. eng. (2015) 3(1), 43-56attribution-noncommercial-noderivatives 4.0 international coupling order release methods with autonomous control methods – an assessment of potentials by literature review and discrete event simulation http://dx.doi.org/10.1016/s0278-6125(00)86630-3 http://dx.doi.org/10.1016/s0925-5273(98)00248-5 http://dx.doi.org/10.1109/ieem.2007.4419360 http://dx.doi.org/10.1007/978-3-540-76862-3_41 http://dx.doi.org/10.1007/978-3-540-76862-3_41 http://dx.doi.org/10.1016/j.cirp.2007.10.002 http://dx.doi.org/10.1016/j.cirp.2007.10.002 http://dx.doi.org/10.1007/3-540-33855-1 http://dx.doi.org/10.1016/j.cirp.2007.05.105 http://dx.doi.org/10.1016/j.cirp.2007.05.105 http://dx.doi.org/10.1080/00207540310001613683 http://dx.doi.org/10.1016/j.procir.2013.06.012 http://dx.doi.org/10.1007/978-3-662-03853-6_5 http://dx.doi.org/10.1080/0951192x.2013.799779 http://dx.doi.org/10.1111/j.1937-5956.2011.01307.x http://dx.doi.org/10.1007/978-3-540-77046-6_33 http://dx.doi.org/10.1007/s10845-010-0426-z http://dx.doi.org/10.1016/j.cie.2009.05.015 http://dx.doi.org/10.1109/66.4384 http://dx.doi.org/10.1287/mnsc.38.7.1018 http://dx.doi.org/10.1080/00207540600969758 http://creativecommons.org/licenses/by-nc-nd/4.0/ wiendahl, h.-h., von cieminski, g., wiendahl, h.-p. (2005). stumbling blocks of ppc: towards the holistic configuration of ppc systems. production planning and control, 16(7): 634-675. doi:10.1080/09537280500249280 wiendahl, h.-p. (ed.). (1991). anwendung der belastungsorientierten auftragsfreigabe. munich, germany: carl hanser. wiendahl, h.-p. (1997). fertigungsregelung. logistische beherrschung von fertigungsabläufen auf basis des trichtermodells. munich, germany: carl hanser. wiendahl, h.-p. (ed.). (2005). betriebsorganisation für ingenieure. munich: hanser. windt, k., becker, t., jeken, o., gelessus, a. (2010). a classification pattern for autonomous control methods in logistics. logistics research, 2(2): 109-120. doi:10.1007/s12159-010-0030-9 wong, t. n., leung, c. w., mak, k. l., fung, r. y. k. (2006). an agent-based negotiation approach to integrate process planning and scheduling. international journal of production research, 44(7): 1331-1351. doi:10.1080/00207540500409723 wyssusek, b. (1999). grundlagen der systemanalyse. in krallmann, h., frank, h. & gronau, n. (eds.), sytemanalyse im unternehmen (pp. 19-43). munich, germany: oldenbourg. xiang,w., lee, h.p. (2008). ant colony intelligence in multi-agent dynamic manufacturing scheduling. engineering applications of artificial intelligence, 21(1): 73-85. doi:10.1016/j.engappai.2007.03.008 zozom, a., hodgson, t. j., king, r. e., weintraub, a. j., cormier, d. (2003). integrated job release and shopfloor scheduling to minimize wip and meet due-dates. international journal of production research, 41(1): 31-45. doi:10.1080/00207540210162992 56 int. j. prod. manag. eng. (2015) 3(1), 43-56 attribution-noncommercial-noderivatives 4.0 international grundstein, s., schukraft, s., scholz-reiter, b. and freitag, m. http://dx.doi.org/10.1080/09537280500249280 http://dx.doi.org/10.1007/s12159-010-0030-9 http://dx.doi.org/10.1080/00207540500409723 http://dx.doi.org/10.1016/j.engappai.2007.03.008 http://dx.doi.org/10.1080/00207540210162992 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2019.8607 received 2017-09-20 accepted: 2018-12-31 value stream mapping as a lean manufacturing tool: a new account approach for cost saving in a textile company carvalho, c.p.a1, carvalho, d.s.b and silva, m.b.a2 adepartment of chemistry, school of engineering of lorena, são paulo state university, lorena, são paulo, brazil bfaculty of human sciences of cruzeiro facic, brazil a1 cleginaldopcarvalho@hotmail.com abstract: companies around the world are under pressure to reduce their prices to be competitive. in order to keep profitability, they are adopting lean manufacturing (lm) and its tools to reduce waste and, consequently, the costs. value stream mapping (vsm) is an efficient lm tool, identifying production flow waste. when it is possible to combine vsm with other lm tool, such as kanban, a substantial impact in inventory reduction is reached. additionally, the conventional account system shows limitations in the costs reduction to keep up with lm. the purpose of this paper is to present the results obtained by the application of the vsm combined with kanban and the new account approach to measure the effectiveness of the cost reduction in the finished goods inventory. considering that lm tools application in the textile segment has not an expressive number of researches, four hypothesis are presented and validated. the findings are discussed and finally, the research limitations and the practical implications originated by this work are presented as well as topics for further researches are suggested. key words: value stream mapping, kanban, activity-based costing, lean manufacturing, textile, lean accounting. 1. introduction in the last decades, many organizations in brazil have implemented lean manufacturing (lm) tools with the objective of increasing their competitiveness. the majority of the applications were for discrete manufacturing (dm) systems in the auto-parts and automobile segments. a small number of applications were considered for continuous manufacturing (cm) systems in textile, steel and chemical segments. the first time the term lean was described was in the book the machine that changed the world (womack, jones and ross, 1990), in which the solid gap between the mass production conception and lean production was exposed. one of the main features of lean is the focus on eliminating waste or muda (in japanese) all things that do not add value to the product or service under the customer point of view by the means of continuous improvement tasks. a lm tool designed over the years was value stream mapping (vsm), which is used to manage and follow spot ways and identify and remove waste (rother and shook, 1998). it is meant to design the material flow through the operations, considering the whole supply chain, with the information of cycle time, downtime, and inventories and other key operations. lopez et al. (2013) presented the value stream costing as an alternative costing technique as an approach to minimize costs in the lean enterprise. unfortunately, two key problems are encountered when talking about lm: most of the research that we have nowadays is based on the outcome of to cite this article: carvalho, c.p., gonçalves, l.w.n. and silva, m.b. (2019). value stream mapping as a lean manufacturing tool: a new account approach for cost saving in a textile company. international journal of production management and engineering, 6(2), 1-12. https://doi.org/10.4995/ijpme.2019.8607 int. j. prod. manag. eng. (2019) 7(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international 1 http://creativecommons.org/licenses/by-nc-nd/4.0/ big corporations after introducing lm into their production line, and the preponderance of discrete manufacturing system (dms) articles and the lack of continuous manufacturing system (cms), as the case of the textile segment (carvalho et al., 2017). one of the main issues introducing lm in brazil is how predominant the traditional cost accounting methods are in the brazilian companies and how behind their methods are. most of these traditional methods clash in some way or another with lm. big companies have already shown some type of resistance in introducing lm, while small companies, either, don’t have the knowledge lm exists or are too comfortable to revamp and make lm a reality. in this paper, we take a small textile company in a small town in brazil as a case study. its current manufacturing process (cvsm) and current cost analysis will be analyzed, and a future stream mapping (fvsm) will be introduced with the goal to minimize cost and maximize profitability. this study aims to present a new account approach for cost saving validation considering a textile company in which had applied two combined lm tools in order to reduce its production wastes. 2. research hypotheses there is not a vast list of literature about the application of a vsm tool in the textile enterprise. a few articles can be found, such as soliman (1998), which describes vsm as an important approach for process industry making it easier to identify where and how to improve. some authors highlight the efficiency and validity of vsm as a lean tool in a continuous process industry: abdulmalek and rajgopal (2007) in a steel mill, ratnayake and chaudry (2015) in the oil and gas industry, seth et al. (2007) in a cottonseed oil company, ragadali (2010) in a plastic extrusion segment and palauro (2014) in a pulp production. in a similar way, all the articles explained points to be improved through vsm techniques in different industries. in a textile industry context, hodge et al. (2011) observed that those who have implemented lean showed significant use of lean tools such as 5s, kaizen, kanban, “value stream mapping” (vsm), and “visual control”. panwar et al. (2013) studied the level of implementation of lean manufacturing in indian process industries and found out that, even though, previous literature exemplifies vsm as a primary activity while adopting lean manufacturing (seth et al., 2007; mahapatra et al., 2006; upadhye et al., 2010), their study revealed that the use of vsm is not common in indian process industries. in the rank of significantly used lean tools in indian process industries, developed by them, vsm is the 16th tool out of 20. considering the topics described above, it is possible to establish the following hypotheses: hypothesis 1a. value stream mapping as an lm tool aids top managers to identify main wastes in the production flow. hypothesis 1b. value stream mapping applied with other lm tool reduces the inventory impacting positively in the enterprise turnover. the world is in a constant state of change and as the global market becomes more competitive companies need to rethink their strategies to attend consumers’ needs. companies not only have to keep up with multinationals but also need to maintain low prices without sacrificing product quality. in order to be competitive, companies need to think outside the box and implement new methodologies and activities. a herculean improvement in the manufacturing process has been the addition of lean activities in a company. what started out as a way to reduce waste in manufacturing suddenly overtook all departments of a company, including the finance and accounting departments. the purpose of implementing lean methods is to reduce waste or costs leading to increase in profit. as a result, the financial aspect of the company needs to be analyzed. in essence, there are two schools of thoughts relating the relationship between the implementation of lean and a company’s net profit. there are several authors who follow the idea that the implementation of lean does not add to the company. however, authors, such as, chenhall (1997), easton and jarrel (1998) and callem et al. (2003) drew a positive association between lean manufacturing and financial performance. following the connection of these authors and to illustrate this positive connection between them two, it is possible to outline the following hypothesis: hypothesis 1c. implementation of lean activities makes the company more liquid, positively impacting net profit. finished goods inventory does not add to cost until the products can be sold (carnes and hedin, 2015). int. j. prod. manag. eng. (2019) 7(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho, c.p., gonçalves, l.w.n. and silva, m.b. 2 http://creativecommons.org/licenses/by-nc-nd/4.0/ besides the cost to storage the finished product, the physical space needed for these products also hinders the possibilities for the business to expand or accept other ventures. hypothesis 1d. implementation of lean activities decrease inventory freeing up space and expanding business opportunities. 3. the limitations of traditional costing systems most companies tend to use outdated cost accounting systems. these cost methods and systems have been used for years and have aided several companies to keep track of their costs and determine their product price. it is important to understand how they work and outline their flaws and limitations to better understand the need to move forwards and the lean transformation for a better fit. 3.1. traditional accounting systems and the overhead problem traditionally, companies did their cost accounting with the solely purpose to supply managers and company’s decision makers with the specific information needed to understand the cost of running the business and what paths to take based on that information. differently than financial accounting, cost accounting is not subjected to general accepted accounting principles (gaap). because of this exemption, several types of cost accounting systems were created. vanderbeck (2010) explains that most of these systems use three main elements: direct materials, direct labor and overhead costs (indirect manufacturing costs). combining these elements with traditional cost accounting methods, gave companies the ability to compute the total cost incurred in the year and the unit cost. having this information benefits accountants to elaborate financial statements used by managers to better their planning and control over the company’s life. however, complications can emerge since managers seek more efficient process methods while still relying on these outdated accounting methods, this conflict can yield distorted costs. martins (2008) explains how most brazilian companies still rely on activity-based costing (abc) and total absorption costing as their main cost system, which generates major conflicts as the companies implement new manufacturing methods. to understand this better, lopez and arbós (2013) illustrated it better in equation (1), which highlights how absorption costing systems or full cost calculates the per unit cost c of each product over a period of time, which accounts for the sum of all different materials used, plus all the direct labor cost and the sum of all the overhead allocated divided by the total units produced that period equivalent units of production material cost irect labor cost overhead allocatedc d = + +|| (1) equation (1) can be misleading given that the more units the company produce the lower the unit cost. however, it does not necessarily mean all the products were sold on that period and that some of them are either still being produced in the inventory. to further examine the unit cost (equation (1)) the cost of material and the equivalent number of units of production (equation (3), where t is the current period and t+1 is the next period) need to be calculated by using inventory and work in process (wip), in order to better analyze if the company has significantly sold less products than it has produced: cost of materialst purchasest–∆inventory(t+1)–t (2) equivalent unitst=units completedt+∆wip(t+1)–t (3) even when taking into account the change in inventory it is observed that the increase in production will decrease the unit cost because the fixed overhead will be divided by a larger number increasing the company’s gross margin. it might look better for the company but the company will suffer with high inventory in the case of its products becoming obsolete or not profitable anymore. walther (2013) explains how arbitrarily choosing the allocation of overhead can either make or break a company’s decision given that it is extremely hard to isolate the total direct costs of products given that companies, nowadays, produce multiple products and it has become more difficult to track direct costs. even with all the communication and information available online, it is worth mentioning the lack of introduction of new cost accounting systems in brazil. a sheer amount of companies are still reluctant to adapt to even the smallest changes and still prefer traditional cost systems over improved methods such as lean or throughput system that reflects the implementation of newer manufacturing systems. int. j. prod. manag. eng. (2019) 7(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international value stream mapping as a lean manufacturing tool: a new account approach for cost saving in a textile company 3 http://creativecommons.org/licenses/by-nc-nd/4.0/ this is occurs especially with small companies, which still use outdated styles of manufacturing and cost accounting. 3.2. the demise of abc as companies expended and commenced producing a variety of products as opposed to only one, traditional accounting systems started becoming obsolete as it became harder to analyze the cost of several products at once. furthermore, johnson and kaplan (1987) conceptualized the idea that cost accounting should account for more than just the cost of material, direct labor and overhead; every little action that in some way brought value to the product should be integrated as cost, even if had novalues attached to the cost of selling the product, that’s when abc was created. abc was created to fix some of the deficiencies created by traditional accounting systems as the companies became larger and the process of selling a product became more robust and more complex. barfield, raiborn and kinney (1994) claim that consumers will not buy products if they perceive the product as not being cost effective or does not bring enough value to the price they are purchasing the product for. abc uses a system that embodies every cost system termed “activities” that allocates costs and identifies cost drivers. the cost driver problem is better comprehended in equation (4), which computers the unit cost c using the total amount of drivers and the amount of drivers consumed by the products in that period. ( ) c materialscost total amount of drivers labor all other amount of drivers cost # # = + +|| (4) even though, abc is a revolutionary system, it still can’t assign all overhead costs to products. the system’s flaw is shown on the reliance on determining the ‘right’ denominator (flanagan, 2008). the misuse of the wrong drivers will lead to unit cost distortions, which will produce inaccurate information. krumwiede (1998) explains how abc is not beneficial for every company. smaller companies might not possess the resources to correctly apply abc due to lack of time, it resources, detailed information. furthermore, this outdated system would negatively impact a small brazilian company due to the system being focused on high profitability and the importance of all variables being integrated in the system. as companies demand a more suitable cost system to complement their new manufacturing style, abc started losing ground in the 1990’s; even with authors such as cooper (1996) contemplating the idea that abc still supports jit principles, other authors such as kaplan and norton (1992) completely discredit abc and go as far as creating a better system that overtakes abc. huntzinger (2007) gives a more recent view to this topic, mentioning how the amount of resources needed (money, it, accurate data, time), actually averts companies from using abc systems. 4. cost management for lean manufacturing 4.1. lean accounting ideology with companies adopting the lean methodology the costing system had to change to match this evolution. the lean accounting incorporates what obviously adds value to the product by eliminating or reducing everything else (womack, 1990), which consists of waste and error elimination, capacity improvement and process acceleration. in essence, lean accounting eliminates wasted time and cost from the back office by simplifying systems. the philosophy behind the lean accounting described by womack and jones (2003) mentions five fundamental principles that eliminate company waste: value creation, value stream analysis, optimizing flows, pull system application and strive for perfection. to start these principles, the organization needs to identify consumers’ needs and expectations and derive the specific added values of the product. as the second principle, womack and jones (2003) determine that all work be organized in value streams. value stream consists of all activities a company needs to develop in order to sell a product, such as marketing, packaging, delivering, manufacturing, etcetera. this will generate a value to the consumer which will determine the profit of the company, besides highlighting the waste. each value stream can be mapped with progress charts. from these maps, the company can design a first map that might include possible obstacles and waste and a second int. j. prod. manag. eng. (2019) 7(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho, c.p., gonçalves, l.w.n. and silva, m.b. 4 http://creativecommons.org/licenses/by-nc-nd/4.0/ map that will provide the intended value stream status (maskell and baggarely, 2003). the third principle regards optimizing flow, which womack and jones (2003) explains three types of flows exist: physical flow of material, information flow and cash flow. the purpose of this breakdown is to facilitate information and make it easier for all involved to work focused on improving these flows. lean philosophy determines that all work must be separated by work cell. each cell organizes workers and equipment together physically and functionally instead of departments. after being organized, the equipment is then placed in a sequence in a continuous flow. womack and jones (2003) explain that all workers must be trained to work on all activities and handle all equipment. the fourth principle – the pull system – determines that the client will determine the level of production. alongside this principle, the company also has to implement just in time, which regards the company to obtain the specific part, and the specific place, and at the specific time (boyle, scherrer-rathje and stuart, 2011). the fifth and last principle, strive for perfection, brings all the other principles together by specifying and identifying the right value stream, the company has a continuous flow for a specific product that will enable consumers to pick the best product for them. this endless cycle will generate a product that is the closest to what the consumers want (womack and jones, 2003). with the best possible product, the company will determine what the waste parts are and will be able to eliminate these processes from the flow. kennedy and brewer (2005) specify that the managers need to empower the employees to also improve the value flows for the consumers instead of exclusively relying on the managers. 4.2. the evolution of cost accounting: the rise of lean accounting the shortcomings of traditional cost systems led to a breakthrough in cost accounting that would aid and match the lean philosophy that companies started implementing: lean accounting. to complement the advances in lean manufacturing, “lean accounting aims to provide information useful to people who are implementing and sustaining lean manufacturing” (maskell, 2000). as the second principle stands, vsc is essential for managers to implement lean accounting. 5. methodology 5.1. the company´s profile founded in 1994, the studied company started its activities in the hot stamped sport shirts segments, when purchased the textile from the different sort of the suppliers. looking forward to amplify its market share the company had started its own production in 1997, when started a new brand. in 2010, with strong investments in news pieces of equipment and installations, the company started its operations in the textile segment, when emerging a vast range of products, initially in the sport segment, focusing in the whole brazilian domestic market. located in the paraiba valley, sao paulo state, brazil, the company currently has 22 employees, and produces 30 tons per moth of textile and stamps in average 4000 sport shirts monthly. 5.2. method the research work conducted in this paper was an exploratory case study. regarding to the scientific approach this research can be classified as qualiquanti and applied as its nature. the data are collected during a period of six months when the company was operated in two shifts. the search of information was the field research and several meetings were conducted with blue colors employees, managers and the company board. 6. findings the first results of this work were regarded to the vsm elaboration for current production flow situation, reached in the beginning of this research. managing the information collected from the production reports, the current value stream map (cvsm) was designed. a series of discussions among the researchers and the studied company´s top managers was conducted and kanban was elected as the lm tool to be applied to reduce the main detected waste – finished goods inventory. after kanban implementation, the future value stream map was designed. the results in the inventory level reduction were considering in a new account approach, and the results also are presented. int. j. prod. manag. eng. (2019) 7(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international value stream mapping as a lean manufacturing tool: a new account approach for cost saving in a textile company 5 http://creativecommons.org/licenses/by-nc-nd/4.0/ 6.1. the current stream mapping the next step is to draw the current state map which is done after the production flow observation. in figure 1 there are some basic icons utilized in the mapping. figure 1. basic icons used for mapping design. cycle time, inventory time and the number of employees in each position were collected by informal interviews with personnel during the visits to the textile company. and some data were collected as shown in tables 1 to 5. table 1. pieces produced (embroidery). machine 01 date machine 01 date 407 01 august 405 09 august 117 02 august 260 10 august 250 03 august 330 11 august 350 04 august 340 12 august 230 05 august 295 13 august 200 06 august 385 15 august 220 08 august 320 16 august table 2. pieces produced (stamping). date machine 01 machine 02 machine 03 machine 05 machine 06 01 august 0 500 790 450 380 02 august 520 550 260 420 420 03 august 320 620 440 600 550 04 august 210 200 180 210 200 05 august 620 610 430 580 530 06 august 550 440 670 520 510 08 august 550 680 550 560 520 09 august 280 220 210 180 200 10 august 0 550 600 0 420 11 august 620 610 440 500 580 12 august 650 480 620 610 590 13 august 350 490 660 480 520 15 august 220 205 200 210 210 16 august 480 520 440 520 520 table 3. embroidery non-programmed stopping times (minutes). date busted line needle exchange 23 august 20 24 august 14 2 25 august 35 2 26 august 36 2 12 september 25 13 september 11 4 14 september 36 5 15 september 38 5 table 5. embroidery setup times (minutes). time (minutes) pieces time (minutes) pieces 15 150 22 300 7 100 7 100 10 200 8 100 16 200 7 100 table 4. embroidery setup times (minutes). top line exchange bottom line exchange glue and paste changing design line cutting/changing sides change line and sides changing frame 4,5 10 51 12 72 3 17 49 52 8 1,5 32 28,5 6 39 6 3 38 33 7 38 5 10 45 15 60 5 18 55 55 9 2,5 35 30,5 8 45 8 5 35 35 8 39 int. j. prod. manag. eng. (2019) 7(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho, c.p., gonçalves, l.w.n. and silva, m.b. 6 http://creativecommons.org/licenses/by-nc-nd/4.0/ based on the information described above the cvsm was drawn as shown in figure 2. as we can see in the map, the manager decides what will be produced based on their inventory in the shop and that is why we used the glasses icon, which means “go see scheduling”. there is no raw material supplier since the fabric is produced in the same company and goes directly to the first step of the production process. the first process block of the flow is spreading the fabric and cut in layers, after this they cut it in the exactly sizes and format that the shop required. then this “work in progress” stays about seven days until go to the stamp machine. the assembly will basically do the preparation to the embroidery that requires a specific assembly to be done. than cleaning and separating the products to be send to the sewing, which is outsourced. the number of employees is represented by the operator icon and if there is more than one it is written on the side. in the data block we can find the cycle time (ct) which is the time spent to finish each process. the output refers to the number of products produced at the end of one cycle time. the setup time is the time to prepare the machine when the product that is being produced is changed. the mr is the machine repair which means period of time that the machine is not working for some problem in each ct, a not programmed maintenance. in the bottom of the map there is a time line presenting the time spent for one product get ready. the values presented on the top are non-value added times such as setup time, mr time and stock time. the bottom values are value added times (cts). lead time (lt) can be found adding all the times, value added time (vat) adding only the bottom values. then efficiency of the cycle can be calculated by dividing vat by lt. for the current state an efficiency of 27.67% was found, which is not a high value. it was observed that some of the employees work in more than one process which can directly affect the productivity, for example the production stops when the truck is ready to be loading because the employees that are in production are responsible for the truck loading as well. there are a high number of stops for maintenance in the embroidery process due to busted line and needle exchange. a relatively high setup time is also observed. a high mr and setup time are also present in the stamping process figure 2. current state value stream map. int. j. prod. manag. eng. (2019) 7(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international value stream mapping as a lean manufacturing tool: a new account approach for cost saving in a textile company 7 http://creativecommons.org/licenses/by-nc-nd/4.0/ which the main reasons are defective parts in the machinery, piston with problems, foam and eucatex exchange. these reasons were all based on historical data provided. 6.2. future stream mapping main wastes could be detected during the design of the current state map as shown in the table 6 below. table 6. main wastes detected in the cvsm. type of waste process inventory between cutting and stamping waiting stamping waiting embroidery due to these actions were chosen to minimize these effects. the future state map was designed through some improvements in the manufacturing process for t-shirts production. analyzing the most representative wastes, an action plan was designed to reach the inventory reduction, which is presented as follows: kanban – the establishment of pull production concept should be conducted to the implementation of control of the finish goods inventory. the introduction of cards in the t-shirts store would be managing the master production schedule and as the consequence the work in process level would decrease. within the actions described above, the t-shirts future state map process is designed and figure 2 shows the results. beyond the changes with the implementation of kanban a reduction of 3 days in the inventory between cutting and stamping could be expected. after this action, the cycle efficiency increased to 40.1%. 6.3. lean account application to the studied company´s production system in the approach to apply lean account concepts, the company´s current and future stream mapping were considered, as well as, the following methodology to get the t-shirts production cost: 1. computing physical values on the current and the future value stream mapping; 2. evaluation of the capacity; 3. research of the costs involved in the both value stream mapping; figure 2. future state value stream map. int. j. prod. manag. eng. (2019) 7(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho, c.p., gonçalves, l.w.n. and silva, m.b. 8 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. accounting data collection; 5. final comparison between the costs before and after the improvements implementation to better understand the impact the implementation of lm tools had in the company’s liquidity, the authors evaluated the total thread in stock of november, before implementing lm tools, this can be seen in table 7. table 7. november thread stock. kg total value thread 75/34 0 8.90 thread 75/36 16765 8.90 149,208.50 thread 75/108 0 9.38 thread 108/36 0 8.17 thread 30/1 466 8.90 4,147.40 total thread in stock 153,355.90 in progress r$ 138,978.20 finished in stock r$ 661,067.16 raw in stock r$ 95,812.42 total stock 1,049,213.68 the top managers of the company implemented kanban during the month of december and to see whether the implementation of the lm tool had a positive or negative impact, the authors checked the total thread in stock at the end of december and compared with the previous month. table 8. december thread stock. kg total value thread 75/34 8874 8.90 78,978.60 thread 75/36 0 8.90 thread 75/108 0 9.38 thread 108/36 0 8.17 thread 30/1 466 8.90 4,147.40 total thread in stock 83,126.00 in progress r$ 111,675.31 finished in stock r$ 566,149.25 raw in stock r$ 53,647.78 total stock 814,598.34 the december final thread stock shows a decrease in the final goods inventory worth r$231,615.34, proving that the implementation of kanban and vsm had a positive impact in the liquidity of the company since all inventories (in progress, finished and raw) decreased after their implementation. 6. conclusion this article gave an understanding of how lm tools can be applied to identify wastes in a textile enterprise production flow. value stream mapping showed the opportunities in waste reduction mainly in the final goods inventory. the work also reported the combined application of two lm tools: vsm and kanban. to complement the investigation of this paper and the impact these tools have, both tools were implemented in the studied company and, in a short period of time, increased inventory turnover, impacting the operation profitability. the new finished goods inventory is lower than previously, with the same amount of assets, the company is, currently, more liquid since inventory is lower and profit higher (evening out the balance sheet). low inventory also means more free space, which increases business opportunities because the enterprise has room to add new machinery or expand production. the four hypothesis presented were validated by the results of this research work, but a limitation can be pointed because this study was conducted in one textile company only. as practical implication, the paper left a new guide for future account analysis regarded to the cost saving when lm tools are applied as well as opened a new field of approach for lm tools application as this case study, reducing the current researches gaps when it takes in consideration continuous manufacturing systems. the group which had conducted this work, strongly suggests for further research a new approach by application of this acquired knowledge in different textile companies as well as in others which operate under continuous manufacturing systems as paper and steel mills. references abdulmalek, f. a., rajgopa, j. (2007). analyzing the benefits of lean manufacturing and value stream mapping via simulation: a process sector case study, int. j. production economics, 107(1), 223-236. https://doi.org/10.1016/j.ijpe.2006.09.009 achanga, p., shehab, e., roy, r., nelder, g. (2005). critical success factors for lean implementation within smes, journal of manufacturing technology management, 17(4), 460-471. https://doi.org/10.1108/17410380610662889 int. j. prod. manag. eng. (2019) 7(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international value stream mapping as a lean manufacturing tool: a new account approach for cost saving in a textile company 9 https://doi.org/10.1016/j.ijpe.2006.09.009 https://doi.org/10.1108/17410380610662889 http://creativecommons.org/licenses/by-nc-nd/4.0/ amorim, a.h. (2011). competitividade internacional do complexo têxtil brasileiro no período 1998 a 2006. redige,vol.2. andrade, p.f, pereira, v.g., conte, e.g.d. (2015). value stream mapping and lean simulation: a case study in automotive company, international journal in advance technology, 85(1-4), 547-555. https://doi.org/10.1007/s00170-015-7972-7 barfield, j.t., raiborn, c.a., kinney, m.r. (1994).cost accounting: traditions and innovations. 4th ed. n.p.: south-western college pub. print. bhamu, j., sangwan, k.s. (2014). lean manufacturing: literature review and research issues. international journal of operations and production management, vol. 34 (7), 876-940. https://doi.org/10.1108/ijopm-08-2012-0315 bortolotti, t., boscari, s., danese, p. (2015). successful lean implementation: organizational culture and soft lean practices. international journal of production economics, 160, 182-201. https://doi.org/10.1016/j.ijpe.2014.10.013 boyle, t., scherrer-rathje, m., stuart, i. (2011). learning to be lean: the influence of external information sources in lean improvements. journal of manufacturing technology management, 28(5), 587-603. https://doi.org/10.1108/17410381111134455 braglia, m., carmignani, g., zammori, f. (2006). a new value stream mapping approach for complex production systems, international journal of production research, 44(18-19), 3929-3952. https://doi.org/10.1080/00207540600690545 callen, j.l., morel, m., fader, c. (2003). the profitability-risk tradeoff of just-in-time manufacturing technologies. managerial and decision economics, 24(5), 393-402. https://doi.org/10.1002/mde.1104 carnes, k., hedin, s. (2005). accounting for lean manufacturing: another missed opportunity?. management accounting quarterly, 7(1), 28-35. carvalho, c.p., gomes, f.m., n.silva, j.p., silva, m.b. (2017). lean manufacturing in continuous manufacturing systems: a literature review. international journal of research studies in science, engineering and technology, v.4, p. 1-13. cirino, s.r.a., gonçalves, h.s., queiroz, f.c.b.p., queiroz, j.v., hekis, h.r. (2013). sistema de produção enxuta: analisando as práticas adotadas em uma indústria têxtil paraibana. gepros. gestão da produção, operações e sistemas, bauru, ano 8, no 1, 9-21. chauhan, g. (2016). an analysis of the status of resource flexibility and lean manufacturing in a textile machinery manufacturing company. international journal of organization analysis, 24, 107-122. https://doi.org/10.1108/ijoa-11-2012-0625 chenhall, r.h. (1997). reliance on manufacturing performance measures, total quality management and organizational performance. management accounting research, 8, 187-206. https://doi.org/10.1006/mare.1996.0038 cooney, r. (2002). is “lean” a universal production system?. international journal of operations and production management, 22(10), 1130-1147. https://doi.org/10.1108/01443570210446342 cooper, r. (1996). activity-based management and the lean enterprise, journal of cost management, 69(10), 20-27. deflorin, p., rathje, m.s. (2012). challenges in the transformation to lean production from different manufacturing – process choices: a path-dependent perspective, international journal of production research, 50(14), 3956-3973. https://doi.org/10.1080/00207543.2011.613862 dhandapani, v., potter, a. and naim, m. (2004). applying lean thinking: a case study of an indian steel plant, international journal of logistics research and applications, 7(3), 239-250. https://doi.org/10.1080/13675560412331298491 easton, g.s., jarrell, s.l. (1998). the effects of total quality management on corporate performance: an empirical investigation. journal of business, 71(2), 253-307. https://doi.org/10.1086/209744 flanagan, r. (2008). the review of policing: final report. london: home office, 9-13. green, j.c., lee, j., kozman, t.a. (2010). managing lean manufacturing in material handling operations, international journal of production research, 48(10), 2975-2993. https://doi.org/10.1080/00207540902791819 gurumurthy, a., kodali, r. (2011). design of lean manufacturing systems using value stream mapping with simulation, journal of manufacturing technology management, 22(4), 444-473. https://doi.org/10.1108/17410381111126409 hedelind, m., jackson, m. (2011). how to improve the use of industrial robots in lean manufacturing systems, journal of manufacturing technology management, 22(7), 891-905. https://doi.org/10.1108/17410381111160951 hines, p., rich, n. (1997). the seven value stream mapping tools, international journal of operations and production, 17(1), 46-64. https://doi.org/10.1108/01443579710157989 hines, p., rich, n., esain, a. (1999). value stream mapping: a distribution industry application. benchmarking: an international journal, 6(1), 60-77. https://doi.org/10.1108/14635779910258157 hines, p., holweg, m., rich, n. (2004). learning to evolve, international journal of operations and production management, 24(10), 9941011. https://doi.org/10.1108/01443570410558049 hodge, g.l., ross, k.g., joines, j.a., thoney, k. (2011). adapting lean manufacturing principles to the textile industry, production planning and control, 22(3), 237-247. https://doi.org/10.1080/09537287.2010.498577 holweg, m. (2003). the three-day car challenge – investigating the inhibitors of responsive order fulfillment in new vehicle supply systems, international journal of logistics: research and applications, 6(3), 165-183. https://doi.org/10.1080/1367556031000123106 int. j. prod. manag. eng. (2019) 7(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho, c.p., gonçalves, l.w.n. and silva, m.b. 10 https://doi.org/10.1007/s00170-015-7972-7 https://doi.org/10.1108/ijopm-08-2012-0315 https://doi.org/10.1016/j.ijpe.2014.10.013 https://doi.org/10.1108/17410381111134455 https://doi.org/10.1080/00207540600690545 https://doi.org/10.1002/mde.1104 https://doi.org/10.1108/ijoa-11-2012-0625 https://doi.org/10.1006/mare.1996.0038 https://doi.org/10.1108/01443570210446342 https://doi.org/10.1080/00207543.2011.613862 https://doi.org/10.1080/13675560412331298491 https://doi.org/10.1086/209744 https://doi.org/10.1080/00207540902791819 https://doi.org/10.1108/17410381111126409 https://doi.org/10.1108/17410381111160951 https://doi.org/10.1108/01443579710157989 https://doi.org/10.1108/14635779910258157 https://doi.org/10.1108/01443570410558049 https://doi.org/10.1080/09537287.2010.498577 https://doi.org/10.1080/1367556031000123106 http://creativecommons.org/licenses/by-nc-nd/4.0/ hu, q., mason, r., williams, s., found, p. (2015). lean implementation within smes: a literature review. journal of manufacturing technology management, 26, 980-1012. https://doi.org/10.1108/jmtm-02-2014-0013 huntzinger, j.r. (2007). lean cost management: accounting for lean by establishing flow, j. ross publishing, inc., fort lauderda, fl. jasti, n., kodali, r. (2014).a literature review of empirical research methodology in lean manufacturing. international journal of operations and production management, 34(8), 1080-1122. https://doi.org/10.1108/ijopm-04-2012-0169 kaplan, r. s. 1990. the four-stage model of cost systems design. management accounting (february): 22-26. kaplan, r. s., norton, d. (1992). “the balanced scorecard: measures that drive performance.” harvard business review, 70, 71–79. karthi, s., devadasan, s.r., selvaraju, k., sivaram, n.m., sreenivasa, c.g. (2013). implementation of lean six sigma through iso 9001:2008 based qms: a case study in a textile mill, the journal of the textile institute, 104(10), 1089-1100. https://doi.org/10.1080/00405000.2013.774945 kennedy, f., brewer, p. (2005). lean accounting: what’s it all about? strategic finance, november, 27-34. krumwiede, k.r. 1998. abc: why it’s tried and how it succeeds. management accounting (april), 32-34, 36, 38. kuniyoshi, d.s. (2006). implementação da metodologia lean seis sigma em uma empresa do setor têxtil. dissertação de graduação. universidade de são paulo. leone, g.s.g. (1996). custos: planejamento, implantação e controle. 2. ed. são paulo: editora atlas. lopez, p.r.d.a., fortuny-santos, j., cuatrecasas-arbós, l. (2013). lean manufacturing: costing the value stream: industrial management and data systems: vol 113, no 5. emerald insight, web. 17 oct. 2016. lourenço jr, j. (2012). a produção enxuta em um sistema de fabricação contínuo:aplicação da simulação discreta estocástica na indústria de condutores elétricos. dissertação de doutorado. universidade estadual paulista-júlio de mesquita filho. guaratinguetá. mahapatra, s.s., mohanty, s.r. (2007). lean manufacturing in a continuous process industry: a case of study, journal of scientific and industrial research, 66, 19-27. martins, e. (2008). contabilidade de custo, atlas, sao paulo, 34, 102-105. maskell, b.h. (2000). “lean accounting for lean manufacturers”, manufacturing engineering, 125(6), 46-53. maskell, b.h., baggaley, b. (2003). practical lean accounting: a proven system for measuring and managing the lean enterprise. new york: productivity press. https://doi.org/10.1201/b11454 mcdonald t., van aken e.m., rentes, a.f. (2002). utilizing simulation to enhance value stream mapping: a manufacturing case application, international journal of logistics: res and appl. 5(2), 213-232. https://doi.org/10.1080/13675560210148696 milnitz, d., tubino, d.f. (2013). aplicação do método de mapeamento de fluxo de valor no setor de engenharia de uma empresa têxtil, exacta, 11(2), 199-212. monica, z. (2015). virtual modeling of components of a production system as the tool of lean engineering. iop conference series: materials science and engineering, 95, (6pp). oberhausen, c., plapper, p. (2015). value stream management in the” lean manufacturing laboratory”. procedia cirp, 32, 144-149. https://doi.org/10.1016/j.procir.2015.02.087 ohno, t. (1997). o sistema toyota de produção, além da produção em larga escala. porto alegre: bookman, 137. ohno, t. (1988). just in time for today and tomorrow. productivity press. onyeocha, c., khoury, j., geragthy, j. (2015), evaluating of multi-product: lean manufacturing systems with setup and erratic demand. computer and industrial engineering, 87, 465. https://doi.org/10.1016/j.cie.2015.05.029 palauro, r.l. (2014). lean manicuring: value stream mapping: map of the current situation and highlights for the problems identified in the line and production of pulplimeira, sp: [s. n.], 26 f. panwar, a., nepal, b. p., jain, r., rathore, a. p. s. (2015). on the adoption of lean manufacturing principles in process industries, production planning and control, 26:7, 564-587. https://doi.org/10.1080/09537287.2014.936532 panwar, a., jain, r., rathore, a. p. s. (2013). lean implementation in indian process industries – some empirical evidence, journal of manufacturing technology management, 26(1), 131-160. https://doi.org/10.1108/jmtm-05-2013-0049 powell, d., riezebos, j., strandhagen, o. (2013). lean production and erp systems in small-medium –sized enterprises: erp support for pull production, international journal of production research, 51(2), 395-409. https://doi.org/10.1080/00207543.2011.645954 ratnayake, r.m.c., chaudry, m.o. (2015). performance improvement of oil and gas industry via lean concept: a case study from valves requisition. 2015 ieee international conference on industrial engineering and engineering management (ieem). https://doi.org/10.1109/ieem.2015.7385892 rother, m., shook, j. (1999). learning to see: value stream mapping to add value and eliminate muda, the lean enterprise institute, inc., brookline, ma. rother, m., shook, j. (2003). creating continuous flow, the lean enterprise institute: massachusetts. schimidtke, d., heiser, u., hinrichsen, o. (2014). a simulation enhanced value stream mapping approach for optimization of complex production environments, international journal of production research, 52:20, 6146-6160. https://doi.org/10.1080/00207543.2014.917770 int. j. prod. manag. eng. (2019) 7(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international value stream mapping as a lean manufacturing tool: a new account approach for cost saving in a textile company 11 https://doi.org/10.1108/jmtm-02-2014-0013 https://doi.org/10.1108/ijopm-04-2012-0169 https://doi.org/10.1080/00405000.2013.774945 https://doi.org/10.1201/b11454 https://doi.org/10.1080/13675560210148696 https://doi.org/10.1016/j.procir.2015.02.087 https://doi.org/10.1016/j.cie.2015.05.029 https://doi.org/10.1080/09537287.2014.936532 https://doi.org/10.1108/jmtm-05-2013-0049 https://doi.org/10.1080/00207543.2011.645954 https://doi.org/10.1109/ieem.2015.7385892 https://doi.org/10.1080/00207543.2014.917770 http://creativecommons.org/licenses/by-nc-nd/4.0/ serrano, i., ochoa, c., castro, r. (2008). evaluation of value stream mapping in manufacturing redesign, international journal of production research, 46(16), 4409-4430. https://doi.org/10.1080/00207540601182302 seth, d., seth, n., goel, d. (2008). application of value stream mapping (vsm) for minimization of wastes in the processing side of supply chain of cottonseed oil industry in indian context. journal of manufacturing technology management , 19(4), 529-550. https://doi.org/10.1108/17410380810869950 soliman, f. (1998). optimum level of process mapping and least cost business process re-engineering. international journal of operations and production management, 18(9/10), 810-816. https://doi.org/10.1108/01443579810225469 susilawati, a., tan, j., bell, d., sarwar, m. (2015). fuzzy logic based method to measure degree of lean activity in manufacturing industry. journal of manufacturing systems, 34, 1-11. https://doi.org/10.1016/j.jmsy.2014.09.007 sutari, o. (2015). process improvement using lean principles on the manufacturing of wind turbine componentsa case study. materials today: proceedings, 2, 3429-3437. https://doi.org/10.1016/j.matpr.2015.07.318 upadhye, n., deshmukh, s.g., garg, s. (2010). “lean manufacturing in biscuit manufacturing plant: a case.” international journal of advanced operations management, 2(1/2): 108-139. https://doi.org/10.1504/ijaom.2010.034589 vanderbeck, e.j. (2010). principles of cost accounting. mason, oh: thomson/south-western, 22-38. print. vinoth, s., arvind, k.r., somanaathan, m. (2010). application of value stream mapping in an indian camshaft manufacturing organization, journal of manufacturing technology management, 21(7), 888-900. https://doi.org/10.1108/17410381011077973 walther, l.m. (2013). financial accounting. san bernardino, ca: principles of accounting.com, 12-18. print. werkema, m. (2011). lean seis sigma: introdução às ferramentas do lean manufacturing -2. edrio de janeiro: elsevier. womack, j.p., jones, d.t., ross, d. (1990). the machine that changed the world, macmillan publishing company, canada. womack, j.p., jones, d.t. (1996). lean thinking: banish waste and create wealth in your corporation, simon and schuster, new york, ny. yang, t., kuo, y., su, c.t., hou, c.l. (2015). lean production system design for fishing net manufacturing using lean principles and simulation optimization, journal of manufacturing systems, 34, 66-73. https://doi.org/10.1016/j.jmsy.2014.11.010 int. j. prod. manag. eng. (2019) 7(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international carvalho, c.p., gonçalves, l.w.n. and silva, m.b. 12 https://doi.org/10.1080/00207540601182302 https://doi.org/10.1108/17410380810869950 https://doi.org/10.1108/01443579810225469 https://doi.org/10.1016/j.jmsy.2014.09.007 https://doi.org/10.1016/j.matpr.2015.07.318 https://doi.org/10.1504/ijaom.2010.034589 https://doi.org/10.1108/17410381011077973 https://doi.org/10.1016/j.jmsy.2014.11.010 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2016.4418 received 2015-12-09 accepted: 2015-12-16 a cloud platform to support collaboration in supply networks beatriz andresa*, raquel sanchisb and raul polerc centro de investigación en gestión e ingeniería de la producción (cigip) escuela politécnica superior de alcoy, universitat politècnica de valència (upv). calle alarcón, 03801 alcoy (españa). a* bandres@cigip.upv.es b rsanchis@cigip.upv.es c rpoler@cigip.upv.es abstract: collaboration is a trend in supply networks management, based on the jointly planning, coordination and integration of processes, participating all network entities. due to the current characteristics of uncertainty in the markets and economic crisis, there is a need to encourage collaboration tools to reduce costs and increase trust and accountability to market requirements. this study presents an overview of the research carried out in the h2020 european project: cloud collaborative manufacturing networks (c2net), which is directed towards the development a cloud platform that consist of, optimization tools, collaboration tools to support and agile management of the network. the collaborative cloud platform allows to collect real time information coming from real-world resources and considering all the actors involved in the process. the collaborative cloud provides real time data gathered from the entire network partners in order to improve their decision-making processes. key words: cloud platform, collaborative networks, collaborative processes, plans optimization. 1. introduction globalization, information and communication technologies (ict) and innovation processes have revolutionised in the recent decades the organisation of value chains. although the small and medium enterprises (smes) represent 99% (european comission, 2013) of european enterprises, their decadence has increased in recent years due to the current economic situation. the european commission notes that collaboration among the network enterprises may be reasonable to prevent the decline of traditional specialisation of smes strategy (european comission, 2012). in the global economy, there is an increasing interest in new organizational structures to flexible enough to respond to market changes and at the same time to perform collaborative projects. enterprise networks have been identified as fundamental instruments to the implementation of the 2020 strategy of the european union (eu). the eu initiative “innovation union” and “an integrated industrial policy for the globalization era” (european commission, 2014), specifically refers to enterprise networks as key tools for achieving this strategy. currently, industry competitiveness and growth leads towards innovative high performance industrial systems and agile companies interconnected through the creation and consolidation of collaborative networks. traditional supply chains are based on centralized decisionmaking, where most of the players must adapt to the constraints defined by a few models. real-world experiences show that when making decisions is not collaboratively performed lower performances are achieved, and in current highly dynamic markets, this generates large inefficiencies in the performance of the entire network. the key factor to address these emerging challenges is the enterprises collaboration. the collaborative manufacturing ensures constant feedback circuit, unbroken communication between product designers, engineers, manufacturing facilities and customers. in collaborative supply networks, manufacturers can offer value-added services (e.g. maintenance, updating...) or even sell their products “as a service”. the remotely management of services helps to int. j. prod. manag. eng. (2016) 4(1), 5-13creative commons attribution-noncommercial-noderivatives 4.0 international 5 http://dx.doi.org/10.4995/ijpme.2016.4418 http://creativecommons.org/licenses/by-nc-nd/4.0/ improve equipment uptime, reduce service costs (e.g. travel expenses...), increasing the efficiency of services and accelerating the processes innovation (e.g., remote update software of the devices). motivated by this situation, this paper describes the planned work to be carry out along the h2020 european project: cloud collaborative manufacturing networks (c2net). to this end, this paper is organised as follows: in section 2 a state of the art is presented, centring its attention on the most relevant research areas related with the collaborative context. section 3 focuses on the description of c2net european project, including the objectives, expected results and expected impacts derived from its exploitation. in section 4 a more detailed description of the c2net contribution regarding the optimization and simulation algorithms for collaborative manufacturing and logistics processes (cmlp) is proposed. finally, section 5 provides the main conclusions and future research work. 2. state of the art globalisation and competitive environments, in which enterprises are embedded, trigger the appeareance of new patterns of operation among manufacturing and service enterprises operation. over the last decade, collaboration has emerged as new way on how enterprises make their business, due to the competitive advantages associated. camarinhamatos and afsarmanesh (2005) established the bases of collaborative networks (cn) discipline and proposed a common definition: cn consist of a variety of heterogeneous autonomous entities, geographically distributed, in which participants collaborate to achieve a common goal and base their interactions through computer networks. in order to face dynamic and turbulent environments enterprises participation in cn allows them to increase their agility, responsiveness and resilience. particularly, smes are commonly characterised by having limited resources and capabilities to efficiently establish collaborative relationships (matopoulus et al., 2007). so that, there is a gap in terms of smes affordable tools (in term of cost and usability) to help them to support their decisions and processes in the collaborative context. in order to face potential barriers that can appear when smes participate in collaborative processes, joint efforts must be performed to achieve the desired collaborative scenarios. the restructuration of smes internal operations, the exchange information, the information systems interoperability, the coordination of smes production processes and planning processes, the alignment of enterprises strategies and goals, the achievement of suitable levels of trust, the agreements in practices, and the alignment of values (bititci et al., 2007; macedo et al., 2010; andres and poler, 2014) are part of the tasks to be performed by enterprises if they are willing to achieve sustainable and stable collaborative relationships. in the light of this, a set of 13 research areas are identified of interest in the collaborative networks context to support enterprises in the participation of collaborative process. these areas are briefly described next as part of the literature review research in the cn context: agile supply chain. according to lee (2004), an agile supply chain is a set of partners that responding quickly to short-term changes in demand (or supply) and handling external disruptions smoothly. recently various initiatives have appeared for analysing and supporting the deployment of agility in supply chain. such initiatives are soa and web service based (zhang et al., 2012). cloud computing. cloud computing is defined as the on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction (mell and grance, 2011). automation in cloud systems makes it possible to manage the data centres so that they deliver portions of themselves (in the form of virtual machines and networks) (amazon ec2, 2012). in the light of this, cloud services must, guarantee what is known as a service level agreement (sla) (buyya et al., 2009). enterprise applications. aim to facilitate the management of the different areas of an enterprise. ranging from micro mechatronics systems (mems), mes and intelligent manufacturing systems (ims) to erp, aps and customer relationship management (crm) (ollero et al., 2003). enterprise interoperability. interoperability is an essential problem of sharing information and exchanging services. it goes far beyond the simple technical problems of computer hardware and software, but encompasses the broad but precise identification of barriers not only concerning data and service but also process and business as well 6 int. j. prod. manag. eng. (2016) 4(1), 5-13 creative commons attribution-noncommercial-noderivatives 4.0 international andres, b., sanchis, r. and poler, r. http://creativecommons.org/licenses/by-nc-nd/4.0/ (chen et al., 2006). different european actions have been carried out in order to support research activities enabling enterprises to seamlessly interoperate: athena fp6-ist 507849; interop-noe fp6ist 508011 event driven architecture (eda). eda is in charge of managing event exchanges between processes, people and machines, filtering and applying business rules to detect relevant events or combinations of events. for example, two events, which are not seen as risks or opportunities when viewed separately, may have a different meaning if they are considered together (and so they create a complex event); this is called complex event processing (cep) (marterer et al., 2012). everything as a service (eaas). provides an increasing number of services through internet under the cloud computing paradigm instead of providing them locally (lenk et al., 2009). the “everything” term comprises from infrastructure, middleware and platforms, software and compound applications (for different domains) human intelligence, and any combination of them. eaas is focused on making cloud-based services, both creating new services or adapting previous ones to the cloud paradigm in such way they take as much advantage as possible of the whole potential provided by cloud computing. internet of everything (ioe). brings together people, process, data, and things to make networked connections more relevant and valuable, turning information into actions that create new capabilities, richer experiences, and business opportunities (cisco, 2013). five key factors are considered in ioe: (i) assets utilization through reduced costs and more effective decision making, (ii) employee productivity empowering people and increase efficiency, (iii) supply chain and logistics, eliminating waste and idle times, (iv) know customer experience and build better relationships, and (v) innovation, to create and expand new markets and services, including productservices concept (fines, 2013). manufacturing processes management. globalisation of businesses has increased enterprises competition. as a consequence companies have to deal with rapid changes in their business or face extinction (banerjee, 2000). in order to face up to these new requirements, companies seek to redefine and improve various aspects of their manufacturing process management (mpm). mes provides shop floor control and simultaneously manufacturing feedback (garetti et al., 2007). therefore, mpm is a strategy that supports formal communication between engineering and production (fortin et al., 2007). manufacturing processes optimization. research in optimization techniques has been a very active field in the last decades. one of the most successful application areas has been the planning and scheduling of operational manufacturing processes, trying to optimize cost, resource allocation, delivery time, etc. different mathematical programming approaches in a deterministic context are proposed in the literature (alemany et al., 2010). intelligent modelling and heuristic modelling approaches have also been proposed as bridging techniques between the problems that theoretical optimization techniques can handle and real world problems. hernandez et al. (2014) developed multi-agent negotiation based algorithms. moreover, fuzzy mathematical programming models (mula et al. 2010a) have been developed for production planning problems under uncertainty. identify the appropriate technique for the different optimisation process is a controversial discussion and is the source of much research. reallife enterprise optimization problems, as for example planning and scheduling, often need both linear and non-linear constraints and may involve hundreds/ thousands of (real/integer/binary) variables and constraints, and are difficult to tackle with genetic algorithms-like techniques. mobile computing. with the emergence of mobile computing and the rapid developments on mobile devices and communication technologies, the possibility of providing information to users when and where is needed is now possible. mobile computing poses a series of unique challenges for the development of mobile information systems (satyanarayanan, 1996) where personalized information is shown to people with different necessities and profiles. ontologies. refers to an explicit and formal specification of a shared conceptualization (gruber, 1993). an ontology represents knowledge of some domain of interest as a set of concepts, their definitions and their inter-relationships; it is shareable and can be understood by a computer. sharing common understanding of the structure of information among people or software agents is one of the more common goals in developing ontologies, as well as, making explicit domain assumptions in order to change these assumptions easily if the 7int. j. prod. manag. eng. (2016) 4(1), 5-13creative commons attribution-noncommercial-noderivatives 4.0 international a cloud platform to support collaboration in supply networks http://creativecommons.org/licenses/by-nc-nd/4.0/ knowledge about the domain changes. different ontologies have been developed since then that uses specific ontology languages (e.g. owl, cgs, oil, and daml) and editor (e.g. protégé). open source technologies/software (oss). the term is traced back to the 1950s and 1960s, but it was richard stallman, who founded the free software foundation that provided the conceptual foundation for open source software. the oss development approach has become a remarkable option to consider for cost-efficient, high quality software development. utilizing oss (merilinna and matinlassi, 2006) has allowed its utilization as part of an in-house software application. main benefits for adopting oss are (dixon, 2004): easy adoption, lower cost, motivation, control and flexibility. some of the main institutions on the oss are: the free software foundation (fsf), the open source initiative (osi), and open source development labs (odsl). the computational infrastructure for operations research (coin-or) project is an initiative to spur the development of open-source software for the operations research community. orchestration / choreography of workflows. orchestration (i) refers to an executable business process and (ii) represents the service composition from one partner’s perspective in the case of a collaborative process (peltz, 2003). choreography refers to a viewpoint of the services composition where the interactions between the services involved into collaborative processes are seen from a global perspective (barros et al. 2005). choreography is used to execute several business processes. various languages have emerged in the last years in order to model the services choreography: (i) interaction modelling: the logic of the choreography is specified as a workflow where the activities represent the messages exchanged between the parties, e.g. wscdl (barros et al. 2005) and (ii) interconnected interfaces modelling: the logic of the choreography is shared across the parties through their roles, which are connected using message flows, channels or equivalent concepts, e.g. bpmn 2.0 (omg, 2015). 3. cloud collaborative manufacturing networks: c2net project 3.1. objective currently, the european smes do not have access to advanced management systems and collaborative tools because of their limited resources (european commission 2005). value chains formed by smes are distributed and dependent on information and complex materials flows that require new approaches to reduce the complexity of manufacturing management systems. in this context, ubiquitous tools are needed to support collaboration between different entities in the value chain and offer advanced algorithms to achieve global and local optimization of manufacturing processes and respond more quickly and efficiently unforeseen changes. the main objective of c2net is building a new architecture in the cloud to provide smes, affordable tools (in terms of cost and ease of use) to help overcome the current economic crisis, improving competitiveness in the economy world. therefore, c2net objective is based on the creation of cloud tools to support optimization of manufacturing networks composed mainly of smes and their logistic assets through demand management, production and supply plans, considering the collaborative network perspective. 3.2. expected results the c2net project will generate a cloud architecture composed by the cloud platform (c2net cpl), the data collection framework (c2net dcf), the optimizer (c2net opt) and the collaboration tools (c2net cot) (see figure 1). the data collection framework (c2net dcf) to provide software components and hardware devices for iot-based continuous data collection from supply network resources. this supports the collaborative manufacturing functionality while taking advantage of cloud environments, which can enable solutions that are highly scalable, available and fault-tolerant. the optimizer (c2net opt) to support manufacturing networks in the optimization of manufacturing and logistics assets by the collaborative computation of production plans, replenishment plans and delivery plans in order to achieve shorter delivery times, better speed and consistency of schedules, higher use of productive resources and energy savings. the collaboration tools (c2net cot) for providing the collaborative manufacturing network platform with a set of tools in charge of managing the agility of the collaborative processes (lauras et al.,. 2015). 8 int. j. prod. manag. eng. (2016) 4(1), 5-13 creative commons attribution-noncommercial-noderivatives 4.0 international andres, b., sanchis, r. and poler, r. http://creativecommons.org/licenses/by-nc-nd/4.0/ the cloud platform (c2net cpl) to integrate the data module, the optimizers and the collaborative tools in the cloud and allow the access to process optimization resources to all the participants in the value chain to support their decisions and process enhancement. it will provide the base for the integration of the different modules for generating a collaborative working environment for manufacturing network partners. 3.3. expected impacts among the major challenges that manufacturing companies face today are the growing complexity of their processes and supply networks, cost pressures, growing user and customer expectations for quality, speed, and custom products, and worker safety and assistance. manufacturing is evolving from being perceived as a production-centred operation to a human-centred business with a greater emphasis on workers, suppliers and customers being in-the-loop (actionplant, 2014). manufacturing networks performance can be significantly improved through more harmonious and equitable peer-to-peer inter-enterprise relationships, conforming a collaborative decision making model and providing major benefits mainly in terms of: enhanced overall competitiveness, innovation and adaptability in today and tomorrow’s enterprise partnership scenario. cross-country and inter-enterprise interchanges, building networked enterprises that are supported by stable relationship schemas and modern cooperation & co-ordination business paradigms. figure 1. overview of c2net cloud architecture. 9int. j. prod. manag. eng. (2016) 4(1), 5-13creative commons attribution-noncommercial-noderivatives 4.0 international a cloud platform to support collaboration in supply networks http://creativecommons.org/licenses/by-nc-nd/4.0/ cost reduction, through overall optimisation and elimination of inefficiencies of processes, stocks, flows, plans, etc. companies’ human resources improved quality of work and skills, through improved knowledge management and dissemination, better understanding of dynamics and flows, and clearer definition of roles and responsibilities. end consumers’ advantages, mainly in terms of diminishment of products time-to-market and costs. smes empowerment and enhanced accessibility to networked enterprises. optimisation of materials, wastes and energy consumption based on more rational and homogeneous production and supply plans, stocks and workforce balance. the european industry needs advanced methods and tools to address economic and risk assessments in order to support complex decision-making as the management of co-evolution of products-services and the related production systems, the evaluation of alternative configurations of the network of actors involved in the global supply chain, or integration of new technologies in the factory. best-in-class companies versus others are up to 1.78 times as likely to have ability to analyse current level of supply chain risk exposure to have online visibility into supply chain disruptions and to have online trading partner collaboration and enablement. 4. collaborative plans optimization c2net optimizer (c2net opt) supports manufacturing networks in the optimization of manufacturing and logistics assets by the collaborative computation of production plans, replenishment plans and delivery plans in order to achieve shorter delivery times, better speed and consistency of schedules, higher use of productive resources and energy savings. the main objective of c2net is to develop tools to support manufacturing networks in the optimization of manufacturing and logistics assets by the collaborative computation of production plans, replenishment plans and delivery plans in order to achieve shorter delivery times, better speed and consistency of schedules, higher use of productive resources and energy savings. c2net opt will incorporate a set of highly specialized optimization algorithms to be applied depending on the characteristics of each unexpected event, e.g., lack of products at a point of sale, a machine breakdown or a delay on a component arrival. using these algorithms together with some business rules agreed between the companies in the supply chain (c2net will incorporate a business rules engine), c2net optimizer will provide a new production plan whose main objective will be to optimize the available capacity of manufacturing assets. in order to design the c2net opt four steps are identified: (i) the creation of a taxonomy to classify the collaborative plans, (ii) the selection of collaborative plans to specify and develop optimization and simulation algorithms, (iii) the validation of optimization and simulation algorithms developed and (iv) the computational implementation of the algorithms. this steps are outlined next. the creation of a taxonomy of optimization and simulation solutions for manufacturing and logistics processes is provided from the (i) analysis of the collaborative manufacturing and logistics processes identified in each of the industrial pilots, and (ii) review of current soa of optimization algorithms and simulation procedures used to solve related manufacturing processes problems, especially in the frame of collaborative processes, considering source (s), make (m) and delivery (d) plans and their combination (e.g. source & make, make & delivery and source & make & delivery). as a result of this activity it will have a deep understanding of the different optimization and simulation approaches existing in the literature and existing tools as the main input to address the development of new optimization and simulation algorithms to support collaborative manufacturing and logistics processes optimization. collaborative problems are selected from industrial scenarios and their requirements analysis in order to specify and develop optimization and simulation algorithms for collaborative manufacturing and logistics processes (cmlp). taking into account the taxonomy created and the existing optimization and simulation tools, the specification and development of new optimization and simulation algorithms and the adaptation of existing algorithms for supporting collaboration between enterprises will be made. the proposed algorithms will be classified concerning its calculation mechanism, which is related with the calculation time and the proximity to the optimum: 10 int. j. prod. manag. eng. (2016) 4(1), 5-13 creative commons attribution-noncommercial-noderivatives 4.0 international andres, b., sanchis, r. and poler, r. http://creativecommons.org/licenses/by-nc-nd/4.0/ optimiser algorithms (oa) employ a technique that ensures to find the optimum solution. these algorithms use to be the slowest ones and, for some kind of problems, the needed time to find the optimum is unacceptable. some examples of oa are: branch and bound (land and doig, 1960), dynamic programming (giegerich et al., 2004), lomnicki (lomnicki, 1965) or simplex (dantzig et al.,1955 ). heuristic algorithms (ah) employ a practical methodology not guaranteed to be optimal or perfect, but sufficient for the immediate goals. heuristic algorithms use to be the quickest ones and normally find one solution (it can not ensure optimum). some examples of ah are: decomposition & aggregation (pervozvanskii and gaitsgori, 2013), greedy (luke, 2013), minimum spanning tree (horowitz and sahni, 1978), nearest neighbour (cover and hart, 1967) or lagrangian (hestenes, 1969). metaheuristic algorithms (am) consist of higherlevel procedures designed to find, generate, or select a heuristic (partial search algorithm) that may provide a sufficiently good solution. meta-heuristic algorithms use to implement random search and can find several solutions (it can not ensure optimum) using a search strategy which needs a stop rule (it can be time). some examples of am are: colony optimization, evolutionary computation, genetic algorithm, iterated local search, neural networks, scatter search, simulated annealing, tabu search, variable neighbourhood search (luke, 2013). matheuristic algorithms (aa): is an optimisation algorithm made by the interoperation of metaheuristics and optimiser techniques. matheuristic can find optimum solutions faster than some optimiser techniques. some examples of aa are: crossentropy progressive hedging (rubinstein, 2001), hybrid reactive grasp (feo and resende, 1989) or steiner tree (garey and johnson, 1977). the optimization and simulation algorithms developed will be validated for cmlp, running test suites of on the industrial pilots. a collection of real collaborative processes instances with their solutions representing production, replenishment and delivery plans will be prepared as a key input to validate the algorithms. after the validation, the optimization and simulation algorithms developed will be implemented in the “processes optimization of manufacturing assets” (poma) bricks, which will be integrated within the c2net cloud architecture. the optimization and simulations algorithms to solve collaborative production, replenishment and delivery plans will be offered in an optimization as a service mode to the firms collaborating in a manufacturing network. poma bricks will include tools to collaboratively calculate production plans (amount of items to be produced), replenishment plans (amount of items to be ordered) and delivery plans (amount of items to be ordered) per periods of a horizon to every actor of the supply chain involved in the collaboration, from a holistic point of view to reach global optimization of manufacturing and logistics assets. poma bricks will be also validated with the industrial pilots. 5. conclusions the c2net project provides a scalable real-time architecture and a cloud platform that will gather different software to support the different network enterprises to: (i) manage the complexity and data security of all the network enterprises; (ii) store and share product data, processes and logistics; (iii) optimize manufacturing assets through collaborative computing production plans; (iv) optimizing logistics assets through efficient delivery plans, and (v) make the complete set information of the network is available on any mobile device (pc, tablets, smartphones...) to support decision makers in the tasks of control and visualization, so that they can share the necessary information and data to efficiently collaborate. future research lines led to implement the steps described, with regards the optimizer (c2net opt) and propose a wide variety of optimization and simulation algorithms. optimisation algorithms will not only be focused on the local scope of enterprises, but also will to support the network enterprises on computing from a collaborative and automated way the plans performed in the value chain, including source, make and delivery plans. collaborative planning, besides being supported by the c2net opt module, will be assisted by the proposals developed in the collaboration module, c2net cot. the tasks of supervision, detection, adaptation and assessment, developed in c2net cot will facilitate the coordination between the network partners. each of the c2net modules will be supported by the cloud platform (c2net cpl). the cloud 11int. j. prod. manag. eng. (2016) 4(1), 5-13creative commons attribution-noncommercial-noderivatives 4.0 international a cloud platform to support collaboration in supply networks http://creativecommons.org/licenses/by-nc-nd/4.0/ collaborative platform will allow smes, with less accessible capabilities and resources, to have access to sophisticated tools in order to deal with collaboration and facilitate the collaborative decision-making performed at both, enterprises and network level. acknowledgements the research leading to these results has received funding from european community’s h2020 programme (h2020/2014-2020) under grant agreement n°636909, “cloud collaborative manufacturing networks (c2net)”. references actionplant. (2014). actionplant the european ict forum for factories of the future. http://cordis.europa.eu/project/rcn/95333_es.html retrieved december 2015. alemany, m.m.e., boj, j.j., mula, j., lario, f.c. (2010). mathematical programming model for centralised master planning in ceramic tile supply chains. international journal of production research, 48(17): 5053-5074. doi:10.1080/00207540903055701 amazon ec2. (2015). amazon web services. https://aws.amazon.com/es/ec2/ retrieved december 2015. andres, b., poler, r. (2014). computing the strategies alignment in collaborative networks. in k. mertins, f. bénaben, r. poler, & j.-p. bourrières (eds.), enterprise interoperability vi (pp. 29–40). cham: springer international publishing. doi:10.1007/978-3-319-04948-9 banerjee, s. k. (2000). developing manufacturing management strategies: influence of technology and other is-sues. international journal of production economics, 64(1): 79-90. doi:10.1016/s0925-5273(99)00046-8 barros, a., dumas, m., oaks, p. (2005). a critical overview of the web services choreography description language. bptrends newsletter, 3: 1-24. bititci, u., turnera, t., mackaya, d., kearneyc, d., parunga, j., waltersb, d. (2007). managing synergy in collaborative enterprises. production planning & control: the management of operations, 18(6): 454–465. doi:10.1080/09537280701494990 buyya, r., yeo, c.s., venugopal, s., broberg, j., brandic, i. (2009). cloud computing and emerging it platforms: vision, hype, and reality for delivering computing as the 5th utility. future generation computer systems, 25(6): 599-616. chen, d., vernadat, f.b. (2003). enterprise interoperability: a standardisation view. in enterprise inter-and intra-organizational integration (pp. 273–282). springer us. doi:10.1007/978-0-387-35621-1_28 cisco. (2013). cisco systems, inc. 2013 annual report. https://www.cisco.com/web/about/ac49/ac20/ac19/ar2013/docs/2013_annual_ report.pdf retrieved december 2015. coin-or. (2012). computational infrastructure for operations research. http://www.coin-or.org/ retrieved december 2015. cover, t.m., hart, p.e. (1967). nearest neighbor pattern classification. ieee transactions, information theory, 13(1): 21–27. doi:10.1109/ tit.1967.1053964 dantzig, g.b., orden, a., wolfe, p. (1955). the generalized simplex method for minimizing a linear form under linear inequality restraints. pacific journal of mathematics, 5(2): 183-195. doi:10.2140/pjm.1955.5.183 dixon, r. (2004). open source software law. artech house. european commission. (2014). an integrated industrial policy for the globalisation era. putting competitiveness and sustainability at centre stage. http://eur-lex.europa.eu/legal-content/en/txt/?uri=uriserv%3aet0005, retrieved december 2015. european commission. (2005). the new sme definition: user guide and model declaration. https://ec.europa.eu/digital-agenda/en/news/ new-sme-definition-user-guide-and-model-declaration, retrieved december 2015. european commission. enterprise and industry. (2012). european competitiveness report 2012. reaping the benefits of globalization. http://wbc-inco.net/object/document/11137# retrieved december 2015. european commission. enterprise and industry. (2013). fact and figures about the eu´s small and medium enterprise (sme). http:// ec.europa.eu/enterprise/index_en.htm, retrieved december 2015. feo, t.a., resende, m.g. (1989). a probabilistic heuristic for a computationally difficult set covering problem. operations research letters, 8(2): 67-71. doi:10.1016/0167-6377(89)90002-3 fines. (2013). future internet enterprise systems. http://cordis.europa.eu/fp7/ict/enet/ei_en.html retrieved december 2015. fortin, c., huet, g. (2007) manufacturing process management: iterative synchronisation of engineering data with manufacturing realities. international journal of product development, 4(3-4): 280-295. doi:10.1504/ijpd.2007.012496 garetti, m., macchi, m., terzi s. (2007). product lifecycle management: state-of-the-art, trends and challenges. advanced manufacturing. an ict and systems perspective. taylor & francis, london: 37-49. doi:10.1201/9781439828328.ch4 garey, m.r., johnson, d.s. (1977). the rectilinear steiner tree problem is np-complete. siam journal on ap-plied mathematics, 32(4): 826-834. doi:10.1137/0132071 giegerich, r., meyer, c., steffen, p. (2004). a discipline of dynamic programming over sequence data. science of computer programming, 51(3): 215-263. doi:10.1016/j.scico.2003.12.005 12 int. j. prod. manag. eng. (2016) 4(1), 5-13 creative commons attribution-noncommercial-noderivatives 4.0 international andres, b., sanchis, r. and poler, r. http://cordis.europa.eu/project/rcn/95333_es.html http://dx.doi.org/10.1080/00207540903055701 https://aws.amazon.com/es/ec2/ http://dx.doi.org/10.1007/978-3-319-04948-9 http://dx.doi.org/10.1016/s0925-5273(99)00046-8 http://dx.doi.org/10.1080/09537280701494990 http://dx.doi.org/10.1007/978-0-387-35621-1_28 https://www.cisco.com/web/about/ac49/ac20/ac19/ar2013/docs/2013_annual_report.pdf https://www.cisco.com/web/about/ac49/ac20/ac19/ar2013/docs/2013_annual_report.pdf http://www.coin-or.org http://dx.doi.org/10.1109/tit.1967.1053964 http://dx.doi.org/10.1109/tit.1967.1053964 http://dx.doi.org/10.2140/pjm.1955.5.183 http://eur-lex.europa.eu/legal-content/en/txt/?uri=uriserv%3aet0005 https://ec.europa.eu/digital-agenda/en/news/new-sme-definition-user-guide-and-model-declaration https://ec.europa.eu/digital-agenda/en/news/new-sme-definition-user-guide-and-model-declaration http://wbc-inco.net/object/document/11137# http://ec.europa.eu/enterprise/index_en.htm http://ec.europa.eu/enterprise/index_en.htm http://dx.doi.org/10.1016/0167-6377(89)90002-3 http://cordis.europa.eu/fp7/ict/enet/ei_en.html http://dx.doi.org/10.1504/ijpd.2007.012496 http://dx.doi.org/10.1201/9781439828328.ch4 http://dx.doi.org/10.1137/0132071 http://dx.doi.org/10.1016/j.scico.2003.12.005 http://creativecommons.org/licenses/by-nc-nd/4.0/ gruber, t.r. (1993). a translation approach to portable ontology specifications. knowledge acquisition, 5(2): 199-220. http://dx.doi. org/10.1006/knac.1993.1008 hernández, j.e., mula, j., poler, r., lyons, a.c. (2014). collaborative planning in multi-tier supply chains sup-ported by a negotiation-based mechanism and multi-agent system. group decision and negotiation, 23(2): 235-269. doi:10.1007/s10726-013-9358-2 hestenes, m.r. (1969). multiplier and gradient methods. journal of optimization theory and applications, 4(5): 303-320. http://dx.doi. org/10.1007/bf00927673 horowitz, e., sahni, s. (1978). fundamentals of computer algorithms. potomac, m.d: computer science press. land, a.h., doig, a.g. (1960). an automatic method of solving discrete programming problems. econometrica, 28(3): 497-520. doi:10.2307/1910129. doi:10.2307/1910129 lauras, m., lamonthe, j. benaben, f., andres, b., poler, r. (2015). towards an agile and collaborative platform for managing supply chain uncertainties. international federation for information processing. iwei 2015, lnbip 213, 64-72. lee, h.l. (2004). the triple-a supply chain. harvard business review, 82(10): 102-113. lenk, a., klems, m., nimis, j., tai, s., sandholm, t. (2009). what’s inside the cloud? an architectural map of the cloud landscape. in proceedings of the 2009 icse workshop on software engineering challenges of cloud computing (pp. 23-31). ieee computer society. doi:10.1109/cloud.2009.5071529 lomnicki, z.a. (1965). a branch-and-bound algorithm for the exact solution of the three-machine scheduling problem. operations research society, 16: 89-100. doi:10.1057/jors.1965.7 luke, s. (2013). essentials of metaheuristics, lulu, second edition. https://cs.gmu.edu/~sean/book/metaheuristics/essentials.pdf retrieved december 2015. macedo, p., abreu, a., camarinha-matos, l.m. (2010). a method to analyse the alignment of core values in collaborative networked organisations. production planning & control, 21(2): 145-159. doi:10.1080/09537280903441930 marterer, r., moi, m., koch, r. (2012). an architecture for distributed, event-driven systems to collect and analyze data in emergency operations and training exercises. proceedings of the 9th international iscram conference, vancouver, canada, april 2012 matopoulos, a., vlachopoulou, m., manthou, v., manos, b. (2007). a conceptual framework for supply chain collaboration: empirical evidence from the agri-food industry. supply chain management: an international journal ,12(3): 177-186. doi:10.1108/13598540710742491 mell, p., grance, t. (2009). cloud computing definition. national institute of standards and technology, 15: 10-7. merilinna, j., matinlassi, m. (2006). state of the art and practice of opensource component integration. in software engineering and advanced applications, 2006. seaa’06. 32nd euromicro conference on. ieee. (pp. 170-177). doi:10.1109/euromicro.2006.61 mula, j., peidro, d., poler, r. (2010). the effectiveness of a fuzzy mathematical programming approach for supply chain production planning with fuzzy demand. international journal of production economics, 128(1): 136-143. doi:10.1016/j.ijpe.2010.06.007 ollero, a., morel, g., bernus, p., nof, s.y., sasiadek, j., boverie, s., erbe, r., goodall, r. (2002). milestone report of the manufacturing and instrumentation coordinating committee: from mems to enterprise systems. annual reviews in control, 26(2): 151-162. doi:10.1016/ s1367-5788(02)00026-3 omg. (2015). object management group. business process model and notation (bpmn) 2.0, http://www.omg.org/spec/bpmn/2.0 retrieved december 2015. peltz, c. (2003). web services orchestration and choreography. computer, (10): 46-52. doi:10.1109/mc.2003.1236471 pervozvanskii, a.a., gaitsgori, v.g. (2013). theory of suboptimal decisions: decomposition and aggregation (vol. 12). springer science & business media. rubinstein, r.y. (2001). combinatorial optimization, cross-entropy, ants and rare events. in stochastic optimization: algorithms and applications (gainesville, fl, 2000), pp. 303-363. dordrecht: kluwer acad. publ. doi:10.1007/978-1-4757-6594-6_14 satyanarayanan, m. (1996). fundamental challenges in mobile computing. in proceedings of the 15th annual acm symposium on principles of distributed computing (pp. 1-7). doi:10.1145/248052.248053 zhang, w., xu, y., dong, x.f. (2012). design and implementation of the agile supply chain information sharing platform in steel industry based on service-oriented architecture and web service. in advanced materials research, 505: 75-81. doi:10.4028/www.scientific. net/amr.505.75 13int. j. prod. manag. eng. (2016) 4(1), 5-13creative commons attribution-noncommercial-noderivatives 4.0 international a cloud platform to support collaboration in supply networks http://dx.doi.org/10.1006/knac.1993.1008 http://dx.doi.org/10.1006/knac.1993.1008 http://dx.doi.org/10.1007/s10726-013-9358-2 http://dx.doi.org/10.1007/bf00927673 http://dx.doi.org/10.1007/bf00927673 http://dx.doi.org/10.2307/1910129 http://dx.doi.org/10.1109/cloud.2009.5071529 http://dx.doi.org/10.1057/jors.1965.7 https://cs.gmu.edu/~sean/book/metaheuristics/essentials.pdf http://dx.doi.org/10.1080/09537280903441930 http://dx.doi.org/10.1108/13598540710742491 http://dx.doi.org/10.1109/euromicro.2006.61 http://dx.doi.org/10.1016/j.ijpe.2010.06.007 http://dx.doi.org/10.1016/s1367-5788(02)00026-3 http://dx.doi.org/10.1016/s1367-5788(02)00026-3 http://www.omg.org/spec/bpmn/2.0 http://dx.doi.org/10.1109/mc.2003.1236471 http://dx.doi.org/10.1007/978-1-4757-6594-6_14 http://dx.doi.org/10.1145/248052.248053 http://dx.doi.org/10.4028/www.scientific.net/amr.505.75 http://dx.doi.org/10.4028/www.scientific.net/amr.505.75 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2015.3038 received 2014-05-29 accepted: 2014-10-22 a performance comparison of single product kanban control systems alvin ang department of systems and engineering management, nanyang technological university 50 nanyang avenue. singapore 639798 angw0038@ntu.edu.sg abstract: this paper presents a simulation experiment comparing the single stage, single product base stock (bs), traditional kanban control system (tkcs) and extended kanban control system (ekcs). the results showed that bs incurs the highest cost in all scenarios; while ekcs is found to be effective only in a very niche scenario. tkcs is still a very powerful factory management system to date; and ekcs did not perform exceptionally well. the only time ekcs did outperform tkcs was during low demand arrival rates and low backorder (c b ) and shortage costs (c s ). that is because during then, it holds no stock. the most important discovery made here is that ekcs becomes tkcs once it has base stock (or dispatched kanbans). the results have also evinced the strength of the pure kanban system, the tkcs over bs. hence managers using bs should consider upgrading to tkcs to save cost. key words: kanban control system (kcs), base stock (bs), extended kanban control system (ekcs), markov chain, arena, simulation. 1. introduction kanban is a japanese word for card. a kanban control system (kcs) is a production mechanism that uses kanbans, or production authorization cards, to control the work-in-process (wip) of the production floor. once a customer demand arrives, the kanban that was previously attached to the finished part is removed and sent back, upstream, to re-initiate the manufacturing production process. this happens simultaneously while the finished part is being shipped to the customer. 1.1. traditional kanban control system (tkcs) the traditional kanban control system (tkcs) was first proposed by sugimori, kusunoki, cho, and uchikawa (1977). figure 1 shows a single stage production line controlled by the tkcs. by single stage, we mean that the manufacturing process (mp) contains only a single server. the number of kanbans limit the wip. this system is the most famous pull mechanism in the world today (monden, 1998). it limits the amount of inventory at each stage, such that the maximum wip is only equal to the number of kanbans circulating in that stage. referring to figure 1, the tkcs operates as follows: when a customer demand arrives at the system it joins customer demand queue d1, requesting the release of a finished product from output buffer b1 to the customer. at that time there are two possibilities: 1. if a part is available in output buffer b1 (which is initially the case), the finished is released to the customer. at the same time, the kanban that was attached to it will be detached. then, this kanban is transferred upstream to the undispatched kanban queue k1, carrying with it a demand signal for the production of a new finished part. mp b0 p b1p + k1 parts to customers k1 p + k1 k1 d1 customer demands p + k1 p figure 1. a single stage, single product traditional kanban control system (ss/sp/tkcs) (source: sugimori et al. (1977)) 57int. j. prod. manag. eng. (2015) 3(1), 57-74creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2015.3038 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. otherwise, if no parts are available in output buffer b1, the demand is backordered and the customer is left waiting in the customer demand queue d1; until such a time a new part is completed and placed in output buffer b1. the newly finished part will be released to the customer instantly and the detached kanban will be transferred back to kanban queue k1. the output buffer b0 represents the raw material inventory buffer, in which it is assumed to carry infinite stock, while the manufacturing process is denoted by mp. the advantage of the kanbans in tkcs is that it acts as a form of feedback, assisting coordination between stages. another advantage is that it sets a limit on wip levels at each stage. however, demand signal blockage can sometimes occur, since demand can only flow upstream if the downstream demand is satisfied. in addition, the kanban provides no instantaneous transmission of demand information to all production stages, neither does it set a limit on the wip for the entire production line. lastly, this system does not respond well to long-term demand fluctuations (monden, 1983). 1.2. base stock (bs) figure 2 shows the base stock (bs) system. it was first proposed by clark and scarf (1960). it does not limit wip but limits inventory stored in the output buffer, b1, with its base stock level, s. it does not use cards as feedback signals to previous stages, but uses instantaneous transmission of demand signals to all production stages. queues di, where i = 1, 2, contain the customer demands. mp b0 p b1p parts to customers p s d1 d2 customer demands p p d1 d2 figure 2. a single stage, single product base stock system (ss/sp/bs) (source: clark and scarf (1960)). the base stock (bs) system operates as follows: when the system is in its initial state (before any demands arrive), buffer b1 contains s number of base stocks of finished products. buffer b0 is the components buffer and is assumed to contain an infinite quantity of components. when a customer demand arrives at the system, it is replicated into multiple demand signals. these signals are immediately transmitted to its respective queues, queues di, where i = 1, 2. the last demand signal joins queue d2, requesting the release of a finished product from buffer b2 to the customer. at this point there can be two possibilities: 1. if a product is available in output buffer b1, it is immediately released to the customer. then, the previous demand signal in customer demand queue d1 will signal the mp to produce a new part to top up the base stock level, s in output buffer b1. 2. otherwise, if no product is available in output buffer b1, the demand is backordered and the customer waits in queue d2 until a new part is completed in the upstream stage. one advantage of the base stock (bs) system is that it sets a target level of production in the output buffer b1, by bounding it with a base stock level, s. that is, the mp will stop producing once the output buffer b1 contains s parts. another advantage of this system is that there is no demand information blockage, due to the assumption of instantaneous transmission of demands to all production stages. one disadvantage of this system is that, though the output buffers are bounded by the base stock level, the wip levels in each stage are unbounded. that is, this system does not set a limit on the wip levels; neither in each stage nor for the entire production line. thus, if a stage fails, the demand process will continue to remove parts from the output buffer, and the machines downstream of the failed machine will operate normally until it becomes starved of parts to process. the upstream stages will continue to receive direct demand information and will operate to release parts as usual leading to an unbounded buildup inventory in front of the failed machine. another disadvantage is that this system has no feedback, and hence there is no coordination between stages. 1.3. extended kanban control system (ekcs) figure 3 shows the extended kanban control system (ekcs). it is a hybrid of both the tkcs and bs. in the initial state, the output buffer b1 contains s amounts of finished parts, with a kanban attached to every part. and the raw material buffer, buffer b0 is assumed to contain infinite number of components. 58 int. j. prod. manag. eng. (2015) 3(1), 57-74 creative commons attribution-noncommercial-noderivatives 4.0 international ang, a. http://creativecommons.org/licenses/by-nc-nd/4.0/ mp b0 p b1p + k1 parts to customers k1 p + k1 k1 s d1 d2 customer demands p + k1 p d d figure 3. a single stage, single product extended kanban control system (ss/sp/ekcs) (source: dallery (2000)) the customer demand queues, queue di, i = 1, 2 are empty; while the undispatched kanban queue, queue k1, contains k number of undispatched kanbans. the ekcs operates as follows: when a customer demand arrives at the system, it is instantaneously split into multiple demands. the first demand joins queue d2, requesting the release of a finished product from output buffer b1 to the customer. at this point, there are two possibilities: 1. if a product is available in output buffer b1, this product is released to the customer after the kanban is detached. simultaneously, this undetached kanban is transferred upstream to the undispatched kanban queue, queue k1.the replicated demand joins customer demand queue d1. since input buffer b0 is assumed to contain infinite raw parts, the previous undetached kanban in queue k1 is now attached to one raw part and sent into the mp for processing. 2. if there is no final product in output buffer b1, the demand is backordered and has to wait in queue d2. however, the replicated demand that went into queue d1 may signal a raw part from input buffer b0 into the mp for processing; presuming that there’s an undispatched kanban lying in the undispatched kanban queue, k1. one advantage of ekcs is that there is an instantaneous transmission of demand. when a demand arrives at the system, it is immediately broadcasted to every stage in the system. this implies that each stage in the system knows immediately the need for production of a new part in order to replenish the finished-product buffer. another advantage of ekcs is the decoupling of kanbans and demand signals. this means that a demand signal moves independently of a kanban, and can be released earlier to upstream stages. a disadvantage of ekcs is that kanbans are freed later in ekcs. a kanban is detached only after it proceeds out of the output buffer. lastly, the ekcs doesn’t respond well to long-term demand fluctuations (dallery, 2000). 2. literature review 2.1. overview of kanban control systems (kcs) the tkcs was first proposed by sugimori et al. (1977). bs was first proposed by clark and scarf (1960) and ekcs was first proposed by dallery (2000). they have been described in the previous section 1 introduction. there are many other types of kcs and they will be briefly mentioned here. conwip stands for constant work–in–process. this system was first proposed by spearman, woodruff, and hopp (1990). this is a pull system as it limits work–in–process (wip) via cards similar to kanbans (w. j. hopp & spearman, 2004). the number of conwip cards represents the total wip allowed. when the preset wip is reached, no new parts can be released into the system until finished parts have been discharged. conwip can also be seen as a single kanban cell encompassing all stages (boonlertvanich, 2005). that is, a ss/tkcs is equivalent to a ss/conwip system. conwip control is executed only at the entry of the manufacturing system. the generalized kanban control system (gkcs) was first proposed by buzacott (1989). gkcs sets a limit on wip levels at each stage. kanbans and demand signals are coupled together and both of them have to be present in an authorization queue before parts can move downstream. however, the presence of the authorization queue means additional waiting time. gkcs does suffer from demand information blockage. demands can only flow upstream if there is a kanban in ki. also, no instantaneous transmission of demand information exists to all production stages. lastly, gkcs doesn’t respond well to longterm demand fluctuations. these factors, including its complicated structure, help explain why gkcs has not become popular (boonlertvanich, 2005). conwip kanban (ck) was first proposed by bonvik, couch, and gershwin (1997). it was proposed to leverage on the advantages of both the conwip and the tkcs. the advantage of conwip kanban is that the wip is controlled for the entire production line as well as the individual stages. this limits 59int. j. prod. manag. eng. (2015) 3(1), 57-74creative commons attribution-noncommercial-noderivatives 4.0 international a performance comparison of single product kanban control systems http://creativecommons.org/licenses/by-nc-nd/4.0/ excessive inventory build-up in front of a machine if it fails. since it has the conwip element, shop floor managers also get to dictate part number sequence and schedule priority jobs first, thereafter allowing the following stages to “pull” the parts downstream. with two kinds of cards, this system has lots of feedback. however, having two types of cards also mean more complications, since workers on the factory floor may have accidental mix ups. there is no instantaneous transmission of demand signals and the systems doesn’t respond well to long-term demand fluctuations (bonvik et al., 1997). extended conwip kanban (eck) was first proposed by (boonlertvanich, 2005). it is a combination of bs, conwip, and tkcs, and is supposed to encapsulate the advantages of all three systems. the kanban is freed earlier in eck than in other systems, since its detached right after the part leaves a mp. there is also an instantaneous transmission of demand. since it has the conwip element, managers get to dictate part number sequence and schedule priority jobs; the following stages “pull” the parts downstream. eck also sets a limit on wip levels for the entire production line, while at the same time maintaining the wip level for each stage. there are lots of feedbacks in this system. conwip cards feedback from the last to the first stage, while kanban cards coordinate every stage. another advantage is that there is a target inventory level (the base stock) set at every stage’s output buffer. however, the eck is more complicated than more traditional systems not only because of its structure, but also because there are two types of cards to handle. lastly, it doesn’t respond well to long-term demand fluctuations (boonlertvanich, 2005). the flexible kanban system (fks) was first introduced by gupta, al-turki, and perry (1999). they introduced this new system to cope with uncertainties and planned/unplanned interruptions. they demonstrated fks’s superiority by conducting four case examples which covered various uncertainties. after comparing the fks’s performance with the traditional jit system, their claim was that in all the cases considered, the performance of their fks was superior. the adaptive kanban system (aks) was first introduced by tardif and maaseidvaag (2001). they introduced a new adaptive kanban-type pull control mechanism which determines when to release or reorder raw parts based on customer demands. they claimed that their system differs from the traditional kanban system in that the number of kanban cards is allowed to change with respect to the inventory and backorder levels. however, the number of cards in the system remains limited, restricting the amount of wip in the system. they showed that their adaptive system can outperform the traditional kanban pull control mechanism while remaining easy to implement. the reactive kanban system (rks) was proposed by takahashi, morikawa, and nakamura (2004). their paper proposed a reactive just-in-time (jit) ordering system for multi-stage production systems with unstable changes in demand. they proposed a reactive controller of the buffer size which can detect unstable changes in the mean and variance of demand. it uses exponentially weighted moving average charts for detection. they placed numerous detection points at each inventory buffer to detect these unstable changes. the performance of their rks is finally analysed using simulation experiments. the knowledge kanban (kk) was proposed by jou lin, frank chen, and min chen (2013). they proposed a kk system to enhance knowledge flow for a virtual research and development (r&d) process. the idea is to employ the kanban philosophy into r&d firms for quicker and easier access to knowledge. they claimed that their proposed system helps employees of these r&d firms reduce the cycle time of their work. first, they created a virtual enterprise (ve). then, they designed a kk model to custom fit it. finally, in their study, they claimed that kk system they created is an effective tool to facilitate knowledge creation, storage, transmission and sharing for r&d firms. the latest e-kanban paper was documented by al-hawari and aqlan (2012). in their paper, they developed a software application for an e-kanban inventory control system; developed to track wip inventory and finished goods in an aluminium factory. they claimed that after the current paper system is replaced with their e-kanban system, data entry errors were minimized. furthermore, their results showed that manufacturing lead time and wip have been reduced by an average of 88% and 50%, respectively. they also built an accountability measure into their system to identify errors. their system can generate reports about order information, aiding managers to make decisions based on realtime information. aghajani, keramati, and javadi (2012) studied a cellular manufacturing system controlled by kanban. in their model, they included the possibility of defective items and rework. they used a mixed60 int. j. prod. manag. eng. (2015) 3(1), 57-74 creative commons attribution-noncommercial-noderivatives 4.0 international ang, a. http://creativecommons.org/licenses/by-nc-nd/4.0/ integer non linear programming (minlp) model to minimize total cost. thereafter, their total cost model was used to determine the optimal number of kanbans and batch size. they also used particle swarm optimization (pso) and simulated annealing (sa) algorithms to lessen the large computational time for solving large minlps. they showed that both pso and sa result in a near optimal solution but the pso algorithm gives a better performance than the sa method. al-tahat, dalalah, and barghash (2012) studied how to synchronize the flow of materials in a kanban controlled serial production line. their production line is described as a queuing network; which they then made use of a dynamic programming (dp) algorithm to solve it by decomposing it into several numbers of single-stage sub-production lines. a performance measure is then developed to determine and compare production parameters. thereafter, they validated their results using the pro model discrete event simulator. they discovered that their performance measure had a very small error compared to their result. thus, they claimed that their method was effective in synchronizing inventory with customer demands. 2.2. comparison studies of kanban control systems (kcs) the true advantages of ekcs over bs and tkcs are still not properly addressed in the research literature. in this section, the literature that compares different types of kcs are reviewed. karaesmen and dallery (2000) used an optimal control framework to study the bs, tkcs and gkcs. they used a two-stage production system where demands arrive according to a poisson process with rate λ and their mp have exponentially distributed service times with rate μi (i=1,2). however, their modeling approach has made it difficult for the analyses of inventory levels in the two separate stages because they used x1 as a random variable to represent a combination of stage 1 output buffer and stage 2 mp. usually, in literature, x1 should denote the wip of the first mp plus the first output buffer while x2 denote the wip of the second mp plus the second output buffer. also, they did not use ekcs in their comparison because under the state space representation approach, the ekcs is a special case of the gkcs. hence, scenarios which ekcs outperforms bs and tkcs are not clearly highlighted. moreover, they did not compare their performances in terms of key performance indicators (kpis). the latest comparison of pull control policies was done by korugan and çadırcı (2008). they studied the four most common pull control systems: bs, tkcs, gkcs and ekcs, using a markov chain model to develop each of the four policies. these models were then analysed using a cost function, which was then minimized with respect to the control parameters of each control mechanism. finally, results are obtained from numerical experiments and conclusions drawn. even though the authors explicitly mentioned that the ekcs and gkcs displayed superior performance over the bs and tkcs, they did not show how this was done. also, the pull models are not of standard tandem process lines. they included an additional remanufacturing process on top of the usual mp, which makes their analysis more complex. khuller (2006) used simulation to compare two types of kanban control systems in different manufacturing environments. although kpis such as fill rate, wip and order fulfillment time were used as a gauge, he did not use the standard ekcs. instead, he used the extended information kanban control system (eikcs) i.e. ekcs with the base stock level equal to the maximum wip capacity at each stage. deokar (2004) also used simulation to compare the tkcs, gkcs and ekcs. she assumed a multiproduct system where the kanbans are either dedicated or shared. by assuming a multi-product system, she increased the complexity of the analysis. even so, she has not specifically mentioned why, and in what scenarios, does the ekcs outperform the tkcs. only four references in the kanban literature exists for evaluating ekcs performance. furthermore, these papers do not distinctly demonstrate how ekcs perform in different scenarios. this clearly shows that there is insufficient analysis on how the ekcs outperforms traditional systems like the bs and tkcs. thus, in order for conventional factory managers to be convinced of ekcs’s performance, a clear and well defined comparison in terms of kpis is needed. 3. kanban controlled systems optimization models this section briefly discusses models to optimize bs, tkcs and ekcs, all of which also assume single stage and single product (ss/sp) for consistency. having models to optimize these systems are important because a fair comparison can then take place. if non-optimal systems were 61int. j. prod. manag. eng. (2015) 3(1), 57-74creative commons attribution-noncommercial-noderivatives 4.0 international a performance comparison of single product kanban control systems http://creativecommons.org/licenses/by-nc-nd/4.0/ compared, their performance results would be inaccurate. the most popular method to optimize bs was proposed by zipkin (2000) later simplified by wallace j. hopp and spearman (2008). it is based on expected total cost (etc), comprised of total holding cost and backorder cost. optimization of each of component leads to a cumulative distribution function (cdf), g(d), of demand during replenishment lead time, which is shown to be equal to the ratio of the individual backorder cost (cb $ per unit) and the sum of holding (ch $ per unit) and backorder costs. with the assumption that cb ≥ ch and g(s*) is poisson distributed, poissoninv ([cb/(cb+ch)], d) is used in matlab to obtain the optimal base stock level, s*. in this research, a matlab program has been written to obtain s* for bs, following zipkin (2000) and wallace j. hopp and spearman (2008)’s approach. many authors have proposed different techniques, mostly based on markov chains, to find the optimal kanban number, k*, for tkcs. one proposed by nori and sarker (1998) is presented, which considers an etc of total holding and shortage costs. the model is based on markov chains, and the state space is fixed at the output buffer b1. demand arrivals follow a poisson process, and exponential processing times are assumed at the mp. in all markov chain-based methods, the most tedious and difficult part is obtaining the steady state probabilities of different states; this typically requires many cross substitutions, arising from simultaneous equations. however, nori and sarker (1998) cleverly devised a coefficient matrix, s, using standard techniques in stochastic processes (ramakumar, 1993) from the rate of departures matrix, r. then they use induction to generalize equations to obtain an expression for the etc. in order to speed up the search process, they ascertain bounds for k. the algorithm proposed by nori and sarker (1998) to obtain, k* for tkcs has been coded in matlab for comparison purposes. ang and piplani (2010) have proposed a model for obtaining optimal base stock, s*, and optimal number of kanbans, k*, in a ss/sp/ekcs. in this research, our method will be used for optimizing the ekcs. likewise, a matlab program for this model has been written. 4. simulation experiment 4.1. arena simulation model tkcs, bs and ekcs were simulated in arena version 12. figures i1 to i3 in appendix i shows their snapshots respectively. these simulation models were developed based on their respective kcs schematics. thus, ss/sp/bs (figure i1) corresponds to figure 1; ss/sp/tkcs (figure i2) corresponds to figure 2 and ss/sp/ekcs (figure i3) corresponds to figure 3. these models were simulated only after their respective optimal parameters were found. for example in bs case, the optimal base stock, s*, was found using the method discussed in section 3, coded in matlab and simulations were run using the parameters described in section 4.2. these steps are repeated for ss/sp/tkcs and ekcs. 4.2. simulation parameters simulation experiments are conducted under three scenarios (table 1): low, medium and high backorder and shortage costs, cb and cs. the holding cost, ch, is kept constant. mp rate is also constant, but the demand arrival rate is varied. before each scenario is simulated, matlab was used to obtain optimal base stock, s* and kanban, k* for the three systems. based on these optimal values, operation of ss/sp/ekcs, tkcs and bs is simulated in arena simulation software. finally, results are tabulated and compared in terms of actual total cost (atc). in this experiment, key performance indicator (kpi) for each kcs is actual total cost (atc). all other kpis are translated into atc. for example, table 1. simulation parameters for ss/sp/kcs. scenario 1(low) scenario 2 (medium, 20x) scenario 3 (high, 200x) holding cost, ch (per unit per day) $10 $10 $10 backorder cost, cb (per unit) $20 $200 $2,000 shortage cost, cs (per day) $20 $200 $2,000 manufacturing process, mp, rate (per day) 20 demand arrival rates (units per day) 10, 12, 14, 16, 18 62 int. j. prod. manag. eng. (2015) 3(1), 57-74 creative commons attribution-noncommercial-noderivatives 4.0 international ang, a. http://creativecommons.org/licenses/by-nc-nd/4.0/ a kpi like fill rate can be indirectly represented by total backorder cost (since number of backorders is simply number of demands unfilled), while a kpi like average inventory level can be represented by the total holding cost. in order to obtain atc for each scenario, the total backorder, shortage and holding costs are added together. thus, actual total cost (atc) = total backorder cost + total shortage cost + total holding cost (1) backorder cost, cb, can be defined as a penalty cost, in dollars per unit, for unmet demand at the end of each period. shortage cost, cs, in dollars per unit day, is defined as the penalty cost for demand kept waiting. for example, if average customer waiting time, as given by arena, is 20 hours, it means that on average, a customer order is required to wait for 20 hours before the demand is met. this translates into a shortage cost of $(20/24×cs), as the system is in shortage mode for 20 hours. finally, holding cost, ch, can be defined as cost (dollars per unit per day) of holding inventory in the output buffer. 4.3. simulation assumptions assumptions made in modelling the ss/sp/kcs were: 1. all systems produced only a single product. 2. one card kanban system was adopted. 3. the system produced no defective parts. 4. all systems had a single stage containing only one mp. 5. each mp contained only one machine/server. 6. machine setup times were zero. 7. no machine failures could occur. 8. each machine could only process one part per unit time. 9. parts were transported with negligible transfer time. 10. demand signals and kanbans flowed instantaneously. 11. parts followed a first in first out (fifo) dispatching policy at all machines and buffers. 12. input material buffers had an infinite supply of component parts. 13. demand arrivals followed a poisson process. 14. all processing times at mps were assumed to be exponentially distributed. 15. each replication was run for one year. 16. each simulation run was replicated 10 times. 17. the warm up period for each replication was three months. the reason for using a three month warm up period was to eradicate any transient behaviour of the systems. since each replication was chosen to run for one year, we discarded the results for the first quarter, selecting the steady-state results only at the beginning of the second quarter. 5. results and discussion figures 7 to 9 show the atc of each system with varying demand arrival rates. the figures show that ekcs and tkcs outperform bs significantly by achieving a lower atc. however, the most interesting insight is that the performance of ekcs does not differ much from that of tkcs. 100,00 120,00 140,00 150,04 121,87 11,07 11,57 21,21 21,91 26,99 2,07 26,38 0,00 50,00 100,00 150,00 200,00 10 12 14 16 18 a ct ua l t ot al c os t ( $) demand arrival rate (units per day) bs tkcs ekcs figure 4. comparison of ss/sp/ekcs, tkcs and bs in low backorder and shortage costs scenario 63int. j. prod. manag. eng. (2015) 3(1), 57-74creative commons attribution-noncommercial-noderivatives 4.0 international a performance comparison of single product kanban control systems http://creativecommons.org/licenses/by-nc-nd/4.0/ referring to figures 7 to 9, the most prominent difference is between bs and the other two systems. bs incurs the highest cost, followed by tkcs and ekcs. bs, by definition, keeps a pre-specified level of stock, thereby incurring higher inventory costs. because bs follows a “push” production strategy, whilst ekcs and tkcs follow a “pull”, bs will produce stock according to the demand arrival rate; the higher the arrival rate, the more stock it will produce. in fact, the popular optimization algorithm for bs (proposed by zipkin (2000)) does not take into account the mp rate, as the idea is to stock up and prevent a stock-out situation. on the other hand, tkcs and ekcs follow a lean philosophy. they produce only when needed and keep inventory low. their optimization methods incorporate mp rates to obtain the utilization rate. the utilization rate represents the level of congestion in the system, and determines the optimal number of kanbans using markov chains. since the number of kanbans defines wip or “congestion” level, controlling it ultimately determines the inventory level. 5.1. ekcs and tkcs performance similarity looking at figures 4 to 6, it can be noted that performance results of ekcs and tkcs differ very little. in fact, ekcs seems to imitate tkcs in almost all scenarios. on the surface, they look vastly different, with ekcs having something that tkcs doesn’t, namely instantaneous transmission of demand. however, deeper analysis reveals that they aren’t as different as they seem to be. both their average inventory levels seem to hold almost equal amount of inventory; hence, their average backorders and customer waiting times are almost equal. this leads to their atc being very close. they hold comparable inventory levels because their optimal number of dispatched kanbans is always the same. dallery (2000) also notes that by setting the number of kanbans for tkcs the same as the base stock level of ekcs, ekcs becomes and behaves like tkcs. 140,00 170,00 180,00 200,00 182,58 20,67 25,67 38,25 89,92 35,50 39,08 83,17 0,00 50,00 100,00 150,00 200,00 250,00 10 12 14 16 18 a ct ua l t ot al c os t ( $) demand arrival rate (units per day) bs tkcs ekcs figure 5. comparison of ss/sp/ekcs, tkcs and bs in medium backorder and shortage costs scenario. 140,00 170,00 180,00 200,00 182,58 25,67 38,25 39,08 89,92 20,67 35,50 39,08 83,17 0,00 50,00 100,00 150,00 200,00 250,00 10 12 14 16 18 a ct ua l t ot al c os t ( $) demand arrival rate (units per day) bs tkcs ekcs figure 6. comparison of ss/sp/ekcs, tkcs and bs in medium backorder and shortage costs scenario. 64 int. j. prod. manag. eng. (2015) 3(1), 57-74 creative commons attribution-noncommercial-noderivatives 4.0 international ang, a. http://creativecommons.org/licenses/by-nc-nd/4.0/ 5.2. dispatched kanbans between ekcs and tkcs the number of kanbans calculated by optimization algorithms for ekcs and tkcs are almost equal. the number of dispatched kanbans in ekcs represents its “base stock” level, just as kanbans in tkcs represent the average inventory level. logically, the more stock a system has, the higher its inventory level, but with lower backorders and customer waiting time, and vice versa. hence, adding one more dispatched kanban is equivalent to increasing the base stock, and increasing holding cost, and thus atc. on the other hand, taking away one kanban lowers base stock by one, lowering holding cost, but incurring longer customer waiting time, thereby (possibly) increasing atc again. this illustrates the need to balance holding costs and backorder/ shortage costs. the optimization algorithms used in this research seek the lowest atc in all scenarios to compute optimal base stock, s* and/or number of kanbans, k*. 5.3. undispatched kanban queue in ekcs: some comments proposers of ekcs have claimed that it is “leaner” than tkcs, as it uses its undispatched kanban queue to lower inventory, yet achieves optimal wip. however, this research has shown that that does not result in ekcs outperforming tkcs; rather, its performance gets worse. this may be due to the following reasons: 1. by reducing the number of kanbans and placing them in the undispatched queue, average onhand inventory level is reduced, leading to higher backorder and shortage costs and ultimately, higher atc, outweighing the benefits of lower stock. the proposers of ekcs had the idea of reducing stock by having the undispatched kanban queue locked away and used only when needed. but the moment ekcs has base stock, optimal ekcs becomes optimal tkcs, and optimal dispatched kanbans in ekcs make it analogous to tkcs. 2. base stock in ekcs makes the undispatched kanbans redundant. referring to figure 3, the only time the undispatched kanbans are allowed into the mp are in the event of demand arrivals. but demand arrivals are always accompanied by kanbans being passed from downstream stages, which are then placed behind these undispatched kanbans. those kanbans in front of queue k1 get attached to component parts, and are sent into mp (since there is already demand arrival in queue d1). this brings the undispatched kanbans back to the original number. looking back, the initial proposed role of undispatched kanbans in ekcs was to allow more wip into mp. but it turns out that absolutely no benefit results, since the bottleneck is the mp rate. in other words, more wip can be allowed into mp for ekcs, but mp can still only process one part per unit time. 3. instantaneous transmission of demands in ekcs does not deliver any benefits, as kanbans in queue k1 already fulfil this role. this is not immediately clear. but upon closer examination of figures 1 and 3, the same thing can be observed for both systems: once a demand arrives, and there is a part in the output buffer, a component part is instantly sent into mp. thus, with or without queue d1, ekcs behaves identically to tkcs. of mp1 mp2 b0 k1 p b1 b2 d p + k1 parts to customers customer demands k1 pp + k1 p + k1 k2 p + k2 p + k2 p + k2 k2 figure 7. two stage, single product tkcs. 65int. j. prod. manag. eng. (2015) 3(1), 57-74creative commons attribution-noncommercial-noderivatives 4.0 international a performance comparison of single product kanban control systems http://creativecommons.org/licenses/by-nc-nd/4.0/ course, this is under the assumption of negligible kanban transfer time. even if this assumption was relaxed and kanban transfer time was taken into account, it would still affect both systems the same way. 5.4. study of multiple stage, single product kanban control systems (ms/sp/kcs) in a two stage sp/tkcs (figure 7), it features instant demand transmission to all stages upon a demand arrival, since its kanbans are transmitted to their individual stage mps immediately (assuming a part is in the output buffer and instantaneous transmission of kanbans), even though demand arrives only at the final stage. this makes the role of instantaneous transmission of demands in a two stage sp/ekcs (figure 8) redundant. that is, demand queues d1 and d2 are unnecessary. this is because if base stock exists in b1 and b2, any demand arrivals will immediately transmit previously attached kanbans upstream, joining the undispatched kanban queues k1 and k2. those kanbans placed in front are then attached to component parts and sent into mp. in the end, the instantaneous demand queues are rendered unnecessary. 5.5. comparison of tkcs and ekcs in low utilization scenario this section analyses scenarios under which ekcs may outperform tkcs. further investigation (figures 9 to 11) shows that below 50% utilization rate, for low backorder, cb, and shortage cost, cs, ekcs outperforms tkcs. however, medium and high cb and cs scenarios show negligible difference. as can be seen in figure 9, the cost difference between ekcs and tkcs is quite significant and worthy of further study. mp1 b0 p b1 d3 p + k1 parts to customers k1 p + k1 k1 s1 d1 k2 d2 k2 customer demands p + k1 p + k2 mp2 p + k2 b2 s2 p + k2 p d d d figure 8. two stage, single product ekcs. 10,16 10,27 10,43 10,68 1,13 1,24 1,42 1,68 0,00 2,00 4,00 6,00 8,00 10,00 12,00 2 4 6 8 a ct ua l t ot al c os t ( $) demand arrival rate (units per day) tkcs ekcs figure 9. comparison of ekcs and tkcs (low utilization rates and low backorder and shortage costs). 66 int. j. prod. manag. eng. (2015) 3(1), 57-74 creative commons attribution-noncommercial-noderivatives 4.0 international ang, a. http://creativecommons.org/licenses/by-nc-nd/4.0/ 5.6. ekcs and tkcs: some comments the reason for the significant cost difference in figure 9 is that ekcs has zero holding cost. ekcs can afford not to hold stock in its output buffer as: 1. the utilization rate is very low –below 50%. this means that, most of the time, mp is idle and whenever demand arrives, it can produce at a fast rate to meet the demand. in contrast, figure 7 shows that the gap starts to close at above 10 demand arrivals per day (or above 50% utilization), as mp gets increasingly congested. stock is now needed to prevent backorders and to shorten customer waiting time. but the moment ekcs has base stock, it starts to behave like optimal tkcs. 2. the ratio of holding cost, ch to backorder, cb, and shortage cost, cs, is low. this implies that the cost incurred in holding stock in ekcs is comparable to the cost for backorders and shortages. so in this scenario, having lower stock proves to be less costly, and making customers wait leads to the same cost penalty as holding stock. tkcs, however, is forced to hold same amount of stock as the number of kanbans. hence, even though ekcs and tkcs can both have same number of kanbans in their system–each having only one, tkcs attaches kanbans to real stock, whereas ekcs holds kanbans in the undispatched queue, which is converted into finished product only upon arrival of a demand. the most important insight here is that the extra demand queue d1 (figure 3) in ekcs prevents undispatched kanbans from entering mp when there is no demand, thereby making ekcs a system with truly no inventory. 5.7. ekcs and tkcs: medium and high backorder, cb, and shortage costs, cs, scenario in medium to high cb and cs scenarios (figure 10 and 11), the gap between ekcs and tkcs is negligible. the key factor is whether ekcs holds base stock. in medium cb and cs scenario (cs=20×ch), ekcs does not hold base stock, thus increasing shortage costs. this leads to higher costs for ekcs. ekcs, in this case, is still capable of remaining stockless and achieving a slightly lower cost than tkcs. 11,58 12,67 14,33 16,83 11,25 12,42 14,17 16,75 0,00 5,00 10,00 15,00 20,00 2 4 6 8 a ct ua l t ot al c os t ( $) demand arrival rate (units per day) tkcs ekcs figure 10. comparison of ekcs and tkcs (low utilization rates and medium backorder and shortage costs) 25,83 31,67 40,83 51,67 25,83 31,67 40,83 51,67 0,00 10,00 20,00 30,00 40,00 50,00 60,00 2 4 6 8 a ct ua l t ot al c os t ( $) demand arrival rate (units per day) tkcs ekcs figure 11. comparison of ekcs and tkcs (low utilization rates and high backorder and shortage costs). 67int. j. prod. manag. eng. (2015) 3(1), 57-74creative commons attribution-noncommercial-noderivatives 4.0 international a performance comparison of single product kanban control systems http://creativecommons.org/licenses/by-nc-nd/4.0/ in a high cb and cs scenario (cs=200×ch), there is no difference in cost between ekcs and tkcs, as ekcs now has to hold base stock. it cannot be stockless anymore as the penalty for shortage is too high, leading to the identical atc curves in figure 11 5.8. ekcs’ performance: final comments if ekcs holds base stock, the undispatched kanbans become ineffective. but when ekcs does not hold base stock, the question becomes: would the undispatched kanbans still be useful? once ekcs is stockless, the maximum number of undispatched kanbans required is only one, as the mp can only produce one part at a time. increasing the number of un-dispatched kanbans allows more wip into mp, but that only makes it more congested. in other words, even if more than one demand arrives simultaneously, the system only really needs one undispatched kanban, as the sole undispatched kanban can be sent back instantly to upstream queue k1 and then into mp for processing. this is also what would happen in a system with multiple undispatched kanbans. the conclusion then is that ekcs outperforms tkcs only when it contains no base stock and during low-demand arrival rate and low-backorder (cb) and shortage cost (cs) scenarios. also, the optimal number of undispatched kanbans in the system is one, assuming negligible kanban transfer times. 6. conclusion and future work this paper presents a performance comparison of single stage, single product kanban control systems (ss/sp/kcs), namely ekcs, tkcs and bs. firstly, optimization models for bs and tkcs are described. these optimization models are used to find the optimal base stock level, s*, and optimal number of kanbans, k*, for the respective systems. then, three scenarios with different simulation parameters are set up to compare the individual kcs performance. the simulation results in this chapter show that bs incurs the highest cost in all cost scenarios, while ekcs is found effective only under very special cases. also, tkcs is still a very powerful system. the only time ekcs outperforms tkcs is when demand arrival rate is low, and backorder, cb, and shortage costs, cs are low as well, since under those circumstances it does not need to hold stock. the most important insight made was that ekcs behaves like tkcs once it contains base stock (or dispatched kanbans). this chapter also supports the superiority of the pure kanban system, the tkcs, over bs. bs was developed in the 1960s, while tkcs was developed in the 1970s and ekcs in the 2000s. naturally, tkcs outperformed bs, because lean production seems to work best for mass-produced products, such as cars, which is where these systems were predominantly implemented. in fact, many publications have described tkcs as the “just–in– time (jit)” revolution that made toyota the biggest car manufacturer in the world (womack, jones daniel t., & roos daniel., 2007). in this chapter, it is shown that bs always incurs the highest cost, as it stocks a higher level of inventory, disregarding the mp processing rate and putting emphasis only on demand arrival rate. all in all, the results clearly illustrate that bs is an inferior control system when compared to pull-type control systems. to summarize, the main findings of this research are: 1. ekcs outperforms tkcs only when the demand rate is low (<50% utilization rate) and backorder, cb, and shortage costs, cs are low. 2. if ekcs has stock in its output buffer, it behaves exactly like tkcs. their performance becomes the same, as the optimal number of dispatched kanbans is the same. 3. if ekcs has stock in its output buffer, its undispatched kanbans become ineffective, and the number of kanbans equal the base stock. 4. the role of the extra demand queue for instantaneous transmission in ekcs (queue d1) is ineffective. this is because, if compared to the tkcs, tkcs also has this functionality but without needing the additional queue. in other words, since we have assumed negligible kanban transfer times, tkcs’ kanban queues also act to instantaneously transmit a demand signal. thus, in this context, adding an extra demand queue for ekcs does not improve its performance at all. 5. extra demand queues are useful only when ekcs has no stock held in the output buffer. extra demand queues help lock up undispatched kanbans, which makes ekcs truly stockless.. 6. it has been shown that ms/sp/ekcs behaves similarly to ms/sp/tkcs, assuming negligible kanban transfer times; the optimal number of undispatched kanbans in such a case is one. 68 int. j. prod. manag. eng. (2015) 3(1), 57-74 creative commons attribution-noncommercial-noderivatives 4.0 international ang, a. http://creativecommons.org/licenses/by-nc-nd/4.0/ 7. implication for practice in practical situations, the ekcs is best applied to managing the production of niche or high net value products. using car manufacturing as an example, niche market cars such as formulae one indy cars, or customized luxury cars, have extremely low demands. manufacturers (or “crafters”) do not produce them until they receive orders (sardi, 2009). these cars are only produced one unit per time and only per order. these cars also have low backorder and shortage costs as compared to holding costs (sardi, 2009). well-known luxury car brands such as ferrari care more for their image than mass production. in other words, their concern is not about backorders nor long customer waiting times; but rather, having too many of their cars on the roads cheapening their image. hence they even create waitlists for customers who wish to purchase their cars. this is very similar to how ekcs behaves – which has an optimal number of undispatched kanban only equal to one and it performs best only during low backorder and shortage costs scenarios. as for economical cars such as toyota or honda, the tkcs would still be the preferred choice of managing their productions. these cars have medium to high backorder and shortage costs compared to holding costs (womack et al., 2007). that is, managers of such production floors cannot afford to keep their customer waiting, or worse, having their customers walk off. these manufacturers stand to lose out if they do not hold stock in the long run. also, these cars usually have high demands (monden, 1993). similarly, the tkcs performs best during medium and high backorder and shortage cost scenarios, coupled with high utilization rates of more than 50%. future work for this research would be to compare single-stage, multiple-product kanban controlled systems which operate and behave very differently from sp/kcs. references aghajani, m., keramati, a., javadi, b. (2012). determination of number of kanban in a cellular manufacturing system with considering rework process. the international journal of advanced manufacturing technology, 63(9-12): 1177-1189. doi:10.1007/s00170-012-3973-y al-hawari, t., aqlan, f. (2012). a software application for e-kanban-based wip control in the aluminium industry. international journal of modelling in operations management, 2(2): 119-137. al-tahat, m. d., dalalah, d., barghash, m. a. (2012). dynamic programming model for multi-stage single-product kanban-controlled serial production line. journal of intelligent manufacturing, 23(1): 37-48. doi:10.1007/s10845-009-0336-0 ang, a., piplani, r. (2010). a model for determining the optimal number of base stock and kanbans in a single stage extended kanban control system (ekcs). paper presented at the proceedings of the 5th aotule international postgraduate students conference on engineering. bonvik, a. m., couch, c. e., gershwin, s. b. (1997). a comparison of production-line control mechanisms. international journal of production research, 35(3), 789-804. doi:10.1080/002075497195713 boonlertvanich, k. (2005). extended-conwip-kanban system: control and performance analysis. (ph.d.), georgia tech, usa. buzacott, j. a. (1989). queuing models of kanban and mrp controlled production systems. engineering costs and production economics, 17(1-4): 3-20. doi:10.1016/0167-188x(89)90050-5 clark, a. j., scarf, h. (1960). optimal policies for a multi-echelon inventory problem. management science, 50(12_supplement): 17821790. doi:10.1287/mnsc.1040.0265 dallery, y. (2000). extended kanban control system: combining kanban and base stock. iie transactions, 32(4): 369-386. doi:10.1080/07408170008963914 deokar, s. s. (2004). performance evaluation of multi-product kanban-like control systems. thesis (m.s.i.e.), university of south florida, 2004., 71. gupta, s. m., al-turki, y. a. y., perry, r. f. (1999). flexible kanban system. international journal of operations & production management, 19(10): 1065-1093. doi:10.1108/01443579910271700 hopp, w. j., spearman, m. l. (2004). to pull or not to pull: what is the question? manufacturing & service operations management, 6(2): 133. doi:10.1287/msom.1030.0028 hopp, w. j., spearman, m. l. (2008). factory physics (3rd ed.). new york, ny: mcgraw-hill/irwin/irwin. jou lin, c., frank chen, f., min chen, y. (2013). knowledge kanban system for virtual research and development. robotics and computerintegrated manufacturing, 29(3): 119-134. doi:10.1016/j.rcim.2012.04.020 karaesmen, f., dallery, y. (2000). a performance comparison of pull type control mechanisms for multi-stage manufacturing. international journal of production economics, 68(1): 59-71. doi:10.1016/s0925-5273(98)00246-1 69int. j. prod. manag. eng. (2015) 3(1), 57-74creative commons attribution-noncommercial-noderivatives 4.0 international a performance comparison of single product kanban control systems http://dx.doi.org/10.1007/s00170-012-3973-y http://dx.doi.org/10.1007/s10845-009-0336-0 http://dx.doi.org/10.1080/002075497195713 http://dx.doi.org/10.1016/0167-188x(89)90050-5 http://dx.doi.org/10.1287/mnsc.1040.0265 http://dx.doi.org/10.1080/07408170008963914 http://dx.doi.org/10.1108/01443579910271700 http://dx.doi.org/10.1287/msom.1030.0028 http://dx.doi.org/10.1016/j.rcim.2012.04.020 http://dx.doi.org/10.1016/s0925-5273(98)00246-1 http://creativecommons.org/licenses/by-nc-nd/4.0/ khuller, b. (2006). a comparison of traditional and extended information kanban control system using dedicated and shared kanbans. master’s thesis, graduate college of oklahoma state university. korugan, a., çadırcı, ö. (2008). a comparison of pull control policies in hybrid production systems. poms 19th annual conference, la jolla, california, u.s.a. lind, d. a., marchal, w. g., wathen, s. a. (2011). basic statistics for business & economics douglas a. lind ; william g. marchal ; samuel a. wathen: boston [u.a.] mcgraw-hill 2011. 7th ed., internat. student ed. monden, y. (1983). toyota production system: practical approach to production management. norcross, ga: industrial engineering and management press, institute of industrial engineers. monden, y. (1993). toyota production system: an integrated approach to just-in-time (2nd ed.). norcross, ga.: industrial engineering and management press. doi:10.1007/978-1-4615-9714-8 monden, y. (1998). toyota production system : an integrated approach to just-in-time (3rd ed.). norcross, georgia: engineering & management press. nori, v. s., sarker, b. r. (1998). optimum number of kanbans between two adjacent stations. production planning & control, 9(1): 60-65. doi:10.1080/095372898234523 ramakumar, r. (1993). engineering reliability : fundamentals and applications (prentice hall international ed.). london: prentice-hall international. sardi, m. (2009). ferrari waiting list so, you want a ferrari? automobile magazine, february, 2009 issue(february, 2009 issue). spearman, m. l., woodruff, d. l., hopp, w. j. (1990). conwip. a pull alternative to kanban. international journal of production research, 28(5): 879-894. doi:10.1080/00207549008942761 sugimori, y., kusunoki, k., cho, f., uchikawa, s. (1977). toyota production system and kanban system: materialization of just-in-time and respect-for-human system. international journal of production research, 15(6): 553-564. doi:10.1080/00207547708943149 takahashi, k., morikawa, k., nakamura, n. (2004). reactive jit ordering system for changes in the mean and variance of demand. international journal of production economics, 92(2): 181-196. doi:10.1016/j.ijpe.2003.10.014 tardif, v., maaseidvaag, l. (2001). an adaptive approach to controlling kanban systems. european journal of operational research, 132(2): 411-424. doi:10.1016/s0377-2217(00)00119-3 womack, j. p., jones daniel t., & roos daniel. (2007). the machine that changed the world : the story of lean production -toyota’s secret weapon in the global car wars that is revolutionizing world industry (1st trade pbk. ed.). new york: free press. zipkin, p. h. (2000). foundations of inventory management. boston: mcgraw-hill appendix i – simulation figures from arena m p1 match b 1 n k 2 create part match b 0 n k 1 batc h b0 n k1 dem ands create cus t cus tom ers di s pos e parts to 1 as s i gn kanban create k1 0 0 0 0 0 0 figure i1. arena snapshot of a single stage, single product traditional kanban control system (ss/sp/tkcs). 70 int. j. prod. manag. eng. (2015) 3(1), 57-74 creative commons attribution-noncommercial-noderivatives 4.0 international ang, a. http://dx.doi.org/10.1007/978-1-4615-9714-8 http://dx.doi.org/10.1080/095372898234523 http://dx.doi.org/10.1080/00207549008942761 http://dx.doi.org/10.1080/00207547708943149 http://dx.doi.org/10.1016/j.ijpe.2003.10.014 http://dx.doi.org/10.1016/s0377-2217(00)00119-3 http://creativecommons.org/licenses/by-nc-nd/4.0/ appendix ii – statistical analysis of simulation results we used jmp to conduct hypothesis tests for the simulation results. 29 out of 30 hypothesis tests conducted on the single product kanban controlled system (sp/kcs) showed that ekcs does outperform tkcs and bs. hence, the claim that ekcs outperforms tkcs and bs is true. there were three main scenarios for the simulation experiments: low, medium and high backorder and shortage costs. for each of these scenarios, the actual total cost (atc) was plotted against the demand arrival rate. since the comparisons done were using the atc mean values; but not their standard deviations, hypothesis tests were done to confirm the results. hence, for each scenario, a hypothesis test was carried out. we follow lind, marchal, and wathen (2011) method of comparing population means with unknown population standard deviations (pooled t-test).these are the assumptions: 1. the samples are independent 2. the two populations follow the normal distribution 3. the population standard deviations are unknown (thus we use the t distribution rather than the z) n k1 m atc h b0 n d1 m p1 se p a ra te cu s t 1 o r iginal duplicat e n k2 m atc h b1 n d2 ba tc h b0 n d1 n k1 cu s to m e rs di s p o s e pa rts to cre a te pa rt cre a te s1 de m a n d s cre a te cu s t di s p o s e 0 0 0 0 0 0 0 0 0 figure i2. arena snapshot of a single stage, single product base stock (ss/sp/bs). k1 m a tc h b0 n d1 n m p1 se p a ra te cu s t 1 o r iginal duplicat e m a tc h b1 n d2k1 ba tc h b0 n d1 n as s i g n ka n b a n 1 cu s to m e rs di s p o s e pa rts to cre a te pa rt cre a te s1 de m a n d s cre a te cu s t cre a te k1 0 0 0 0 0 0 0 0 0 figure i3. arena snapshot of a single stage, single product extended kanban control system (ss/sp/ekcs). 71int. j. prod. manag. eng. (2015) 3(1), 57-74creative commons attribution-noncommercial-noderivatives 4.0 international a performance comparison of single product kanban control systems http://creativecommons.org/licenses/by-nc-nd/4.0/ step 1: taking samples we take a specific case as an example. for the low backorder and shortage cost scenario, and for a demand arrival rate of 10 units per day, the following ten samples were taken: step 2: stating the claim since there are three systems but we can only do one comparison per time, we have to make two claims here h h ekcs kcs ekcs tkcs t0 1 | | 2 #n n n n where μekcs: refers to mean of ekcs’s actual total cost (atc) μtkcs: refers to mean of tkcs’s actual total cost (atc) h h ekcs bs ekcs bs 3 4 | | 2 #n n n n where μekcs: refers to mean of ekcs’s actual total cost (atc) μbs: refers to mean of bs’s actual total cost (atc) this means that if h0 is accepted, ekcs’s atc is lower than tkcs. and if h3 is accepted, it means that ekcs’s atc is lower than bs. step 3: selecting level of significance we choose a significance level of 0.05, or rather, α=0.05. in other words, if h0 and h3 are accepted, the claim that ekcs does outperform tkcs and bs is proven at a 95% confidence interval (this is using the t distribution for a one-tailed test). according to lind et al. (2011), the p-value gives the probability of observing a sample value as extreme as, or more extreme than, the value observed, given that the null hypothesis is true. a p-value is frequently compared to the significance level to evaluate the decision regarding the null hypothesis. it is a means of reporting the likelihood that h0 is true. if the p-value is greater than the significance level, then h0 is not rejected. but if the p-value is less than the significance level, then h0 is rejected. step 4: perform a two-sample pooled t-test for difference of two means using statistical software jmp. we enter the above sampled data into jmp figure ii1. filling up the y and x axis for jmp. table ii1. simulation parameters for ss/sp/kcs. sample number actual total cost (atc) $ extended kanban control system (ekcs) traditional kanban control sys-tem (tkcs) base stock (bs) 1 1.91 11.17 107.89 2 2.31 12.12 76.49 3 2.14 11.85 100.68 4 1.96 11.91 98.73 5 2.30 12.19 102.03 6 2.00 11.50 100.03 7 2.09 11.20 102.94 8 2.51 9.87 90.45 9 2.47 10.57 106.78 10 1.69 11.11 108.21 72 int. j. prod. manag. eng. (2015) 3(1), 57-74 creative commons attribution-noncommercial-noderivatives 4.0 international ang, a. http://creativecommons.org/licenses/by-nc-nd/4.0/ after doing a t-test, we obtained: figure ii2. jmp output showing a data plot and t test of the data samples. since we are investigating if mean atc for ekcs is less than tkcs, we are performing a left tailed test. as such, the respective p-value is 1. as p-value >0.05 (or α), we accept h0 and reject h1. therefore, we are 90% confident that ekcs outperforms tkcs since its atc is lower. the following steps above are repeated for comparing ekcs versus bs. and its p-value is also 1, which is >0.05 (or α), we accept h0 and reject h1. therefore, we are 95% confident that ekcs outperforms bs since its atc is lower. results of hypothesis test thirty hypothesis tests were conducted at 95% confidence interval for sp/kcs systems. ten tests were conducted for each for the three scenarios: low, medium and high backorder and shortage cost. overall, in almost all cases, it showed that ekcs does outperform tkcs and bs; except for a few unique cases. referring to table ii2 below, there are ten hypothesis tests conducted for low backorder, shortage cost scenario. each of it showed a p-value of over 0.05 (or α); which means that the claim of ekcs atc is lower than tkcs and bs is true at a 95% confidence level. referring to table ii3 below, there are ten hypothesis tests conducted for medium backorder, shortage cost scenario. most of it showed a p-value of over 0.05 (or α); which means that the claim of ekcs atc is lower than tkcs and bs is true at a 95% confidence level. we now examine the two special cases where ekcs does not outperform. table ii2. results of hypothesis test for ss/sp/kcs: low backorder and shortage cost scenario. demand arrival rate (lambda) (units per day) ekcs vs. tkcs p-value ekcs vs. bs p-value conclusion at 95% confidence interval 10 1 1 ekcs outperforms tkcs and bs 12 0.3143 1 ekcs outperforms tkcs and bs 14 0.9721 1 ekcs outperforms tkcs and bs 16 0.462 1 ekcs outperforms tkcs and bs 18 0.6581 1 ekcs outperforms tkcs and bs table ii 3. results of hypothesis test for ss/sp/kcs: medium backorder and shortage cost scenario. demand arrival rate (lambda) (units per day) ekcs vs. tkcs p-value ekcs vs. bs p-value conclusion at 95% confidence interval 10 0.5767 1 ekcs outperforms tkcs and bs 12 0.0515 1 ekcs outperforms tkcs and bs. but it only outperforms tkcs slightly since the p-value is very close to 0.05 (or alpha) 14 0.9794 1 ekcs outperforms tkcs and bs 16 0.0193 1 ekcs outperforms bs but not tkcs because its p-value is lower than alpha. 18 0.7666 1 ekcs outperforms tkcs and bs 73int. j. prod. manag. eng. (2015) 3(1), 57-74creative commons attribution-noncommercial-noderivatives 4.0 international a performance comparison of single product kanban control systems http://creativecommons.org/licenses/by-nc-nd/4.0/ case 1: for the scenario of demand arrival rate of 12 units per day. we examine the jmp output: figure ii3. case 1: where ekcs only outperforms tkcs by a little. referring to figure ii3, looking at the data plot we see that the values are very close to one another. this explains why the p-value is only 0.0515, very close to α of 0.05. case 2: for the scenario of demand arrival rate of 16 units per day. we examine the jmp output: figure ii4. case 2: where ekcs does not outperform tkcs. referring to figure ii4, this is a rare case, and indeed the only case that the p-value (0.0193) has fallen below 0.05. hence, only in this case we are 95% confident that ekcs does not outperform tkcs. referring to table ii4 below, there are ten hypothesis tests conducted for high backorder, shortage cost scenario. all of them showed a p-value of over 0.05 (or α); which means that the claim of ekcs atc is lower than tkcs and bs is true at a 95% confidence level. table ii4. results of hypothesis test for ss/sp/kcs: high backorder and shortage cost scenario. demand arrival rate (lambda) (units per day) ekcs vs. tkcs p-value ekcs vs. bs p-value conclusion at 95% confidence interval 10 0.9988 1 ekcs outperforms tkcs and bs 12 0.997 1 ekcs outperforms tkcs and bs 14 0.9031 1 ekcs outperforms tkcs and bs 16 0.1666 1 ekcs outperforms tkcs and bs 18 0.6768 0.2255 ekcs outperforms tkcs and bs 74 int. j. prod. manag. eng. (2015) 3(1), 57-74 creative commons attribution-noncommercial-noderivatives 4.0 international ang, a. http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2016.4599 received 2015-09-02 accepted: 2016-04-29 applying total quality management tools using qfd at higher education institutions in gulf area (case study: alhosn university) adnan al-bashir industrial engineering department. the hashemite university. jordan. abashir@hu.edu.jo abstract: human power’s quality plays the key role in the growth and development of societies where the quality of human powers can be enriched with the high quality education provided by the higher education institutions. the higher education institutions are hereby an important sector of any society since it defines the overall quality of human lives. this research will investigate the application of total quality management (tqm) tools at the higher education institutions; specifically at alhosn university. in this study five tools were implemented at alhosn university’s engineering college including: quality function deployment, affinity diagrams, tree diagrams, pareto charts, and fishbone diagrams. the research will reveal that the implementation of tqm tools has a great benefit for higher education institutions where they have uncovered many area of potential improvement as well as the main causes of some of the problems the faculty of engineering is facing. also, it will show that the implementation of tqm tools on higher education institution systems will enhance the performance of such institutions. key words: quality function deployment, higher education, total quality management. 1. introduction the construction of a modern and knowledgeable society requires governments to draw their attention on the type of education provided to their society. this would certainly require governments to devote their efforts and resources in achieving an educational system of high quality. governments around the world are trying to develop strong relationships between the education sector and the society (zaman, 2016). one of the ways in which educational systems at higher education institutions (hei) can be developed and improved is the use and implement of total quality management (tqm) tools. higher education is now a real part of the globalization process in which supply and demand are matched, qureshei et al., 2014. higher education institutions are institutions which seek to achieve defined objectives and goals in their academic, research, and community development aspects. this has given the institutions’ achievements in science, technology, and services a high importance because of the benefits such achievements have on graduates and society which puts higher education institutions under more pressure to enhance their abilities in adaptation and development in respect with the fast development in both education and technology (abou chahine et al., 2008) the usage of such tools would give the higher educational institution a strong position among other educational institutions since it would afford a higher quality educational system in the competitive field the higher education institution is being competent in. in order to assure that the institution is being competitive enough, total quality management tools allows institutions to review and assess their performance to see whether int. j. prod. manag. eng. (2016) 4(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international 87 http://dx.doi.org/10.4995/ijpme.2016.4599 http://creativecommons.org/licenses/by-nc-nd/4.0/ they follow the required conditions of learning and teaching (tarawneh and mubaslat, 2011). the application of total quality management is done through using qualitative and quantitative tools as means for higher education institutions to assess the performance of the educational institution where they can define their strengths in order to enhance it and their weaknesses in order to eliminate them. besides, the institution needs to take into consideration the improvement for the institution over time, and compare its performance with the performance of other local and foreign universities. there are many factors that would affect the learning environment and the quality of the services provided by the higher educational institutions. the application of total quality management at higher education institutions is not an easy task; it consumes lots of time, money and efforts. the total quality management would require the institution to be fully committed to a continuous cycle of improvement and development. as per tarawnwh and mubaslat (2011) applying the concepts, principles and tools of tqm at higher education institutions face many problems/ obstacles such as the following: the difficulty of determining the quality of non-tangible educational services in higher education institutions. most of the higher education institutions focus on introducing and applying several laws and standards to develop the educational system; however, little work is done to measure and control the quality of their system. limited budget and resources the institution or government can afford which leads to the decrement in the performance of the institutions and the decrement in the quality level of its outputs. some studies, such as rosa et al. (2012) study, have agreed on the non-applicability of the tqm practices in hei’s claiming that the difficulty comprises from the fact that the higher education is a service which is intangible. on the other hand, many other studies were against this idea and have highly supported the idea of implementing tqm practices on the hei’s such as al-tarawneh and mubaslat (2011) and abou chahine et al. (2008). salameh et al. (2011) stated in their research that the traditional administrative methods are no longer useful for higher education institutions and it should be replaced by the tqm philosophy to improve the performance of the hei’s. the reason behind the ineffectiveness of the traditional methods stems from environmental forces such as growing number of students, competition between higher educational institutions and the flexibility of programs in both postgraduate and undergraduate levels which certainly have made the need for effective implementation of quality in higher education more essential (najafabadi et al. 2008). as abou chahine et al. (2008) have showed, the effectiveness of the educational quality would first depend on the proper identification of the clients of the hei’s. hereby, there are two primary clients: students and other stakeholders such as parents/ guidance, business, society, etc; where students are performing a dual role: the role of the client, and the role of the product of higher education. salameh et al. (2011) have declared that the student satisfaction is the crucial factor in the success of the higher education and it must be the focal point in all tqm practices. the main objective of this research is to show the importance of assessing and continuously improving the performance of the higher educational institutions by applying several qualitative and quantitative tqm tools, and this will allow determining the areas of potential improvement and obstacles that it might face (tarawneh and mubaslat, 2001). 2. overview of qfd a large number of tqm tools are applied in different manufacturing and service disciplines to improve quality. one of the most common approaches for assessing quality in higher education institutions is the quality function deployment (qfd). the qfd, which was first developed by dr. yoji akoa in the 1960’s, is a quantitative tool that translates a set of customer requirements into operation requirements to be met by a new product and process design. the qfd is applied by using a sort of matrix called the house of quality (hoq) 2.1. house of quality according to mukaddes et al. (2010), “quality function deployment helps to maintain a correct focus on true requirements and minimizes int. j. prod. manag. eng. (2016) 4(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-bashir, a. 88 http://creativecommons.org/licenses/by-nc-nd/4.0/ misinterpreting customer needs”. thus, this matrix allows organizations to prioritize the customer requirements, identify their position within the market, and identify their position against other competitors (benchmarking). moreover, shyamal (2011) has stated that relating customer requirements to technical requirements through the qfd support the idea of work team between design engineers, marketing staff, and manufacturing workers. qfd also support the concept of continuous improvement since organizations are continually required to research and develop in accordance with the changes in customers’ needs and the rapid technological changes. as per qureshei et al., 2012, applying qfd in higher education institutions has hereby recorded a remarkable success since it is capable of maintaining a customer oriented service. according to mukaddes et al. (2010), “quality function deployment helps to maintain a correct focus on true requirements and minimizes misinterpreting customer needs”. thus, this matrix allows organizations to prioritize the customer requirements, identify their position within the market, and identify their position against other competitors (benchmarking). moreover, shyamal (2011) has stated that relating customer requirements to technical requirements through the qfd support the idea of work team between design engineers, marketing staff, and manufacturing workers. qfd also support the concept of continuous improvement since organizations are continually required to research and develop in accordance with the changes in customers’ needs and the rapid technological changes. applying qfd in higher education institutions has hereby recorded a remarkable success since it is capable of maintaining a customer oriented service. 3. methodology the qfd has been used as a tool of quality assessment and improvement for the faculty of engineering at alhosn university and as a tool of benchmarking for some of the higher education institutions in the city of abu dhabi. the qfd figure 1. house of quality. int. j. prod. manag. eng. (2016) 4(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international applying total quality management tools using qfd at higher education institutions in gulf area (case study: alhosn university) 89 http://creativecommons.org/licenses/by-nc-nd/4.0/ was applied by constructing the house of quality matrix through the following steps: 3.1. step 1: identification of the customers’ requirements the quality function deployment process starts by identifying the needs and requirements of customers, or what is called the voice of customer, which is certainly considered as the most important step through the qfd process qureshei et al., 2014. as it has been mentioned before, any higher educational institution would have two basic types of customers: the students and the market. hereby, students were encouraged to describe their needs in their own words by having two focus group sessions with both male and female students from the engineering college at alhosn university. 3.1.1. affinity and tree diagrams the students’ and market’s requirements were entered into the customer requirements portion in the house of quality matrix after constructing an affinity diagrams as shown in figure 2. the affinity diagram previously shown is one of the total quality management tools that is aimed to summarize the large number of ideas and issues created at the focus group sessions. according to breyfogle (2003) the affinity diagrams helps individuals in better understanding the essence of the problems and breakthrough solution alternatives. the requirements were then entered into a tree diagram figure 3, which is used “to communicate a logical relationship that is hierarchal between events” (breyfogle, 2003:124). such diagram made requirements easily arranged to be directly entered into the house of quality. figure 2. affinity diagram for customers’ requirements. int. j. prod. manag. eng. (2016) 4(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-bashir, a. 90 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.2. step 2: identification of the technical requirements. this step deals with the technical specifications that alhosn staff and employees would require through the educational process at the university. at this step, a meeting with academic and administrative staff from alhosn university was held in order to clarify how customer requirements can be achieved; or what is called the voice of engineering. identifying the technical requirements would certainly describe the quality of the university’s products who are the students. following are the technical requirements. facility availability of parking lots availability of sport clubs and availability of study rooms availability of food court safer and advanced labs educational fees affordable book prices affordable credit hours fees scholarships and discounts courses improve research opportunities more practical experience availability of distance learning strict exam grading staff qualified faculty staff availability language and presentation skills enhance administrative procedures students skills language skills managerial skills communication and presentation skills computer skills graduation the ability of solving real-life engineering problems graduation ceremony relations with local companies and factories recognition continuous improvement students' apprectiation better class and exam schedules variety of engineering programs customers' requirements figure 3. customers’ requirements tree diagram. int. j. prod. manag. eng. (2016) 4(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international applying total quality management tools using qfd at higher education institutions in gulf area (case study: alhosn university) 91 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.3. step 3: quantifying the relative importance of each of the customer requirements. at this step, the house of quality deals with finding out the relative importance of each of the customer requirements. this has been done through distributing a survey that has a scale from 1 to 5 where 1 represents an extremely unimportant requirement and 5 represent an extremely important requirement. the survey was distributed among customers –female students and male students from alhosn engineering college and 95 surveys were collected back. according to the results of the surveys, an average importance value was calculated for each of the requirements. 3.4. step 4: performing benchmarking assessment of the customer requirements this step performs a benchmarking assessment between alhosn university and two of their competitors within the city of abu dhabi; they would be named as competitor 1 and competitor 2. hereby, we have distributed another survey to determine the satisfaction level of students of each of alhosn, competitor 1, and competitor 2 regarding each if the requirements mentioned at step 1. another time an average value for each of the satisfaction weightings for each of the three universities has been calculated as shown in figure 4. table 1. technical requirements. faculty skills examination skills communication skills lecturing skills students’ skills hard-working students active research students: teaching staff ratio sufficient number of teaching staff appropriate size of class facilities strong library sport activities for students engineering equipments for labs bigger campus good environment for teaching students’ evaluation entry exam exit exam faculty’s evaluation faculty selection criteria for recruitment evaluation for courses encourage students to read educational field trips internship time flexibility for faculty decrease the work-load on faculty financial incentives online procedures int. j. prod. manag. eng. (2016) 4(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-bashir, a. 92 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.5. step 5: relating the students’ requirements with the technical requirements this step forms the inside of the house of quality and used to detect the relationship between the customers’ needs and the staff requirements. these correlations are determined qfd team and represented using one of the three following symbols: + 9 strong + 3 medium + 1 week 3.6. step 6: identification of the interrelationship among the technical requirements. this step is responsible for building the roof of the house of quality, the roof which is also called the correlation matrix is used to examine the interrelationships between the technical requirements. the interrelationships usually determined by the qfd team and expressed by variety of symbols such as: represents a positive relationship represents a negative relationship 3.7. step 7: calculations of the planning matrix here, the students’ requirements planned satisfaction rate and sales point were set in order to calculate the improvement factor and the overall weighting of each of the requirements. in case of the planned satisfaction rate it was indicated in accordance with the university’s current capabilities in providing students with the requirements they need. the rates used at this step are the same as those used in step 3. then, the sales point were determined based on the importance weighting of the requirement where each requirement was given a point within the range of 1 to 1.5 in such a way that high sales points are given to requirements which have high importance weighting. sales values are used to add weight to the requirement where it can be utilized by the university to promote the programs and accordingly attract more students to the university. the improvement factor was then calculated using the following formula: if=([psr–sw]×ii)+… … …eq. (1) where: if: improvement factor psr: planned satisfaction rate sw: satisfaction weighting of alhosn university students ii: improvement increment, which had a fixed value of 0.2 lastly, the overall weighting represented in the last column of the house of quality was calculated by multiplying the importance weighting by the sales point and the improvement factor. 3.8. step 8: prioritizing the technical requirements in meeting the customers’ expectations this is the final section of the hoq matrix and it sum up the conclusion drawn from the data contained in the entire matrix. this section is often made up of three essential parts: the target value, absolute weight and the relative weight. target value – is an objective measure that defines by how much each technical requirement should be improve in order to meet or exceed the customer’s expectation. the target value should not be determined in vain, the qdf team should fully understand the customer needs, the competitor’s performance and the organization’s current performance before determining any of the target values. the last two rows of the hoq matrix are the absolute weight and the relative weight. those weights show the impact of technical requirements on the customer requirements. the formulas used to calculate each of these weights are as follows: int. j. prod. manag. eng. (2016) 4(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international applying total quality management tools using qfd at higher education institutions in gulf area (case study: alhosn university) 93 http://creativecommons.org/licenses/by-nc-nd/4.0/ absolute weight …… .a r c eqj ij i i j n f= = | (2) where: aj: row vector of absolute weights for the technical requirements (j=1, 2, …, m) rij: weights assigned to the relationship matrix (i=1, 2, …, n), (j=1, 2, …, m) ci: column vector of importance weighting (i=1, 2, …, n) m: number of technical requirements n: number of customers requirements relative weight .r eqb d ……j ij i i j n f= = | (3) where: bj: row vector of relative weights for the technical requirements (j=1, 2, …, m) di: column vector of overall weighting for the customer requirements (i=1, 2, …, n), (j=1, 2, …, m) higher absolute and relative weights indicate the critical areas where the engineering efforts need to be concentrated. the results of applying qfd at the college of engineering at alhosn would be discussed in the following chapter: results and discussion. 4. results of applying quality function deployment at alhosn university the qfd has been used in identifying the characteristics that needs to be enhanced in order to meet customers’ expectations. as shown in figure 4, the customers’ requirements were identified and had included requirements regarding the offered courses, facilities, educational fees, and others. similarly, the technical requirements (staff requirements) were identified and had included requirements such as students’ evaluation, faculty evaluation, etc. according to the survey done at step 3, it was observed that all of the requirements mentioned in figure 4 have got a high rank by students; a rank of 4 (important) and a rank of 3 (neutral). but regarding the satisfaction weighting of alhosn students which was done at step 4, it was shown that most of the requirements has got a rank of 2 (unsatisfied) and 3 except one of the requirements that got a rank of 4 (satisfied); satisfaction with the qualifications the faculty have. by comparing the satisfaction weightings of alhosn university with the competitor universities it was noticed that alhosn satisfaction weightings were either behind the satisfaction weightings of the competitors or similar to those of the competitors except that alhosn has got a higher rank in satisfying students with their qualified faculty than competitor 1 has (rank of 3) and a similar rank to competitor 2 (rank of 4). the relationship matrix between customer and technical requirements done at step 5 shows that all students’ requirements regarding the facility has a strong relationship with possessing a bigger campus. it also shows that students who are hardworking and actively research are the students’ who get appreciated. another observation is that active research cannot be done unless students’ were provided with advanced and safe labs, appropriate language skills courses to write research papers, etc. moreover, making relations with local companies and factories is strongly associated with the educational field trips and internships done by university. the house of quality (hoq) provided in figure 4 shows other relationships between both types of requirements. the roof of the house of quality (step 6) has shown the interrelation of the technical requirements. it has shown that the results of the students’ exit exam are strongly associated with students’ internships and educational field trips. also, it shows that the provision of abigger campus would allow staff to have the appropriate number of students in class, provide students with sport activities, and possess a larger and stronger library which would certainly lead to a better learning and teaching environment. other relationships can be seen in the hoq in figure 4. int. j. prod. manag. eng. (2016) 4(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-bashir, a. 94 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 4. house of quality. int. j. prod. manag. eng. (2016) 4(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international applying total quality management tools using qfd at higher education institutions in gulf area (case study: alhosn university) 95 http://creativecommons.org/licenses/by-nc-nd/4.0/ at step 7, it can be seen that many of the customers’ requirements planned satisfaction rate is much higher than the current satisfaction weighting of alhosn students. for instance, the requirements of advanced labs, practical experience, relations with local companies and factories have a satisfaction weighting of 3 while the planned satisfaction rate is 5. also, the column of the sales point has included high sales point which the university must use in marketing the engineering programs at alhosn such as the qualified faculty which was given a sales point of 5. lastly, the overall weighting was calculated to find out that the ability of solving real life engineering problems, the relations with local companies and factories, the recognition and the continuous improvement have got the highest overall weighting value. on the other hand, the availability of distance learning has got the least overall weighting value which was 3.75. lastly, the target value was set and the absolute and relative values were calculated in the last three rows in the house of quality. it is worth mentioning that the higher the absolute and the relative values of the requirements are, the more technical efforts are needed in those areas. in this case study, the requirement of having a bigger campus has got the highest absolute value which is 315, and the highest relative value which is 500.1. 5. discussion the study has revealed that there is a good awareness from top management of the importance of applying quality practices at their higher education institution. in this regard, the university has given the quality of the programs a great attention through implementing quality standards imposed by the ministry of education and the abet. they also implement some of the quality practices in order to assure the continuous development and the customers’ satisfaction such as distributing surveys and getting the feedback so they can develop action plans to follow. one of the tqm practices that the university must consider is benchmarking in order for the university to be one of the best competent among local and hopefully foreign universities. at this study, benchmarking has been performed through the utilization of the quality function deployment matrix and it was seen that the university is still behind other competitors at some places. also, the study has shown that there is a general feeling that the students are not satisfied with the services they are getting; this was concluded from the application of the qfd tool where satisfaction weighting have smaller values than the importance weighting indicating the university must give the student’s requirements more focus. 6. conclusion total quality management (tqm) is an integrated organizational effort designed to improve quality at every level to achieve excellence. tqm has a remarkable application in hei’s where the adaption of tqm will help the higher educational institution to maintain their competitive position, satisfy all stakeholders, focus on the market needs and achieve higher performance. there are varieties of qualitative and quantitative tools and techniques that can be used in order to implement tqm principles such as benchmarking, statistical process control, quality function deployment, failure mode and effect analysis, pareto charts, cause and effect diagram and others. 7. recommendations there is a need for big change in higher education institution quality management style. tqm must hereby be used as a quality system to be followed in any institution in order to make development. academic and administrative staff should be provided with total quality management training courses to assure tqm’s success in higher education institutions. in this regard, financial resources must be properly allocated for training staff. the staff should be provided with the necessary skills to practice tqm tools, especially statistical control methods so that tqm practices can be monitored and controlled. the following recommendations were proposed to be implemented in areas that need more attention to be paid: the provision of a bigger campus for students and faculty in order to have a better learning environment. there is a big need for establishment of a research center (including labs) in the university’s campus in order to increase the research opportunities for both students and faculty. relationships with local companies and factories should be available in order to gain recognition int. j. prod. manag. eng. (2016) 4(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-bashir, a. 96 http://creativecommons.org/licenses/by-nc-nd/4.0/ and provide students with the necessary practical experience. moreover, it would increase the chance for graduated students to find jobs after graduation. the university must have selective criteria in choosing students. students who want to register at the university must pass an entry exam; in case of engineering college, the exam should include questions from math, chemistry, physics, and english subjects. there is also a need to have an exit exam to verify whether they possess the proper qualifications to practice real life jobs. the exam would include questions from a variety of courses the student has studied in the program. references abet – accreditation. (2011). abet. retrieved march 14, 2013. abou chahine, s., jammal, a., kaissi, b., loutfi, m. (2008).guide i: introduction to quality management in higher education in lebanon. (project id: scm-m014a05). abu hasan, h. f. (2008). service quality and student satisfaction: a case study at private higher education institutions. international business research. 1(3), 163-175. ahmed, r., ali, s. (2011). implementing tqm practices in pakistani higher education institutions. pak. j. eng. technol. sci., 2(1), 1-26. ahmed, a., hamdoon, b. (2007). the challenges and obstacles of tqm implementation in the higher education institutions: the case of sharjah university in uae. international journal of management sciences and business research, 3(3), 8-23. akao, y. (1997). qfd: past, present and future, international symposium on qfd, 1997, linkopping, sweden. al-amri, a. h., bin bon, a. t. (2012). measuring the total quality management in the yemeni universities. international journal of research and reviews in applied sciences, 10(1), 37-45 al-tarawneh, h., mubaslat, m. (2011). the implementation of total quality management (tqm) on the higher educational sector in jordan. international journal of industrial marketing, 1(1), 1-10. besterfield, d. h., besterfield-michna, c., besterfield, g. h., besterfield-sacere, m. (2003). total quality management. new jersey: pearson education. breyfogle, f. w. (2003). implementing six sigma. hoboken, new jersey: john wiley & sons, inc. dan reid, r., sanders, n. (2007). operations management. united states: wiley. durlabhji, s. g., fusilier, m. r. (1999). the empowered classroom: applying tqm to college teaching. managing service quality, 9(2), 110–115. http://dx.doi.org/10.1108/09604529910257939 gonzalez, r. c., woods, r. e. (2011). process quality (1st ed.). upper saddle river, nj: pearson prentice hall. gryna, f. m., chua, r. h., defeo, j. a. (2007). juran’s quality planning and analysis: foe enterprise quality (5th ed.). new york: mcgraw-hill companies. hauser, j., clausing, d. (1988). the house of quality, harvard business review, 66(3), 63-73. hughey, a. w. (2000). application of the deming philosophy to higher education. industry of higher education. 14(1), 40-44. http://dx.doi. org/10.5367/000000000101294841 jamison, p. (n.d.). teacher decision making in student evaluation. ssta research centre report #02-07. lai-kow chan, ming-lu wu, (2002). invited review quality function deployment: a literature review, european journal of operational research, 143, 463–497. http://dx.doi.org/10.1016/s0377-2217(02)00178-9 mukaddes, a. m., bagum, m. n., islam, m. a., bashar, m., a., chakrabarty, v. (2010). translating the student’s voice into teaching techniques: a quality function deployment approach. in proceedings of the 2010 international conference on industrial engineering and operations management, dhaka, bangladesh, 237-241. najafabadi, h., sadeghi, s., habibzadeh, p. (2008). total quality management in higher education case study: quality in practice at university college of borås. university college of borås institution for engineering school thesis for graduation in master of science in industrial engineering. qureshi, m., khan, k., bhatti, m., khan a., zaman, k. (2012). quality function deployment in higher education institutes of pakistan. middle-east journal of scientific research 12 (8): 1111-1118. qureshi, m. i., janjua, s. y., zaman, k., lodhi, m. s., tariq, y. b. (2014). internationalization of higher education institutions: implementation of dmaic cycle. scientometrics, 98(3), 2295-2310. http://dx.doi.org/10.1007/s11192-013-1163-9 rosa, m., sarrico, c., amaral, a. (2012). implementing quality management systems in higher education institut. quality assurance and management, intech publisher. http://dx.doi.org/10.5772/33922 salameh, r., alzyadat, m., alnsour, j. (2011). implementation of (tqm) in the faculty of planning & management at al-balqa applied university. international journal of business and management, 6(3), 194-207. http://dx.doi.org/10.5539/ijbm.v6n3p194 int. j. prod. manag. eng. (2016) 4(2), 87-98creative commons attribution-noncommercial-noderivatives 4.0 international applying total quality management tools using qfd at higher education institutions in gulf area (case study: alhosn university) 97 http://dx.doi.org/10.1108/09604529910257939 http://dx.doi.org/10.5367/000000000101294841 http://dx.doi.org/10.5367/000000000101294841 http://dx.doi.org/10.1016/s0377-2217(02)00178-9 http://dx.doi.org/10.1007/s11192-013-1163-9 http://dx.doi.org/10.5772/33922 http://dx.doi.org/10.5539/ijbm.v6n3p194 http://creativecommons.org/licenses/by-nc-nd/4.0/ shyamal, g. (2011). quality function deployment (qfd). https://xisspm.files.wordpress.com/2011/07/chap-7-qfd.pdf. retrieved on 15, december 2015. winn, r. c., green, r. s. (1998). applying total quality management to the educational process. international journal of engineering education, 14(1), 24-29. zaman, k. (2016). quality guidelines for good governance in higher education across the globe. pacific science review b: humanities and social sciences, 1(1), 1-7. http://dx.doi.org/10.1016/j.psrb.2016.01.001 int. j. prod. manag. eng. (2016) 4(2), 87-98 creative commons attribution-noncommercial-noderivatives 4.0 international al-bashir, a. 98 https://xisspm.files.wordpress.com/2011/07/chap-7-qfd.pdf http://dx.doi.org/10.1016/j.psrb.2016.01.001 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2015.3459 received 2014-06-29 accepted: 2014-12-15 some trends and applications of operational research/management science to operations management ramón companysa and imma ribasb a escola politècnica superior d’edificació de barcelona, upc, dr. gregori marañón, 44-50, 08028 barcelona ramon.companys@upc.edu b escola tècnica superior d’enginyers industrials de barcelona, upc. avda. diagonal 647, 08028 barcelona imma.ribas@upc.edu abstract: the editor suggested us to write about our point of view on the current use of operations research techniques applied to the operations management and about its future evolution. with some of unconsciousness we accept it, but it is obvious that our vision, even though we try to do our best, will be partial and biased. hence the title chosen shows signs of prudence. more caution have been applied to the development where, after a glance at the past and reflection on the abundance of new denominations without content, we consider five aspects that, nowadays, acquire increasing importance and that will strongly influence in future developments. among the five aspects two correspond to trends in the field of operations research techniques, one is a philosophy in the field of operations management, another to an area of the company and the last one to an industrial sector in which operations management, supported by operations research methods, is taking a predominant role. key words: operational research, management science. 1. introduction it is impossible to try to see the present and the future without looking, occasionally, to the past. in a recent paper titled “operational research versus operations management” (petrovic and mccarthy, 2014) published in the operational research society bulletin, the connections between both disciplines are analysed. “o. r. is as a discipline that deals with the application of advanced analytical methods to help make better decisions” whereas “om deals with the activities, decisions and responsibilities of managing the design, production and delivery of goods and services”. they conclude that there are many points of contact, a large area of overlap, but differ in some aspects and approaches. for us, the concept of or defined by ackoff and sasieni (1968) was and is still valid. they said that “or can be considered as being: the application of scientific method, by interdisciplinary teams to problems involving the control of organized (manmachine) systems so as provide solutions which best serve the purpose of the organization as a whole.” our only qualification is to interpret “interdisciplinary” as the existence of a team with members with different views and experiences. therefore, for us, or is not a toolbox, not a collection of recipes, not a profession or an academic entertainment, it is an attitude, an approach, a philosophy to focus and solve problems connected with the operation of the systems. however, we have nothing to object to the above definition of om, although we used to distinguish, according to buffa (1961), the strategic decisions (design) from the tactical ones (operations), and we identified om with the latter ones. hence, for us, or and om are not the same: or is a methodology while om is an area of management. or can, and probably should be, used to solve problems of om. 1int. j. prod. manag. eng. (2015) 3(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2015.3459 mailto:ramon.companys@upc.edu mailto:imma.ribas@upc.edu http://creativecommons.org/licenses/by-nc-nd/4.0/ if the problem is recurrent, or will design quantitative methods and techniques that will be associated with om. the title of this paper refers to these procedures and techniques. for perspective we have included figure 1. a few years ago, a graph, which appeared in the first edition of dervitsiotis, impacted us. the graph was titled: chronological development of production technology (hardware) and production methodology (software). this figure, which originally ranged from 1750 to 1980, suggested that the development of both technology and methodology going in ascending order, with an exponential growth, but the second one, methodology, was behind technology with an apparently growing gap. a new updated version is shown in figure 1 where the time domain is up today. in the figure, the abscissa scale is linear and refers to the time, the ordinate scale is logarithmic and in it is measured the development in an arbitrary unit; given the nature of the scale, the exponential growth is represented as a line segment. the upper line refers to technological development, which dervitsiotis baptized as “hardware”, and the lower, the methodological development or “software”. most milestones reflected in the figure are widely known, what excuse us of giving an explanation, which unnecessarily lengthen the text. we just indicate some details that seem significant to us. the initial milestone, which appears in 1775, refers to the watt’s steam engine improved, contemporary with the ideas of adam smith on the advantages of the division of labour. in 1798 eli whitney introduced the notion of interchangeability, standardization, and brought them to practice. a follower of whitney was colonel samuel colt who invented the famous revolvers, with absolutely interchangeable parts. in 1837 general baron antoine henri jomini published precis de l’art de la guerre ou nouveau tableau analytique des principales combinaisons de la grande tactique et de la politique militaire; in which what he calls the art of war is divided into six parts and the fourth one is the logistics, defined as “practical art of moving armies.” charles babbage appears in the graph, not for designing an automatic calculating machine but for writing “on the economy of machinery and manufacturers” (1832) that noted an advantage of the division of labor that had not been mentioned by 1 1750 1800 1850 1900 1950 2000 2050 10 100 1000 10000 j. watt (1775) steam engine m. boulton & j.watt jr. (1775) a. smith (1776) the wealth of nation e. withney (1798) interchangeable parts jacquard (1801) loom g. stepheson (1821) father of railways morse(1844) morse code n. otto (1860) a.g. bell (1876) (1900) lathe ford (1913) (1951) computer (1948) nuclear power station (1942) atomic pile (1976) apple-i (1971) intel 8008 (1970s) rfid (1969) apolo 11 (1968) plc's arpanet (1967) (1992-94) www (1995) gps (2003) graphene (2005) 3d printer ch. babbage (1832) a.h. jomini (1837) f. taylor (1895) h.gantt (1910) gilbreth (1912) f.harris (1913) h.fayol (1916) shewart (1920) tippet (1930) o.r. (1940) l.p. (1947) jonhson (1954) bowman & fetter (1957) pert (1958) h.m.m.s. (1960) j.forrester (1962) qfd (1972) mrp (1975) jit (1981) mrp ii (1982) the goal (1984) tqm (1985) m. baldrige (1989) reenginering (1993) internet (1995) supply chain (1995) analytics (2007) (1935) radar figure 1. chronological development of production technology and methodology. 2 int. j. prod. manag. eng. (2015) 3(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international companys, r. and ribas, i. http://creativecommons.org/licenses/by-nc-nd/4.0/ adam smith. he noted that the bottleneck job was, surely, the one that perceived a higher salary, and if all the work is carried on by a single operator, he should have such category. there is no room for stopping in f. taylor, in gilbreth and h. gantt, whose contributions are widely known. ford harris published his formula eoq around 1913 (although some authors situate this fact in 1911). less known in our environments is henry fayol, mining engineer of the ecole de saint-etienne who entered, with 19 years old, in the societé commentryfourchambault. he published, in 1908, the report “discussion des principes de l’administration generale” which is the first draft of his seminal work, “administration industrielle et génerale: prevoyance, organisation, commandement, coordination, contrôle”, published in 1916. in 1913, henry ford installed the first assembly line to produce the t model. this is neither the first nor the only experience, but was the one that popularized the procedure, probably because it was linked to the perception of the workers of perceiving an excellent salary for those times. walter shewart invented, in 1924, the statistical quality control and in 1934 l. h. c. tippet, in england, developed a new application of statistics to the production, the work sampling, accepted, albeit with great reluctance. harold f. dodge y harry g. romig, from bell telephone laboratories, worked in the reception control and published their sampling tables in 1944. the first commercially available computer, the univac i (universal automatic computer) was developed by a company founded by mauchly and eckert, who had developed at the university of pennsylvania the eniac (electronic numerical integrator and computer), in 1945. other methodological milestones include the publication, in 1957, of the book of bowman and fetter (1957) that was one of the first to link quantitative techniques with production management, the linear rules by holt et al (1960), in 1960 and the models based in industrial dynamics by forrester (1962). the advanced research projects agency network (arpanet, 1967) was one of the world’s first operational packet switching networks, the first network to implement tcp/ip, and the progenitor of what was to become the global internet (1995). although the beginnings of rfid (radio frequency identification) is around 1970 is not until 1990s when the intense use of rfid revolutionized logistics operations. after seen the past, we can look at the present and extrapolate into the near future. a first observation we make is the feeling that the methodological innovations have been slowing since the 80s of last century. many new names have appeared, but few new concepts. moreover if the observations we have made on the delay of the methodology with regard the technology are true, the current technological advances, which are quite significant in the field of information systems, will be which define the methodological developments in the near future. three factors that, from our point of view, will influence the structure of new quantitative methods that serve as a support to operations management are: the increasing complexity of the situations they face, the risk caused by the high variability of many influential elements and the sustainability policy and respect for the environment, which is necessary in a world of limited resources. in particular, they must incorporate mechanisms to deal with two conflicting concepts: robustness to respond to the uncertainty generated by the short-term variability, and flexibility to adapt to structural changes at higher levels. we do not know what the future will be but to discuss some connections between operations management and quantitative methods and techniques, we dealt with the following headings: analytics exotic heuristics the lean manufacturing the supply chain or in healthcare 2. analytics there are three kinds of lies: lies, damned lies, and statistics. benjamin disraeli if you torture the data long enough, it will confess. ronald coase (professor of economics, university of chicago) you may prove anything by figures. thomas carlyle, chartism (1840) if enough data is collected, anything may be proven by statistical methods. williams and holland’s law, in arthur bloch (1977) 3int. j. prod. manag. eng. (2015) 3(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international some trends and applications of operational research/management science to operations management http://en.wikipedia.org/wiki/packet_switching http://en.wikipedia.org/wiki/tcp/ip http://en.wikipedia.org/wiki/internet http://creativecommons.org/licenses/by-nc-nd/4.0/ since some years ago, analytics and big data have become popular, and in many cases linked to operational research environments. possibly one of the triggers of this interest is the knowledge that, in the 2002 campaign for president of the usa, in the obama team worked 50 experts in analytics. one of its missions was to identify voters likely to be influenced by a contact (persuadable voters swing) and those where contact could be counterproductive. under the orders of chief scientist data, developed a persuasion voter model, with the following characteristics: 1. what’s predicted: which voter will be positively persuaded by political campaign contact such a call, door knock, flyer, or tv ad. 2. what’s done about it: persuadable voters are contacted, and voters predicted to be adversely influenced by contact are avoided. news in the same direction was the work of cern to locate the theorized higgs boson in 1964. scientists analysed more than 800 trillion collisions to confirm their hypothesis. during this process they amassed more than 200 petabytes of data, what was scrutinized billions of times doing statistical analysis to confirm and corroborate that the particle whose trace was found was indeed higgs boson. the scientists concluded that there were one-in-550 million chances that the results may have been a statistical coincidence. what is analytics? according to nestler et al. (2012) “analytics is the scientific process of transforming data into insights for making better decisions”. but also it is said “an important aspect of teaching decision skills is the observation that we are dealing with a decision-maker, a person. people, not data, make decisions” (abbas, 2014). since old times there are two schools of thought, which according to the times and fashions are imposed on one another. one of them argue “analyze what we do (decide) and based on this analysis we will see what data are needed” and the other “capture and keep all the data you can and then we will see what we do with them.” we belong to the first one because we are of the opinion that you cannot fish smaller than the size of the fish net used. therefore the network must adapt to the fish sought and any search results, capture and storage of data are biased by definition. unfortunately, or fortunately, depending how you look, now the technology allows the capture, storage and processing of data on quantities and deadlines before implausible. in the 1960s, statisticians used terms like “data fishing” or “data dredging” to refer to what they considered the bad practice of analysing data without an a-priori hypothesis. however, the term “data mining” that appeared around 1990 in the database community, although it is similar to the data fishing term became popular in the business and press communities. since about 2007, “predictive analytics” (name that first appeared in 1999) and since 2011, “data science” were also used to describe this field. predictive analytics is defined as the technology that learns from experience (data) to predict the future behaviour of individuals in order to drive better decisions (siegel, 2013). millions of decisions a day determine whom to call, mail, approve, test diagnose, warn, investigate, incarcerate, set up on a date, and medicate. pa is the means to drive perperson decisions empirically, as guided by data. by answering the mountain of smaller questions pa may, in fact, answer the biggest question of all: how can improve the effectiveness of all the massive functions across government, healthcare, business, non-profit and law reinforcement work. continuing with the proliferation of terms to name the accumulation of data, the name big data, was first introduced in an article by two researchers from nasa in 1997, and is usually linked to analytics in many contexts. we consider that big data refers to the capture, storage and retrieval of data and analytics to their treatment. analytics has demonstrated tremendous potential to extend and improve human life by facilitating better diagnosis, promoting preventive treatment and introducing a future of personalized medicines (rao and jain, 2013). we assume that we should also have to talk about detection when the object of study deals with the proper functioning of a machine or a system (physical or conceptual). some authors claim that big data, accompanied by advanced analytics, enable organizations to identify meaningful materials and applicable knowledge that are buried in the data, without having to understand its reason. it is a respectable point of view, but antithetical to the scientific tradition that always asks about the why of things, what obtains, in some cases, a satisfactory answer. how the operations research community’s has welcomed analytics? infors considers, almost 4 int. j. prod. manag. eng. (2015) 3(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international companys, r. and ribas, i. http://creativecommons.org/licenses/by-nc-nd/4.0/ as synonyms, operations research, management science and analytics. ors in the header of its newsletter writes “inside or: the science of better at the heart of analytics”. in the courses given by these professional societies to train experts in analytics appear very well known themes: 1. define de client’s problem properly. surfacing the “stakeholders” criteria for quality and action 2. problem structuring and work plans. smart decomposition of complex problems 3. managing a project or team. navigating client interactions. bridging the gap between desired and available data 4. making the case for change and implementation of the analysis and recommendations through persuasive communication. analysing the levers of persuasion. presenting results based on complex analytical work that is, the same that is explained, since long time ago, as phases to follow to develop a study of or. analytics and its treatment of large volumes of data open new possibilities to quantitative methods. inspired by mortenson et al. (2014) we can list some; traditionally the difficulty inherent in the models and applications of or/ms was to have sufficient data to make the results meaningful. currently the situation tends to become reverse and therefore models and applications should be adapted to manage large volumes of data. at the same time, the procedures to be used in hypothesis testing and validation of the models must be defined before this sea of data. the models and techniques for handling them were developed taking into account the limitations of the data. consequently, they must be adapted (or redesigned) for the current situation. as ackoff (1967) indicated, it will be desirable to develop methods for reducing the data (singular value decomposition, principal components, cores or kernels) maintaining the significant insights. some or / ms models are potential producers of big data (e.g. simulation), which until now were not utilized fully due to the difficulty of processing so condensed. circumstances have changed; this may be subjected to a criticism review. working with large volumes of data will be convenient to use the new data architectures, developed for this circumstance. the real-time applications can now harness the wealth of data available. the existing ones must be redesigned, and others new will be able to be designed. moreover, in the interactive processes, it will be interesting to use intensively the display of data. the analytics obviously has many positive aspects, but do not hide the danger of misuse of its possibilities through a comprehensive snooping of all our acts similar to the predicted by orwell (orwell, 1949). however, being optimistic, we are sure that soon it will be designed and disseminated (and applied) some strategies that serve to poison the big brother and to get he leaves us reasonably quiet. 3. exotic heuristics. the progress of science varies inversely with the number of journals published. parkinson’s sixth law, in arthur bloch (1977) most of the problems found when modelling combinatorial production systems are classified as np-hard so it is often used heuristic procedures for resolution. racs and raes (dannenbring, 1977) were two pioneering heuristics to solve flow shop scheduling problems. a first step to take was the neighbourhood definition of a given solution. in continuous functions with the solution defined as a point in a space of n dimensions, the procedure was apparently natural. in the discrete context, with the often solution described as a permutation of n elements, the things are not so clear. however, definitions of neighbourhood were found and two of the most popular were the insertion and swap ones: a transformation or movement allowed moving from one solution to another considered its neighbour and the succession of movements could identify with a path. soon, it became obvious that, in some cases, the objective function associated with these neighbourhoods could not be considered unimodal, consequently, after some movements a better solution could be reached, but not necessarily the optimal. it was a local minimum with respect to the neighbourhood used, and it was desirable to seek ways out of this impasse. the first idea that comes to overcome this difficulty is to perform multiple searches which start in several skilfully chosen points to increase the chances of finding the global optimum. this idea leads naturally to the heuristics named multistart or grasp (feo and resende, 1989). a natural extension is to keep the various paths in parallel with interaction between, them such as genetic algorithms (holland, 1975 and goldberg, 1989) or ant colony algorithms (dorigo, 1992) a second idea is to provide the exploration of searching mechanisms that allow continuing the 5int. j. prod. manag. eng. (2015) 3(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international some trends and applications of operational research/management science to operations management http://creativecommons.org/licenses/by-nc-nd/4.0/ trajectory by escaping from local minima without be trapped in a circle, i.e. systematically visit the same points. this second idea has several aspects; on one hand, it leads to extensions of exhaustive descends (ed) search and non-exhaustive descends (ned) search known as simulated annealing (kirkpatrick, 1983 based on metropolis, 1953) and tabu search (glover, 1986), respectively. on the other hand, to apply a random perturbation to the solution, causing a “jump” in the trajectory in order to move away (but not too much) from the local minimum, as is done in the iterated local search algorithm (ils) (stützle, 1998) or iterated greedy algorithm (iga) (ruiz and stützle, 2005) and finally, since the neighbourhood structure induces that a point be a local minimum, to change the neighbourhood in order to extend the path is on another plane (variable neighbourhood search (vns) (mladenovic and hansen, 1999). it should not hide that several of these mechanisms can be used simultaneously. newer concepts have been ts, sa, ga and aco. ts tries to avoid circuits and improve the trajectory by remembering segments, recent and not so recent, to exploit this knowledge when appropriate. it is, to some extent, imitation of human behaviour. sa is based on a metaphor from metallurgy, and presents the novelty of accepting a movement that leads to a worse solution with some probability, this probability decreases with the amount of worst solution and path length (the latter through a parameter called temperature, which decreases with time). the basic idea of genetic algorithms is older, but we perceive it as a competitor of above schemes. here the metaphor is the natural evolution. it is considered a set of solutions, called population. each solution, called individual or chromosome, has a particular health (fitness) (corresponding to the value of the objective function), and assumes that the better health, the more positive aspects comprise the chromosome and, therefore, there is greater interest in preserving new individuals. given two individuals, properly selected, an operator called crossover allows to generate new individuals that maintain some of the characteristics of the parents. to increase the diversity of the population another operator called mutation is introduced. this one can be identified with what we called above movement, which is applies to certain selected individuals. finally, to keep the population within reasonable limits, a natural selection that conserves only the most promising individuals intervenes. in the aco several agents (ants) build heuristically parallel solutions and after each shift, agents exchange information about the quality of the constructed solutions. there exists a memory that from this information or pheromone assigns a value to each of the elements likely to be part of a solution so that those who have been part of good solutions are more likely to be part of future solutions. from certain point of view, the ants learned through the common experience of the colony. in the early 90s all these new procedures were named metaheuristics. in fact was glover (1986) who used this name to qualify the ts. meta is a prefix comes from the greek μετά and that usually means, in most languages, “after” or “beyond” next to the item to which it is attached as “metaphysics”, “metaphor” or “ metalanguage”. hence, it was surprising to find a classification where metaheuristics are a particular category of heuristics as in diaz and teeng (1996). perhaps the answer lies in the following ambiguous definition “a metaheuristic is a high-level problemindependent algorithmic framework that provides a set of guidelines or strategies to develop heuristic optimization algorithms. that term is also used to refer a problem-specific implementation of a heuristic optimization algorithm according to the guidelines expressed in such framework” (sörensen and glover, 2013). this has reminded us that over 40 years ackoff complained that the nomenclature was not well established “different terms are used to refer to the same thing and the same term is used to refer to different things” he referred to the field of systems, more than 40 years ago, but today, we could applied it to the field of metaheuristics. he concluded “defining concepts is treated by scientist as an annoying necessity to be complete as quickly and thoughtlessly as possible. a consequence of this disinclination to define is often research carried out like surgery performed with dull instruments. the surgeon has to work harder, the patient has to suffer more, and the chances for success are decreased”. we believe that a solution should be adopted. otherwise everything will be qualified as metaheuristic if not as hypermetaheuristic. ant colonies inspired many learners of researcher. glover and laguna were surprised in 1977 that, given their social behavior, no one would have noticed the bees of so much literary tradition (maeterlinck, 1901). they would have done better to leave irony to later since the bees appeared in 2005 (karaboga, 2005), followed by flies, termites, fireflies, worms and many other insects. once the insects were sold out, researchers were not discouraged and continued studying animal biology: kangaroos, sharks, jumping frogs and bats and even more exotic things: 6 int. j. prod. manag. eng. (2015) 3(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international companys, r. and ribas, i. http://creativecommons.org/licenses/by-nc-nd/4.0/ intelligent water drops, cuckoo eggs and jazz. any of these analogies allowed, with a hermetic language, presenting an algorithm labelled “new” which solved better some old problems and new ones invented for the occasion. the reasons for the explosion are many. every guy wants to be recognized as the inventor and owner of a procedure (which provides many citations and references). if the procedure appears promising new (at least the first article about it was published) is easy to make changes that results also in publishable articles: by applying the procedure to another problem, by adding a mechanism taken from another procedure (with the which the adjective “hybrid” can be added), etc.. if the referees still swallowing, this provides more references that make the process more attractive and more attempts to use innovatively. sörensen (2013) has denounced the fact “¼the behaviour of virtually any species of insects, the flow of water, musicians playing together – it seems that no idea is too far-fetched to serve as inspiration to launch yet another metaheuristic. in this paper, we will argue that this line of research is threatening to lead the area of metaheuristics away from scientific rigor¼”. there is a very old story concerning an emperor who two clever swindlers who pose as weavers and tailors, sell a non-existent costume by convincing both the emperor and his courtiers, that the fabric was invisible to those who were in a position greater than their merits. the best known version is the tale by andersen (1837). the story ends when the emperor wearing his supposed clothing, followed by his court makes a parade before the people and an innocent child throws the cry “the emperor has no clothes”. sörensen, like the child in andersen’s tale, has shouted “the emperor has no clothes”. will it happen as in this story? we believe that in the near future the designers of heuristic algorithms will abandon the current tendency to use the same scheme for all problems (which favours the publication of articles and cross references that increase the impact of publications) to design heuristics “to measure”, i.e. adapted to the specific problem they aim to solve and they will investigate what is the reason because certain structures and operators perform well in certain problems and less so in others. thanks to the analytics, experimentation of various alternatives for a given algorithm will be able to find the most suitable version with credible guarantees of confidence. a new field very promising as sörensen indicates are the matheuristics, which tries to combine exact algorithms with heuristics. the resulting procedures generally use exact algorithms for solving subproblems that guide the behavior of heuristic algorithms, thus leveraging the best of both worlds. on the other hand, heuristics will be developed to find solutions in real time of many situations, by establishing a proper balance between volumes of data, the complexity of the algorithm, the quality of the solution and the time available. 4. the lean manufacturing lean manufacturing is a production philosophy that considers the expenditure of resources in any aspect other than direct creation of value for the end customer to be wasteful, and thus a target for elimination. therefore, the lean movement leads to highlight what adds value (which the customer is willing to pay) and ignore the rest. lean manufacturing is a philosophy that comes mostly from the toyota production system (tps). the japanese production systems had to be rebuilt after the war during the american occupation following occidental patterns, with support and advice of renowned specialists such nationality such as deming and juran, but later evolved in its original form. there are specific social characteristics of japan and perhaps the level of automation of japanese companies is differential, but that does not explain its high efficiency. it should be included, among these causes, not only the technical production management, especially the management of materials, but also the very concept of productive activity. the occidental rationalism has sought a formalization of procedures that cause the japanese industrial success (which, for some time, was considered somewhat mysterious) and this has led to the lean manufacturing. in a document obtained in the 80s of last century, toyota defined the final motivation of its focus on the design and management of the production system in cost reduction by ruthless elimination of all waste (considered waste all that does not add value to the product) and the maximum use of the capabilities of operators (and not just their hands). the document added that the toyota production system is based on two procedures: the just-in-time and jidoka. in addition, with an added conviction, belonging to the oriental culture, that fully achieve the goal marked requires a continuous improvement effort that never ends (kaizen). toyota distinguished 7 large groups of inefficiencies or “muda”: 7int. j. prod. manag. eng. (2015) 3(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international some trends and applications of operational research/management science to operations management http://creativecommons.org/licenses/by-nc-nd/4.0/ 1. due to overproduction. 2. due to downtime. 3. due to transport. 4. due to inadequate processes. 5. due to stocks. 6. due to unproductive movements. 7. due to defects in the product. although lean manufacturing is seen as an evolution of tps to apply to other sectors and other different environments than the automotive one, some authors point out some differences, which, in our opinion, we do not seem so significant: 1. lean focuses the emphasis on value, rather than waste. the value is produced by the manufacturer but appreciated by the customer. lean thinking starts with a conscious attempt to define, accurately, the value in relation to specific products, with specific prices for established markets. 2. lean identifies the value stream, i.e. the set of all actions required to go from raw material to finish product, with the concept of value from the point of view of the customer. 3. tool orientation is a tendency in many programs to elevate mere tools (standardized work, value stream mapping, visual control, etc.) to an unhealthy status beyond their pragmatic intent. the tools are just different ways to work around certain types of problems but they do not solve them for you or always highlight the underlying cause of many types of problems. no one tool can do all of improvements. but, despite the alleged differences, it remains valid the tps traditional aspects such as tpm, smed, staff participation, quality circles, etc. as it can be seen in recent articles (shah and ward, 2007; elmaraghy and deif, 2014; chiarini, 2014) lean philosophy, with its emphasis on maintaining a continuous flow along what we call supply chain, instead of focusing on the nodes (plants or warehouses) in isolation, represents a necessary change in the way to raise the management in planning short and medium term, in the sizing of buffers, in scheduling, in dispatching (either with kanban or not), in monitoring, etc. this impact will gradually move to the quantitative methods and techniques that support to such management, which must be reviewed, and in some cases designed again, to which will contribute all the support provided by the big data, analytics and new heuristics. 5. supply chain supply chain was a term used by keith oliver, a consultant at booz allen hamilton, in an interview for the financial times on june 4, 1982. regardless of which were behind the name, the term took hold and, during the 1990s, appeared many papers that used it. oliver said in 1982: “supply chain management (scm) is the process of planning, implementing, and controlling the operations of the supply chain with the purpose to satisfy customer requirements as efficiently as possible. supply chain management spans all movement and storage of raw materials, work-in-process inventory, and finished goods from point-of-origin to point-of-consumption”. oliver does not properly define supply chain, he defines supply chain management, whereas chopra and meindl do it in an acceptable form “a supply chain consists of all parties involved, directly or indirectly, in fulfilling a customer request. the supply chain not only includes the manufacturer and suppliers, but also transporters, warehouses, retailers, and customers themselves. within each organization, such as a manufacturer, supply chain includes all functions involved on receiving and filling a customer request. these functions include, but are not limited to new product development, marketing, operations, distribution, finance and customer service”. in any case, these definitions remind us to the definition of logistics. magge (1968) defined: “logistics is the process of monitoring and managing the flow of materials and products from its sources to its point of consumption”. in all definitions of logistics, this is interpreted as a highly focused management activities to manage the flow of materials, but, what is the system on which the logistical management is applied? virtually, in the same system than the supply chain management applies, i.e. in the supply chain. why the concept of supply chain is important? probably, because defines the object of concern regardless of what one wants to do with it, which is the mainly difference from logistics. surely, we will want to govern, through effective management to pursue certain goals, but this is supplementary. we could use again ackoff and his insistence on the importance of names and definitions. 8 int. j. prod. manag. eng. (2015) 3(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international companys, r. and ribas, i. http://creativecommons.org/licenses/by-nc-nd/4.0/ one of the most important challenges in supply chain management is obviously the global or integrated planning, and much more difficult to perform as more organisms, with a high degree of autonomy, are present in the chain. we assume that management may not be completely centralized, because in that case the existence of the role of various organizations would have no palpable effect and the supply chain would not represent any contribution to the traditional management of production systems such as had been developed. if existing organizations in the chain have a certain degree of autonomy accompanied by some degree of power the question that is relevant is how the harmonious functioning of the whole system can be achieved in order to share, satisfactorily the obtained performance to all agents in order to motivate their collaboration. most of the studies published so far treat this topic very partially. we believe that in the future should be deepened in some aspects related to the theory of games and that the equilibrium solutions will be a very interesting field to develop (see yue and you, 2014). moreover schemes based on the establishment of a comprehensive planning framework within which the member organizations establish their own plans harmonically must deepen and extend the traditional decompositions of dantzig-wolfe and benders, or similar techniques in context of wider programming. the three factors mentioned as guidelines for the future development of quantitative methods, which will serve as support for operations management (complexity, risk and sustainability), begin to appear in papers that revolve around the supply chain. the supply chain is already a more complex object than the traditional production system and this complexity increases when considering the reverse logistics. the authors distinguish between three types of complexity: static, dynamic and decisionmaking (serdarasan, 2013). static (structural) complexity is associated with the structure of the supply chain, the variety of its components and strengths of interactions. dynamic (operational) complexity is associated with the uncertainty in the sc and involves the aspects of time and randomness. decision-making complexity is associated with the volume and nature of the information that should be considered when making a supply sc related decision. the three complexity types are interrelated, and they should not be considered in isolation. it would be interesting to have a measure of the complexity of a system in order to evaluate the efficiency of various actions thereon. sivadaran et al. (2006) uses the concept of entropy developed by shannon (1948). a similar scheme is used by isik (2010). with less amplitude cheng et al. (2014) measure the static or structural complexity through entropy, but we consider that more important than the static concept of structural complexity is the dynamic operational complexity in which there is still a long way to go. there are three generic approaches when dealing with complexity in the sc: complexity reduction, complexity management, and complexity prevention. this reminds us the three things that, according to ackoff (1981), can be done with problems: “there are three kinds of things that can be done about problems – they can be resolved, solved or dissolved. to resolve a problem is to select a course of action that yields an outcome that is good enough, that satisfices (satisfies and suffices)…to solve a problem is to select a course of action that is believed to yield the best possible outcome, that optimizes…to dissolve a problem is to change the nature, and/or the environment, of the entity in which it is imbedded so as to remove the problem. problem dissolvers idealize rather than satisfice or optimize…” the steps to follow when dealing with the complexity in the sc are: to classify the complexity in necessary and unnecessary, eliminate or reduce unnecessary complexity, manage complexity and take the necessary measures to prevent the emergence of more unnecessary complexity. the use of tools based on or / ms will be very helpful to manage the necessary complexity. the concept of risk associated overall to the supply chain is a concept that still remains vague, ambiguous and enemy of all quantification, as concluded by the documented work by heckmann et al (2014). one possibility would be to adapt the concept of system reliability to the sc, which continues to be a system, but the complexity of sc from the real world complicates its practical application. therefore most authors consider the notion of risk in parts. the policy of sustainability and environmental friendliness also is connected, as could not be otherwise, to the guidelines for the design and management of the supply chain (winter and knemeyer, 2013) both govindan et al. (2013) as brandenburg et al. (2014) conducted two separate analyzes of the literature from which detected some gaps that provide opportunities for research, including the interrelationship between sustainability and green in the sc, new procedures in consideration of uncertainty, multiobjective approaches etc. 9int. j. prod. manag. eng. (2015) 3(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international some trends and applications of operational research/management science to operations management http://creativecommons.org/licenses/by-nc-nd/4.0/ 6. or in healthcare industry we believe that the healthcare industry is an industry where the application of the methods of om and or techniques has more room to go because their actual use has been, so far, rather anecdotal. in usa and uk, where there is more tradition in om / or, these begin to be used to provide solutions to various problems of management. proof of this is that we recently received an issue of the industrial engineer journal where appears, on the back cover, an advertisement of a simulation software designed specifically to healthcare which serves to avoid risks when making decisions in an environment changing. perhaps the little tradition of using these techniques is due to poor knowledge that the administrators of healthcare institutions have of them and due to entry barriers for professionals belonging to other industrial sectors. however, the pressure exerted by governments on these institutions for being more efficient is forcing them to seek new ways to reduce costs and to address management problems. in this way, one of the areas where hospitals have significant opportunities for improvement is in the management of the entire supply chain, from planning to programming. or techniques can be used as support for both strategic and tactical decisions. in the field of strategic decisions, demand forecasting to meet the capacity requirements, decide locations to serve the greatest number of people, assess the needs of the departments by simulation or by queue models. at the tactical level, or can help to set levels of stock of drugs, budget allocation to a set of resources, allocation of medical equipment, among others. and, for short-term decisions in the field of monitoring and control, or allows resource scheduling, patient, operations…the reader can find a more comprehensive variety of optimization problems and the techniques used in the research being conducted in this sector by rais and viana (2010). but, as evidenced in brailsford and vissers (2011), most of the published papers are researches and models proposed that have not been implemented in the real world. according to the authors, this is due, inter alia, to “academics need to publish in peer-reviewed journals and must therefore demonstrate theoretical or methodological advances. this tends to lead to complex, sophisticated mathematical models which can take years to develop, in stark contrast with the objective of the end-user: a simple, easy-to-use model”. another area with significant opportunities is the processes improvement where the lean tools are often applied to address this problem. implementing lean in healthcare started around 2002 and most of these projects have occurred in the usa, 57%, then the uk, 29%, followed by australia, 4% (de souza, 2009). these, and other lean healthcare projects have achieved some great results, for example reduction in waiting times, increased quality by reducing errors, increased employee’s motivation and customer satisfaction (radnor et al, 2012). the process improvement approach focus on three areas, first defining the value from the patient point of view, mapping value streams and create continues flow by eliminating waste (bozena, 2010) the main challenges in implementing lean healthcare projects are, as in other industrial sectors, getting the managers involved in the lean transformation, the lack of communication between divisions, the identification of the customer as there are many end users in health and the patient not always the one that pays for the service and the lack of understanding of lean methods within the hospitals which often results in poorly implemented lean tools and techniques in the healthcare industry. finally, the use of analytics in healthcare will play a great role in the healthcare system. analytics can have numerous applications. for example, it can help to develop predictive models to forecast patient behaviour and provide preventive care, can help to compare the cost and effectiveness of interventions and treatments or can improve the response in front of disasters by having real-time data on the availability of critical resources. however, according to ward et al (2014), some challenges need to be overcome. these include overcoming privacy concerns, collecting high quality data and making it available or developing data standards to facilitate the extraction of information from the system, among others. 7. conclusions. this paper tries to establish trends in the application of quantitative techniques to operations management. we have not been exhaustive as this would have led to a long list of problems and techniques, without sufficient space for developing them. we have chosen to focus on five aspects of different nature that we consider crucial in the current and future developments of the subject. 10 int. j. prod. manag. eng. (2015) 3(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international companys, r. and ribas, i. http://creativecommons.org/licenses/by-nc-nd/4.0/ in the heuristics section we noted a handicap to the true development of the subject, the complacency of the authors to establish sophisticated analogies without contributions of real value. however, we believe that the designers of heuristics will abandon the current tendency to use the same scheme for all problems to design heuristics adapted to the specific problem to deal with and that they will investigate the reason because certain structures and operators functions perform better in some problems and worse in others. moreover, a new field very promising will be the matheuristics, which tries to combine exact algorithms with heuristics. in analytics we have indicated some positive aspects: the possibility of capturing, storing and processing large amounts of data that open a wide range of possibilities (next to some dangerous aspects). in the section of lean management, we have noted that it is a management philosophy which possibly will be a paradigm that must accommodate all aspects of operations management and therefore, the supporting quantitative tools. with the sc concept, we have defined the system in which most of the developments of or / ms take place and we have pointed that the most important challenge in supply chain will be the coordination between the organisms belonging to it. finally, to illustrate the precedent points, we include the healthcare industry where the om is acquiring a high degree of development. references abbas, a. e. (2014). teaching decision-making with social networks. or/ms today, 40(6). ackoff, r.l. (1967). management misinformation systems. management science, 14(4): 147-156. doi:10.1287/mnsc.14.4.b147 ackoff, r. l., sasieni, m. w. (1968). fundamentals of operations research. wiley ackoff, r. l. (1981). the art and science of mess management. interfaces, 11(1): 20-26. doi:10.1287/inte.11.1.20 andersen, h. ch. (1837). keiserens nye klaeder. bloch, a. (1977). murphy’s law and other reasons why things go wrong. price/stern/sloan publishers bowman, e., fetter, r. (1957). analysis for production and operations management. irwin. brailsford, s., vissers, j. (2011). or in healthcare: a european perspective. european journal of operational research, 212(2): 223-234. doi:10.1016/j.ejor.2010.10.026 brandenburg, m., govindan, k., sarkis, j., seuring, s. (2014). quantitative models for sustainable supply chain management: developments and directions. european journal of operational research, 233(2): 299-312. doi:10.1016/j.ejor.2013.09.032 buffa, e. s. (1961). modern production management. wiley cheng, c., chen, t., chen, y. (2014). an analysis of the structural complexity of supply chain networks. applied mathematical modelling, 38(9-10): 2328-2344. doi:10.1016/j.apm.2013.10.016 chiarini, a. (2014). sustainable manufacturing-greening processes using specific lean production tools: an empirical observation from european motorcycle component manufacturers. journal of cleaner production, 85(15): 226-233. doi:10.1016/j.jclepro.2014.07.080 chopra, s., meindl, p. (2001). supply chain management. strategy, planning and operation. pearson prentice-hall dannenbring, d. g. (1977). an evaluation of flow-shop sequencing heuristics. management science, 23(11): 1174-1182. doi:10.1287/ mnsc.23.11.1174 de souza, l. b. (2009). trends and approaches in lean healthcare. leadership in health services, 22(2): 121-139. doi:10.1108/17511870910953788 diaz, b. a., tseng, f. t. (1996). introducción a las técnicas heurísticas, in diaz, b. a. (ed) optimización heurística y redes neuronales, paraninfo, capítulo 1: 19-36 elmaraghy, h., deif, a.m. (2014). dynamic modelling of impact of lean policies on production levelling feasibility. cirp annals – manufacturing technology, 63(1): 389-392. doi:10.1016/j.cirp.2014.03.108 forrester. j. (1961). industrial dynamics. mit press glover, f. (1986). future paths for integer programming and links to artificial intelligence. computers and operations research, 13(5): 533549. doi:10.1016/0305-0548(86)90048-1 govindan, k., soleimani, h., kannan, d. (2013). reverse logistics and closed-loop supply chain: a comprehensive review to explore the future. european journal of operational research, 240(3): 603-626. doi:10.1016/j.ejor.2014.07.012 heckmann, i., comes, t., nickel, s. (2015). a critical review on supply chain risk – definition, measure and modeling. omega, 52: 119132. doi:10.1016/j.omega.2014.10.004 11int. j. prod. manag. eng. (2015) 3(1), 1-12creative commons attribution-noncommercial-noderivatives 4.0 international some trends and applications of operational research/management science to operations management http://dx.doi.org/10.1287/mnsc.14.4.b147 http://dx.doi.org/10.1287/inte.11.1.20 http://dx.doi.org/10.1016/j.ejor.2010.10.026 http://dx.doi.org/10.1016/j.ejor.2013.09.032 http://dx.doi.org/10.1016/j.apm.2013.10.016 http://dx.doi.org/10.1016/j.jclepro.2014.07.080 http://dx.doi.org/10.1287/mnsc.23.11.1174 http://dx.doi.org/10.1287/mnsc.23.11.1174 http://dx.doi.org/10.1108/17511870910953788 http://dx.doi.org/10.1016/j.cirp.2014.03.108 http://dx.doi.org/10.1016/0305-0548(86)90048-1 http://dx.doi.org/10.1016/j.ejor.2014.07.012 http://dx.doi.org/10.1016/j.omega.2014.10.004 http://creativecommons.org/licenses/by-nc-nd/4.0/ holt, c. c., modigliani, f., muth, j. f., simon, h. a. (1960). planning production, inventories and workforce, prentice-hall isik, i. (2010). an entropy-based approach for measuring complexity in supply chains. international journal of production research , 48(12): 36813696. doi:10.1080/00207540902810593 karaboga, d. (2005). an idea based on honey bee swarm for numerical optimization. technical report-tr06, erciyes university, engineering faculty, computer engineering department, 10. maeterlinck, m. (1901). la vie des abeilles. magee, f. (1968). industrial logistics. mcgraw-hill. mortenson, m. j., doherty, n. f., robinson, s. (2015). operational research from taylorism to terabytes: a research agenda for the analytics age. european journal of operational research, 241(3): 583-595. doi:10.1016/j.ejor.2014.08.029 nestler, s., lewis, j., klimack, b. (2012). certified analytics professional-informs prepares to launch first-of-its-kind program aimed at a variety of practicing professionals. or/ms today, 39(5). orwell, g. (1949). nineteen eighty-four. secker and warburg, london petrovic, s., mccarthy, b. (2014). operational research versus operations management. inside o. r., 525: 26-27 poksinska, b. (2010).the current state of lean implementation in health care: literature review. quality management in health care, 19(4): 319-329. doi:10.1097/qmh.0b013e3181fa07bb radnor, z. j., holweg, m., waring, j. (2012). lean in healthcare: the unfilled promise? social science & medicine, 74(3): 364-371. doi:10.1016/j.socscimed.2011.02.011 rais, a., viana, a. (2010). operations research in healthcare: a survey. international transactions in operational research, 18(1): 1-31. doi:10.1111/j.1475-3995.2010.00767.x rao, h., jain, d. (2013). the evolution of analytics. or/ms today, 40(6): 30-32. serdarasan, s. (2013). a review of supply chain complexity drivers. computers & industrial engineering, 66(3): 533-540. doi:10.1016/j. cie.2012.12.008 shah, r., ward, p. t. (2007). defining and developing measures of lean production. journal of operations management, 25(4): 785-805. doi:10.1016/j.jom.2007.01.019 shannon, c. e. (1948). the mathematical theory of communication. the bell technical journal, 27: 379-423, 623-651 shannon, c.e., weaver, w. (1971). the mathematical theory of communication. university of illinois press siegel, e. (2013). predictive analytics. wiley s. sivadasan, s., efstathiou, j., calinescu, a., huatuco, l. h. (2006). advances on measuring the operational complexity of supplier-customer systems. european journal of operational research, 171(1): 208-226. doi:10.1016/j.ejor.2004.08.032 sörensen, k., glover, f. (2013). metaheuristics, in glass, s. and fu, m. (eds) encyclopedia of operations research and management science, 3rd ed. springer. doi:10.1007/978-1-4419-1153-7_1167 sörensen, k. (2015). metaheuristics – the metaphor exposed. international transactions in operational research, 22(1): 3-18. doi:10.1111/ itor.12001 ward, m.j., marsolo, k.a., froehle, c.m. (2014). applications of business analytics in healthcare. business horizons, 57(5): 571-582. doi:10.1016/j.bushor.2014.06.003 winter, m., knemeyer, a. m. (2013). exploring the integration of sustainability and supply chain management: current state and opportunities for future inquiry. international journal of physical distribution & logistics management, 43(1): 18-38. doi:10.1108/09600031311293237 yue, d., you, f. (2014). game-theoretic modeling and optimization of multi-echelon supply chain design and operation under stackelberg game and market equilibrium. computers and chemical engineering, 71: 347-361. doi:10.1016/j.compchemeng.2014.08.010 12 int. j. prod. manag. eng. (2015) 3(1), 1-12 creative commons attribution-noncommercial-noderivatives 4.0 international companys, r. and ribas, i. http://dx.doi.org/10.1080/00207540902810593 http://dx.doi.org/10.1016/j.ejor.2014.08.029 http://dx.doi.org/10.1097/qmh.0b013e3181fa07bb http://dx.doi.org/10.1016/j.socscimed.2011.02.011 http://dx.doi.org/10.1111/j.1475-3995.2010.00767.x http://dx.doi.org/10.1016/j.cie.2012.12.008 http://dx.doi.org/10.1016/j.cie.2012.12.008 http://dx.doi.org/10.1016/j.jom.2007.01.019 http://dx.doi.org/10.1016/j.ejor.2004.08.032 http://dx.doi.org/10.1007/978-1-4419-1153-7_1167 http://dx.doi.org/10.1111/itor.12001 http://dx.doi.org/10.1111/itor.12001 http://dx.doi.org/10.1016/j.bushor.2014.06.003 http://dx.doi.org/10.1108/09600031311293237 http://dx.doi.org/10.1016/j.compchemeng.2014.08.010 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2014.2326 received 2014-04-30 accepted 2014-06-05 enterprise modeling in the context of enterprise engineering: state of the art and outlook vernadat, f.b. laboratory for industrial engineering, production and maintenance (lgipm), university of lorraine, ile du saulcy, metz f-57042 cedex 1, france. francois.vernadat@eca.europa.eu abstract: enterprise modeling is a central activity in enterprise engineering which can facilitate production management activities. this state-of-the-art paper first recalls definitions and fundamental principles of enterprise modelling, which goes far beyond process modeling. the cimosa modeling framework, which is based on an event-driven process-based modeling language suitable for enterprise system analysis and model enactment, is used as a reference conceptual framework because of its generality. next, the focus is on new features of enterprise modeling languages including risk, value, competency modeling and service orientation. extensions for modeling collaborative aspects of networked organizations are suggested as research outlook. major approaches used in enterprise modeling are recalled before concluding. key words: enterprise engineering, enterprise modeling, process modeling, capability/competency modeling, risk modeling, value modeling, collaborative networked organization, cimosa . 1. introduction nowadays, companies are facing drastic competition, unstable business conditions and serious efficiency problems. they must rationalize and optimize their daily operations in a productive and cost-effective way. they must be reactive and agile to quickly face changing conditions. their operations must be highly interoperable with the partners’ ones. to achieve this, these operations must be properly defined, described and put under control. thus, precise and up-to-date models of these operations are necessary to understand how they work or are organized. enterprise modeling (em) is concerned with representing and describing the structure, the organization and the behavior of a business entity to evaluate its performances, reengineer its various internal and external flows or optimize them in order to make the enterprise more efficient and effective. a business entity is whole or part of an enterprise or of a group of enterprises (e.g., an extended enterprise, a virtual enterprise, a networked enterprise or a supply chain). over the last two decades, em has proved to be a central activity in enterprise engineering and enterprise integration projects. enterprise engineering (ee) deals with design or redesign of business entities (kosanke & nell, 1997). it concerns all activities, except enterprise operations, involved in the enterprise life cycle, i.e., mission identification, strategy definition, requirements definition, conceptual design, implementation description, installation, maintenance and continuous improvement as defined in geram (ifac-ifip task force, 1999). it mostly focuses on engineering and optimizing business processes of enterprises in terms of their related flows (namely, product/ material flows, information/decision flows and control flows), resources (human agents, technical agents, components) as well as time and cost aspects. hence, em techniques for ee must cater for the representation and analysis of function, information, resource and organization aspects of business entities (amice, 1993). they can also cover cost/economic, performance or collaboration aspects. enterprise integration (ei) is concerned with breaking down organizational barriers to create a synergistic whole to improve competitiveness and sustain growth 57int. j. prod. manag. eng. (2014) 2(2), 57-73creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2014.2326 mailto:francois.vernadat@eca.europa-eu http://creativecommons.org/licenses/by-nc-nd/4.0/ of an enterprise or an networked enterprise. the goal is to make interoperable all elements of the enterprise (i.e., humans, machines as well as it applications) to facilitate system-wide the 4c’s (communication, co-operation, co-ordination and collaboration). an enterprise can be considered to be integrated when the right information is delivered at the right place at the right time. enterprise interoperability is therefore an essential enabler of ei (vernadat, 1996, 2007a). the paper provides a state of the art review of enterprise modeling in the context of ee and ei and is organized as follows. first, definitions and fundamental principles of enterprise modeling are recalled. the cimosa modeling framework, which is based on an event-driven process-based modeling language suitable for enterprise system analysis and model enactment, will be used as a reference conceptual framework. next, the focus is on new features that need to be covered by enterprise modeling languages including risk, value, capability and competency aspects as well as service orientation. extensions to model collaborative aspects of networked organizations are also discussed as research outlook. a short panorama of tools is then given followed by the conclusion. 2. enterprise modeling principles and cimosa 2.1. enterprise modeling definition enterprise modeling (em), not to be confused with process modelling (curtis et al. 1992, dalal et al. 2004), can be defined as the art of developing models, i.e., abstract representations of a definite part of an enterprise (in a more or less formal way), to accurately represent the structure, behavior and organization of a business entity. it is a generic term which covers the set of activities, methods and tools related to developing models for various aspects of an enterprise or a network of enterprises (vernadat, 1996; owen & walker, 2013; wikipedia, 2013). enterprise models can be used in practice to represent, visualize, understand, communicate, design, rengineer and improve enterprise operations with a focus on quality, cost or delays as well as system efficiency and effectiveness. especially, these models are useful (but should not be limited) to: understand and analyze the structure and behavior of an enterprise domain, reengineer a part of the enterprise, evaluate the behavior and performances of business processes before their implementation (either in terms of cycletime or cost), choose the best solution among various implementation alternatives (‘what-if” scenarios), evaluate implementation risks and costs, optimize resource selection and management, support modelbased integration or support continuous process improvement. however, the prime advantage of em in industry is to provide a shared view or “picture” of the enterprise that can be communicated to the various actors, i.e., to build a consensus that enforces a common enterprise culture. enterprise modeling techniques equally apply to industrial firms, service companies, administrative organizations or even government agencies. 2.2. cimosa early em methods, i.e., built before the 90’s, were mostly activity centric and based on the functional decomposition principle (e.g., the idef and grai methods). at the turn of the 90’s, business processcentric methods emerged advocating causal and precedence relationships among activities and object flows. the modeling framework of cimosa, an open system architecture for integrated manufacturing enterprises (amice, 1993), paved the way in this field by introducing an event-driven process-based modeling approach and formalizing the concept of business process (jorysz & vernadat, 1990). in addition to the usual functional and information modeling aspects covered by early em methods, it soon became obvious that resource and organization aspects also had to be addressed to properly assess business processes and related concepts as found in manufacturing companies or in industrial supply chains. because cimosa has been the root for geram and several european and iso standards for em, e.g. iso 19439 (2006) and iso 19440 (2007), it is used as a reference in the paper. 2.3. fundamental principles of em due to the complexity and multi-faceted nature of enterprise organizations and especially industrial organizations, enterprise modeling frameworks should respect the following principles: principle #1: plural nature of enterprise models. this means that there is no such thing as an “enterprise model”. there are enterprise models. indeed, any business entity, be it a manufacturing plant, a r&d department, a branch of a company, a supply chain or 58 int. j. prod. manag. eng. (2014) 2(2), 57-73 creative commons attribution-noncommercial-noderivatives 4.0 international vernadat, f.b. http://creativecommons.org/licenses/by-nc-nd/4.0/ a virtual enterprise, is so complex that it is impossible to represent it by one single model expressed in one language. several models will be necessary and, indeed, the enterprise model is an assemblage of submodels, each depicting some specific aspects. principle #2: concept of modeling views. the concept of modeling view or viewpoint is a mechanism that allows to focus on some aspects of a system while discarding others to manage structural complexity. a modeling framework will be powerful, complete and consistent if it provides a minimal set of non-overlaping views to cover all essential aspects of the system. in enterprise modeling, four basic views have been defined by cimosa and adopted in geram and iso 19439, namely: the function/process view, which defines the enterprise functionality (i.e., what has to be done) and the enterprise behavior (i.e., in which order work has to be done). the information/object view, which defines what are the objects to be processed and to be used and what are their states over time. the resource/infrastructure view, which describes who/what does what, what is needed to execute operations, which roles, capabilities and skills are required or available. the organization/decision view, which describes organization units and decision levels of a business entity and their relationships, who is responsible for what or whom, who decides and who has authority on what. other views can be defined but iso 19439:2006 and geram have retained the four cimosa views as the essential ones. figure 1 provides a common illustration of principles 1 and 2. figure 1. illustration of the concept of modeling view. principle #3: three fundamental types of flows. there are three fundamental types of flows circulating within or across any type of enterprises (excluding financial flows): material flows (made of physical objects such as raw materials, semi-finished parts, products, components, tools…), information flows (made of information and decision objects such as orders, documents, data, computer files, emails, phone calls…), control flows (or workflow, i.e., logical execution sequences of tasks). principle #4: processes versus agents. at a macro-level, any enterprise can be viewed as a large collection of: concurrent processes (called business processses) executed to achieve business goals and competing for resources on one hand, and interacting agents, or functional entities (i.e., human or technical resources), executing processes on the other hand. the art of management is to make sure that in fine the agents (i.e., the doers) execute the processes in an efficient, effective and economic way to achieve business objectives. principle #5: business process synchronization. there are three fundamental types of business process synchronization: 1. event-based synchronization: events in the form of messages, requests, orders or timers can be generated by one process and used to trigger other processes (e.g., ev2 in figure 2). 2. object-based synchronization: a step in a process control flow may need availability of objects produced by one or more steps of different processes. 3. resource-based synchronization: execution of a step in a process may need availability of some resource(s) that can also be needed by another step in another process (this is a resource conflict that needs to be solved by means of priority rules or a conflict resolution mechanism). figure 2 illustrates the three possible process synchronization mechanisms as well as the three types of flows (principles 3 and 4). 59int. j. prod. manag. eng. (2014) 2(2), 57-73creative commons attribution-noncommercial-noderivatives 4.0 international enterprise modeling in the context of enterprise engineering:state of the art and outlook http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 2. process synchronization schemes. principle #6: the concept of modeling levels. enterprise models can be developed at three generic levels as proposed in cimosa, geram and iso 14258 (1998). these are: requirements definition: used to represent “the voice of the users”, i.e., what is needed, expressed in a detailed and unambiguous way in a user-orientated or descriptive language. design specification: used to formally specify one or more solutions satisfying the set of requirements, to analyze their properties and to select the “best” one. these models are expressed by means of a formal and computer executable language (prone to model validation, performance analysis or system simulation). implementation description: used to state in detail the implementation solution taking into account physical technical constraints. operational models are defined at this level. due to the complexity of models handled in practice (both in terms of number of components and number of relationships existing among these components), a modular and incremental modeling approach is often recommended. to achieve this, most modeling languages use a building block approach (i.e., they are made of a limited set of constructs, defined as object classes or template structures (see appendix) which can be assembled in a “lego” fashion) and support the definition of libraries of partial models (i.e., predefined, reusable modules which can be customized and put together) to develop the particular model of a specific business entity. 2.4. simple illustrative example the example used in the paper concerns a simple functional domain made of a business process dealing with customer order processing. figure 3 depicts a process process_customer_orders triggered by the newcustomerorder event and made of three process steps (enter_order, create_ customer and process_order). each step employs a number of sub-steps or functional operations as indicated. these operations are performed by two functional entities (orderentryclerk being a human agent and orderprocessing system being a computer application). figure 3. a simple business entity example. 3. fundamental em constructs to follow the cimosa and iso 19440:1997 presentations, essential constructs of em for enterprise engineering are presented according to the four basic modeling views introduced in section 2.3. the discussion is based on (vernadat, 1998) and templates for the constructs are given in the appendix of the paper. 3.1. function/process view the goal is to describe the enterprise functionality and the enterprise behavior, i.e., what has to done and in which order. essential constructs are domains, events, business processes, enterprise activities and functional operations interpreted as follows: a functional domain is a cluster of business processes which are triggered by occurrences of events and which control the execution of a sequence of enterprise activities to achieve enterprise business goals. activities are made of functional operations. 60 int. j. prod. manag. eng. (2014) 2(2), 57-73 creative commons attribution-noncommercial-noderivatives 4.0 international vernadat, f.b. http://creativecommons.org/licenses/by-nc-nd/4.0/ a domain is a functional area of the enterprise. it must not be confused with an organizational unit or department. it is used to break down the enterprise into manageable functional areas to cope with system complexity. in that sense, domains must not overlap. examples of domains could be r&d functions, marketing, production planning, production, change management, industrial process improvement, etc. domains are made of a set of consistent business processes. the golden rule is that when a process is assigned to a domain, it must be fully contained in this domain and must not span over several domains. interactions among domains are called domain relationships and consist of exchanges of events or materialized objects. an event depicts a change in the state of the system. it is a fact, a happening, solicited or not, that happens at a given time and that implies an action to be taken (i.e., the start of a process or activity chain). examples of expected events can be the arrival of an order, the sending of a request, the end of an activity or even a timer, i.e., a clock-time (e.g., 5:00 pm). examples of unexpected events are a fire alarm, the breakage of a machine, an accident, etc. events can be endogeneous, i.e., generated within the context of the enterprise (e.g., by its resources or its activities) or exogeneous, i.e., generated by the environment of the enterprise (e.g., customers, suppliers, insurance companies, banks, revenue & taxation offices, etc.). a business process can be defined as a partially ordered sequence of process steps executed to fulfil some business goals of the enterprise. process steps are either elementary (enterprise activities) or composite (sub-processes). end-to-end processes fully contained in a domain are called domain processes. business processes describe the enterprise behavior, i.e., in which sequence things are done. a process must have a start point (start), defining the process trigeering condition, and an end point (finish), defining the results to be achieved by the process. all steps in between must be modeled as part of the process. occurrences of processes are triggered by occurrences of events. however, the triggering conditions of processes (or start conditions) can be much more complex. they can be made of a single event (occurrence of the event will trigger an occurrence of the process), they can be a combination of events using boolean logical operators (e.g., arrival of a customer order and 8:00 am) or they can be a logical combination of events and boolean conditions (e.g., start process_customer_orders and non-empty (customerorder file)). the enterprise behavior in a process is described by the sequencing of the process steps. for instance, in cimosa, this is done in the form of when (condition) do action rules, where the condition part of the rule describes the triggering condition defined above and the action part indicates the next step to be activated. the formalism used is based on the eca (event[condition]/action) rules in which the [condition] clause is optional (chakravarthy, 1989). with this formalism, it is possible to describe the following basic situations in a control flow: pure sequence, conditional branching, parallelization (synchronous or asynchronous), rendez-vous (synchronous or asynchronous) and the feedback loop (constructed with a go-to). for instance, using the business process template provided in the appendix, the description of the process of the example of figure 3 is: business process identifier: bp-031 name: process_customer_orders description: used to process customer orders objectives: process customer orders within 24 hours constraints: process orders in arrival order declarative rules: operates from 8:00 am until 5:00 pm triggering events: ev-001/newcustomerorder process behavior: when (start with newcustomerorder and past (8:00am) and before (5:00pm)) do enter_order when (es(enter_order) = new customer) do create_customer when (es(enter_order) = known customer) do process_order when (es(create_customer) = customer created) do process_order when (es(process_order) = done) do finish ending statuses: done comprises: enter_order, create_customer, process_ order where-used: nil convention: the name of business processes and enterprise activities must always start with an action verb (because they represent functionality, the nature of which is “to do”). an enterprise activity is the locus of action. it transforms inputs x into outputs y using time t and a set of resources r. it can be modeled by a transfer function fr such that y = fr (x, t), where fr depicts the activity behavior. fr can be an algorithm, a procedure, a recipe or a protocol. it models enterprise functionality. 61int. j. prod. manag. eng. (2014) 2(2), 57-73creative commons attribution-noncommercial-noderivatives 4.0 international enterprise modeling in the context of enterprise engineering:state of the art and outlook http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 4 provides generic representations of the activity. figure 4a is the original representation proposed by ross (1977) in sadt while figure 4b is the cimosa representation which generalizes the sadt box. figure 4. sadt and cimosa activity representation. the vectors of inputs x and ouputs y are made of manifestations of enterprise objects called object views and representing objects in a certain state. three types of inputs and three types of outputs of an activity can be distinguished as follows: function input (fi): these are all the objects processed or modified by the activity. control input (ci): these are information objects constraining the execution of the activity but not modified by the activity. resource input (ri): these are all the objects used in support to the execution of the activity. function output (fo): these are all the objects created or transformed by the activity. control output (co): these are the events generated by the activity, if any. resource output (ro): returns the status of resource objects at the end of the activity. for instance, the representation of the activity enter_order of figure 3 using the enterprise activity template of the appendix will be: enterprise actvity identifier: ea-001 name: process_order description: get order details and check data objectives: process an order in less than 1 hour constraints: process orders in arrival order declarative rules: operates from 8:00 am until 5:00 pm required capabilities: cs-001/op_capabilities inputs: function: ov-002/customer_order, ov-022/customer control: ov-032/product_file, ov-033/customer_file resource: fe-010/orderentryclerk outputs: function: ov-003/checked_customer_order control: nil resource: nil activity behavior: begin fe-010.review-order (ov-002), fe-010.check-order (ov-002), fe-010.check-customer (ov-022, ov-033) end ending statuses: {known customer, new customer} where-used: bp-031/ process_customer_orders the required capabilities indicate the set of capabilities expected from the resources needed to execute the activity. the ending statuses are all possible states in which the activity can finish its execution (also called output or intermediate events in other modeling languages). finally, enterprise activities can be further broken down into functional operations to explicit the activity behavior, i.e., how their transfer function is actually implemented. in the previous example, the activity process_order is made of three sequential operations. a functional operation (fo) is an atom of functionality in the model in the sense that it corresponds to one instruction interpretable and executed by one resource object, called a functional entity (fe). it must be denoted by a verb. using a dotted notation, it can be specified as: fe-name.fo-name (argument list) fe-name is the name of the functional entity requested to execute the operation, fo-name is the label of the operation and argument list is list input and output parameters passed to or returned by the operation during execution. 3.2. information view the information view describes inputs and outputs of enterprise activities in the form of objects. more precisely, these inputs and outputs are objects in certain state at a certain time, hereby called object views. the various states of the enterprise objects can be modeled in the form of state-transition diagrams, in which states are the object views and transitions are activities transforming an input state into an output state. these diagrams are also called object state transition networks (ostn) in some methods. 62 int. j. prod. manag. eng. (2014) 2(2), 57-73 creative commons attribution-noncommercial-noderivatives 4.0 international vernadat, f.b. http://creativecommons.org/licenses/by-nc-nd/4.0/ a typical example concerns a machining process in which a part is sequentially machined by several machines. for instance, the blank part is first mounted on a milling machine for surface milling. then, the milled part is mounted on a drilling machine to be drilled. finally, the drilled part goes to a third machine for surface finishing to produce the finished part. the part is the enterprise object and “blank part”, milled part”, “drilled part” and “finished part” are four object views of the part object. according to cimosa and iso 19440, the two fundamental constructs of the information view are: enterprise object and object view. templates proposed for these constructs are given in the appendix. an enterprise object is any entity of the enterprise having a life cycle of its own. it must be uniquely identifiable (i.e., it has an identity) and is described by a set of descriptive properties. each property is defined by a name and a data type that can itself be an object or a set or list of objects. two abstraction mechanisms borrowed from the information system modeling theory can be used to represent semantic links among enterprise object classes: (1) the generalization/specialization hierarchy (or is-a link) to relate a super-class to its more specialized classes, and (2) the aggregation mechanism (or part-of link) to relate a compound class to its component classes. these are the same mechanisms as those found in object class diagrams of the unified modeling language or uml (http://www.uml.org/). an object view is a manifestation or state of an enterprise object or a combination of enterprise objects at a certain time. usually, an object view is defined over one enterprise object as a projection over the set of properties of the object, i.e., whole or part of the properties of the object (some properties can be derived or calculated properties from the properties of the object) are used in the view. this would be the case for the example of the machined part just mentioned. however, an object view can be a manifestation constructed over several enterprise objects. a typical example is the customer order. a customer order is not an entity of the enterprise in its own right because it depends on two other enterprise objects. indeed, any customer order is the relationship between a customer (an enterprise object) with a product (another enterprise object) ordered in some quantity on a certain date. the object view construct template therefore distinguishes between the so-called leading object (the main object on which the object view is defined) and related objects (the other enterprise objects providing properties of the object view). last but not least, two fundamental types of object views must be distinguished: physical object views and information object views. physical object views represent the physical manifestations of the object, i.e., the ones that can be situated in shape, space and time. information object views are purely descriptive and are usually computer or database representations of the objects. this distinction of nature is essential to separately represent the physical or material flows from the information flows as defined in section 2.3 (see figure 2). as a matter of example, the description of a customer order object class could be: enterprise object identifier: eo-002 name: customer_order description: orders received from company customers is-a: eo-001/order part-of: nil properties: date: date customer: eo-100/customer billing address: address delivery address: address list of items: list [1,30] of eo-101/order_item 3.3. resource view the resource view is intended to describe all enterprise objects used in support to the execution of enterprise activities (i.e., technical agents or human agents). it can also be used to describe the physical infrastructure of the enterprise. ontologically speaking, cimosa distinguishes active resources from passive resources. active resources, called functional entities, are resources that can receive and send messages and can execute functional operations. therefore, they have a controller device or function that can interpret the functional operation received as a command in the form of a message in a certain format transported by some protocol via a communication channel or media. passive resources, called components, are any kind of resources other than functional entities that cannot execute any functional operation (e.g., a tool, a cart, a truck, a telephone, etc.). functional entities can be further partitioned into three fundamental sets: humans, machines and applications. humans are all kinds of human agents. machines are machines with a controller (e.g., robots, 63int. j. prod. manag. eng. (2014) 2(2), 57-73creative commons attribution-noncommercial-noderivatives 4.0 international enterprise modeling in the context of enterprise engineering:state of the art and outlook http://creativecommons.org/licenses/by-nc-nd/4.0/ cnc machines, a truck with its driver equipped with a gsm or cellular phone, a printer connected to a computer network, etc.). applications are it systems or computer applications including database systems. each category can be further decomposed. to model resources and their charateristics, at least two essential constructs are needed: resource and capability set. templates of these constructs are given in the appendix. the resource construct will be used to model either functional entities or components as well as aggregations of these in the form of resource sets or cells as found in manufacturing systems. the capability set construct is used to model additional specific aspects of the resource, usually in the form of technical constraints, and dealing with functional capabilities, object related capabilities, performance capabilities of operational capabilities of the resource. from an ontological point of view, an active resource or functional entity is an enterprise object fundamentally defined by its non-empty list of functional operations and the list of capabilities that it can provide. an object view defined on the resource object can be used to describe additional intrinsic characteristics of the resource such as availability, capacity, location, reliability (mttr, mtbf…), mono or multi-tasking capability, priority rules, etc. as an example, let us consider the description of the orderentryclerck, a human functional entity in figure 3, together with its capability set. resource identifier: fe-010 name: orderentryclerk description: employee processing orders class: functional entity cardinality: 1 object view: ov-010/oeclerk_characteristics capabilities: cs-123/oeclerk_capabilities operations: review-order, check-order, check-customer, create-customer, create-daily-report made-of: nil where-used: nil capability set identifier: cs-123 name: oeclerk_capabilities description: capabilities for order entry & verification capabilities: functional: to receive customer orders, to check customer data, to check orders, to create customer records, to make daily reporting object-related: can deal with regular oders, special orders and sub-contracting orders performance: order processing time < 60 mn operational: must produce a daily report before leaving the office 3.4. organization view the organization view is used to describe the organizational structure (i.e., administrative and functional structures) of the enterprise as a collection of organizational entities (or decision centers) related to one another either by hierarchical (i.e., decisional) links or co-lateral (i.e., network) links. an organization entity may be a workposition, a work-station, a team, a department, a division, a plant or a branch, a direction or the entire enterprise. organizational entities are usually assigned responsibilities and authorities on tasks or processes, systems, data or resources and people of the enterprise. cimosa and iso 19440 provide two basic constructs for the organization view (see appendix): organization unit and organization cell. an organization unit is a single decision centre reduced to one functional entity assigned to a specific job (defined by its job description) and provided with well-defined responsibilities and authorities. a responsibility is a duty for which a person is accountable (e.g., the order o1 of customer x must be delivered by the end of the week) while an authority is a right (e.g., to hire someone, to sign a contract, to delete data in corporate databases, etc.). an organization cell is an aggregation of organization units and/or cells to form a higher level decision center in the organization structure. it is placed under the management of one functional entity (a person, who is the manager) and has a set of well-defined responsibilities and authorities on components or people of the enterprise. this way, jobs can be aggregated in teams, teams in departments, departments in divisions, divisions in directions and directions in the entire enterprise to reflect the organization chart of the enterprise. using the “belongs to”, “comprises” and “related to” links of the two construct templates, it is possible to describe any kind of enterprise organizations, be it a single enterprise, a multi-branch enterprise or a networked enterprise. 4. em language extensions in the previous sections, the essential constructs for em as understood all over the 90’s have been recalled. since then, a number of extensions to capture more information or to cover new aspects have been proposed. they are reviewed in this section. 64 int. j. prod. manag. eng. (2014) 2(2), 57-73 creative commons attribution-noncommercial-noderivatives 4.0 international vernadat, f.b. http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.1. competency/skill modeling human aspects were poorly addressed in the cimosa, geram or iso 19440 frameworks. moreover, human agents, considered as a specific type of functional entites, are modeled somewhere in the resource view, aside to machines and applications in the enterprise architecture (see section 3.3). pera (purdue enterprise reference architecture), another enterprise architecture framework for industrial engineering (williams, 1992), enforced and positioned the human architecture at the heart of the framework. to capture more human-related aspects in the model, harzallah & vernadat (2000) first proposed to add the job profile field in the organization unit construct. the job profile (or job description) is defined by a list of skills or competencies that the job holder must have. the job holder for this organization view is indicated in the functional entity field (must be a person). next, the capability set construct has been expanded to also cope with competency or skill description of a human agent. indeed, technical agents (i.e., machines and applications) are described in terms of their capabilities while human agents can be described by their capabilities and competencies. thus, the capability set construct as been augmented with a competencies section to become the capability/ competency set construct presented in the appendix. to generalize these concepts, a formal model of individual competencies has been proposed by harzallah et al. (2006). using this new template, the capability/competency set of the orderentryclerk human resource of figure 3 becomes: capability/competency set identifier: cs-123 name: oeclerk_capabilities description: capabilities for order entry & verification capabilities: functional: to receive customer orders, to check customer data, to check orders, to create customer records, to make daily reporting object-related: can deal with regular oders, special orders and sub.contracting orders performance: order processing time < 60 mn operational: must produce a daily report before leaving the office competencies: knowledge: advanced order processing training, college accounting diploma know-how: minimum 2-years experience in customer order processing, know how to create customer records, know how to check customer data, know how to produce daily reports know-whom: to be organized, to be meticulous, to be service oriented, to meet deadlines, high integrity 4.2. risk and value modeling the aim of a business process is to achieve or fulfil a business goal. should this goal have been rated as reasonably or highly valuable for the enterprise, its achievement must be monitored in terms of performance management. industrial performance can be measured and assessed from four main dimensions: benefits, costs, value and risk. therefore, risk management and value management are gaining ever increasing importance in industrial engineering. indeed, value reflects the expectations with regard to goal or objective satisfaction while risk reflects the concerns/fears of not achieving the goal or objective. value is very much related to stakeholders’ satisfaction (including usual quality, cost and delay performance criteria). in other words, the concept of value is based on the relationship between satisfying needs and expectations and the resources required to achieve them. the aim of value management is to reconcile all stakeholders’ views and to achieve the best balance between satisfied needs and resources (ivm, 2014). according to iso 31000 (2009), risk is the effect of uncertainty on objectives, whether positive or negative. risk management is the identification, assessment, and prioritization of risks followed by coordinated and economical application of resources to minimize, monitor and control the probability and/or impact of unfortunate events or to maximize the realization of opportunities. on the basis of this understanding of value and risk, shah et al. (2014) have proposed a conceptual valuerisk model based on enterprise modeling concepts that can support decision-making and performance management in manufacturing systems. value and risk are analyzed at the enterprise activity level and can be aggregated to the business process level. the proposed conceptual value-risk model is depicted by figure 5. it can be interpreted as follows: an activity i has to achieve objectives that will be materialized by results (output(s) of activity i). the activity is subject to one or more risk factors that need to be analyzed and evaluated. risk judgements will result on how well risks have been mitigated and value judgements will result on how good or bad objectives have been met. the conceptual model can be complemented by a quantitative model to compute aggregated values of risk and value at the process level. to quantify the process risks rp, the risk assessment approach of larson & kusiak (1996) is adopted and 65int. j. prod. manag. eng. (2014) 2(2), 57-73creative commons attribution-noncommercial-noderivatives 4.0 international enterprise modeling in the context of enterprise engineering:state of the art and outlook http://creativecommons.org/licenses/by-nc-nd/4.0/ modified. first, the likelihood of all risk events of event factors multiplied by their corresponding consequences (or severity) is modelled for each activity using equation 1. risk of activity i p c c cij ij q ij ij c t g= + +^ h (1) pij is the probability of a risk event j on activity i and , ,c c cij q ij ij c t are consequences on quality, cost and time objectives, respectively. the global risk for activity i subject to j risk events is given by equation 2: r p ci ij ij obj j 1 j #= = ^ h/ (2) where, ri is the risk magnitude of activity i. however, all risk events are not equally important in any given scenario. some risks are more important than others depending on what objective(s) they are influencing. therefore, a weighting factor or importance index dij is introduced to model the importance of the risk events in the scenario s. so, equation 2 can be rewritten as equation 3: dr p cij ij obj i ijj 1 j #= = ^ h/ (3) therefore, the global risk of the process path pk is given by equation 4: dr p p cr ijk ij ijj 1 j i pi p i kk #= = 66 !! =^ ^h h/// (4) the probability pr (pk)of the path set (or scenario) must verify equation 5: p p 1r kk 1 k = = ^ h/ (5) so, the expected risk of the process p made of k path sets is given by equation 6 which gives the expected risk of the whole process: e rp p dp p cr k ij ij ij j 1 j k 1 k i pk #= 6 ! == ^ ^ ^fh h hp/// (6) to compute an interim value (value at an activity level) using the conceptual model, the output of the activity i is judged with regard to the objective which is appraised by a performance measure, using utility theory principles. as a result, value functions v(c) are developed for each performance measure c at the activity level. the overall value of a process is then obtained by aggregating the interim values of all its activities using an aggregation operator (e.g., weighted sum or choquet integral) (clivillé et al., 2007). similarly, the outputs of activity i are judged using risk measures to develop the interim risk function v(r) and aggregated to obtain the global risk indicator. 4.3. business service orientation the business process approach as practiced throughout the 90’s has introduced a natural and horizontal way in organizing business systems as opposed to previous vertical and activity centric approaches. however, rapidly changing market conditions and business requirements tend to increase the gap between what the business requires and what it can deliver. to build agile, i.e., more flexible, extensible and evolvable environments in which it can be more quickly and easily aligned with the business, many organizations are adopting service-oriented architecture (soa) principles to close the gap (herzum, 2001, vernadat, 2007a). soa is emerging as a new wave for building agile and interoperable enterprise systems. it is an it strategy consisting in exposing as encapsulated services the business functions of an application. broadly speaking, an soa is essentially a collection of services. in technical terms, soa is about designing and building it systems using heterogeneous network addressable software components (preferably communicating over internet). these interoperable standards-based components or services (i.e., callable and reusable functions accessible by their interface) can be directly invoked by business users or executed as steps of business processes. they can be combined, modified, or reused quickly to meet business needs. they can be implemented as web services (ws) or functions of web applications and, therefore, be located anywhere on the web (herzum, 2001; khalaf et al., 2005). from an it perspective, a service is a piece of business functionality that can be invoked by its locator and its interface(s) as publicly published in a standard format. details of its implementation are hidden. the service can be implemented using any it technology (c program, c++, pl/sql, java, .net/c#…). from a business user perspective, a service is an invokable piece of functionality that will return a result (i.e., provide a service to a client) under the conditions defined in its service level agreement (sla). it has higher grain than a ws. practically, business services are functional or informational business components designed to be 66 int. j. prod. manag. eng. (2014) 2(2), 57-73 creative commons attribution-noncommercial-noderivatives 4.0 international vernadat, f.b. http://creativecommons.org/licenses/by-nc-nd/4.0/ accessed as such by service clients or consumers. they represent stand-alone and reusable business functions of the real world (e.g., declare birth of a child, change personal address data or create a customer). compared to a business process, a business service can be one step in a business process, a business process can be a choregraphy of execution of a set of business services and a business service can be implemented by a business process. soa is an it technology that can be used to implement business services. a business service is a discrete piece of functionality that appears to be platform-independent, logically addressable and self-contained from the point of view of the service caller. it must be uniquely identified within the enterprise and has a service owner. logically addressable means that it can be dynamically invoked simply by calling its logical address or universal resource identifier (uri), thus without having to know where it is physically located. self-contained means that the service exists as a whole and that it maintains its own state. in practice, it is recommended to develop only stateless services (i.e., each service call is independent of previous calls), especially if they are implemented as web services. however, stateful services may also exist (they require that both the consumer and the provider share the same consumer-specific context – passed in the message) (vernadat, 2007b; omg, 2012). physically, a business service can either be performed by a human agent or a technical agent. for instance, the create_customer activity could be a stateless business service offered by the customer order processing domain. in this case, the service can either be invoked within the course of the process_customer_orders process or be called as a separate function by raising the create_customer_request event as illustrated by figure 6. figure 6. business process with business service. 5. research outlook: collaborative network organizations enterprise networks and collaborative networked organizations (cnos) are made of a finite set of collaborating entities from different partner companies working together. this implies among other things business process synchronization, sending events, exchanging material or information objects and even, in some cases, sharing resources. a collaboration view has recently been proposed as an extension of em frameworks to describe the context and characteristics of the collaboration from the point of view of each partner (kosanke et al., 2014). indeed, the enterprise architecture of each partner belongs to this partner while the architecture of the whole network does not belong to any specific partner. following the model view concept of cimosa/ iso 14258, this view identifies three new language constructs with their respective elements to globally describe collaboration as commonly depicted in the literature (camarinha-matos & afsarmanesh, 2007; camarinha.matos et al., 2009; jagdev & thoben, 2010). these are (see appendix for construct details): collaboration domain collaborating partner collaboration point the collaboration domain construct is used to describe a given collaboration area between the enterprise at hand (us) and its partner companies identified as its collaborating partners. it indicates the collaboration entities (i.e., processes, activities, resources or organization units described in the other modeling views of the enterprise architecture), the exchanged objects in terms of events and object views as well as the list of collaboration points, i.e., gateways supporting the various exchange flows with the different partners. collaborating partners of the enterprise at hand are business entities involved in the collaborative exchanges with this enterprise (us). they are defined in terms of their role in the collaboration (e.g., supplier, provider, consumer or retailer). a description of their ict environment can be made. collaboration points, as defined by li et al. (2013), represent the collaboration interfaces between collaborating entities of an enterprise and those of one of its collaborating partners. the type of collaboration can be unidirectional or bidirectional, 67int. j. prod. manag. eng. (2014) 2(2), 57-73creative commons attribution-noncommercial-noderivatives 4.0 international enterprise modeling in the context of enterprise engineering:state of the art and outlook http://creativecommons.org/licenses/by-nc-nd/4.0/ synchronous or asynchronous or based on mutual adjustment. the exchange media or transportation means supporting the exchange flows must be specified. a partner enterprise, or one of its branches, can be involved in several collaboration domains. each collaboration domain may comprise several collaboration points. note: the use of these three modeling constructs enables the representation of a collaboration model from the point of view of a given company with reference to the models of its individual partner enterprises by describing its collaboration domains and interacting partners. figure 7 illustrates a collaboration domain between two partner enterprises (abc and xyz) containing two collaboration points (cp1 and cp2) for the exchange of orders and invoices. the model belongs to the architecture of the xyz company. the “+” in functional boxes indicates that they represent business processes. figure 7. collaboration example. 6. panorama of em methods enterprise modeling techniques have their roots in the entity-relationship (er) model of chen (1976) and in sadt (structured analysis and design technique) of ross (1977), both developed for software engineering. sadt has later been renamed idef0 in the 80’s and has provided a generic definition of the activity in the form of the icom (input – control – output – mechanism) box (see figure 4) to the idef method (www.idef. com). sadt/idef0 has formalized the functional decomposition process to break down complex systems into sub-systems (golden rule: a system can be broken down into at least three sub-systems and no more than six) while the er model has formalized the concept of entity in information systems, which has later been extended to the concept of object class in the object-oriented approach. the first real enterprise modeling techniques and languages appeared in the 80’s with the idef method (icam, 1981) and the grai method (doumeingts, 1984) which then became grai-gim (roboam et al., 1992). idef method: the idef (icam definition) method was developed within the framework of the integrated computer-aided manufacturing (icam) program of the us air force and sponsored by the us department of defence (dod) to model large manufacturing systems. it is made of a series of modeling methods: idef0 to model enterprise activities with sadt, idef1 to model data structures in the form of extended entity-relationship models and idef2 to model system behavior using the queueing formalism of the simulation language slam (icam, 1981). unfortunately, it poorly addresses resource aspects, does not cover organizational aspects and cannot model business processes. furthermore, idef2 has been abandoned. idef3: because idef0 can only model enterprise activities and not business processes, the idef3 method has been added to idef in 1992 (mayer et al., 1992). idef3 is a popular method in north america to represent process control flows in manufacturing systems. process steps in a control flow are called units of behavior (uobs) and are connected by junction boxes (and, or or xor boxes). the major strength of idef3 is that it can represent any kind of business processes (well-structured, semi-structured or illstructured) although its syntax and semantic are not precisely defined. grai-gim: it is a methodology and enterprise architecture framework containing enterprise modeling tools, including idef0 for activity modeling. it covers data (i.e., information), process (i.e., function) and operational aspects (mixing resource and organization aspects). the major advantage of grai-gim is its ability to analyze the decision system of an enterprise for which it provides the grai grid to identify decision centers and the grai nets to model operational and decision activities. in addition, grai-gim has been provided with a performance evaluation framework for economic evaluation (doumeingts & vallespir, 1995; doumeingts & ducq, 2001). 68 int. j. prod. manag. eng. (2014) 2(2), 57-73 creative commons attribution-noncommercial-noderivatives 4.0 international vernadat, f.b. http://creativecommons.org/licenses/by-nc-nd/4.0/ aris toolset: aris (architecture for information systems) was originally, as its name suggests, an architecture framework for building it systems (programs and databases). it has then been proposed as an enterprise architecture framework (scheer, 1992, 1999). although made of four views (data, control, function and on top organization), it mostly covers function and information aspects and very partially organization and resource aspects. each view comes with its modeling constructs. especially, business processes are modeled as sequences of events and activities called event-process chains (epc). epc has been a popular business process language because it has been promoted by aris toolset, the modeling tool of aris which is number one sales in the world, and recommended by sap for implementing the sap erp. recently, the epc notation has been replaced by bpmn and its use is declining in industry. uml: although clearly announced as a modeling method for software-based systems by omg (object management group), uml (unified modeling language) can be useful in enterprise modeling, especially for building models of enterprise objects and object views in the information view but also to specify computer and it systems to be implemented in the particular architecture of an enterprise. uml provides object class diagrams, activity diagrams, sequence diagrams and collaboration diagrams that can be useful to analyze, design and specify it components of the enterprise. starting with version 1.3 (omg, 2000), the current version of uml is 2.4 (omg, 2011a). version 2.5 is now available in beta version. bpmn: because there were just too many similar but different and non interoperable formalisms, languages and specific tools to model business processes, there was a need for a de facto standard language or notation. bpmn (business process model and notation), coming from it, has become the esperanto in the field. most commercial systems for modeling business processes provide a bpmn interface, including aris toolset. as the name says, it is a business process description notation to be used to capture process structure and behavior. this is a graphical notation. models are boxes connected by arrows and documented by text. it is however a simple and sufficiently precise notation to build models of process control flows that can easily and quickly be communicated to many stakeholders of the enterprise. bpmn is not an enterprise modeling language. it is a process modeling language which covers only enterprise behavior description (i.e., process control flow and some related elements). originally, proposed by bpmi (business process management initiative), bpmn is now supported by omg. it is widely used in industry. the current version is bpmn 2.0 (omg, 2011b). archimate: archimate (the open group, 2012) is an enterprise architecture modeling language to support the description, analysis and visualization of particular enterprise architectures within and across business domains. it is a de facto standard based on concepts of the ieee 1471 standard intended for describing the architecture of a “software-intensive system”. as such, it has a strong it flavor and can only partially cover the four essential views of enterprise modeling as defined in iso 19439. the aim of archimate is to become an open standard in enterprise architecture to generalize previous modeling frameworks. it is currently gaining popularity. mapping archimate with togaf (www.opengroup.org/togaf/), one of the currently prominent reference architecture frameworks, has been addressed. both togaf and archimate are supported by the open group (http://www.opengroup.org) and find their way in industrial or administrative applications. soaml: originally intended as an open source specification for describing a uml profile and metamodel for the modeling and design of services in a service oriented architecture, soaml (service oriented architecture modeling language) provides a consistent framework for (1) modeling business services and business processes and (2) for implementing these services and processes with web services or soa technology (omg, 2012). in that sense, it is the perfect framework for implementing the business service orientation presented in section 4.3. 7. conclusion enterprise modeling emerged in the late 80’s as a technique for describing various aspects of an enterprise, especially for the purpose of analysis, design, reengineering, optimization, performance evaluation and even control of a given business entity. at that time, cimosa demonstrated that this can be achieved in an integrated and uniform way by developing modeling language constructs and templates. these templates can then easily be 69int. j. prod. manag. eng. (2014) 2(2), 57-73creative commons attribution-noncommercial-noderivatives 4.0 international enterprise modeling in the context of enterprise engineering:state of the art and outlook www.opengroup.org/togaf http://www.opengroup.org http://creativecommons.org/licenses/by-nc-nd/4.0/ implemented in the form of object classes to be stored in relational or object databases and used by advanced modeling and simulation tools. this what most ea and em commercial tools now do. since then, em has evolved and has been constantly enriched with new constructs to capture more details or cover additional aspects. in the paper, fundamental principles and constructs of em have been first reminded before presenting recent extensions and a short panorama of well-known modeling techniques to make a state of the art. further developments can still be proposed to cover additional aspects. we can mention soft issues (for instance, trust or human behavior), security and legal aspects in collaborative environments or value in networked enterprises, just to name a few. references amice. (1993). cimosa: open system architecture for cim, 2nd revised and extended edition. berlin: springer-verlag. 234 pages. camarinha-matos, l.m., afsarmanesh, h. (2007). a comprehensive modeling framework for collaborative networked organizations. journal of intelligent manufacturing, 18(5): 529-542. doi:10.1007/s10845-007-0063-3 camarinha-matos, l., afsarmanesh, h., galeano, n., molina, a. (2009). collaborative networked organizations concepts and practice in manufacturing enterprises. computers & industrial engineering, 57(1): 46-60. doi:10.1016/j.cie.2008.11.024 chakravarthy, u.s. (1989). rule management and evaluation: an active dbms perspective. acm sigmod record, 18(3): 20-28. doi:10.1145/71031.71034 chen, p.p.s. (1976). the entity-relationship model: toward a unified view of data. acm transactions on database systems, 18(1): 9-36. doi:10.1145/320434.320440 clivillé, v., berrah, l., mauris, g. (2007). quantitative expression and aggregation of performance measurements based on the macbeth multi-criteria method. international journal of production economics, 105(1): 171-189. doi:10.1016/j.ijpe.2006.03.002 curtis, b., kellner, m.i., over, j. (1992). process modeling. communications of the acm, 35(9): 75-90. doi:10.1145/130994.130998 dalal, n.p., kamath, m., kolarik, w.j., sivaraman, e. (2004). toward an integrated framework for modeling enterprise processes. communications of the acm, 47(3): 83-87. doi:10.1145/971617.971620 doumeingts, g. (1984). méthode grai : méthode de conception des systèmes en productique. phd thesis, university of bordeaux i, france. 519 pages (in french). doumeingts, g., vallespir, b. (1995). a methodology supporting design and implementation of cim systems including economic evaluation. in p. brandimarte & a. villa, eds. optimization models and concepts in production management (pp. 307-331). new-york, ny: gordon and breach science publishers. doumeingts, g., ducq, y. (2001). enterprise modelling techniques to improve efficiency of enterprises. production planning & control, 12(2): 146-163. doi:10.1080/09537280150501257 harzallah, m., berio, g., vernadat, f.b. (2006). analysis and modeling of individual competencies: toward better management of human resources. ieee trans. on systems, man, and cybernetics, 36(1): 187-207. doi:10.1109/tsmca.2005.859093 harzallah, m., vernadat, f. (2000). validation of modelling constructs for the organisational aspects in cimosa. proc. of ifac/ieee/inria int. conf. on manufacturing control and production logistics (mcpl’2000), grenoble, france, 4-7 july 2000. cd-rom. herzum, p. (2001). web services and service-oriented architecture. cutter consortium, executive report 4, no. 10. icam. (1981). integrated computer-aided manufacturing (icam) architecture, part ii, vol. iv (idef0), vol. v (idef1), vol. vi (idef2). waltham, ma: softech inc. ifac-ifip task force. (1999). geram: generalised enterprise reference architecture and methodology, version 1.6.3. http://www.ict. griffith.edu.au/~bernus/taskforce/geram/versions/geram1-6-3/v1.6.3.html. iso 14258. (1998). industrial automation systems – concepts and rules for enterprise models. geneva, switzerland: international standards organization. iso 19439. (2006). enterprise integration – framework for enterprise modelling. geneva, switzerland: international standards organization. iso 19440. (2007). enterprise integration – constructs for enterprise modelling. geneva, switzerland: international standards organization. iso 31000. (2009). risk management: guidelines on principles and implementation of risk management. geneva, switzerland: international standards organization. ivm. (2014). the institute of value management. http://ivm.org.uk. jagdev, h.s., thoben k.d. (2010). anatomy of enterprise collaborations. production planning & control, 12(5): 437-451. doi:10.1080/09537280110042675 jorysz, h.r., vernadat, f. (1990). cim-osa part i: total enterprise modelling and function view. int. j. of computer integrated manufacturing, 3(3): 144-167. doi:10.1080/09511929008944444 70 int. j. prod. manag. eng. (2014) 2(2), 57-73 creative commons attribution-noncommercial-noderivatives 4.0 international vernadat, f.b. http://dx.doi.org/10.1007/s10845-007-0063-3 http://dx.doi.org/10.1016/j.cie.2008.11.024 http://dx.doi.org/10.1145/71031.71034 http://dx.doi.org/10.1145/320434.320440 http://dx.doi.org/10.1016/j.ijpe.2006.03.002 http://dx.doi.org/10.1145/130994.130998 http://dx.doi.org/10.1145/971617.971620 http://dx.doi.org/10.1080/09537280150501257 http://dx.doi.org/10.1109/tsmca.2005.859093 http://www.ict.griffith.edu.au/~bernus/taskforce/geram/versions/geram1-6-3/v1.6.3.html http://www.ict.griffith.edu.au/~bernus/taskforce/geram/versions/geram1-6-3/v1.6.3.html http://ivm.org.uk http://dx.doi.org/10.1080/09537280110042675 http://dx.doi.org/10.1080/09511929008944444 http://creativecommons.org/licenses/by-nc-nd/4.0/ khalaf, r., curbera, f., nagy, w.a., mukhi, n., tai, s., duftler, m. (2005). understanding web services. in m. singh, ed. practical handbook of internet computing (chap. 27). boca raton, fl: chapman & hall/crc press. kosanke, k., nell, j.g. (eds.). (1997). enterprise engineering and integration: building international consensus. berlin: springer-verlag. doi:10.1007/978-3-642-60889-6 kosanke, k., vernadat, f.b., zelm, m. (2014). means to enable enterprise interoperation: cimosa object capability profiles and cimosa collaboration view, proc. of the 19th world congress of the ifac, cape town, south africa, 24-19 august 2014. larson, n., kusiak, a. (1996). managing design processes: a risk assessment approach. ieee trans. on systems, man, and cybernetics – part a: systems and humans, 26(6): 749-759. doi:10.1109/3468.541335 li, q., wang, z., li, w., li, j., wang c., du, r. (2013). application integration in a hybrid cloud computing environment: modelling and platform. enterprise information systems, 7(3): 237-271. doi:10.1080/17517575.2012.677479 mayer, r.j., cullinane, t.p., dewitte, p.s., knappenberger, w.b., perakath, b., wells, m.s. (1992). information integration for concurrent engineering (iice) idef3 process description capture method report. armstrong laboratory, wright-patterson afb, ohio 45433, altr-1992-0057. omg. (2000). omg unified modeling language specification, version 1.3. object management group. http://www.omg.org/spec/uml. omg. (2011a). omg unified modeling language specification (omg uml), version 2.4. object management group. http://www.omg.org/ spec/uml. omg (2011b). business process model and notation (bpmn), version 2.0. object management group, http://www.omg.org/spec/bpmn. omg. (2012). service oriented architecture modeling language (soaml) specification, version 1.0.1. object management group. http:// www.omg.org. the open group. (2012). archimate 2.1 specification. the open group, http://pubs.opengroup.org/architecture/archimate2-doc/toc.html. owen, s., walker, z. (2013). enterprise modelling and architecture. new dehli, india: ocean media pvt. ltd. roboam, m., zanettin, m., pun, l. (1989). grai-idef0-merise (gim): integrated methology to analyse and design manufacturing systems. int. j. computer-integrated manufacturing systems, 2(2): 82-98. doi:10.1016/0951-5240(89)90021-9 ross, d.t. (1977). structured analysis (sa): a language for communicating ideas. ieee trans. on software engineering, se-3(1): 6-15. doi:10.1109/tse.1977.229900 shah, l.a., etienne, a., siadat, a., vernadat, f. (2014). decision-making in the manufacturing environment using a value-risk graph. journal of intelligent manufacturing, 25, 2. doi:10.1007/s10845-014-0895-6 scheer, a.-w. (1992). architecture of integrated information systems: foundations of enterprise modelling. berlin: springer-verlag. doi:10.1007/978-3-642-97389-5 scheer, a.-w. (1999). aris – business process modeling, 2nd editionberlin: springer-verlag. 218 pages. doi:10.1007/978-3-642-97998-9 vernadat, f.b. (1996). enterprise modeling and integration: principles and applications. london: chapman & hall. 528 pages. vernadat, f.b. (1998). the cimosa languages. in p. bernus, k. mertins and g. schmidt, eds. handbook on architectures of information systems (pp. 243-263). berlin: springer-verlag. reedited in 2006. vernadat, f.b. (2007a). interoperable enterprise systems: principles, concepts, and methods. annual reviews in control, 31(1): 137-145. doi:10.1016/j.arcontrol.2007.03.004 vernadat, f.b. (2007b). reengineering the organization with a service orientation. in c. hsu, ed. service enterprise integration – an enterprise engineering perspective (pp. 77-101). new york, ny: springer science+business media, llc. williams, t.j. (1992). the purdue enterprise reference architecture. charlotte, nc: instrument society of america. wikipedia. (2013). enterprise modeling. http://en.wikipedia.org/wiki/enterprise_modelling. 71int. j. prod. manag. eng. (2014) 2(2), 57-73creative commons attribution-noncommercial-noderivatives 4.0 international enterprise modeling in the context of enterprise engineering:state of the art and outlook http://dx.doi.org/10.1007/978-3-642-60889-6 http://dx.doi.org/10.1109/3468.541335 http://dx.doi.org/10.1080/17517575.2012.677479 http://www.omg.org/spec/uml http://www.omg.org/spec/uml http://www.omg.org/spec/uml http://www.omg.org/spec/bpmn http://www.omg.org http://www.omg.org http://pubs.opengroup.org/architecture/archimate2-doc/toc.html http://dx.doi.org/10.1016/0951-5240(89)90021-9 http://dx.doi.org/10.1109/tse.1977.229900 http://dx.doi.org/10.1007/s10845-014-0895-6 http://dx.doi.org/10.1007/978-3-642-97389-5 http://dx.doi.org/10.1016/j.arcontrol.2007.03.004 http://en.wikipedia.org/wiki/enterprise_modelling http://creativecommons.org/licenses/by-nc-nd/4.0/ appendix: em construct templates note: the following templates have been derived from the formal reference base of cimosa and iso 19440:2007. “xxx” denotes a reference number that uniquely identifies each construct class. convention: reference to another construct in a construct is made by <identifier/name> of the referenced construct. function/process view domain identifier: dm-xxx name: name of the domain description: text domain objectives: business objectives to be fulfilled by this domain domain constraints: applicable constraints business processes: list [1,n] of business processes belonging to this domain boundary: list [1,n] of domain relationships between this domain and other functional domains (defined as binary relationships) object views: list [1,n] of object views used or generated within this domain events: list [1,n] of events used or raised within this domain event identifier: ev-xxx name: name of the event description: text source: “external” or identifier/name of a resource or an enterprise activity triggers: list [1,n] of business processes object view: one-or-zero object view attached to the event predicate: boolean expression defining the happening condition of the event timestamp: time at which the event occurrence will happen business process identifier: bp-xxx name: name of the business process description: text objectives: to be fulfilled by the process constraints: to be met by the process declarative rules: list [0,n] of business rules triggering events: list [0,n] of events process behavior: non-empty set of when (condition) do action rules ending statuses: set [1,n] of ending statuses comprises: list[1,n] of sub-processes or activities used in this process where-used: list [0,n] of processes employing this process enterprise actvity identifier: ea-xxx name: name of the enterprise activity description: text objectives: to be fulfilled by the activity constraints: to be met by the activity declarative rules: list [0,n] of business rules required capabilities: list [1,n] of capability sets defining required capabilities inputs: function: list [0,n] of object views control: list [0,n] of information object views resource: list [1,n] of resources outputs: function: list [0,n] of object views control: list [0,n] of generated events resource: information on resource status activity behavior: {pre-conditions, sequence of functional opeerations, post-conditions} ending statuses: set [0,n] of all possible ending statuses for this activity where-used: list [1,n] od processes employing this activity information view object view identifier: ov-xxx name: name of the object view description: text leading object: one enterprise object related objects: list [0,n] of enterprise objects properties: list [1,n] of object properties in the form: property_name: data type enterprise object identifier: eo-xxx name: name of the enterprise object description: text is-a: zero-or-one enterprise object part-of: list [0,n] of enterprise objects properties: list [1,n] of object properties in the form: property_name: data type resource/infrastructure view resource identifier: fe-xxx or cp-xxx name: name of the resource description: text class: “functional entity” or “component” or “resource set” or “resource cell” cardinality: number of occurrences of this class object view: object view containing descriptive information about the resource (e.g. availability, capacity, location, rtbf, mttr…) capabilities: list [1,n] of capability sets operations: list [1,n] of functional operations defined as: fo-name (in parameters, out parameters) made-of: list [0,n] of sub-resources where-used: list [0,n] of parent resources capability/competency set identifier: cs-xxx name: name of the capability set description: text capabilities: functional: functional capabilities object-related: object related capabilities performance: performance capabilities operational: operational capabilities 72 int. j. prod. manag. eng. (2014) 2(2), 57-73 creative commons attribution-noncommercial-noderivatives 4.0 international vernadat, f.b. http://creativecommons.org/licenses/by-nc-nd/4.0/ competencies: knowledge: theoretical knowledge know-how: acquired skills know-whom: personal traits and behavior organization view organization unit identifier: ou-xxx name: name of the organization unit description: text functional entity: assigned functional entity task description: text job profile: list [1,n] of required skills/competencies responsibilities: list [1,n] of responsibilities authorities: list [0,n] of authorities belongs to: organization cell organization cell identifier: ov-xxx name: name of the organization cell description: text level: “team”, “workstation”, “service”, “department”, “division”, “direction”, entity” manager: human functional entity managing the organization cell responsible for: list [1,n] of model components under the responsibility of the cell manager authority on: list [1,n] of activities, resources or enterprise objects on which the cell manager has authority with indication of the authority type belongs to: upper organization cell comprises: list [1,n] of organization units or cells belonging to this cell related to: list [0,n] of organization cells in relation with this cell collaboration view collaboration domain identifier: cd-xxx name: name of the collaboration domain description: text collaborating partners: list [1,n] of partners collaborating entities: list [1,n] of entities exchanged object views: list [1,n] of object views collaboration points: list [1,n] of collaboration points collaboration partner identifier: pc-xxx name: name of the collaboration partner description: text partner identification: legal name, legal status, location parent entity: name of parent entity, if any partner role: “supplier”, “service provider”, “retailer”, “costumer” or “other” partner ict environment: description text collaboration point identifier: cp-xxx name: name of the collaboration point description: text collaboration partner: collaboration partner collaboration type: asynchronous, synchronous, mutual adjustment exchange flows: list [1,n] of (sender, receiver, flow: list [1,n] of events and/or object views) exchange media: type of media supporting the exchange (e.g. file transfer, e-mail, https, virtual private network, courier, transportation, etc.) 73int. j. prod. manag. eng. (2014) 2(2), 57-73creative commons attribution-noncommercial-noderivatives 4.0 international enterprise modeling in the context of enterprise engineering:state of the art and outlook http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2015.3595 received 2015-02-05 accepted: 2015-05-28 a robust evaluation of sustainability initiatives with analytic network process (anp) lanndon ocampoa and christine omela ocampob a department of mechanical engineering, university of san carlos, cebu city, 6000 cebu, philippines don_leafriser@yahoo.com b department of industrial engineering, university of san carlos, cebu city, 6000 cebu, philippines omehler89@yahoo.com abstract: this paper presents a methodology on evaluating sustainable manufacturing initiatives using analytic network process (anp) as its base. the evaluation method is anchored on the comprehensive sustainable manufacturing framework proposed recently in literature. a numerical example that involves an evaluation of five sustainable manufacturing initiatives is shown in this work. results show that sustainable manufacturing implies enhancing customer and community well-being by means of addressing environmental issues related to pollution due to toxic substances, greenhouse gas emissions and air emissions. to test the robustness of the results, two approaches are introduced in this work: (1) using monte carlo simulation and (2) introducing structural changes on the evaluation model. it suggests that the results are robust to random variations and to marginal changes of the network structure. the contribution of this work lies on presenting a sustainable manufacturing evaluation approach that addresses complexity and robustness in decision-making. key words: analytic network process, evaluation, manufacturing, robustness, sustainability. 1. introduction in sustaining manufacturing industry, purely profitbased strategies became insufficient brought about by various issues that concern environmental degradation, resource depletion, carbon emissions, and social responsibility. these issues are associated with the interests of various stakeholders who are capable of influencing salient decisions of manufacturing firms (pham and thomas, 2012). these stakeholders, which include customers, employees, investors, suppliers, communities and governments (theyel and hofmann, 2012) directly or indirectly compel manufacturing firms to manage the performance of their products and processes in order to satisfy persistent issues on resource depletion, socio-economic concerns and human health problems. when these demands from stakeholders are integrated in mainstream decisionmaking, manufacturing firms could establish long term relations with these stakeholders (harrison et al., 2010). this is believed to be beneficial from the perspective of the manufacturing industry as stakeholders play a crucial role in the sustainability of manufacturing firms (kassinis and vafeas, 2006; paloviita and luoma-aho, 2010). ocampo and clark (2014a) implied that these demands from stakeholders are pushing firms to gear up towards a more holistic concept of the triplebottom-line – a term first coined by elkington (1997) – which interprets sustainability into three main dimensions: environmental stewardship, economic growth and social well-being. labuschagne et al. (2005) claimed that optimal approaches of manufacturing firms towards sustainability are only possible when these three dimensions are taken into consideration. from the perspective of sustainability of manufacturing firm emerges specialized framework popularly known as ‘sustainable manufacturing’ and is defined as the “creation of manufactured products that use processes that minimize negative environmental impact, conserve energy and natural resources, are safe for employees, communities and consumers, and are economically sound” (international trade administration, 2007). operationally, manufacturing firms must: 123int. j. prod. manag. eng. (2015) 3(2), 123-133creative commons attribution-noncommercial-noderivatives 4.0 international http://dx.doi.org/10.4995/ijpme.2015.3595 http://creativecommons.org/licenses/by-nc-nd/4.0/ (1) design and manufacture eco-efficient products with processes that possess minimal environmental footprint using a life cyclce assessment (lca) approach, (2) develop initiatives on cost reduction and return on investment maximization across organizational levels, and (3) maintain programs that enhance well-being of stakeholders (ocampo, 2015). recent studies claim that firms that promote sustainability in their decision-making are more likely to be successful in their respective industries (jayal et al., 2010). among various research domains in this area, evaluation of manufacturing initiatives that promote sustainability is popularly taken (joung et al., 2013; ocampo and clark, 2015). the basis of evaluation is usually anchored on some established indicator sets (jayal et al., 2010; ocampo and clark, 2014b; ocampo and clark, 2015). these indicator sets provide verifiable standards in evaluating products, processes, firm, economic sectors or even countries and regions in the context of sustainable manufacturing (joung et al., 2013). a review of these indicator sets were discussed in mayer (2008), joung et al. (2013), ocampo and clark (2014b), ocampo and clark (2015), ocampo (2015) and will not be repeated here. the challenge of these indicators sets is twofold: (1) being comprehensive, and (2) being operational. a plausible integration of these indicators sets that attempts to cover sustainability areas in great detail was proposed by joung et al. (2013) and this framework was used by ocampo and clark (2014b), ocampo and clark (2015) and ocampo (2015). ocampo (2015) utilized the framework of joung et al. (2013) in index computation to assess sustainability of manufacturing at firm level. ocampo and clark (2015) used the same structure to evaluate sustainable manufacturing of a case firm using analytic hierarchy process (ahp). ocampo and clark (2014b) extended the former evaluation to include causal relationships between criteria and across the decision model using the general analytic network process (anp). despite of these recent works, the specific problem that is advanced in this paper is an evaluation framework that captures complexity and robustness of decision-making in the framework of sustainable manufacturing. this paper extends previous works by embedding robustness in sustainable manufacturing evaluation in the context of the anp. following the argument of ocampo and clark (2014b) on the use of anp, this work imposes such use due to the complexity and multi-dimensionality of the evaluation problem associated with the issues that concern sustainability. developed by thomas saaty, anp generalizes any decision-making problem by overcoming the hierarchic assumption mostly characterized by other decision-making tools (saaty, 2001). the use of anp in sustainable manufacturing evaluation allows comprehensiveness of addressing the complexity inherent in the decision-making process. chen et al. (2012) agreed that ahp and anp are appropriate analytical tools for addressing location, program or strategy selection problems. among various applications that highlight the use of anp include developing sustainability index for a manufacturing enterprise (garbie, 2011), developing multi-actor multi-criteria approach in complex sustainability project evaluation (de brucker et al., 2013), evaluating industrial competitiveness (sirikrai and tang, 2006), evaluating energy sources (chatzimouratidis and pilavachi, 2009), developing an impact matrix and sustainability-cost benefit analysis (chiacchio, 2011), etc. the departure of this work include: (1) evaluating robustness of the results of the evaluation problem and, (2) determining the impact of structural changes of the evaluation problem on the results of the anp. the contribution of this work is on presenting a sustainable manufacturing evaluation approach that addresses complexity and robustness in decision-making. this paper is organized as follows: section 2 presents the methodology of the study. section 3 highlights the evaluation model along with the results of the anp and robustness tests. section 4 provides the discussion and ends with concluding remarks in section 5. 2. methodology the proposed evaluation approach can be generally described in the following procedure: 1. incorporate feedback and dependence relationships on the hierarchical sustainable manufacturing evaluation framework proposed by ocampo and clark (2015). this is presented in the parallel work of ocampo and clark (2014b). the ten sustainable manufacturing initiatives under evaluation were described in the concept paper of ocampo and clark (2014a). although, they attempt to develop an evaluation method following the demands of stakeholders and the triple-bottom line, the approach was not generalizable (ocampo and clark, 2014a). by convention, an arrow that emanates from one component to another component implies 124 int. j. prod. manag. eng. (2015) 3(2), 123-133 creative commons attribution-noncommercial-noderivatives 4.0 international ocampo, l and ocampo, c. o. that the latter influences the former. introducing these dependence relationships is based from theory and practice of sustainability as discussed by ocampo and clark (2014a). 2. based from the resulting network of step 1, corresponding pairwise comparisons matrices are constructed. a detailed discussion on this topic was provided by saaty (2001). in eliciting pairwise comparisons, generally we ask this question: “given a control element, a component (element) of a given network, and given a pair of component (or element), how much more does a given member of the pair dominate other member of the pair with respect to a control element?” (promentilla, et al., 2006). saaty’s fundamental scale (saaty, 1980), as shown in table 1, is used to compare elements pairwise. note that a pairwise comparisons matrix possesses a reciprocal characteristic, i.e. 1 a ji aij = . table 1. saaty fundamental scale (adopted from saaty, 1980). definition explanation 1 equal importance two elements contribute equally to the objective 2 weak between equal and moderate 3 moderate importance experience and judgment slightly favor one element over another 4 moderate plus between moderate and strong 5 strong importance experience and judgment strongly favor one element over another 6 strong plus between strong and very strong 7 very strong or demonstrated importance an element is favored very strongly over another; its dominance demonstrated in practice 8 very, very strong between very strong and extreme 9 extreme importance the evidence favoring one element over another is one of the highest possible order or affirmation determining the priority vector of a pairwise comparisons matrix involves solving an eigenvalue problem in the form aw=λmaxw (1) where a is the positive reciprocal of the pairwise comparisons matrix and w is the principal eigenvector associated with the maximum eigenvalue λmax. saaty (1980) claimed that w is the best estimate of the priority vector of the pairwise comparisons matrix. for consistent judgment, λmax=n; otherwise, λmax>n where n is the number of elements being compared. consistency of judgment is measured using consistency index (ci) and consistency ratio (cr). ci is a measure of the degree of consistency of judgment and is denoted by ci= λmax–n n–1 (2) cr is computed as ci cr ri = (3) where ri is the mean random consistency index. cr≤0.10 is an acceptable degree of consistency (saaty, 1980). otherwise, decision-makers will be asked to reconsider their judgments. 3. form the initial supermatrix based from the network developed in step 1. see saaty (1980) on the discussion of supermatrix. populate this initial supermatrix with the local priority vectors obtained in step 2. then, transform the initial supermatrix to column stochastic supermatrix by normalizing column values such that column sum is unity. finally, raise the stochastic supermatrix to sufficiently large powers until row values become identical. each column of this limiting supermatrix is likewise identical and is known as the global eigenvector of the supermatrix. this is used to describe the overall dominance of the elements in the decision network. 4. to test the robustness of the results, this paper adopted two approaches. first, monte carlo simulation was performed to determine the effect of repeated decisions on the final ranking of results. second, structural changes of the decision network were introduced to evaluate their impact on the final ranking. comparison of the results with the findings of ocampo and clark (2014b) and ocampo and clark (2015) were reported. 3. results the evaluation problem proposed by ocampo and clark (2015) was based from the hierarchical 125int. j. prod. manag. eng. (2015) 3(2), 123-133creative commons attribution-noncommercial-noderivatives 4.0 international a robust evaluation of sustainability initiatives with analytic network process (anp) http://creativecommons.org/licenses/by-nc-nd/4.0/ sustainability indicators set proposed by joung et al. (2013) along with the sustainable manufacturing initiatives discussed by ocampo and clark (2014b). this problem is composed of the goal, the triplebottom line (environmental stewardship, economic growth and social well-being), 10 sub-criteria, 33 attributes and 5 sustainability initiatives. using analytic hierarchy process (ahp), the work was able to assign priority ranking of sustainability initiatives that the case firm must adopt to further promote sustainability. although the dependence relationships were shown in ocampo and clark (2014b), the motivations behind these relationships are discussed in this paper. in this work, the hierarchical structure of joung et al. (2013) was still used while feedback and dependence relations in the criteria and sub-criteria components were introduced. this approach of introducing feedback and dependence relationships in the criteria and sub-criteria components, excluding the attribute component, was done to provide interrelationships at an intermediate level while maintaining hierarchical dependence at lower level. this allows control from upper level decision components to the lower level components. figure 1 shows the evaluation problem and table 2 presents the decision components and elements along with their corresponding codes. the details of this coding system were discussed by ocampo and clark (2014b). as shown in figure 1, attribute component contains no dependence relationships as they only become redundant due to the existing relationships in higher level components. the hierarchical dependence relationships from goal – criteria – sub-criteria – attributes were based from the work of ocampo and clark (2015). note that all decision components have feedback control loop towards the goal component. this is a structural issue as it guarantees that the the goal component takes control over all other components in the evaluation problem. in this paper, pairwise comparisons matrices of the hierarchical dependence relationships from goal – criteria – sub-criteria – attributes were obtained from ocampo and clark (2015). generally, there are three sets or levels of pairwise comparisons matrices performed in this work. first is the dependence relationships among elements in the criteria component and table 3 shows a sample of these ocampo, l. & ocampo, c.o. creative commons attribution-noncommercial 4.0 int. j. prod. manag. eng. (2015) 3(2), ppp-ppp | 3 international license figure 1. decision network of the evaluation of sustainable manufacturing initiatives criteria g a b c sub criteria a1 a2 a3 a4 b3b1 b2 c1 c3c2 initiatives i1 i2 i4 i5i3 attributes a1 a1 a1 a1 a1 a2 a2 a2 a2 a4 a4 a4 a3 a3 a3 a3 b1 b1 b2 b2 b2 b2 b3 b3 c1 c1 c1 c2 c2 c2 c3 c3 c3 figure 1. decision network of the evaluation of sustainable manufacturing initiatives. 126 int. j. prod. manag. eng. (2015) 3(2), 123-133 creative commons attribution-noncommercial-noderivatives 4.0 international ocampo, l and ocampo, c. o. pairwise comparisons matrices. the question being asked in table 3 is: “comparing environmental dimension (a) and economic dimension (b), which one more dominates environmental dimension (a) and by how much?” the resulting priority vector is reported using equation (1). second is the dependence relationships among elements in the sub-criteria component and table 4 shows a sample of these pairwise comparisons matrices. the question being asked in table 4 is: “comparing pollution (a1) and emission (a2), which one more influences the community (c3) and by how much?” the resulting priority vector is again reported. lastly, pairwise comparisons were performed on the hierarchical dependence relationships of sub-criteria to sustainable manufacturing initiatives. table 5 shows a sample of these pairwise comparisons matrices. the question being asked in table 5 is: “comparing health and wellness program (i1) and employee compensation and benefits (i2), which one more characterizes toxic substance (a11) and by how much?” the resulting priority vector is reported. the supermatrix in table 6 is populated by the priority vectors provided by ocampo and clark (2015) on hierarchical dependence relationships of the network model and the resulting vectors obtained in this work. to facilitate discussion, let a, b, c, d and e be the goal, criteria, sub criteria, attributes and initiatives decision components. generally, based from the network presented in fig. 1, the supermatrix can be structured as in table 6. table 3. pairwise comparisons of the dominance of criteria with respect to environmental criterion (a). a a b c priority vector a 1 3 2 0.5396 b 1/3 1 1/2 0.1634 c 1/2 2 1 0.2970 λmax=3.009, cr=0.009 table 4. pairwise comparisons of the dominance of subcriteria with respect to community (c3). c3 a1 a2 a3 a4 priority vector a1 1 2 4 3 0.4673 a2 1/2 1 3 2 0.2772 a3 1/4 1/3 1 1/2 0.0954 a4 1/3 1/2 2 1 0.1601 λmax=4.031, cr=0.012 table 2. decision elements and their codes (adopted from ocampo and clark, 2015). decision components and elements code decision components and elements code decision components and elements code evaluation of sustainable manufacturing g effluent a21 employees health and safety c11 environmental stewardship a air emissions a22 employees career development c12 economic growth b solid waste emissions a23 employee satisfaction c13 social well-being c waste energy emissions a24 health and safety impacts from manufacturing and product use c21 pollution a1 water consumption a31 customer satisfaction from operations and products c22 emissions a2 material consumption a32 inclusion of specific rights to customer c23 resource consumption a3 energy/electrical consumption a33 product responsibility c31 natural habitat conservation a4 land use a34 justice/equity c32 profit b1 biodiversity management a41 community development programs c33 cost b2 natural habitat quality a42 health and wellness program i1 investment b3 habitat management a43 employee compensation and career development i2 employee c1 revenue b11 occupational health and safety i3 customer c2 profit b12 elimination of lead in plating process i4 community c3 materials acquisition b21 lean six sigma initiatives i5 toxic substance a11 production b22 greenhouse gas emissions a12 product transfer to customer b23 ozone depletion gas emissions a13 end-of-service-life product handling b24 noise a14 research and development b31 acidification substance a15 community development b32 127int. j. prod. manag. eng. (2015) 3(2), 123-133creative commons attribution-noncommercial-noderivatives 4.0 international a robust evaluation of sustainability initiatives with analytic network process (anp) http://creativecommons.org/licenses/by-nc-nd/4.0/ table 5. pairwise comparisons of the dominance of sustainable manufacturing initiatives with respect toxic substance (a11). a11 i2 i3 i4 i5 i9 priority vector i2 1 4 2 1/2 4 0.2697 i3 1/4 1 1/3 1/5 1 0.0682 i4 1/2 3 1 1/3 3 0.1688 i5 2 5 3 1 5 0.4252 i9 1/4 1 1/3 1/5 1 0.0682 λmax=5.062, cr=0.014 the supermatrix in table 6 is populated by the priority vectors provided by ocampo and clark (2015) on hierarchical dependence relationships of the network model and the resulting vectors obtained from this work. table 6. blocks of the supermatrix. a b c d e a 1 1 1 1 1 b ba bb 0 0 0 c 0 diag [cb] cc 0 0 d 0 0 diag [dc] i 0 e 0 0 0 dc i note that the first row in the supermatrix which is composed of blocks aa, ab, ac, ad, and ae is a unity vector. this is the representation of the feedback control loop from components to the goal element. block ba, i.e. b dominates a, is a hierarchical dependence relation from goal to criteria component. blocks cb and dc are diagonal matrices resulting from dominance relationships of lower level elements to their parent criteria. cb denotes dominance relations of sub-criteria component to their parent criteria element while dc is the dominance of attributes to their parent sub-criteria. blocks bb and cc denote interdependencies in the criteria and sub-criteria component, respectively. block dc is a hierarchical dependence relation of attribute component to sustainable manufacturing initiatives. identity matrices represented by blocks dd and ee show inner dependence relationships of the elements in the attributes and initiatives components, respectively. null matrices for the rest of the blocks in the supermatrix represent nonexistent feedback and dependence relationships on the elements of decision components. the initial supermatrix is presented in appendix 1. a stochastic matrix is formed by dividing column values of the initial supermatrix with their corresponding column sums. then, the stochastic matrix is raised to large powers until it converges to its cesaro sum. convergence exists if row values are identical. each column is the global priority vector and is used to measure the overall dominance of each element in the supermatrix. priority ranking of elements was performed per decision component. this was obtained by normalizing values per component. table 7 shows the ranking of the elements per component. table 7. priority ranking of decision elements. elements raw vector distributive ranking ideal ranking rank g 0.39578 1 1 1 a 0.06823 0.22986 0.59151 3 b 0.11535 0.38861 1 1 c 0.11325 0.38153 0.98180 2 a1 0.01920 0.10449 0.59689 5 a2 0.02279 0.12408 0.70875 3 a3 0.01337 0.07278 0.41571 8 a4 0.00428 0.02330 0.13308 10 b1 0.01758 0.09568 0.54653 6 b2 0.02279 0.12406 0.70864 4 b3 0.02454 0.13358 0.76305 2 c1 0.01447 0.07875 0.44986 7 c2 0.03216 0.17506 1 1 c3 0.01253 0.06822 0.38967 9 a11 0.00334 0.04495 0.40775 6 a12 0.00334 0.04495 0.40775 6 a13 0.00115 0.01554 0.14097 22 a14 0.00062 0.00837 0.07592 31 a15 0.00115 0.01554 0.14097 22 a21 0.00263 0.03544 0.32152 10 a22 0.00526 0.07089 0.64305 2 a23 0.00263 0.03544 0.32152 10 a24 0.00088 0.01181 0.10717 27 a31 0.00201 0.02702 0.24516 15 a32 0.00067 0.00901 0.08172 30 a33 0.00201 0.02702 0.24516 15 a34 0.00201 0.02702 0.24516 15 a41 0.00107 0.01442 0.13080 26 a42 0.00053 0.00721 0.06540 32 a43 0.00053 0.00721 0.06540 32 b11 0.00330 0.04441 0.40289 8 b12 0.00330 0.04441 0.40289 8 b21 0.00228 0.03071 0.27861 13 b22 0.00228 0.03071 0.27861 14 b23 0.00114 0.01536 0.13930 24 b24 0.00114 0.01536 0.13930 24 b31 0.00409 0.05512 0.50000 3 b32 0.00818 0.11023 1 1 c11 0.00260 0.03509 0.3184 12 c12 0.00087 0.01170 0.1061 28 c13 0.00087 0.01170 0.1061 28 c21 0.00193 0.02600 0.2359 18 c22 0.00386 0.05201 0.4718 5 c23 0.00386 0.05201 0.4718 4 c31 0.00157 0.02111 0.1915 19 c32 0.00157 0.02111 0.1915 21 c33 0.00157 0.02111 0.1915 19 i1 0.00876 0.17697 0.5898 3 i2 0.00701 0.14160 0.4719 5 i3 0.00837 0.16919 0.5639 4 i4 0.01484 0.30004 1 1 i5 0.01050 0.21220 0.70726 2 128 int. j. prod. manag. eng. (2015) 3(2), 123-133 creative commons attribution-noncommercial-noderivatives 4.0 international ocampo, l and ocampo, c. o. in order to test the robustness of these results, two approaches were performed. first, a monte carlo simulation of 500 runs is used to show the impact of randomness on the final results. this is done in a pom for windows application software which is available in public domain. second, structural revisions of the decision network were introduced to assess the impact of dependence relationships on the anp results. in this approach, interdepence relationships of the sub-criteria component were eliminated and then results were subsequently reported. furthermore, all interdependence relationships of criteria and sub-criteria components were removed and results were reported. table 8 summarizes the monte carlo simulation results. it shows that the anp order ranking of i4-i5i1-i3-i2 in decreasing priority is fairly robust after 500 random simulation runs which yield the order ranking of i4-i5-i1-i2-i3 in decreasing priority with rank reversal in the last two initiatives. table 8. comparison with monte carlo simulation results. sustainable manufacturing initiatives anp results monte carlo simulation priority rank priority rank i1 0.18 3 0.16 3 i2 0.14 5 0.15 4 i3 0.17 4 0.12 5 i4 0.30 1 0.29 1 i5 0.21 2 0.28 2 table 9 presents a comparison of anp results with the results from structural changes. it shows that the absence of interdependencies in the subcriteria component changes the ranking of i1 and i3. on the other hand, the complete absence of interdependencies in the decision network changes the top priority, i.e. i5 instead of i4. table 9. impact of structural changes in the decision network. anp results absence of sub-criteria interdepencies complete absence of interdependencies priority rank priority rank priority rank i1 0.18 3 0.17 4 0.17 4 i2 0.14 5 0.14 5 0.15 5 i3 0.17 4 0.18 3 0.18 3 i4 0.30 1 0.26 1 0.25 2 i5 0.21 2 0.25 2 0.26 1 finally, the results of this paper were compared with the results of ocampo and clark (2014b) and ocampo and clark (2015). table 10 highlights the comparison. table 10. comparison of the results. current results with monte carlo simulation ocampo and clark (2015) with ahp ocampo and clark (2014b) with anp rank rank rank i1 3 4 3 i2 4 5 5 i3 5 3 4 i4 1 2 1 i5 2 1 2 table 10 shows that the results of the methodology are not consistent with the results of ocampo and clark (2015) but are fairly consistent with ocampo and clark (2014b). 4. discussion valuable insights could be gained from the results of this paper. anp provides insightful approach in better understanding the evaluation of sustainable manufacturing initiatives. in the criteria component, economic dimension (b) is preferred over social dimension (c) which ranks second and environmental dimension (a) which ranks third. this ranking supports the results of ocampo and clark (2015) with minor differences on the priority weights. economic and social dimensions have almost equal weights which means that manufacturing firms must focus on economic gains and their corresponding social impacts, i.e. welfare of stakeholders which may include employees, customers and community. addressing social issues as results of economic decisions could be achieved via environmental impact on manufactured products and manufacturing processes. this claim is supported by the ranking in the sub-criteria component. customer (c2), investment (b3), emissions (a2), cost (b2), and pollution (a1) are sub-criteria on top priority. the details of this ranking could be examined by taking a look at the priority attributes in the lower level decision component. community development (b32), air emissions (a22), investment to research and development (b31), inclusion of customer rights (c23), customer satisfaction (c22), toxic substance (a11), and ghg emissions (a12) are on top priority in the attribute component. thus, manufacturing decision-making must focus on maximizing revenue and profit by maximizing investment on research and development in technology and investment that contributes community development. investments on community development implies developing 129int. j. prod. manag. eng. (2015) 3(2), 123-133creative commons attribution-noncommercial-noderivatives 4.0 international a robust evaluation of sustainability initiatives with analytic network process (anp) http://creativecommons.org/licenses/by-nc-nd/4.0/ and implementing initiatives that minimize environmental impact of toxic substance, ghg and air emissions. revenue and profit are maximized by reinforcing customer satisfaction strategies and by inclusion of customer rights on manufactured products. developing initiatives that simultaneously enhance customer satisfaction and community development by addressing environmental concerns on toxic substance, ghg emissions and air emissions is fundamentally important to increase revenue and profit. this ranking influences the priority ranking of sustainable manufacturing initiatives. the rank is as follows: elimination of lead in plating process (i4), lean six sigma initiatives (i5), health and wellness program (i1), occupational health and safety (i3) and employee compensation and career development (i2). the first initiative, which is a cleaner production technology, is developed to satisfy customer requirements and at the same time promotes community development through embedding decreased risks associated with occupational sarety and health. cleaner production in a wider scale could promote greater social welfare as the society becomes a direct stakeholder on the environmental issues related to manufactured products and manufacturing process. these results differ marginally with the results of ocampo and clark (2015) using ahp of the same research problem. their results provide less emphasis on environmental impact and greater emphasis on minimizing costs due to the pure independence assumption in the criteria component. when feedback and dependence are taken into account, environmental issues must be addressed to enhance social impact which is vital for sustainability. future research must direct how to develop strategies in designing products and processes that will provide long term benefits to the customer and to the community as well. these results were subjected to test of robustness using monte carlo simulation that attempts to repeat the results over several simulation runs, i.e. 500 runs in this study. results show that these anp results are fairly robust with the exception in the bottom two initiatives. it implies that this priority ranking is dependable and the case firm could use this as an input in prioritizatizing investments, for instance. the absence of interdependence relationships among sub-criteria could also change the ranking except for the first two initiatives. this indicates that the first two decisions are robust enough such that minor changes in the decision model could hardly change their priority ranking. lastly, it is interesting to note that the ranking with complete absence of interdependencies are consistent with the results of ocampo and clark (2015) using ahp. this is due to the inherent structure of the decision network. when interdependencies are removed, the decision network approaches the structure of a hierarchy such that the appropriate methodology becomes the ahp. 5. conclusion this paper demonstrates the use of analytic network process (anp) in evaluating sustainable manufacturing initiatives. the decision problem is structured as a hierarchical network which is built upon the model of ocampo and clark (2014b) and ocampo and clark (2015). results show that cleaner production technologies, i.e. elimination of lead in the plating process, are considered on topmost priority. this work suggests that sustainable manufacturing is achieved by formulating strategies that address issues on customer and community well-being by means of focusing on environmental concerns, e.g. toxic substance, ghg emissions and air emissions. to test the robustness of these results, this work adopts two approaches: (1) using monte carlo simulation, (2) introducing structural changes on the evaluation model. results show that the first two topmost sustainable manufacturing initiatives are robust enough for the case firm to subscribe in these results. future work must focus on formulating specific policies regarding the design of products and processes that could enhance customer and community welfare. acknowledgements we are grateful with the insightful comments from two anonymous reviewers that helped us improve the quality of this paper. references chatzimouratidis, a. i., pilavachi, p. a. (2009). technological, economic and sustainability evaluation of power plants using the analytic hierarchy process. energy policy, 37(3): 778-787. doi:10.1016/j.enpol.2008.10.009 chen, c. c., shih, h. s., shyur, h. j., wu, k. s. (2012). a business strategy selection of green supply chain management via an analytic network process. computers and mathematics with applications, 64(8): 2544-2557. doi:10.1016/j.camwa.2012.06.013 130 int. j. prod. manag. eng. (2015) 3(2), 123-133 creative commons attribution-noncommercial-noderivatives 4.0 international ocampo, l and ocampo, c. o. http://dx.doi.org/10.1016/j.enpol.2008.10.009 http://dx.doi.org/10.1016/j.camwa.2012.06.013 chiacchio, m. s. (2011). early impact assessment for sustainable development of enabling technologies. total quality management and excellence, 39(3): 1-6. de brucker, k., macharis, c., verbeke, a. (2013). multi-criteria analysis and the resolution of sustainable development dilemmas: a stakeholder management approach. european journal of operational research, 224(1): 122-131. doi:10.1016/j.ejor.2012.02.021 elkington, j. (1997). cannibals with forks: the triple bottom line of 21st century business. oxford: capstone. garbie, i. h. (2011). framework of manufacturing enterprises sustainability incorporating globalization issues. in: proceedings of the 41st international conference on computers and industrial engineering, los angeles, ca. harrison, j., bosse, d., philips, r. (2010). managing for stakeholders, stakeholder utility functions, and competitive advantage. strategic management journal, 31(1): 58-74. doi:10.1002/smj.801 international trade administration. (2007). how does commerce define sustainable manufacturing? u.s. department of commerce. available: http://www.trade.gov/competitiveness/sustainablemanufacturing/how_doc_defines_sm.asp. jayal, a. d., baburdeen, f., dillon, o. w. jr., jawahir, i. s. (2010). sustainable manufacturing: modelling and optimization challenges at the product, process and system levels. cirp journal of manufacturing science and technology, 2(3): 144-152. doi:10.1016/j. cirpj.2010.03.006 joung, c. b., carrell, j., sarkar, p., feng, s. c. (2013). categorization of indicators for sustainable manufacturing. ecological indicators, 24: 148-157. doi:10.1016/j.ecolind.2012.05.030 kassinis, g., vafeas, n. (2006). stakeholder pressures and environmental performance. academy of management journal, 49(1): 145-159. doi:10.5465/amj.2006.20785799 labuschagne, c., brent, a. c., van erck, r. p. g. (2005). assessing the sustainability performances of industries. journal of cleaner production, 13(4): 373-385. doi:10.1016/j.jclepro.2003.10.007 mayer, a.l. (2008). strengths and weaknesses of common sustainability indices for multidimensional systems. environment international, 34(2): 277-291. doi:10.1016/j.envint.2007.09.004 ocampo, l. a. (2015). a hierarchical framework for index computation in sustainable manufacturing. advances in production engineering & management, 10(1): 40-50. doi:10.14743/apem2015.1.191 ocampo, l. a., clark, e. e. (2014a). developing a framework for sustainable manufacturing strategies selection. dlsu business & economics review, 23(2): 115-131. ocampo, l. a., clark, e. e. (2014b). a comprehensive evaluation of sustainable manufacturing programs using analytic network process (anp). multiple criteria decision making, 9: 101-122. ocampo, l. a., clark, e. e. (2015). an analytic hierarchy process (ahp) approach in the selection of sustainable manufacturing initiatives: a case of a semiconductor manufacturing firm in the philippines. international journal of the analytic hierarchy process, 7(1): 32-49. doi:10.13033/ijahp.v7i1.223 paloviita, a., luoma-aho, v. (2010). recognizing definitive stakeholders in corporate environmental management. management research review, 33(4): 306-316. doi:10.1108/01409171011030435 pham, d. t., thomas, a. j. (2012). fit manufacturing: a framework for sustainability. journal of manufacturing technology management, 23(1): 103-123. doi:10.1108/17410381211196311 promentilla, m. a. b., furuichi, t., ishii, k., tanikawa, n. (2006). evaluation of remedial countermeasures using the analytic network process. waste management, 26(12): 1410-1421. doi:10.1016/j.wasman.2005.11.020 saaty, t. l. (1980). the analytic hierarchy process. new york: mcgraw-hill. saaty, t. l. (2001). decision making with dependence and feedback: the analytic network process (2nd ed.) pittsburg: rws publications. sirikrai, s. b., tang, j. c. s. (2006). industrial competitiveness analysis: using the analytic hierarchy process. journal of high technology management research, 17(1): 71-83. doi:10.1016/j.hitech.2006.05.005 theyel, g., hofmann, k. (2012). stakeholder relations and sustainability practices of us small and medium-sized manufacturers. management research review, 35(12): 1110-1133. http://dx.doi.org/10.1108/01409171211281255 131int. j. prod. manag. eng. (2015) 3(2), 123-133creative commons attribution-noncommercial-noderivatives 4.0 international a robust evaluation of sustainability initiatives with analytic network process (anp) http://dx.doi.org/10.1016/j.ejor.2012.02.021 http://dx.doi.org/10.1002/smj.801 http://www.trade.gov/competitiveness/sustainablemanufacturing/how_doc_defines_sm.asp http://dx.doi.org/10.1016/j.cirpj.2010.03.006 http://dx.doi.org/10.1016/j.cirpj.2010.03.006 http://dx.doi.org/10.1016/j.ecolind.2012.05.030 http://dx.doi.org/10.5465/amj.2006.20785799 http://dx.doi.org/10.1016/j.jclepro.2003.10.007 http://dx.doi.org/10.1016/j.envint.2007.09.004 http://dx.doi.org/10.14743/apem2015.1.191 http://dx.doi.org/10.13033/ijahp.v7i1.223 http://dx.doi.org/10.1108/01409171011030435 http://dx.doi.org/10.1108/17410381211196311 http://dx.doi.org/10.1016/j.wasman.2005.11.020 http://dx.doi.org/10.1016/j.hitech.2006.05.005 http://creativecommons.org/licenses/by-nc-nd/4.0/ appendix 1. initial supermatrix g a b c a1 a2 a3 a4 b1 b2 b3 c1 c2 c3 a11 a12 a13 a14 a15 a21 a22 a23 a24 a31 a32 a33 a34 a41 a42 a43 b11 b12 b21 b22 b23 b24 b31 b32 c11 c12 c13 c21 c22 c23 c31 c32 c33 i1 i2 i3 i4 i5 g 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 a 0.2000 0.5396 0.2297 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b 0.4000 0.1634 0.6483 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c 0.4000 0.2970 0.1220 0.6000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a1 0 0.3511 0 0 0.667 0 0 0.1601 0 0 0 0.3333 0.6667 0.4673 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a2 0 0.3511 0 0 0.333 1 0 0.095 0 0 0 0.6667 0.3333 0.277 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a3 0 0.1609 0 0 0 0 1 0.277 0 1 0 0 0 0.095 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a4 0 0.1368 0 0 0 0 0 0.467 0 0 0 0 0 0.16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b1 0 0 0.4000 0 0 0 0 0 0.5000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b2 0 0 0.4000 0 0 0 0 0 0.2500 0.7500 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b3 0 0 0.2000 0 0 0 0 0 0.2500 0.2500 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c1 0 0 0 0.2500 0 0 0 0 0 0.297 0 1 0 0.2500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c2 0 0 0 0.5000 0 0 0 0 1 0.54 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c3 0 0 0 0.2500 0 0 0 0 0 0.163 0 0 0 0.7500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a11 0 0 0 0 0.3475 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a12 0 0 0 0 0.3475 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a13 0 0 0 0 0.1201 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a14 0 0 0 0 0.0647 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a15 0 0 0 0 0.1201 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a21 0 0 0 0 0 0.2308 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a22 0 0 0 0 0 0.4615 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a23 0 0 0 0 0 0.2308 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a24 0 0 0 0 0 0.0769 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a31 0 0 0 0 0 0 0.3000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a32 0 0 0 0 0 0 0.1000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a33 0 0 0 0 0 0 0.3000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a34 0 0 0 0 0 0 0.3000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a41 0 0 0 0 0 0 0 0.5000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a42 0 0 0 0 0 0 0 0.2500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a43 0 0 0 0 0 0 0 0.2500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b11 0 0 0 0 0 0 0 0 0.5000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b12 0 0 0 0 0 0 0 0 0.5000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b21 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b22 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b23 0 0 0 0 0 0 0 0 0 0.1667 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b24 0 0 0 0 0 0 0 0 0 0.1667 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b31 0 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b32 0 0 0 0 0 0 0 0 0 0 0.6667 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c11 0 0 0 0 0 0 0 0 0 0 0 0.6000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 c12 0 0 0 0 0 0 0 0 0 0 0 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 c13 0 0 0 0 0 0 0 0 0 0 0 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 c21 0 0 0 0 0 0 0 0 0 0 0 0 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 c22 0 0 0 0 0 0 0 0 0 0 0 0 0.4000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 c23 0 0 0 0 0 0 0 0 0 0 0 0 0.4000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 c31 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 c32 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 c33 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 i1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.2697 0.2000 0.2156 0.2264 0.2889 0.1869 0.1750 0.1250 0.1237 0.1084 0.1084 0.1237 0.1667 0.1429 0.1429 0.1429 0.1199 0.0450 0.1237 0.0655 0.1667 0.2000 0.0780 0.3309 0.3325 0.1367 0.1367 0.1429 0.1237 0.1667 0.1237 0.2000 0.2500 1 0 0 0 0 i2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.0682 0.0780 0.0801 0.0558 0.0727 0.0534 0.0636 0.1250 0.1237 0.1084 0.1084 0.1237 0.1667 0.1429 0.1429 0.1429 0.1199 0.0670 0.1237 0.0655 0.1667 0.2000 0.0780 0.3309 0.0896 0.4030 0.4030 0.1429 0.1237 0.1667 0.1237 0.2000 0.2500 0 1 0 0 0 i3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.1688 0.1255 0.1293 0.4904 0.1228 0.1213 0.1077 0.1250 0.1237 0.1084 0.1084 0.2343 0.1667 0.1429 0.1429 0.1429 0.2101 0.2191 0.1237 0.2500 0.1667 0.2000 0.1342 0.1985 0.3325 0.1367 0.1367 0.1429 0.1237 0.1667 0.1237 0.2000 0.2500 0 0 1 0 0 i4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.4252 0.5459 0.5231 0.0867 0.4429 0.5871 0.5900 0.5000 0.2343 0.4737 0.2011 0.1237 0.3333 0.4286 0.4286 0.4286 0.0706 0.1480 0.2343 0.1094 0.1667 0.2000 0.4882 0.0844 0.2012 0.0791 0.0791 0.4286 0.2343 0.1667 0.3945 0.2000 0.1250 0 0 0 1 0 i5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.0682 0.0507 0.0518 0.1407 0.0727 0.0513 0.0636 0.1250 0.3945 0.2011 0.4737 0.3945 0.1667 0.1429 0.1429 0.1429 0.4795 0.5210 0.3945 0.5096 0.3333 0.2000 0.2215 0.0553 0.0443 0.2444 0.2444 0.1429 0.3945 0.3333 0.2343 0.2000 0.1250 0 0 0 0 1 132 int. j. prod. manag. eng. (2015) 3(2), 123-133 creative commons attribution-noncommercial-noderivatives 4.0 international ocampo, l and ocampo, c. o. g a b c a1 a2 a3 a4 b1 b2 b3 c1 c2 c3 a11 a12 a13 a14 a15 a21 a22 a23 a24 a31 a32 a33 a34 a41 a42 a43 b11 b12 b21 b22 b23 b24 b31 b32 c11 c12 c13 c21 c22 c23 c31 c32 c33 i1 i2 i3 i4 i5 g 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 a 0.2000 0.5396 0.2297 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b 0.4000 0.1634 0.6483 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c 0.4000 0.2970 0.1220 0.6000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a1 0 0.3511 0 0 0.667 0 0 0.1601 0 0 0 0.3333 0.6667 0.4673 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a2 0 0.3511 0 0 0.333 1 0 0.095 0 0 0 0.6667 0.3333 0.277 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a3 0 0.1609 0 0 0 0 1 0.277 0 1 0 0 0 0.095 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a4 0 0.1368 0 0 0 0 0 0.467 0 0 0 0 0 0.16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b1 0 0 0.4000 0 0 0 0 0 0.5000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b2 0 0 0.4000 0 0 0 0 0 0.2500 0.7500 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b3 0 0 0.2000 0 0 0 0 0 0.2500 0.2500 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c1 0 0 0 0.2500 0 0 0 0 0 0.297 0 1 0 0.2500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c2 0 0 0 0.5000 0 0 0 0 1 0.54 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c3 0 0 0 0.2500 0 0 0 0 0 0.163 0 0 0 0.7500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a11 0 0 0 0 0.3475 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a12 0 0 0 0 0.3475 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a13 0 0 0 0 0.1201 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a14 0 0 0 0 0.0647 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a15 0 0 0 0 0.1201 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a21 0 0 0 0 0 0.2308 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a22 0 0 0 0 0 0.4615 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a23 0 0 0 0 0 0.2308 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a24 0 0 0 0 0 0.0769 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a31 0 0 0 0 0 0 0.3000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a32 0 0 0 0 0 0 0.1000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a33 0 0 0 0 0 0 0.3000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a34 0 0 0 0 0 0 0.3000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a41 0 0 0 0 0 0 0 0.5000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a42 0 0 0 0 0 0 0 0.2500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a43 0 0 0 0 0 0 0 0.2500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b11 0 0 0 0 0 0 0 0 0.5000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b12 0 0 0 0 0 0 0 0 0.5000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b21 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b22 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b23 0 0 0 0 0 0 0 0 0 0.1667 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b24 0 0 0 0 0 0 0 0 0 0.1667 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b31 0 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b32 0 0 0 0 0 0 0 0 0 0 0.6667 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c11 0 0 0 0 0 0 0 0 0 0 0 0.6000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 c12 0 0 0 0 0 0 0 0 0 0 0 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 c13 0 0 0 0 0 0 0 0 0 0 0 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 c21 0 0 0 0 0 0 0 0 0 0 0 0 0.2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 c22 0 0 0 0 0 0 0 0 0 0 0 0 0.4000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 c23 0 0 0 0 0 0 0 0 0 0 0 0 0.4000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 c31 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 c32 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 c33 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 i1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.2697 0.2000 0.2156 0.2264 0.2889 0.1869 0.1750 0.1250 0.1237 0.1084 0.1084 0.1237 0.1667 0.1429 0.1429 0.1429 0.1199 0.0450 0.1237 0.0655 0.1667 0.2000 0.0780 0.3309 0.3325 0.1367 0.1367 0.1429 0.1237 0.1667 0.1237 0.2000 0.2500 1 0 0 0 0 i2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.0682 0.0780 0.0801 0.0558 0.0727 0.0534 0.0636 0.1250 0.1237 0.1084 0.1084 0.1237 0.1667 0.1429 0.1429 0.1429 0.1199 0.0670 0.1237 0.0655 0.1667 0.2000 0.0780 0.3309 0.0896 0.4030 0.4030 0.1429 0.1237 0.1667 0.1237 0.2000 0.2500 0 1 0 0 0 i3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.1688 0.1255 0.1293 0.4904 0.1228 0.1213 0.1077 0.1250 0.1237 0.1084 0.1084 0.2343 0.1667 0.1429 0.1429 0.1429 0.2101 0.2191 0.1237 0.2500 0.1667 0.2000 0.1342 0.1985 0.3325 0.1367 0.1367 0.1429 0.1237 0.1667 0.1237 0.2000 0.2500 0 0 1 0 0 i4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.4252 0.5459 0.5231 0.0867 0.4429 0.5871 0.5900 0.5000 0.2343 0.4737 0.2011 0.1237 0.3333 0.4286 0.4286 0.4286 0.0706 0.1480 0.2343 0.1094 0.1667 0.2000 0.4882 0.0844 0.2012 0.0791 0.0791 0.4286 0.2343 0.1667 0.3945 0.2000 0.1250 0 0 0 1 0 i5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.0682 0.0507 0.0518 0.1407 0.0727 0.0513 0.0636 0.1250 0.3945 0.2011 0.4737 0.3945 0.1667 0.1429 0.1429 0.1429 0.4795 0.5210 0.3945 0.5096 0.3333 0.2000 0.2215 0.0553 0.0443 0.2444 0.2444 0.1429 0.3945 0.3333 0.2343 0.2000 0.1250 0 0 0 0 1 133int. j. prod. manag. eng. (2015) 3(2), 123-133creative commons attribution-noncommercial-noderivatives 4.0 international http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2018.8662 received 2017-09-28 accepted: 2018-03-21 product wheels for scheduling in the baking industry: a case study alexandria lee trattnera1, zaza nadja lee herbert-hansenb and lars hvama2 atechnical university of denmark broyal danish archives, rigsdagsgården, 9, copenhagen a1 alemo@dtu.dk, b znhh@sa.dk, a2 lahv@dtu.dk abstract: this paper illustrates current challenges and suggests solutions within the area of scheduling in the baking industry. the analysis applies the product wheel heuristic approach of king (2009) and tests the production cycles generated using actual sales and production data from a manufacturer of frozen baked goods. the product wheel method showed to be a suitable method for application at the baked goods manufacturer and generated a 23% reduction in setup and inventory cost at the case company. despite the benefits, the product wheel method proved difficult to apply in a high variety setting, where an operations research model may have achieved more significant results. key words: production planning, scheduling, product wheel, process industry. 1. introduction rising labour costs, increased price competition, and higher demand for customised products and quick delivery times are just a few of the pressures placed on the modern baked goods manufacturer (higgins, 2013). bread, a staple in the diet of the western world, has multiplied its forms through the years, spanning from baguettes to buns and flatbreads and assorted sweet treats. behind the increased variety on the store shelf are manufacturers struggling in a competition to maintain their low-cost infrastructure while accommodating a higher product mix. producing a wider array of products can require investment in technology in manufacturing, but can also require adjustment of operational procedures for tasks like production scheduling. it is a common goal for production scheduling to maximise service level towards the customer while minimising costs for the company (christou et al., 2007). in doing this, the production schedule acts as a critical link between the needs of the market and the physical output of a manufacturing system in the baking industry, helping a company achieve greater flexibility (nakhla, 1995). production scheduling is a decision-making process whereby the lot sizes, start and end times, and order sequence for a production operation are determined (lütke entrup, 2005; stadtler & kilger, 2005). additional factors determined by plans and schedules include which products to produce, which machines to make the products on, which machines to overload, when to schedule maintenance, and which demands to satisfy (fordyce et al., 2015; pinedo, 2009). factors such as the complexity of the production process, the number of products being produced and the variability of product demand all influence the planning and scheduling process. aspects of the baking industry that complicate production scheduling are the multi-stage fermentation process, the presence of active yeast in the dough, handling of allergens and organic ingredients, and the use of large, capital-intensive production equipment to cite this article: trattner, a.l., herbert-hansen, z.n.l., hvama, l. (2018). product wheels for scheduling in the baking industry: a case study. international journal of production management and engineering, 6(2), 65-78. https://doi.org/10.4995/ijpme.2018.8662 int. j. prod. manag. eng. (2018) 6(2), 65-78creative commons attribution-noncommercial-noderivatives 4.0 international 65 mailto:alemo@dtu.dk mailto:znhh@sa.dk mailto:lahv@dtu.dk http://creativecommons.org/licenses/by-nc-nd/4.0/ which often requires lengthy setup times (akkerman & van donk, 2009a; modal & datta, 2008). additionally, production scheduling can be a cost driver for a company if the batch sizes, sequencing and finishing times are not optimal. due to the broad range of effects of planning and scheduling on overall business performance and given the increasing demand for flexibility from customers, research around scheduling in the baking industry is needed to help baked goods manufacturers gain and maintain a competitive advantage. the food industry is an essential component of the european and global economies, accounting for roughly 13% of the turnover in the manufacturing sector in the eu in 2014 (statistical office of european communities, 2017). manufacturers of baked or farinaceous products are one of the principle niches within the broader food industry and have shown steady growth in revenue since 2010 in major european markets such as germany (12%) and italy (8%) (statistical office of european communities, 2017). despite this, the academic literature offers limited research on scheduling methods within the food sector and, more specifically, the baking industry. when discussing the industrial landscape of europe, it is necessary to consider small to medium-sized enterprises (smes) which employer fewer than 250 people and, yet, account for over 99% of the number of enterprises, 57% of total value added and over 66% of employment in the eu (european commission, 2016). smes in the food sector require a different approach for operations management due to the capabilities of management and limited resources (dora & gellynck, 2015; rymaszewska, 2014). however, scheduling tools discussed in the literature have primarily catered towards large enterprises, focusing on the installation of software systems and optimisation models and algorithms (van donk & van dam, 1994), both options which can be out of reach for an sme in the baking industry. this article, therefore, addresses the following research question: how can scheduling techniques be applied to improve production in the baking industry? this question is explored using a literature review of planning and scheduling methods in various food contexts, including baked goods manufacturing. next, the product wheel methodology of king (2009) is selected for application and testing at a small to medium-sized baked goods manufacturer. the product wheel is a heuristic method for gaining economies of repetition while simultaneously responding to needs for increased variety and flexibility towards the end customer (king, 2009; wilson & ali, 2014). this paper contributes to research by testing the product wheel in a food context, an industry with a documented need for better scheduling and which to date as seen limited research (o’reilly et al., 2015). the paper contributes to practice by using a reallife case, hence illustrating the practicability of the approach. the paper is structured as follows: first, a literature review is carried out exploring the baking industry, food sector and scheduling methods. second, the research methodology and empirical data are described. third, the findings from the case study are analysed and theoretical and practical implications are described. finally, conclusions and notes regarding further research are outlined. 2. literature review a preliminary look at the literature revealed an absence of studies focusing on scheduling in the baking industry. therefore, the literature search was expanded to explore scheduling methods in the broader food industry. 2.1. baked goods manufacturing manufacturers of baked goods, products such as bread, cakes, and pastries that are baked in an oven (oxford dictionaries, 2015). baked goods manufacturers can vary from large-scale to smallscale and may deliver either fresh or frozen products. fresh bread products typically are provided to markets daily to ensure the product is of the acceptable quality level when purchased (zhou & therdthai, 2006) while frozen products are often held in cold storage and delivered with longer lead times (ribotta et al., 2006). these manufacturers may also make semi-processed baked goods such as refrigerated dough, frozen dough, and partially baked dough (ribotta et al., 2006). a yeast bread-making process typically consists of the following stages: dough making, dividing, proving, baking, cooling, slicing and packaging (zhou & therdthai, 2006). characteristic production systems for baking and other food processes are flow shops, a configured set of dedicated machines which process all jobs in a fixed order (dewa et al., 2013; gupta & kumar, 2016). int. j. prod. manag. eng. (2018) 6(2), 65-78 creative commons attribution-noncommercial-noderivatives 4.0 international trattner, a.l., herbert-hansen, z.n.l. and hvama, l. 66 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2.2. scheduling in the food industry the current needs of the market and technological constraints of food manufacturers require production planners to consider several factors when scheduling jobs in production. these factors include the use of both batch and continuous processes within one production line, processing of perishable goods, sequence-dependent setup times, and high variability of yields and process duration (akkerman & van donk, 2009b). for the larger context of consumer goods manufacturers, recent trends in operational scheduling have included grouping products into families which use a comparable production setup, the use of regular production schedules for high running products to increase customer responsiveness, the change of emphasis from low production cost to fast delivery and logistics performance, and the shift from make-to-stock (mts) products to make-to-order (mto) products (bilgen & günther, 2010). products with a shorter shelf-life will often be scheduled with an mto strategy since the products cannot be stored in inventory for prolonged periods (akkerman & van donk, 2009b). in contrast, products with a longer shelf life (i.e. frozen baked goods) are made with an mts strategy as they have a longer shelflife in inventory. complicating the scheduling task further is the growing variety of products within the food industry (nakhla, 1995). 2.3. product mix flexibility the flexibility literature has been around since the 1980s and aims to define and measure how organisations can adapt to quickly to changes in the market. the flexibility of a production line directly impacts the production scheduling practices which can be used. given the level and uncertainty of customer demand, a manufacturer may respond by building flexibility capabilities within volume, product mix, quality, new product introduction, and delivery (slack, 1983). product mix flexibility has been defined as the “ability to manufacture a particular mix of products within the minimum planning period used by the company” (slack, 1983, p. 9). the process technology used, the design of the products, and the scheduling of the products are known to affect product mix flexibility (slack, 1983). product mix flexibility is inversely proportional to the difficulty to change between products on a manufacturing line. changeover time can be defined as the time needed to transition from producing one product to another. often, changeover time includes the time to turn the line on or off, time for the system to adjust to temperature or pressure and time to remove material from a previous production run (king, 2009). changing between products on a food production line is a critical task since food safety, allergens, production line arrangement, packaging type, and labour utilisation all must be considered. equipment changes and cleaning procedures must be executed thoroughly to avoid contamination and endangering end customers. from a cost perspective, changeovers should be as short and infrequent as possible since they pose a significant cost for companies in waste product generated and reduced speeds during the start-up phase of a new product (akkerman & van donk, 2008). one method used in the food industry to improve mix flexibility is natural sequencing, a technique where similar products are scheduled in succession to minimise overall changeover time (bilgen & günther, 2010). when applying natural sequencing, sequences of products are often developed based on product families and consolidated into schedule blocks which are then arranged to minimise changeover time. a product wheel, like a schedule block, is a method for natural sequencing which uses a flexible scheduling sequence for production that is based on demand, changeover times, production rates, and inventory carrying costs (king, 2009). product wheels can also serve as a tool for continuous improvement whereby reducing the duration (or lot size) of a product wheel allows greater flexibility and responsiveness to customers (king, 2009). 2.4. existing scheduling models for the food industry several operations research models and heuristics have been generated for specific food production industries, including the baking industry, as shown in table 1. since there were limited articles on planning problems in the baking industry, the literature search was expanded for those involving a food industry application. food process applications represented among these models range from producing candy to baked goods to beverages. different modelling formulations in the literature describe the specific production systems which vary depending on the number of production stages, the number and size of inventory locations, the presence of bottleneck resources, as well as the perishability of the products made. one model by hecker et al. (2013) includes a no-wait constraint for a baked good production line to model the time-sensitivity of the fermentation int. j. prod. manag. eng. (2018) 6(2), 65-78creative commons attribution-noncommercial-noderivatives 4.0 international product wheels for scheduling in the baking industry: a case study 67 http://creativecommons.org/licenses/by-nc-nd/4.0/ process. this model was the only one to incorporate such a constraint specifically for the baking industry. 2.5. literature summary this literature review shows that scheduling in the food industry is challenging and requires consideration of several factors. furthermore, it shows that research into planning and scheduling in the baking industry is limited. the models currently developed and in use within the food industry are operations research models, heuristics, and lean scheduling methods, such as the product wheel. these models and tools vary depending on the nature of the production system being scheduled and are often customised to incorporate the specific constraints facing the food industry, such as perishability of the products and sequence-dependent changeover time. this literature review reveals a lack of literature exploring both the quantitative and qualitative aspects of scheduling in the baking industry. there are limited applications of quantitative scheduling models made for the baking companies, with only three instances identified in table 1. also, lacking in the literature was the application of the heuristic methods, the product wheel more specifically, within the baking industry. given the limited work in the food industry to meet the needs of food smes in europe, the product wheel heuristic is selected for application in this study. this heuristic approach offers an approach to improve scheduling that is within the resources of a typical sme and accommodating the product mix flexibility required in the baking industry. the contribution of this paper is a map of literature on the scheduling practices and research in food companies as well as a real case application. 3. methodology a mixed methods approach was used to frame the empirical study of how scheduling methods can optimise performance in the baking industry. both qualitative data and quantitative data were gathered from a case company to understand the scheduling process, schedule performance and productivity. different data sources were also used for triangulation purposes. the explorative nature of the research question allows for an in-depth understanding of the research area, suitable for the qualitative work in this project, while the performance of the scheduling methods can be assessed using quantitative methods (creswell, 2014). the case company, here called baking company, is an sme located in denmark with approximately 200 full-time employees. the company was selected as it produces a wide range of baked goods with over 200 products serving the convenience bread market (also known as “bake-off” market). baking company qualifies as being an sme per the definition of the european commission (european commission, 2016). all products are either fully baked or partially baked in the process and all are frozen before being sold. the company was also selected as their production 1. system was seen to be typical for the baking industry, consisting of two automated, flexible flow shop production lines.based on the research question and findings from the literature review, a research framework was taken from the product wheel heuristic from king (2009). the method includes 10 steps which assess various aspects of the production system and scheduling practices. the steps are: 2. decide which assets would benefit from product wheels. 3. analyse product demand variability. 4. determine the optimum production sequence. 5. calculate the shortest wheel time based on time available for changeovers. 6. estimate the economic optimum wheel time based on economic lot size (els) model. 7. determine the basic wheel time and determine which products are made on every cycle and the frequency for other products. 8. calculate inventory levels to support the wheel. 9. repeat steps 3-7 to fine-tune the design. 10. revise all scheduling processes, as appropriate. create a visual display (heijunka) to manage the levelled production. these steps are shown graphically in figure 1. steps 1 to 7 of the method will be applied to the data set from one production line at baking company and the appropriateness of the approach will be assessed. the high-volume production line was selected as it is the production line with the highest capacity int. j. prod. manag. eng. (2018) 6(2), 65-78 creative commons attribution-noncommercial-noderivatives 4.0 international trattner, a.l., herbert-hansen, z.n.l. and hvama, l. 68 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 1. operations research models and heuristics for scheduling in the food industry. author model and objective case solution method and findings silva, et al. (2014) model a flow-shop problem with parallel production, setup times, batch production, due date and include transport capacity. baking company apply a greedy heuristic and genetic algorithm to solve the problem. hecker et al. (2013) milp for hybrid flow-shop. uses no-wait constraints to model time-sensitivity with dough fermentation. objective: min. total makespan or idle time. bakery (germany) solve for local optimal (not global) using particle swarm theory and ant colony optimisation programmed in matlab. dewa et al. (2013) finite capacity scheduling system applied to flow shop tested with 5 heuristics. objective: min. cost of earliness & lateness. bakery (zimbabwe) simulate 5 heuristics in arena including earliest due date, first come first served, first in system last served, shortest processing time first, and a random procedure. edd heuristic gave the optimal schedule. mehrotra et al. (2011) milp for creating production patterns in the processed food industry. objective: minimise setup and inventory costs. conagra foods two-stage heuristic which groups and assigns products to lines and sets sequence of each group. use a heuristic-based planning tool which reduces the cost of setup and inventory by 15% with a 4-week long cyclic schedule. bilgen & günther (2010) milp applied to multi-site production and distribution. objective: minimise setup, inventory and transportation costs. fruit juice, 3 plants (germany) uses block planning to model natural sequencing of products using randomly generated demand data. solve models to optimality. christou et al. (2007) ip with 3 levels of scheduling granularity for aggregate planning on multi-product lines. objectives: maximise customer service level and minimise extra labour costs and inventory costs (maximise freshness). beverage manufacturer with 3 plants (greece) solve to optimality using lp relaxation and a custom two-part optimisation for solving the shift allocation first and the scheduling second. code programmed in ansi. doganis & sarimveis (2007) milp to optimise production, customised for sequencing of yoghurt products. objective: minimise changeover, inventory, and labour costs. yoghurt production line (greece) solve milp using cplex to global optimality in less than 15 seconds. the output is a daily production schedule and resulting inventory levels soman et al. (2007) short-term batch scheduling heuristic using economic lot-sizing for mto/mts products. objective: minimise overall makespan while meeting demand. threestage food manufacturer apply heuristic within a hierarchical planning framework. suggest using the heuristic alongside detailed manual scheduling. mendéz & cerdá (2002) milp application to two-stage make-andpack production with unlimited intermediate and final storage. objective: minimise total makespan. candy producer (theoretical) solved to optimality using cplex. applied pre-ordering rules (e.g. shortest intermediate processing time, general first processed first served) to reduce problem size. tadei et al. (1995) two model approach: (1) medium planning using an lp to allocate labour and (2) shortterm planning using an ip to determine shift schedule. objectives: min. inventory, meet demand. alimentary preserves (portugal) solve using a decomposition heuristic was implemented using c++. the method proved as a consistent tool to evaluate whatif scenarios. evaluate schedules based on average stock levels. randhawa et al. (1994) scheduling heuristic for a multi-stage system with parallel machines. uses shortest processing time first on bottleneck resources. objective: min. average flow times/ lateness. freeze-dried food producer (usa) solve with a computer model which creates a schedule for each stage along with kpis (% utilisation, % idle time, time in the system, etc.) claassen & van beek (1993) milp model with two tiers (tactical and operational planning) applied to a flow shop with n-jobs and m-machines. objective: minimise penalty for lateness, setup, overtime, and costs. packaging line (netherlands) solved to local optimality using a decomposition heuristic. implemented programs into a decision support tool which generated higher quality schedules than the manually generated ones. *integer programming (ip) using only integer variables, linear programming (lp) uses continuous variables, and mixed-integer linear programming (milp) which uses both integer and continuous variables. int. j. prod. manag. eng. (2018) 6(2), 65-78creative commons attribution-noncommercial-noderivatives 4.0 international product wheels for scheduling in the baking industry: a case study 69 http://creativecommons.org/licenses/by-nc-nd/4.0/ and is often running at full utilisation, therefore showing potential to benefit from further optimised production planning. the primary methods of data collection were company visits, interviews, presentations by the senior staff, product documentation, sales records, and data from the manufacturing execution system. production data used to create the product wheels is from the full year 2014. the product wheels generated will be tested using a simulation built in microsoft excel and four months of actual sales data from 2014 to see the impact on delivery service to the customers. the effect of the product wheel on delivery service to customers was measured as days with stock outs per product. analyze product demand variability determine the optimum production sequence calculate the shortest wheel time estimate the economic optimum wheel time calculate inventory levels to support the wheel create visual display (heijunka) to manage the levelled production determine the basic wheel time is the design as desired? decide which assets would benefit from product wheels yes no figure 1. steps in the product wheel method. adapted from (king, 2009). two days were spent on site in production observing the system and gaining an understanding of the process flow. during these visits, the production manager was interviewed for 30 minutes to understand the production data. qualitative data on the planning process was gathered through two, onehour semi-structured interviews with the production planner to gain deeper insight and supplement the quantitative data. the questions focused on how schedules were created, rules for production sequence, and interactions between the planner and production team when creating the schedule. the proposed research methodology allows for exploration of the gap in the literature regarding the impact of the scheduling via the product wheel in an sme in the food industry by using quantitative methods such as the els model and production cycles. the qualitative aspects are addressed by analysing the scheduling process of the production planner. this research methodology will also address the lack of the application of the heuristic scheduling methods within the baking industry with the development of production cycles at the case company. 4. findings observations collected during the on-site visits at baking company revealed that the company uses a batch production system that is available for production approximately 90 hours per week, operating from monday to friday (i.e. maintenance time excluded). the product assortment at baking company consists of six major product groups: sausage rolls, danish pastries, focaccia bread, buttermilk horns, pastry bars, and pastry rolls. each product is assigned to a specific line for production as the preferred line. the production line selected for study produces mainly sausage rolls and buttermilk horns. all products at baking company are frozen and then baked at the retail location for final sale. figure 2 shows the primary process stages on the line. in mixing phase, the wet and dry ingredients for the dough are weighed and mixed. the mixed dough is then placed into a hopper and guided through an extruder and onto a conveyor to the lamination stage where the dough is rolled flat and layered. after lamination, the dough moves via conveyor to the makeup stage where it is cut and formed into the final shape with additional ingredients, such as sausages, cheese, and cream filling. the formed products are int. j. prod. manag. eng. (2018) 6(2), 65-78 creative commons attribution-noncommercial-noderivatives 4.0 international trattner, a.l., herbert-hansen, z.n.l. and hvama, l. 70 http://creativecommons.org/licenses/by-nc-nd/4.0/ placed on trays and then moved to a proofing step where the dough can rise before either being frozen or baked. the products which are fully or partiallybaked can cool before being frozen. once frozen, all products are placed in boxes and palletised before moving into cold storage. the products have a oneyear shelf life in cold storage. 4.1. creating the master production schedule to assess the qualitative nature of the planning tasks, a task decomposition of the production scheduling process at baking company was created. at the highest planning level, a master production schedule (mps) is created by the production planner which shows the aggregated volumes of each product to be produced on both lines. each week, the production planner develops the mps for a six-week planning horizon and then revises it based on rush orders and orders for mto products. mts products are selected for production based on their inventory levels and expected demand. mto products are scheduled with a lead time of 3-4 weeks while mts products must be delivered in one day due to the competitive nature of the convenience bread market. with such short lead times, the mts products must have a sufficient stock level to cover demand with an acceptable level of service. 4.2. creating the detailed schedule once the aggregate planning values are determined in the mps, the planner generates the detailed production schedule with a one-week planning horizon. a minimum run length is set to 3 hours and a target run length is 7.5 hours, which is the approximate length of one shift. the planner uses a combination of rules of thumb along with the planning interface of the erp system to determine the run length and achieve the desired volume in the mps. once the planner determines the run lengths, she assigns the products to the line, sequences the products based on scheduling rules and then estimates the changeover time. these tasks are often executed in parallel as the planner attempts to make the schedule fit demand while complying with capacity constraints. as is common in the food industry, many planning rules are used to create schedules for baking company to reduce production costs. the following list contains a set of planning rules that are currently used at baking company to form the mps and the detailed production schedule: organic items are scheduled as the first of the week to avoid contamination. light-coloured doughs are scheduled before dark-coloured doughs to avoid colour mixing. items with sauce are scheduled after items without sauce and toward the end of the week to avoid excessive cleaning. products with allergens (i.e. sesame or almonds) are run as the last product for the week. fully-baked products are scheduled separately from partially-baked products. chicken products are always made before pork products to reduce the risk of cross-contamination. once the schedule is made for the week, the planner reviews the schedule with the production team leader for feasibility. the planner requests feedback on the planned changeover time, sequencing, time for new products, and other factors before updating the schedule and releasing it to production. during onsite visits, it was seen that the production planner at baking company maintains daily communication with the production workers and is seated in an office that is close to the factory operations. 4.3. production volume a pareto method called the glenday sieve was applied to production data with a summary shown in table 2 to visualise the distribution of the production volume among the products at baking company (glenday, 2005). the glenday sieve figure 2. the production process at baking company. int. j. prod. manag. eng. (2018) 6(2), 65-78creative commons attribution-noncommercial-noderivatives 4.0 international product wheels for scheduling in the baking industry: a case study 71 http://creativecommons.org/licenses/by-nc-nd/4.0/ reveals that half of the production time was spent making only 19 products on both lines at baking company in 2014. these represent the high running products such as a danish pastry product and various products from the sausage roll group. the glenday sieve also shows that 27 products contribute to the red category which accounts for only 1% of production time at the facility. with the highest concentration of stock keeping units (skus) residing in the yellow group, it appears that the company is spending 95% of its time making only 65% products. the glenday sieve reveals a clear distinction between the products that are high runners and those that are “low runners” at baking company. furthermore, the production time is unevenly distributed amongst the products with a moderately long tail taking only limited capacity. table 2. glenday sieve for products in 2014. product colour cumulative % production h number of products % of products green 50% 19 12% yellow 95% 88 53% blue 99% 31 19% red 100% 27 16% total 100% 166 100% 5. analysis in the following sections, the first seven steps of the heuristic presented by king (2009) is applied to the 2014 production and sales data to test the applicability of the method to the baking sector. the details of the implementation are listed in the following sections. 1. decide which assets would benefit from product wheels. the asset at the baking company selected for study is a high-volume, flow shop production line (gupta & kumar, 2016). the production line is scheduled as a single unit since all machines connect via a conveyor system with low work in process inventory between machines. 2. analyse product demand variability. an analysis of the variation of monthly demand for a full year of sales data from 2014 was performed for products at baking company based the method of d’alessandro and beveja (2000). in this method, the average monthly demand and coefficient of variation (cv) of monthly demand are calculated for 86 products produced on the line and graphed to segregate the products into mto and mts categories. the findings for baking company are presented in figure 3. 0,00 0,50 1,00 1,50 2,00 2,50 1 10 100 1.000 10.000 c oe ffi ci en t of v ar ia nc e of m on th ly d em an d average monthly demand (cartons) (logarithmic scale) q2 mto q1 mts or mto q3 mts or mto q4 mts figure. 3 analysis of the variation of monthly demand (logarithmic scale with quadrants). the threshold cv value was set to 1.0 since most of the green and yellow products (high volume products from the glenday sieve) had cv values between 0 and 1.0. the threshold value for average monthly demand was set to 300 cartons as this equates to roughly one 3-hour production run every two months for most products. as 3 hours is the minimum run length in production, making this quantity every two months was deemed to be “low running.” as is typical for this demand analysis, the products in q2 are classified as mto, and the products in q4 are classified as mts. the products in q1 and q3 can be classified as either mto or mts depending on the company sales and operations strategy. in this analysis, the q1 and q3 products are designated as mts as the frozen nature of the food allows them to stay in inventory with a low risk of being scrapped. table 6 shows a summary of the four quadrants including the number of products within each and strategies for scheduling them, where 81 of the 86 products classify as mts. the 81 mts products will be carried through the remaining steps of the product wheel heuristic since the 5 mto products should be scheduled only when an order is received. when comparing the results of the glenday sieve to the results of the demand variability analysis, all green products from the glenday sieve were designated as mts based on the demand variability analysis as they fell within q4 (high volume, low demand variability). int. j. prod. manag. eng. (2018) 6(2), 65-78 creative commons attribution-noncommercial-noderivatives 4.0 international trattner, a.l., herbert-hansen, z.n.l. and hvama, l. 72 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 3. mto and mts segregation for products. q2 – low volume, high variability strategy = mto # products = 5 q1 – high volume, high variability strategy = mts # products = 4 q3 – low volume, low variability strategy = mts # products = 26 q4 – high volume, low variability strategy = mts # products = 51 3. determine the optimum production sequence. using the 81 mts products as a basis and the scheduling rules gathered from the interviews production planner, the optimal production sequence was determined. the data from the interview was triangulated by an assessment of the planned changeover time between the product groups in 2014 since changeovers at baking company are sequence dependent. a changeover time matrix was made on the product level but was not utilised since there were many missing combinations of products; therefore, the average changeover time between product groups was used. both analyses showed that changeover time is minimised when scheduling items from the same product group next to each other. 4. calculate the shortest wheel time based on time available for changeovers. this step advises the user to place all products in a single production cycle and estimate the number times a changeover could be performed. using equation (1) presented by king (2009), the maximum number of cycles of a product wheel in one year which contains all mts products for the line is computed as 1.7 (see equation 2). this is calculated assuming the total available production time on the line is 4,320 hours (90 hours per week, 48 weeks per year), the total production time equals the production time from 2014 for the mts products and that the product wheel changeover time is the sum of the average changeover time for the mts products. the calculation shows that the production cycle should be run between one and two times every year if all products are made once in every cycle. .maxcycles productwheelchangeover time total avail time productwheeltime–= (1) ( ) .maxcycles 53 4230 4232 1 7 – = = (2) 5. estimate the economic optimum wheel time based on els model. continuing with the product wheel heuristic, the economic lot sizing (els) analysis was performed to calculate the optimal batch sizes for production of the 81 mts products on the line at baking company. the els calculates the cycle length xj (the amount of time between production runs for each product j) and the lot size. variables in the calculation were determined using the sales and production data from 2014. the inventory holding cost, hj, was determined by taking the costs per carton for storage and handling finished goods at an external warehouse close to baking company. the calculation of lot size (cycle length multiplied by demand rate) is shown in equation 3 where: j = product number elsj = economic lot size dj = demand rate for product j in cartons per hour from 2014 sales data qj = production rate for product j in cartons per hour from 2014 production data hj = inventory holding costs for product j in eur per carton per hour cj = setup cost for product j in eur per setup ( ) els xjdj h q c 1 2 j j d q j j j j= = (3) for each product, the optimal run length, lot size and production frequency were determined. the base units for volume and time in this analysis are cartons and hours. this model assumes a constant demand rate and production rate and utilises sequenceindependent setup times in the calculations. while the production at the baking company experiences sequence-dependent setup times, the sequenceindependent setup time was calculated by taking the average setup time for each product to simplify the model. three of the mts products had cycle lengths greater than once every year, which is not feasible given the one-year shelf life of the products. 6. determine the basic wheel time; determine which products are made on every cycle and the frequency for others. since there were 81 mts products, a different strategy for generating production cycles was required. the basic wheel time is usually set by the int. j. prod. manag. eng. (2018) 6(2), 65-78creative commons attribution-noncommercial-noderivatives 4.0 international product wheels for scheduling in the baking industry: a case study 73 http://creativecommons.org/licenses/by-nc-nd/4.0/ high-volume products (king, 2009). a histogram of the cycle lengths for the line studied (see figure 4) shows that most products have a 5-6-week cycle length, while most products have a cycle length of less than 11 weeks. no products had a cycle length less than 3 weeks. high volume, green products have cycle lengths from the els analysis which ranged from 3.3 – 6.2 weeks. figure 4. the economic cycle lengths for mts products. for the first iteration, the product wheel uses a 4-week cycle time which repeats two times so that products are made every 4 or 8 weeks while other mts products are made when the stock is nearly depleted. all product cycle lengths were rounded to the nearest multiple of four, and the lots sizes were updated. the product wheels generated are shown in figure 5 and figure 6. the green and yellow colours in the figures and tables indicate that the colour classification per the glenday sieve and serves as a reference for the high volume and mid-volume products. figure 5. product wheel for weeks 1-4. the wheels were designed so that the product cycle always takes 70% of the production time for the week to allow room in the schedule for mto products which have a lead time of 3-4 weeks. as suggested by king (2009), products which have a cycle length over eight weeks should be placed in empty grey spokes to accommodate both mts and mto production strategies. only 35 products with an mts strategy which had cycle lengths of less than eight weeks were included in the product wheel. since products made every 12 or more weeks are made only 5 or fewer times per year, these are not included in the product wheel but are given space to be produced in time designated for “other sausage rolls”, etc. shown as one of the grey spokes in the wheel in figures 5 and 6. figure 6. product wheel for weeks 5-8. 7. calculate inventory levels to support the wheel. for testing the production plan, the safety stock levels were set at two weeks of the product demand which is the current safety stock target for high running products at baking company. simulating the product wheels with actual demand data for four months (march 3, 2014 – june 25, 2014) at baking company was possible for the 32 products which were in the product wheel. demand data was not available for three of the products in this time frame, so only 32 of the 35 products were assessed. the simulation demand period was selected to minimise the influence of seasonality in demand. in the simulation, stocks were initialised to be equal to the economic lot size plus the safety stock for each product. for each day in the simulation, the production quantities from the product wheels and demand quantities from the demand data were added or deducted from the stock in the previous day accordingly for each product. int. j. prod. manag. eng. (2018) 6(2), 65-78 creative commons attribution-noncommercial-noderivatives 4.0 international trattner, a.l., herbert-hansen, z.n.l. and hvama, l. 74 http://creativecommons.org/licenses/by-nc-nd/4.0/ the simulation of the first version of the product wheel showed that two products faced stock-outs over the four-month period: scones g has 13 days without stock and sandwich r has two days without stock. the lot sizes were increased and the simulation ran a second time where there were no stock-outs. 6. results and discussion it is estimated that implementing the product wheels will lead to a 42-hour reduction (-18%) in changeover time on the line, which equates to roughly 2 days of additional production time per year. the impact of the product wheel on setup and inventory costs is also found and summarised in table 3. as can be seen, implementing production cycles could potentially lead to eur 103,600 (23%) decrease in annual production costs from 2014. as this applies to the products that are already running in long production series already, some of the benefits are overlooked. for example, if all products are scheduled in economic lot sizes, the company could potentially save eur 145,000 in costs. such results show that this manual heuristic did accommodate the need for mix flexibility and minimisation of production costs in baking company. table 3. impact of the product wheel on annual setup and inventory costs for 35 mts products. changeover cost (eur) inventory holding cost* (eur) total (eur) original schedule (2014) 226,400 217,100 443,500 product wheel 180,200 149,000 339,900 savings 46,200 57,400 103,600 * inventory cost includes the cost of safety stock. note: costs and savings calculated only for the 35 products assessed in the product wheel. the theoretical savings calculated for the production cycles at baking company are slightly higher than other research studies which used production cycles, such as the 15% reduction in setup and inventory costs found by mehrotra et al. (2011) using their optimisation model. a study of product wheels in the process industry showed mixed results as to the impact of scheduling on changeover time, increasing the time in some cases and decreasing time in others (wilson & ali, 2014). where the product wheel had less favourable results in the case application was in the meeting demand requirements to the market for the mts products. the lack of widespread stock-outs suggests that the inventory levels and batch sizes can be reduced slightly, but seasonality of demand should be considered. the stock out situation of baking company is comparable to the product wheel implementation at a chemical manufacturer (wilson & ali, 2014). it is worth noting that while making the product wheels in a real production scheduling scenario, the stock levels and corresponding lot sizes would need to be adjusted per the changing market needs and demand seasonality. however, the sequence of the production runs should remain the same since it is designed to reduce the total changeover time and inventory costs based on natural sequencing. looking beyond the production and warehouse impact, implementing the production cycles at the case in this study requires changes to the scheduling process, as well. in one potential redesign of the process, the product wheels are the first items to be planned when creating the mps. the cycles are allocated across the weeks and production lines based on their cycle length. if there is capacity remaining in each week after production cycles have been allocated, the items are selected for production based on the original process. this process is expected to reduce the decision-making load of the planner. the specific heuristic presented by king (2009) was difficult to apply to the case company for various reasons. the presence of sequence-dependent changeover times made the manual tasks in step 3 of determining the optimum production sequence quite tricky. step 4 of calculating the shortest wheel time was not readily applicable given the high product variety 81 mts products. the heuristic assumes that all products are made in every cycle, which would mean that a cycle would be developed for up to 81 products which would be scheduled over the course of 5-6 months, compared to the 35 products in the proposed 8-week product wheel set. this would be complicated and reduce the chances that the benefits of the product wheel, such as economies of repetition and making faster changeovers in production, would be realised. running a production cycle twice per year will pose many practical challenges, such as very high batch sizes and stock levels for products. this shows that the product wheel method is not the best fit for production scenarios in the process industry which have a high number of make to stock products. in a study on product wheels at a chemical manufacturer, only eight products were included in int. j. prod. manag. eng. (2018) 6(2), 65-78creative commons attribution-noncommercial-noderivatives 4.0 international product wheels for scheduling in the baking industry: a case study 75 http://creativecommons.org/licenses/by-nc-nd/4.0/ the product wheel design, so this was much simpler to generate the schedule for (wilson & ali, 2014). this suggests that the product wheel is more suitable for smaller scheduling problems. among the collection of scheduling methods tested in the food sector, the product wheel is an overly simplistic approach for the high variety company which was studied in this case. the product wheel offered an approach by which variety and sequencing could be addressed in the scheduling process in the baking company. however, based on the application example of baking company in this study, it can be concluded that the method should be reserved for small problems where few mts items are required to be integrated into the product wheel. the other optimisation models presented in table 1 which utilised operations research methods to solve the issues of natural sequencing via schedule blocks might be more suitable for solving applications with higher variety (bilgen & günther, 2010; günther et al., 2006; mehrotra et al., 2011; mendéz & cerdá, 2002; pinedo, 2009). just like the product wheel, the operations research production cycles aim to increase production efficiency by using pre-defined sequences of production orders. however, their solving ability for more complex problems makes them superior to the product wheel. regardless of the issues with implementation, the simulation of the product wheel at baking company showed savings in changeover and inventory costs. 7. conclusions and future research through a literature review and application of the product wheel methodology to a case company, it was found that the production cycles are a suitable scheduling method for improving the production performance in the baking industry, particularly at small to medium-sized enterprises. however, when testing the product wheel method proposed by king (2009) at a case company with high product variety, the method was found difficult to apply due to its manual nature. the results suggest that when scheduling production in a baking company with high variety, a more sophisticated technique for scheduling based on operations research methods should be utilised. despite the drawbacks with the number of products, the product wheels generated in the study led to a 23% reduction in changeover and inventory costs for the products simulated at the case company. this work contributes to the current gap in the literature aspects of scheduling in the baking industry by providing a case-based approach to show the applicability of production cycles as a scheduling method to a baked goods manufacturer and significant benefits of production cycles in this sector are estimated. this research study is limited in generalizability due to the nature of the case study. however, it is reasonable to assume that applying the product wheel in another food company with similar variety and seasonality would yield comparable results. the primary area of future work is to implement and assess the effectiveness of the proposed production cycles. such research would provide the data needed to evaluate the actual performance of the cycles against the estimated performance and compare them to the simulated values. references akkerman, r., van donk, d. p. (2008). development and application of a decision support tool for reduction of production losses in the foodprocessing industry. journal of cleaner production, 16(3), 335-342. https://doi.org/10.1016/j.jclepro.2006.07.046 akkerman, r., van donk, d. p. (2009a). product mix variability with correlated demand in two-stage food manufacturing with intermediate storage. international journal of production economics, 121(2), 313-322. https://doi.org/10.1016/j.ijpe.2006.11.021 akkerman, r., van donk, d. p. (2009b). analyzing scheduling in the food-processing industry: structure and tasks. cognition, technology and work, 11(3), 215-226. https://doi.org/10.1007/s10111-007-0107-7 bilgen, b., günther, h. o. (2010). integrated production and distribution planning in the fast moving consumer goods industry: a block planning application. or spectrum, 32(4), 927-955. https://doi.org/10.1007/s00291-009-0177-4 christou, i. t., lagodimos, a. g., lycopoulou, d. (2007). hierarchical production planning for multi-product lines in the beverage industry. production planning and control: the management of operations, 18(5), 367-376. https://doi.org/10.1080/09537280701340631 claasen, g. d., van beek, p. (1993). planning and scheduling packaging lines in the food industry. european journal of operational research, 70(2), 150-158. https://doi.org/10.1016/0377-2217(93)90034-k creswell, j. w. (2014). research design: qualitative, quantitative, and mixed methods approaches (4rd ed.), thousand oaks, ca, usa: sage publications, inc. int. j. prod. manag. eng. (2018) 6(2), 65-78 creative commons attribution-noncommercial-noderivatives 4.0 international trattner, a.l., herbert-hansen, z.n.l. and hvama, l. 76 https://doi.org/10.1016/j.jclepro.2006.07.046 https://doi.org/10.1016/j.ijpe.2006.11.021 https://doi.org/10.1007/s10111-007-0107-7 https://doi.org/10.1007/s00291-009-0177-4 https://doi.org/10.1080/09537280701340631 https://doi.org/10.1016/0377-2217(93)90034-k http://creativecommons.org/licenses/by-nc-nd/4.0/ d’alessandro, a. j., baveja, a. (2000). divide and conquer: rohm and haas’ response to a changing specialty chemicals market. interfaces, 30(6), 1-16. https://doi.org/10.1287/inte.30.6.1.11627 dewa, m. t., mhlanga, s., masiyazi, l., museka, d. (2013). design of a finite capacity scheduling system for bakery operations (flow shop environment). international journal of innovative research in science, engineering and technology, 2(11), 6631-6640. doganis, p., sarimveis, h. (2007). optimal scheduling in a yoghurt production line based on mixed integer linear programming. journal of food engineering, 80, 445-453. https://doi.org/10.1016/j.jfoodeng.2006.04.062 dora, m., gellynck, x. (2015). house of lean for food processing smes. trends in food science & technology, 44(3), 272-281. https://doi.org/10.1016/j.tifs.2015.03.008 european commission (2016). annual report on european smes 2015/2016. retrieved from https://ec.europa.eu/jrc/sites/jrcsh/files/ annual_report_-_eu_smes_2015-16.pdf (last access: july 2018). fordyce, k., milne, j.r., wan, c.t. zisgen, h. (2015). modelling and integration of planning, scheduling, and equipment configuration in semiconductor manufacturing part ii. fab capability assessment. international journal of industrial engineering: theory, applications and practice, 22(5), 601-607. glenday, i. (2005). breaking through to flow. lean enterprise academy, ross-on-wye, uk. günther, h. o., grunow, m., neuhaus, u. (2006). realizing block planning concept in make-and-pack production using milp modelling and sap apo. international journal of production research, 44(18-19), 3711-3726. https://doi.org/10.1080/00207540600589127 gupta, a., kumar, k. (2016). flow shop scheduling decisions through techniques for order preference by similarity to an ideal solution (topsis). international journal of production management and engineering, 4(2), 43-52. https://doi.org/10.4995/ijpme.2016.4102 hecker, f. t., hussein, w. b., paquet-durand, o., hussein, m. a., becker, t. (2013). a case study on using evolutionary algorithms to optimize bakery production planning. expert systems with applications, 40(17), 6837-6847. https://doi.org/10.1016/j.eswa.2013.06.038 higgins, k. t. (2013). food processors must balance high throughput with demand for fast changeovers and flexible production. retrieved 02 july 2015 from https://www.foodprocessing.com/ king, p. l. (2009). lean for process industries. michigan: crc press. lütke entrup, m. (2005). advanced planning fresh food industries: integrating shelf life into production planning. heidelberg, germany: physica-verlag. mehrotra, m., dawande, m., gavirneni, s., demirci, m., tayur, s. (2011). or practice-production planning with patterns: a problem from processed food manufacturing. operations research, 59(2), 267-282. https://doi.org/10.1287/opre.1100.0871 mendéz, c. a., cerdá, j. (2002). an milp-based approach to the short term scheduling of make-and-pack continuous production plants. or spectrum, 24(4), 403-429. https://doi.org/10.1007/s00291-002-0103-5 modal, a., datta, a. k. (2008). bread baking a review. journal of food engineering, 86(4), 465-474. https://doi.org/10.1016/j. jfoodeng.2007.11.014 nakhla, m. (1995). production control in the food processing industry: the need for flexibility in operations scheduling. international journal of operations & production management, 15(8), 73-88. https://doi.org/10.1108/01443579510094107 oxford dictionaries. (2015). retrieved 02 july 2015 from http://www.oxforddictionaries.com/ o’reilly, s., kumar, a, adam, f. (2015). the role of hierarchical production planning in food manufacturing smes. internation journal of operations & production management, 35(10), 1362-1385. https://doi.org/10.1108/ijopm-04-2014-0157 pinedo, m. (2009). planning and scheduling manufacturing and services. new york, ny: springer. randhawa, s. u., juwono, c., burhanuddin, s. (1994). scheduling in multistage flowshop systems: an application in the food processing industry. industrial management & data systems, 94(5), 16-24. https://doi.org/10.1108/02635579410063270 ribotta, p. d., leon, a. f., anon, m. c. (2006). frozen dough. in y.h. hui, bakery products: science and technology. ames, iowa, usa: blackwell publishing. https://doi.org/ 10.1002/9780470277553.ch21 rymaszewska, a. d. (2014). the challenges of lean manufacturing implementation in smes. benchmarking: an international journal, 21(6), 987-1002. https://doi.org/10.1108/bij-10-2012-0065 silva, f.a.m. da, moretti, a. c., azevedo, a. t. (2014). a scheduling problem in the baking industry. journal of applied mathematics, 2014. https://doi.org/10.1155/2014/964120 slack, n. (1983). flexibility as a manufacturing objective. international journal of operations & production management, 3(3), 4-13. https://doi.org/10.1108/eb054696 soman, c. a., van donk, d. p., gaalman, g. j. (2007). capacitated planning and scheduling for combined make-to-order and make-tostock production in the food industry: an illustrative case study. international journal of production economics, 108(2), 191-199. https://doi.org/10.1016/j.ijpe.2006.12.042 stadtler, h., kilger, c. (2005). supply chain management and advanced planning (3rd ed.). heidelberg, germany: springer berlin. statistical office of european communities (2017). eurostat: industry, trade and services statistics. luxembourg: eurostat. retrived from http://ec.europa.eu/eurostat int. j. prod. manag. eng. (2018) 6(2), 65-78creative commons attribution-noncommercial-noderivatives 4.0 international product wheels for scheduling in the baking industry: a case study 77 https://doi.org/10.1287/inte.30.6.1.11627 https://doi.org/10.1016/j.jfoodeng.2006.04.062 https://doi.org/10.1016/j.tifs.2015.03.008 https://ec.europa.eu/jrc/sites/jrcsh/files/annual_report_-_eu_smes_2015-16.pdf https://ec.europa.eu/jrc/sites/jrcsh/files/annual_report_-_eu_smes_2015-16.pdf https://doi.org/10.1080/00207540600589127 https://doi.org/10.4995/ijpme.2016.4102 https://doi.org/10.1016/j.eswa.2013.06.038 https://www.foodprocessing.com https://doi.org/10.1287/opre.1100.0871 https://doi.org/10.1007/s00291-002-0103-5 https://doi.org/10.1016/j.jfoodeng.2007.11.014 https://doi.org/10.1016/j.jfoodeng.2007.11.014 https://doi.org/10.1108/01443579510094107 http://www.oxforddictionaries.com/ https://doi.org/10.1108/ijopm-04-2014-0157 https://doi.org/10.1108/02635579410063270 https://doi.org/ 10.1002/9780470277553.ch21 https://doi.org/10.1108/bij-10-2012-0065 https://doi.org/10.1155/2014/964120 https://doi.org/10.1108/eb054696 https://doi.org/10.1016/j.ijpe.2006.12.042 https://ec.europa.eu/eurostat http://creativecommons.org/licenses/by-nc-nd/4.0/ tadei, r., trubian, m., avendano, j. l., della croce, f., menga, g. (1995). aggregate planning and scheduling in the food industry: a case study. european journal of opertional reserach, 87, 564-573. https://doi.org/10.1016/0377-2217(95)00230-8 wilson, s. ali, n. (2014). product wheels to achieve mix flexibility in the process industries. journal of manufacturing technology management, 25(3), 371-392. https://doi.org/10.1108/jmtm-03-2012-0026 van donk, d. p., van dam, p. (1994). structuring complexity in scheduling: a study in a good processing industry. international journal of operations & production management, 16(5), 54-63. https://doi.org/10.1108/014435796101 zhou, w., therdthai, n. (2006). manufacture. in y. h. hui, bakery products: science and technology. ames, iowa, usa: blackwell publishing. https://doi.org/ 10.1002/9780470277553.ch17 int. j. prod. manag. eng. (2018) 6(2), 65-78 creative commons attribution-noncommercial-noderivatives 4.0 international trattner, a.l., herbert-hansen, z.n.l. and hvama, l. 78 https://doi.org/10.1016/0377-2217(95)00230-8 https://doi.org/10.1108/jmtm-03-2012-0026 https://doi.org/10.1108/014435796101 https://doi.org/ 10.1002/9780470277553.ch17 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2016.4618 received 2016-02-01 accepted: 2016-05-31 solving the traveling salesman problem based on the genetic reactive bone route algorithm whit ant colony system majid yousefikhoshbakht a*, nasrin malekzadeh b, mohammad sedighpour c a young researchers & elite club, hamedan branch, islamic azad university, hamedan, iran. a khoshbakht@iauh.ac.ir b young researchers & elite club, ardabil branch, islamic azad university, ardabil, iran. c hamedan branch, islamic azad university, hamedan, iran. abstract: the tsp is considered one of the most well-known combinatorial optimization tasks and researchers have paid so much attention to the tsp for many years. in this problem, a salesman starts to move from an arbitrary place called depot and after visits all of the nodes, finally comes back to the depot. the objective is to minimize the total distance traveled by the salesman. because this problem is a non-deterministic polynomial (np-hard) problem in nature, a hybrid meta-heuristic algorithm called reacsga is used for solving the tsp. in reacsga, a reactive bone route algorithm that uses the ant colony system (acs) for generating initial diversified solutions and the genetic algorithm (ga) as an improved procedure are applied. since the performance of the metaheuristic algorithms is significantly influenced by their parameters, taguchi method is used to set the parameters of the proposed algorithm. the proposed algorithm is tested on several standard instances involving 24 to 318 nodes from the literature. the computational result shows that the results of the proposed algorithm are competitive with other metaheuristic algorithms for solving the tsp in terms of better quality of solution and computational time respectively. in addition, the proposed reacsga is significantly efficient and finds closely the best known solutions for most of the instances in which thirteen best known solutions are also found. key words: reactive bone route algorithm, genetic algorithm, ant colony system, traveling salesman problem, np-hard problems. 1. introduction the traveling salesman problem (tsp) is one of the most important problems in industrial problems which is concerned with searching for the minimum hamiltonian cycle in a network of nodes by a salesman, based at one fixed node called depot, to serve a set of nodes such that: the total distance travelled by the salesman is minimized, this salesman must leave and also return to the depot. in literature, this minimum hamiltonian cycle means the closed walk that traverses every node once and only once in a graph traversing the minimum path in terms of the length of the edges. this problem belongs to hard combinatorial optimization problems that calls for the determination of the optimal sequence of deliveries conducted. their importance relies upon the fact that they are difficult to be solved but are intuitively used for modelling several real world problems (masutti et al., 2009). in practice, the basic tsp is extended with constraints, for instance, on the allowed capacity of the salesman, the length of the route, arrival, departure and service time, the time of collection and delivery of goods. it should be noted that the main goal of all tsp problems is to obtain the minimum transportation cost. in another view, there are several reasons for choosing the tsp as the problem to examine the efficiency of the new algorithm: 1. although the tsp is easily understandable, this problem is an important np-hard optimization int. j. prod. manag. eng. (2016) 4(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international 65 http://dx.doi.org/10.4995/ijpme.2016.4618 mailto:khoshbakht@iauh.ac.ir http://creativecommons.org/licenses/by-nc-nd/4.0/ problem like the n-queens problem (masehian et al., 2014) and k-sat (jaafar and samsudin, 2013). furthermore, tsp arises in several applications including the computer wiring, designing hardware devices and radio electronic devices, etc. 2. it is easily understandable, since the algorithm behavior is not obscured by too many technicalities. 3. because new algorithms are easily applied in the tsp, it is a standard test bed for new algorithmic ideas. in most of the time, a good performance of the tsp is often taken as a proof of their usefulness. for the first time, tarantilis and kiranoudis (2002) presents an adaptive memory-based method for solving the cvrp, called bone route. this concept is a population-based method producing a new solution out of components of routes of previous solutions. the components of routes used in this method are sequences of nodes, called bones. rochat and taillard (1995) introduced this concept to describe a pool of good solutions that is dynamically updated throughout the solution search process. some components of these solutions are extracted from the pool periodically and combined to construct a new solution. furthermore, tarantilis and kiranoudis (2007) also presented a more flexible adaptive memory-based algorithm for real-life transportation based on the bone route method framework for the classical vehicle routing problem. the tsp is a problem of high computational complexity (np-hard). this means that a polynomial time algorithm does not exist for it and the computational attempt required to solve this problem increases exponentially with the size of the problem. in more details, large-scale tsps (involving usually more than 100 nodes) are unlikely to be solvable in a reasonable amount of time by exact algorithms. this has led researchers to develop metaheuristics that manage to find high quality solutions in a reasonable amount of computer time. therefore, an efficient reactive bone route algorithm is combined with a modified genetic algorithm and ant colony system (acs) is proposed. in this work, a new solution from a component of the other solutions is produced while using new diversification and intensification mechanisms. the reacsga employs the acs for generating initial diversified solutions and modified ga improvement procedure for improving the initial solutions. some test problems of tsplib are considered and the results of reacsga are compared to the several metaheuristic algorithms. the results show that the proposed algorithm in comparison with these algorithms can provide better solutions. in the following parts of this paper, background and literature review is presented in section 2. in section 3, the proposed algorithm is described in more details. in section 4, the proposed algorithm is compared with some of the famous metaheuristic algorithms on standard tsp problems. finally in section 5, the conclusions are presented. 2. background and literature review the traveling salesman problem (tsp) is a wellknown optimization problem in operations research that has nowadays received much attention because of its practical applications in industrial and service problems. for these reasons, different algorithms including exact, heuristic and metaheuristic algorithms have been explored during the several decades ago. for example, the exact optimization methods which can guarantee optimality based on different techniques have been proposed for solving small-size problems with relatively simple constraints. these techniques use algorithms that generate both a lower and an upper bound on the true minimum value of the problem instance. if the upper and lower bound coincide, a proof of optimality is achieved (yadlapalli et al., 2009; cordeau et al., 2010). although optimal solutions can be obtained using exact methods for small size of tsp problems, the computational time required to solve adequately large problem instances is still prohibitive. so, the heuristic and meta-heuristic algorithms have been proposed by researchers and scientists in order to provide near-optimal solutions with reasonable computational time for large-size problems. the proposed heuristics can be separated into three categories including tour construction heuristics, tour improvement heuristics, and composite heuristics. the most powerful algorithms in this group are composite algorithms consisting of the construction algorithm to produce an initial solution and the improvement procedure for finding the best solutions (renaud et al., 1998). as the tsp is an np-complete problem, most of the heuristic algorithms do not have high quality int. j. prod. manag. eng. (2016) 4(2), 65-73 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakhta, m., malekzadeh, n., sedighpoura, m. 66 http://iajit.org/index.php?option=com_content&task=view&id=835&itemid=364 http://creativecommons.org/licenses/by-nc-nd/4.0/ for solving the tsp and then great effort has been devoted to metaheuristics that produce a good tour, if not optimal. johnson and mcgeoch (2002) conclude that these algorithms can provide remarkably good results in reasonable amounts of time for practical instances. since the meta-heuristic approaches are very efficient for escaping from local optimum, they are one of the best algorithms for solving combinatorial optimization problems. that is why the recent publications are more based on meta-heuristic approaches such as gravitational emulation search (balachandar and kannan, 2007), neural network (thiago et al., 2009), ant colony optimization (aco) (yousefikhoshbakht et al., 2013), imperialist competitive algorithm (yousefikhoshbakht and sedighpour, 2012) and particle swarm optimization (zhong et al., 2007). although heuristic and metaheuristic algorithms can obtain better quality solution compared with the exact algorithm, many researchers have found that the employment of hybridization in optimization algorithms can improve the quality of the problem in comparison with these algorithms. for examples, ant colony optimization (aco) and beam algorithm (lópez-ibáñez et al., 2010), ga with a local search (créput and koukam, 2009), aco with sweep algorithm (yousefikhoshbakht and sedighpour, 2013), threshold accepting and edge recombination (liu, 2007), particle swarm optimization and local search (shi et al., 2007), variable neighborhood descent search and grasp (hernandez-perez et al., 2009) have greater ability for finding an optimal solution to solve the complex problems. wang (2010) presented a hybrid algorithm in which ga, aco and a new strategy called gsa were proposed aiming at the key link in the algorithm. this algorithm converts the genetic solution from ga into information pheromone to distribute in aco. furthermore, gsa takes a new matrix formed by the combination of the former 90% of individuals from genetic solution and 10% of an individual by random generation as the basis of the transformation of pheromone value. the best combination of genetic operators in ga was also discussed. besides, weber (2006) proposed a distributed algorithm in which ant colony and genetic algorithms work independently of each other and only communicate when better solutions are discovered. in this algorithm, ant colonies are used to explore the solution space, while genetic algorithms are used to improve the convergence rate of the search. 3. the proposed algorithm in this section, our algorithm in the name of reacsga is presented. it should be noted that the proposed bone route algorithm is a method producing a new solution out of components of routes called the bones of previous solutions. this component of used routes is sequences of nodes. in other words, the fact behind the bone route is to extract bone nodes with predefined size and frequency of the adaptive memory (am). the size and frequency specified by the algorithm-designer restricts the number of nodes in a bone (bonesize) and the minimum number of stored routes in the am that must include a bone (bone-freq-min). it is noted that the value of the bone-freq-max and bone-size express the degree of similarity among the new constructed solution and the previously stored solutions in the am. therefore, if these values are considered high, the new solution is more similar to other solutions in am. in the proposed reacsga, n different diversified initial solutions are generated by the modified acs. the goal of building multiple initial solutions is to spread the search in order to explore different regions of the solution space of the problem (diversification strategy). at each iteration of the proposed acs, each ant builds a solution of the tsp step by step. at each step, the ant makes a move in order to complete the actual solution by following a probability function. the probability of ant k which moves from city i to city j which has not been visited yet is presented in formula (1). ( ) & & ( ) ( ) ( ) ( ) p t if q q j j if q q j j t t t t otherwise 1 0ijk ir irr j ij ij 0 0 i k ! # # x h x h = = ) ) ! a b a b z [ \ ] ] ] ] ] ] ]]| (1) where ( ) ( )argmaxj t t* ij ijr jikx h= ! a b identifies the unvisited node in ji k that maximizes pij k(t). τij(t): the value of pheromone on the arc (i,j) ηij(t): the heuristic information the arc (i,j) defined here as the reciprocal of the distance between node i and node j. int. j. prod. manag. eng. (2016) 4(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international solving the traveling salesman problem based on the genetic reactive bone route algorithm whit ant colony system 67 http://ieeexplore.ieee.org/xpl/articledetails.jsp?arnumber=5689028 http://creativecommons.org/licenses/by-nc-nd/4.0/ q: a uniformly distributed random number between 0 to 1. q0: a variable assumed 0.2 at the beginning of the algorithm. in every iteration of the algorithm, this variable is increased 0.01 until q0=0.9. the smaller q0, the higher the probability to make a random choice (0≤q0≤1). α,β: the controlling parameters by the user. in order to improve future solutions, the pheromone trails of the ants must be updated to reflect the ant’s performance and the quality of the solutions found. this updating is a key element of adaptive learning technique of acs and helps to ensure the improvement of subsequent solutions. pheromone trail is updated in global updating. after all ants have completed their schedule, the pheromone level is updated by applying the global updating rule only on the paths that belong to the best found schedule since the beginning as follows: τij(t+1)=(1-ρ)· τij(t)+ρ(1/cb) if {edge(i,j)!tb} (2) where cb: the cost of the best tour ρ: a parameter in the range [0, 1] that regulates the reduction of pheromone on the edges. tb: the best tour found by ants. this rule is intended to provide a greater amount of pheromone on the paths of the best schedule, thus intensifying the search around this schedule. in other words, only the best ant that took the shortest route is allowed to deposit pheromone. after n solutions of the tsp are built by the modified acs, they are considered as a group of initial population for ga. then crossover and mutation operations are applied to them for improving the obtained solution of the modified acs. order is still one of the best crossovers in terms of quality a speed. therefore, this method that is simple to implement has been considered here and the modified order crossover is proposed. in order crossover, a randomly chosen crossover point divides the parent strings into left and right substrings. the right substrings of the parents are selected and replace these genes based on the arrangement in other chromosomes. the only difference between the proposed crossovers with the order crossover is that instead of all the positions to the right of the selected chromosome, several random positions of all the genes in the parent are selected (figure 1). clearly this method allows only the generation of valid strings. 2 3 1 6 5 4 2 * 1 * * 4 2 3 1 5 6 4 1 3 4 2 5 6 1 3 * * * 6 1 3 2 5 4 6 (a) (b) (c) figure 1. (a). randomized selection number of genes on each chromosome. (b). finding arrangement these genes in another chromosome. (c). replacement these genes on based new arrangement. moreover, two mutations are used in the proposed algorithm. these operators randomly select two points in the string, and it replaces together or reverses the substring between these two cut points (figure 2). 2 5 1 6 3 4 (a) 2 3 1 6 5 4 2 5 6 1 3 4 (b) figure 2. two used mutations in a chromosome. in this step, the am is constructed using improved solutions of previous step in which their routes are sorted by increasing costs of relative solutions. in each iteration, the s randomly detects bone-like sequences of a predetermined number of nodes using the nearest neighborhood heuristic. then bone sequences are used to construct a new solution. it is noted that selected bones must not contain common nodes between them. in order to avoid this, the bones are selected that belong to the high quality solution. furthermore, if those solutions have the same cost, that one is choosing with the highest frequency. then, combining the extracted bones by applying the modified acs generates new solutions. in this step, three different neighborhood operators including insert, swap and 2-opt algorithms are used to improve the new constructed solution generated in the previous step and set as the best solution, if it is better than the previous elite solution. in insert move a node from its position is changed to another position. in the swap move, two nodes from the route are changed. in 2-opt move, two non-adjacent edges int. j. prod. manag. eng. (2016) 4(2), 65-73 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakhta, m., malekzadeh, n., sedighpoura, m. 68 http://creativecommons.org/licenses/by-nc-nd/4.0/ are replaced by two other edges. it should be noted that there are several routes for connecting nodes in order to produce the tour again, but a state that satisfies the problem’s constraints is acceptable. these operators are shown in figure 3. figure 3. insert (left), swap (middle) and 2-opt exchanges (right). after, if am is not full, the am is updated and the new improved solution is added, otherwise am is updated by inserting the routes of new improved solution and removing routes that belong to the worst solution when the new improved solution is better than the low quality solution in am. if the best solution of algorithm in current iteration has been improved, the search process is intensified in neighborhood of current solutions by applying these local search algorithms again. then replace the new solution in this procedure instead of the best solution in am if it has higher quality. this technique leads to increase the convergence of the algorithm to the best solution. to effectively implement a metaheuristic, one of the most important elements is how to employ the search memory to get a compromise between the diversification and intensification in the search space of the problems. the intensification forces the search to check the neighborhood of some good solutions. such procedure is a kind of utilization and learning from the accumulated experience. nevertheless, the diversification is to explore the unexplored regions of the search space. in this step, if the meta-heuristic has not updated the best solution for a prespecified number of consecutive iterations, we must drive the search towards a part of the solution space that has not explored yet (diversification policy). so, we decrease the similarity among the solutions in am with decreasing the value of bone-size in a number of consecutive iterations. after the diversification policy, the search process is intensified by increasing the value of bone-size for a number of consecutive iterations. this reactive behavior of the adaptive memory based metaheuristic provides a diversified solution to explore the solution space more precisely. therefore, at this stage, at first the parameters are updated and then the stop condition is checked. if this condition is met, the algorithm ends. otherwise, if stop conditions are not satisfied, the algorithm is iterated. in the proposed algorithm, if the best known solution is iterated 20 times, the algorithms ends and the obtained results and values up to now are considered as the best values and the results of the algorithm. figure 4 shows the main steps of the proposed algorithm. 1: generate n initial solutions by employing the acs. 2: improve the quality of solutions produced in step 1 by using the ga and determine the best solution. 3: construct the adaptive memory (am) and sort the routes by increasing the costs of their relative solutions. repeat 4: extract the promising bone sequences of nodes according to some selection criteria. 5: extract the selected bones from the am, according to the values of the bone-size and bone-freq-min. 6: employ the acs to generate n new solutions using the extracted bones. 7: improve the quality of the newly constructed solution generated in step 6 by using ga and obtain the best current solution. 8: update am. 9: if the best solution until now is improved, apply several local search algorithms and then update am (if necessary) 10: update parameters. until the best solution has not been changed for 20 iterations. figure 4. outline of the reacsga. 4. experiments and computational results the whole algorithmic approach was implemented in c and implemented on a pentium 4, 3 ghz (2 gb ram) pc with windows xp. like every metaheuristic algorithm, the quality solutions produced by the proposed algorithm have been dependent on the seed used to generate the sequence of pseudorandom numbers and on the different values of the parameters. therefore, a number of different alternative values were tested and the taguchi int. j. prod. manag. eng. (2016) 4(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international solving the traveling salesman problem based on the genetic reactive bone route algorithm whit ant colony system 69 http://creativecommons.org/licenses/by-nc-nd/4.0/ method is implemented to tune the parameters of the proposed algorithm. finally, the ones selected are those that gave the best computational results concerning the quality of the solution. for achieving this goal, orthogonal arrays provided in this method are designed to perform examinations with varied numbers of levels. it should be noted that, increase of levels numbers results the increase of experiments numbers. so, because 13 parameters are in the proposed algorithm, the most suitable design is three-echelon experiments in this paper. in other words, according to the standard orthogonal arrays of taguchi, 𝐿27 array is considered as the suitable experimental design to tune the parameters of the proposed algorithm. 𝐿27 array is an experimental design with 27 times runs. if the experiments were run completely, it was necessary to run 37=2187 experiments to run experiments to tune the parameters of the acs and 36=729 experiments to tune the parameters of tabu search and ga. the proposed ga and tabu search algorithm run for each taguchi experiment. then the ration of s/n is calculated by minitab software v.16 and the optimum combinations of the parameters are shown in table 1 for each algorithm. although the results confirm that our parameters setting worked well, it is also possible that the better solutions may exist. thus, the selected parameters are given in table 1. all of the parameter values have been determined on the eil51 by the numerical experiments. furthermore, the proposed algorithm stops after no improvements are found for 20 iterations. the algorithm was tested on a set of 19 problems with sizes ranging from 24 to 318 nodes. the selected test problems are symmetric and euclidean tsp instances. the word ‘‘symmetric’’ means that the travel cost from city a to city b is the same as the travel cost from city b to city a. each instance is described by its tsplib name and size, e.g., in table 2 the instance named gr48 has size equal to 48 nodes. the sets of data used for the experiment are tsp instances available on the tsplib (gutin and punnen, 2002). in table 2, the first and second columns show the number and the name of each instance. the third column specifies the instance’s size referenced. moreover, the fourth, fifth, sixth, seventh, eighth, ninth and tenth columns show the cpu time and the best results of seven meta-heuristic algorithms and the eleventh column present the best solutions of the proposed algorithm through 10 independent runs. additionally, in order to recognize the performance of the method, the best solutions published (bks) in the literature, are presented in twelfth column. the last column of table 2 shows the gap value of the reacsga, where the gap is defined as the percentage of deviation from the best known solution in the literature. the gap is equal to: 100[c(s**) – c(s*)]/c(s*) (3) where s** is the best solution found by the algorithm for a given instance, and s* is the overall best known solution for the same instance on the web. a zero gap indicates that the best known solution is found by the algorithm. as can be seen in this column in table 2, the reacsga finds the optimal solution for thirteen out of nineteen problems that are published in the literature. for instance lin318, the gap is relatively as high as 1 percent. however, in other instances, the proposed algorithm finds nearly the best known table 1. optimum parameter values. parameter candidate optimum value alpha 1 3 5 1 beta 1.5 2.5 3.5 2.5 rho 0.7 0.8 0.9 0.9 q0 0.10 0.15 0.20 0.20 number of ants n/3 n/2 n n number of population created by the acs n/3 n/2 n n number of iterations as stop condition which the acs is stopped after no improvements 5 7 9 7 number of iterations as stop condition which the ga is stopped after no improvements 5 7 9 7 amsize 5 6 7 7 bonefreq_min 2 3 4 2 bonefreq_max 5 6 7 7 number of consecutive iterations. diversification policy is done 2 3 4 4 number of consecutive iterations. intensification policy is done 2 3 4 4 int. j. prod. manag. eng. (2016) 4(2), 65-73 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakhta, m., malekzadeh, n., sedighpoura, m. 70 http://creativecommons.org/licenses/by-nc-nd/4.0/ ta bl e 2. c om pa ri so n of a lg or ith m s fo r s ta nd ar d pr ob le m s of t sp . n um be r in st an ce n a c s g a n n g sa p ps o b c o a c o r e a c sg a b k s g ap b es t ti m e b es t ti m e b es t ti m e b es t ti m e b es t ti m e b es t ti m e b es t ti m e b es t ti m e 1 g r 24 24 12 72 12 72 12 72 2. 5 12 72 0 2 b ay g2 9 29 16 10 16 10 16 10 2. 36 16 10 0 3 g r 48 48 50 46 50 47 50 46 3. 87 50 46 0 4 a t t 48 48 10 62 8 10 64 3 10 66 1 10 62 8 3. 99 10 62 8 0 5 e il5 1 51 42 7 42 9 42 7 7. 95 42 7 42 7 4. 06 42 8 42 6 5. 77 42 6 4. 65 42 6 0 6 b er lin 52 52 75 42 75 48 75 42 8. 91 75 42 75 42 4. 12 75 42 5. 90 75 42 5. 13 75 42 0 7 e il7 6 76 54 2 54 9 54 1 15 .0 7 53 8 54 0 11 .5 9 53 9 54 3 18 .2 3 53 8 9. 19 53 8 0 8 k ro a 10 0 10 0 21 30 9 21 54 0 21 33 3 15 .8 7 21 28 2 21 29 6 23 .9 5 21 76 3 21 34 1 41 .6 4 21 28 2 18 .2 1 21 28 2 0 9 k ro b 10 0 10 0 22 18 3 22 43 1 22 34 3 22 .0 6 22 14 1 22 63 7 22 14 1 21 .7 2 22 14 1 0 10 k ro c 10 0 10 0 20 78 7 22 99 3 20 91 5 21 .6 5 20 74 9 20 85 3 20 75 4 25 .5 3 20 74 9 0. 02 11 k ro d 10 0 10 0 21 34 1 21 59 1 21 37 4 21 .9 2 21 30 9 21 64 3 21 33 5 21 .1 5 21 29 4 0. 19 12 k ro e 10 0 10 0 22 10 9 22 19 8 22 39 5 21 .4 9 22 06 8 22 45 0 22 06 8 19 .9 9 22 06 8 0 13 e il1 01 10 1 63 6 64 3 63 8 21 .7 2 63 0 63 5 62 9 20 .9 3 62 9 0 14 l in 10 5 10 5 14 53 4 14 70 3 14 37 9 20 .6 2 14 37 9 15 28 8 14 37 9 25 .7 7 14 37 9 0 15 k ro a 15 0 15 0 26 74 9 27 05 4 26 67 8 39 .8 9 26 52 4 27 85 8 26 61 1 55 .1 9 26 52 4 0. 33 16 k ro b 15 0 15 0 26 43 1 26 65 9 26 26 4 40 .6 1 26 13 0 26 53 5 26 20 2 54 .3 2 26 13 0 0. 28 17 k ro a 20 0 20 0 29 76 2 30 27 6 29 60 0 61 .8 8 29 38 3 29 56 3 19 8. 55 29 96 1 30 08 3 33 8. 98 29 36 8 67 .1 6 29 36 8 0 18 k ro b 20 0 20 0 29 65 3 31 98 0 29 63 7 57 .9 1 29 54 1 30 35 0 29 50 9 90 .1 0 29 43 7 0. 25 19 l in 31 8 31 8 42 83 4 11 0. 28 42 48 7 44 68 5 42 54 3 16 4. 13 42 02 9 1. 22 int. j. prod. manag. eng. (2016) 4(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international solving the traveling salesman problem based on the genetic reactive bone route algorithm whit ant colony system 71 http://creativecommons.org/licenses/by-nc-nd/4.0/ solution, i.e., the gap is below 0.35, and overall, the average difference is 0.13. a computational experiment has been conducted to compare the performance of the proposed algorithm for tsp with some of the best techniques designed including aco, ga, and particle swarm optimization (pso) (zhong et al., 2007), acs, bee colony optimization (bco) (wong et al., 2008), selforganizing neural network (nn) (masutti and castro, 2009) and genetic algorithm combined by simulated annealing and ant colony system and particle swarm optimization (gsap) (chen and chien, 2011) in table 2. the results also show that the reacsga has the ability to escape from local optimum and find the best solutions for most of the instances. the results of this comparison show that the proposed algorithm gains equal solutions to the ga in gr24 and bayg29, and it gains better solutions than the ga in other problems from gr48 to krob200. furthermore, the results indicate that although the acs gives an equal solution to the proposed algorithm for 5 instances including gr24, bayg29, gr48, att48 and berlin52, this algorithm cannot gain optimal solutions for others and yields worse solutions than the proposed reacsga algorithm. besides, in general the proposed algorithm gives better results compared to other three algorithms including pso, aco and bco algorithms in terms of the solution’s quality. the performance comparison of results shows that the proposed algorithm method clearly yields better solutions than the bco. moreover, the computational results of the reacsga and pso shows that these algorithms have a close competition and the proposed algorithm gives better 4 solutions than pso. in other words, the performance of the proposed algorithm is better in reaching the sub-optimal solution than the pso. furthermore, the results indicate that although the nn yields solutions equal to the proposed algorithm for lin105, this algorithm cannot maintain this advantage in the other examples. the proposed algorithm yields better solutions than the nn for other 18 instances. moreover, computational results of the proposed algorithm and gsap show that these algorithms have a close competition, but the proposed algorithm produces seven better solutions more than gsap. in other words, the reacsga performs better in reaching the suboptimal solution than the gsap. as a result, the proposed algorithm yields better solutions than the ga, acs, gsap, nn, pso, acs, and bco. although direct comparisons of the required computational times cannot be exactly compared, as they closely depend on various factors such as the processing power of the computers, the programming languages, the coding abilities of the programmers and the running processes on the computers, the cpu times of common instances obtained by the proposed algorithm are compared with nn, pso and aco in figure 5. it is noted that because the cpu time of all instances were not reported, only five instances eil51, berlin52, eil76, kroa100 and kroa200 are considered here. by comparing the obtained results in this figure and table 2, it is concluded that the proposed algorithm can obtain high quality solutions in acceptable time. figure 5. comparison of cpu time of algorithms for standard problems of tsp. 5. conclusions in this paper, a reactive algorithm which uses the modified acs for generating initial diversified solutions and ga for intensification mechanisms was proposed for solving the tsp as a non-deterministic polynomial (np-hard) problem. furthermore, taguchi method is used to set the parameters of the proposed algorithm in order to obtain high quality solution both in cpu time and cost. experiments are implemented to evaluate the algorithm’s performance on some test instances of tsplib. computational results demonstrate that our algorithm is effective for solving tsp. as shown, in almost 19 cases of instances, the results of mentioned algorithms can be improved by our algorithm and the gap between the bks reported and the solution found by the proposed algorithm for 13 instances is zero. using this proposed algorithm for other versions of the tsp and also applying this method in other combinational optimization problems including the vehicle routing problem, school bus routing problem and the sequencing of jobs are suggested for future research. int. j. prod. manag. eng. (2016) 4(2), 65-73 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakhta, m., malekzadeh, n., sedighpoura, m. 72 http://creativecommons.org/licenses/by-nc-nd/4.0/ references balachandar, s. r., kannan, k. (2007). randomized gravitational emulation search algorithm for symmetric traveling salesman problem, applied mathematics and computation, 192(2), 413-421. http://dx.doi.org/10.1016/j.amc.2007.03.019 chen, s. m., chien, c. y. (2011). solving the traveling salesman problem based on the genetic simulated annealing ant colony system with particle swarm optimization techniques, expert systems with applications, 38(12), 14439-14450. http://dx.doi.org/10.1016/j. eswa.2011.04.163 cordeau, j. f., dell’amico, m., iori, m. (2010). branch-and-cut for the pickup and delivery traveling salesman problem with fifo loading, computers & operations research, 37(5), 970-980. http://dx.doi.org/10.1016/j.cor.2009.08.003 créput, j. c., koukam, a. (2009). a memetic neural network for the euclidean traveling salesman problem, neurocomputing, 72(4-6), 1250-1264. http://dx.doi.org/10.1016/j.neucom.2008.01.023 gutin, g., punnen, a. (2002). the traveling salesman problem and its variations, kluwer academic publishers, dordrecht. hernández-pérez, h., rodríguez-martín, i., salazar-gonzález, j. j. (2009). a hybrid grasp/vnd heuristic for the one-commodity pickupand-delivery traveling salesman problem, computers & operations research, 36(5), 1639-1645. http://dx.doi.org/10.1016/j. cor.2008.03.008 jaafar, a., samsudin, a. (2013). an improved version of the visual digital signature scheme the international arab journal of information technology, the international arab journal of information technology, 10(6), 20-32. johnson, d.s., mcgeoch, l.a. (2002). experimental analysis of heuristics for the stsp, in g. gutin and a.p. punnen (eds.), the traveling salesman problem and its variations, kluwer academic publishers, the netherlands, 369–443. lópez-ibáñez, m., blum, c. (2010). beam-aco for the traveling salesman problem with time windows, computers & operations research, 37(9), 1570-1583. http://dx.doi.org/10.1016/j.cor.2009.11.015 masehian, e., akbaripour, h., mohabbati-kalejahi, n. (2014). solving the n-queens problem using a tuned hybrid imperialist competitive algorithm, the international arab journal of information technology, 11(6). masutti, t. a. s., castro, l. n. d. (2009). a self-organizing neural network using ideas from the immune system to solve the traveling salesman problem, information sciences, 179(10), 1454-1468. http://dx.doi.org/10.1016/j.ins.2008.12.016 renaud, j., fayez, f., boctor, f. (1998). an efficient composite heuristic for the symmetric generalized traveling salesman problem, european journal of operational research, 108(3), 571-584. http://dx.doi.org/10.1016/s0377-2217(97)00142-2 rochat, y., taillard, e.d. (1995). probabilistic diversification and intensification in local search for vehicle routing, journal of heuristics, 1(1), 47–167. http://dx.doi.org/10.1007/bf02430370 shi, x. h., liang y. c., lee h. p., lu c., wang, q.x. (2007). particle swarm optimization-based algorithms for tsp and generalized tsp, information processing letters, 103(5), 169-176. http://dx.doi.org/10.1016/j.ipl.2007.03.010 tarantilis, c. d., kiranoudis, c. t. (2002). boneroute: an adaptive memory-based method for effective fleet management.  annals of operations research, 115(1-4), 227-241. http://dx.doi.org/10.1023/a:1021157406318 tarantilis, c. d., kiranoudis, c. t. (2007). a flexible adaptive memory-based algorithm for real-life transportation operations: two case studies from dairy and construction sector, european journal of operational research, 179(3), 806-822. http://dx.doi.org/10.1016/j. ejor.2005.03.059 thiago, a.s., masutti, l., castro, n. (2009). a self-organizing neural network using ideas from the immune system to solve the traveling salesman problem. information sciences, 179(10), 1454-1468. http://dx.doi.org/10.1016/j.ins.2008.12.016 yadlapalli, s., malik, w.a., darbha, s., pachter, m. (2009). a lagrangian-based algorithm for a multiple depot, multiple traveling salesmen problem, nonlinear analysis: real world applications, 10(4), 1990-1999. http://dx.doi.org/10.1016/j.nonrwa.2008.03.014 yousefikhoshbakht, m., didehvar, f., rahmati, f. (2013). modification of the ant colony optimization for solving the multiple traveling salesman problem, romanian journal of information science and technology, 16(1), 65-80. yousefikhoshbakht, m., sedighpour, m. (2013). new imperialist competitive algorithm to solve the travelling salesman problem, international journal of computer mathematics, 90(7), 1495-1505. http://dx.doi.org/10.1080/00207160.2012.758362 yousefikhoshbakht, m., sedighpour, m. (2012). a combination of sweep algorithm and elite ant colony optimization for solving the multiple traveling salesman problem, proceedings of the romanian academy a, 13(4), 295-302. wang, c. (2010). a hybrid algorithm based on genetic algorithm and ant colony optimization for traveling salesman problems. the 2nd international conference information science and engineering, 4257-4260. http://dx.doi.org/10.1109/icise.2010.5689028 weber, b. (2006). distributed hybrid metaheuristics for optimization, work paper, 1-12. wong, l. p., low, m. y. h., chong, c. s. (2008). a bee colony optimization algorithm for traveling salesman problem. modeling & simulation, 2008. aicms 08. second asia international conference, 818–823. http://dx.doi.org/10.1109/ams.2008.27 zhong, w., zhang j., chen, w. (2007). a novel discrete particle swarm optimization to solve traveling salesman problem. evolutionary computation, 3283–3287. int. j. prod. manag. eng. (2016) 4(2), 65-73creative commons attribution-noncommercial-noderivatives 4.0 international solving the traveling salesman problem based on the genetic reactive bone route algorithm whit ant colony system 73 http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6ty8-4n987x7-m&_user=10&_coverdate=09%2f15%2f2007&_alid=1464318684&_rdoc=14&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5612&_st=13&_docanchor=&_ct=198&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=fcdf082d2786b3597f0180f62c3e5abd&searchtype=a http://dx.doi.org/10.1016/j.amc.2007.03.019 http://dx.doi.org/10.1016/j.eswa.2011.04.163 http://dx.doi.org/10.1016/j.eswa.2011.04.163 http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6vc5-4x1ycm2-1&_user=10&_coverdate=05%2f31%2f2010&_alid=1464303336&_rdoc=19&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5945&_st=13&_docanchor=&_ct=1030&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=0b1fb9a2454c922a18bd672b02f051fe&searchtype=a http://dx.doi.org/10.1016/j.cor.2009.08.003 http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6v10-4s0204s-1&_user=10&_coverdate=01%2f31%2f2009&_alid=1464303336&_rdoc=11&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5660&_st=13&_docanchor=&_ct=1030&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=aa69cb80088b2f1c05eebaf154a8f7df&searchtype=a http://dx.doi.org/10.1016/j.neucom.2008.01.023 http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6vc5-4s5fjjb-1&_user=10&_coverdate=05%2f31%2f2009&_alid=1593540283&_rdoc=1&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5945&_sort=r&_st=13&_docanchor=&view=c&_ct=1635&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=7bf1385951f9bf596d6a83448109c6b1&searchtype=a http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6vc5-4s5fjjb-1&_user=10&_coverdate=05%2f31%2f2009&_alid=1593540283&_rdoc=1&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5945&_sort=r&_st=13&_docanchor=&view=c&_ct=1635&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=7bf1385951f9bf596d6a83448109c6b1&searchtype=a http://dx.doi.org/10.1016/j.cor.2008.03.008 http://dx.doi.org/10.1016/j.cor.2008.03.008 http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6vc5-4xvc4vp-2&_user=10&_coverdate=09%2f30%2f2010&_alid=1464303336&_rdoc=18&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5945&_st=13&_docanchor=&_ct=1030&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=9bbc8b97cf3766540b30bac060a46167&searchtype=a http://dx.doi.org/10.1016/j.cor.2009.11.015 http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6v0c-4v88fsk-1&_user=10&_coverdate=04%2f29%2f2009&_alid=1464318684&_rdoc=10&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5643&_st=13&_docanchor=&_ct=198&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=7a889f16356d2e4612cbfc2ed657ed1b&searchtype=a http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6v0c-4v88fsk-1&_user=10&_coverdate=04%2f29%2f2009&_alid=1464318684&_rdoc=10&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5643&_st=13&_docanchor=&_ct=198&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=7a889f16356d2e4612cbfc2ed657ed1b&searchtype=a http://dx.doi.org/10.1016/j.ins.2008.12.016 http://econpapers.repec.org/article/eeeejores/ http://econpapers.repec.org/article/eeeejores/ http://dx.doi.org/10.1016/s0377-2217(97)00142-2 http://dx.doi.org/10.1007/bf02430370 http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6v0f-4ncsgty-2&_user=10&_coverdate=08%2f31%2f2007&_alid=1593535328&_rdoc=57&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5645&_sort=r&_st=13&_docanchor=&view=c&_ct=1635&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=f829879b220712f530d9249c48d680b2&searchtype=a http://dx.doi.org/10.1016/j.ipl.2007.03.010 http://dx.doi.org/10.1023/a:1021157406318 http://dx.doi.org/10.1016/j.ejor.2005.03.059 http://dx.doi.org/10.1016/j.ejor.2005.03.059 http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6v0c-4v88fsk-1&_user=10&_coverdate=04%2f29%2f2009&_alid=1464318684&_rdoc=10&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5643&_st=13&_docanchor=&_ct=198&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=7a889f16356d2e4612cbfc2ed657ed1b&searchtype=a http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6v0c-4v88fsk-1&_user=10&_coverdate=04%2f29%2f2009&_alid=1464318684&_rdoc=10&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=5643&_st=13&_docanchor=&_ct=198&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=7a889f16356d2e4612cbfc2ed657ed1b&searchtype=a http://dx.doi.org/10.1016/j.ins.2008.12.016 http://www.sciencedirect.com/science?_ob=articleurl&_udi=b6w7s-4s2f5p7-1&_user=10&_coverdate=08%2f31%2f2009&_alid=1464303336&_rdoc=7&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=6634&_st=13&_docanchor=&_ct=1030&_acct=c000050221&_version=1&_urlversion=0&_userid=10&md5=c71a63c2441ea104cc822d94f8783394&searchtype=a http://dx.doi.org/10.1016/j.nonrwa.2008.03.014 http://scholar.google.nl/citations?view_op=view_citation&hl=nl&user=0ioq81aaaaaj&citation_for_view=0ioq81aaaaaj:u5hhmvd_uo8c http://dx.doi.org/10.1080/00207160.2012.758362 http://ieeexplore.ieee.org/xpl/articledetails.jsp?arnumber=5689028 http://ieeexplore.ieee.org/xpl/mostrecentissue.jsp?punumber=5680733 http://dx.doi.org/10.1109/icise.2010.5689028 http://ieeexplore.ieee.org/xpl/mostrecentissue.jsp?punumber=4530427 http://ieeexplore.ieee.org/xpl/mostrecentissue.jsp?punumber=4530427 http://dx.doi.org/10.1109/ams.2008.27 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering http://dx.doi.org/10.4995/ijpme.2016.4177 received 2015-10-20 accepted: 2015-12-16 cost modelling as decision support when locating manufacturing facilities christina windmark* and carin andersson production and materials engineering, faculty of engineering, lund university, 22100 box 118, lund, sweden * christina.windmark@iprod.lth.se abstract: this paper presents a methodology for cost estimation in developing decision support for production location issues. the purpose is to provide a structured work procedure to be used by practitioners to derive the knowledge needed to make informed decisions on where to locate production. this paper present a special focus on how to integrate cost effects during the decision process. the result is a structure of cost estimation tools aligned to different steps in the work procedure. the cost models can facilitate both cost estimation for new production configurations and cost simulations to analyse the risks of wrong estimations and uncertainties in the input parameters. future research aims to test the methodology in ongoing transfer projects to further understand difficulties in managing global production systems. cost is usually estimated, in existing models and methods presented in the literature, on a too aggregated level to be suitable for decision support regarding production system design. the cost estimation methodology presented here provides new insights on cost driving factors related to the production system. key words: cost analysis, location decision, production cost. 1. introduction to be competitive and increase profitability, many manufacturing companies have to enter the global arena in both sales and production (aspelund and butsko, 2010; rusten and bryson, 2010; bell et al., 2003). this need for internationalisation results in board decisions on manufacturing relocation in the form of both outsourcing and manufacturing of products at plants in new locations. in discussions of moving manufacturing activities, three distinct terms are used: relocating, offshoring, and outsourcing. relocating refers to moving manufacturing activities within the company, between national or international sites (kinkel et al., 2007). outsourcing refers to transferring manufacturing activities from internal control to external control, mainly to reduce the production cost by letting a subcontractor to produce the product at a lower cost than the contractee can (nordigården, 2007). offshoring refers to moving the manufacturing activities of a company abroad. note that offshoring can refer to both manufacturing relocation and outsourcing (kinkel et al., 2007). this paper will concentrate on company relocation and the establishment of new processes and facilities within the company. the aim is to support companies in making informed decisions on production of key products and to facilitate the make or buy process. a german survey conducted by kinkel et al. (2007) compared the motives for offshoring with those for back-sourcing. the survey found that 87% of the studied companies considered production cost factors as the main drivers of offshoring, while 52% considered production costs as the main drivers for back-sourcing. the survey indicated that costs were a main reason for corporate offshoring and back-sourcing. companies also back-sourced in order to supply particular sites and customers and to coordinate costs. in the case of offshoring, the ability to supply customers was a key factor, whereas coordination costs were of minor importance. the survey found that companies frequently overestimated the cost benefits of offshoring and did not completely understand the conditions at the new location. platts and song (2010) interviewed informants from several companies that had outsourced to china, finding that costs ended up 25–50% higher than quoted. some studies find that relocation decisions are based on inadequate and uninformed consideration, often resulting in manufacturing activities eventually being repatriated to their original location (whitten and leidner, 2006; int. j. prod. manag. eng. (2016) 4(1), 15-27creative commons attribution-noncommercial-noderivatives 4.0 international 15 http://dx.doi.org/10.4995/ijpme.2016.4177 http://creativecommons.org/licenses/by-nc-nd/4.0/ kinkel, 2009). the survey performed by kinkel et al. (2007) indicated that approximately 20% of the 1450 surveyed german companies had conducted some sort of back-sourcing. as evident from the above studies, one important reason for relocation is cost. one possible conclusion from the above investigations is that companies could benefit from structured decision support which integrates various costs associated with relocating the production. this paper will present a cost analysis methodology which can be integrated in a decision process for production location. the decision support was developed during a research project, aiming to facilitate manufacturing footprint decisions. the project, which aimed to develop a structured costbased decision support for production relocation issues, was a three-year collaboration between five companies and two universities. the developed decision-support process, based on stage-gate principles and previously published in andersson et al. (2013) and bellgran et al. (2013). the work procedure aims to support the decision process preceding the realisation of a production location decision, and involves selecting and comparing various location alternatives. the user is guided through a series of activities in each step supported by a variety of tools and templates for analysing, for example, the costs and risks in each step. the process consists of five sequential phases representing the main activities in a location decision process: 1. initiation establish motives and goals for changing the manufacturing footprint. 2. scoping establish a project organization and plan for executing the relocation project and roughly estimate the costs and benefits. 3. pre-study analyse the current footprint and the requirements for and consequences of changing the footprint. 4. generation of alternatives establish and analyse various location alternatives. 5. location selection prepare to select the most suitable production location. the last four steps of the work require different types of cost analysis at different levels of detail. the aim is to capture all costs affected by a change in location. the scoping phase (2) requires rough estimates of market potential, investment range, and project organisation. for this purpose, a rough business case design has been presented by windmark and andersson (2014). when analysing opportunities within the current manufacturing footprint (phase 3) and different location alternatives (phase 4), a more detailed cost analysis is required. costs associated with production ramp-up, skills provision, and the impacts of moving products from an existing production site also need to be considered, motivating the development of several cost analysis tools for the different phases. the cost analysis methodology and tools presented here are connected to phase 3 (p3) and phase 4 (p4), as showed in the list below. those marked with “*” are presented in this article, while the others are presented more as concepts: current manufacturing cost analysis (p3)* cost for supporting processes (p3, p4)* cost impact of relocation (p3) checklist of location factors (p3)* scenario cost analyses of manufacturing (p4)* installation and ramp-up cost analysis (p4) costs for knowledge & skills provision (p4) 2. parameters and models the following section reviews the literature on the parameters used when evaluating locations and designing decision supports. the purpose is to investigate the parameters, categories, functions, and stages included in existing location models and decision-support models. 2.1. cost parameters and factors for location decisions the importance of considering cost minimisation and other cost-driving factors is highlighted in boloori et al. (2012). they have classified facility location models in logistics and production, demonstrating that 48 of the 66 reviewed papers had the minimisation of cost, time, distance, and risk as the main objectives. numerous studies address the importance of cost awareness (maccarthy and atthirawong, 2003; fang and weng, 2010; platts and song, 2010), identifying the important cost groups as personnel costs, project costs, and investments, but no thorough analysis or decision support focuses on the cost influence of the combined financial impact of relocation. mccarthy and attirawong (2003) identified five major factors that may influence location decision: costs, int. j. prod. manag. eng. (2016) 4(1), 15-27 creative commons attribution-noncommercial-noderivatives 4.0 international windmark, c. and andersson, c. 16 http://creativecommons.org/licenses/by-nc-nd/4.0/ infrastructure, labour characteristics, governmental and political situation, and economic factors. kinkel et al.’s (2007) survey of german companies revealed that the main driving forces of offshoring in the investigated companies were access to lowcost production factors, access to markets, support of trade and distribution, proximity to customers, support of services, access to technologies, access to resources and materials, ability to counter-attack competitors, the search for strategic assets, tax incentives and benefits, and access to excellent infrastructure. ellram et al. (2013) compiles a list of the driving factors of current global manufacturing location decisions; these factors include logistic costs, shipping time, supply chain response and recovery time, labour costs, labour productivity, environmental issues, currency stability, and theft of intellectual property. 2.2. models and processes for location decisions yang and lee (1997) present a decision model based on an analytical hierarchy process (ahp) consisting of seven steps: (0) justifying and identifying the facilities, (1) identifying location factors, (2) developing priority weighting, (3) collecting data and ranking the potential locations, (4) analysing comparative results, (5) identifying preferred site(s), and (6) making final recommendations. the model acknowledges that often no location is “optimal” but that different sites can have different advantages so that, after some compromises, the best option can be found. the model focuses on the relationships between factors for each site, presenting relative rather than absolute preference information. one of the difficult steps in the model is that of developing adequate priority weighting, which can be facilitated by factual knowledge and actual cost information. the model only indicates the best alternative based on the inputs and not whether it is profitable to move manufacturing to the actual location. christodoulou et al. (2007) present a decision support for relocation, providing companies with strategies for gathering and evaluating data. it is a wide-ranging tool, focusing more on qualitative than quantitative parameters and thus lacking in-depth general manufacturing analysis and cost calculations and estimates. a decision model for production location presented by dogan (2012) combines bayesian networks and total cost of ownership. their literature study is extensive and cites several examples of relevant papers in the field. the model has four steps identifying: (1) supplier selection criteria, (2) factors, (3) cost elements, and (4) total costs. the model combines qualitative parameters such as labour skills and worker motivation with quantitative parameters such as wages, when estimating the labour costs. many production location models include costs, but often omit guidelines for making the needed estimates (yang and lee, 1997; christodoulou et al., 2007; dogan, 2012). the present literature survey identifies a need for improved cost estimation when relocating and outsourcing, due to higherthan-expected final costs (platts and song, 2010). costs have been demonstrated to be crucial when relocating, so a model based on cost estimation can be considered very useful. the literature survey identifies a gap in how cost estimates are integrated into current support frameworks. the location decision support tools found in the literature do not involve extensive cost analysis taking production performance into consideration and therefore do not closely quantify the cost impacts. 3. method and motivation the literature review indicates that cost is considered a key factor when locating production. nevertheless, the literature overlooks how manufacturing costs are integrated into current decision support models for production location. the development of the location decision support presented here was motivated by industry statements on the need to integrate cost effects into the decision process. research methods involving the industrial partners were selected because the research performed has a strong industrial motivation. the working process and its tools and templates were developed in parallel through interviews, literature studies, observations at companies, participation in case studies at companies, frequent discussions and workshops at the participating companies, and validations in the research group. during the workshops and interviews at the companies, the companies’ requirements in terms of costs, risks, and strategic analyses were discussed. this resulted in a work procedure to support the decision making in a production location project; this procedure is a five-step stage-gate model together with a selection of cost-based tools and templates, all documented in a handbook (andersson et al. (2013). int. j. prod. manag. eng. (2016) 4(1), 15-27creative commons attribution-noncommercial-noderivatives 4.0 international cost modelling as decision support when locating manufacturing facilities 17 http://creativecommons.org/licenses/by-nc-nd/4.0/ the main purpose of this paper is to describe the cost tools developed in connection with phases p 3 and p 4. the literature presented above and a previous interview study (windmark and andersson, 2012) indicated that a wide range of parameters and factors are needed, suggesting that a locationdecision support model must include extensive analyses. the developed tools were then presented to the participating companies during workshops and projects in which input was given to ensure industrial relevance and possible implementation. the paper presents economic estimates used in a decisionsupport process taking specific manufacturing considerations from a particular location into account and combining manufacturing performance with location-specific parameters. in addition, analyses of costs associated with other supporting activities in the supply chain are included in the model. the goal is to present a method for deciding if a new location provides more cost advantages than the current one or which of several alternative locations provides the most cost advantages. this paper proposes a structure for analysing production costs to be used when making location decisions. 4. cost estimations and simulations supporting location decisions the costs affected by a change in manufacturing footprint depend on the organisational changes driven by the footprint change and by the costs of implementing them. when the production of a certain product is moved, individual sites might be subjected to changes in product ranges and capacity, causing increased costs for the remaining products manufactured at those sites. according to simons et al. (2000), the work-process functions included in the most basic organisations are a marketing and sales unit, controller’s department, information technology (it) department, and production unit. since the location analysis method presented here focuses on the design and performance of the production system, the cost of the production unit is broken down into manufacturing process costs and costs of the support functions necessary to deliver products to customers on time. figure 1 illustrates the basic structure of the manufacturing system cost drivers used to develop the cost analysis tool structure in phases p 3 and p 4 of the process presented here. the consolidated cost analysis assembled in phase p 5 of the work procedure includes the integrated impact of the cost drivers; (1) manufacturing processes, (2) production support functions, (3) impact on current location, (4) knowledge provision and (5) ramp-up and testing. the costs of knowledge provision in the new site, process testing, and ramp-up are cost drivers that must also be accounted for in a complete analysis. these costs can be regarded as generated before production starts and can be treated as investments in equipment and, in the end, be included in the process cost. in this paper, the cost analysis methods developed for manufacturing process costs and production support process costs are implemented in separate excel tools. the concept of part cost is used, meaning that all costs are presented per unit. different distribution keys (e.g., annual production volume) are therefore used to transform department and personnel costs when necessary. a crucial factor when establishing a knowledge-based decision support is the quality of input data. data acquisition can be a challenge for both existing and new production sites. to adopt a structured approach, bjelkemyr et al. (2013) suggest categorising location parameters based on the various functions of a corporate organisation: 1. sales and marketing: market price, market share, costs of the marketing and sales division, etc. 2. sourcing and purchasing: raw material price, costs of purchasing division, etc. 3. legal and finance: interest rates, tax levels, costs of regulation investigations, etc. 4. facilities and it: building costs, hardware & software costs, infrastructure costs, etc. 5. human resources: salary costs, insurance costs, moving costs, pensions, etc. figure 1. cost structure of a manufacturing operation used for cost tool development. int. j. prod. manag. eng. (2016) 4(1), 15-27 creative commons attribution-noncommercial-noderivatives 4.0 international windmark, c. and andersson, c. 18 http://creativecommons.org/licenses/by-nc-nd/4.0/ 6. r&d: additional personnel and office costs, moving costs, etc. 7. operations: salary costs, equipment investments, performance, energy costs, etc. 8. installation and ramp-up: travel, living, and personnel costs; testing costs, installation costs, etc. the last item on the list is an activity, rather than a corporate function, often necessary when establishing a new manufacturing unit and changing locations. based on the above categories, we propose a checklist to provide the various cost analysis tools with the required input data. the checklist serves as a gross list of input data and is aligned with the data requirements of the cost analysis tools. a tradeoff always exists between the effort spent acquiring information about the production system and the possibility of making well-informed decisions; the purpose of the checklist is to provide insight into important cost-driving factors and parameters. the checklist contains a total of 69 quantitative parameters structured in the eight categories listed above. in addition to the estimated or measured values, the checklist also contains a column for estimating the quality (or risk of making an erroneous estimate) of each input data item. the quality of the estimated data is also dependent on whom or what corporate function responsible for data acquisition. the checklist should therefore be specified with the support of employees of the various functions associated with the location project, and include information about who is responsible for gathering data. as the location project proceeds and the amount of information increases, the checklist can be updated to improve the quality of the input data. figure 2 shows the various cost categories in the production system included in the analysis in phase p 3 and p 4. the analysis in these two phases is based on the same methodology, but in phase p 3 the current manufacturing site (if there is one) is analysed, while in phase p 4 the selected alternative locations are investigated. it is a challenge to estimate investments, operator costs, and production performance for facilities not yet realised. to overcome the risk of over or underestimating input parameters, scenario simulation is a powerful methodology. this enables the analysis of both the bestand worst-case scenarios as well as the sensitivity analysis of individual cost drivers. the following sections present the methodology for analysing manufacturing and support costs and use the scenario capability to analyse various fictive location scenarios. 4.1. phase 3: current footprint in phase p 3 – current footprint, the decision support encourages the analysis of the existing organisation, systems, and products. to achieve this, cost models have been integrated into two tools, one focusing on the manufacturing process and one on the supporting activities. manufacturing cost estimation the foundation of the cost analysis structure is a performance-driven manufacturing part cost model designed for production development (ståhl et al., 2007). the model outcome is the cost per part in a manufacturing process. a special feature of this model is that production performance, in terms of quality, availability, and production speed rate losses, are taken into consideration and directly connected to the costs integrated in the cost model. other factors central to the model are the set-up time, cycle time, and batch size. several processes are often involved in manufacturing the products. this cost model is based on the principles of activitybased costing (cooper and kaplan, 1998), making it possible to allocate costs to particular activities and to visualize the cost drivers in an organisation, so that the economics of operations can be understood and improved. the process part cost is influenced by five cost categories (see figure 3), each consisting of several cost drivers that together constitute the total  figure 2. costs affecting the final part cost. int. j. prod. manag. eng. (2016) 4(1), 15-27creative commons attribution-noncommercial-noderivatives 4.0 international cost modelling as decision support when locating manufacturing facilities 19 http://creativecommons.org/licenses/by-nc-nd/4.0/ production part cost. the maintenance costs can be seen as costs connected either to a specific product or to general equipment not connected to specific manufactured products. since a product is usually machined and assembled in a series of steps, and the total production cost comprises the accumulated costs of all processing steps, the cost model is applied to each individual step. in the first step, the material cost corresponds to the purchased raw material. in the second step, the material cost corresponds to the manufacturing costs in the first step and any additional raw materials used in this step. this procedure is repeated throughout the manufacturing chain to yield the total manufacturing cost for the finished part. with this methodology, the cost of poor quality can be determined after each process step, visualising the cost effects of quality defects occurring early or late in the manufacturing chain. when calculating hourly equipment and personnel costs, the available production time must be established, since it is the basis on which costs are allocated. available production time is all the time when personnel is paid and/or equipment is in operation. e.g. a 24 hour operation 7 days a week gives an available production time of 8760 hours per year. if production downtime occurs, the total time is unaffected. instead, the performance parameters connected to availability and equipment utilisation are affected. personnel costs are highly dependent on where the production is located and are often one of the main reasons for a relocation decision (brouwer et al., 2004; windmark and andersson, 2012), as the location determines remuneration, employer contributions, and whether personnel receive free meals and free housing. other factors affecting personnel costs are the daily working hours and the policies on employees’ standard of living. figure 4 shows the design of the excel tool for the manufacturing cost analysis. in the right-hand column, general input data and data on equipment and maintenance are specified. the equipment cost per hour and the manufacturing cost per part are calculated, based on the input specified in the white cells. in the left-hand column, factors connected to the actual products are taken into consideration. here the production performance parameters of availability, performance, and quality, are the three constituents of the overall equipment efficiency (oee) measure, because they are associated with a specific product. these parameters are often regarded as equipment specific, but analysis made by stål et al. (2012) shows that oee can vary between products manufactured using the same equipment. estimating costs of supporting processes relocating or setting up a new production facility will likely affect the costs of various production-support functions. here we regard support functions as those required to ensure that products reach customers on time, as follows: it support, marketing, purchasing, quality assurance, internal/inbound logistics, external logistics, management, and additional costs (to capture other costs not connected to the specified figure 3. factors and parameters affecting the process part costs. int. j. prod. manag. eng. (2016) 4(1), 15-27 creative commons attribution-noncommercial-noderivatives 4.0 international windmark, c. and andersson, c. 20 http://creativecommons.org/licenses/by-nc-nd/4.0/ functions). these were identified and discussed together with the five industrial partners. table 1 shows parameters for different support functions that are considered in estimating the cost of each function. the purpose is to estimate the cost per part incurred by the support functions necessary for production systems operation, in order to capture costs likely to be affected by a change in location. opening a new plant could entail the opening of a new purchasing department or the expansion of an existing one. the new location could also require new marketing units for the local market. the logistics cost will also depend on where the plant is located, which will affect the plant design and hence the internal logistics configuration. a local it-support unit might be needed and the cost of local management should also be included. the support cost for management includes both financial personnel and managers connected with production. a new location would also require new local suppliers, were costs for identifying and quality figure 4. a tool for estimating process part cost. table 1. input costs and parameters in the economic tool for process support costs. a nn ua l p ro du ct io n a nn ua l p ro du ct io n p ro du ct pe rs on ne l c os ts r en ta l c os ts c om pu te rs e qu ip m en t c os ts m ai nt en an ce l ic en si ng c os ts tr av el c os ts e du ca tio n co st s a dv er tis in g in su ra nc e co st s n um be r o f p ar ts in d el iv er y d el ay c os ts tr an sp or ta tio n co st s (t ra ns po rt ) tr an sp or ta tio n co st s (p er so nn el ) d ut y/ de liv er y ti ed c ap ita l ( sh ip pi ng )/ pa rt ti ed c ap ita l ( st or ag e) /p ar t pl an ni ng p er so nn el c os t r en t c os t p la nn in g pe rs on ne l a dd iti on al c os ts in bo un d lo gi st ic s co st s it support x x x x x x x x x x marketing x x x x x x x x x x purchasing x x x x x x x x x management x x x x x x x x x quality assurance x x x x x x x x x x external logistics x x x x x x x x x x x x x x x inbound logistics x x int. j. prod. manag. eng. (2016) 4(1), 15-27creative commons attribution-noncommercial-noderivatives 4.0 international cost modelling as decision support when locating manufacturing facilities 21 http://creativecommons.org/licenses/by-nc-nd/4.0/ assuring these could vary substantially, depending on the maturity in the region. the personnel costs comprise wage costs and additional costs for facility, equipment and travels. if personnel are relocated, living costs should also be included. these are separated into individual cost drivers to allow analysis of the impact of each of them on the total cost. the annual costs of these functions and the annual production volume are specified to determine the estimated part cost. cost models for inbound logistics have previously been developed by windmark and andersson (2015) and are therefore not considered in detail here. the external logistic costs vary considerably depending on the product, business contract, and type of logistics transportation. due to the wide range of possible cost estimates, the users can choose to use either the predetermined parameters or a fixed cost per part based on their own estimates. the predetermined parameters are of four types: (1) costs of insurance and delay, (2) costs connected to product transportation, (3) costs connected to inventory and storage outside the manufacturing plant, and (4) costs connected to planning the external logistics. burns et al. (1985) present an extensive cost model for calculating transportation costs which is close in comparison to the input factors for external logistics in table 1. 4.2. phase p4: generating alternatives in phase p4, various location alternatives are compared. in some cases, the estimation of costs and productivity for new locations is made problematic by data collection difficulties. due to the high risk of inaccurate data, the tools used in this phase are constructed to allow for scenario analysis. this makes it possible to analyse both the impact of different factors on the total cost and the cost range of the produced product. the cost analysis methodology used in phase p3 can be reused in this phase for estimating the cost of alternative production systems and for estimating the production costs of the remaining production in existing plants. when a well-functioning performance measurement system is not in place, the oee data is recommended to be estimated based on experience. in the case of location comparison, gathering data for the various alternatives can be challenging. when configuring a new production cell or line, many estimates are needed, for example, of machine costs. to be able to design the production system and its capacity, a thorough analysis of the potential market is necessary. decisions on whether to buy new or use existing equipment is also required together with estimates of the required number of annual operation hours to meet the market demands. the available production time depends on market demand and operators’ daily working hours (latino et al., 2013). the level of education and skills in a country or region are also important when deciding what process technology to use (brynjolfsson and hitt, 2000). if new equipment will be used, it´s recommended to acquire information on equipment performance to be able to analyse different cost scenarios in in phase p4. if the company already has experience from similar equipment, the estimates could be more accurate than if the technology is new to the company. scenario analyses of three fictive cases in the following sub-section, three fictive production alternatives are analysed to illustrate how economic tools can be used for location decision support. the cases involve both a scenario on improvements by investing in a current production facility and an analysis of a new location in an attractive area for swedish companies to relocate. the main objective is to reduce costs. the three cases are: (1) sweden – current: an existing assembly line, (2) sweden – new: same assembly line as the current one but with additional investments to reduce cycle time, improve performance parameters, and increase the level of automation, (3) china: a new assembly line close in configuration to the current assembly line but involving more manual work. in the three cases, raw material is obtained from each location’s region, captured by using differentiated cost levels for the raw material. due to greater distance from the design department and the fact that new suppliers are needed, the quality rate is estimated to be lower for the production site in china. the personnel costs include assembler wages, employer contributions, and technician salaries. the ingoing parameters needed to perform the scenario analyses are shown in tables 2 and 3. to simplify the estimation and simulation, the occupancy level is set to 100%. from the data above, the following part costs are calculated: sweden – current 21.6 €/part sweden – new 18.2 €/part china 15.9 €/part to reduce the risk of erroneous estimates, various bestor worst-case scenarios can be simulated to int. j. prod. manag. eng. (2016) 4(1), 15-27 creative commons attribution-noncommercial-noderivatives 4.0 international windmark, c. and andersson, c. 22 http://creativecommons.org/licenses/by-nc-nd/4.0/ provide information on how sensitive the results are to changes in different parameter values. figure 5 shows examples, of how the total part costs are influenced by the changes in quality rate and the downtime rate in all three scenarios. the diagram shows that the impact of downtime differs depending on the equipment setup. the influence of other isolated parameters or the combined effect of changes in more than one parameter can be simulated as well. a typical location scenario could be a change in automation level that would increase the equipment cost and at the same time reduce the operator cost. boston consulting group (2011) shows that the wage costs in china are increasing at much faster rates than in the usa. the annual wage cost increase by 2% for sweden and 10% for china was used as the base for this scenario analysis. table 2. data concerning equipment costs in the assembly line. process equipment costs c os t o f c ap ita l ( in te re st ) ( % ) r en t ( €/ m 2/ ye ar ) e ne rg y co st ra te (€ /k w h) to ta l i nv es tm en t ( €) a re a of o cc up at io n (m 2) te ch ni ca l l if et im e (y ea rs ) e ne rg y co ns um pt io n – o pe ra tio n (k w /h ) e ne rg y co ns um pt io n – id le (k w /h ) sh ar e of id le ti m e (% ) pl an ne d m ai nt en an ce ( se rv ic e) ti m e (h ) pe rs on ne l c os ts ( m ai nt en an ce ) ( €/ h) c os t o f s pa re p ar ts (€ ) sweden – current 8 100 0.08 400,000 450 15 18 12 100 110 25.5 2000 sweden – new 8 100 0.08 500,000 450 15 24 12 100 100 25.5 2000 china 8 50 0.10 350,000 600 15 18 12 100 150 5 2000 table 3. data concerning general input, performance parameters, tools, and personnel costs in the assembly line. general input performance parameters tools personnel pl an ne d an nu al p ro du ct io n (u ni ts ) a va ila bl e pr od uc tio n tim e (h ) r aw m at er ia l ( €/ pa rt ) b at ch s iz e o cc up an cy le ve l ( % ) q ua lit y ra te (% ) sp ee d pe rf or m an ce (% ) a va ila bi lit y (n ot in cl ud in g se tup ) ( % ) id ea l c yc le ti m e (m in ) se tup ti m e (m in ) to ol c os t ( €/ pa rt ) c os t o f l iq ui ds a nd p ro ce ss a dd iti ve s (€ /p ar t) pe rs on ne l c os ts (€ /p er so nn el ) n um be r o f o pe ra to rs sweden – current 100,000 3600 15 1 200 100 95 93 85 2 40 0 0.02 25 3 sweden – new 100,000 3600 15 1 200 100 98 93 90 1.5 40 0 0.02 25 3 china 100,000 3600 12 1 200 100 90 93 80 2 40 0 0.02 4 7 int. j. prod. manag. eng. (2016) 4(1), 15-27creative commons attribution-noncommercial-noderivatives 4.0 international cost modelling as decision support when locating manufacturing facilities 23 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 5. simulated part costs when the quality rate, downtime rate, and personnel cost vary. the result in figure 5 shows that the quality rate has a greater effect on the final cost than does the downtime rate. a conclusion is that it is more critical to estimate the quality in a correct way. with different salary growth rates, the cost of the part produced in china will approach the level of the “sweden – new” case. these simulations only consider the costs of assembly, while the support functions such as logistics and it are not included. a relocation of production is likely to cause changes in more than one parameter; therefore scenarios involving a set of changes for the new location in china are performed. scenario: the quality rate in china is estimated to decrease from 90% to 80%. the assembler wage costs are estimated to increase by 10% annually. for the other parameters, the conditions are as stated in tables 2 and 3. the results of the simulations, shown in figure 6, indicate that the manufacturing part cost at the new site in china will exceed the one at the improved swedish assembly line after approximately six years. the above simulations include only the process part cost and not the costs connected to the manufacturing support functions. including these costs might considerably change the outcome. when comparing only the process costs between china and sweden, the relocation to china appears to be preferable, but if the risks of overestimated performance and underestimated wage growth are included, it is not obviously the best location. figure 6. quality rate in the assembly line in china is estimated to be 80% and the wage growth 10%. in addition to the quantitative risk, illustrated by the simulations above, qualitative factors, such as access to skilled personnel, infrastructure reliability, and quality and delivery reliability of incoming goods, could have a substantial impact on the success of production relocation. 5. results and discussion decisions on relocation and outsourcing require a clear understanding of the driving forces for changing the manufacturing footprint. for example, previous studies (kinkel et al., 2007; aspelund and butsko, 2010; windmark and andersson, 2012) have demonstrated that costs and strategic factors such as distance from key customers and markets are highly important for companies considering changes in the manufacturing footprint. even if relocation decisions are mainly based on strategic motives, an analysis of costs and benefits should always (and is usually) performed and considered in the decision process. the motivation for the research presented here are both the gap found in the literature on cost analysis int. j. prod. manag. eng. (2016) 4(1), 15-27 creative commons attribution-noncommercial-noderivatives 4.0 international windmark, c. and andersson, c. 24 http://creativecommons.org/licenses/by-nc-nd/4.0/ methodology in production relocation and the industrial practitioners view of lacking a structure to develop decision support in production relocation issues. the development of a work procedure to facilitate the development of a structured decision support for production relocation, were performed in close collaboration with five industrial partners. the research methods used during the development work were interviews, case studies, workshops and group validation. the research limitation is that the development is based on a limited set of empirical data. a broader constellation of case study companies would have the opportunity to capture a broader set of parameters influencing location. a research limitation is also the selected scope of operational focus. this will incur a risk of limiting the consideration of e.g. sustainability factors. in the cost tools it is possible to simulate currency fluctuation, but this have not been included in the scenario analysis presented here. the involved companies differed in size and in experience regarding location projects, which contributed to different needs for support in the location process. this group of industrial partners included both companies with international operations and those planning to establish such operations; this motivated the development of a wide-ranging modularised decision support, enabling users to select parts of the procedure or tools to complement their already established work procedures. this paper presents a methodology for estimating costs in the process of developing a comprehensive support for production location decisions. the principle for cost calculations is based on a cost model presented by ståhl et al. (2007), integrating technical performance parameters with financial parameters. a feature of this cost model is the inclusion of equipment performance (overall equipment efficiency), parameters that substantially affect costs (see figure 4). the structure of the cost model (see figure 3) incurs the possibility of scenario simulation of different production setups, making the model suitable for analysis of different alternative in a relocation decision process. to capture the total costs of relocation, an array of parameters needs to be considered. figure 4 show the set of parameters needed to estimate the part cost for operations including equipment, facility and employees. the additional costs concerning support functions such as quality assurance, market etc. were identified together with the industrial partners, see table 1. tools for estimating support costs were developed. models for estimating inbound logistics costs are presented in windmark and andersson (2014, 2015). the use of the cost analysis methodology is demonstrated by the scenario simulations presented in section 4.2. these analyses show (see figure 5) that production costs are influenced by performance (downtime and quality rates), indicating that theses parameters should not be neglected when analysing location alternatives. figure 5 also shows that wage cost fluctuations can change the costs and benefits completely. this indicates that analyses of future potential changes in wage costs, currency and market stability should be made prior to a location decision. the success of using the presented methodology is highly dependent on the availability of data for the parameters included in the cost models. the models are fairly comprehensive and the work needed to gather the necessary input data could be time consuming. however, in view of the huge investments needed to establish a new production location, we argue that the workload is justified to be able to make decisions on a thorough analysis of costs and benefits. as companies constantly increase the quality and availability of data and the data is digitally available to a greater extent, the effort needed for developing comprehensive decision supports, is likely to decrease. a challenge is however to retrieve data to estimate costs for a new location. these efforts can however be supported by experiences from an existing production system. the methodology presented here only includes quantitative parameters, and would require additional concern of strategic non-quantifiable parameters, e.g. legal, cultural, social, political, and economic factors prior to a decision. in addition, proximity to suppliers, markets/customers, parent company facilities, and competitors must be taken into consideration (maccarthy and atthirawong, 2003). other cost drivers important in production relocation, not included in the cost analyses here, are costs of knowledge provision, process testing and ramp-up at a new site. these costs could be viewed as the costs generated before the production starts and can be treated as the investment costs and be included in the process costs. it is also important to analyse the cost int. j. prod. manag. eng. (2016) 4(1), 15-27creative commons attribution-noncommercial-noderivatives 4.0 international cost modelling as decision support when locating manufacturing facilities 25 http://creativecommons.org/licenses/by-nc-nd/4.0/ references andersson, c., bellgran, m., bruch, j., rösiö, c., wiktorsson, m., windmark, c. (2013) production location handbook – forming your strategic manufacturing footprint, lund, sweden: media-tryck. aspelund, a., butsko, v. (2010), small and middle-sized enterprises’ offshoring production: a study of firm decisions and consequences, tijdschrift voor economische en sociale geografie, 101(3): 262-275. doi:10.1111/j.1467-9663.2009.00585.x bell, j., mcnaughton, r., young, s., crick, d. (2003). towards an integrative model of small firm internationalisation. journal of international entrepreneurship, 1(4): 339-362. doi:10.1023/a:1025629424041 bellgran, m., burch,j., rösiö, c., wiktorsson, m. (2013). decision support for production localization: process activities and location factors, proceedings of the 20th euroma conference, 7th-12th june, 2013, dublin. bjelkemyr, m., wiktorsson, m., rösiö, c., bruch, j., bellgran, m. (2013). production localization factors: an industrial and literature based review. in proceedings of the 11th international conference on manufacturing research (icmr2013) 489-494. boloori arabani, a., farahani, r. z. (2012). facility location dynamics: an overview of classifications and applications. computers & industrial engineering, 62(1): 408-420. doi:10.1016/j.cie.2011.09.018 boston consulting group. (2011). made in america, again. report boston consulting group inc. brouwer, a. e., mariotti, i., van ommeren, j. n. (2004). the firm relocation decision: an empirical investigation. the annals of regional science, 38(2): 335-347. doi:10.1007/s00168-004-0198-5 brynjolfsson, e., hitt, l. m. (2000). beyond computation: information technology, organizational transformation and business performance. the journal of economic perspectives, 14(4): 23-48. doi:10.1257/jep.14.4.23 burns, l. d., hall, r. w., blumenfeld, d. e., daganzo, c. f. (1985). distribution strategies that minimize transportation and inventory costs. operations research, 33(3), 469-490. doi:10.1287/opre.33.3.469 christodoulou p., fleet´s, d., hanson, p., phaal, r., probert, d., shi, y. (2007). manufacturing the right things in the right places – a structured approach to developing and exploiting manufacturing footprint strategy, university of cambridge institute for manufacturing. cooper, r., kaplan, r. s. (1998). cost and effect. harvard business school press, boston. dogan, i. (2012). analysis of facility location model using bayesian networks. expert systems with applications, 39(1): 1092-1104. doi:10.1016/j.eswa.2011.07.109 ellram, l. m., tate, w. l., petersen, k. j. (2013). offshoring and reshoring: an update on the manufacturing location decision. journal of supply chain management, 49(2): 14-22. doi:10.1111/jscm.12019 fang, d., weng, w. (2010). kpi evaluation system of location decision for plant relocation from the view of the entire supply chain optimization. in automation and logistics (ical), 2010 ieee international conference on automation and logistics, hong kong and macau, pp. 659663. doi:10.1109/ical.2010.5585366 yang, j., lee, h. (1997). an ahp decision model for facility location selection. facilities, 15(9/10): 241-254. doi:10.1108/02632779710178785 kinkel, s. (2009). erfolgskritische standortfaktoren ableiten–eine erfahrungsbasierte auswahlhilfe. in erfolgsfaktor standortplanung, 57-80. springer berlin heidelberg. doi:10.1007/978-3-540-88471-2_4 kinkel, s., lay, g., maloca, s. (2007). development, motives and employment effects of manufacturing offshoring of german smes. international journal of entrepreneurship and small business, 4(3): 256-276. doi:10.1504/ijesb.2007.013251 latino, r. j., latino, k. c., latino, m. a. (2013). root cause analysis: improving performance for bottom-line results. crc press. maccarthy b.l., atthirawong, w. (2003). factors affecting location decisions in international operations – a delphi study, international journal of operations & production management, 23(7): 794-818. doi:10.1108/01443570310481568 nordigården, d. (2007). outsourcing in the wood product manufacturing sector a combined customer and supplier perspective. phd thesis, linköping university. platts, k.w., song, n. (2010). overseas sourcing decisions: the total cost of sourcing from china, supply chain management: an international journal, 15(4): 320–331. doi:10.1108/13598541011054689 impact on a current site if the location project results in production relocation. further development of the cost analysis method will include these cost drivers. the aim for the future is also to test the work methodology and tools in ongoing transfer projects to further understand and grasp issues and difficulties in managing global production systems. acknowledgements the authors want to acknowledge vinnova for financially supporting the research. special thanks are also sent to the employees at the case study companies for their valuable information and support and to dr volodymyr bushlya for valuable comments on this work. int. j. prod. manag. eng. (2016) 4(1), 15-27 creative commons attribution-noncommercial-noderivatives 4.0 international windmark, c. and andersson, c. 26 http://dx.doi.org/10.1111/j.1467-9663.2009.00585.x http://dx.doi.org/10.1023/a:1025629424041 http://dx.doi.org/10.1016/j.cie.2011.09.018 http://dx.doi.org/10.1007/s00168-004-0198-5 http://dx.doi.org/10.1257/jep.14.4.23 http://dx.doi.org/10.1287/opre.33.3.469 http://dx.doi.org/10.1016/j.eswa.2011.07.109 http://dx.doi.org/10.1111/jscm.12019 http://dx.doi.org/10.1109/ical.2010.5585366 http://dx.doi.org/10.1108/02632779710178785 http://dx.doi.org/10.1007/978-3-540-88471-2_4 http://dx.doi.org/10.1504/ijesb.2007.013251 http://dx.doi.org/10.1108/01443570310481568 http://dx.doi.org/10.1108/13598541011054689 http://creativecommons.org/licenses/by-nc-nd/4.0/ rusten, g., bryson, j. r. (2010). placing and spacing services: towards a balanced economic geography of firms, clusters, social networks, contracts and the geographies of enterprise. tijdschrift voor economische en sociale geografie, 101(3): 248-261. doi:10.1111/j.14679663.2009.00584.x simons, r., dávila, a., kaplan, r. s. (2000). performance measurement & control systems for implementing strategy. upper saddle river, nj: prentice hall. stål c., andersson, a., gabrielson, p., ståhl, j.-e. (2012). a production performance analysis regarding downtimes and downtime pattern, 22nd international conference on flexible automation and intelligent manufacturing, faim 2012, 10-13 june, helsinki, finland. ståhl, j.-e., andersson, c., jönsson, m. (2007). a basic economic model for judging production development, paper presended at 1st swedish production symposium, 28–30 august. gothenburg, sweden. whitten, d., leidner, d. (2006). bringing it back: an analysis of the decision to backsource or switch vendors. decision sciences, 37(4): 605-621. doi:10.1111/j.1540-5414.2006.00140.x windmark, c., andersson, c. (2012). business case as a decision support when relocating manufacturing, presented at 5th swedish production symposium 2012, 6th-8th november, linköping windmark, c., andersson, c. (2014). a business case tool as decision support in early production location project stages. in the 6th international swedish production symposium 2014. windmark, c., andersson, c. (2015). cost models of inbound logistics activities: supporting production system design. international journal of supply chain and operations resilience, 1(2): 181-200. doi:10.1504/ijscor.2015.069927 int. j. prod. manag. eng. (2016) 4(1), 15-27creative commons attribution-noncommercial-noderivatives 4.0 international cost modelling as decision support when locating manufacturing facilities 27 http://dx.doi.org/10.1111/j.1467-9663.2009.00584.x http://dx.doi.org/10.1111/j.1467-9663.2009.00584.x http://dx.doi.org/10.1111/j.1540-5414.2006.00140.x http://dx.doi.org/10.1504/ijscor.2015.069927 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2017.6618 received 2016-09-20 accepted: 2016-10-03 a hybrid algorithm for flexible job-shop scheduling problem with setup times ameni azzouz*, meriem ennigrou and lamjed ben said lab. soie. stratégies d’optimisation et informatique intelligente, isg, institut supérieur de gestion, université de tunis, tunisie. ameni.azzouz@isg.rnu.tn abstract: job-shop scheduling problem is one of the most important fields in manufacturing optimization where a set of n jobs must be processed on a set of m specified machines. each job consists of a specific set of operations, which have to be processed according to a given order. the flexible job shop problem (fjsp) is a generalization of the above-mentioned problem, where each operation can be processed by a set of resources and has a processing time depending on the resource used. the fjsp problems cover two difficulties, namely, machine assignment problem and operation sequencing problem. this paper addresses the flexible job-shop scheduling problem with sequence-dependent setup times to minimize two kinds of objectives function: makespan and bi-criteria objective function. for that, we propose a hybrid algorithm based on genetic algorithm (ga) and variable neighbourhood search (vns) to solve this problem. to evaluate the performance of our algorithm, we compare our results with other methods existing in literature. all the results show the superiority of our algorithm against the available ones in terms of solution quality. key words: job-shop scheduling problem, flexible manufacturing systems, sequence-dependent setup times, genetic algorithms, local search. 1. introduction flexible-job-shop scheduling problem (fjsp) is a well-known np-hard problem (garey et al., 1976), which reflect a wide range of scheduling problems encountered in real manufacturing systems. for this reason, fjsp continues to attract the interests of researchers both in academia and industry. this problem mainly cover two difficulties: the first one is resource assignment problem where each operation can be processed by more than one resource from a set of available resource and has, consequently, a processing time depending on the resource used. the second one is operation sequencing problem with maintaining the feasibility conditions. recently, many researches have been made to find the near optimal solution of fjsp using a varied range of tools and techniques such as branch and bound (fatahi et al., 2007; zribi et al., 2007) and heuristics (wang and yu, 2010; ziaee, 2014). fjsp is known to be strongly np-hard. consequently, most of the literature related to the fjsp is based on metaheuristic methods like genetic algorithms (gas) (zhou et al., 2006; pezzella et al., 2008; zhang et al., 2011; zambrano rey et al., 2014), particle swarm optimization (pso) (zhang et al., 2009; nouiri et al., 2015) simulated annealing (sa) (najid et al., 2002; yazdani et al., 2009), tabu search (ts) (brandimarte, 1993; fatahi et al., 2007; vilcot and billaut, 2011) and beam search (bs) (wang et al., 2008). most job-shop scheduling researches reported in the literature ignore the setup times or consider them as a part of the processing time. however, in many real-life situations such as chemical, printing, pharmaceutical and automobile manufacturing (kim and bobrowski, 1994), the setup times are int. j. prod. manag. eng. (2017) 5(1), 23-30creative commons attribution-noncommercial-noderivatives 4.0 international 23 https://doi.org/10.4995/ijpme.2017.6618 http://creativecommons.org/licenses/by-nc-nd/4.0/ not only often required between jobs but they are also strongly dependent on job itself (sequence independent) and the previous job that ran on the same machine (sequence dependent). hence, reducing setup times is an important task to improve shop performance. the fjsp has been widely studied. however, few papers have considered this problem with setup times. among these, cheung and zhou (2001) propose a hybrid algorithm based on genetic algorithm and heuristic rules to solve sdst-jsp with minimizing the makespan. for the same problem, zhou et al. (2006) propose an immune algorithm which certifies the diversity of the antibody. moghaddas and houshmand (2008) develop a mathematical and heuristic model based on priority rules. naderi et al. (2009) consider the job shop scheduling with sequence-dependent setup times and preventive maintenance policies using four meta-heuristics based on simulated annealing and genetic algorithms. considering the flexibility constraints, flexible job-shop problem presents additional difficulty than the classical jsp and requires more effective algorithms. in recent decades, many attempts have been made to find the near optimal solution of sdstfjsp using a varied range of tools and techniques. imanipour (2006) was the first one who investigates the sdst-fjsp. the author modeled the problem as a non linear mixed integer programming model and proposes a tabu search for the same problem. saidimehrabad and fattahi (2007) presented a tabu search for solving the sdst-fjsp to minimize makespan. they assumed in their research that each operation can be performed by two machine alternatives. they compared their obtained results with the results of the lingo software. bagheri and zandieh (2011) propose a variable neighborhood search (vns) based on integrated approach to minimize an aggregate objective function (aof) where aof = αf1+(1-α)f2 and α denote the weight given respectively to makespan (f1) and mean tardiness (f2). to evaluate this model, the authors generate randomly 20 problem instances under four different classes. using the same aof, sadrzadeh (2013) present an artificial immune system algorithm (ais) and a particle swarm optimization algorithm (pso) and prove that both algorithms works better than vns of bagheri and zandieh (2011). mousakhani (2013) formulate the sdst-fjsp as a mixed integer linear programming model to minimize total tardiness and present a meta-heuristic based on iterated local search for the same problem. oddi et al. (2011) considers the sdst-fjsp to minimize the makespan using the iterative flattering search (ifs) and propose a new benchmark which is denoted sdst-hudata. it consists of 20 instances produced as an extension of the existing wellknown benchmarks of fjsp of hurink et al. (1994). gonzàlez et al. (2013) develop memetic algorithm to minimize the makespan which the tabu search was applied to every chromosome generated by the genetic algorithm. in order to evaluate their model, they used the same benchmark as in oddi et al. (2011) and prove that the memetic algorithm has obtained a better result than the ifs. recently, rossi (2014) investigate the sdst-fjsp with transportation times using ant-colony algorithm with reinforced pheromone. the most recent comprehensive survey of scheduling problem with setup times is given by allahverdi (2015). nevertheless, most of the above-mentioned research considered only one method optimization to solve sdst-fjsp. however, the literature reviews show that none of these methods are sufficient on their own to solve this np-hard problem. for that, in this paper, we propose a hybrid genetic algorithm (hga) based on ga and vns for the sdst-fjsp. then, we show that our algorithm can be very effective with respect to the state of the art. the remainder of this paper is organized as follows. in section 2, we formulate the problem and we give an illustrative example. section 3 presents the proposed algorithm to solve the sdst-fjsp. section 4 describes the performance of our algorithm on a set of benchmark problems and explains the most interesting results. conclusions and some future works are presented in section 5. 2. problem definition the sdst-fjsp can be defined as follows: this problem consists in performing n jobs on m machines. the set machines is noted m, m={m1,... ,mk}. each job i consists of a sequence of ni operations (routing). each routing has to be performed to complete a job. the execution of each operation j of a job i (noted oij) requires one machine out of a set of given machines mi,j (i.e. mi,j is the set of machines available to execute oij). the problem is to define a sequence of operations together with assignment of start times and machines for each operation. int. j. prod. manag. eng. (2017) 5(1), 23-30 creative commons attribution-noncommercial-noderivatives 4.0 international azzouz, a., ennigrou, m. and ben said, l. 24 http://creativecommons.org/licenses/by-nc-nd/4.0/ assumptions considered in this paper are the following: (1) jobs are independent of each other; (2) machines are independent of each other; (3) one machine can process at most one operation at a time; (4) no preemption is allowed; (5) all jobs are available at time zero; (6) setup times are dependent on the sequence of jobs. when one of the operations of a job t is processed before one of those of job i (t≠i) on machine mk, the sequence dependent setup time is st,i,k>0. the current sdst-fjsp based on these assumptions is aimed to minimize two kinds of objective functions: minimize the makespan ( i.e. the time required to complete all jobs) minimize aggregate objective function (aof) where aof = αf1+(1-α)f2 and α denote the weight given respectively to makespan (f1) and mean tardiness (f2). fjsp is classified as total fjsp and partial fjsp (kacem et al., 2002). in total fjsp (t-fjsp), each operation can be processes by all machines. however, in partial fjsp (p-fjsp), at least one operation may not be processed on all machines. several researches pointed out that the p-fjsp is more complex as compared to t-fjsp on the same scale. in this paper, we consider the p-fjsp. to illustrate this problem, we consider an instance with three jobs and three machines. in table 1, we show the processing time of each operation. the symbol “-” means that the machine can not execute the corresponding operation. table 2 presents the sequence dependent setup times of each job. for instance, the setup time for job3 after job2 on machine m1 is s2,3,1=2 and the setup time for job2 after job1 on machine m3 is s1,2,3=2. table 1. processing times. job operation m1 m2 m3 j1 o11 4 5 o12 3 4 o13 6 5 j2 o21 3 4 o22 4 5 o23 4 7 j3 o31 5 3 o32 4 o33 4 5 3 table 2. sequence-dependent setup times. job machine1 machine2 machine3 job1 job2 job3 job1 job2 job3 job1 job2 job3 dummy 3 2 1 2 4 1 4 3 2 job1 0 1 3 0 2 4 0 2 3 job2 1 0 2 4 0 3 2 0 4 job3 1 3 0 2 1 0 3 2 0 dummy job signifies the starting of a job on each machine. when one of the operations of a jobi is the first operation executed on machine mk, the dummy job is di,k. for example, if o3,1 is the first operation executed on machine m2, then, dummy job value would be d3,1=1. in order to explain better this problem, we represent a gantt chart of solution on figure 1.     figure 1. gantt chart of solution figure 1. gantt chart of solution. int. j. prod. manag. eng. (2017) 5(1), 23-30creative commons attribution-noncommercial-noderivatives 4.0 international a hybrid algorithm for flexible job-shop scheduling problem with setup times 25 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3. hybrid genetic algorithm for sdst-fjsp since the discovery of the genetic algorithms by holland (1975), they have been recognized as a powerful methods for solving combinatorial optimization problems such as scheduling problems. ga can find problem solutions by imitating natural selection mechanisms, using choosing, crossing and mutation operations. in this section, we present the step by step algorithm included our adaptation of the different hga parameters to our problem. 3.1. hga procedures step1 (initialization scheme): generate an initial parent population of n solutions. here, we propose an improved function of initial population which is based on three traditional dispatching rules as following: 20% using shortest processing time (spt); 20% using longest processing time (lpt); 20% using heuristic rules based on local search algorithm; the remaining with random solution. we coded the solution as binary matrix where the rows represent furthermore, the order in which they appear in the matrix describes the sequence of operations present in the solution. the columns correspond to the used machines as described in the figure 2. step2 (selection): choose (n/2) members from the parent population using tournament selection. step3 (reproduction phase): regarding to a certain probability, we perform crossover, mutation or vns algorithm in order to create the offspring population. for that, we use the crossover operator order 1 which consists on selecting randomly two positions xp1 and xp2 in parent1 (which is coded by a binary matrix). subsequently, the middle part is copied to the offspring1. the rest of this child is filled from the parent2 starting with position xp2+1, and jumping elements that are already presented in offspring1. the same steps are repeated for the second offspring by starting with the parents2. figure 2 shows an example of this procedure. till now to the mutation operator, we use the mutation technique proposed by pezzella et al. (2008), in which, we select one operation with the maximum workload (i.e. the amount of work that a machine produces in a specified time period). then we assign it to the machine with the minimum workload if possible. the vns algorithm is used as a reproduction operator. in the next section, we will focus our attention on the specific vns we have used. step4 (environmental selection): fill the new population using a replacement strategy. the new population is formed with the n best solutions. if the stopping criterion is met then return the best solution; otherwise, return to step 2. figure 2. crossover operator. 3.2. vns algorithm traditionally, the hybridization between ga and local search algorithm is based on simple local search. however, we adapted an advanced local search called variable neighborhood search (vns) inspired from bagheri and zandieh (2011) and ennigrou and ghedira (2008) as a local search for our algorithm. more in detail, this vns based on three kinds of neighborhood structures presented as following: step1 (initial solution): we conserve the same solution representation as in ga. step2 (generate neighborhood): in this step, we determine the neighborhood of the current solution. then, we use three types of moves: two positions p1 and p2 are randomly selected on the solution and then, the operations between them are randomly reordered. int. j. prod. manag. eng. (2017) 5(1), 23-30 creative commons attribution-noncommercial-noderivatives 4.0 international azzouz, a., ennigrou, m. and ben said, l. 26 http://creativecommons.org/licenses/by-nc-nd/4.0/ operation is chosen randomly and then, changing the assignment of the selected operation to another machine. the combination between the two precedent neighborhood structures step3 (neighborhood evaluation): after applying each of the neighborhood structures described above, we execute a local search for certain iteration. we use the same local search as in. step4 (final stage): after a number of iteration of vns algorithm, the best solution founded is selected to the next population. 4. experimental study this section evaluates the performance of our proposed hybrid algorithm for two kinds of objective functions: makespan and aggregate objective function (aof). for that, we compare our hga against the available algorithms in the literature including variable neighourbood search (vns) proposed by bagheri and zandieh (2011), an adapted tabu search (ts) proposed by ennigrou and ghedira (2008), artificial immune system (ais), particle swarm optimization (pso) from sadrzadeh (2013) and our ga. our proposed algorithm has been implemented using java and run on pc with core2duo, 2,6ghz and 2gb ram. in our experiment, we tested different values for our hga parameters, and computational experience proves that the following values are more effective for the two problems (i.e. sdst-fjsp with or without learning effects: population size: 150 crossover probability: 0.6 mutation probability: 0.2 local search probability: 0.2 number of iteration of hga (stopping condition): 150 or cpu time limit fixed to n×ni×m×0.1s number of no improvement (stopping condition):20 number of iteration of vns:30. for the makespan objective function, we consider the same benchmark as in oddi a. et al. (2011) and gonzález, m. et al. (2013) which is denoted sdsthudata. it consists of 20 instances derived from the first 20 instances of the fjsp benchmark proposed in hurink et al. (1994). each instance was created by adding to the original instance one setup time matrix st,k for each machine k. the same setup time matrix was added for each machine in all benchmark instances. each matrix has size n×n, and the value st,i,k. indicates the setup time needed to reconfigure the machine k when switches from job t to job i. these setup times are sequence dependent and they fulfill the triangle inequality. the non-deterministic nature of our algorithm makes it necessary to carry out multiple runs on the same instance in order to obtain meaningful results. after ten runs of each generated instance by the abovementioned algorithms, the best solutions obtained for each instance (which is named solmin) are calculated. we use the relative percentage deviation (rpd) measure to compare the performance of algorithms. rpd is obtained as follows: rpd = solalgo−solmin/solmin×100 where solalgo is the makespan of each algorithm. table 3 show the performance of the proposed hga compared with others algorithms. the instance names are listed in the first column, the second column show the size (n×m) of each instance. the third column table 3. summary of results in the sdst-fjsp to minimize the makespan: sdst-hudata benchmark. instance problem size n×m vns ts ga hga la01 10×5 2.54 7.36 0.00 0.00 la02 1.58 8.53 2.57 0.33 la03 2.66 5.33 0.13 0.54 la04 3.50 3.66 4.07 0.90 la05 1.11 3.58 0.16 0.24 la06 15×5 4.45 2.79 2.04 0.32 la07 2.92 2.99 1.97 0.34 la08 2.37 2.41 2.86 0.65 la09 3.82 3.21 0.41 0.33 la10 1.84 4.02 0.00 0.62 la11 20×5 3.32 2.08 1.61 0.34 la12 5.33 2.71 2.38 0.60 la13 3.30 3.14 3.02 1.02 la14 1.70 1.76 1.26 0.93 la15 2.37 1.18 1.97 1.95 la16 10×10 2.37 4.04 0.00 0.76 la17 4.34 2.39 0.07 0.14 la18 1.86 5.81 0.91 0.35 la19 6.37 6.67 0.11 0.17 la20 4.84 9.36 2.48 0.35 average 3.12 4.15 1.40 0.54 int. j. prod. manag. eng. (2017) 5(1), 23-30creative commons attribution-noncommercial-noderivatives 4.0 international a hybrid algorithm for flexible job-shop scheduling problem with setup times 27 http://creativecommons.org/licenses/by-nc-nd/4.0/ represents the flexibility (i.e. the average number of alternative machines for each operation). the third, fourth, fifth and sixth columns report the obtained results of vns algorithm (bagheri and zandieh, 2011), ts algorithm (ennigrou and ghedira, 2008), our ga only and hga. the obtained results show that the proposed hga performs better than the others algorithms in 13 instances. only in instance la03, la05 la10, la16, la17 and la18 ga has gained better results. however, in instance la15, ts obtained the better results. the proposed hga outperforms the others algorithms with average rpd of 0.54 while the worst performing algorithm is ts with average rpd of 4.15. moreover, we notice that hga obtained the best average rpd of 0.35 in the largest number of machine against 4.58, 6.48 and 0.71 for vns, ts and ga respectively. to further evaluate the performance of our algorithm, we study the interaction between the performance of the algorithm and the problem size in figure 3. we remark that our algorithm keeps its robust performance in different problem sizes. furthermore, for the aof, we consider artificial benchmarks according to the function proposed by bagheri and zandieh (2011). we propose four classes of instances.these classes are different in number of jobs, n, number ofoperations for each job i, ni, and number of machines, m, that are denoted as (n×ni×m). the generated instances have partial flexibility and the number of available machines for each operation (amo) is generated randomly according to uniform distribution. table 4 summarizes the characteristics of the artificial benchmarks used in this paper. in table 4. characteristics of the artificial benchmarks n×ni×m amo processing time sdst dummy jobs class1 10×5×5 u(1,5) u(20,100) u(20,60) u(20,40) class2 15×5×8 u(1,8) class3 10×10×5 u(1,5) class4 15×10×10 u(1,10) figure 3. the average rpd of the algorithms versus the number of jobs. figure 4. the average rpd of the algorithms of each type of problem class for α=0.25. figure 5. the average rpd of the algorithms of each type of problem class for α=0.5 figure 6. the average rpd of the algorithms of each type of problem class for α=0.75. int. j. prod. manag. eng. (2017) 5(1), 23-30 creative commons attribution-noncommercial-noderivatives 4.0 international azzouz, a., ennigrou, m. and ben said, l. 28 http://creativecommons.org/licenses/by-nc-nd/4.0/ order to introduce due dates, we consider the same formula as in (bagheri & zandieh, 2011). overall, compared to vns, ais, pso and ga, our hga has a superiority result to minimize the aof for all α values. moreover, from the results shown in figures 4, 5 and 6, we remark that hga is more effective with α =0.25 then α = 0.75. otherwise, our algorithms have the best results with mean tardiness against makespan objective function. 5. conclusions in this paper, we focus on solving the flexible job shop scheduling problem where sequence dependent setup times are also taken into account. we have proposed a hybrid genetic algorithm to minimize two kinds of objective functions: makespan and aggregate objectives function. for that, we tested hga on two kinds of benchmark. results showed that the present hga is better than other algorithms. in future works, it will be interesting to investigate the dynamic scheduling problem to closely reflect the real flexible job shop scheduling environment. for the same reason, we will consider the multi-criteria scheduling problem and the scheduling problems with learning effects considerations. acknowledgements this he authors would like to say thanks to miguel a. gonzalez for providing us with the sdst-fjsp instances. references allahverdi, a. (2015). the third comprehensive survey on scheduling problems with setup times/costs. european journal of operational research, 246(2), 345-378. https://doi.org/10.1016/j.ejor.2015.04.004 azzouz, a., ennigrou, m., jlifi, b. (2015). diversifying ts using ga in multi-agent system for solving flexible job shop problem. in proceedings of the 12th international conference on informatics in control, automation and robotics, 94-101. https://doi. org/10.5220/0005511000940101 azzouz, a., ennigrou, m., jlifi, b., ghedira, k. (2012). combining tabu search and genetic algorithm in a multi-agent system for solving flexible job shop problem. in artificial intelligence (micai), 2012, 11th mexican international conference on, pp. 83-88. ieee. https:// doi.org/10.1109/micai.2012.12 bagheri, a., zandieh, m. (2011). bi-criteria flexible job-shop scheduling with sequence-dependent setup times variable neighborhood search approach. journal of manufacturing systems, 30(1), 8-15. https://doi.org/10.1016/j.jmsy.2011.02.004 brandimarte, p. (1993). routing and scheduling in a flexible job shop by tabu search. journal annals of operations research, 41(3), 157183. https://doi.org/10.1007/bf02023073 cheung, w., zhou, h. (2001). using genetic algorithms and heuristics for job shop scheduling with sequence-dependent setup times. annals of operations research, 107(1), 65-81. https://doi.org/10.1023/a:1014990729837 fattahi, p., saidi-meradbad, m., jolai, f. (2007). mathematical modeling and heuristic approaches to flexible job shop scheduling problems. journal of intelligent manufacturing, 18(3), 331-342. https://doi.org/10.1007/s10845-007-0026-8 gonzález, m. a., rodriguez vela, c., varela, r. (2013). an efficient memetic algorithm for the flexible job shop with setup times. in twentythird international conference on automated, pp. 91-99. hurink, j., jurisch, b., thole, m. (1994). tabu search for the job-shop scheduling problem with multi-purpose machines. operationsresearch-spektrum, 15(4), 205-215. https://doi.org/10.1007/bf01719451 imanipour, n. (2006). modeling solving flexible job shop problem with sequence dependent setup times. international conference on service systems and service management. ieee, 2, 1205-1210. https://doi.org/10.1109/icsssm.2006.320680 kim, s.c., bobrowski, p.m. (1994). impact of sequence-dependent setup time on job shop scheduling performance. the international journal of production research, 32(7), 1503-1520. https://doi.org/10.1080/00207549408957019 moghaddas, r., houshmand, m. (2008). job-shop scheduling problem with sequence dependent setup times. proceedings of the international multiconference of engineers and computer scientists, 2, 978-988. mousakhani, m. (2013). sequence-dependent setup time flexible job shop scheduling problem to minimise total tardiness. international journal of production research, 51(12), 3476-3487. https://doi.org/10.1080/00207543.2012.746480 naderi, b., zandieh, m., ghomi, s.f. (2009). scheduling sequence-dependent setup time job shops with preventive maintenance. the international journal of advanced manufacturing technology, 43, 170-181. https://doi.org/10.1007/s00170-008-1693-0 int. j. prod. manag. eng. (2017) 5(1), 23-30creative commons attribution-noncommercial-noderivatives 4.0 international a hybrid algorithm for flexible job-shop scheduling problem with setup times 29 https://doi.org/10.1016/j.ejor.2015.04.004 https://doi.org/10.5220/0005511000940101 https://doi.org/10.5220/0005511000940101 https://doi.org/10.1109/micai.2012.12 https://doi.org/10.1109/micai.2012.12 https://doi.org/10.1016/j.jmsy.2011.02.004 https://doi.org/10.1007/bf02023073 https://doi.org/10.1023/a:1014990729837 https://doi.org/10.1007/s10845-007-0026-8 https://doi.org/10.1007/bf01719451 https://doi.org/10.1109/icsssm.2006.320680 https://doi.org/10.1080/00207549408957019 https://doi.org/10.1080/00207543.2012.746480 https://doi.org/10.1007/s00170-008-1693-0 http://creativecommons.org/licenses/by-nc-nd/4.0/ najid, n.m., dauzere-peres, s., zaidat, a. (2002). a modified simulated annealing method for flexible job shop scheduling problem. in proceedings of the ieee international conference on systems, man and cybernetics, 5, 6-9. https://doi.org/10.1109/ icsmc.2002.1176334 nouiri, m., bekrar, a., jemai, a., niar, s., ammari, a. c. (2015). an effective and distributed particle swarm optimization algorithm for flexible job-shop scheduling problem. journal of intelligent manufacturing, 1-13. https://doi.org/10.1007/s10845-015-1039-3 oddi, a., rasconi, r., cesta, a., & smith, s. (2011). applying iterative flattening search to the job shop scheduling problem with alternative resources and sequence dependent setup times. in coplas 2011 proceedings of the workshopon constraint satisfaction techniques for planning and scheduling problems, pp. 15-22. pezzella, f., morganti, g., ciaschetti, g. (2008). a genetic algorithm for the flexible job-shop scheduling problem. computers & operations research, 35(10), 3202-3212. https://doi.org/10.1016/j.cor.2007.02.014 rossi, a. (2014). flexible job shop scheduling with sequence-dependent setup and transportation times by ant colony with reinforced pheromone relationships. international journal of production economics, 153, 253-267. https://doi.org/10.1016/j.ijpe.2014.03.006 sadrzadeh, a. (2013). development of both the ais and pso for solving the flexible job shop scheduling problem. arabian journal for science and engineering, 38(12), 3593-3604. https://doi.org/10.1007/s13369-013-0625-y saidi-mehrabad, m., fattahi, p. (2007). flexible job shop scheduling with tabu search algorithms. the international journal of advanced manufacturing technology, 32(5), 563-570. https://doi.org/10.1007/s00170-005-0375-4 vilcot, g., billaut, j.c. (2011). a tabu search algorithm for solving a multicriteria flexible job shop scheduling problem. international journal of production research, 49(23), 6963-6980. https://doi.org/10.1080/00207543.2010.526016 wang, s.j., zhou, b.h., xi, l.f. (2008). a filtered-beam-search-based algorithm for flexible job-shop scheduling problem. international journal of production research, 46(11), 3027-3058. https://doi.org/10.1080/00207540600988105 wang, s., yu, j. (2010). an effective heuristic for flexible job-shop scheduling problem with maintenance activities. computers and industrial engineering, 59(3), 436-447. https://doi.org/10.1016/j.cie.2010.05.016 yazdani, m., gholami, m., zandieh, m., mousakhani, m. (2009). a simulated annealing algorithm for flexible job-shop scheduling problem. journal of applied sciences, 9(4), 662-670. https://doi.org/10.3923/jas.2009.662.670 zambrano rey, g., bekrar, a., prabhu, v., trentesaux, d. (2014). coupling a genetic algorithm with the distributed arrival-time control for the jit dynamic scheduling of flexible job-shops. international journal of production research, 52(12), 3688-3709. https://doi.org/10 .1080/00207543.2014.881575 zhang, g., gao, l., shi, y. (2011). an effective genetic algorithm for the flexible job-shop scheduling problem. expert systems with applications, 38(4), 3563-3573. https://doi.org/10.1016/j.eswa.2010.08.145 zhang, g., shao, x., li, p., gao, l. (2009). an effective hybrid particle swarm optimization algorithm for multi-objective flexible job-shop scheduling problem. computers and industrial engineering, 56(4), 1309-1318. https://doi.org/10.1016/j.cie.2008.07.021 zhou, y., beizhi, l., yang, j. (2006). study on job shop scheduling with sequence-dependent setup times using biological immune algorithm. the international journal of advanced manufacturing technology, 30(1), 105-111. https://doi.org/10.1007/s00170-005-0022-0 ziaee, m. (2014). a heuristic algorithm for solving flexible job shop scheduling problem. the international journal of advanced manufacturing technology, 71(1), 519-528 https://doi.org/10.1007/s00170-013-5510-z zribi, n., kacem, i., el kamel, a., borne, p. (2007). assignment and scheduling in flexible job-shops by hierarchical optimization. in systems, man, and cybernetics, part c: applications and reviews, ieee (4), 652-661. https://doi.org/10.1109/tsmcc.2007.897494 int. j. prod. manag. eng. (2017) 5(1), 23-30 creative commons attribution-noncommercial-noderivatives 4.0 international azzouz, a., ennigrou, m. and ben said, l. 30 https://doi.org/10.1109/icsmc.2002.1176334 https://doi.org/10.1109/icsmc.2002.1176334 https://doi.org/10.1007/s10845-015-1039-3 https://doi.org/10.1016/j.cor.2007.02.014 https://doi.org/10.1016/j.ijpe.2014.03.006 https://doi.org/10.1007/s13369-013-0625-y https://doi.org/10.1007/s00170-005-0375-4 https://doi.org/10.1080/00207543.2010.526016 https://doi.org/10.1080/00207540600988105 https://doi.org/10.1016/j.cie.2010.05.016 https://doi.org/10.3923/jas.2009.662.670 https://doi.org/10.1080/00207543.2014.881575 https://doi.org/10.1080/00207543.2014.881575 https://doi.org/10.1016/j.eswa.2010.08.145 https://doi.org/10.1016/j.cie.2008.07.021 https://doi.org/10.1007/s00170-005-0022-0 https://doi.org/10.1007/s00170-013-5510-z https://doi.org/10.1109/tsmcc.2007.897494 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2016.5209 received 2016-04-06 accepted: 2016-06-13 a mathematical programming model for tactical planning with set-up continuity in a two-stage ceramic firm david pérez peralesa1, m. m. eva alemanya2 a research centre on production management and engineering. universitat politècnica de valència. a1 dapepe@omp.upv.es a2 mareva@cigip.upv.es . abstract: it is known that capacity issues in tactical production plans in a hierarchical context are relevant since its inaccurate determination may lead to unrealistic or simply non-feasible plans at the operational level. semi-continuous industrial processes, such as ceramic ones, often imply large setups and their consideration is crucial for accurate capacity estimation. however, in most of production planning models developed in a hierarchical context at this tactical (aggregated) level, setup changes are not explicitly considered. their consideration includes not only decisions about lot sizing of production, but also allocation, known as capacitated lot sizing and loading problem (clslp). however, clslp does not account for set-up continuity, specially important in contexts with lengthy and costly set-ups and where product families minimum run length are similar to planning periods. in this work, a mixed integer linear programming (milp) model for a two stage ceramic firm which accounts for lot sizing and loading decisions including minimum lot-sizes and set-up continuity between two consecutive periods is proposed. set-up continuity inclusion is modelled just considering which product families are produced at the beginning and at the end of each period of time, and not the complete sequence. the model is solved over a simplified twostage real-case within a spanish ceramic firm. obtained results confirm its validity. key words: set-up continuity, ceramic firm, tactical planning, mixed integer linear programming. 1. introduction in the majority of the production planning models developed in a hierarchical context at the tactical level, the capacities at each stage are aggregated and setup changes are not explicitly considered. however, if at this level the setup times involve an important consumption capacity and have been completely ignored, this may lead to an overestimation of the real capacity availability which, in turn, may lead to unrealistic or unfeasible events during the subsequent disaggregation of tactical plans (pérez, 2013). considerable savings may be also be achieved through optimum lot-sizing decisions, known in the literature as capacitated lot sizing problem (clsp) problem. but standard clsp does not sequence products within a period and also assumes that setup cost occur for each lot in a period, even if the last product to be produced in a period is the first one in the period that follows. in addition to that, most of them focus on the operational (disaggregated) level. many works have addressed the standard clsp problem such as: barani et al., 1984; eppen and martin, 1987; chen and tizy, 1990; maes et al., 1991; chung et al., 1994; hindi, 1996; belvaux and wolsey, 2001. standard clsp may lead therefore to inaccurated capacity estimations at a tactical (aggregated) level, specially relevant in semicontinuous production environments with lengthy and costly set-ups and where minimum run lengths are similar to planning periods. in these contexts, setup continuity must be incorporated. these models are known as clsp with setup carryovers or simply clsp with linked int. j. prod. manag. eng. (2016) 4(2), 53-64creative commons attribution-noncommercial-noderivatives 4.0 international 53 http://dx.doi.org/10.4995/ijpme.2016.5209 mailto:dapepe@omp.upv.es mailto:mareva@cigip.upv.es http://creativecommons.org/licenses/by-nc-nd/4.0/ lot-sizes (haase, 1994). these models have not been as intensively studied as the standard clsp, mainly due to their model complexity and computational difficulty (sox and gao, 1999). just a few works have addressed the clsp with linked lot-sizes, all of them with constant sequence independent setup times and /or setups, with a setup carryover. no sequence is considered within a period. they just focus on determining the products produced last and first in two consecutive periods, and also the configuration of the machine at the end of the period. some examples may be found in kang et al., 1999; gopalakrishnan et al., 2001; porkka et al., 2003; suerie and stadtler, 2003. but accounting accurately for setup times at the tactical level would mean simultaneously including not only lot sizing decisions, but also allocation of production. this later problem is known as capacitated lot sizing and loading problem (clslp) (özdamar and birbil, 1998; özdamar and bozyel, 1998). although the above quoted works consider both allocation and lot sizing issues in a tactical planning level, there is a lack of tactical models in a hierarchical context that consider this clslp problem, so that capacities are aggregated and no product families allocation takes place, leading to inaccurate estimation of the real capacity availability that clearly affects to the operational level (mustafa et al., 1999; grieco et al., 2001). in addition to that, despite considering product families allocation and lot sizing issues, no setup continuity issues are included, specially in multistage systems. this is particularly important in industrial sectors with semicontinuous processes such as: ceramic (alemany et al., 2009, 2011) food (van donk, 2001; soman et al., 2004, 2007; romsdal et al., 2011; kopanos et al., 2012a, 2012b). textile (ishikura, 1994; de toni and meneghetti, 2000; guo et al., 2006; min and cheng, 2006; wong and leung, 2008; ngaia et al., 2014) chemical (meijboom and obel, 2007; ulstein et al., 2007; teimoury et al., 2010; fumero et al., 2012; shabani and sowlati, 2013; van elzzaker et al., 2014) all of them cope with very lengthy setup times in their manufacturing processes and at the same time their product families minimum run length are almost, equal or even higher than the planning period. many firms in these sectors only work with planning overviews based on spreadsheets. however, given the increasing complexity of product catalogues and current market pressure to reduce supply times, more rigorous methods are needed to optimise resources, as the one mathematical programming-based proposed in this work. furthermore, given the dramatic increase of end products, the possibilities for assigning and establishing lots on production lines multiply. therefore, the expected reduction of tactical production planning costs stands out as the proposed model establishes the product families to be produced on each line in an attempt to save changeovers as far as possible, this being an important objective, among others, in the aforementioned sectors. in this article, an approach to accurately model the capacity in tactical (aggregated) plans in a hierarchical context for a ceramic firm is proposed. for that, not only the clslp problem is considered, but also setup continuity issues and a two-stage system. some of this paper authors already approached this issue in pérez et al. (2014), but in a single stage one. the differences that result from this consideration justify this new scenario. this setup continuity is made over discrete periods of time, that is, it assumes that if a product family is manufactured two periods of time in the same production line just one set-up should be considered. besides, it accounts for minimum lot sizes even if the product family was produced in a production line in different periods. the set-up continuity consideration along with the minimum lot sizes requirement allows the model to produce the minimum lot-sizes over two consecutive periods being another contribution of the paper. within this model more efficient and realistic plans will be achieved at the tactical level, reducing later plan modifications due to internal aspects of the firm. the rest of the paper is arranged as follows. section 2 describes the problem being studied, as is the case of a spanish ceramic plant. in section 3, a deterministic milp model to solve the problem is presented. section 4 reports a numerical example to validate the model. section 5 offers some conclusions and future research lines, some of them already being undertaken. int. j. prod. manag. eng. (2016) 4(2), 53-64 creative commons attribution-noncommercial-noderivatives 4.0 international pérez perales, d., alemany, m.m.e. 54 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. problem description this case involves a ceramic spanish plant based in the province of castellón, dedicated to the manufacture of different types of tiles (floorings and coverings) since 1975. although this plant forms part of a broader industrial group (tiles sc) which is made up of different plants dedicated to the design, manufacture, marketing and distribution of finished goods, this work is single-company based, and the decisional problem to be addressed just focuses in mid term/tactical production planning issues. each production plant follows a make-to-stock strategy and it can be classified as a hybrid flow shop composed of several stages (presses-glazing lines, kilns and sorting-packing) uncoupled by buffers .each stage is integrated by similar machines and different finished goods can be processed by each machine at each stage. the main characteristics of each one of such stages are: 1. presses-glazing lines: is made up of one or several production lines in parallel with a limited capacity. production lines may process different product families. changeovers between product families incur setup costs owing to the time spent in changing, for example, moulds. a product family is defined as a group of finished goods of identical use (flooring or coverings), format (size), grout (white or red), and whose preparation on production lines is similar. this grouping into product families is crucial not only for commercial reasons but also to minimise setup times and costs. glazing lines may not be standardised, in that case, each product family can be processed according to specific facilities with the appropriate technical features. therefore, not all glazing lines are capable of processing all the product families, although a product family that may be processed on each line is known. technological factors involved in the production process mean that when a certain family is manufactured on a specific line, it should be produced in an equal or greater amount than the minimum lot size. this is partly because a certain percentage of defects occur during the production process, and only a percentage of the manufactured items may be sold as first quality finished goods. 2. kilns: represent the bottleneck section and imply a high energy consumption and cost. changeovers also occur in this section but are not as important as in the presses-glazing lines. 3. sorting-packing: this section always has excess capacity and does not represent any critical resource. at the tactical level, an aggregate plan (ap) for capacity-related decisions is defined for product families in the first two stages (sorting-packing is not taken into account). in this context not only is important the consideration of setup times but also its continuity over consecutive planning periods, because the set-up are lengthy and the minimum lot sizes of product families imply a run length (3 weeks) similar to the planning periods (1 month). these aspects are crucial to get accurate capacity availability estimation in the ap, which will constraint the master plan. 3. problem modeling a milp model has been developed to solve this ceramic tactical production planning problem. the objective is to minimize the total cost (set-up and inventory) over the time periods of the planning horizon. decisions will have to simultaneously deal with not only the allocation of product families to production lines and kilns with a limited capacity, but also with the determination of lot sizing and other decisions regard to set-up continuity modelling. for example those which allow to know the first and the last product family processed on each production line and kiln in a planning period, so that one changeover can be saved if the last one processed in t and the first one in t+1 are the same. or those which allow processing the minimum lot size between two consecutive periods with no changeover. all of them are later explained. the indexes, parameters, and decision variables are described in tables 1-3, respectively. table 1. indexes. f product families (f) (f=1…f) l production lines (l) (l=1…l) k kilns (k) (k=1….k) t periods of time (pt) (t=1…t) int. j. prod. manag. eng. (2016) 4(2), 53-64creative commons attribution-noncommercial-noderivatives 4.0 international a mathematical programming model for tactical planning with set-up continuity in a two-stage ceramic firm 55 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 2. parameters. dft demand of f f in pt t. cif inventory cost of a f f in a pt. ciif inventory cost (intermediate) of a f f in a pt. cslfl setup cost for f f on l l. cshfk setup cost for f f on k k. tflfl time to process a f f on l l. tfhfk time to process a f f on k k. tslfl setup time for f f on l l. tshfk setup time for f f on k k. lmlfl minimum lot size of f f on l l. lmhfk minimum lot size of f f on k k. capllt production capacity available (time) of l l during pt t. caphkt production capacity available (time) of k k during pt t. i0f inventory of f f at the start of the first pt. ii0f inventory (intermediate) of f f at the start of the first pt. m1,m2,m3,m4 very large integers. nf number of f βl0fl the l l is prepared to produce the f f at the start of the first pt. βh0fk the l l is prepared to produce the k k at the start of the first pt. table 3. decision variables. ift inventory of f f at the end of pt t. iift inventory (intermediate) of f f at the end of pt t. pflflt amount of f f produced on l l in pt t. pfhfkt amount of f f produced on k k in pt t. ylflt binary variable with a value of 1 if f f is produced on l l in pt t, and with a value of 0 otherwise. yhfkt binary variable with a value of 1 if f f is produced on k k in pt t, and with a value of 0 otherwise. xlflt binary variable with a value of 1 if l l is ready to produce the f f in pt t, and with a value of 0 otherwise. xhfkt binary variable with a value of 1 if k k is ready to produce the f f in pt t, and with a value of 0 otherwise. zlflt binary variable with a value of 1 if l l if a setup takes place of f f on l l in pt t, and with a value of 0 otherwise. zhfkt binary variable with a value of 1 if k k if a setup takes place of f f on k k in pt t, and with a value of 0 otherwise. wllt binary variable with a value of 1 if more than one f f is produced on l l in pt t, and with a value of 0 otherwise. whkt binary variable with a value of 1 if more than one f f is produced on k k in pt t, and with a value of 0 otherwise. αlflt binary variable with a value of 1 if l l is prepared to produce the f f at the start of pt t, and with a value of 0 otherwise. αhfkt binary variable with a value of 1 if k k is prepared to produce the f f at the start of pt t, and with a value of 0 otherwise. βlflt binary variable with a value of 1 if l l is prepared to produce the f f at the end of pt t, and with a value of 0 otherwise. βhfkt binary variable with a value of 1 if k k is prepared to produce the f f at the end of pt t, and with a value of 0 otherwise. int. j. prod. manag. eng. (2016) 4(2), 53-64 creative commons attribution-noncommercial-noderivatives 4.0 international pérez perales, d., alemany, m.m.e. 56 http://creativecommons.org/licenses/by-nc-nd/4.0/ objective function: c * c * c * c i * fl flt fh fht t l f t h f f ft f ft t f t f sl zl sh zh i i i ii + + + ∑ ∑ ∑ ∑ ∑ ∑ ∑ ∑ ∑ ∑ (1) subject to: 0ft f fkt ft k i i pfh d= + −∑ , ∀ f, t = 1 (2) 1ft ft fkt ft k i i pfh d−= + −∑ , ∀ f, t > 1 (3) * * fk fkt f fk fkt kt f tfh pfh tsh zh caph + ≤ ∑ ∑ , ∀ k, t (4) 1*fkt fktpfh m xh≤ , ∀ f, k, t (5) 2 *fkt fktpfh m yh≤ , ∀ f, k, t (6) 1 1m * ( )fk fkt fkt fkt fkt l h zh zh yh pfh + ++ − ≤ , ∀ f, k, t (7) 1 1 1 m * ( 2) fk fkt fkt fkt fkt fkt fkt l h zh zh yh yh pfh pfh + + + + + + − ≤ + , ∀ f, k, t (8) fkt fktyh pfh≤ , ∀ f, k, t (9) fkt fktyh xh≤ , ∀ f, k, t (10) fkt fktzh yh≤ , ∀ f, k, t (11) αhfkt–βh0fk ≤∑zhfkt, ∀ f, k, t = 1 (12) αhfkt–βhfkt–1 ≤∑ f zhfkt, ∀ f, k, t >1 (13) βhfkt–αhfkt ≤(∑ f xhfkt)–1, ∀ f, k, t (14) ∑ f αhfkt=1, ∀ k, t (15) ∑ f βhfkt=1, ∀ k, t (16) αhfkt ≤ xhfkt , ∀ f, k, t (17) βhfkt ≤ xhfkt , ∀ f, k, t (18) 3*xhfkt–∑ f xhfkt ≤ αhfkt +βhfkt, ∀ f, k, t (19) 2*xhfkt– αhfkt –βh0fk ≤ 2*zhfkt, ∀ f, k, t (20) 2*xhfkt– αhfkt –βhfkt–1 ≤ 2*zhfkt, ∀ f, k, t (21) ∑ f zhfkt ≤ nf *(3– αhfkt– βhfkt –βh0fk), ∀ f, k, t =1 (22) ∑ f zhfkt ≤ nf *(3– αhfkt– βhfkt –βhfkt–1), ∀ f, k, t >1 (23) 2–∑ f yhfkt ≤ 2*(1–whkt), ∀ k, t (24) ( ) 1 *fkt kt f yh nf wh− ≤∑ , ∀ k, t (25) αhfkt+βhfkt ≤ (2–whkt), ∀ f, k, t (26) iift=ii0f +∑ l pflflt – ∑ k pfhfkt , ∀ f, t = 1 (27) iift= iift–1 +∑ l pflflt – ∑ k pfhfkt , ∀ f, t = 1 (28) ∑ l tflfl *pflflt +∑ f tslfl *zlflt ≤ capllt ∀ l,t (29) pflflt ≤ m 3*xlflt, ∀ f, l, t (30) pflflt ≤ m 4*ylflt, ∀ f, l, t (31) lmlfl*(zlflt+zlflt+1–ylflt+1) ≤pflflt, ∀ l, f, t (32) lmlfl*(zlflt+zlflt+1+ylflt+ylflt+1–2)≤pflflt+pflflt+1, ∀ l, f, t (33) ylflt  ≤ pflflt , ∀ f, l, t (34) ylflt  ≤ xlflt , ∀ f, l, t (35) zlflt  ≤ ylflt , ∀ f, l, t (36) αlflt– βl0fl ≤ ∑ f zlflt , ∀ f, l, t = 1 (37) αlflt– βlflt–1 ≤ ∑ f zlflt , ∀ f, l, t >1 (38) βlflt– αlflt ≤ (∑ f xlflt )–1, ∀ f, l, t (39) ∑ f αlflt=1, ∀ l, t (40) int. j. prod. manag. eng. (2016) 4(2), 53-64creative commons attribution-noncommercial-noderivatives 4.0 international a mathematical programming model for tactical planning with set-up continuity in a two-stage ceramic firm 57 http://creativecommons.org/licenses/by-nc-nd/4.0/ ∑ f βlflt=1, ∀ l, t (41) αlflt ≤ xlflt , ∀ f, l, t (42) βlflt ≤ xlflt , ∀ f, l, t (43) 3*xlflt–∑ f xlflt ≤ αlflt+ βlflt , ∀ f, l, t (44) 2*xlflt– αlflt+ βl0fl ≤ 2*zlflt , ∀ f, l, t (45) 2*xlflt– αlflt+ βlflt–1 ≤ 2*zlflt , ∀ f, l, t (46) ∑ f zlflt ≤ nf *(3–αlflt–βlflt–βl0fl) , ∀ f, l, t = 1 (47) ∑ f zlflt ≤ nf *(3–αlflt–βlflt–βlflt–1) , , ∀ f, l, t > 1 (48) 2–∑ f ylflt ≤ 2 *(1–wllt), ∀ l, t (49) (∑ f zlflt )–1≤ nf *wllt, ∀ l, t (50) αlflt+ βlflt ≤(2–wllt), ∀ f, l, t (51) the objective function (1) expresses the minimization of the setup costs of the fs on the ls and ks (both stages) and the inventory costs of the fs at the middle (intermediate) and the end of the manufacturing process. constraints (2) and (3) are the inventory balance equations of in-process and finished fs, respectively. constraint (4) ensures that the capacity required for the setup of fs and the manufacturing of the lots assigned to each k do not exceed the capacity available on each k in each pt. constraint (5) indicates that a f can only be produced on a k in a pt if the k has previously be prepared to produce the f in such a pt. constraint (6) indicates that a f can only be produced on a k in a pt if it has previously been decided to produce the f on the k in such a pt. constraint (7) guarantees that should a certain amount of a f be produced on a k, it is equal to or above the minimum lot size established for the f on that k if the f is just produced in a single pt. constraint (8) allows not to produce the minimum lot size established for a f on a k in a pt, if either the f was the last one produced in the previous pt and the first one produced in the next pt, or the f is the only one produced during two consecutive pts. however, it guarantees in both cases that the total amount of f produced will be superior to its minimum lot size. constraint (9) establishes that if there is no amount of f produced on a k in a pt then it is not allowed to produce the f on the k in such a pt. constraint (10) establishes that if a f is produced on a k in a pt, then the k has been previously prepared to produce the f in such a pt. constraint (11) establishes that if a f is not produced on a k in a pt, then there is no setup on the k in such a pt. constraints (12) and (13) ensure that if a k “status” at the start of a pt is different from the “status” at the end of the previous pt, then at least one setup has to be made on the k in such a pt. constraint (14) indicates that if a k does not change its “status” during a pt, then it is already prepared (either at the start or the end of such a pt) to produce the same f. constraints (15) and (16) guarantee that a k can be only prepared to produce just one f, in the start and in the end of a pt, respectively. constraints (17) and (18) ensures that if a k is not prepared to produce a f in a pt, then that f can not be either the first or the last, respectively, for which the k was prepared in such a pt. constraint (19) indicates that if a k is only prepared to produce just one f in a pt, then the k should be prepared either at the start or the end of such a pt to produce the f. constraints (20) and (21) indicate that it is only possible to save a single changeover on a k in a pt if the k is prepared at the start of the current pt to manufacture the same f for which it was prepared at the end of the previous pt. constraints (22) and (23) indicate that if the “status” of a k at the start and the end of a current pt is equal to the “status” at the end of the previous pt, then just one or no f is manufactured. constraint (24) assures that if one or no f is manufactured on a k in a pt, then wl=0, although the contrary case does not imply wl=1. for this it is implemented constraint (25). constraint (26) guarantees that if more than one f is manufactured on a k in a pt, none of them can be the first and the last at the same time in such a pt. therefore, only in the case in which one or no f is manufactured on a k in a pt is possible that αh=1 and βh=1 for that f. constraints (27) and (28) are the inventory balance equations of intermediate products (between ls and ks). constraints from (29) to (51) are the same as constraints from (4) to (26) but in this case regarding to the ls. int. j. prod. manag. eng. (2016) 4(2), 53-64 creative commons attribution-noncommercial-noderivatives 4.0 international pérez perales, d., alemany, m.m.e. 58 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. numerical example the model validation is made by its application to a simplified two-stage real case within a ceramic firm. the input data and the solution obtained are described in the following sections. 4.1. input data description the model data are based on historical information (demand data) and on the mean real values (times and costs). the physical configuration has been slightly changed for confidentiality reasons, considering a problem of a size that represents the main relevant characteristics, but not excessively large so that it could be described in detail here. the model’s planning horizon is assumed to be half a year and it is divided into six monthly planification periods, from t1 to t6. six product families labelled f1 to f4 were included, each of them corresponding to different formats. only a single plant is considered, made up of two stages. first one, presses-glazing lines stage consists of three production lines , from l1 to l3. second one, kilns stage, consists of two kilns, k1 and k2. both stages are uncoupled by buffers. other relevant fgs data for the model can be consulted in tables 4-5. data of product families demand, and production capacity in each of the production lines (l) and kilns in each period of time are shown in table 4. table 4. data of product families (f) demand and production capacity of production lines (l) and kilns (k) in each pt (t). f d ft t1 t2 t3 t4 t5 t6 f1 100 125 135 140 150 130 f2 125 110 135 150 125 115 f3 140 125 110 130 115 125 f4 100 125 135 140 150 130 f5 125 110 135 150 125 115 f6 140 125 110 130 115 125 l capl lt t1 t2 t3 t4 t5 t6 l1 50 70 70 50 70 70 l2 70 50 70 50 50 70 l3 50 50 50 70 70 50 k caph kt t1 t2 t3 t4 t5 t6 k1 1000 1250 1000 1000 1200 1200 k2 1200 1100 1100 1200 1300 1000 in addition to the former table, some specific data of product families on production lines and kilns are shown in table 5. no backorder is permitted. the proposed model was translated to the mpl language, v4.2. the resolution was carried out with optimisation solver gurobi 4.5.1. the input data and the model solution values were processed with the microsoft access database (2007). the experiment was run on a pc with a 2.40 ghz processor and 2 gb of ram. table 5. specific data of product families (f) in each of the production lines (l) and kilns (k). f l i0 ciif tslf cslf tflf lmlf βl0lf f1 l1 50 0.1 2 35 0.1 160 0 f2 50 0.15 2.5 30 0.25 180 1 f3 50 0.2 3 40 0.2 175 0 f4 50 0.15 3.5 45 0.2 160 0 f5 50 0.25 2.5 30 0.1 180 0 f6 50 0.1 3 45 0.15 170 0 f1 l2 50 0.1 2 35 0.1 160 0 f2 50 0.15 2.5 30 0.25 180 0 f3 50 0.2 3 40 0.2 175 1 f4 50 0.15 3.5 45 0.2 160 0 f5 50 0.25 2.5 30 0.1 180 0 f6 50 0.1 3 45 0.15 170 0 f1 l3 50 0.1 2 35 0.1 160 1 f2 50 0.15 2.5 30 0.25 180 0 f3 50 0.2 3 40 0.2 175 0 f4 50 0.15 3.5 45 0.2 160 0 f5 50 0.25 2.5 30 0.1 180 0 f6 50 0.1 3 45 0.15 170 0 f k ii0 cif tshf cshf tfhf lmhf βh0kf f1 k1 50 0.1 10 120 1.5 160 0 f2 50 0.15 15 115 1.8 180 1 f3 50 0.2 18 100 2.5 175 0 f4 50 0.15 16 125 2 160 0 f5 50 0.25 15 110 3.5 180 0 f6 50 0.1 17 100 1.5 170 0 f1 k2 50 0.1 10 120 1.5 160 0 f2 50 0.15 15 115 1.8 180 0 f3 50 0.2 18 100 2.5 175 1 f4 50 0.15 16 125 2 160 0 f5 50 0.25 15 110 3.5 180 0 f6 50 0.1 17 100 1.5 170 0 4.2. evaluation of results the values of the decision variables linked to the production lines and kilns that lead to the optimum solution and help to validate the set-up continuity are shown in tables 6-9. int. j. prod. manag. eng. (2016) 4(2), 53-64creative commons attribution-noncommercial-noderivatives 4.0 international a mathematical programming model for tactical planning with set-up continuity in a two-stage ceramic firm 59 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 6. amount (m2) of product families (f) manufactured on production lines l1 and l2 in each pt. t1 t2 t3 t4 t5 t6 l1 pfl f1 f2 f3 f4 f5 f6 164 40 465 230 140 115 125 xl f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 1 yl f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 βl0=f2 αl βl f2 f6 f6 f6 f3 f6 f3 f3 f3 f3 f3 f3 zl f1 f2 f3 f4 f5 f6 1 1 wl 1 1 l2 pfl f1 f2 f3 f4 f5 f6 175 125 35 350 140 170 280 125 115 xl f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 1 1 yl f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 1 1 βl0=f3 alfal betal f3 f4 f4 f4 f4 f4 f4 f1 f1 f5 f5 f5 zl f1 f2 f3 f4 f5 f6 1 1 1 wl 1 1 1 table 7. amount (m2) of product families (f) manufactured on production line l3 in each pt. t1 t2 t3 t4 t5 t6 l3 pfl f1 f2 f3 f4 f5 f6 260 25 185 106 210 150 240 xl f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 1 yl f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 βl0=f1 αl βl f1 f5 f5 f5 f5 f2 f2 f2 f2 f2 f2 f2 zl f1 f2 f3 f4 f5 f6 1 1 wl 1 1 table 8. amount (m2) of product families(f)manufactured on kilns k1 and k2 in each pt. t1 t2 t3 t4 t5 t6 k1 pfh f1 f2 f3 f4 f5 f6 198 175 75 185 122 210 150 240 115 125 xh f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 1 1 1 yh f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 1 1 1 βh0=f2 αh βh f2 f5 f5 f5 f5 f2 f2 f2 f2 f3 f3 f3 zh f1 f2 f3 f4 f5 f6 1 1 1 1 wh 1 1 1 k2 pfh f1 f2 f3 f4 f5 f6 310 225 90 465 230 135 140 140 420 280 125 115 xh f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 1 1 1 1 1 yh f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 1 1 1 1 1 1 βh0=f3 αh βh f3 f6 f6 f6 f6 f4 f4 f1 f1 f5 f5 f5 zh f1 f2 f3 f4 f5 f6 1 1 1 1 1 1 wh 1 1 1 1 table 9. intermediate and final inventory of product families (f) at each pt (t). f iift ift t1 t2 t3 t4 t5 t6 t1 t2 t3 t4 t5 t6 f1 0 0 0 0 0 0 260 135 0 0 130 0 f2 16 16 0 0 0 0 123 13 0 0 115 0 f3 0 0 0 0 0 0 135 10 130 0 0 0 f4 0 35 250 0 0 0 125 0 0 280 130 0 f5 0 0 0 0 0 0 0 75 150 0 0 0 f6 0 0 0 0 0 0 0 340 370 240 125 0 the assessment method used consisted in analyzing if the model accounts for set-up continuity issues. however, the model operation was also assessed by two parameters: computational efficiency and the total costs in terms of tactical planning. these results confirm that the described constraints are valid to model the set-up continuity over discrete periods of time. it implies that if a product family is manufactured in two periods of time just one set-up is considered. this occurs just in case f is the last to be manufactured on a l or a k in a pt t and the first int. j. prod. manag. eng. (2016) 4(2), 53-64 creative commons attribution-noncommercial-noderivatives 4.0 international pérez perales, d., alemany, m.m.e. 60 http://creativecommons.org/licenses/by-nc-nd/4.0/ to be manufactured on the same l or k in pt t+1. in addition to that, the model allows the minimum lot size may be completed without any changeover during these two consecutive periods. a representative example may be seen in table 6 for f1, which is manufactured on l2 in two consecutive pts t=4 and t=5. f1 is manufactured at the end of pt t=4 in an amount less than its minimum lot size although it is also manufactured at the start of pt t=5, therefore meeting that minimum lot size of 160 and just considering one set-up instead of two. another example may be seen in table 8 for f4, which is manufactured on k2 in two consecutive pts t=3 and t=4. f4 is manufactured at the end of pt t=3 in an amount less than its minimum lot size although it is also manufactured at the start of pt t=4, therefore meeting that minimum lot size of 160 and just considering one set-up instead of two. in table 10 , the values of the total costs are shown. this paper focuses on the validation of set-up continuity issues so that some simplifications in the example are assumed, leading to approximated results of the reality. as aforementioned, tactical production planning in real ceramic sc scenarios includes a wide variety of production mix and other additional variables/costs, mainly related with the number of shifts planned in the press-glazing lines, the activation / desactivation of kilns or the subcontracting of some supplementary capacity for certain products families. in our example, the model generated a total cost of € 1894.4. the different components of the objective function appear in table 10: intermediate and final inventory costs and setup costs in both stages. backorder costs are not reflected in the model assuming that all the demand has to be fulfilled. finally, problem size characteristics and computational efficiency can be consulted in table 11. table 10. total costs. total costs intermediate inventory costs 47.55 final inventory costs 389.15 press-glazing lines set-up costs 255 kilns set-up costs 1125 1816.7 table 11. computational efficiency. computational efficiency iterations 133745292 variables 1182 integers 1182 constraints 2922 non-zero 12804 density (%) 0.4 time (hours) 30 mip best bound 1437.15 the computational efficiency parameter measures the computational effort required to solve models. the indicators are: the number of iterations needed by the solver and used to reach the final solution. table 11 shows the number of model variables, the number of integers in the model, the number of constraints in the model, the number of non-zero elements in the constraints matrix that the model contains, the density of the constraints matrix that the model contains, the cpu time required to obtain the model solution and the mip best bound. in this case, the model was solved by the standard solver setting the parameter “limit time” to 30 hours, obtaining a gap of 20% regarding to the optimal solution (table 11). more efficient solutions could be reached, applying other solution techniques. for instance, from the validation of the model, the authors have observed that the solution time of the model substantially decreases by fixing the value of the binary variables ylfltand yhfkt. therefore, the development of heuristics or metaheuristics similar to motta et.al (2013) that evaluate different solutions generated by fixing the value of the binary variables ylflt and yhfkt, transferring them as input data to the model and optimize the value of the remaining decision variables, could substantially reduce the solution time and the gap. however, this issue is out of scope of this work and constitutes a future research line. 5. conclusions this work presents a mix integer linear programming (milp) model to solve the tactical planning problem in a two stage production system in the ceramic sector for the purpose of minimizing product families int. j. prod. manag. eng. (2016) 4(2), 53-64creative commons attribution-noncommercial-noderivatives 4.0 international a mathematical programming model for tactical planning with set-up continuity in a two-stage ceramic firm 61 http://creativecommons.org/licenses/by-nc-nd/4.0/ references alemany, m.m., alarcón, f., lario, f.c., boj, j.j. (2009). planificación agregada en cadenas de suministro del sector cerámico. iii internacional conference on industrial engineering and industrial management. barcelona, spain 3-42. alemany, m.m., boj, j.j., mula, j., lario, f.c. (2011). mathematical programming model for centralised master planning in ceramic tile supply chains. international journal of production research, 48, 5053-5074. http://dx.doi.org/10.1080/00207540903055701 barany, i., van roy, t.j., wolsey, l.a. (1984). strong formulations for multi-item capacitated lot sizing. management science, 30(10), 1255–1261. http://dx.doi.org/10.1287/mnsc.30.10.1255 belvaux, g., wolsey, l.a. (2001). modelling practical lot-sizing problems as mixed-integer programs. management science, 47(7), 724-38. http://dx.doi.org/10.1287/mnsc.47.7.993.9800 chen, w.h., thizy, j.m. (1990). analysis of relaxations for the multi-item capacitated lot-sizing problem. annals of operations research, 26, 29-72. http://dx.doi.org/10.1007/bf02248584 chung, c., flynn, j., lin, c.m. (1994). an efective algorithm for the capacitated single item lot size problem. european journal of operational research, 75(2), 427-40. http://dx.doi.org/10.1016/0377-2217(94)90086-8 de toni, a., meneghetti, a. (2000).the production planning process for a network of firms in the textile-apparel industry. international journal of production economics, 65(1), 17-32. http://dx.doi.org/10.1016/s0925-5273(99)00087-0 eppen, g.d., martin, r.k. (1987). solving multi-item capacitated lot sizing problems using variable redefinition. operations research, 35(6), 832–48. http://dx.doi.org/10.1287/opre.35.6.832 fumero, y., montagna, j.m., corsano, g. (2012). simultaneous design and scheduling of a semicontinuous/batch plant for ethanol and derivatives production. article computers & chemical engineering, 36, 342-357. http://dx.doi.org/10.1016/j. compchemeng.2011.08.004 gopalakrishnan, m., ding, k., bourjolly, j.m., mohan, s. (2001). a tabu-search heuristic for the capacitated lot-sizing problem with set-up carryover. management science, 47(6), 851–863. http://dx.doi.org/10.1287/mnsc.47.6.851.9813 grieco, s., quirico, s., tullio, t. (2001). a review of different approaches to the fms loading problem. international journal of flexible manufacturing systems, 13(4), 361-384. http://dx.doi.org/10.1023/a:1012290630540 guo, z.x., wong, w.k., leung, s.y.s., fan, j.t., chan, s.f. (2006). mathematical model and genetic optimization for the job shop scheduling problem in a mixedand multi-product assembly environment: a case study based on the apparel industry. computers & industrial engineering, 50(3), 202-219. http://dx.doi.org/10.1016/j.cie.2006.03.003 haase, k. (1994). lotsizing and scheduling for production planning. in: lecture notes in economics and mathematical systems. springerverlag, berlin. http://dx.doi.org/10.1007/978-3-642-45735-7 set-up and inventory costs, while considering set-up continuity and a given forecasted mid-term demand. the model contemplated was validated by a ceramic real-world case example, but one on a smaller scale for the purpose of providing details of all the input data and of the solution obtained, focusing on product families allocation and set-up continuity aspects. although it is just considered a two-stage production process, it might be adapted to larger models for specific situations replicating the links between the additional stages and extrapolated to other semi-continuous production sectors. its main contributions are on one hand the accounting for explicit setup times at the tactical (aggregated) level which implies including decisions about the product families allocation and lot sizing of production. on the other hand the consideration of set-up continuity constraints, especially important in contexts with lengthy set-ups and where product families minimum run length are almost, equal or even higher than the planning period. the set up continuity modelling also allows the consideration of minimum lot sizes produced during two consecutive periods. both contributions help to achieve a more accurate capacity availability estimation in the tactical level so it may lead to feasible and more efficient events during the subsequent disaggregation into operational plas. the model has been validated by its application to a realistic ceramic firm. the obtained results confirm that the proposed model accounts for both issues: product families allocation and set-up continuity. for larger real problems with more time periods and/or products it should be necessary to develop solution techniques to reduce the computational time. for this reason, future research lines could develop efficient solution methods by means heuristic or metaheuristics applied to this problem. int. j. prod. manag. eng. (2016) 4(2), 53-64 creative commons attribution-noncommercial-noderivatives 4.0 international pérez perales, d., alemany, m.m.e. 62 http://dx.doi.org/10.1080/00207540903055701 http://dx.doi.org/10.1287/mnsc.30.10.1255 http://dx.doi.org/10.1287/mnsc.47.7.993.9800 http://dx.doi.org/10.1007/bf02248584 http://dx.doi.org/10.1016/0377-2217(94)90086-8 http://dx.doi.org/10.1016/s0925-5273(99)00087-0 http://dx.doi.org/10.1287/opre.35.6.832 http://dx.doi.org/10.1016/j.compchemeng.2011.08.004 http://dx.doi.org/10.1016/j.compchemeng.2011.08.004 http://dx.doi.org/10.1287/mnsc.47.6.851.9813 http://dx.doi.org/10.1023/a:1012290630540 http://dx.doi.org/10.1016/j.cie.2006.03.003 http://dx.doi.org/10.1007/978-3-642-45735-7 http://creativecommons.org/licenses/by-nc-nd/4.0/ hindi, k.s. (1996) .solving the clsp by a tabu search heuristic. journal of the operational research society, 47(1), 151–61. http://dx.doi. org/10.1057/jors.1996.13 ishikura, h. (1994). study on the production planning of apparel products: determining optimal production times and quantities. computers & industrial engineering, 27(1/4), 19-22http://dx.doi.org/10.1016/0360-8352(94)90227-5 kang, s., malik, k., thomas, l.j. (1999). lotsizing and scheduling on parallel machines with sequence dependent setup costs. management science, 45(2), 273-289. http://dx.doi.org/10.1287/mnsc.45.2.273 karacapilidis, n., pappis, c. (1996). production planning and control in textile industry: a case study. computers in industry, 30(2), 127-144. http://dx.doi.org/10.1016/0166-3615(96)00038-3 kopanos, g.m., puigjaner, l., georgiadis, m.c. (2012a). simultaneous production and logistics operations planning in semicontinuous food industries. omega, 40(5), 634-650. http://dx.doi.org/10.1016/j.omega.2011.12.002 kopanos, g.m., puigjaner, l., georgiadis, m.c. (2012b). single and multi-site production and distribution planning in food processing industries. computer aided chemical engineering, 31, 1030-1034. http://dx.doi.org/10.1016/b978-0-444-59506-5.50037-7 maes, j., mcclain, j.o., van wassenhove, l.n. (1991). multilevel capacitated lot sizing complexity and lp-based heuristics. european journal of operational research, 53(2), 131-48. http://dx.doi.org/10.1016/0377-2217(91)90130-n newson, e.f. (1975). multi-item lot size scheduling by heuristic, part i: with fixed resources. management science, 21(10), 1186-1193. http://dx.doi.org/10.1287/mnsc.21.10.1086 meijboom, b., obel, b. (2007). tactical coordination in a multi-location and multi-stage operations structure: a model and a pharmaceutical company case. omega, 35(3), 258-273. http://dx.doi.org/10.1016/j.omega.2005.06.003 min, l., cheng, w. (2006). genetic algorithms for the optimal common due date assignment and the optimal scheduling policy in parallel machine earliness/tardiness scheduling problems. robotics and computer-integrated manufacturing, 22(4), 279-287. http://dx.doi. org/10.1016/j.rcim.2004.12.005 motta c.f., resendo r., morelato p. (2013). a hybrid multi-population genetic algorithm applied to solve the multi-level capacitated lot sizing problem with backlogging, computers & operations research, 40 910–919. http://dx.doi.org/10.1016/j.cor.2012.11.002 mustafa, k., sinan, k., nesim, e. (1999). a generic model to solve tactical planning problems in flexible manufacturing systems. international journal of flexible manufacturing systems, 11(3), 215-243. http://dx.doi.org/10.1023/a:1008182411581 ngaia, e.w.t., penga, s., alexander, p., moon, k. (2014). decision support and intelligent systems in the textile and apparel supply chain: an academic review of research articles. expert systems with applications, 41(1), 81-91. http://dx.doi.org/10.1016/j.eswa.2013.07.013 özdamar, l., birbil, s.i. (1998). hybrid heuristics for the capacitated lot sizing and loading problem with setup times and overtime decisions. european journal of operational research,110(3), 525-547. http://dx.doi.org/10.1016/s0377-2217(97)00269-5 özdamar, l., bozyel, a. (1998).simultaneous lot sizing and loading of product families on parallel facilities of different classes. international journal of production research, 36(5), 1305-1324. http://dx.doi.org/10.1080/002075498193336 romsdal, a., thomassen, m.k., dreyer, h.c., strandhagen, j.o. (2011). fresh food supply chains; characteristics and supply chain requirements. 18th international annual euroma conference. cambridge, uk, cambridge university. pérez, d. (2013). framework and methodology proposal for the modeling of the supply chain collaborative planning process based on mathematical programming. application to the ceramic sector. dissertation, universitat politècnica de valència pérez, d., alemany m.m.e., lario, f.c., fuertes, v.s. (2014).set-up continuity in tactical planning of semi-continuous industrial processes. managing complexity. springer international publishing, 165-173. porkka, p., vepsalainen, a.p.j., kuula, m. (2003). multiperiod production planning carrying over set-up time. international journal of production research, 41(6), 1133–1148. http://dx.doi.org/10.1080/0020754021000042995 shabani, n., sowlati, t. (2013). a mixed integer non-linear programming model for tactical value chain optimization of a wood biomass power plant. applied energy, 104, 353-361. http://dx.doi.org/10.1016/j.apenergy.2012.11.013 soman, c.a., van donk, d.p., gaalman, g.j.c. (2004). combined make-to-order and make-to-stock in a food production system. international journal of production economics, 90(2), 223–235. http://dx.doi.org/10.1016/s0925-5273(02)00376-6 soman, c.a., van donk, d.p., gaalman, g.j.c. (2007). capacitated planning and scheduling for combined make-to-order and make-tostock production in the food industry: an illustrative case study. international journal of production economics, 108(1-2), 191–199. http://dx.doi.org/10.1016/j.ijpe.2006.12.042 sox, c.r., gao, y.b. (1999). the capacitated lot sizing problem with setup carry-over. iie transactions, 31(2), 173–181. http://dx.doi. org/10.1080/07408179908969816 suerie, c., stadtler, h. (2003). the capacitated lot-sizing problem with linked lot sizes. management science, 49(8), 1039–1054. http:// dx.doi.org/10.1287/mnsc.49.8.1039.16406 teimoury, e., modarres, m., ghasemzadeh, f., fathi, m. (2010). a queueing approach to production-inventory planning for supply chain with uncertain demands: case study of pakshoo chemicals company. journal of manufacturing systems, 29(2/3), 55-62. http://dx.doi. org/10.1016/j.jmsy.2010.08.003 int. j. prod. manag. eng. (2016) 4(2), 53-64creative commons attribution-noncommercial-noderivatives 4.0 international a mathematical programming model for tactical planning with set-up continuity in a two-stage ceramic firm 63 http://dx.doi.org/10.1057/jors.1996.13 http://dx.doi.org/10.1057/jors.1996.13 http://dx.doi.org/10.1016/0360-8352(94)90227-5 http://dx.doi.org/10.1287/mnsc.45.2.273 http://dx.doi.org/10.1016/0166-3615(96)00038-3 http://dx.doi.org/10.1016/j.omega.2011.12.002 http://dx.doi.org/10.1016/b978-0-444-59506-5.50037-7 http://dx.doi.org/10.1016/0377-2217(91)90130-n http://dx.doi.org/10.1287/mnsc.21.10.1086 http://dx.doi.org/10.1016/j.omega.2005.06.003 http://dx.doi.org/10.1016/j.rcim.2004.12.005 http://dx.doi.org/10.1016/j.rcim.2004.12.005 http://dx.doi.org/10.1016/j.cor.2012.11.002 http://dx.doi.org/10.1023/a:1008182411581 http://dx.doi.org/10.1016/j.eswa.2013.07.013 http://dx.doi.org/10.1016/s0377-2217(97)00269-5 http://dx.doi.org/10.1080/002075498193336 http://dx.doi.org/10.1080/0020754021000042995 http://dx.doi.org/10.1016/j.apenergy.2012.11.013 http://international journal of production economics http://international journal of production economics http://dx.doi.org/10.1016/s0925-5273(02)00376-6 http://dx.doi.org/10.1016/j.ijpe.2006.12.042 http://dx.doi.org/10.1080/07408179908969816 http://dx.doi.org/10.1080/07408179908969816 http://dx.doi.org/10.1287/mnsc.49.8.1039.16406 http://dx.doi.org/10.1287/mnsc.49.8.1039.16406 http://dx.doi.org/10.1016/j.jmsy.2010.08.003 http://dx.doi.org/10.1016/j.jmsy.2010.08.003 http://creativecommons.org/licenses/by-nc-nd/4.0/ ulstein, n.l., nygreen, b., sagli, j.r. (2007). tactical planning of offshore petroleum production. european journal of operational research, 176(1), 550-564. http://dx.doi.org/10.1016/j.ejor.2005.06.060 van donk, d.p. (2001). make to stock or make to order: the decoupling point in the food processing industries. international. journal of production economics, 69(3), 297–306. http://dx.doi.org/10.1016/s0925-5273(00)00035-9 van elzakker, m.a.h., zondervan, e., raikar, n.b., hoogland, h., grossmann, i.e. (2014). an sku decomposition algorithm for the tactical planning in the fmcg industry. computers & chemical engineering, 62(5), 80-95. http://dx.doi.org/10.1016/j.compchemeng.2013.11.008 wong, w.k., leung, s.y.s. (2008). genetic optimization of fabric utilization in apparel manufacturing. international journal of production economics, 114(1), 376-387. http://dx.doi.org/10.1016/j.ijpe.2008.02.012 int. j. prod. manag. eng. (2016) 4(2), 53-64 creative commons attribution-noncommercial-noderivatives 4.0 international pérez perales, d., alemany, m.m.e. 64 http://dx.doi.org/10.1016/j.ejor.2005.06.060 http://dx.doi.org/10.1016/s0925-5273(00)00035-9 http://dx.doi.org/10.1016/j.compchemeng.2013.11.008 http://international journal of production economics http://international journal of production economics http://dx.doi.org/10.1016/j.ijpe.2008.02.012 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j https://ojs.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2016.5780 received 2016-05-15 accepted: 2016-06-09 hybrid genetic algorithms: solutions in realistic dynamic and setup dependent job-shop scheduling problems rogério m. brancoa*, antônio s. coelhob1, sérgio f. mayerleb2 a instituto federal do rio grande do sul – ifrs campus rio grande, rua alfredo huch, 475, centro, cep 96201-460, rio grande, rs, brasil. b universidade federal de santa catarina, caixa postal 476, campus universitário ufsc trindade, cep 88040-900, florianópolis, sc, brasil. a rogerio.branco@gmail.com b1 a.s.coelho@ufsc.br b2 sergio.mayerle@ufsc.br abstract: this paper discusses the application of heuristic-based evolutionary technique in search for solutions concerning the dynamic job-shop scheduling problems with dependent setup times and alternate routes. with a combinatorial nature, these problems belong to an np-hard class, with an aggravated condition when in realistic, dynamic and therefore, more complex cases than the traditional static ones. the proposed genetic algorithm executes two important functions: choose the routes using dispatching rules when forming each individual from a defined set of available machines and, also make the scheduling for each of these individuals created. the chromosome codifies a route, or the selected machines, and also an order to process the operations. in essence , each individual needs to be decoded by the scheduler to evaluate its time of completion, so the fitness function of the genetic algorithm, applying the modified giffler and thomson’s algorithm, obtains a scheduling of the selected routes in a given planning horizon. the scheduler considers the preparation time between operations on the machines and can manage operations exchange respecting the route and the order given by the chromosome. the best results in the evolutionary process are individuals with routes and processing orders optimized for this type of problem. key words: genetic algorithms, dispatching rules, realistic job-shop scheduling. 1. introduction the growing competition from companies , arising from the globalized market, reinforces their attention to quality and productivity, focusing in the relationships in the supply chain and flexibility, increasing the efficiency of manufacturing. in this context, the flexible manufacturing system (fms) combines high flexibility, productivity and low levels of stock: characteristics that accept the alternative routes of production and make it more agile and robust in face of failures. so, if a machine breaks during a task, a reschedule to find an alternate route is done to finish this job, respecting due dates already planned (porter et al., 1999; chan, 2003). in respect to manufacturing systems directly involved with cells and fms, porter et al. (1999) and matsuzaki (2004) point to the job-shop class in the production of small volumes and more variety of concurrent processes. in general, the job-shops are processoriented production systems and obey a pre-defined sequence of processing. the scheduling problems are widely studied because they assume difficult conditions to solve in polynomial time (np-hard), due to their combinatorial nature (allocating machines to int. j. prod. manag. eng. (2016) 4(2), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international 75 http://dx.doi.org/10.4995/ijpme.2016.5780 mailto:rogerio.branco@gmail.com mailto:a.s.coelho@ufsc.br mailto:sergio.mayerle@ufsc.br http://creativecommons.org/licenses/by-nc-nd/4.0/ produce parts). also, the flexibility of alternate routes increases the combinations of resources to compose sequences and therefore, the problem complexity. the literature presents researches involving the sdst-jssp sequence dependent setup times jobshop problems, which are classic jssp extensions and in which a setup time between two consecutive operations is required. these extensions make the classic cases closer to realistic situations but more complex. in dynamic cases, another extension commonly seen is the ndd-jssp non deterministic dynamic job shop problems, which differs from classical because the process does not start at time 0, with random characteristics over starting times and thus, close to realistic cases. regarding this perspective, this paper propounds combined heuristics techniques and genetic algorithms to solve the combination of these two kind of problems, called ndd-sdst-jssp. the scheduling algorithm contemplates the goals of shorter processing time and delivery due dates, late start times, variations in processing times of tasks and dependent setup times. also, there’s another one that forms routes and their internal sequencing, integrated in the ga’s evaluation function. 2. the jssp problem the scheduling problems belong to the np-hard problems and exact methods are applied only for relatively small examples of the problem (araújo, 2006). furthermore, real problems have additional details which involve more combinations than the classics (herrmann et al., 1995). the classical jssp is a group of n jobs to be processed into a set of m machines. each task has a number of operations and a technological sequence of process. these operations require an uninterrupted processing time over a designed machine. therefore, it is a timecompletion problem that satisfies the constraints: the goal is the minor total completion time makespan (vazquez and whitley, 2000). in the sdst-jssp, there is a setup time between consecutive operations in the same machine. thus, once the operation ojv leaves the machine mv, before the okv process starts, a setup time soiv,okv is added (gonzales et al., 2005). 3. methods to solve jssps in the literature, there is a great diversity of methods applied to solve the jssps. jain and meeran (1998) cite johnson (1954) as one of the firsts significant works in the theory of scheduling, which aimed to minimize the makespan. several other studies have followed him, where the variety of techniques involved and the forms to modelling these problems greatly increased over these nearly six decades. following, some methods applied to solve the jssp can be viewed, without the intention to terminate the discussion about the subject, but only listing the most expressive techniques quite evident in the literature which belong to this state of the art. 3.1. exact and aproximative methods several strategies are presented in the literature to solve the classic jssp. in decision problems as these ones, the critical path method cpm is one of the most mentioned. also, formulations involving linear programming (lp), integer linear programming (ilp) or mixed integer programming (mip) are used. furthermore, the enumerative methods such as branch and bound (bb) are strong highlights and the dynamic programming is evidenced in the optimal solution of the classic jssps. in general, many simplifications are required to problems in order to find solutions, and they still are little adaptable to variations in size, where applications are restricted to a few small problems (wall, 1996). it is not difficult to imagine that for real jssps, more complex than the classic ones, these implementation strategies will be aggravated. so, a non exact method is now treated, due to its potential in solving jssps: the metaheuristics. they are able to search solutions, consisting in the application, at each step, of a subordinate heuristic, which has to be modeled for each specific problem. for them, the principle adopted to explore the solution in the space search can be local or populational. in the first case the operation is performed by means of movement applied to each step on the current solution, generating another promising approach in int. j. prod. manag. eng. (2016) 4(2), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international branco, r. m., coelho, a. s., mayerle, s. f. 76 http://creativecommons.org/licenses/by-nc-nd/4.0/ its vicinity, like the tabu search or the simulated annealing techniques. already, the methods based on population search, are to maintain a set of good solutions and combine them in order to try to produce even better solutions. classic examples are the genetic and also, the memetic algorithms. the following will be arranged some more detailed operating information of these metaheuristics, focusing on the problem dealt with in this work. 3.2. tabu search this metaheuristic is considered an iterative global optimization technique. having originated from the search for integer programming problem solving, it was later extended to almost all combinatorial problems (goldbarg and luna, 2000). in general, the tabu search (tabu search ts) is a procedure that restricts the search and tries to find optimal solutions, storing the search history in memory. it prohibits (tabu) movements in the neighbourhood with certain attributes, in order to guide the search process as well as solutions (based on available information) have double or are similar to previously stored solutions / obtained. the work of jain and meeran (1998) takes the ts as one of the most efficient search for good solutions in classical job-shop systems. they note that methods like branch and bound, if combined, show improvements in search, yet with greater computational cost. like most local search strategies, ts requires many parameters that must be carefully adjusted. considering the imminent application of differences in actual cases studied, this may be a difficult barrier to be overcome. 3.3. simulated annealing belonging to the random-guided search techniques, simulated annealing sa presents random components, but also employ current status information to guide the search of the solution of the problem studied. it is a local search method that accepts worsening movements to escape from local optima. it is based on an analogy with thermodynamics, simulating the cooling of a heated set of atoms. for the use of sa should be defined a priori, a method for generating an initial solution s, a method for generating the surrounding solutions s (neighborhood structure) and an objective function f(s) to be optimized (mauri and lorena, 2006). some contributions to neighborhood functions for jssp were showed by jain and meeran (1998). basically they consist of the reversal of processing orders from a pair of adjacent critical operations for the same machine. this method of sa proposed appears to be quite robust to the jssp, but jain and meeran (1998) mentioned that the results are, also, poor. only when incorporated into other techniques (eg .: genetic algorithm) is that the quality of the results is improved. the authors also mention the excessive consumption of computational time for good solutions can be found, and the high dependence of the parameters to the algorithm’s nature. it adds that slow colds also potentiate the best results, but also generate a considerable computational time consumption. 3.4. evolutionary algorithms the evolutionary algorithms (eas) are heuristic search techniques based on natural selection mechanisms, computationally simulating the environments which use the principles of evolution and heredity. they operate with a population of solutions (chromosomes, or individuals), applying selection techniques guided by the ability of each one and subsequently genetic operators such as reproduction and mutation act on them, generating new individuals, new solutions. according to linden (2008), there are several proposed computational models based on the concept of simulation of evolution through selection and breeding and mutation operators, all dependent on each individual’s fitness in their species and the environment in which it’s inserted. barboza (2005) cite some of these methods, as the evolutionary strategies (ee), genetic programming (pg), classifier systems (cs) and transgenetic computational (ct), among others, saying that the most widespread and researched is the genetic algorithm (ga), given their flexibility and effectiveness in performing global search in different environments. 4. applied methods in general, the solution method proposed for nddsdst-jssp is a combination of dispatching rules, int. j. prod. manag. eng. (2016) 4(2), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international hybrid genetic algorithms: solutions in realistic dynamic and setup dependent job-shop scheduling problems 77 http://creativecommons.org/licenses/by-nc-nd/4.0/ the genetic and the modified gt algorithms (giffler and thompson, 1960). thus, several considerations were made, starting from the chromosome coding, application of genetic operators and evaluation of individuals. 4.1. the priority dispatching rules the most popular heuristic techniques applied to scheduling problems, the priority dispatching rules (pdr´s), have demonstrated their importance in several works and even today, are still widely used in combined methods. some examples are the work of singh, mehta and jain (2006), el-bouri and shah (2006), branco (2010), among others. such importance lies in the easy implementation and low computational cost required. in general, the procedure is to choose a set of operations, not scheduled yet. according to a criterion of choice, a set formed by operations that can be processed in a specific machine will have one of them selected by this adopted criterion, which will be inserted into the scheduling. according to the tests, are used the following know rules from the literature: rnd (random): rule based in random uniformly distributed variable; spt (shortest process time): gives greater priority to the task that presents smaller processing time. s/rpt (slack per remaining processing time): gives priority to operations based on the composite index by the ratio of the delivery date, subtracts the task completion and remaining processing time. there are several other rules, such as srpt, ltwk and spt/twk, commented by chiang and fu (2006). extensions of spt incorporate, under combined conditions, other goals, also with good efficiency and late operations. jain and meeran (1998) apud chang et al. (1996) show a study evaluating 42 pdrs applied in an integer linear programming model, where the spt rule showed the best performance. regarding work interests, it is important to consider the due time when implementing processes, given the characteristics of demand oriented to orders/ requests that the processes are subject to, but without forgetting the relevant conditions concerning flow time in processes, that leads to combined rules. thus, the aim is to combine some simpler rules to promote better results in low computational time. therefore, sptq (less time needed to complete a process) is selected in equation 1 and s/rpt, in equation 2, motivated by the success of the first, in a wide variety of jobs and, for the latter, considering the time needed to complete the ongoing processes. mi q q j ispt piq = = ∑ (1) mi i q j i q mi q q j d t piq d t ispt isprt isptpiq = = − − − − = = ∑ ∑ (2) where: di = jobi due date; pij = processing time of the operation j in job i; t = current time; mi = number of operations remaining to finish jobi; mi q j piq = ∑ = isptq = process time remaining (jobi); both indexes are inversely proportional to the priority value, i.e., higher priority to lower value and lower priority to higher values, so the algorithm is built to prioritize operations with lower rates. the ichr is given by the equations below: . . i qq q q d t ispt ichr ispt isprt ispt ispt − − = = (3) i qichr d t ispt= − − (4) 4.2. the genetic algorithm solving a wide variety of problems in class np complete, evolutionary algorithms (eas) make heuristic search techniques based on natural mechanisms of selection, simulating computational environments based on these principles of evolution and heredity (goldberg, 1989). int. j. prod. manag. eng. (2016) 4(2), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international branco, r. m., coelho, a. s., mayerle, s. f. 78 http://creativecommons.org/licenses/by-nc-nd/4.0/ the proposed structure to chromosome has two known parts: head and body. the “head” contains the information of the route and the “body” will act in the operation sequence, both with the same dimensions and, for each operation, the locus contains a machine index. the figure 1 shows the relationship between the operations and machines available to process them and the structure of the “head” of the chromosome.   figure 1. example of relation between the original table of operations and the chromossome structure – phenotype×genotype (source: own). also, the “body” contains whole alleles, not repeated, in the interval ai=[0,total operations-1] and, in the scheduling, it indexes the order of operations, as the figure 2 shows.   figure 2. example of scheduling process from a given chromosome – phenotype×genotype (source: own). 4.3. creating the initial population since the “head” must be built first, this is made using random numbers in the range of [0,nmij-1], where nmij is the number of available machines to process operation j of the process i. the figure 1 above shows this construction part, where resource allocation starts building the “body”, an ordered 0 to nmij-1 array. 4.4. evaluation of the population the function corresponding to the evaluation of the population, individual by individual, is the fitness function. each of these values are the quantification of its adaptations. in other words, it means to apply the modified gt algorithm to make the active scheduling for each chromosome. the time completion can be the objective function of the search. where: n = number of tasks; oik = operation k of job i; tik = time able to start operation k of job i; pik = processing time of the operation k belonging to the job i. as follows, the modified gt algorithm schedulings: modified gt algorithm: step 1: place the first schedulable operation of each task (of the active planning horizon) in the set of candidate operations c={oi1|1 ≤ i ≤ n}; step 2: choose an operation o’ of c, with earliest completion time; step 3: determine the machine m’, in which o’ must be processed and thus build the set g (the conflict set of m’), consisting of all operations of c to be executed in m’; step 4: remove operations that do not start before o’ finish, g = {oik ∈ g | tik <t'+p'} step 5: run the subalgorithm to select an operation o*ik of g; step 6: remove o*ik from c, where c=c\{oik} step 7: insert the operation o*ik in the schedule and calculate start time; step 8: insert the successor operation of the o*ik in the set c (if any); step 9: if c ≠ ∅ , go to step 2, if not end. end of sub-algorithm. int. j. prod. manag. eng. (2016) 4(2), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international hybrid genetic algorithms: solutions in realistic dynamic and setup dependent job-shop scheduling problems 79 http://creativecommons.org/licenses/by-nc-nd/4.0/ and, to calculate the step 5 (chosing the o*ik of g), the sub-algorithm is now presented: sub-algorithm step 5 create cr: set of operations using m’, unscheduled, and with previous scheduled; if g ∩ cr ≠ {} then choose operation g∩cr with higher priority index (rhp’s) else create pg: set of processes belonging to the g operations; create crg: subset of cr, with the operations ∈ to the processes of pg; if crg ≠ {} then find o: operation of crg with higher priority index; find j*: process that contains the task o; find o*: operation of g belonging to j*; else find o*: oper. of g with low priority; end if; end if; end of sub-algorithm. for the m’ machine, if the operation is to be the one sequenced by thex “body” of the chromosome and the one with highest priority index, it will also be the operation o*, i.e., the candidate operation elected to be scheduled. otherwise, if the sequence does not match with the job-shop problem, conflicts may occur. because there are two conflicting interests, both of them without the creation of unfeasible individuals must be considered. the goal is to find, in g, an operation that most closely matches to that suggested by the chromosome’s “body”, i.e., to the machine m’, it is tried to schedule an operation of g which belongs to the same process of the highest indexed operation from cr. if no intersection of g with the operation indicated by the sequencing for m’ exists, then the one with less priority is selected, relaxing in some choice criterion. 4.5. selecting individuals the selection process must list individuals, which will be part of the reproduction. the selection method adopted was proposed by mayerle (1994), which consists in a ordered stochastic selection, having, in maximization, individuals in decreasing order according to their fitness, as follows). ( )sel r r r j m rnd m m 1 2 1 4 j 2$ != = + + +^ h; e) 3 (5) where: r is the set of the m individuals; rj is the j-th chromosome; rnd is random uniformly distributed ∈ [0,1); ⌈x⌉ is the smallest integer greater than x. the method provides less selective pressure then monte carlo’s selection (roulette), and also allows the best individuals to have more chance of crossing than the less able ones with the merit of recovering the super-individual effect, which will possibly exist due to the elitist strategy regarding the population formation. 4.6. strategy to form a population the population is created considering four different formation processes: cloning, random formation, greedy formation and reproduction. because there is a sorting process, the best are cloned to the current population, but in a small fraction of the total population. other small fraction is designed to individuals generated by the original algorithm of the first population. the fourth way of formation consists in generating a very small population part using dispatching rules (drs) widely discussed in the literature, as: spt, s/prt, chr and rnd (shortest processing time, slack per remaining time, combined heuristic rule and random, respectively). the user also defines the number of individuals to compete for reproduction. this strategy is based on gonçalves et al. (2005), whose intention is to avoid a premature convergence of the population. 4.7. genetic operators applied to ga the operators applied in the ga are: crossover, mutation and cloning. int. j. prod. manag. eng. (2016) 4(2), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international branco, r. m., coelho, a. s., mayerle, s. f. 80 http://creativecommons.org/licenses/by-nc-nd/4.0/ starting with the crossover operator, the main operator of genetic algorithm, perpetuates the characteristics of the fittest individuals through the exchange of parents’ information, passing it to the offspring individuals. the application is different to the “head” and to the “body”. to the first, their alleles are copied to the offsprings. for this, the uniform crossover is applied, based on a binary mask formed by 0/1 digits. figure 3. example of the “mask” used into the uniform crossover operator (source: own). in crossover process, a mask allele “zero” means, to the offspring, that the gene’s donor is father 1. otherwise, the donor is father 2. to another offspring, the reversal of the mask is required before the process. no harmonization is required. after, individuals are subjected to the second phase of the crossing, now using order-based operators. this is necessary due to the desire to keep the sequence proposed by the parents as faithful as possible. the operator now is the ppx (precedence preservative crossover), acting in the chromosomes’ “bodies” and using the masks involved in the previous step. inheriting from his father 1 all the genes situated in his respective locus of the “zero” allele mask, it starts to complete the sequence based on the father 2. if the allele obtained from the father 2 already exists in some locus of offspring 1, it sought the next, until the gap can be completed.   figure 4. example of the ppx crossover operator used to from a new offspring (source: own). this operator was applied based on the observations of gonçalves et al. (2005), such operators produce good effects when applied to schedules, instead of the traditional one or two cut points. moreover, during the crossover process, the mutation operator will actuate changing the value sampled by another one, randomly generated. in phenotype terms, it’s a new route, meaning a new machine assignment to that operation. in the “body”, the mutation will exchange genes between two randomly selected points, forcing task sequence changes. 5. applied tests with interest in the analisys of how the algorithm solves the proposed problems types and their characteristics, the original data from chan (2003) and kumar et al. (2003) (appendix i and ii, respectively) were applied in the tests. although, some modified problems proposed by araujo (2006) were, also applied. 5.1. from simple routes or classic jssps in these cases, the combined dispatching rule presented in equation 4 contributes to form the route with specific criteria. this chr is choosen based in a previous survey of branco (2010) among others, when it is used in classic jssps with good results to these problems which have the characteristic of just one route to choose. thus, the chr is also applied to the ndd-sdst-jssp presented in this work. 5.2. expansion of alternative routes in order to evaluate the algorithm ability with alternative routes combinations, araujo (2006) expanded the original chan (2003) problem (data in appendix 1), introducing another machine and so, turning some task processes more flexible. the average makespan with the proposed conjuncture was reduced, at minimum values from 931 ut to 786 ut, with average computational time close to 20 s using a dual core cpu with 2 gbytes of ram. it was expected, since the increase of machine number also increases the number of routes and thus the processing options. araujo (2006), observing it has raised the issue that, since it increases the amount of routes and thus reduces the total processing time, reduces, at the same time, the frequency of use of machines. int. j. prod. manag. eng. (2016) 4(2), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international hybrid genetic algorithms: solutions in realistic dynamic and setup dependent job-shop scheduling problems 81 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5.3. non-deterministic dynamic characteristics for the tests involving ndd-jssp (realistic) simulation routines in the original algorithm were introduced with the aim to generate random variables, and the simulated times are: instant start operation processing, transport time between machines, operation processing time in the machine. all these times are crucial for defining the end time of each operation, in each job. the simulation of turbulent environment scheduling consists in to generate random variables with defined average and standard deviation. the average is equal to the original time and standard deviation will be between 5, 10 and 20%. these indeterminisms created during the scheduling, they vary setup times (dependent), processing times, start and traveling time are important to evaluate the behavior of the solution ndd-jjsp proposed, since it reacts at the time when the change is detected, adapting itself. considering the objective function and the chan (2003) original problem, the minimum and average data obtained for each case are: 960 and 990, 1013 and 1045, 995 and 1022, respectively, for the turbulence simulations with 5, 10 and 20%. it is noticed that they vary little from one class to another, according to the disturbances generated, which presents good scheduling in dynamic environments, as can be viewed in table 1. table 1. results of mean makespans from the solved problems, considering disturbances of 0, 5, 10 and 20% and 1 and 5% mutations rates. mean makespan kumar chan chan (expand.) disturb 1% 5% 1% 5% 1% 5% 0% 385 383 944 949 842 830 5% 431 432 997 996 859 840 10% 457 459 1049 1051 899 909 20% 527 523 1168 1167 1048 1032 in a non turbulent environment and, the makespan obtained by chan (2003) is close to 1000 ut, araujo (2006) a mean of 1070 ut and this work reached 931/944 of minimal/mean values. the same performance analysis is done to the kumar et al. (2003) problem (appendix 2), where the closeness of the makespans results is also observed, which were: 381, 424, 501 ut. the adopted conjuncture is robust with regard to schedules in turbulent environments. araujo (2006) has obtained 410 ut for the same demand of pieces. comparing and considering no disturbances, this work got a 350/383 ut of minimum/mean makespans values, meaning good response even with disturbances. 5.4. test of dependent setup times to evaluate the dependent setup times in the jssps, the problem in table 2, is approached with the goal to analise the solutions to this classic problem with and without setup times. table 2. data from a randomly formed 3x3 jssp (source: yamada and nakano (1997)). job oper. machine order tstart tstopl pij 3 3 3 0 0 1 1 3 1 2 1 0 3 3 1 1 1 2 0 3 3 2 3 2 3 0 4 4 2 2 3 4 0 3 3 1 3 3 5 0 3 3 1 2 2 6 0 3 3 2 1 1 7 0 2 2 3 2 1 8 0 2 2 unique of each chromosome, a processing order is established for each operation (“body”), as the machines that will process each task. this test shows how the scheduler creates a agenda from a given chromosome and a table of dependent setup times among operations, where the scheduler can obtain a result as the figure 5, with 15 ut to be finished. with a setup dependent time between operations (1,1,1) and (2,1,1), in the assigned sequence of the machine 1 (iteration 3), a second scheduling can be obtained (figure 6). figure 5. example of solution of yamada and nakano 3×3 jssp (source: own). int. j. prod. manag. eng. (2016) 4(2), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international branco, r. m., coelho, a. s., mayerle, s. f. 82 http://creativecommons.org/licenses/by-nc-nd/4.0/ it is observed that the depend setup time made the scheduler choose, in its 3rd iteration, the operation (3,2,1), instead of (2,1,1). being dependent on a predecessor operation, the processing time will increase from 2 ut to 5 ut, leading the the completion time to increase from 5 ut to 8 ut   figure 6. example of solution of yamada, nakano 3×3 jssp considering dependent setup time (source: own). the scheduler has taken the best choice, considering de cost to accept, as in the figure 3, the (2,1,1) after the (1,1,1). this choice saved 1 ut in this special case, so the goal was achieved when the scheduler did this exchange of tasks aimed at performing the best schedule, avoiding the time costs of the operations for the same order established by the same chromosome. 6. discussions of results this paper proposes a scheduling technique using combined heuristic rule (chr) in schedules based in a modified version of the gt algorithm. the focus is to promote efficient scheduling in realistic job-shop problems (no determinism in times of operations), and, at the same time, consider the dependent setup times between operations in the same machine. with the implemented algorithms in object-pascal language was possible to observe the general behavior in turbulent environments (varying time of operations process), with good scheduling capabilities without exceeding the proposed completion times. also the good results found when compared to those obtained by other authors made further encouraging the implementation in future decision-making process. the heuristic rule chr was efficient in the test results, which usually get good solutions with time quite satisfactory, probably due to components based on rules: spt and s/prt. regardless, the adopted situation for the genetic algorithm, which inserts individuals formed by other heuristics, could “dope” at low rates the population that was being built at each iteration. the super-individual absence was important to the good performance of the ga, as a result of the selection proposed by mayerle (1994). with focus on problems with multiple route options, as the problems posed by kumar et al. (2003) and chan (2003), algorithms for route selection (using coding “head/body” of the chromosome and equipment tables available on the shop floor, as well as their processing times for each operation and setup times) and scheduling (inserted into the ga as the fitness function based on the modified algorithm gt), the results reached makespan, in some cases, better than the original solution problems. the expectations have not changed when the results were contrasted with those obtained by araujo (2006), solved by other circumstances, also based on genetic algorithms. considering the problems demonstrates the proposal ability to organize, for a given planning horizon, the tasks to be scheduled, whether belonging to a physical or virtual manufacturing cell, which may form an alternate route, also considering dependent setup times. about dependent setup times, the tests can show good results, exchanging operations in order to find best completion times, scapping from the costs imposed by the times among operations, respecting the order predefined by the chromosome. references araujo, l. o. (2006). método de programação de sistemas de manufatura do tipo job shop dinâmico não determinístico. tese (doutorado), universidade de são paulo, são paulo. barboza, a. o. (2005). simulação e técnicas da computação evolucionária aplicadas à problemas de programação linear inteira mista, tese (doutorado), universidade tecnológica federal do paraná. paraná. branco, r. m. (2010). agendamento de tarefas em sistemas de manufatura job-shop realista com demanda por encomenda: solução por algoritmo genético, tese (doutorado), universidade federal de santa catarina. chan, f. t. s. (2003). effects of dispatching and routeing decisions on the performance of flexible manufacturing system. international journal of advanced manufacturing technology, 21(5), 328-338. doi:10.1007/s001700300038 int. j. prod. manag. eng. (2016) 4(2), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international hybrid genetic algorithms: solutions in realistic dynamic and setup dependent job-shop scheduling problems 83 http://dx.doi.org/10.1007/s001700300038 http://creativecommons.org/licenses/by-nc-nd/4.0/ chiang, t-c., fu, l-c. (2006). using dispatching rules for job shop scheduling with due date-based objectives, proceedings of the 2006 ieee international conference on robotics and automation, orlando, florida. doi:10.1109/robot.2006.1641909 el-bouri, a., shah, p. (2006). a neural network for dispatching rule selection in a job shop. the international journal of advanced manufacturing technology, 31(3), 342-349. doi:10.1007/s00170-005-0190-y giffler, b., thompson, g. (1960). algorithms for solving production scheduling problems. operations research, 8(4), 487-503. doi:10.1287/ opre.8.4.487 goldbarg, m. c., luna, h. p. l. (2000). otimização combinatória e programação linear. 2 ed. editora campus. goldberg, d. e. (1989). genetic algorithms in search, optimization, and machine learning. us: new york gonçalves, j. f., de magalhães mendes, j. j., resende, m. g. (2005). a hybrid genetic algorithm for the job shop scheduling problem. european journal of operational research, 167(1), 77-95. doi:10.1016/j.ejor.2004.03.012 herrmann, j. w., lee, c.-y., hinchman, j. (1995). global job shop scheduling with a genetic algorithm. production and operations management, 4(1), 30-45. doi:10.1111/j.1937-5956.1995.tb00039.x jain, a. s., meeran, s. (1998). a state-of-the-art review of job-shop scheduling techniques, technical report, department of applied physics, electronics and mechanical engineering, university of dundee, scotland. johnson, s. m., (1954). optimal two and three-stage production schedules with setup times included. naval research logistics quarterly, 1(1), 61-68. doi:10.1002/nav.3800010110 kumar, r., tiwari, m. k., shankar, r. (2003). scheduling of flexible manufacturing systems: an ant colony optimization approach. proceedings of the institution of mechanical engineers, part b: journal of engineering manufacture, 217(10), 1443-1453. doi:10.1243/095440503322617216 linden, r. (2008). algoritmos genéticos. brasport, 2 edição. 2008. mauri, g. r., lorena, l. a. n. (2006). simulated annealing aplicado a um modelo geral do problema de roteirização e programação de veículos. xxxviii sbpo simpósio brasileiro de pesquisa operacional. goiânia, go. matsusaki, c. t. m. (2004). modelagem de sistemas de controle distribuídos e colaborativos de sistemas produtivos. tese (doutorado), universidade de são paulo, são paulo. doi:10.11606/t.3.2004.tde-20122004-112454 mayerle, s. f. (1994). um algoritmo genético para a solução do caixeiro viajante. tecnical report, departamento de engenharia de produção e sistemas, ufsc. porter, k., little, d., peck, m., rollins, r. (1999). manufacturing classifications: relationship with production control systems. integrated manufacturing systems, 10(3-4): 189-198. doi:10.1108/09576069910280431 singh, a., mehta, n. k., jain, p. k. (2007). multicriteria dynamic scheduling by swapping of dispatching rules. international journal of advanced manufacturing technology, 34(9), 988-1007. doi:10.1007/s00170-006-0674-4 vázquez, m., whitley, d. (2000). a comparison of genetic algorithms in solving the dynamic job shop scheduling problem. gecco. available in https://docs.google.com/file/d/0b0xb4crovcgtmxpkrnp6nfo3btq/edit wall, m. b. (1996). a genetic algorithm for resource-constrained scheduling, tesis, departament of mechanical engineering, massachusetts institute of technology. int. j. prod. manag. eng. (2016) 4(2), 75-85 creative commons attribution-noncommercial-noderivatives 4.0 international pérez perales, d., alemany, m.m.e. 84 http://dx.doi.org/10.1109/robot.2006.1641909 http://dx.doi.org/10.1007/s00170-005-0190-y http://dx.doi.org/10.1287/opre.8.4.487 http://dx.doi.org/10.1287/opre.8.4.487 http://dx.doi.org/10.1016/j.ejor.2004.03.012 http://dx.doi.org/10.1111/j.1937-5956.1995.tb00039.x http://dx.doi.org/10.1002/nav.3800010110 http://dx.doi.org/10.1243/095440503322617216 http://dx.doi.org/10.11606/t.3.2004.tde-20122004-112454 http://dx.doi.org/10.1108/09576069910280431 http://dx.doi.org/10.1007/s00170-006-0674-4 https://docs.google.com/file/d/0b0xb4crovcgtmxpkrnp6nfo3btq/edit http://creativecommons.org/licenses/by-nc-nd/4.0/ appendix i table i. data from the original (without gray data) and modified problem of chan (2003) (source: araujo(2006)). job operation 1 2 3 4 1 m1 (105) m3 (168) m5 (70) m2 (210) m2 (126) m4 (140) m3 (140) m1 (175) m6 (91) m4 (175) 2 m2 (140) m3 (70) m5 (245) m4 (175) m3 (168) m2 (112) m2 (266) m5 (203) m7 (154) 3 m5 (200) m1 (125) m4 (150) m2 (75) m1 (100) m2 (60) m3 (135) m4 (90) m7 (60) 4 m4 (150) m2 (150) m5 (100) m3 (125) m2 (75) m5 (75) m4 (50) m1 (75) m6 (100) 5 m1 (50) m3 (100) m2 (75) m4 (150) m3 (40) m4 (150) m5 (100) m3 (125) m6 (125) 6 m3 (175) m2 (84) m1 (175) m5 (70) m5 (140) m4 (56) m4 (140) m3 (161) m5 (70) 7 m4 (245) m5 (70) m1 (70) m2 (105) m1 (266) m4 (126) m4 (105) m5 (170) m3 (245) m7 (105) 8 m5 (105) m4 (280) m3 (175) m1 (140) m4 (70) m5 (210) m2 (140) m2 (56) appendix ii table ii. data of the problem proposed by kumar et al. (2003) (source: araujo(2006)). job operation 1 2 3 4 1 m1 (7) m2 (3) m1 (3) m1 (2) m3 (4) m3 (6) m2 (4) 2 m1 (8) m3 (4) m1 (7) m1 (8) m2 (12) m2 (14) m3 (4) 3 m1 (10) m2 (2) m1 (2) m1 (6) m2 (15) m3 (6) m3 (4) m2 (3) m3 (8) 4 m2 (9) m1 (6) m2 (7) m1 (9) m3 (5) m3 (2) m3 (12) m2 (6) m3 (3) 5 m1 (10) m2 (7) m1 (5) m1 (4) m3 (15) m3 (14) m2 (8) m2 (6) m3 (8) int. j. prod. manag. eng. (2016) 4(2), 75-85creative commons attribution-noncommercial-noderivatives 4.0 international a mathematical programming model for tactical planning with set-up continuity in a two-stage ceramic firm 85 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2017.6807 received 2016-11-03 accepted: 2017-01-16 integrated production-distribution planning optimization models: a review in collaborative networks context beatriz andresai*, raquel sanchisaii, jaques lamotheb, leila saaric and frederic hauserd aresearch centre on production management and engineering (cigip). escuela politécnica superior de alcoy, universitat politècnica de valència (upv). centre d’innovació i investigació. calle alarcón, 03801 alcoy, spain. ai bandres@cigip.upv.es, aii rsanchis@cigip.upv.es buniversity of toulouse – mines albi, industrial engineering centre, campus jarlard, route de teillet. 81013 albi, france. b lamothe@mines-albi.fr cvtt technological research centre of vtt ltd., kaitoväylä 1, fi-90570 oulu, finland. c leila.saari@vtt.fi dlaboratoire pierre fabre dermo cosmetique. les cauquillous 81506 lavaur, france. d frederic.hauser@pierre-fabre.com abstract: researchers in the area of collaborative networks are more and more aware of proposing collaborative approaches to address planning processes, due to the advantages associated when enterprises perform integrated planning models. collaborative production-distribution planning, among the supply network actors, is considered a proper mechanism to support enterprises on dealing with uncertainties and dynamicity associated to the current markets. enterprises, and especially smes, should be able to overcome the continuous changes of the market by increasing their agility. carrying out collaborative planning allows enterprises to enhance their readiness and agility for facing the market turbulences. however, smes have limited access when incorporating optimization tools to deal with collaborative planning, reducing their ability to respond to the competition. the problem to solve is to provide smes affordable solutions to support collaborative planning. in this regard, new optimisation algorithms are required in order to improve the collaboration within the supply network partners. as part of the h2020 cloud collaborative manufacturing networks (c2net) research project, this paper presents a study on integrated production and distribution plans. the main objective of the research is to identify gaps in current optimization models, proposed to address integrated planning, taking into account the requirements and needs of the industry. thus, the needs of the companies belonging to the industrial pilots, defined in the c2net project, are identified; analysing how these needs are covered by the optimization models proposed in the literature, to deal with the integrated production-distribution planning. key words: collaborative planning, supply chain, production planning, distribution planning. 1. introduction the usual way to proceed in many manufacturing enterprises involves dealing with the optimisation of different types of plans separately (e.g. production, and distribution). nevertheless, this approach has associated a set of drawbacks, because it does not consider the limitations of the partners downstream. this limits potential improvement at the enterprises level and discards the total supply chain profit. researchers are more aware of analysing and designing approaches based on the combination of different functions in the supply chain (sc) into integrated optimisation models; according to park (2005) the collaborative planning is significantly relevant for companies that adopted it. the research area of integrating production and distribution int. j. prod. manag. eng. (2017) 5(1), 31-38creative commons attribution-noncommercial-noderivatives 4.0 international 31 https://doi.org/10.4995/ijpme.2017.6807 http://creativecommons.org/licenses/by-nc-nd/4.0/ planning is increasing its impact, due to its decisive contribution in the establishment of collaborative relationships among the enterprises of the same supply chain. in this paper, we consider the distribution planning problem and its evolution towards the collaborative production and distribution planning in supply chain. a literature review is carried out with the main aim of identifying solutions that deal with both plans, including the isolated perspective (in which each plan, production and distribution are separately calculated) and the collaborative perspective, which jointly computes production-distribution plans. moreover, it is analysed how the integrated production-distribution planning approaches provided in the literature deal with the requirements of the real industry. the analysis of industrial planning needs was possible with the pilot partners participating in the h2020 cloud collaborative manufacturing networks (c2net) research project. the goal of c2net project is the creation of cloudenabled tools for supporting the smes supply network optimization of manufacturing and logistic assets based on collaborative demand, production and delivery plans. c2net project will provide a scalable real-time architecture, platform and software to allow the supply network partners: to master complexity and data security of the supply network, to store and share product, process and logistic data, to optimize the manufacturing assets by the collaborative computation of production plans, to optimize the logistics assets through efficient delivery plans and to render the complete set of supply chain management information on the any digital mobile device (pc, tablets, smartphones) of decision makers enabling them to monitor, visualize, control, share and collaborate. the objectives of the project are grouped in three categories: technical, functional and exploitation. at the technical level c2net objectives are focused on the (i) implementation of c2net data collection framework (c2net dcf) for iot-based continuous data collection from supply network resources; (ii) implementation of c2net optimizer (c2net opt) for the optimization of manufacturing and logistics assets of the supply network by the collaborative computation of production, replenishment and delivery plans; (iii) implementation of c2net collaboration tools (c2net cot) for providing support to the collaborative processes of the supply network; and (iv) implementation of c2net cloud platform (c2net cpl) to integrate the data module, the optimizers and the collaborative tools in the cloud. at the functional level, the objectives lead to (i) provide new ways to securely store relevant information from supply network partners in public cloud, community cloud or private cloud, ensuring data security and provide the ability of data sharing and data analytics; (ii) provide new tools for supporting the supply network optimization of manufacturing and logistic assets; (iii) provide new tools to support decision-makers in collaborative processes with intuitive uis that display in real-time the right data at the right time, built in advanced apps for mobile devices for selective management functionalities across the supply network; and (iv) provide a new cloud space for allowing the access to all the participants in the value chain to support their decisions and processes enhancement. finally, the objectives related with the exploitation center its attention in (i) technological exploitation of new ict tools developed by open its integration on existing and future software and its evolution as open source technologies; (ii) commercial exploitation of c2net platform and tools providing the services to companies of any manufacturing sector and licensing the software to consultancy firms; and (iii) know-how exploitation of the new knowledge created during the research and innovation activities developed in the project providing consultancy and training to companies. this paper focuses on the research of optimization of manufacturing and logistics assets of the supply network by the collaborative computation of production and delivery plans. accordingly, this paper presents a study among production and distribution plans with the main aim of identifying gaps in the current optimisation models, taking into account the requirements and needs coming from the specifications defined in the c2net industrrial pilots . specifically, this paper reports the literature review regarding optimisation models for integrated production-distribution planning decisions. a summary of a review of optimization models for integrated production and delivery planning decisions is presented. in this regard, this paper will provide researchers and practitioners a starting point for generating optimization models in the context of integrated production-distribution planning. int. j. prod. manag. eng. (2017) 5(1), 31-38 creative commons attribution-noncommercial-noderivatives 4.0 international andres, b., sanchis, r., lamothe, j., saari, l. and hauser f. 32 http://creativecommons.org/licenses/by-nc-nd/4.0/ to this regard, the paper is organised as follows: in section 2 a brief description of c2net project is provided, focusing on the dermo-cosmetics sector. in section 3 a review of the combined models for production and distribution planning is provided and the gaps between the literature and the industry are identified. finally, section 4 presents the conclusions. 2. dermo-cosmetics pilot in c2net project the current research work carries out a state of the art to collect and analyse some solutions in the context of distribution and production planning. this research is part of the c2net h2020 project (2015-2017) that addresses the potential of cloud technologies providing a collaboration infrastructure for the real-world data of different supply chain components such as manufacturing assets status, inventory levels or current demand at consumption points (lauras et al., 2015). c2net h2020 project is motivated by the fact that currently, the european smes do not have access to advanced management systems and collaborative tools because of their limited resources (european comission, 2005, 2012, 2013, 2014). value chains formed by smes are distributed and dependent on information and complex materials flows that require new approaches to reduce the complexity of manufacturing management systems. in this context, ubiquitous tools are needed to support collaboration between different entities in the value chain and offer advanced algorithms to achieve global and local optimization of manufacturing processes and respond more quickly and efficiently unforeseen changes. the main objective of c2net is building a new architecture in the cloud to provide smes, affordable tools (in terms of cost and ease of use) to help overcome the current economic crisis, improving competitiveness in the economy world. therefore, c2net objective is based on the creation of cloud tools to support optimization of manufacturing networks composed mainly of smes and their logistic assets through demand management, production and supply plans, considering the collaborative network perspective (andres et al., 2016). c2net will provide specific tools for optimisation and collaboration in the cloud, allowing enterprises to be responsive and flexible against changes in both the demand and unexpected events that take place during products production and distribution. moreover, c2net will provide a technological infrastructure that allows collecting real-world data from processes and products within the supply chain. finally c2net project provides collaboration tools with a set of services able to facilitate the decisionmaking processes, increasing the enterprises agility. accordingly, the c2net project will generate a cloud architecture composed by the cloud platform (c2net cpl), the data collection framework (c2net dcf), the optimizer (c2net opt) and the collaboration tools (c2net cot) (see figure 1). the validation of c2net outputs (collaborative platform) will be possible with the four industrial pilots participating in the project. each pilot is devoted to improve different collaborative levels, due to their characteristics. thus, pilots corresponding to the automotive sector and oem manufacturing are devoted to enhance production collaborative processes. the dermo-cosmetics pilot is focused to enhance deliver collaborative processes. the metalworking pilot is devoted to enhance source collaborative processes. this paper focuses on the planning needs of dermo-cosmetics pilot regarding production and distribution. the supply chain under study in contains one production unit, its international warehouse, two subsidiaries’ distribution centres and 20 retailers (pharmacists’ drugstores). in a first stage, the supply chain under study concerns one production unit, his international warehouse, two subsidiaries’ distribution centres and 20 retailers (pharmacists’ drugstores). in a second phase, this pilot propose to test the scalability at a bigger size, the european supply chain composed of 15 distribution subsidiaries and 50000 selling points. enterprises of dermo-cosmetics pilot will use the c2net platform (cloud infrastructure) and software to perform collaborative production and distribution network from first-tier suppliers of materials to final resellers all over the world. the use of c2net optimiser will optimise the availability of products and limiting the inventory levels all along the supply chain. in this pilot the collaboration tools will be used to manage the production and delivery collaborative processes between the manufacturing plant, the subsidiaries, the transportation companies and the retailers. this will particularly take effect on the capability of these stakeholders to better manage the uncertainties such as sudden demand variation or transportation failures. the c2net platform will improve the agility capabilities of the whole network. int. j. prod. manag. eng. (2017) 5(1), 31-38creative commons attribution-noncommercial-noderivatives 4.0 international integrated production-distribution planning optimization models: a review in collaborative networks context 33 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3. integrated optimization models for production-distribution planning the consideration of a collaborative production and distribution planning problem is a determined decision to improve the net profit in a multi-plant, multi-retailer, multi-item, and multi-period logistic environment. different optimisation models and heuristic solutions have been proposed in the literature the to deal with both integrated and decoupled plans of production and distribution. the main problems to face are the high inventory levels required to cover the demand, the long lead times that freeze a lot of assets, the bullwhip effect, the back orders, changes in short-term forecasts, etc. next subsections will analyse the current solutions to of distribution plans and solutions that propose optimisation models, algorithms and simulation approaches to deal with the collaborative production and distribution planning. 3.1. distribution planning distribution planning deals with the plan delivery (d) identified in the scor views (supply chain council, 2012), and focuses on the calculation of the amount of items to be delivered per periods in a planning horizon. the main input data are the customers’ orders, with amounts and locations to deliver the goods. the delivery will use limited resources (own, from the customer or from a third party). distribution plans can be arranged considering four category levels: d/demand planning: plan that enables sales teams to develop demand forecasts as input to the service planning processes, production, inventory planning and revenue planning; d/distribution planning: plan that concerns the movement of a finished product or services to the customers. the customer is the final destination of a selling channel. this plan examines demographic characteristics and travel patterns evaluates alternative improvements for the distribution of finished products. distribution va lu e ch ai n opt dcf plans optimization services cot c2net sharing distribution assests sharing logistics assets sharing product stocks sharing components stocks sharing manufacturing assests sharing logistics assets stocks and manufacturing real time data collection suppliers events procurement plans collaborative decision making production plans collaborative decision making distribution plans collaborative decision making stocks and sales real time data collection suppliers events manufacturing real time data collection manufacturers events automotive smes 2nd tier suppliers use case metalworking sme’s network use case dermo-cosmetics smes retailers use case collaboration services data collection & storage services sharing manufacturing assests sharing logistics assets oem micro-suppliers use case sharing manufacturing assests sharing logistics assets / sharing components stocks figure 1. c2net project results. int. j. prod. manag. eng. (2017) 5(1), 31-38 creative commons attribution-noncommercial-noderivatives 4.0 international andres, b., sanchis, r., lamothe, j., saari, l. and hauser f. 34 http://creativecommons.org/licenses/by-nc-nd/4.0/ planning includes steps like: (i) calculation of distances, times and distribution costs; (ii) optimisation of distribution routes; (iii) analysis of logistic flows; and (iv) design and analysis of a distribution network; d/order promising: plan that matches customer orders against available manufacturing resources and then replies with promised quantities and due dates; d/transport planning: plan that concerns the movement of raw materials or semielaborated products to manufacturing enterprises. plan that deals with procurement transport activities for accomplishing production requirements. table 1 below presents the mapping between dermo-cosmetics industrial pilot optimisation needs and plans described in literature. the first column carries the reference of a plan in literature. for any optimisation case the input data, objectives and output data are crucial and need to match sufficiently. input data was classified into eight classes, objectives five and output data six. the last column identifies the algorithm type presented in the literature reference: ao is an optimisation algorithm, ah is a heuristic algorithm and am is a matheuristic algorithm. each x in the table presents a match between the algorithm in literature and industrial pilot optimisation needs. correspondingly, each empty cell identifies a gap between literature and real case. 3.2. collaborative production and distribution planning make & deliver (md) plans deal with the transformation of a product into a finished good and provide finished goods and services to meet planned or actual demands. dermo-cosmetics pilot allowed defining two collaborative processes regarding md plans: m/production planning & d/distribution planning: plan that deals with production and table 1. matching pilot 2 industrial needs and deliver literature plans. industry needs input data objectives output data algorithm typeid1 id2 id3 id4 id5 id6 id7 id8 o1 o2 o3 o4 o5 od1 od2 od3 od4 od5 od6 martin (1995) x x x x x x x x x ah/ greedy fliedner (2003). x x x x x x x ao/ dynamic programming venkatadri et al. (2006) x x x x x x x x x x ao/ simplex gelper et al. (2010) x x x ao/ dynamic programming fildes et al. (2009) x x x ao/ dynamic programming chan et al. (2006) x x x x x x x x x x x x am/ genetic alg. arntzen et al. (1995) x x x x x x x x x x x x x ao/ simplex wang et al. (2011) x x x x x x x x ao/ branch and bound florez et al. (2015) x x x x x x x x x x x ao/ branch and bound santoso et al. (2005) x x x x x x x x x x x x ah/decomposition and aggregation rim et al. (2014) x x x x x x x x x x am/ genetic alg. torabia and hassini (2008) x x x x ao/ fuzzy programming meijboom and obel (2007) x x x x x x x x ao/ gams id1: demand, id2: inventory, id3: capacity, id4: production time, id5: transport/distribution cost, id6: backorders, id7: lead time of supply, id8: supplier prices. o1: transport cost min, o2: sales max, o3: inventory min, o4: backorders min, o5: service level max. od1: transport cost, od2: backorders, od3: inventory, od4: delivery time, od5: total cost, od6: product to transport. int. j. prod. manag. eng. (2017) 5(1), 31-38creative commons attribution-noncommercial-noderivatives 4.0 international integrated production-distribution planning optimization models: a review in collaborative networks context 35 http://creativecommons.org/licenses/by-nc-nd/4.0/ distribution operations to customers, over the planning horizon. the plan identifies the production and distribution quantities of each item for each period at each site, subject to consider the capacity constraints and distribution lead-time. the total cost includes the costs of production, inventory, stockout, and distribution; m/production planning & d/transport planning: plan that deals with two phases, (i) the manufacturing phase that focuses on the efficient allocation of the production capacity at the various production sites, and (ii) the transportation phase that focuses on procurement transport activities for accomplishing production requirements. table 2 identifies the matches between the dermocosmetics industrial pilot optimization needs and combined collaborative md planning algorithms found in literature. this table follows the same structure as table 1 in section distribution planning. 4. conclusions the research work carried out in this paper identifies the concrete solutions proposed in the literature to support the collaborative process of production and distribution planning. the literature plans are compared with the real requirements collected from the industry. in this regard, the needs of one of the pilots of c2net project, the dermo-cosmetics pilot, are analysed. the comparison allowed identifying some gaps between the literature and the industrial needs regarding collaborative production and distribution planning. the algorithms found in the literature will need massive modifications before they can be applied in the optimisation cases because both the amount and nature of inputs, objectives and outputs have very low correlation with the industry needs. some features of solutions found in literature will be adapted when implementing collaborative table 2. matching pilot 2 industrial needs and integrated production-distribution literature plans. industry needs input data objectives output data algorithm typeid1 id2 id3 id4 id5 id6 id7 id8 o1 o2 o3 o4 o5 od1 od2 od3 od4 od5 od6 park (2005) x x x x x x x x ao/ simplex jung and jeong (2005) x x x x x x x x x x colaborative agents jang et al. (2002) x x x x x x x x x heuristic gupta and maranas (2003) x x x x am/ genetic alg. ekşioğlu et al. (2006) x x x x x x x x ah/ primaldual chen and wang (1997) x x x x x ao/ simplex dhaenens and finke (2001) x x x x ao/ simplex jayaraman and pirkul (2001) x x x x x x ah/lagrangian sakawa et al. (2001) x x x x x ao/ simplex lee and kim (2002) x x x x x ao/ gams selim et al. (2008) x x x x x x ao/ simplex aliev et al. (2007) x x x am/ genetic alg. rizk et al. (2006) x x x x x x ao/ simplex id1: demand, id2: inventory, id3: capacity, id4: production time, id5: transport/distribution cost, id6: backorders, id7: lead time of supply, id8: supplier prices. o1: transport cost min, o2: sales max, o3: inventory min, o4: backorders min, o5: service level max. od1: transport cost, od2: backorders, od3: inventory, od4: delivery time, od5: total cost, od6: product to transport. int. j. prod. manag. eng. (2017) 5(1), 31-38 creative commons attribution-noncommercial-noderivatives 4.0 international andres, b., sanchis, r., lamothe, j., saari, l. and hauser f. 36 http://creativecommons.org/licenses/by-nc-nd/4.0/ optimization algorithms for the requirements defined by the industry. acknowledgements the research leading to these results is in the frame of the “cloud collaborative manufacturing networks” (c2net) project, which has received funding from the european union’s horizon 2020 research and innovation programme under grant agreement no 636909. references aliev, r.a., fazlollahi, b., guirimov, b.g., aliev, r.r. (2007). fuzzy-genetic approach to aggregate production–distribution planning in supply chain management. information sciences, 177(20), 4241-4255. https://doi.org/10.1016/j.ins.2007.04.012 andres, b., sanchis, r., poler, r. (2016). a cloud platform to support collaboration in supply networks. international journal of production management and engineering, 4(1), 5-3. https://doi.org/10.4995/ijpme.2016.4418 arntzen, b.c., brown, g.g., harrison, t.p., trafton, l.l. (1995). global supply chain management at digital equipment corporation. interfaces, 25(1), 69-93. https://doi.org/10.1287/inte.25.1.69 chan, f.t., chung, s.h., choy, k.l. (2006). optimization of order fulfillment in distribution network problems. journal of intelligent manufacturing, 17(3), 307-319. https://doi.org/10.1007/s10845-005-0003-z chen, m., wang, w. (1997). a linear programming model for integrated steel production and distribution planning. international journal of operations & production management, 17(6), 592-610. https://doi.org/10.1108/01443579710167276 dhaenens, c., finke, g. (2001). an integrated model for an industrial production–distribution problem. iie transactions, 33(9), 705-715. https://doi.org/10.1080/07408170108936867 ekşioğlu, s.d., romeijn, h.e., pardalos, p.m. (2006). cross-facility management of production and transportation planning problem. computers & operations research, 33(11), 3231-3251. https://doi.org/10.1016/j.cor.2005.02.038 european commission. (2005). the new sme definition: user guide and model declaration. https://ec.europa.eu/digital-agenda/en/news/ new-sme-definition-user-guide-and-model-declaration, retrieved december 2015. european commission. (2014). an integrated industrial policy for the globalisation era. putting competitiveness and sustainability at centre stage. http://eur-lex.europa.eu/legal-content/en/txt/?uri=uriserv%3aet0005, retrieved december 2015. european commission. enterprise and industry. (2012). european competitiveness report 2012. reaping the benefits of globalization. http://wbc-inco.net/object/document/11137# retrieved december 2015. european commission. enterprise and industry. (2013). fact and figures about the eu's small and medium enterprise (sme). http:// ec.europa.eu/enterprise/index_en.htm, retrieved december 2015. fildes, r., goodwin, p., lawrence, m., nikolopoulos, k. (2009). effective forecasting and judgmental adjustments: an empirical evaluation and strategies for improvement in supply-chain planning. international journal of forecasting, 25(1), 3-23. https://doi.org/10.1016/j. ijforecast.2008.11.010 fliedner, g. (2003). cpfr: an emerging supply chain tool. industrial management & data systems, 103(1), 14-21. https://doi. org/10.1108/02635570310456850 florez, j.v., lauras, m., okongwu, u., dupont, l. (2015). a decision support system for ro-bust humanitarian facility location. engineering applications of artificial intelligence. 46, 326-335. http://dx.doi.org/10.1016/j.engappai.2015.06.020 gelper, s., fried, r., croux, c. (2010). robust forecasting with exponential and holt-winters smoothing. journal of forecasting, 29(3), 285300. https://doi.org/10.1002/for.1125 gupta, a., maranas, c.d. (2003). managing demand uncertainty in supply chain planning. computers & chemical engineering, 27(8), 12191227. https://doi.org/10.1016/s0098-1354(03)00048-6 jang, y.j., jang, s.y., chang, b.m., park, j. (2002). a combined model of network design and production/distribution planning for a supply network. computers & industrial engineering, 43(1), 263-281. https://doi.org/10.1016/s0360-8352(02)00074-8 jayaraman, v., pirkul, h. (2001). planning and coordination of production and distribution facilities for multiple commodities. european journal of operational research, 133(2), 394-408. https://doi.org/10.1016/s0377-2217(00)00033-3 jung, h., jeong, b. (2005). decentralised production-distribution planning system using col-laborative agents in supply chain network. the international journal of advanced manufacturing technology, 25(1-2), 167-173. https://doi.org/10.1007/s00170-003-1792-x lauras, m., lamothe, j., benaben, f., andres, b., poler, r. (2015). towards an agile and collaborative platform for managing supply chain uncertainties. in enterprise interoperability, 64-72. springer berlin heidelberg. https://doi.org/10.1007/978-3-662-47157-9_6 lee, y.h., kim, s.h. (2002). production–distribution planning in supply chain considering capacity constraints. computers & industrial engineering, 43(1), 169-190. https://doi.org/10.1016/s0360-8352(02)00063-3 martin, a.j. (1995). drp: distribution resource planning: the gateway to true quick response and continuous replenishment. john wiley & sons. int. j. prod. manag. eng. (2017) 5(1), 31-38creative commons attribution-noncommercial-noderivatives 4.0 international integrated production-distribution planning optimization models: a review in collaborative networks context 37 https://doi.org/10.1016/j.ins.2007.04.012 https://doi.org/10.4995/ijpme.2016.4418 https://doi.org/10.1287/inte.25.1.69 https://doi.org/10.1007/s10845-005-0003-z https://doi.org/10.1108/01443579710167276 https://doi.org/10.1080/07408170108936867 https://doi.org/10.1016/j.cor.2005.02.038 https://ec.europa.eu/digital-agenda/en/news/new-sme-definition-user-guide-and-model-declaration https://ec.europa.eu/digital-agenda/en/news/new-sme-definition-user-guide-and-model-declaration http://eur-lex.europa.eu/legal-content/en/txt/?uri=uriserv%3aet0005 http://wbc-inco.net/object/document/11137# http://ec.europa.eu/enterprise/index_en.htm http://ec.europa.eu/enterprise/index_en.htm https://doi.org/10.1016/j.ijforecast.2008.11.010 https://doi.org/10.1016/j.ijforecast.2008.11.010 https://doi.org/10.1108/02635570310456850 https://doi.org/10.1108/02635570310456850 http://dx.doi.org/10.1016/j.engappai.2015.06.020 https://doi.org/10.1002/for.1125 https://doi.org/10.1016/s0098-1354(03)00048-6 https://doi.org/10.1016/s0360-8352(02)00074-8 https://doi.org/10.1016/s0377-2217(00)00033-3 https://doi.org/10.1007/s00170-003-1792-x https://doi.org/10.1007/978-3-662-47157-9_6 https://doi.org/10.1016/s0360-8352(02)00063-3 http://creativecommons.org/licenses/by-nc-nd/4.0/ meijboom, b., obel, b. (2007). tactical coordination in a multi-location and multi-stage oper-ations structure: a model and a pharmaceutical company case. omega, 35(3), 258-273. https://doi.org/10.1016/j.omega.2005.06.003 park, y.b. (2005). an integrated approach for production and distribution planning in supply chain management. international journal of production research, 43(6), 1205-1224. https://doi.org/10.1080/00207540412331327718 rim, s.c., jiang, j., lee, c.j. (2014). strategic inventory positioning for mto manufactur-ing using asr lead time. in logistics operations, supply chain management and sustainability, 441-456. springer international publishing. rizk, n., martel, a., d’amours, s. (2006). multi-item dynamic production-distribution planning in process industries with divergent finishing stages. computers & operations research, 33(12), 3600-3623. https://doi.org/10.1016/j.cor.2005.02.047 sakawa, m., nishizaki, i., uemura, y. (2001). fuzzy programming and profit and cost allocation for a production and transportation problem. european journal of operational research, 131(1), 1-15. https://doi.org/10.1016/s0377-2217(00)00104-1 santoso, t., ahmed, s., goetschalckx, m., shapiro, a. (2005). a stochastic programming approach for supply chain network design under uncertainty. european journal of operational research, 167(1), 96-115. https://doi.org/10.1016/j.ejor.2004.01.046 selim, h., araz, c., ozkarahan, i. (2008). collaborative production–distribution planning in supply chain: a fuzzy goal programming approach. transportation research part e: logistics and transportation review, 44(3), 396-419. https://doi.org/10.1016/j.tre.2006.11.001 supply chain council. (2012). supply chain operations reference model (scor), supply chain operations management retrieved from: http://www.apics.org/sites/apics-supply-chain-council/frameworks/scor torabi, s.a., hassini, e. (2008). an interactive possibilistic programming approach for multiple objective supply chain master planning. fuzzy sets and systems, 159(2), 193-214. https://doi.org/10.1016/j.fss.2007.08.010 venkatadri, u., srinivasan, a., montreuil, b., saraswat, a. (2006). optimization-based decision support for order promising in supply chain networks. international journal of production economics, 103(1), 117-130. https://doi.org/10.1016/j.ijpe.2005.05.019 wang, f., lai, x., shi, n. (2011). a multi-objective optimization for green supply chain network design. decision support systems, 51(2), 262-269. https://doi.org/10.1016/j.dss.2010.11.020 int. j. prod. manag. eng. (2017) 5(1), 31-38 creative commons attribution-noncommercial-noderivatives 4.0 international andres, b., sanchis, r., lamothe, j., saari, l. and hauser f. 38 https://doi.org/10.1016/j.omega.2005.06.003 https://doi.org/10.1080/00207540412331327718 https://doi.org/10.1016/j.cor.2005.02.047 https://doi.org/10.1016/s0377-2217(00)00104-1 https://doi.org/10.1016/j.ejor.2004.01.046 https://doi.org/10.1016/j.tre.2006.11.001 http://www.apics.org/sites/apics-supply-chain-council/frameworks/scor https://doi.org/10.1016/j.fss.2007.08.010 https://doi.org/10.1016/j.ijpe.2005.05.019 https://doi.org/10.1016/j.dss.2010.11.020 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2018.7898 received 2017-07-21 accepted: 2017-12-06 what we know and do not know about organizational resilience cristina ruiz-martin a,b*, adolfo lópez-paredes a and gabriel wainer b a grupo insisoc. escuela técnica superior de ingenieros industriales. universidad de valladolid, pº. del cauce s/n, 47011-valladolid, spain b department of systems and computer engineering. carleton university, canada * cruiz@eii.uva.es abstract: we present a literature review about organizational resilience, with the goal of identifying how organizational resilience is conceptualized and assessed. the two research questions that drive the review are: (1) how is organizational resilience conceptualized? and (2) how is organizational resilience assessed? we answer the first question by analyzing organizational resilience definitions and the attributes or characteristics that contribute to develop resilient organizations. we answer the second question by reviewing articles that focus on tools or methods to measure organizational resilience. although there are three different ways to define organizational resilience, we found common ideas in the definitions. we also found that organizational resilience is considered a property, ability or capability that can be improved over time. however, we did not find consensus about the elements that contribute to improving the level of organizational resilience and how to assess it. based on the results of the review, we propose a conceptualization of organizational resilience that integrates the three views found in the literature. we also propose a four-level maturity model for organizational resilience – mmor. using this model, the organization can be in one of the following levels based on its ability and capacity to handle disruptive events: fragile, robust, resilient or antifragile. key words: antifragility, resilience, robustness, organizational resilience, conceptualization, assessment. 1. introduction during the last years, the study of resilience has become more important because people are more aware of the consequences of natural and humanmade disasters (tukamuhabwa, stevenson, busby, zorzini, 2015). some authors think that the study of resilience is gaining importance due to the speed of changes in the economy, society, and technology (e.g. horne iii, 1997). due to this speed of changes, survival is now considered a critical aspect of business, and being resilient is important for such survival. although there is increasing interest in this research area, there is no agreement about where these ideas were first introduced. some (coutu, 2002) say that it was in psychology. others (henry & ramirezmarquez, 2010; annarelli & nonino, 2016) say that the concept was popularized after holling (1973), “resilience and stability of ecological systems”. today, research on resilience is important in many different fields such as management, ecology, psychology, disaster management, organization management, sociology, and engineering. as research in resilience has been attacked in many areas, there is no widely accepted definition, even in the same area (bergström, van winsen, henriqson, 2015). at first sight, one may think that there is no relation among the different research areas on resilience. for instance, one could believe that resilience against disasters is not related to building resilient systems, to cite this article: ruiz-martin, c., lópez-paredes, a., wainer, g. (2018). what we know and do not know about organizational resilience. international journal of production management and engineering, 6(1), 11-28. https://doi.org/10.4995/ijpme.2018.7898 int. j. prod. manag. eng. (2018) 6(1), 11-28creative commons attribution-noncommercial-noderivatives 4.0 international 11 http://creativecommons.org/licenses/by-nc-nd/4.0/ organizations or individuals. however, several authors have already identified relationships between the different fields. for example, we need resilient individuals to build resilient organizations (biggs, hall, stoeckl, 2012; doe, 1994; mallak, 1997). resilient organizations also need resilient supply chains (sheffi, 2007) or resilient infrastructure (bell, 2002; erol, mansouri, sauser, 2009). resilient organizations contribute to creating resilient communities (kendra & wachtendorf, 2003; lee, vargo, seville, 2013) or societies (beermann, 2011). resilient organizations also contribute to developing resilient territory (gilly, kechidi, talbot, 2014). resilience engineering principles contribute to developing resilient organizations (righi, saurin, wachs, 2015). figure 1 represents the relationship among these areas centered on their relation to organizational resilience. organizational resilience influences the resilience research areas painted in grey color, and it is influenced by research areas depicted in white color. organizational resilience is influenced by resilient individuals, resilience engineering, infrastructure resilience, cyber resilience, system resilience, supply chain resilience and business resilience. organizational resilience influences community resilience, societal resilience, economic resilience, city or urban resilience, territory resilience and socio-ecological resilience. figure 1. relations between resilience concepts and organizational resilience. here, we focus on resilience at the organizational level. more specifically, our objective is to identify what we know and what we do not know about the conceptualization and assessment of organizational resilience. two fundamental questions drive the research: (1) how is organizational resilience conceptualized? and (2) how is organizational resilience assessed? at the organizational level, resilience has simultaneously emerged from different fields such us enterprise risk management, business continuity management, emergency management, crisis management, physical security, and cyber-security (braes & brooks, 2010, 2011; gibson & tarrant, 2010). in these fields, researchers and practitioners have studied how to protect the organizations against disruptive events. louisot (2015) considers resilience as a main issue in risk management, and jackson, firtko, and edenborough (2007) view resilience as a new way of thinking about risk. as systems and organizations cannot be designed to anticipate all possible risks (fiksel, 2003), we need resilient organizations to deal with events that will have serious consequences, even when they have low probability of occurrence (ambulkar, blackhurst, grawe, 2015; dalziell & mcmanus, 2004). we also need to study when policies, procedures, practices, and tools fail during an emergency response (kendra & wachtendorf, 2003). although there are several reviews about resilience at the organizational level (annarelli & nonino, 2016; bhamra, burnard, dani, 2015; bhamra, dani, burnard, 2011; linnenluecke, 2017) the research questions we discussed earlier are not yet answered. bhamra et al. (2011) introduced a general review about resilience based on 74 papers published before 2011. they identified five perspectives for resilience studies (ecological, individual, socio-ecological or community, organizational and supply chain). they focused on the conceptualization of resilience based on these perspectives. however, only three of the definitions they presented were in the context of resilience of organizations. bhamra et al. (2015) presented an updated version of (bhamra et al., 2011) including 100 articles and five definitions valid for organizations. annarelli & nonino (2016) investigated the research domains of organizational resilience based on a literature review and co-citation analysis. they aimed to understand the actual state of development of organizational resilience and the future research directions in this area. they also reviewed several definitions of resilience and organizational resilience, and proposed a new one. however, they did not analyze what are the differences in the conceptualization of organizational resilience. linnenluecke (2017) focused on the evolution of organizational resilience theory. she acknowledged that there is no unified theory and proposed several future research questions such as int. j. prod. manag. eng. (2018) 6(1), 11-28 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz-martin, c., lópez-paredes, a. and wainer, g. 12 http://creativecommons.org/licenses/by-nc-nd/4.0/ “what capacities bring about resilience really?” or “how resilience can/should be operationalized?” although these reviews have dealt with many important issues, there are still many open questions regarding the conceptualization and elements that contribute to resilience and how it is assessed. to address these issues, we propose a conceptualization of resilience following suddaby (2010), which shows how to construct clarity in theories of management and organization. we also introduce a maturity model for organizational resilience. finally, we present basic dimensions to assess and measure organizational resilience. the rest of the paper is organized as follows. in section 2, we explain the methodology we followed to select and analyze the papers. in section 3, we review how organizational resilience is conceptualized. in section 4, we review how resilience is assessed at the organizational level. in section 5, we conclude the paper with a discussion about the conceptualization, assessment and future research lines in the study of organizational resilience. we also introduce our conceptualization and maturity model for organization resilience (mmor). 2. research methodology as discussed in the introduction, we aim to identify what do we know and do not know about the conceptualization and assessment of organizational resilience. to do so, we conducted a comprehensive search in the literature, and we conducted a systematic review of those research papers. we followed the research protocol depicted in figure 2, based on (tranfield, denyer, smart, 2003). we first defined the research questions. to do this we conducted a preliminary search for organizational resilience. this search indicated a lack of consensus about how organizational resilience is defined, what are the factors or attributes that contribute to improve the level of organizational resilience and how we can measure it (bhamra et al., 2011; braes & brooks, 2011; smith & fischbacher, 2009). based on this search, we agreed that, in order to understand organizational resilience better, we needed to study what do we know and what do not know about the following questions: (1) how is resilience conceptualized at the organizational level, and, (2) how is resilience assessed in practice at the organizational level. in a second step, we defined the scope of our research. we are interested in the study of resilience in management at the organizational level. therefore, we included the fields of management and business. we also included engineering because our preliminary search indicated that some of the management results come from engineering. we discarded psychology and human resources because the study of resilience in these fields is mainly focused on resilience of individuals. we also discarded finance and economics because they do not specifically focus on the management of the organization. in a third step, we defined the search criteria and selected a database to do the search. we considered four databases: ieee xplore, sciencedirect, scopus and web of science. scopus includes both ieee and sciencedirect papers, and scopus has a larger coverage than web of science, though it does not cover some journals (chadegani et al., 2013). additionally, it provides more search filters. step 1 defining research questions • how is resilience conceptualized at organizational level? • how is resilience assessed in practice at organizational level? step 2 defining boundaries • business and management • engineering step 3 defining search criteria data base selection • scopus search boundaries • business & management busi • engineering engi cover period • unitil january 2016 search terms • (resilient or resilienc*) and (business or organi?ation* or enterprise* or firm* or compan*) and ( system*) • keyword refinement • title and abstract refinement step 4 defining exclusion criteria • replicate the search two times independently step 5 validating search results step 6 analyzing selected papers • content in the selected paper • title • abstract step 7 including criteria for additional papers figure 2. research protocol. int. j. prod. manag. eng. (2018) 6(1), 11-28creative commons attribution-noncommercial-noderivatives 4.0 international what we know and do not know about organizational resilience 13 http://creativecommons.org/licenses/by-nc-nd/4.0/ based on the scope defined in step 2, we searched any document indexed in scopus, written in english, belonging to the subject areas engineering or business, management and accounting. before defining the search term criteria, we performed another preliminary scan of the literature related to organizational resilience. this showed that some authors also talk about resilient organizations, enterprises or firms. therefore, we included these keywords in our search. an organization can be defined as “an organized group of people with a particular purpose, such as business or government department” (oxford university press, 2017). therefore, we decided to include company and business. additionally, some authors view the organization as a system (beer, 1989), so we also included the keyword system. however, system can also refer to other types of systems, such as control systems, computer network systems or mechanical systems among others. since the keyword system has several meanings, especially in engineering, it was only included in management to avoid results not related to our topic. this search returned 3322 results that cover a time-period from january 1970 until january 2016. this result may contain some duplicates results as the search was carried in a twostep process. in a fourth step, we defined the exclusion criteria. we refined our results based on the papers keywords and analyzing their title and abstract. we realized that there were papers that did not include a keyword section (e.g. mallak, 1998b). if we refined the results using the keyword option in scopus, those papers would be discarded. to address this issue, we allowed the keywords to be either in the title or in the keyword section. the preliminary scan of the literature showed that resilience is related to other concepts such as risk, reliability, disasters, redundancy, vulnerability, uncertainty, recovery, prevention, robustness or adaptation. however, we refined the search only using the keyword “*resilien*” (e.g., all the keywords that contain “resilien”), as we wanted to focus on works talking about definitions, characteristics, and measures of organizational resilience. after the keyword refinement, we obtained 1352 papers. we refined again these 1352 papers using their title and abstract. we excluded all the papers whose title or abstract was not related to organizational resilience and management. we included all the papers that did not specify the type of system that was studied, and discarded those that focused on specific systems not related to management or organizations. when the title and abstract did not help to decide if the paper should be included or not, we took into consideration the journal or conference title. in other cases, we read the full paper. in the end, 143 papers were selected. we noticed that many papers displayed in our search, despite having the keyword “*resilien*” were not related to the topic. therefore, we discarded them. in a fifth step, we replicated the search and the refinement two times to validate our results. we considered that reading the title and abstract once was not enough to obtain reliable results. finally, we analyzed all the papers. during this step, we realized that some cited relevant papers did not appear in our search because of the content of the title, abstract and keywords. to address this issue, we reviewed the references of those 143 papers to find relevant literature that our search did not display. we included those papers whose title, abstract or the content referenced in the paper was related to the definitions, characteristics or measurement of organizational resilience. we included 48 additional papers. the final number of papers was 191 (143 papers displayed by the search and 48 additional papers based on the references). this core of papers widely represents the works that address the conceptualization and assessment of organizational resilience. 3. how is resilience conceptualized at the organizational level? suddaby (2010) stated that a clear conceptualization of any construct has four elements: (1) a good definition, (2) scope conditions or contextual circumstances, (3) semantic relations with other concepts, and (4) coherence and logical consistency. he defined three properties for good definitions: (1) capture the essential properties or characteristics of the concept, (2) avoid tautology, and (3) be parsimonious (i.e. focus narrowly on the meaning but being relevant). in order to conceptualize resilience at the organizational level, called organizational resilience we reviewed over 50 definitions. bhamra et al. (2015) presented 18, and only 5 within the organizational context. they indicated that, although there seems to be a common core understanding of what organizational resilience is, there are issues to be discussed. we also found a lack of a clear conceptualization of the term organizational resilience. int. j. prod. manag. eng. (2018) 6(1), 11-28 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz-martin, c., lópez-paredes, a. and wainer, g. 14 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.1. a review about how organizational resilience is understood there are three main streams in the conceptualization of resilience: (1) resilience as a feature of an organization (i.e., something that an organization has), (2) resilience as an outcome of the organization’s activities (i.e., something that an organization does); and (3) resilience as a measure of the disturbances that an organization can tolerate. we found that all of them have the same basic meaning: they have an emphasis either on the organization survival, or in dealing with jolts, risks or changes. however, there is no consensus about the following issues: (1) if the risks are only related to threats or also to opportunities, (2) what survival means, (3) if the risks are already known by the organization or not, and (4) if resilience is always a desirable property. we discuss these points in section 3.1.2. 3.1.1. organizational resilience conceptualization resilience as a feature of an organization most authors (including acquaah, amoakogyampah, jayaram, 2011; alblas & jayaram, 2015; ates & bititci, 2011; bauernhansl, mandel, diermann, 2012; bhamidipaty, lotlikar, banavar, 2007; chand & loosemore, 2016; danes et al., 2009; demmer, vickery, calantone, 2011; freeman, hirschhorn, triad, 2003; t o grøtan & asbjørnslett, 2007; hamel & valikangas, 2003; hollnagel, 2010; horne iii, 1997; yao hu, li, holloway, 2008; jaaron & backhouse, 2014; s. jackson, 2007; lengnickhall & beck, 2005; lengnick-hall, beck, lengnickhall, 2011; mafabi, munene, ahiauzu, 2015; mallak, 1997, 1998a; milanzi & weeks, 2014; sheffi & rice jr., 2005; starr, newfrock, delurey, 2003; tillement, cholez, reverdy, 2009; tse, couturier, roux, 2012; winston, 2014), understand organizational resilience as an ability to deal with internal and external changes, risks or jolts. others, defines it as a capacity to deal with them (alexiou, 2014; dewald & bowen, 2010; fiksel, 2006; gilly et al., 2014; linnenluecke, griffiths, winn, 2012; manyena, 2006; ortiz-demandojana & bansal, 2015; powley, 2009; proper & pienaar, 2011; stewart & o’donnell, 2007; tierney, 2003). finally, some others define it as a capability to deal with these issues (annarelli & nonino, 2016; bell, 2002; kamalahmadi & parast, 2016; reinmoeller & van baardwijk, 2005; robb, 2000; zhang & van luttervelt, 2011). ability, capacity, and capability have different connotations; however, the authors of these papers do not clarify why they choose one term or the other. likewise, they do not define ability, capability or capacity. these three words are sometimes used as synonyms to refer to the power to perform an action or a task. therefore, we will assume that these terms are interchangeable in the definitions. without specifying why they use ability, capacity or capability, some authors combine these terms with a specific adjective to define resilience. for instance, manyena (2006) and hollnagel (2010) consider resilience as something intrinsic to the organization. powley (2009) defines resilience as a latent capacity. gilly et al. (2014) state that the resilience of an organization is both an active and a reactive capacity. resilience can be also considered something dynamic (alexiou, 2014; kamalahmadi & parast, 2016) or incremental (ortizde-mandojana & bansal, 2015). a group of authors (burnard & bhamra, 2011; hilton, wright, kiparoglou, 2012) considers resilience as an emergent property that the organization exhibits when it encounters setbacks. other authors consider resilience as a process to recover from a disruption (van breda, 2016). horne iii & orr (1998) understand resilience as a quality to respond to significant change. some authors (erol et al., 2009; gunasekaran, rai, griffin, 2011; mcmanus, seville, vargo, brunsdon, 2008) define organizational resilience as a function of specific capabilities or abilities. for instance, mcmanus et al. (2008) define resilience as a function of three abilities or capabilities: adaptive capacity, situation awareness, and management of keystone vulnerabilities. erol et al. (2009) include enterprise flexibility, adaptability, agility, and efficiency as attributes for enterprise resilience. gunasekaran et al. (2011) include adaptability, responsiveness, sustainability, and competitiveness. the essence of these capabilities is the same: dealing with change, environmental jolts or risks. defining resilience as a function of characteristics indicates that resilience is a complex concept. some authors also use different dimensions of resilience in their definitions. for example, välikangas & romme (2012) distinguish two dimensions of resilience: operational and strategic. operational resilience is understood as a way to bounce back after a crisis. strategic resilience not only means to bounce back but also to turn threats int. j. prod. manag. eng. (2018) 6(1), 11-28creative commons attribution-noncommercial-noderivatives 4.0 international what we know and do not know about organizational resilience 15 http://creativecommons.org/licenses/by-nc-nd/4.0/ into opportunities. lengnick-hall & beck (2005) identify three components of resilience: cognitive, behavioral and contextual. each component provides the organization with some capabilities. cognitive resilience enables the organization to noticing, interpreting, analyzing and formulating responses that go beyond survival. the elements that contribute to building cognitive resilience include having a constructive sense-making and a strong ideological identity. constructive sense is related to seeking information and its interpretation to take actions. ideological identity refers to having value-driven behavior. behavioral resilience is the engine that makes the organization works. the elements that contribute to building behavioral resilience are a complex and varied action inventory and the functional habits or routines. complex and varied action inventory means having a number of alternatives to deal with different scenarios. functional habits or routines help to open new communication channels or seeking resources automatically. contextual resilience is the framework where cognitive and behavioral resilience takes places. the characteristics that are used to build contextual resilience are social capital and broad resource network. resilience as an outcome of an organization several authors define resilience with focus on what a resilient organization does. for instance, resilience is defined as “the maintenance of positive adjustment under challenging conditions such that the organization emerges from those conditions strengthened and more resourceful” (sutcliffe & vogus, 2003; vogus & sutcliffe, 2007). a resilient organization can return to its performance level at any key performance metric (sheffi, 2007). it is able to achieve its objectives and realized opportunities in the face of predicted or unpredicted disruptive events (hilton et al., 2012; whitehorn, 2010; wright, kiparoglou, williams, hilton, 2012). resilience as a measure of the disturbance that an organization can tolerate resilience has also been defined as a magnitude. under this view, resilience is the level of disturbance an organization can tolerate and still survive (linnenluecke & griffiths, 2010; mamouni limnios, mazzarol, ghadouani, schilizzi, 2014) 3.1.2. open issues in resilience conceptualization many authors consider resilience as a property related to events that may have a negative impact on the organizations. for example, resilience can be related to surviving or adapting to: disruptions (bell, 2002; horne iii & orr, 1998; yao hu et al., 2008; lengnick-hall et al., 2011; sheffi & rice jr., 2005), disasters or catastrophic events (tierney, 2003); (alblas & jayaram, 2015); challenging conditions (sutcliffe & vogus, 2003; vogus & sutcliffe, 2007); disturbances (hollnagel, 2010; linnenluecke & griffiths, 2010; mamouni limnios et al., 2014; tillement et al., 2009); threats (bhamidipaty et al., 2007; dewald & bowen, 2010) or changes (fiksel, 2006; t o grøtan & asbjørnslett, 2007; mafabi et al., 2015; milanzi & weeks, 2014; stewart & o’donnell, 2007). however, some authors consider that these changes can also be opportunities (ates & bititci, 2011; bhamidipaty et al., 2007; dewald & bowen, 2010), and resilient organizations take advantage of these opportunities. regarding the discussion about what is the meaning of “surviving” in the context of resilience, some conceptualizations state that an organization is resilient if it bounces back to a prior point of stability (freeman et al., 2003; sheffi, 2007). others acknowledge that an organization is resilient if it returns to the same point or if it achieves another state of stability (i.e., it changes, while minimizing the effects due to changes and hazards) (acquaah et al., 2011; burnard & bhamra, 2011; demmer et al., 2011). some authors consider that a resilient organization can also bounce forward, grow or become stronger (bell, 2002; fiksel, 2006; vogus & sutcliffe, 2007). woods (2015) identifies four meanings of resilience that bring four interpretations of “surviving”. these four streams are using resilience as rebound (i.e. returning to previous or normal activities after a disruption), robustness (i.e. absorbing disturbances), graceful extensibility (i.e. how to extend adaptive capacity in the face of disruptions) and sustaining adaptability (i.e. the ability to adapt to future disruptions as the conditions change and evolve). these four meanings can be understood as different forms of survival. many authors do not define the type of disruptions that resilient organizations are prepared to deal with. others state that the disruption or change is turbulent (ates & bititci, 2011; bauernhansl et al., 2012; burnard & bhamra, 2011; fiksel, 2006). others int. j. prod. manag. eng. (2018) 6(1), 11-28 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz-martin, c., lópez-paredes, a. and wainer, g. 16 http://creativecommons.org/licenses/by-nc-nd/4.0/ consider that resilience refers to both expected and unexpected events (hilton et al., 2012; hollnagel, 2010; wright et al., 2012). most of the authors consider resilience as a desirable ability or capability for the organizations. although this is not specifically stated in the definitions, it can be inferred from their work. however, a few authors consider that resilience is not always desirable, depending on the state of the system or organization (mamouni limnios, 2011; mamouni limnios et al., 2014). for example, in a cournot duopoly, after an increase in the production cost for both firms in the same amount, companies are not willing to exhibit resilience (understood as bouncing back to the previous state of cost) (lambertini & marattin, 2016). the reason is that the new equilibrium in the market may satisfy both companies and they will not be willing to invest money to return to the previous level of costs. 3.2. resilience and related concepts: fragile, robust and antifragile to clarify the divergences we presented in section 3.1.2, we need to analyze the concepts related to resilience. resilience is related to fragility, robustness, and antifragility. the concept of fragility is related to how a system is broken or damaged in the case of variations (taleb, 2012; taleb & douady, 2013). robustness is the capacity of a system to absorb disturbances (woods, 2015). antifragility is a new concept introduced by taleb (2012), which is defined as the property of a system that, when facing challenges such as failures or volatility, it improves. he differentiates fragile, robust/resilient and antifragile entities, although he uses indistinctly the words resilient and robust. woods (2015) pays attention to the difference between robustness and resilience. being different, using them indistinctly creates confusion when studying resilience. a robust organization absorbs disturbances, but it does not necessarily recover in case of disruptions. read (2005) provides an illustrative example comparing trees. in case of wind, both a palm and a sycamore tree moves from their equilibrium position. when both trees are exposed to the same wind intensity, the sycamore tree movements are much smaller than the palm tree. therefore, it is more robust. however, the palm tree is more resilient as it is able to recover easier from bigger disturbances (i.e., the sycamore tree will probably break). by focusing on the type of disruption that the resilient organizations are prepared to face, these organizations should be able to survive to both known and unknown disturbances. a robust organization is designed to cope and absorb a set of known disturbances. therefore, a resilient organization is more prepared to survive than a robust one. following this view of resilience, we consider it as a desirable property in any organization although in section 3.1.2., we show an example (cournot duopoly) where an organization is not willing to exhibit resilience. being resilient is not only related to bouncing back to the same previous point of stability; being resilient is also achieving another desirable point of stability. if this new point is better than the previous one, and the organization is stronger, we consider that this organization is not only resilient but also antifragile. the distinction between resilient and antifragile organizations clarifies the open question about if resilient organization responds just to threats or also to opportunities. if the organization is able to recover or survive to threats, it is resilient. if this same organization takes advantage of the threats and opportunities to become stronger, it is resilient and antifragile. 3.3. organizational resilience and its attributes as discussed in section 3.1, resilience is a complex and dynamic concept. complex concepts are characterized by different elements or attributes (suddaby, 2010). to identify these elements, we analyzed over 110 works that address different models and frameworks proposed to build or improve organizational resilience. this review revealed that there is a great variety regarding the factors and mechanisms that contribute to resilience. sometimes, the authors refer to the same concept with different words. for instance, improvisation (coutu, 2002; kendra & wachtendorf, 2002), creativity and innovation (dervitsiotis, 2004) are used to refer to bricolage skills; face down reality (coutu 2002) is used to refer to situation awareness. despite the different terminology, we also found some common and repeated characteristics or factors that contribute to enhance resilience. the most cited attributes or elements of a resilient organization include building situation awareness (afgan, 2010; braes & brooks, 2010; coutu, 2002; int. j. prod. manag. eng. (2018) 6(1), 11-28creative commons attribution-noncommercial-noderivatives 4.0 international what we know and do not know about organizational resilience 17 http://creativecommons.org/licenses/by-nc-nd/4.0/ mcmanus et al., 2008), managing organization’s vulnerabilities (erol, sauser, mansouri, 2010; mcmanus et al., 2008; whitehorn, 2010), having resources (aleksic, stefanović, arsovski, tadić, 2013; ates & bititci, 2011; brewton, danes, stafford, haynes, 2010; crichton, ramsay, kelly, 2009; kendra & wachtendorf, 2002; mallak, 1998a; orchiston, prayag, brown, 2016), improvisation capacity (coutu, 2002; tor olav grøtan, størseth, rø, skjerve, 2008; kendra & wachtendorf, 2002; mallak, 1997; rerup, 2001; weick, 1993), ability to anticipate events (apneseth, wahl, hollnagel, 2013; berman, 2009; hardy, 2014; rerup, 2001; wright et al., 2012), agility (gibson & tarrant, 2010; ismail, poolton, sharifi, 2011; megele, 2014; starr et al., 2003; thomas, byard, francis, fisher, white, 2016), learning capacity (aguirre, dynes, kendra, connell, 2005; burnard & bhamra, 2011; hilton et al., 2012; robb, 2000; zhang & van luttervelt, 2011), collaboration (alonso & bressan, 2015; boza & poler, 2013; proper & pienaar, 2011; winston, 2014), resiliency of individuals (doe, 1994; mallak, 1997; riolli & savicki, 2003), flexibility (berman, 2009; kendra & wachtendorf, 2002; megele, 2014; pal, torstensson, mattila, 2014; proper & pienaar, 2011), robustness (heinicke, 2014; s. jackson, 2007; kendra & wachtendorf, 2002; pal et al., 2014; tierney, 2003; tompkins, 2007) and redundancy (chopra & khanna, 2014; yao hu et al., 2008; johnsen & veen, 2012; powley, 2009; tierney, 2003; winston, 2014). it is necessary to remark that other suggested elements may also be important attributes for organizational resilience. a resilient organization includes a mix of several capabilities and actions to be performed. it is this mix what makes an organization resilient (gibson & tarrant, 2010). 4. how is resilience assessed in practice? in this section, we focus on the assessment of organizational resilience, and we study how it can be measured in practice. with this purpose, we reviewed over 30 works that propose tools or methods to assess organizational resilience. the number of articles reviewed is fewer than those reviewed for organizational resilience conceptualization because there are fewer works in the literature in this area. the review of these works indicates a lack of consensus about how to measure organizational resilience. we can classify these works in the same three streams discussed in section 3: those assessed using the features of the organization, those assessed on the organizational outcomes, and those based on how the organization recovers from failure. 4.1. assessment using the features of the organization we classified the works in this area based on how the problem is assessed: using indicators, or other techniques (such as fuzzy cognitive maps or assessment of organizational processes). 4.1.1. measuring organizational resilience based on indicators mcmanus et al. (2007) and seville (2009) suggested 23 indicators followed by a description to evaluate four factors (situation awareness, management of keystone vulnerabilities, resilience ethos and adaptive capacity) that contribute to enhancing resilience. whitehorn (2010) defined a subset of 15 indicators among the previous ones. he did not consider resilience ethos. the indicators proposed by lee et al. (2013) are a subset of the ones suggested by mcmanus et al. (2007) and seville (2009). they suggest to evaluate each factor using several items (see lee et al., 2013 for the list of items). they tested the model proposed by mcmanus et al. (2007), and found that using their sample and scale, the three factors model was not supported. they defined a new version with four factors and they suggested evaluating the factors through 73 items. however, their sample data and scale did not support this new model. they finally suggested a model with two factors (adaptive capacity and planning), 13 indicators and 53 items to be evaluated. whitman et al. (2013) proposed a shorter version of the assessment tool by lee et al. (2013). they suggested using just 13 items (one per indicator). they justified this short version based on two reasons: the low rate response they got while measuring resilience with a long questionnaire, and the correlation between the two assessment tools. the indicators proposed by lee et al. (2013) included some of the characteristics for resilient organizations presented in sections 3.4, such as innovation and creativity (which matches improvisation capacity), collaboration (which matches with partnerships) or situation monitoring and reporting (which matches with situation awareness and the ability to anticipate events). int. j. prod. manag. eng. (2018) 6(1), 11-28 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz-martin, c., lópez-paredes, a. and wainer, g. 18 http://creativecommons.org/licenses/by-nc-nd/4.0/ other authors suggested different factors to assess resilience. for example, starr et al. (2003) suggested using eight points: (1) organization transparency, (2) understanding of risk interdependencies, (3) development of viability studies in the organization, (4) alignment between the strategy in the organization and the objectives, (5) organizational knowledge about the efforts on resilience, (6) situation awareness, (7) how the organization uses situation awareness to react in a timely manner, and (8) measures to evaluate resilience and the progress of the organization. however, they did not define a scale for these eight points. tompkins (2007) suggested using robustness, responsiveness, resourcefulness, rapidity, and redundancy (the five r’s) to evaluate resilience. however, the items to be evaluated in each category were not discussed. sanchis & poler (2013) proposed to measure resilience based on the vulnerability of the organization, its adaptive capacity and recovery ability. kohno et al. (2012) suggested evaluating resilience by taking into account the areas where the organization’s facilities are located, the infrastructure the organization needs, the organization facilities and the supply chains. apneseth et al. (2013) proposed to assess organizational resilience based on how good the organization is at monitoring, responding, anticipating and learning. somers (2009) suggested to measure the organizational resilience potential based on the six organizational resilience attributes defined in mallak (1998a). these factors can be organized in three levels, and the overall resilience of the organization is evaluated from 1 (low) to 7 (high). hollnagel (2010) proposed to assess resilience based on the ability of the organization to respond, monitor, anticipate and learn. van trijp et al. (2012a); van trijp et al. (2012b) suggested to evaluate resilience as a function of four factors: situation awareness, management of keystone vulnerabilities, adaptive capacity and quality. to evaluate these factors, they defined a performance measure based on the attributes they depend on. for example, to measure situation awareness, they evaluate: level of awareness about expectations, obligations and limitations; ability to look forward opportunities and potential crises; level of awareness about resource availability; ability to identify the crises and their consequences; the level of comprehension about the factors that trigger a crisis, and the level of comprehension about the minimum operating requirements for recovery. rigaud et al. (2013) proposed evaluating resilience based on the capacity of the organization to (1) respond, (2) monitor short-term developments and threats, (3) anticipate long-term threats and opportunities and (4) learn from past events. they suggested several indicators for each one. however, they did not describe the indicators proposed. other authors focus on specific sectors. for instance, danes et al. (2009) determined resilience in family firms by evaluating the following items: (1) role clarity, (2) who has the decision authority, (3) ownership equality, (4) fairness of compensation, (5) failure to resolve firm conflicts, (6) unfair workloads and (7) competition for resources between the family and firm. wicker et al. (2013) developed an organizational resilience scale to measure resilience in sports clubs. they develop items (ranged from 1 to 5) to evaluate each factor of resilience defined by bruneau et al. (2003): robustness, redundancy, resourcefulness, and rapidity. for example, to measure rapidity, they evaluate the capability of the organization to achieve goals in a timely manner, adapt quickly to changing circumstances, meet priorities in a timely manner, restore services quickly during unexpected events and respond quickly to disruptive events. 4.1.2. measuring organizational resilience based on other techniques other authors use fuzzy cognitive maps (fcm) and fuzzy sets to analyze these characteristics. for example, grande & trucco (2008) suggested analyzing the resilience of an organization using fuzzy cognitive maps (fcm) to capture the relations between the variables that contribute to resilience. a fcm is a graph with edges and nodes representing causal relations between concepts. to study a civil defense system, they suggested evaluating 17 variables and their relations. asgary et al. (2009) developed a fuzzy-jess expert system based on 17 variables and a set of rules that takes into account these variables to determine the level of resilience in the business. the variables include the existence of a strategic plan, the existence of a business continuity committee or a number of potential hazards among others. aleksić et al. (2013) suggested assessing the organization resilience potential of smes using fuzzy sets, and evaluating the contributing factors for each business process. the importance of each factor in the process is weighted to calculate the resilience of the process. then, the importance of int. j. prod. manag. eng. (2018) 6(1), 11-28creative commons attribution-noncommercial-noderivatives 4.0 international what we know and do not know about organizational resilience 19 http://creativecommons.org/licenses/by-nc-nd/4.0/ each process in the organization is also weighted to measure the overall organizational resilience. they proposed to evaluate internal factors (planning strategies, capability and capacity of internal resources, internal situation monitoring and reporting, human factors and quality), external factors (external situation monitoring and reporting and capability and capacity of external resources) and enabling resilience factors (design of the organization, detection potential, emergency response and safety management system). tadić et al. (2014); macuzić et al. (2016) suggested to evaluate resilience using a fuzzy approach. they suggest the following steps: (1) creating an organizational reference model and to identify the factors that contribute to resilience, (2) weighing the importance of these factors and processes using a fuzzy approach, (3) determining linguistic expressions to evaluate these factors, (4) calculating the resilience factors’ values using a fuzzy approach and (5) ranking the organizational resilience factors. other authors follow other approaches to assess the organizational characteristics. for example, hu et al. (2008); hu et al. (2009); hu et al. (2010) proposed to solve an optimization problem in a network model of the enterprise to determine the effect of a disruption and the resilience of the enterprise. the objective is to understand the balance between operational redundancy and inventory redundancy to achieve resilience. they do not provide items to be analyzed to evaluate resilience. caralli et al. (2010) proposed the cert® resilience management model to assess resilience. it defines 26 process areas with specific goals and practices. these areas include asset definition and management, resilience requirement development, risk management, people management or monitoring. the position of the organization in these processes can be used as a benchmark for identifying organizational capability for managing operational resilience. 4.2. measurement based on the organizational outcomes this stream is less popular, as fewer authors use this approach. for example, watanabe et al. (2004) proposed to use the operating income to sales to measure resilience. dalziell & mcmanus (2004) suggested to measure resilience based on key performance index (kpis) defined taking into account the organization’s objectives. however, these authors did not state the items, attributes, components or kpis to be measured. afgan (2010) proposed an index to measure resilience based on the change of company profit, the change of total company income, the change of product cost and the change of manpower (i.e. human resources availability). markman & venzin (2014) suggested to measure resilience based on the return on equity (roe) and volatility. jackson (2007) suggested to measure resilience potential based on the statistical correlation between minor and major incidents. he found that minor accidents are positively correlated to major accidents. 4.3. measurement based on the organizational recovery in this case, the authors measure resilience based on how the organization recovers from failure. the drawback is that the organization needs to suffer failures to assess its resilience. therefore, this way to measure resilience is only valid after the organization has suffered some shocks. there are two main ways to measure resilience following this approach. henry & ramirez-marquez (2010) suggest to measure resilience quantitatively as the ratio of recovery and loss. here, loss is the deterioration from the original state after the disruption and recovery is the amount it bounces back from the disruptive state to the recovered state. the authors acknowledge that the limitation is to not to consider the money and time to recover. they do not consider what we should evaluate to measure loss and recovery. erol, henry & sauser (2010); erol, henry, sauser, et al. (2010) proposed to measure resilience based on recovery time, level of recovery, initial vulnerability and potential loss averted. however, they do not indicate how to assess these items 5. discussion and conclusions following the guides in suddaby (2010) to construct clarity in theories of management and organization, this section introduces a conceptualization of resilience. we also present a maturity model for organizational resilience (mmor). finally, we present basic dimensions to measure organizational resilience. int. j. prod. manag. eng. (2018) 6(1), 11-28 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz-martin, c., lópez-paredes, a. and wainer, g. 20 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5.1. organizational resilience conceptualization from our analysis on how organizational resilience is understood presented in section 4, we can see that, at present, there is no clear conceptualization of organizational resilience. following suddaby (2010), a clear conceptualization should have a good definition, scope conditions or contextual circumstances, semantic relations with other concepts, and coherence and logical consistency. we propose a conceptualization that integrates the three views about resilience presented in section 4. resilience, at the organizational level, is the measurable combination of characteristics, abilities, capacities or capabilities that allows an organization to withstand known and unknown disturbances and still survive. resilience is not a static concept. the degree of resilience an organization evolves over time. an organization evolves from fragile to antifragile, and resilience is a middle estate in this evolution (taleb, 2012). focusing on how well the organization has developed its abilities to survive in changing or turbulent environments, we suggest a four-level maturity model for organizational resilience – mmor(figure 3). the organization can be at any of the following levels: fragile, robust, resilient and antifragile. fragilefragile antifragileantifragile resilientresilient robustrobust time figure 3. four-level maturity model for organizational resilience (mmor). the organization evolves from one level to another over time based on the improvement on its abilities, characteristics or capabilities to deal with disturbances. a fragile organization is not able to withstand changing environments: it collapses. a robust organization is able to survive to some set of changes in the environment. however, if these changes are outside the designed parameters, the organization will probably collapse. a resilient organization is not only robust, but it is also able to survive to unforeseen events. an antifragile organization is able to not only to survive, but also to prosper or thrive in turbulent environments. regarding the attributes, elements or characteristics for resilience, we propose the ones presented in section 3.3 as an initial combination: building situation awareness, managing organization’s vulnerabilities, having resources, improvisation capacity, ability to anticipate events, agility, learning capacity, collaboration, resilient individuals, flexibility, robustness and redundancy. future research lines in this area should aim to develop a framework to understand how good the organization is dealing with turbulent environments. to be consistent with the previous literature, the organization should be classified in one of the following four levels: fragile, robust, resilient or antifragile. the characteristics and attributes the organization has in each of these levels should be identified. as we proposed in our definition of organizational resilience, these attributes must be measurable to be able to assess resilience and provide an estimate of the resilience potential. these works should combine theoretical and empirical work to test the results. 5.2. discussion about organizational resilience assessment after reviewing how to measure organizational resilience, we consider that two main dimensions to evaluate organizational resilience should coexist. the first one should aim to provide an estimate of organizational resilience potential (i.e. evaluate resilience before a disruptive event occurs). the second one should aim to evaluate the level of resilience an organization has exhibited after a disruptive event has occurred. to provide an estimate of organizational resilience potential, we suggest assessing organizational resilience based on the features of the organizational (section 4.1). the organizational elements, attributes or characteristics to be evaluated should include, at least, the ones presented in section 5.1. to measure the level of resilience an organization has exhibit after a disruptive event, we recommend assessing organizational resilience based on how the organization recovers from failure. we propose int. j. prod. manag. eng. (2018) 6(1), 11-28creative commons attribution-noncommercial-noderivatives 4.0 international what we know and do not know about organizational resilience 21 http://creativecommons.org/licenses/by-nc-nd/4.0/ evaluating a recovery ratio that measures the organizational loses against the recovery and the recovery time. the recovery ratio should include both organizational capabilities and organizational performance. measuring resilience after a disruptive event has occurred will help to provide better estimates of the resilience potential studying the correction between the two measures. we want to clarify that our aim is to provide the basis to build an organizational resilience measurement scale. we are no focused on providing the measurement scale. we suggest that future research directions in this area should aim to identify a common framework that includes indicators and a scale to measure not only organizational resilience potential but also the maturity level the organization has to deal with turbulent environments. although there are some works in this research direction, the works that measure resilience do not take into account the antifragility concept and vice versa. for example, there are some criteria to measure antifragility (johnson & gheorghe, 2013; kennon, schutte, lutters, 2015). however, the link to resilience and the other levels has not been analyzed yet. a scale to measure the level of resilience after a disruptive event has occurred should also be provided. additionally, we also suggest that future research lines should aim to identify paths to develop resilient and antifragile organizations. following this research line, several works have started to address the issue. for example, ruiz-martin et al. (2017) proposed to apply the viable system model to design resilient organizations. ruiz-martin et al. (2015) remarked the importance of communication systems in organizational resilience and suggested evaluating the communication system using network theory. tolk & johnson iv (2013) proposed to develop antifragile systems focusing on their components and interactions and jones (2014) stated that a change in the way we design systems is needed. tseitlin (2013) explained how antifragility is developed in a real organization: netflix. this research area has just started its development and more theoretical and empirical research is needed. 5.3. further discussion about organizational resilience as discussed in the introduction, the study of resilience covers different related areas. future research directions should aim to identify these relations. some questions to be answered are: (1) what is the lowest level of resilience? is it having resilient individuals? (2) what kinds of resilience (i.e. infrastructure resilience, resilient individual, etc.) affect organizational resilience and how? (3) what kind of resilience (i.e. community resilience, city resilience, and so on) are influenced by organizational resilience? and (4) how all these areas of resilience are integrated to develop a more resilient world? in section 5.1, we introduced the mmor model. future research direction should also aim to investigate if this concept and the four-level mmor to develop antifragile organizations (i.e. from fragile to antifragile organizations) are also applicable to the other concepts such as infrastructures, individuals, communities or territories. acknowledgements this research is partially supported by university of valladolid, banco santander and nserc references acquaah, m., amoako-gyampah, k., jayaram, j. (2011). resilience in family and nonfamily firms: an examination of the relationships between manufacturing strategy, competitive strategy and firm performance. international journal of production research, 49(18), 5527-5544. https://doi.org/10.1080/00207543.2011.563834 afgan, n. h. (2010). resilience of company management system. in picmet ’10 portland international center for management of engineering and technology, proceedings technology management for global economic growth (pp. 1–8). aguirre, b. e., dynes, r. ., kendra, j., connell, r. (2005). institutional resiliance and disaster planning for new hazards: insights from hospitals. journal of homeland security and emergency management, 2(2), 1547-7355. https://doi.org/10.2202/1547-7355.1113 alblas, a., jayaram, j. (2015). design resilience in the fuzzy front end (ffe) context: an empirical examination. international journal of production research, 53(22), 6820–6838. https://doi.org/10.1080/00207543.2014.899718 int. j. prod. manag. eng. (2018) 6(1), 11-28 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz-martin, c., lópez-paredes, a. and wainer, g. 22 https://doi.org/10.1080/00207543.2011.563834 https://doi.org/10.2202/1547-7355.1113 https://doi.org/10.1080/00207543.2014.899718 http://creativecommons.org/licenses/by-nc-nd/4.0/ aleksic, a., stefanović, m., arsovski, s., tadić, d. (2013). an assessment of organizational resilience potential in smes of the process industry, a fuzzy approach. journal of loss prevention in the process industries, 26(6), 1238–1245. https://doi.org/10.1016/j. jlp.2013.06.004 alexiou, a. (2014). taming the waves of adversity: exploring the multidimensional construct of organizational resilience. in assimakopoulos, d. g., oshri, i., pandza, k. (eds.). managing emerging technologies for socio-economic impact. cheltenham, uk: edward elgar publishing. https://doi.org/10.4337/9781782547884.00026 alonso, a. d., bressan, a. (2015). resilience in the context of italian micro and small wineries: an empirical study. international journal of wine business research, 27(1), 40–60. https://doi.org/10.1108/ijwbr-08-2014-0035 ambulkar, s., blackhurst, j., grawe, s. (2015). firm’s resilience to supply chain disruptions: scale development and empirical examination. journal of operations management, 33–34, 111–122. https://doi.org/10.1016/j.jom.2014.11.002 annarelli, a., nonino, f. (2016). strategic and operational management of organizational resilience: current state of research and future directions. omega, 62, 1–18. https://doi.org/10.1016/j.omega.2015.08.004 apneseth, k., wahl, a., hollnagel, e. (2013). measuring resilience in integrated planning. in albrechtsen, e & besnard, d (eds.) oil and gas, technology and humans: assessing the human factors of technological change. boca raton, fl: crc press. asgary, a., kong, a., levy, j. (2009). fuzzy-jess expert system for indexing business resiliency. in tic-sth’09: 2009 ieee toronto international conference science and technology for humanity, 153-158. https://doi.org/10.1109/tic-sth.2009.5444516 ates, a., bititci, u. (2011). change process: a key enabler for building resilient smes. international journal of production research, 49(18), 5601–5618. https://doi.org/10.1080/00207543.2011.563825 bauernhansl, t., mandel, j., diermann, s. (2012). evaluating changeability corridors for sustainable business resilience. in procedia cirp (vol. 3, pp. 364–369). https://doi.org/10.1016/j.procir.2012.07.063 beer, s. (1989). the viable system model: its provenance, development, methodology and pathology. in r. espejo & r. harnden (eds.), the viable system model, interpretations and applications of stafford beer’s vsm. chichester: wiley. https://doi.org/10.1057/jors.1984.2 beermann, m. (2011). linking corporate climate adaptation strategies with resilience thinking. journal of cleaner production, 19(8), 836–842. https://doi.org/10.1016/j.jclepro.2010.10.017 bell, m. (2002). the five principles of organizational resilience. gartner research.[online], 2–4. bergström, j., van winsen, r., henriqson, e. (2015). on the rationale of resilience in the domain of safety: a literature review. reliability engineering & system safety, 141, 131–141. https://doi.org/10.1016/j.ress.2015.03.008 berman, e. (2009). small business resilience. industrial management, 51(1), 6. bhamidipaty, a., lotlikar, r., banavar, g. (2007). rmi: a framework for modeling and evaluating the resiliency maturity of it service organizations. in ieee international conference on services computing (scc 2007), 300–307. salt lake city, ut. https://doi.org/10.1109/scc.2007.94 bhamra, r., burnard, k., dani, s. (2015). resilence. the concept, a literature review and future directions. in r. bhamra (ed.), organizational resilience. concepts, integration and practice, 3–29. crc press. https://doi.org/10.1201/b19305-3 bhamra, r., dani, s., burnard, k. (2011). resilience: the concept, a literature review and future directions. international journal of production research, 49(18), 5375–5393. https://doi.org/10.1080/00207543.2011.563826 biggs, d., hall, c. m., stoeckl, n. (2012). the resilience of formal and informal tourism enterprises to disasters: reef tourism in phuket, thailand. journal of sustainable tourism, 20(5), 645–665. https://doi.org/10.1080/09669582.2011.630080 boza, a., poler, r. (2013). enhancing enterprise resilience through enterprise collaboration. in ifac proceedings volumes (ifac-papersonline), 688–693. https://doi.org/10.3182/20130619-3-ru-3018.00283 braes, b., brooks, d. (2010). organisational resilience: a propositional study to understand and identify the essential concepts. proceedings 3rd australian security and intelligence conference, (november), 14–22. https://doi.org/10.4225/75/579ec432099ca braes, b., brooks, d. (2011). organisational resilience: understanding and identifying the essential concepts. in safety and security engineering iv, 117. wit press. https://doi.org/10.2495/safe110111 brewton, k. e., danes, s. m., stafford, k., haynes, g. w. (2010). determinants of rural and urban family firm resilience. journal of family business strategy, 1(3), 155–166. https://doi.org/10.1016/j.jfbs.2010.08.003 bruneau, m., chang, s. e., eguchi, r. t., lee, g. c., o’rourke, t. d., reinhorn, a. m., shinozuka, m., tierney, k., wallace, w. a., von winterfeldt, d. (2003). a framework to quantitatively assess and enhance the seismic resilience of communities. earthquake spectra, 19(4), 733–752. https://doi.org/10.1193/1.1623497 burnard, k., bhamra, r. (2011). organisational resilience: development of a conceptual framework for organisational responses. international journal of production research, 49(18), 5581–5599. https://doi.org/10.1080/00207543.2011.563827 caralli, r. a., curtis, p. d., allen, j. h., white, d. w., young, l. r. (2010). improving operational resilience processes: the cert® resilience management model. proceedings socialcom 2010: 2nd ieee international conference on social computing, passat 2010: 2nd ieee international conference on privacy, security, risk and trust, 1165–1170. https://doi.org/10.1109/socialcom.2010.173 int. j. prod. manag. eng. (2018) 6(1), 11-28creative commons attribution-noncommercial-noderivatives 4.0 international what we know and do not know about organizational resilience 23 https://doi.org/10.1016/j.jlp.2013.06.004 https://doi.org/10.1016/j.jlp.2013.06.004 https://doi.org/10.4337/9781782547884.00026 https://doi.org/10.1108/ijwbr-08-2014-0035 https://doi.org/10.1016/j.jom.2014.11.002 https://doi.org/10.1016/j.omega.2015.08.004 https://doi.org/10.1109/tic-sth.2009.5444516 https://doi.org/10.1080/00207543.2011.563825 https://doi.org/10.1016/j.procir.2012.07.063 https://doi.org/10.1057/jors.1984.2 https://doi.org/10.1016/j.jclepro.2010.10.017 https://doi.org/10.1016/j.ress.2015.03.008 https://doi.org/10.1109/scc.2007.94 https://doi.org/10.1201/b19305-3 https://doi.org/10.1080/00207543.2011.563826 https://doi.org/10.1080/09669582.2011.630080 https://doi.org/10.3182/20130619-3-ru-3018.00283 https://doi.org/10.4225/75/579ec432099ca https://doi.org/10.2495/safe110111 https://doi.org/10.1016/j.jfbs.2010.08.003 https://doi.org/10.1193/1.1623497 https://doi.org/10.1109/socialcom.2010.173 http://creativecommons.org/licenses/by-nc-nd/4.0/ chadegani, a. a., salehi, h., yunus, m. m., farhadi, h., fooladi, m., farhadi, m., ebrahim, n. a. (2013). a comparison between two main academic literature collections: web of science and scopus databases. asian social science, 9(5), 18–26. https://doi.org/10.5539/ ass.v9n5p18 chand, a. m., loosemore, m. (2016). hospital learning from extreme weather events: using causal loop diagrams. building research and information, 44(8), 875-888. https://doi.org/10.1080/09613218.2016.1097805 chopra, s. s., khanna, v. (2014). understanding resilience in industrial symbiosis networks: insights from network analysis. journal of environmental management, 141(april 2014), 86–94. https://doi.org/10.1016/j.jenvman.2013.12.038 coutu, d. l. (2002). how resilience works. harvard business review, 80(5), 46–56. crichton, m. t., ramsay, c., kelly, t. (2009). enhancing organizational resilience through emergency planning: learnings from cross-sectoral lessons. journal of contingencies and crisis management, 17(1), 24–37. https://doi.org/10.1111/j.1468-5973.2009.00556.x dalziell, e. p., mcmanus, s. t. (2004). resilience, vulnerability, and adaptive capacity: implications for system performance. international forum for engineering decision making, 17. danes, s. m., lee, j., amarapurkar, s., stafford, k., haynes, g., brewton, k. e. (2009). determinants of family business resilience after a natural disaster by gender of business owner. journal of developmental entrepreneurship, 14(4), 333–354. https://doi.org/10.1142/ s1084946709001351 demmer, w. a., vickery, s. k., calantone, r. (2011). engendering resilience in small-and medium-sized enterprises (smes): a case study of demmer corporation. international journal of production research, 49(18), 5395–5413. https://doi.org/10.1080/00207543.201 1.563903 dervitsiotis, k. n. (2004). navigating in turbulent environmental conditions for sustainable business excellence. total quality management & business excellence, 15(5–6), 807–827. https://doi.org/10.1080/14783360410001680251 dewald, j., bowen, f. (2010). storm clouds and silver linings: responding to disruptive innovations through cognitive resilience. entrepreneurship: theory and practice, 34(1), 197–218. https://doi.org/10.1111/j.1540-6520.2009.00312.x doe, p. j. (1994). creating a resilient organization. canadian business review, 21, 22. erol, o., henry, d., sauser, b. (2010). exploring resilience measurement methodologies. in 20th annual international symposium of the international council on systems engineering, incose 2010, 1, 302–322. erol, o., henry, d., sauser, b., mansouri, m. (2010). perspectives on measuring enterprise resilience. in 2010 ieee international systems conference proceedings, syscon 2010, 587–592. https://doi.org/10.1109/systems.2010.5482333 erol, o., mansouri, m., sauser, b. (2009). a framework for enterprise resilience using service oriented architecture approach. in 2009 ieee international systems conference proceedings, 127–132. https://doi.org/10.1109/systems.2009.4815785 erol, o., sauser, b. j., mansouri, m. (2010). a framework for investigation into extended enterprise resilience. enterprise information systems, 4(2), 111–136. https://doi.org/10.1080/17517570903474304 fiksel, j. (2003). designing resilient, sustainable systems. environmental science and technology, 37(23), 5330–5339. https://doi.org/10.1021/es0344819 fiksel, j. (2006). sustainability and resilience: toward a systems approach. sustainability: science, practice, policy, 2(2), 14–21. https://doi.org/10.1080/15487733.2006.11907980 freeman, s. f., hirschhorn, l., triad, m. m. (2003). moral purpose and organizational resilience: sandler o’neill & partners, lp in the aftermath of september 11, 2001. in academy of management proceedings, 2003, b1-b6. https://doi.org/10.5465/ambpp.2003.13792457 gibson, c., tarrant, m. (2010). a “conceptual models” approach to organisational resilience. the australian journal of emergency management, 25(2), 8–14. gilly, j.p., kechidi, m., talbot, d. (2014). resilience of organisations and territories: the role of pivot firms. european management journal, 32(4), 596–602. https://doi.org/10.1016/j.emj.2013.09.004 grande, o., trucco, p. (2008). resilience analysis of civil defence organization: a fuzzy cognitive map based approach. in 9th international conference on probabilistic safety assessment and management 2008, psam 2008, 2, 1542–1549. grøtan, t. o., asbjørnslett, b. e. (2007). ict in resilient global logistics. in proceedings of esrel 2007, 2349–2356. grøtan, t. o., størseth, f., rø, m. h., skjerve, a. b. (2008). resilience, adaptation and improvisation – increasing resilience by organising for successful improvisation. 3rd symposium on resiience engineering, 1–7. gunasekaran, a., rai, b. k., griffin, m. (2011). resilience and competitiveness of small and medium size enterprises: an empirical research. international journal of production research, 49(18), 5489–5509. https://doi.org/10.1080/00207543.2011.563831 hamel, g., valikangas, l. (2003). the quest for resilience. harvard business review, 81(9), 52–65. hardy, t. l. (2014). resilience: a holistic safety approach. in proceedings annual reliability and maintainability symposium, 1–6. https//doi.org/10.1109/rams.2014.6798494 heinicke, m. (2014). implementation of resilient production systems by production control. in procedia cirp, 19, 105–110. https://doi.org/10.1016/j.procir.2014.05.001 int. j. prod. manag. eng. (2018) 6(1), 11-28 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz-martin, c., lópez-paredes, a. and wainer, g. 24 https://doi.org/10.5539/ass.v9n5p18 https://doi.org/10.5539/ass.v9n5p18 https://doi.org/10.1080/09613218.2016.1097805 https://doi.org/10.1016/j.jenvman.2013.12.038 https://doi.org/10.1111/j.1468-5973.2009.00556.x https://doi.org/10.1142/s1084946709001351 https://doi.org/10.1142/s1084946709001351 https://doi.org/10.1080/00207543.2011.563903 https://doi.org/10.1080/00207543.2011.563903 https://doi.org/10.1080/14783360410001680251 https://doi.org/10.1111/j.1540-6520.2009.00312.x https://doi.org/10.1109/systems.2010.5482333 https://doi.org/10.1109/systems.2009.4815785 https://doi.org/10.1080/17517570903474304 https://doi.org/10.1021/es0344819 https://doi.org/10.1080/15487733.2006.11907980 https://doi.org/10.5465/ambpp.2003.13792457 https://doi.org/10.1016/j.emj.2013.09.004 https://doi.org/10.1080/00207543.2011.563831 http://https//doi.org/10.1109/rams.2014.6798494 https://doi.org/10.1016/j.procir.2014.05.001 http://creativecommons.org/licenses/by-nc-nd/4.0/ henry, d., ramirez-marquez, j. e. (2010). a generic quantitative approach to resilience: a proposal. in 20th annual international symposium of the international council on systems engineering, incose 2010, 1, 291–301. hilton, j., wright, c., kiparoglou, v. (2012). building resilience into systems. in syscon 2012 2012 ieee international systems conference, proceedings, 638–645. https://doi.org/10.1109/syscon.2012.6189449 holling, c. s. (1973). resilience and stability of ecological systems. annual review of ecology and systematics, 1–23. https://doi.org/10.1146/annurev.es.04.110173.000245 hollnagel, e. (2010). how resilient is your organisation? an introduction to the resilience analysis grid (rag). sustainable transformation: building a resilient organization. horne iii, j. f. (1997). the coming age of organizational resilience. business forum, 22(2/3), 24–29. horne iii, j. f., orr, j. e. (1998). assessing behaviors that create resilient organizations. employment relations today, 24(4), 29–39. hu, y., li, j., holloway, l. e. (2008). towards modeling of resilience dynamics in manufacturing enterprises: literature review and problem formulation. in 4th ieee conference on automation science and engineering, case 2008, 279–284. https://doi.org/10.1109/ coase.2008.4626539 hu, y., li, j., holloway, l. e. (2009). a modeling and aggregation approach for analyzing resilience of manufacturing enterprises. in conference proceedings ieee international conference on systems, man and cybernetics, 692–697. https://doi.org/10.1109/ icsmc.2009.5346703 hu, y., li, j., holloway, l. e. (2010). achieving resilience for a class of serial production networks. in proceedings of the 2010 american control conference, acc 2010, 5326–5331. https://doi.org/10.1109/acc.2010.5530745 ismail, h. s., poolton, j., sharifi, h. (2011). the role of agile strategic capabilities in achieving resilience in manufacturing-based small companies. international journal of production research, 49(18), 5469–5487. https://doi.org/10.1080/00207543.2011.563833 jaaron, a. a. m., backhouse, c. j. (2014). service organisations resilience through the application of the vanguard method of systems thinking: a case study approach. international journal of production research, 52(7), 2026–2041. https://doi.org/10.1080/0020754 3.2013.847291 jackson, d., firtko, a., edenborough, m. (2007). personal resilience as a strategy for surviving and thriving in the face of workplace adversity: a literature review. journal of advanced nursing, 60(1), 1–9. https://doi.org/10.1111/j.1365-2648.2007.04412.x jackson, s. (2007). a multidisciplinary framework for resilence to disasters and disruptions. journal of integrated design and process science, 11(2), 91–108. johnsen, s. o., veen, m. (2012). risk assessment and improvement of resilience of critical communication infrastructure. in advances in safety, reliability and risk management proceedings of the european safety and reliability conference, esrel 2011, 2739–2747. johnson, j., gheorghe, a. v. (2013). antifragility analysis and measurement framework for systems of systems. international journal of disaster risk science, 4(4), 159–168. https://doi.org/10.1007/s13753-013-0017-7 jones, k. h. (2014). engineering antifragile systems: a change in design philosophy. procedia computer science, 32, 870–875. https://doi.org/10.1016/j.procs.2014.05.504 kamalahmadi, m., parast, m. m. (2016). a review of the literature on the principles of enterprise and supply chain resilience: major findings and directions for future research. international journal of production economics, 171, 116–133. https://doi.org/10.1016/j. ijpe.2015.10.023 kendra, j., wachtendorf, t. (2002). elements of community resilience in the world trade center attack, 97–103. kendra, j., wachtendorf, t. (2003). elements of resilience after the world trade center disaster: reconstituting new york city’s emergency operations centre. disasters, 27(1), 37–53. https://doi.org/10.1111/1467-7717.00218 kennon, d., schutte, c. s. l., lutters, e. (2015). an alternative view to assessing antifragility in an organisation: a case study in a manufacturing sme. cirp annals manufacturing technology, 64(1), 177–180. https://doi.org/10.1016/j.cirp.2015.04.024 kohno, y., masuda, y., nagahashi, h., tanaka, k., tashiro, k. (2012). form development for self-rating an organization’s vulnerability and resilience to disruption. journal of disaster research, 7(4), 392–407. https://doi.org/10.20965/jdr.2012.p0392 lambertini, l., marattin, l. (2016). to adjust or not to adjust after a cost-push shock? a simple duopoly model with (and without) resilience. economics of innovation and new technology, 25(2), 172–181. https://doi.org/10.1080/10438599.2015.1031344 lee, a. v, vargo, j., seville, e. (2013). developing a tool to measure and compare organizations’ resilience. natural hazards review, 14, 29–41. https://doi.org/10.1061/(asce)nh.1527-6996.0000075. lengnick-hall, c. a., beck, t. e. (2005). adaptive fit versus robust transformation: how organizations respond to environmental change. journal of management, 31(5), 738–757. https://doi.org/10.1177/0149206305279367 lengnick-hall, c. a., beck, t. e., lengnick-hall, m. l. (2011). developing a capacity for organizational resilience through strategic human resource management. human resource management review, 21(3), 243–255. https://doi.org/10.1016/j.hrmr.2010.07.001 linnenluecke, m. k. (2017). resilience in business and management research: a review of influential publications and a research agenda. international journal of management reviews, 19(1), 4–30. https://doi.org/10.1111/ijmr.12076 int. j. prod. manag. eng. (2018) 6(1), 11-28creative commons attribution-noncommercial-noderivatives 4.0 international what we know and do not know about organizational resilience 25 https://doi.org/10.1109/syscon.2012.6189449 https://doi.org/10.1146/annurev.es.04.110173.000245 https://doi.org/10.1109/coase.2008.4626539 https://doi.org/10.1109/coase.2008.4626539 https://doi.org/10.1109/icsmc.2009.5346703 https://doi.org/10.1109/icsmc.2009.5346703 https://doi.org/10.1109/acc.2010.5530745 https://doi.org/10.1080/00207543.2011.563833 https://doi.org/10.1080/00207543.2013.847291 https://doi.org/10.1080/00207543.2013.847291 https://doi.org/10.1111/j.1365-2648.2007.04412.x https://doi.org/10.1007/s13753-013-0017-7 https://doi.org/10.1016/j.procs.2014.05.504 https://doi.org/10.1016/j.ijpe.2015.10.023 https://doi.org/10.1016/j.ijpe.2015.10.023 https://doi.org/10.1111/1467-7717.00218 https://doi.org/10.1016/j.cirp.2015.04.024 https://doi.org/10.20965/jdr.2012.p0392 https://doi.org/10.1080/10438599.2015.1031344 https://doi.org/10.1061/(asce)nh.1527-6996.0000075 https://doi.org/10.1177/0149206305279367 https://doi.org/10.1016/j.hrmr.2010.07.001 https://doi.org/10.1111/ijmr.12076 http://creativecommons.org/licenses/by-nc-nd/4.0/ linnenluecke, m. k., griffiths, a. (2010). beyond adaptation: resilience for business in light of climate change and weather extremes. business and society, 49(3), 477–511. https://doi.org/10.1177/0007650310368814 linnenluecke, m. k., griffiths, a., winn, m. (2012). extreme weather events and the critical importance of anticipatory adaptation and organizational resilience in responding to impacts. business strategy and the environment, 21(1), 17–32. https://doi.org/10.1002/ bse.708 louisot, j.-p. (2015). risk and/or resilience management. risk governance & control: financial markets and institutions, 5(2), 84–91. https://doi.org/10.22495/rgcv5i2c1art2 macuzić, i., tadić, d., aleksic, a., stefanović, m. (2016). a two step fuzzy model for the assessment and ranking of organizational resilience factors in the process industry. journal of loss prevention in the process industries, 40, 122–130. https://doi.org/10.1016/j. jlp.2015.12.013 mafabi, s., munene, j. c., ahiauzu, a. (2015). creative climate and organisational resilience: the mediating role of innovation. international journal of organizational analysis, 23(4), 564–587. https://doi.org/10.1108/ijoa-07-2012-0596 mallak, l. a. (1997). how to build a resilient organization. in proceedings of the industrial engineering solutions 1997 conference, 170–177. miami. mallak, l. a. (1998a). measuring resilience in health care provider organizations. health manpower management, 24(4), 148–152. https://doi.org/10.1108/09552069810215755 mallak, l. a. (1998b). putting organisational resilience to work. industrial management, 40(6), 8–13. mamouni limnios, e. (2011). resilient organizations : offense versus defense. in in 25th annual anzam conference, 7–9. mamouni limnios, e., mazzarol, t., ghadouani, a., schilizzi, s. g. (2014). the resilience architecture framework: four organizational archetypes. european management journal, 32(1), 104–116. https://doi.org/10.1016/j.emj.2012.11.007 manyena, s. b. (2006). the concept of resilience revisited. disasters, 30(4), 433–450. https://doi.org/10.1111/j.0361-3666.2006.00331.x markman, g. m., venzin, m. (2014). resilience: lessons from banks that have braved the economic crisis-and from those that have not. international business review, 23(6), 1096–1107. https://doi.org/10.1016/j.ibusrev.2014.06.013 mcmanus, s., seville, e., vargo, j., brunsdon, d. (2007). resilience management: a framework for assessing and improving the resilience of organisations. resilient organisations research report 2007/01. mcmanus, s., seville, e., vargo, j., brunsdon, d. (2008). facilitated process for improving organizational resilience. natural hazards review, 9(2), 81–90. https://doi.org/10.1061/(asce)1527-6988(2008)9:2(81) megele, c. (2014). resilient organizations turning challenges into opportunities: hr occupies a central place in preparing companies for change. human resource management international digest, 22(5), 1–4. https://doi.org/10.1108/hrmid-07-2014-0092 milanzi, d., weeks, r. (2014). understanding servitization: a resilience perspective. in picmet 2014 portland international center for management of engineering and technology, proceedings: infrastructure and service integration, 2332–2342. orchiston, c., prayag, g., brown, c. (2016). organizational resilience in the tourism sector. annals of tourism research, 56, 145–148. https://doi.org/10.1016/j.annals.2015.11.002 ortiz-de-mandojana, n., bansal, p. (2015). the long-term benefits of organizational resilience through sustainable business practices. strategic management journal. https://doi.org/10.1002/smj.2410 oxford university press. (2017). oxforddictionaries. retrieved november 17, 2017, from https://en.oxforddictionaries.com/definition/ organization pal, r., torstensson, h., mattila, h. (2014). antecedents of organizational resilience in economic crises an empirical study of swedish textile and clothing smes. international journal of production economics, 147(part b), 410–428. https://doi.org/10.1016/j.ijpe.2013.02.031 powley, e. h. (2009). reclaiming resilience and safety: resilience activation in the critical period of crisis. human relations, 62(9), 1289–1326. https://doi.org/10.1177/0018726709334881 proper, j. w., pienaar, w. j. (2011). resilience as an imperative in public transport organisations. corporate ownership and control, 8(4 d), 373–388. https://doi.org/10.22495/cocv8i4c3art6 read, d. (2005). some observations on resilience and robustness in human systems. cybernetics & systems, 36, 773–802. https://doi.org/10.1080/01969720500306253 reinmoeller, p., van baardwijk, n. (2005). the link between diversity and resilience. mit sloan management review, 46(4), 61. rerup, c. (2001). “houston, we have a problem”: anticipation and improvisation as sources of organizational resilience. comportamiento organizacional y gestão, 7(1), 27–44. rigaud, e., neveu, c., duvenci-langa, s., obrist, m.-n., rigaud, s. (2013). proposition of an organisational resilience assessment framework dedicated to railway traffic management. in rail human factors: supporting reliability, safety and cost reduction, 727–732. https://doi.org/10.1201/b13827-97 righi, a. w., saurin, t. a., wachs, p. (2015). a systematic literature review of resilience engineering: research areas and a research agenda proposal. reliability engineering & system safety, 141, 142–152. https://doi.org/10.1016/j.ress.2015.03.007 int. j. prod. manag. eng. (2018) 6(1), 11-28 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz-martin, c., lópez-paredes, a. and wainer, g. 26 https://doi.org/10.1177/0007650310368814 https://doi.org/10.1002/bse.708 https://doi.org/10.1002/bse.708 https://doi.org/10.22495/rgcv5i2c1art2 https://doi.org/10.1016/j.jlp.2015.12.013 https://doi.org/10.1016/j.jlp.2015.12.013 https://doi.org/10.1108/ijoa-07-2012-0596 https://doi.org/10.1108/09552069810215755 https://doi.org/10.1016/j.emj.2012.11.007 https://doi.org/10.1111/j.0361-3666.2006.00331.x https://doi.org/10.1016/j.ibusrev.2014.06.013 https://doi.org/10.1061/(asce)1527-6988(2008)9:2(81) https://doi.org/10.1108/hrmid-07-2014-0092 https://doi.org/10.1016/j.annals.2015.11.002 https://doi.org/10.1002/smj.2410 https://en.oxforddictionaries.com/definition/organization https://en.oxforddictionaries.com/definition/organization https://doi.org/10.1016/j.ijpe.2013.02.031 https://doi.org/10.1177/0018726709334881 https://doi.org/10.22495/cocv8i4c3art6 https://doi.org/10.1080/01969720500306253 https://doi.org/10.1201/b13827-97 https://doi.org/10.1016/j.ress.2015.03.007 http://creativecommons.org/licenses/by-nc-nd/4.0/ riolli, l., savicki, v. (2003). information system organizational resilience. omega, 31(1), 227–233. https://doi.org/10.1016/s03050483(03)00023-9 robb, d. (2000). building resilient organizations. od practitioner, 32(3), 27–32. ruiz-martin, c., lópez-paredes, a., wainer, g. (2015). applying complex network theory to the assessment of organizational resilience. ifac-papersonline, 48(3), 1224–1229. https://doi.org/10.1016/j.ifacol.2015.06.251 ruiz-martin, c., pérez ríos, j., wainer, g., pajares, j., hernández, c., lópez-paredes, a. (2017). the application of the viable system model to enhance organizational resilience. in advances in management engineering (pp. 88–100). https://doi.org/10.1007/978-3319-55889-9_5 sanchis, r., poler, r. (2013). definition of a framework to support strategic decisions to improve enterprise resilience. in ifac proceedings volumes (ifac-papersonline), 700–705. https://doi.org/10.3182/20130619-3-ru-3018.00600 seville, e. (2009). resilience: great concept... but what does it mean for organizations? tephra, july, 9–14. sheffi, y. (2007). building a resilient organization. the bridge national academy of engineering, 37(1), 30–36. sheffi, y., rice jr., j. b. (2005). a supply chain view of the resilient enterprise. mit sloan management review, 47(1), 41–48+94. smith, d., fischbacher, m. (2009). the changing nature of risk and risk management: the challenge of borders, uncertainty and resilience. risk management, 11(1), 1–12. https://doi.org/10.1057/rm.2009.1 somers, s. (2009). measuring resilience potential: an adaptive strategy for organizational crisis planning. journal of contingencies and crisis management, 17(1), 12–23. https://doi.org/10.1111/j.1468-5973.2009.00558.x starr, r., newfrock, j., delurey, m. (2003). enterprise resilience: managing risk in the networked economy. strategy and business, 30, 70–79. stewart, j., o’donnell, m. (2007). implementing change in a public agency. international journal of public sector management, 20(3), 239–251. https://doi.org/10.1108/09513550710740634 suddaby, r. (2010). editor’s comments: construct clarity in theories of management and organization. academy of management review, 35(3), 346–357. https://doi.org/10.5465/amr.2010.51141319 sutcliffe, k. m., vogus, t. j. (2003). organizing for resilience. in k. cameron, j. e. dutton, r. e. quinn (eds.), positive organizational scholarship, 94–110. san francisco. tadić, d., aleksic, a., stefanović, m., arsovski, s. (2014). evaluation and ranking of organizational resilience factors by using a two-step fuzzy ahp and fuzzy topsis. mathematical problems in engineering, 1–13. https://doi.org/10.1155/2014/418085 taleb, n. n. (2012). antifragile: things that gain from disorder (vol. 3). new york: random house. taleb, n. n., douady, r. (2013). mathematical definition, mapping, and detection of (anti)fragility. quantitative finance, 13(11), 1677–1689. https://doi.org/10.1080/14697688.2013.800219 thomas, a., byard, p., francis, m., fisher, r., white, g. r. t. (2016). profiling the resiliency and sustainability of uk manufacturing companies. journal of manufacturing technology management, 27(1), 82–99. https://doi.org/10.1108/jmtm-06-2014-0086 tierney, k. j. (2003). conceptualizing and measuring organizational and community resilience: lessons from the emergency response following the september 11, 2001 attack on the world trade center. disaster research center preliminary papers. tillement, s., cholez, c., reverdy, t. (2009). assessing organizational resilience: an interactionist approach. management, 12(4), 230–264. https://doi.org/10.3917/mana.124.0230 tolk, a., johnson iv, j. j. (2013). implementing antifragiles: systems that get better under change. in american society for engineering management. 2013 international annual conference. tompkins, j. a. (2007). 4 steps to business resilience. industrial management (norcross, georgia), 49(4), 14–18. tranfield, d., denyer, d., smart, p. (2003). towards a methodology for developing evidence-informed management knowledge by means of systematic review. british journal of management, 14, 207–222. https://doi.org/10.1111/1467-8551.00375 tse, t., couturier, j., roux, l. (2012). resilience of a business: the case of yellow pages directories. international journal of management practice, 5(2), 149–169. https://doi.org/10.1504/ijmp.2012.047582 tseitlin, a. (2013). the antifragile organization. communications of the acm, 56(8), 40. https://doi.org/10.1145/2492007.2492022 tukamuhabwa, b. r., stevenson, m., busby, j., zorzini, m. (2015). supply chain resilience: definition, review and theoretical foundations for further study. international journal of production research, 53(18), 5592–5623. https://doi.org/10.1080/00207543.2015.1037934 välikangas, l., romme, a. g. l. (2012). building resilience capabilities at “big brown box, inc.” strategy & leadership, 40(4), 43–45. https://doi.org/10.1108/10878571211242948 van breda, a. d. (2016). building resilient human service organizations. human service organizations management, leadership and governance, 40(1), 62–73. https://doi.org/10.1080/23303131.2015.1093571 van trijp, j. m. p., ulieru, m., van gelder, p. h. a. j. m. (2012a). quantitative approach of organizational resilience for a dutch emergency response safety region. in advances in safety, reliability and risk management proceedings of the european safety and reliability conference, esrel 2011 (pp. 173–180). int. j. prod. manag. eng. (2018) 6(1), 11-28creative commons attribution-noncommercial-noderivatives 4.0 international what we know and do not know about organizational resilience 27 https://doi.org/10.1016/s0305-0483(03)00023-9 https://doi.org/10.1016/s0305-0483(03)00023-9 https://doi.org/10.1016/j.ifacol.2015.06.251 https://doi.org/10.1007/978-3-319-55889-9_5 https://doi.org/10.1007/978-3-319-55889-9_5 https://doi.org/10.3182/20130619-3-ru-3018.00600 https://doi.org/10.1057/rm.2009.1 https://doi.org/10.1111/j.1468-5973.2009.00558.x https://doi.org/10.1108/09513550710740634 https://doi.org/10.5465/amr.2010.51141319 https://doi.org/10.1155/2014/418085 https://doi.org/10.1080/14697688.2013.800219 https://doi.org/10.1108/jmtm-06-2014-0086 https://doi.org/10.3917/mana.124.0230 https://doi.org/10.1111/1467-8551.00375 https://doi.org/10.1504/ijmp.2012.047582 https://doi.org/10.1145/2492007.2492022 https://doi.org/10.1080/00207543.2015.1037934 https://doi.org/10.1108/10878571211242948 https://doi.org/10.1080/23303131.2015.1093571 http://creativecommons.org/licenses/by-nc-nd/4.0/ van trijp, j. m. p., ulieru, m., van gelder, p. h. a. j. m. (2012b). quantitative modeling of organizational resilience for dutch emergency response safety regions. proceedings of the institution of mechanical engineers, part o: journal of risk and reliability, 226(6), 666–676. https://doi.org/10.1177/1748006x12466371 vogus, t. j., sutcliffe, k. m. (2007). organizational resilience: towards a theory and research agenda. conference proceedings ieee international conference on systems, man and cybernetics, 3418–3422. https://doi.org/10.1109/icsmc.2007.4414160 watanabe, c., kishioka, m., nagamatsu, a. (2004). resilience as a source of survival stategy for high-technology firms experiencing megacompetition. technovation, 24(2), 139–152. https://doi.org/10.1016/s0166-4972(02)00048-2 weick, k. e. (1993). the collapse of sensemaking in organizations: the mann gulch disaster. administrative science quarterly, 628–652. https://doi.org/10.2307/2393339 whitehorn, g. (2010). building organisational resilience in the public sector. in comcover insurance and risk management conference. whitman, z. r., kachali, h., roger, d., vargo, j., seville, e. f. (2013). short-form version of the benchmark resilience tool (brt-53). measuring business excellence, 17(3), 3–14. https://doi.org/10.1108/mbe-05-2012-0030 wicker, p., filo, k., cuskelly, g. (2013). organizational resilience of community sport clubs impacted by natural disasters. journal of sport management, 27(6), 510–525. https://doi.org/10.1123/jsm.27.6.510 winston, a. (2014). resilience in a hotter world. harvard business review, 92(4), 56–64,132. woods, d. d. (2015). four concepts for resilience and the implications for the future of resilience engineering. reliability engineering and system safety, 141, 5–9. https://doi.org/10.1016/j.ress.2015.03.018 wright, c., kiparoglou, v., williams, m., hilton, j. (2012). a framework for resilience thinking. procedia computer science, 8, 45–52. https://doi.org/10.1016/j.procs.2012.01.012 zhang, w. j., van luttervelt, c. a. (2011). toward a resilient manufacturing system. cirp annals manufacturing technology, 60(1), 469–472. https://doi.org/10.1016/j.cirp.2011.03.041 int. j. prod. manag. eng. (2018) 6(1), 11-28 creative commons attribution-noncommercial-noderivatives 4.0 international ruiz-martin, c., lópez-paredes, a. and wainer, g. 28 https://doi.org/10.1177/1748006x12466371 https://doi.org/10.1109/icsmc.2007.4414160 https://doi.org/10.1016/s0166-4972(02)00048-2 https://doi.org/10.2307/2393339 https://doi.org/10.1108/mbe-05-2012-0030 https://doi.org/10.1123/jsm.27.6.510 https://doi.org/10.1016/j.ress.2015.03.018 https://doi.org/10.1016/j.procs.2012.01.012 https://doi.org/10.1016/j.cirp.2011.03.041 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2017.6512 received 2016-08-18 accepted: 2017-01-09 an effective branch-and-cut algorithm in order to solve the mixed integer bi-level programming arsalan rahmani a and majid yousefikhoshbakht b* adepartment of mathematics, university of kurdistan, sanandaj, iran. byoung researchers & elite club, hamedan branch, islamic azad university, hamedan, iran. a arsalan.rah@gmail.com b khoshbakht@iauh.ac.ir abstract: in this paper, a new branch-and-cut algorithm for mixed integer bi-level programming is proposed. for achieving this purpose, a historical perspective of the development of enumeration methods in the field of bi-level linear programming is considered. then, we present some obstacles for using branch and bound method based on them, and an algorithm is developed to solve for mixed integer bi-level problem. finally, we use a preference function to determine the choice of branching and specialized cuts in a branch and cut tree. computational results are reported and compared favorably to those of previous methods and then implications discussed. the results show that not only the proposed algorithm can find high quality solutions for solving a number of the problems, but also it is competitive with other famous published algorithms. key words: mixed-integer bi-level programming, branch and cut method, fathoming branch. 1. introduction an interactive process in which a central unit (leader) coordinates a lower level unit (follower) is called hierarchical organizations. when the follower afforded some level of autonomy, this process becomes more complex to implement, coordinate and optimize. moreover, in some instances, the objectives of the follower may conflict with those of the leader. in mathematical programming environment, these interactive processes are existed and known as bi-level or multi-level programming. linear bi-level programming problems (blpp) generally involve a hierarchy of two optimization problems, in the following form: 1 1 1 1:maxx p z c x d y= + 1 1 1. .s t a x b y g+ ≤ x x∈ 2 2max y c x d y+ 2 2 2. .s t a x b y g+ ≤ y y∈ where x and y are the variables of problem p1 divided into two classes, namely the upper-level variables x and the lower-level variables y. besides, the functions int. j. prod. manag. eng. (2017) 5(1), 1-10creative commons attribution-noncommercial-noderivatives 4.0 international 1 https://doi.org/10.4995/ijpme.2017.6512 mailto:khoshbakht%40iauh.ac.ir?subject= http://creativecommons.org/licenses/by-nc-nd/4.0/ 1 1 2 2,c x d y c x d y+ + are called upper level and lower level objective function, 1 1 1a x b y g+ ≤ and 2 2 2a x b y g+ ≤ are known as upper level and lower level constraints respectively. upper-level constraints involve variables from both levels. in this paper, a special class of blpp is considered and its model is supposed to have the following characteristics: in the upper level problem, the constraints do not include any variable from the lower level problem. the most competitive facility location problems lay in this group, in fact decisions about the location of facilities (for the first decision maker (leader)), capacities of facilities, transportations and so on are full free of lower level variables, the lower level variables may take into accounts in upper level’s objective function. for any feasible solution of upper-level problem, there is a feasible solution for the lower-level problem. in the problems, which a level of service is objective, when one company is not able to service a customer, the other company do it. as a result, in this case for any feasible solution of upper-level problem, there is a feasible solution for the lower-level problem. a special case of this type of problem is discrete competitive facility location problem (beresnev, 2009). jeroslow showed that bi-level programming problems are np-hard even in the “simplest case”, the linear blpp (jeroslow, 1985) while (hansen et al., 1992) proved that the problem is strong nphardness. regarding to solution approaches, many algorithms were proposed in literature. gümüş studied global optimization of bi-level programming problems (gümüş and floudas, 2005), and proposed a convex relaxation of the inner problem followed by its equivalent representation via necessary and sufficient optimality conditions. the approximated branch and bound global optimal principles presented a branch and bound framework. the first precise global optimization approach for the calculation of the flexibility test that is bi-level nonlinear optimization model is introduced in (floudas et al., 1999). then they demonstrate its applicability to a heat exchanger network problem, a pump and pip problem, a reactor-cooler system and a prototype process flow sheet model. pistikopoulos introduced methods based on parametric programming to transform the bi-level problem into a family of single level optimization problems which can be solved to global optimality (pistikopoulos et al., 2007), and presented computational results on several small benchmarks linear-linear, linear-quadratic, quadratic-linear and quadratic-quadratic type problems. moreover, two new methods for bi-level programming problem are proposed (gümüş and floudas, 2005). in the case of bi-level linear programming different algorithms proposed in (vicente, 2001), (colson et al., 2005) and (domínguez and pistikopoulos, 2010). dempe considered the case characterized by continuous upper-level variables and integer lowerlevel variables, and used a cutting-plane approach to approximate the lower-level feasible region (dempe, 2001). wen considered the opposite case, where the lower-level problem is a linear program and the upper-level problem is an integer program (yang, 1990). linear programming duality employed to derive exact solutions. in the discrete case, moore developed a basic enumeration scheme to identify feasible solutions (bard and moore, 1990). beside these methods an applicable lagrangean based algorithm also proposed in (rahmani a, 2015) and the metaheuristic method using particle swarm optimization method has been used for solving a special class of bi-level problem (mirhassani, 2015). a well-known approach to solve a linear blp problem is transferring into a nonlinear programming problem using kuhn-tucker conditions. bard and falk (1982) reported several obstacles to the development of algorithms for solving blmip problems. in this regard they deliver a branch and bound method (bard and moore, 1990) and denegre proposed a simple branch-and-cut algorithm (denegre and ralphs, 2009). for others related works see hansen et al. (1992), shi et al. (2006), vicente et al. (1996), beresnev (2013) and beresnev and melnikov (2014). the goal of this work is to demonstrate that it is possible, in principle, to generalize the greatly successful branch-and-cut (by using suitable branching and cutting rules) framework commonly used to solve mixed integer linear programs to this very challenging computational setting. our implementation is quite rudimentary and intend only as a demonstration of the concept. the algorithm int. j. prod. manag. eng. (2017) 5(1), 1-10 creative commons attribution-noncommercial-noderivatives 4.0 international rahmani, a. and yousefikhoshbakht, m. 2 http://creativecommons.org/licenses/by-nc-nd/4.0/ demonstrates a better performance respect to the branch-and bound algorithm (bard and moore, 1990; beresnev and melnikov, 2014), and branch and cut algorithm (denegre and ralphs, 2009). it can be implemented in a straightforward way using existing software. the computational results show the algorithm efficiency in terms of solution quality and running time. the rest of the paper is as follows. in section 2, we describe the mathematical models and the challenge of solving the models by generalizing solution methods for single-level mathematical programming problems. in section 3, we propose a branch-and-cut algorithm for miblps and section 4 illustrates the algorithm via an example and provides some preliminary computational results. finally, in section 5, we provide conclusions and directions for future work. 2. definitions and notation bard and falk (bard and falk, 1982) utilized the following notation and definitions in their work. blpp constraint region: ( , ) | , , ,x y a x b y g a x b y g x x y y1 1 1 2 2 2# # ! !x = + +" , projection of ω onto the leader’s decision space: ( ) | , ,x x x y y a x b y g a x b y g1 1 1 2 2 27! ! # #x = + +" , follower’s feasible region for x x∈ fixed: ( ) |x y y a x b y g2 2 2! #x = +" , follower’s rational reaction set for ( )x x∈ ω : ( ) { | argmax[ | ( )]}m x y y y c x d y y x y 2 2! ! !x= + inducible region: ( , ) | ( , ) , ( )x y x y y m xr ! !x= " , in order to make p1 well posed it is assumed that ω is non-empty and compact, and for each decision taken by the leader there is some room to move for the follower, or ( )x q!x . definition 1: if ( )y m x∈ then y is said to be optimal with respect to x; such a pair is said to be bi-level feasible. definition 2: a point (x*, y*) is said to be an optimal solution to the blpp if a) (x*, y*) is bi-level feasible; and, b) for all feasible pairs (x,y); c1x+d1y≤c1x*+d1y*. definition 3: we said (x, y) is valid for a cut (set); if it satisfied in this cut (set). definition 4: an inequality defined by (a,b,g) is said valid for set s, if for all pairs ( , )x y s∈ ; ax+by ≤ g bard and moore postulated several obstacles to the development of algorithms to solve p1 (bard and moore, 1990). they established 3 observations for general mixed integer programming problems. fathoming in normal linear programming scenarios presents problems and follows 3 observations. observation1: the solution of the relaxed blpp does not provide a valid upper bound on the solution of the mixed integer blpp. observation2: solutions to the relaxed blpp that are in the inducible region cannot, in general, be fathomed. observation3: not all integer solutions to the relaxed blpp with some of the follower’s variables restricted can, in general, be fathomed. in the blpp, unfortunately, only observation1 can be applied with any degree of confidence. observation 2 needs some strong qualification and observation3 must be discarded altogether. to initialize the cutting plane procedure, first introduce some notations: with any loss of the generality let all variables are binary, and n the number of binary variables of upper level problem. k: the order number of a generated node in a branchand-cut tree: jk 0 ={j | xj is a free binary variable, j=1,2,…,n} jk + ={j | xj is a fixed at 1, j=1,2,…,n} jk – ={j | xj is a fixed at 0, j=1,2,…,n} int. j. prod. manag. eng. (2017) 5(1), 1-10creative commons attribution-noncommercial-noderivatives 4.0 international an effective branch-and-cut algorithm in order to solve the mixed integer bi-level programming 3 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3. branch and cut algorithm one common route with many classes of mathematical programs for achieving global optimality in the branch and bound and branch and cut algorithm, is the development of a bounding strategy. based on the observations, the bounding, fathoming, and branching procedures employed in traditional lp-based branch-and-bound algorithms is not applicable in a straightforward way. in this section, we use previous works that describe how to overcome these challenges to develop a generalized branch-and-cut algorithm for miblp that follows the same basis used in milp. 3.1. bounding wen in (yang, 1990) proved the following lemmas: lemma 1: given two linear programming problems: p2: max z2=cx p3: max z3=cx s.t. ax ≤ b s.t. ax ≤ b+θ x ≤ 0 x ≤ 0 where, θ is a parameter vector. let z2 * be the optimal objective value of p2 , v2 * the dual optimal solution of p2 , z3 * the optimal objective value of p3 ; and v3 * the dual optimal solution of p3 . then z3 * ≤ z2 * +v2 * θ proof: see (yang, 1990) lemma 2: the optimal value of the leader’s objective function in the p1 is less than or equal to the optimal objective function value in the following problem p4. p4: max z2=c 1x+d1y s.t. a1x+b1y ≤ g1 a2x+b2y ≤ g2 ,x x y y∈ ∈ proof: see (yang, 1990) theorem 1: consider the following problem p5( x ): p5( x ): max z5=d 1y s.t. a1x+b1y ≤ g1 a2x+b2y ≤ g2 y≥0 let z5 * be the optimal objective function value for the problem p5 and v5 * the optimal dual solution of the problem p5 . then the following upper bound,z5 u is established for the leader’s objective function value in problem p1 when x = x is fixed. 0 5 5 5 5( ) max{( ), 0} k k u j j j j j j j j z z c v a c v a + ∗ ∗ ∗ ∈ ∈ = + − + −∑ ∑ that is z5 u≥d1y*+c1x* where aj is the j th column vector of the matrix 1 2 a a       . proof: see (yang, 1990) generating valid inequalities there is two more observation, which is related to feasible cuts: observation4: if an inequality is valid for set ω, it is also valid for the main bi-level problem, i.e. set r observation5: let (x,y) ∈ ω, but (x,y) is not bi-level feasible (i.e. y ∉m(x)), then if one inequality is valid for ω -{(x,y)}, it is also valid for the main bi-level problem (r) because of the relationship ω⊆r, observation4 is derived. so, we can remove fractional solutions which are lp (removal of the lower-level optimality and integrally restrictions) resulting from the r; and based on observation5 we can separate points from the r that are integer but not bi-level feasible. for generating valid cuts, first of all, the relaxation problem p5 is solved, let (x,y) be the optimum solution to the p5 then if the solution is integer feasible, then feasibility condition (y ∈ m (x)) must be verified. this is done by solving the lower-level problem with the fixed upper-level solution x. assume the solution is ŷ, if d 2ŷ=d 2 y then (x,y) is bi-level feasible solution, otherwise in the next step, an inequality removing (x,y) from ω. the point (x, ŷ) is bi-level feasible and provides a valid lower bound on the optimal solution value of the original iblp. in the case of d 2ŷ > d 2 y (x,y) is not bi-level feasible, because it does not satisfy y ∈ m( x ) and we may still use (x, ŷ) to bound int. j. prod. manag. eng. (2017) 5(1), 1-10 creative commons attribution-noncommercial-noderivatives 4.0 international rahmani, a. and yousefikhoshbakht, m. 4 http://creativecommons.org/licenses/by-nc-nd/4.0/ the original problem, but we would like to add an inequality to p5 that is valid for p1 and violated by (x, y). the following simple procedure shows how to generate such an inequality. let (x,y) be a feasible point in upper and lower levels constraints without integer constraints and s be the set of constraints binding at (x,y) , then following cut is valid for the main bi-level problem p1. 1fx ey g+ ≤ − which , ,i i i i s i s i s f a e b g g ∈ ∈ ∈ = = =∑ ∑ ∑ and ai , bi , gi are the coefficient of x, y and right hand side respectively. 3.2. branching the algorithm which delivered by moore and bard (bard and moore, 1990), is forced to branch after producing an infeasible integer solution but here we are free to employ the well-developed branching strategies used in traditional algorithms for ilp, such as pseudo-cost branching, or the recently introduced reliability branching (achterberg et al., 2005). a branching technique for bi-level problems is discussed in following paragraph. let a solution of p5 be in hand and 0 ki j∈ i.e. xi be a free variable, from theorem 1, an upper bound is obtainable by 5 uz . now, let 5 uz + , 5 uz − be the value of 5 uz for 1ix = or 0ix = respectively. the upper bound is 5 uz = max { 5 uz + , 5 uz − }. as part of the iterative process, the upper bounds have to be checked against the current upper bound on the objective function z*. if the upper bound on that particular branch is not greater than the current best solution then that branch fathomed. now we propose the following algorithm to solve the mixed integer bi-level linear programming problem. 4. the proposed algorithm the algorithm depends heavily on the preceding observations, lemmas and theorem. especially the relaxation of the problem from a two-level problem to a simple mip, which is easy and quickly solvable compared to solving a complex bi-level linear programming problem. establishing the relaxed problems is the first priority of the algorithm and is completed in steps 1 and 2. this provides a lower bound for the problem by fixing all leaders’ binary variables to zero, in both the leader’s objective function and the follower’s objective function. by the way the follower’s objective function does not contain any leader’s binary variables. therefore, all terms in follower’s objective function related to leader reduce to a constant at the time of optimization and hence will not affect optimum solution. this allows the ignoring of the leader’s variables in the formulation of the follower’s objective function. the algorithm outlined in 7 steps. step 1: initialization n = 0; k = 0 n is a place-keeper of the current level in the tree, k is the counter for evaluating nodes 0 {1, 2,..., }kj n= {}kj + = {}kj − = 0, 1, 2,...,jt j n= = this indicates that all the leader’s variables are free. step 2: relaxed solution let x is the solution related to the kth node. solve problem p5 with the fixed x and obtains y: 5 ( , )p x y→ . these results in 5z ∗ , the optimal objective function value, and 5v ∗ , the optimal dual solution. follower solution: solve follower problem by fixing x and obtain ŷ if the problem results in 1 1ˆ( , )z x y z ∗≥ ; then z1 *=z1 (x, ŷ ) otherwise 1 1z z ∗ ∗= step 3: branching calculate the upper bound of the leader’s objective function from the previous node, (k-1), and xn=0. this denoted as 5 u nz − . similarly calculates 5 u nz + where 1nx = . ,maxz z zn u n u n u 5 5 5 –= +" , and 1n nt t= + . step 4: cut generation generate the following cuts: z c x d y1 1 1#f+ +) 1fx ey g+ ≤ − int. j. prod. manag. eng. (2017) 5(1), 1-10creative commons attribution-noncommercial-noderivatives 4.0 international an effective branch-and-cut algorithm in order to solve the mixed integer bi-level programming 5 http://creativecommons.org/licenses/by-nc-nd/4.0/ which s is the set of constraints binding at 1( , )x y , , ,i i i i s i s i s f a e b g g ∈ ∈ ∈ = = =∑ ∑ ∑ and ai ,bi ,ci are the coefficient of x, y and right hand sight respectively. step 5: optimality check if 5 1 u nz z + ∗< then set tn=2; go to step 6. the next step requires that if the algorithm has arrived at a node at the bottom of the tree (there is no free variable) then it can proceed back up the tree, examining branches and their upper bounds along the way. each upper bound compared to the current best solution to determine whether the branch fathomed or must be considered further. this described in the next step. step 6: backtracking if tn=2 then set tn=0, n=n–1. if n=0 (i.e. we came back to the top of the tree, and all possible nodes are evaluated) go to step 7. else, tn=tn+1. if 1 0nx = then the upper bound 5 5 u u n nz z + = and 1 1nx = ; else 5 5 u u n nz z − = and 1 0nx = ; go to step 4. step 7: termination stop algorithm execution and output the solution. the following simple numerical example illustrates the algorithm. in this example we use the notation p(i,–j), it means that the variable xi is fixed to one, the variable xj is fixed to zero and the other variables are free. example 1: p6: max 15x1+2x2+20x3+10x4+10y1+15y2+20y3+5y4+12y5 max 5y1+3y2 +8y3+4y4+y5 6x1+5x2+10x3+12x4+6y1+3y2+9y3+2y4+2y5≤12 2x1+4x2+13x3+7x4+5y1+y2+3y3+3y4+y5≤19 3x1+8x2+9x3+9x4+10y1+5y2+6y3+4y4+6y5≤15 4x1+3x2+12x3+14x4+4y1+3y2+5y3+y4+6y5≤30 xi, yj ∈ {0,1} in the initialization phase let 1 2 3 4( , , , )x x x x x= are free and try to solve the following problem where x=0: 7 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 () : max10 15 20 5 12 6 3 9 2 2 12 (1) 5 3 3 19 (2) 10 5 6 4 6 15 (3) 4 3 5 1 6 30 (4) 0 1, (5) 0 1, (6) 0 1, (7) 0 1, (8) 0 1, (9) p y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y + + + + + + + + ≤ + + + + ≤ + + + + ≤ + + + + ≤ ≤ ≤ ≤ ≤ ≤ ≤ ≤ ≤ ≤ ≤ this gives (0,1,0.809,0,0.857) with the objective value 41.487 and dual solution v7 *()=(1.143,0,1.619,0,0,3.476,0,0,0) the related optimal solution for the following problem is y=(0,0,1,1,0) with the objective value 25. the binding constraints of this solution are: constraint one: (6x1+5x2+10x3+12x4+6y1+3y2+9y3+2y4+2y5≤12), constraint three: (3x1+8x2+9x3+9x4+10y1+5y2+6y3+4y4+6y5≤15) and the constraint six: (0≤y2≤1), so, f=(6,5,10,12)+(3,8,9,9), e=(6,3,9,2,2)+(10,5,6,4,6)+(0,1,0,0,0) and g=12+15+1, therefore the binding cut is: 9x1+13x2+19x3+21x4+16y1+9y2+15y3+6y4+8y5≤27 (10) and the objective cut is: 15x1+2x2+20x3+10x4+10y1+15y2+20y3+5y4+12y5≥26 (11) the first choice facing the algorithm is processing with x1=0 or x1=1. (our choice variable is random; one can use an appropriate heuristic to select sequence of variables like greedy algorithms) the choice is made dependent on the relative values of the upper bounds for each branch. in this particular case under examination these values are 7 = 41.476 uz − and 7 = 44.762 uz + . at this point it would be a useful exercise to show the development of these numbers. now consider that if 1 0x = then 0 {2, 3, 4}kj = and int. j. prod. manag. eng. (2017) 5(1), 1-10 creative commons attribution-noncommercial-noderivatives 4.0 international rahmani, a. and yousefikhoshbakht, m. 6 http://creativecommons.org/licenses/by-nc-nd/4.0/ {}kj + = , since all 7j jc v a ∗− ’s are negative excepting 1 7 1 3.286c v a ∗− = then 7 = 41.476 uz − , and if 1 1x = then 0 {2, 3, 4}kj = , {1}kj + = and 7 = 44.762 uz + . so, for the first iteration 0 {2, 3, 4}kj = , {1}kj + = and {}kj − = is considered. using (1, 0, 0, 0)x = the integer linear programming problem 7z , becomes: 7 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 (1) : max10 15 20 5 12 6 3 9 2 2 6 5 3 3 17 10 5 6 4 6 12 4 3 5 1 6 26 16 9 15 6 8 17 10 15 20 5 12 11 0 1j p y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y + + + + + + + + ≤ + + + + ≤ + + + + ≤ + + + + ≤ + + + + ≤ − − − − − ≤ − ≤ ≤ this gives (0,1,0.1,1,0) with the objective value 7.8 and dual solution 7 (1) (0.8, 0, 0,0,0, 0, 0,0.3,0,2.2, 0)v ∗ = , the related optimal solution for the follower problem is (0,1, 0,1, 0) with the objective value 35, so till now 35z ∗ = . the binding constraints of this solution are one, eights and tenth, so the next cuts are: 6x1+5x2+10x3+12x4+6y1+4y2+9y3+3y4+2y5≤9 15x1+2x2+20x3+10x4+10y1+15y2+20y3+5y4+12y5≥36 now, if 2 0x = then 0 {3, 4}kj = and {1}kj + = , then 5 = 44.762 uz − , and if 2 1x = then 0 {3, 4}kj = , {1, 2}kj + = and 7 = 28.09 uz + . so, for the next iteration (k=2) 0 {3, 4}kj = , {1}kj + = and {2}kj − = is considered. using (1, 0, 0, 0)x = the integer linear programming problem 7 (1, 2)p − , in the example, now becomes: 7 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 (1, 2) : max10 15 20 5 12 6 3 9 2 2 6 5 3 3 17 10 5 6 4 6 12 4 3 5 1 6 26 16 9 15 6 8 3 10 15 20 5 12 21 0 1j p y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y − + + + + + + + + ≤ + + + + ≤ + + + + ≤ + + + + ≤ + + + + ≤ − − − − − ≤ − ≤ ≤ this problem is infeasible and then this branch is fathomated. the backtracking can now take place. it will examine the node associated with x2=1 and conclude that since 7 28.09 uz + = , is less than the current z *=35, the node is fathomated. by examining the other nodes, the following results obtained: node k=3: 0 {3, 4}kj = , {}kj + = , {1, 2}kj − = , 7 24.8 uz + = , 7 41.47 uz − = , 35z ∗ = . node k=4: 0 {4}kj = , {}kj + = , {1, 2, 3}kj − = , 7 24.8 uz + = , 7 41.47 uz − = , 35z ∗ = . node k=5: 0 {}kj = , {}kj + = , {1, 2, 3, 4}kj − = , 7 41.47 uz − = , 35z ∗ = : the end of branch node k=6: 0 {4}kj = , {4}kj + = , {1, 2}kj − = , 7 23.19 uz + = , 35z ∗ = : fathoming node k=7: 0 {4}kj = , {3}kj + = , {1, 2}kj − = , 7 17 uz + = , 7 35.47 uz − = , 35z ∗ = : 7 ( 1, 2, 3)p − − is infeasible and fathomed. therefore, in this example, only 7 of the 30 nodes were considered, and only one of the possible 16 leaves was met. in compared to the branch and bound method [12] that 18 of the 30 nodes were considered, and 4 of the possible 16 leaves were formulated is promising. this measure will be further discussed in the computational results section. 5. computational results the branch-and-cut algorithm was implemented in aimms 12, utilizing cplex 11 as solver. to our knowledge, the best general algorithm proposed in the literature is the one of bard and moore (1990), beresnev (2013) or beresnev and melnikov (2014). a comprehensive comparison of these algorithms is not at hand. we also evaluated our algorithm on a set of problems in which the leader’s objective function int. j. prod. manag. eng. (2017) 5(1), 1-10creative commons attribution-noncommercial-noderivatives 4.0 international an effective branch-and-cut algorithm in order to solve the mixed integer bi-level programming 7 http://creativecommons.org/licenses/by-nc-nd/4.0/ variable coefficients were established randomly between –30 and +30. the follower’s objective function variable coefficients were placed between –12 and +12. the constraint matrix coefficients were all between –18 and +18 and the bj , or resource values were restricted to be within the range 0.5 to 0.75 of the sum of the aj for the j th constraint. the instances were classified based on the number of upper level variables and the number of lower level variables. in table 2, 10 randomly constructed problems were solved for each problem type and compared with an algorithm proposed in beresnev (2013). a larger sample size would be deemed statistically more significant. in these tables, constructed problems were randomly solved for each problem type, and a combination of n (the number of upper level binary variables) and m (the number of lower level binary variables) for n=5,10,15 and m=5,10,15 are considered. also, the following notations were used in the table 2. e.n: the number of evaluated nodes as a percentage of total nodes in the tree n.i: number of mip problems solved as a percentage of leaves in the tree n.o: the number of nodes where the optimal solution was obtained as a percentage of nodes in the tree av.t: the average cpu time (sec) for algorithms in order to compare the performance of the proposed methods, a set of test problems was generated as described in table 1. the instances were classified based on the number of potential facilities and the number of customers. table 1. characteristics of randomly generated problems. no prob. size total nodes leaves 1 5 5× 62 32 2 5 10× 62 32 3 5 15× 62 32 4 10 5× 2046 1024 5 10 10× 2046 1024 6 10 15× 2046 1024 7 15 5× 65534 32768 8 15 10× 65534 32768 9 15 15× 65534 32768 as shown in table 2, the number of iterations in branch and cut based method was less than the number of iterations in branch and bound algorithm and it was able to solve the problem faster because of using the appropriate cuts. it is well known that in regards to the time solution, the algorithm is superior that solves less mip cases, and usually enumeration methods are slow, because they encounter too many mip sub problems, in the above and based on our computational results we fairly reduce the mip sub problems and it let to achieve optimal solutions in more reasonable time. 6. conclusions through the paper some of the difficulties regarding to the solving mixed integer bi-level linear programming problems were described and a branch-and-cut algorithm proposed. the algorithm is based on two different cuts for mixed integer bilevel linear programming problems. the first one is the binding cut and the second is the objective cut. for the branching and fathoming rule, the extensions of an upper bound theorem of mip problem are table 2. results of 10 samples for each type problems. no. branch and cut based method branch and bound based method from (beresnev, branch-andbound algorithm for a competitive facility location problem, 2013) e.n n.i n.o av.t e.n n.i n.o av.t 1 33% 21% 14% 227 55% 39% 27% 513 2 45% 18% 15% 245 75% 73% 34% 678 3 37% 34% 21% 281 64% 44% 21% 691 4 15% 7% 4% 268 15% 7% 4% 839 5 38% 21% 7% 331 51% 41% 11% 880 6 43% 38% 11% 393 43% 38% 22% 818 7 22% 13% 7% 395 37% 27% 13% 1320 8 15% 11% 4% 442 43% 33% 25% 1495 9 21% 14% 5% 496 54% 48% 30% 1618 int. j. prod. manag. eng. (2017) 5(1), 1-10 creative commons attribution-noncommercial-noderivatives 4.0 international rahmani, a. and yousefikhoshbakht, m. 8 http://creativecommons.org/licenses/by-nc-nd/4.0/ applied. the first advantage of this approach is the ability to exploit the vast solvers for solving mixed integer bi-level linear programming problems. more than it, we believe that the proposed method has the ability of adopts itself with the other algorithm for improvement itself or the other algorithms that are good cases for developing the algorithm. besides that, one can using upper bound theorems to the general bi-level programming problem to develop the algorithm would seem to be the most logical course, or even works on primal heuristics, additional classes of valid inequalities, branching rules based on disjunctions involving more than one variable, and so on are good cases in the future works. competing interests the author(s) declare(s) that there is no conflict of interest regarding the publication of this paper. references achterberg, t., koch, t., martin, a. (2005). branching rules revisited. operations research letters, 33(1), 42-54. https://doi.org/10.1016/j. orl.2004.04.002 colson, b., marcotte, p., savard, g. (2005). bilevel programming: a survey. 4or, 3(2), 87-107. https://doi. org/10.1007/s10288-005-0071-0 bard, j.f., falk, j.e. (1982). an explicit solution to the multi-level programming problem. computer and operations research, 9(1), 77-100. https://doi.org/10.1016/0305-0548(82)90007-7 bard, j.f., moore, j.t.a. (1990). a branch and bound algorithm for the bilevel programming problem. siam journal on scientific and statistical computing, 11(2), 281-292. https://doi.org/10.1137/0911017 beresnev, v.l. (2009). upper bounds for objective functions of discrete competitive facility location problems. journal of applied and industrial mathematics, 3(4), 419-432. https://doi.org/10.1134/s1990478909040012 beresnev, v.l. (2013). branch-and-bound algorithm for a competitive facility location problem. computers & operations research, 40(8), 2062-2070. http://dx.doi.org/10.1016/j.cor.2013.02.023 beresnev, v.l., melnikov, a.a. (2014). branch-and-bound method for the competitive facility location problem with prescribed choice of suppliers, diskretn. anal. issled. oper., 21(2), 3-23. dempe, s. (2001). discrete bilevel optimization problems. technical report d-04109, institut fur wirtschaftsinformatik, universitat leipzig, leipzig, germany. denegre, s., ralphs, t.k. (2009). a branch-and-cut algorithm for bilevel integer programming. in proceedings of the 11th informs computing society meeting, 65-78. domínguez, l.f., pistikopoulos, e.n. (2010). multiparametric programming based algorithms for pure integer and mixed-integer bilevel programming problems. computers and chemical engineering, 34(12), 2097-2106. https://doi.org/10.1016/j. compchemeng.2010.07.032 floudas, c.a., pardalos, p.m., adjiman, c., esposito, w.r., gümüş, z.h., harding, s.t., klepeis, j.l., meyer, c.a., schweiger, c. a. (2013). handbook of test problems in local and global optimization (vol. 33). springer science & business media. gümüş, z.h., floudas, f. (2005). global optimization of mixed-integer bilevel programming problem. computational management science, 2(3), 181-212. https://doi.org/10.1007/s10287-005-0025-1 hansen, p., jaumard, b., savard, g. (1992). new branch-and-bound rules for linear bilevel programming. siam journal on scientific and statistical computing, 13(5), 1194-1217. https://doi.org/10.1137/0913069 jeroslow, r.g. (1985). the polynomial hierarchy and a simple model for competitive analysis. mathematical programming, 32(2), 146-164. https://doi.org/10.1007/bf01586088 mirhassani, s.a., raeisi, s., rahmani, a. (2015). quantum binary particle swarm optimization-based algorithm for solving a class of bi-level competitive facility location problems. optimization methods and software, 30(4), 756-768. https://doi.org/10.1080/10556788.201 4.973875 pistikopoulos, e.n, georgiadis, m.c, dua, v. (2007). multi-parametric programming: theory, algorithms, and applications, volume 1, weinheim: wiley-vch, 1. https://doi.org/10.1002/9783527631216 rahmani, a., mirhassani, s. a. (2015). lagrangean relaxation-based algorithm for bi-level problems. optimization methods and software, 30(1), 1-14. https://doi.org/10.1080/10556788.2014.885519 shi, c., lu, j., zhang, g., zhou, h. (2006). an extended branch and bound algorithm for linear bilevel programming. applied mathematics and computation, 180(2), 529-537. https://doi.org/10.1016/j.amc.2005.12.039 vicente, l. (2001). bilevel programming: introduction, history and overview. in c. a. floudas and p. m. pardalos (eds.) encyclopedia of optimization, 178-180. springer: us. https://doi.org/10.1007/0-306-48332-7_38 int. j. prod. manag. eng. (2017) 5(1), 1-10creative commons attribution-noncommercial-noderivatives 4.0 international an effective branch-and-cut algorithm in order to solve the mixed integer bi-level programming 9 https://doi.org/10.1016/j.orl.2004.04.002 https://doi.org/10.1016/j.orl.2004.04.002 https://doi.org/10.1007/s10288-005-0071-0 https://doi.org/10.1007/s10288-005-0071-0 https://doi.org/10.1016/0305-0548(82)90007-7 https://doi.org/10.1137/0911017 https://doi.org/10.1134/s1990478909040012 http://dx.doi.org/10.1016/j.cor.2013.02.023 https://doi.org/10.1016/j.compchemeng.2010.07.032 https://doi.org/10.1016/j.compchemeng.2010.07.032 https://doi.org/10.1007/s10287-005-0025-1 https://doi.org/10.1137/0913069 https://doi.org/10.1007/bf01586088 https://doi.org/10.1080/10556788.2014.973875 https://doi.org/10.1080/10556788.2014.973875 https://doi.org/10.1002/9783527631216 https://doi.org/10.1080/10556788.2014.885519 https://doi.org/10.1016/j.amc.2005.12.039 https://doi.org/10.1007/0-306-48332-7_38 http://creativecommons.org/licenses/by-nc-nd/4.0/ vicente, j., savard, l., judice, g. (1996). discrete linear bilevel programming problem. journal of optimization theory and applications, 89(3) 597-614. https://doi.org/10.1007/bf02275351 wen, u., yang. y. (1990). algorithms for solving the mixed integer two-level linear programming problem. computers & operations research, 17(2), 133-142. https://doi.org/10.1016/0305-0548(90)90037-8 int. j. prod. manag. eng. (2017) 5(1), 1-10 creative commons attribution-noncommercial-noderivatives 4.0 international rahmani, a. and yousefikhoshbakht, m. 10 https://doi.org/10.1007/bf02275351 https://doi.org/10.1016/0305-0548(90)90037-8 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2018.8913 received 2017-11-06 accepted: 2018-01-17 a milp for multi-machine injection moulding sequencing in the scope of c2net project beatriz andrésa1*, raquel sanchisa2, raúl polera3, manuel díaz-madroñedoa4 and josefa mulaa5 a research centre on production management and engineering (cigip), universitat politècnica de valència, calle alarcón, n°1, alcoy, 03801 alicante, spain a1* bandres@cigip.upv.es, a2 rsanchis@cigip.upv.es, a3 rpoler@cigip.upv.es, a4 fcodiama@cigip.upv.es, a3 fmula@cigip.upv.es abstract: the goal of c2net european h2020 funded project is the creation of cloud-enabled tools for supporting the smes supply network optimization of manufacturing and logistic assets based on collaborative demand, production and delivery plans. in the scope of c2net project, and particularly in the optimisation module (c2net opt), this paper proposes a novel holistic mixed integer linear programing (milp) model to optimise the injection sequencing in a multi-machine case. the results of the milp will support the production planner decision-making process in the calculation of (i) moulds setup in certain machines, and (ii) the amount of products to produce in order to minimise the setup, inventory, and backorders costs. the designed milp takes part of the algorithms repository created in c2net european funded project to solve realistic industry planning problems. the milp is verified in realistic data considering three data sets with different sizes, in order to test its the computation efficiency. key words: milp, sequencing, injection, moulds, multi-machine, automotive. 1. introduction the current situation in the european industrial park highlights the restricted access to advanced management systems and collaborative tools for small and medium sized enterprises (smes), being this tools limited to large enterprises due to the high levels of costs and complexity associated (european commission, 2013). nevertheless, the importance of having a applying this tools within the smes is considered a key factor, since the utilisation of advanced management tools systems and services will enhance enterprises competitiveness by increasing their agility and adaptability to deal with rapid evolutions of existing and future markets (european commission, 2008). smes manufacturing value chains are distributed and dependent on complex information and material flows requiring new approaches to reduce the complexity of manufacturing management systems. they need ubiquitous tools supporting collaboration among value chain partners and providing advanced algorithms to achieve holistic global and local optimization of manufacturing assets and to respond faster and more efficiently to unforeseen changes. c2net aims to collect the most promising results from r&d eu framework program projects and build a novel cloud architecture to provide smes affordable tools (in term of cost and usability) to help them to overcome the current economic crisis and to enhance their competitiveness in the global economy. in general, c2net (cloud collaborative manufacturing networks) european project aims at to cite this article: andrés, b., sanchis, r., poler, r., díaz-madroñedo, m., mula, j. (2018). a milp for multi-machine injection moulding sequencing in the scope of c2net project. international journal of production management and engineering, 6(1), 29-36. https://doi.org/10.4995/ijpme.2018.8913 int. j. prod. manag. eng. (2018) 6(1), 29-36creative commons attribution-noncommercial-noderivatives 4.0 international 29 mailto:bandres@cigip.upv.es mailto:rsanchis@cigip.upv.es mailto:rpoler@cigip.upv.es mailto:fcodiama@cigip.upv.es mailto:fmula@cigip.upv.es http://creativecommons.org/licenses/by-nc-nd/4.0/ the creation of cloud-enabled tools for supporting smes supply network in the optimisation of their manufacturing and logistic planning (andres et al., 2016c) based on the jointly planning, coordination and integration of processes, participating all network entities. due to the current characteristics of uncertainty in the markets and economic crisis, there is a need to encourage collaboration tools to reduce costs and increase trust and accountability to market requirements. this study presents an overview of the research carried out in the h2020 european project: cloud collaborative manufacturing networks (c2net. 1.1. c2net objectives specifically, the goal of c2net project is the creation of cloud-enabled tools for supporting the smes supply network optimization of manufacturing and logistic assets based on collaborative demand, production and delivery plans. c2net project will provide a scalable real-time architecture, platform and software to allow the supply network partners: to master complexity and data security of the supply network, to store and share product, process and logistic data, to optimize the manufacturing assets by the collaborative computation of production plans, to optimize the logistics assets through efficient replenishment, manufacturing and delivery plans and to render the complete set of supply chain management information on the any digital mobile device (pc, tablets, smartphones) of decision makers enabling them to monitor, visualize, control, share and collaborate. the specific technical and functional objectives of the project are described next: implementation of c2net data collection framework (c2net dcf) for iot-based continuous data collection from supply network resources. implementation of c2net optimizer (c2net opt) for the optimization of manufacturing and logistics assets of the supply network by the collaborative computation of production, replenishment and delivery plans. implementation of c2net collaboration tools (c2net cot) for providing support to the collaborative processes of the supply network. implementation of c2net cloud platform (c2net cpl) to integrate the data module, the optimizers and the collaborative tools in the cloud. to provide new ways to securely store relevant information from supply network partners in public cloud, community cloud or private cloud, ensuring data security and provide the ability of data sharing and data analytics. to provide new tools for supporting the supply network optimization of manufacturing and logistic assets. to provide new tools to support decision-makers in collaborative processes with intuitive uis that display in real-time the right data at the right time, built in advanced apps for mobile devices for selective management functionalities across the supply network. to provide a new cloud space for allowing the access to all the participants in the value chain to support their decisions and processes enhancement. 2. c2net results and impacts globalisation and new types of innovation processes have over the last few decades reshaped in new ways the organisation of value chains. smes have an incentive to identify emerging activities that will give them a new competitive advantage. cooperation within a network may be a sensible strategy for preventing the decay of their traditional specialisation (commission, 2012). networks of firms are structures specifically created for active collaboration (camarinha-matos and afsarmanesh, 2005). this collaboration could be open-ended or focused on a specific project task. they may or may not be confined to a specific geographical location and set of industries (camarinha-matos and afsarmanesh, 2008). enterprises belonging to a network make autonomous decisions, according int. j. prod. manag. eng. (2018) 6(1), 29-36 creative commons attribution-noncommercial-noderivatives 4.0 international andrés, b., sanchis, r., poler, r., díaz-madroñedo, m. and mula, j. 30 http://creativecommons.org/licenses/by-nc-nd/4.0/ to what is in their best interest, being sometimes difficult to be inside the network when contradictions among enterprises appear (andres and poler, 2016). in the global economy, there is a growing interest in new organisational structures, which are flexible enough to respond to market changes and at the same time solid enough, with strategies and objectives aligned, to take on cooperative projects. networks have been identified as crucial instruments for implementing the eu’s europe 2020 strategy. c2net fulfils the european industry needs providing advanced methods and tools to address economic and risk assessments in order to support complex decision-making as the management of co-evolution of products-services and the related production systems, the evaluation of alternative configurations of the network of actors involved in the global supply chain, or integration of new technologies in the factory. in the light of this, c2net architecture is proposed with the aim of providing affordable tools to smes in order to deal with complex management systems and establish stable and sustainable collaborative processes within the enterprises of the network. 2.1. c2net architecture c2net architecture is composed by the cloud platform (c2net cpl), the data collection framework (c2net dcf), the optimizer (c2net opt) and the collaboration tools (c2net cot) (figure 1). the c2net cpl will provide cloud access to information collected from heterogeneous and distributed sources to create global and local production plans to optimize the processes using the data received from customers, suppliers and manufacturers. this will result on faster and more efficient decision making which have to be made figure 1. c2net architecture (lauras et al., 2015) (andres et al., 2016c). int. j. prod. manag. eng. (2018) 6(1), 29-36creative commons attribution-noncommercial-noderivatives 4.0 international a milp for multi-machine injection moulding sequencing in the scope of c2net project 31 http://creativecommons.org/licenses/by-nc-nd/4.0/ due to market changes, high competition and customization requirements. c2net cpl will allow collaborative production as production, distribution, supply and customers plans will be calculated based on real time information coming from real-world resources and considering all the actors involved in the process. as information will be available in the cloud, c2net cpl will enable the possibility to provide relevant information to different personnel using mobile devices when and where is needed, to increase the capability for better and faster decision making when changes in production need to be done. the c2net dcf will enable real time information collection from the different value chain enterprise systems and/or physical devices, thus providing the c2net platform with a holistic view on the network. this will stimulate a fast knowledge feedback loop, which will enable companies to become more effective and react faster to market changes. the c2net opt will provide advanced optimization algorithms for the collaborative computation of production, replenishment and delivery plans with the aim of optimize the use of manufacturing and logistics assets of the supply network from a holistic point of view. the c2net cot will propose a concrete solution to support the collaborative value chain by facilitating the diagnosis of any source of divergence of the collaboration with regard to expected situation. moreover the c2net cot will be able to support the adaptation of the stakeholders’ behaviours by implementing reaction mechanisms based on global and local optimization algorithms. for more information about each c2net module we refer readers to (andres et al., 2016c) 3. c2net industry validation c2net project is directly connected with the industrial application perspective. it is designed to comprehensively cover real entire supply chains considering all stages of manufacturing, distribution and sales to supply a product to market. c2net project has the purpose to connect the industrial, the research and the development perspectives. to this end four industrial pilots are considered in the project, belonging to different industrial sectors, including the automotive, dermo-cosmetics, metalworking smes and oem production of hydraulic and lubrication systems. the automotive pilot centers its attention in the collaborative production planning and synchronized replenishment of car components. the dermo cosmetics pilot focuses on collaboration for an agile production and distribution of dermo-cosmetics. the metalworking sme’s network deals with sheet metal supply chain to metalworking industry. finally, the oem focuses on the sub supplier value stream collaboration for product on time delivery (otd) for customers. the validation of the c2net cloud services in the industry will not only serve for the verification of the c2net architecture but also for validation of the integrated modules when implemented in real networks of enterprises. the implementation c2net cloud services will provide developers the required feedback for the platform improvement. sharing the experiences and best practices will contribute to the adjustment and refinement of c2net, and will contribute to the identification of avenues of exploitation, and strengths and weaknesses of the cloud platform according to the expectations and needs of the intended final users. this paper is going to focus in the automotive pilot, to study the needs that the enterprises belonging to this pilot have in their production plans, and how the literature plans cover the potential integration of collaborative replenishment, production and delivery plans. 4. the automotive pilot the automotive pilot is framed in the automotive industrial sector and deals with collaborative production planning and synchronised replenishment of car components. the studied pilot consists of the first and second-tiers fully dedicated to the car assembly. the automotive industrial pilot comprises two manufacturing companies of the automotive supply chain. one is a first tier and the other its supplier, a second tier (figure 2). the first and second-tier supplier will use the c2net cloud infrastructure, platform and software to optimally calculate both, individual and collaborative production, replenishment and delivery plans. enterprises of automotive pilot will int. j. prod. manag. eng. (2018) 6(1), 29-36 creative commons attribution-noncommercial-noderivatives 4.0 international andrés, b., sanchis, r., poler, r., díaz-madroñedo, m. and mula, j. 32 http://creativecommons.org/licenses/by-nc-nd/4.0/ share relevant information and will use the c2net advanced algorithms to optimise production, supply and replenishment plans. 4.1. problem description the research work carried out in these papers (andres et al., 2016a, 2016b, 2017, 2018; orbegozo et al., 2018) identified concrete solutions proposed in the literature to support the (i) individual production, replenishment and delivery planning; and (ii) collaborative production and replenishment plans, collaborative production and delivery plans, collaborative replenishment and delivery plans and collaborative replenishment, production and delivery plans. the literature plans were compared with the real requirements collected from the industry. in this regard, the needs of all the industrial pilots of c2net project are analysed. the comparison allowed identifying some gaps between the literature and the industrial needs regarding individual and collaborative replenishment, production and distribution planning. this paper focuses on the automotive pilot, and a set of plans, with interest of study due to their complexity and the lack of automation (so far are manually solved), have been identified within the automotive pilot, including the injection replenishment plan, emergency replenishment plan, master injection production plan, spare parts production plan, injection scheduling plan, painting scheduling plan and delivery plan. furthermore, among the recorded plans there have been identified which ones are willing to be collaboratively performed within the two enterprises covering the pilot, according to the enterprises needs and criteria. c2net project has focused on identifying the needs and requirements of both companies in the automotive pilot when dealing with their production, replenishment and delivery plans. based on the requirements’ analysis of the first and second-tier suppliers of the automotive pilot (whose production is based on injection moulding) the need of a multi-machine injection moulding sequencing optimisation was detected for reducing the setup costs and forestalling product delays, by not only using the stocks, but also by obtaining a more efficient injection sequencing plan. according to the needs identified in the enterprises of the automotive pilot, and considering the results of the literature carried out, it can be concluded that the algorithms found in the literature will require adaptations before they can be applied in the optimisation cases of the automotive pilot because both the amount and nature of the inputs, objectives and outputs have very low interdependence with the industry requirements. in order to better understand the calculation of the multi-machine injection moulding sequencing plan in the automotive suppliers, the process carried out by second-tier supplier planner is explained next, in the same way as the enterprise computes it currently. the second-tier supplier receives a new demand plan from the first-tier supplier. in order to fulfil the demand of the first-tier supplier, the company planner has to compute the quantity of products to produce, the moulds with the cavities of the demanded product, and the machines in to which setup the mould and carry out the injection process. the second-tier supplier planner has to schedule the products in a set of 20 different machines. in each machine can be setup a specific mould, which through the process of injection generates a specific product. the multi-machine injection moulding sequencing plan is calculated by the planner in a in a spreadsheet sequencing first one matching and considering the sequence of one machine after another. the secondtier supplier planner tries to minimise the products delayed, to the extent that the missing deliveries per million (mpm) equals zero. in the automotive sector, to deliver the products in the quantity and time periods demanded is a key factor, due to the oem works in a just in time (jit) regime. a delayed product could involve a stop in the production line of the oem and consequently millenary losses. in order to avoid this problem, the second-tier supplier works in a 3 days of coverage, of the demand plan released by the first-tier supplier. at the same time, the company planner tries to reduce as much as possible the inventory, but this oem (0) original equipment manufacturer first tier (1) second tier (2) figure 2. c2net automotive pilot. int. j. prod. manag. eng. (2018) 6(1), 29-36creative commons attribution-noncommercial-noderivatives 4.0 international a milp for multi-machine injection moulding sequencing in the scope of c2net project 33 http://creativecommons.org/licenses/by-nc-nd/4.0/ is a hard task, due to the inventory is used to deal with the continuous changes on the demand, which is a common pattern in the automotive sector (diazmadroñero et al., 2018). although the second-tier supplier is interested in minimising the costs of the multi-machine injection moulding sequencing plan, no optimisation is performed and the process is manually solved. according to this, the main aim is to provide a milp to optimally solve the multi-machine injection moulding sequencing plan in the second-tier supplier. the main aim is to include the proposed milp to compute the multi-machine injection moulding sequencing plan in the (c2net opt), connected with the continuous data collection services (c2net dcf) and embedded in the cloud platform (c2net cpl). 5. a multi-machine milp for injection sequencing a milp is proposed to deal with a multi-machine injection moulding sequencing plan, which has 4 indexes for modelling: the products (i), moulds (j), machines (k) and periods (t). the parameters and decision variables are described in table 1. the objective function (equation 1) minimises the sum of the cost of setting up the mould j in a machine k, the inventory and backorder costs of product i. table 1. nomenclature of the multimachine milp. index i j k t set of products (p) set of moulds (o) set of machines (m) set of periods (t) parameters cij cat dit invmaxi invmini invi0 cii csti cdi csj nj nst production rate of the product i using the mould j capacity of production in hours for period t demand of product i during period t maximum available inventory capacity for product i minimum inventory for product i initial inventory of product i inventory cost of product i stockout cost of product i delay cost of one unit of product i setup cost of the mould j amount of mould j available maximum amount of setups allowed in period t decision variables yjkt sjkt pit invit dit 1 if the mould j is producing products in the machine k in the period t, 0 otherwise 1 if the mould j starts producing products in the machine k in the period t and in period t-1 if it was another mould j setup in the machine k, 0 otherwise amount produced of product i in the period t inventory level of product i at the end of period t delay amount of product i at the end of period t int. j. prod. manag. eng. (2018) 6(1), 29-36 creative commons attribution-noncommercial-noderivatives 4.0 international andrés, b., sanchis, r., poler, r., díaz-madroñedo, m. and mula, j. 34 http://creativecommons.org/licenses/by-nc-nd/4.0/ 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 = ���𝑐𝑐𝑐𝑐𝑗𝑗 · 𝑆𝑆𝑗𝑗𝑗𝑗𝑗𝑗 𝑗𝑗𝑗𝑗 + ��𝑐𝑐𝑀𝑀𝑀𝑀 · 𝐼𝐼𝐼𝐼𝐼𝐼𝑀𝑀𝑗𝑗 𝑗𝑗𝑀𝑀 + ��𝑐𝑐𝑐𝑐𝑀𝑀 · 𝐷𝐷𝑀𝑀𝑗𝑗 𝑗𝑗𝑀𝑀𝑗𝑗 (1) constrains are presented in equations 2 – 10: in each period t one or any mould j can be producing (2) in each machine k it can only be setup the amount of moulds j available (3) production of product i in period t, constrained by the capacity (4) first activation of mould j in machine k at first period t=1 𝑆𝑆𝑗𝑗𝑗𝑗𝑗𝑗 = 𝑌𝑌𝑗𝑗𝑗𝑗𝑗𝑗 ∀ 𝑗𝑗, 𝑗𝑗, 𝑗𝑗 = 1 (5) setup restriction: 1 when the mould j in machine k is active in the period t, and it was not active at period (t-1), 0 if the mould j in machine k 𝑆𝑆𝑗𝑗𝑗𝑗𝑗𝑗 ≥ 𝑌𝑌𝑗𝑗𝑗𝑗𝑗𝑗 − 𝑌𝑌𝑗𝑗𝑗𝑗𝑗𝑗 −1 ∀ 𝑗𝑗, 𝑗𝑗, 𝑗𝑗 > 1 (6) mould j setups in machine k is limited by the maximum amount of setups allowed in period (7) inventory balance and bounds 𝐼𝐼𝐼𝐼𝐼𝐼𝑀𝑀𝑗𝑗 = 𝐼𝐼𝐼𝐼𝐼𝐼𝑀𝑀0 + 𝑃𝑃𝑀𝑀𝑗𝑗 − 𝑐𝑐𝑀𝑀𝑗𝑗 + 𝐷𝐷𝑀𝑀𝑗𝑗 ∀ 𝑀𝑀, 𝑗𝑗 = 1 (8) 𝐼𝐼𝐼𝐼𝐼𝐼𝑀𝑀𝑗𝑗 = 𝐼𝐼𝐼𝐼𝐼𝐼𝑀𝑀𝑗𝑗−1 + 𝑃𝑃𝑀𝑀𝑗𝑗 − 𝑐𝑐𝑀𝑀𝑗𝑗 + 𝐷𝐷𝑀𝑀𝑗𝑗 − 𝐷𝐷𝑀𝑀𝑗𝑗−1 ∀ 𝑀𝑀, 𝑗𝑗 ≠ 1 (9) 𝐼𝐼𝐼𝐼𝐼𝐼𝑀𝑀𝐼𝐼𝐼𝐼𝑀𝑀 ≤ 𝐼𝐼𝐼𝐼𝐼𝐼𝑀𝑀𝑗𝑗 ≤ 𝐼𝐼𝐼𝐼𝐼𝐼𝑀𝑀𝐼𝐼𝐼𝐼𝑀𝑀 ∀ 𝑀𝑀, 𝑗𝑗 (10) the main index of the model is the mould j, allowing to sequence moulds in a multi-machine environment. the number of mould setups in each period is limited by the capacity constrains related to the availability of moulds (3). the validation considers 3 different size of realistic data to verify the results, solving realistic problems in reasonable time. the small dataset (i=2, j=2, k=2 and t=2) needs less than 1 sec to be optimised, the medium dataset (i=29, j=10, k=18 and t=14) is solved in 60 s (gap=1.7×10-5). finally, the large dataset (i=100, j=100, k=18 and t=50) is computed in 18 h (gap=0,168). 6. conclusions in this paper, the c2net european h2020 funded project is described, focusing on its objectives, architecture, results and impacts resulting from the implementation of the results obtained: c2net cloud platform. the production planning problem has been identified within the automotive pilot, taking part in c2net project. the research carried out has allowed to state that to the best of our knowledge, the problem identified has not been addressed in the literature due to the particularities handled by the second-tier supplier analysed. in this regard, and with the aim of generating novel solutions in the scope of the c2net opt, a new milp model for multi-machine injection moulding sequencing is proposed. the new model considers the demand backorders, the setups limits per time period and the security stocks defined by the second-tier supplier to face the demand variations to which it is subject. the validation of the milp model has been performed with three sets of data: small, medium and large, in order to confirm the computational efficiency levels and corroborate its applicability in a real enterprise. future developments are lead to include new restrictions related with machine setups, route priorities in machines, and setup dependencies when performing carrying out a change over from one product to another. moreover, it would be useful to include restrictions related with the materials requirement plan in order to sequence the final products, whose raw materials are available in the inventory. finally, the proposed novel milp model for multi-machine injection moulding sequencing will be implemented and validated in an enterprise, considering real data. acknowledgements the research leading to these results is in the frame of the “cloud collaborative manufacturing networks” (c2net) project which has received funding from the european union’s horizon 2020 research and innovation programme under grant agreement no 63690. int. j. prod. manag. eng. (2018) 6(1), 29-36creative commons attribution-noncommercial-noderivatives 4.0 international a milp for multi-machine injection moulding sequencing in the scope of c2net project 35 http://creativecommons.org/licenses/by-nc-nd/4.0/ references andres, b., poler, r. (2016). a decision support system for the collaborative selection of strategies in enterprise networks. decision support systems, 91, 113-123. https://doi.org/10.1016/j.dss.2016.08.005 andres, b., poler, r., saari, l., arana, j., benaches, j.v., salazar, j. (2018). optimization models to support decision-making in collaborative networks : a review, in: closing the gap between practice and research in industrial engineering, lecture notes in management and industrial engineering. 249-258. https://doi.org/10.1007/978-3-319-58409-6_28 andres, b., saari, l., lauras, m., eizaguirre, f. (2016a). optimization algorithms for collaborative manufacturing and logistics processes, in: zelm, m., doumeingts, g., mendonça, j.p. (eds.), enterprise interoperability in the digitized and netwroked factory of the future. iste, 167-173. andres, b., sanchis, r., lamothe, j., saari, l., hauser, f. (2016b). combined models for production and distribution planning in a supply chain, in: building bridges between researchers and practitioners. book of abstracts of the international joint conference cio-icieomiise-aim (ijc2016), 71. andres, b., sanchis, r., lamothe, j., saari, l., hauser, f. (2017). integrated production-distribution planning optimization models : a review in collaborative networks context. international journal of production management and engineering, 5, 31-38. https://doi.org/10.4995/ ijpme.2017.6807 andres, b., sanchis, r., poler, r. (2016c). a cloud platform to support collaboration in supply networks. international journal of production management and engineering, 4, 5-13. https://doi.org/10.4995/ijpme.2016.4418 camarinha-matos, l.m., afsarmanesh, h. (2008). collaborative networks: reference modelling. springer international publishing. camarinha-matos, l.m., afsarmanesh, h. (2005). collaborative networks : a new scientific discipline. journal of intelligent manufacturing, 16(4-5), 439-452. https://doi.org/10.1007/s10845-005-1656-3 commission, e. (2012). european comission [www document]. eur. compet. report. 2012. reaping benefits glob. comm. staff work. doc. swd(2012)299 final. diaz-madroñero, m., mula, j., andres, b., poler, r., sanchis, s. (2018). capacitated lot-sizing and scheduling problem for second-tier suppliers in the automotive sector, in: closing the gap between practice and research in industrial engineering. springer international publishing, 121-129. https://doi.org/10.1007/978-3-319-58409-6 european commission (2008). work programme. cooperation theme 4 nanosciences, nanotechnologies, materials and new production technologies – nmp (european commission c (2007) 5765 of 29 november 2007). european commission (2013). fact and figures about the eu's small and medium enterprise (sme) [www document]. lauras, m., lamothe, j., benaben, f., andres, b., poler, r. (2015). towards an agile and collaborative platform for managing supply chain uncertainties. in: international ifip working conference on enterprise interoperability, 64–72. https://doi.org/10.1007/978-3-66247157-9_6 orbegozo, a., andres, b., mula, j., lauras, m., monteiro, c., malheiro, m. (2018). an overview of optimization models for integrated replenishment and production planning decisions, in: closing the gap between practice and research in industrial engineering, lecture notes in management and industrial engineering, 239-247. https://doi.org/10.1007/978-3-319-58409-6_27 int. j. prod. manag. eng. (2018) 6(1), 29-36 creative commons attribution-noncommercial-noderivatives 4.0 international andrés, b., sanchis, r., poler, r., díaz-madroñedo, m. and mula, j. 36 https://doi.org/10.1016/j.dss.2016.08.005 https://doi.org/10.1007/978-3-319-58409-6_28 https://doi.org/10.4995/ijpme.2017.6807 https://doi.org/10.4995/ijpme.2017.6807 https://doi.org/10.4995/ijpme.2016.4418 https://doi.org/10.1007/s10845-005-1656-3 https://doi.org/10.1007/978-3-319-58409-6 https://doi.org/10.1007/978-3-662-47157-9_6 https://doi.org/10.1007/978-3-662-47157-9_6 https://doi.org/10.1007/978-3-319-58409-6_27 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering doi:10.4995/ijpme.2017.5857 received 2016-05-31 accepted: 2017-01-09 appropriate criteria set for personnel promotion across organizational levels using analytic hierarchy process (ahp) charles noven castilloa, france kevin degamoa, faith therese gitganoa, louie alfred looa, shaira mae pacaanasa, nika toroya, lanndon ocampoa*, christine omela ocampob, leahlizbeth siaa aschool of management, university of the philippines cebu, 6000 cebu city, philippines. bdepartment of industrial engineering, university of san carlos, 6000 cebu city, philippines. a lanndon.ocampo@up.edu.ph abstract: currently, there has been limited established specific set of criteria for personnel promotion to each level of the organization particularly among retail industries. this study is conducted in order to develop a personnel promotion strategy by identifying specific sets of criteria for each level of the organization. the complexity of identifying the criteria set along with the subjectivity of these criteria require the use of multi-criteria decision-making approach particularly the analytic hierarchy process (ahp). results show that for first line management, the specific criteria set are: trustworthiness, job involvement, creative and innovative skills, younger generation, and family of businesses. for the middle management, the specific criteria set are: job involvement, creative and innovative skills, trustworthiness, family of business, and education credentialing. and for the top management, the specific criteria set are: strategic and organizational skills, organizational commitment, creative and innovative skills, content specific knowledge and skills, and strategic entrepreneurial connections. these intend also to help avoid mismatch of employee skills and competencies and their job, and at the same time eliminate the issues in personnel promotion such as favoritism, glass ceiling, and gender and physical attractiveness preference. the contribution of this work is in identifying relevant criteria in developing a personnel promotion strategy across organizational levels. key words: personnel promotion, criteria, ahp, organizational levels, retail industry. 1. introduction promotion is a naturally occurring phenomenon in any established organization with hierarchical organizational structure. personnel promotion represents a changing process in the working place to a new one that requires more competence and qualification skills and bigger responsibilities and offers moral and material satisfaction (dumitrescu, 1995). however, promoting employees also poses various challenges to the human resource department and to the top management due to the considerable number of the criteria to choose from in evaluating which employee to promote. furthermore, there were several identified unprofessional workplace practices such as nepotism and favoritism (arasli and tumer, 2008), men being favorable over women (tinkler et al., 2015), physical attractiveness (dion et al., 1972), and the typical structure of the internal labor market wherein bottom-rank jobs are entry ports and top rank positions are usually done internally (kwon and milgrom, 2014). early strategies and models for personnel promotion developed evaluative criteria which are generalized and are not specific to the different levels of the organization first line management, middle int. j. prod. manag. eng. (2017) 5(1), 11-22creative commons attribution-noncommercial-noderivatives 4.0 international 11 https://doi.org/10.4995/ijpme.2017.5857 http://creativecommons.org/licenses/by-nc-nd/4.0/ management, and top management (kwon and milgrom, 2014). however, each level of the organization requires different sets of skills, competencies, and capabilities. thus, each level definitely requires different sets of criteria. with unsuitable criteria for a particular organizational level, employees being promoted might not match the skills, capabilities, and competencies needed for a certain job position. thus, there is a need to develop a different set of criteria for each organizational level to match appropriately the competencies of employees and their job position. this study attempts to identify the criteria in personnel promotion to help practitioners such as the selection committee in selecting who among the candidates are best fit for the position. since each level have different responsibilities, the specific set of criteria for each level allows the firm to identify candidates that have the potential to be promoted to top-level management and those candidates that can best perform on first line management position. due to the number of criteria and the underlying complexity in the decision-making process brought about by the subjectivity and difficulty in eliciting judgments, a multiple criteria decision-making (mcdm) approach particularly the analytic hierarchy process (ahp) is used in identifying appropriate criteria set for personnel promotion for each level of the organization. ahp was developed by saaty (1980) to address an mcdm problem that is structured as a hierarchy. several applications of the ahp have emerged in current literature across various disciplines. for the current developments, theoretical extensions, and applications of the ahp, see the reviews of vaidya and kumar (2006), ho (2008), sipahi and timor (2010), ishizaka and labib (2011), subramanian and ramanathan (2012), herva and roca (2013). the contribution of this work lies in identifying general appropriate criteria for each level of an organization. 2. personnel promotion criteria 2.1. personality/traits employee satisfaction depends, in a large part, on how much individuals like their superiors (harter et al., 2002). thus, it is important that firms consider the employee’s opinions particularly on what the traits they desire in their leaders are, whether direct or top management leaders. leaders in the workplace also influence the values and actions of followers by setting a personal example of conduct which refers to the process of modelling (bandura, 1986). trustworthy and emotionally intelligent leaders who are capable of leading well are desired by individuals (nichols and cottrell, 2014). as followers develop trust in their leaders, they tend to perform better, display more organizational citizenship behaviors (ocbs) and greater organizational commitment, experience greater job satisfaction and have less intent to leave the organization (dirks and ferrin, 2002). interpersonal traits are more desired in lowlevel leaders than in high-level leaders; whereas dominant traits are more desired in high-level leaders than in low-level leaders (nichols and cottrell, 2014). 2.2. educational attainment profession is defined as having the specific characteristics such as a code of ethics, a service orientation, a body of competency and knowledge, credentials, the backing of a professional society, educational requirements and continuing professional development (cohen, 2015). educational credentialing happens in today’s workplaces (baker, 2011; bills, 2003; bills and brown, 2011). educational credentialing theories explain the processes of how societies allocate individuals to slots in the occupational hierarchy on the basis of their educational achievements. thus, candidates with better educational achievements have higher chance on being selected and promoted for a job. the younger generation today are into the pursuit of graduate and post-graduate education in order to heighten the chance of being accepted or promoted in their respective works. educational credentials, e.g. degrees, diplomas, certificates and licenses, have become major instruments for allocating individuals in the labor market and for serving as job requirements in the occupational structure (baker, 2011). on the other hand, aging workers (older generation) return to universities to catch up with the younger workers with better educational credentials. in order to keep up with the new generation and maintain the level of relevance vis-à-vis the relevance of the young ones, each new generation needs more education to keep up in positional competition (brown et al., 2003). 2.3. social capital more than the quantity of social network is the strength of the ties; that is how strong the connection of the candidate to the executive or high-ranking int. j. prod. manag. eng. (2017) 5(1), 11-22 creative commons attribution-noncommercial-noderivatives 4.0 international castillo, c. n., degamo, f. k., gitgano, f. t., loo, l. a., pacaanas, s. m., toroy, n., ocampo, l., ocampo c. o., sia l. 12 http://creativecommons.org/licenses/by-nc-nd/4.0/ official (kim and canella, 2008). thus, ceos tend to likely choose employees with connections to other people who hold high position in other companies. thus, employers tend to likely choose employees with connections to other people who hold a high or significant position in other companies. firms that direct their attention to human and social capital gives more attention to the qualifications and characteristics of candidates that have more strategic entrepreneurial networks (fernandez and mabel, 2011). in family firms, there is a preference of the parting ceo to see a family heir steer the strategic direction of the family firm which might lead to a succession contest whose constrains favor family members (ahrens et al., 2015). with the presence of a son for instance, the current ceo would likely pass on the position to his son over a non-family member candidate even though that candidate has a higher human capital (dahl and moretti, 2008). 2.4. job satisfaction managers all over the world from top to bottom managerial positions have devised various ways to implement mechanisms that promote job satisfaction. essentially, job satisfaction is one field in industrial psychology that has been widely discussed in current literature (lu et al., 2012). there are three factors on the job satisfaction model: first is the psychological state of employees, then the characteristics of jobs that can create these psychological states and lastly, the attributes of individuals that determine how positively a person will respond to a complex and challenging job (hackman, 1976). related to job satisfaction is job involvement which measures the degree to which people identify psychologically with their job and consider their perceived performance level important to self-worth (robbins and judge, 2013). a significant relationship has been identified between job satisfaction and organizational commitment (umi narimawati, 2007). organizational commitment has been identified as critical for organizational success. they consider the firm with which they are working as an extension of themselves rather than just a workplace. therefore, employees with organizational commitment do a better job because they believe that the organization’s success also parallels to their success. employee engagement is an individual’s involvement with, satisfaction with, and enthusiasm for, the work one does (robbins and judge, 2013). thus, employees who are engaged likely enjoy their job and are willing to put extra effort for better job performance. 2.5. skills and experience it is important that a manager possesses the right skill set that would enable him to properly manage and lead the employees. employers prefer hiring graduates with higher levels of professional expertise content-specific knowledge and skills needed to solve occupation-specific problems (humburg and der velden, 2015). fresh graduates aiming for entrylevel managerial position must possess such skills. from the employer’s perspective, types of skills such as innovative/creative skills and strategic/ organizational skills are more important and better developed after having acquired a few years of work experience (humburg and van der velden, 2015). moreover either task-specific or firm-specific experience contributes to productivity (chowdhury et al., 2014). the job of a “leader” is one that requires, in general, a higher level of cognitive demands (fleishman et al., 1991; mumford, 1986) such that a manager should possess high analytical and cognitive skills as a manager is expected to make wise decision. in the planning skill, managers do not only lead the employees but at the same time involved in the planning process of the firm may it be tactical or strategic (mumford et al., 2002). 3. methodology 3.1. analytic hierarchy process ahp is a powerful mcdm tool especially in hierarchical decision-making where the decision problem is structured into components of different levels. decision-makers elicit pairwise comparisons, based from their value judgments, of the elements in the same level with respect to an element in higher immediate level. the strength of the ahp is in capturing subjective judgments of decision-makers and integrating them into the decision-making process. the theoretical discussion was presented by saaty (1980) and a simple tutorial was developed by dolan et al. (1989). various applications of the ahp have been reported extending from airline industry applications (garg, 2016; delbari et al., 2016), healthcare (yuen, 2014), climate policies (castelloint. j. prod. manag. eng. (2017) 5(1), 11-22creative commons attribution-noncommercial-noderivatives 4.0 international appropriate criteria set for personnel promotion across organizational levels using analytic hierarchy process (ahp) 13 http://creativecommons.org/licenses/by-nc-nd/4.0/ branco et al., 2012) to reverse logistics model (barker and zabinsky, 2011), strategy evaluation (ocampo and clark, 2015), critical sustainability indicators identification (ocampo et al., 2015), and sustainable manufacturing index computation (ocampo, 2015; ocampo et al., 2016) among several hundreds of applications. for the personnel promotion domain, the earliest adoption of the ahp can be traced back from saaty and ramanujam (1983) and the succeeding works were carried out by taylor et al. (2006), islam and rasad (2006), mittal et al. (2009) and bali et al. (2015). this development is rather slow when compared to other applications of the ahp. for the current developments, theoretical extensions, issues and applications of the ahp, see the reviews of zahedi (1986), vaidya and kumar (2006) and sipahi and timor (2010) on the overview of applications, ho (2008) for the integration of other approaches and their applications, ishizaka and labib (2011) on its main developments, subramanian and ramanathan (2012) for its various applications in operations management, (schmidt et al., 2015) for healthcare research, anis and islam (2015) for applications in higher learning institutions, chandio et al. (2013) on gis-based integration, maleki and zahir (2013) on rank reversal phenomenon issue, and godinho et al. (2011) issues with r&d project selection application, among others. generally, the procedure of ahp can be described as follows: 1. develop the problem structure problem structures are developed hierarchically in a top-down approach (saaty, 1980). oftentimes, there is an explicit definition and representation of goal, criteria and alternatives. in various cases, criteria are described in more than one level so that further details are explicitly represented in the problem structure. the decision of the inclusion of components and alternatives is usually carried out either through a critical review of literature or through an expert or group of experts who have sufficient knowledge and experience of the problem under consideration. decision components and elements are usually a combination of both objective and subjective ones with measurements in different multiple dimensions. 2. eliciting judgment in paired comparisons pairwise comparisons of elements in the same level with respect to an element in the immediate higher level are carried out with expert decision-makers. the generic question in making pairwise comparisons goes like this: “given a parent element and given a pair of elements, how much more does a given member of the pair dominate other member of the pair with respect to a parent element?” (promentilla et al., 2006). to achieve a uni-dimensional scaling property of the pairwise comparisons, saaty (1980) established the famous saaty fundamental 9-point ratio scale as shown in table 1. suppose that akij represents the decision of k th decision-maker on the influence of element i on j. to aggregate individual judgments, saaty (1980) proposed the weighted geometric mean method (wgmm) as shown in (1): a aij ijk k k= a^ h% (1) wihere aij is the aggregated judgment, αk is the decision-maker’s importance to the decision making process with αk>0 and 1k k a =| . the values of aij ,i j6 form the pairwise comparisons matrix. in pairwise comparisons, reciprocity is maintained. priority vectors (w) are obtained from the pairwise comparison matrix (a) by solving an eigenvalue problem in the following equation: table 1. saaty fundamental scale. rating scale definition explanation 1 equal importance two elements contribute equally to the objective 2 weak between equal and moderate 3 moderate importance experience and judgment slightly favor one element over another 4 moderate plus between moderate and strong 5 strong importance experience and judgment strongly favor one element over another 6 strong plus between strong and very strong 7 very strong or demonstrated importance an element is favored very strongly over another; its dominance demonstrated in practice 8 very, very strong between very strong and extreme 9 extreme importance the evidence favoring one element over another is one of the highest possible order or affirmation int. j. prod. manag. eng. (2017) 5(1), 11-22 creative commons attribution-noncommercial-noderivatives 4.0 international castillo, c. n., degamo, f. k., gitgano, f. t., loo, l. a., pacaanas, s. m., toroy, n., ocampo, l., ocampo c. o., sia l. 14 http://creativecommons.org/licenses/by-nc-nd/4.0/ aw wmaxm= (2) where λmax is the maximum eigenvalue of the positive reciprocal square matrix (a). the approach also provides a way to measure the consistency of judgments in the pairwise comparison matrix. when decision-making in the pairwise comparisons matrix is consistent λmax=n; otherwise, λmax>n where n is the number of elements being compared. the consistency index (ci), as a measure of degree of consistency, was calculated using the formula ci=(λmax-n)/(n-1) (3) the consistency ratio (cr) is computed as cr=ci/ri (4) where ri is the mean random consistency. acceptable cr values must be less than 0.1. decision-makers were asked to repeat the pairwise comparisons for cr values greater than 0.1. 3. synthesizing judgments saaty (1980) described that synthesizing judgments in ahp is done by weighting the elements being compared in the lower level component to an element in the next immediate level component, referred to as the parent element, by the priority of that element and adding all parents for each element in the lower level. this is referred to as the distributive mode of the ahp. this can be represented in the form for two levels in a hierarchy wt=x3 t(x2 ti)(x1 ti) (5) where w is is the global (synthesized) weight vector of the elements in the lowest (or third level in this case), x3 is the local priority vector of the third level elements (the lowest level), x2 is the local priority vector of the second level elements, x1 is the local priority vector of the first level elements, and i is an identity matrix. 3.2. decision model this study aims to develop a promotion strategy consisting the priority of each criteria used for promotion. with this, the goal is to identify the criteria set for first line management, middle management and top management. from the review of related literature on this field, the problem structure presented in figure 1 was then developed. goal first line management middle management social capital personality and traits educational attainment experience job satisfaction top management • younger generation • older generation • educational credentialing • trustworthiness • emotional intelligence • interpersonal traits • dominant traits • quantity of school ties • strategic entrep connections • family of business • content specific knowledge and skills • innovative and creative skills • strategic and organizational skills • job involvement • employee engagement • organizational commitment figure 1. criteria identification decision model. int. j. prod. manag. eng. (2017) 5(1), 11-22creative commons attribution-noncommercial-noderivatives 4.0 international appropriate criteria set for personnel promotion across organizational levels using analytic hierarchy process (ahp) 15 http://creativecommons.org/licenses/by-nc-nd/4.0/ the levels of management are independent with each other which explains the broken lines between the main goal and the levels of management. this means that the selection of criteria set is done for each management level. the promotion strategy in each goal is derived from the pool of criteria; this explains the segments from the each goal to each of the criteria. for instance, the criteria set for the first line management promotion is selected from the pool of criteria which consists of personality/ traits, educational attainment, social capital, skills and experience, and job satisfaction as discussed in section 2. the sub-criteria for each criterion are compared pairwise against each other to determine the ranking within each criterion. for instance younger generation, older generation and educational credentialing are compared against each other to determine which one ranks the highest with respect to educational attainment. the list of criteria and their corresponding sub-criteria set is shown in table 2. due to the hierarchical nature of the decision-making problem and the presence of multiple criteria, ahp is considered to be the most appropriate mcdm method. the mutual exclusiveness of the criteria indicates that the set of criteria used to determine a certain level of management does not affect the set of criteria for another level of management. the existence of several criteria in personnel promotion makes mcdm a relevant approach in selecting the most appropriate criteria for each organizational level. the decision structure could likewise aid in addressing the negative issues affecting personnel promotion hence it is able to generate a standardized set of criteria. 4. results and discussion table 3 to table 5 show the aggregate pairwise comparison matrix of the first line, middle and top management, respectively. these aggregate matrices were computed using equation 1 with αk=1/k. the priority vector or weight vector was computed using equation 2. the consistency ratio (cr), computed using equation 4, is also shown in each table. the rank of each criterion based on its priority vector provides a way in determining what particular significant criterion for each level of the organization. the prioritization of the sub-criteria for their parent criterion was also performed using the same process. a total of 180 pairwise comparisons were conducted in order to come up with the aggregated weights of the different main criteria and sub-criteria for all levels of management. table 2. coding of the criteria and sub-criteria. decision criteria decision sub-criteria code educational attainment c1 young generation c1.1 older generation c1.2 educational credentialing c1.3 personality and traits c2 trustworthiness c2.1 emotional intelligence c2.2 interpersonal traits c2.3 dominant traits c2.4 social capital c3 quantity of school ties c3.1 quality of school ties c3.2 strategic entrepreneurial connections c3.3 family of businesses c3.4 experience c4 content specific knowledge and skills c4.1 creative and innovative skills c4.2 strategic and organizational skills c4.3 job satisfaction c5 job involvement c5.1 employee engagement c5.2 organizational commitment c5.3 int. j. prod. manag. eng. (2017) 5(1), 11-22 creative commons attribution-noncommercial-noderivatives 4.0 international castillo, c. n., degamo, f. k., gitgano, f. t., loo, l. a., pacaanas, s. m., toroy, n., ocampo, l., ocampo c. o., sia l. 16 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 3. aggregate pairwise comparison of the main criteria for the first line management. goal c1 c2 c3 c4 c5 priority vector c1 1.00 0.38 0.78 0.39 0.32 0.09 c2 2.60 1.00 4.20 1.94 1.11 0.33 c3 1.28 0.24 1.00 0.39 0.31 0.09 c4 2.53 0.52 2.55 1.00 0.70 0.20 c5 3.12 0.90 3.25 1.43 1.00 0.29 cr = 0.012 the aggregate judgment of the respondents for the first line management placed personality and traits as the major consideration in promoting personnel for the first line management. as seen in table 3, personality and traits criterion is more important than any other criteria with respect to first line management. moreover, the respondents were consistent in their judgment given that the cr=0.012, which is within the acceptable range for crs. table 4. aggregate pairwise comparison of the main criteria for the middle management. goal c1 c2 c3 c4 c5 priority vector c1 1.00 0.71 0.80 0.41 0.37 0.11 c2 1.41 1.00 1.31 0.62 0.51 0.16 c3 1.25 0.76 1.00 0.26 0.39 0.11 c4 2.42 1.62 3.80 1.00 0.88 0.31 c5 2.67 1.94 2.58 1.13 1.00 0.31 cr = 0.010 table 4 shows the aggregate judgment of the respondents with respect to the middle management promotion where experience and skills and job satisfaction were closely prioritized by the respondents. with a weight of 0.31 respectively, these two criteria sufficiently supplement the necessary human and conceptual skills needed for the middle management post. at the same time, the remaining three criteria were also closely prioritized by the respondents. however, there is a wide gap between the first two criterion and the last three ones. middle management has to engage in departmental or divisional decisions, experience and organizational commitment play vital roles in managing the department. this indicates that experience with the technical nature of the job and the necessary organization and job commitment to the organization are two important criteria that are required of a manager wanting to be or is ready to be promoted to the middle management post. the rest of the criteria are, despite the weights, still necessary in being a manager for the middle management. personality and traits are necessary in order to handle and manage employees well. following personality and traits, educational attainment and social capital are equally important. in this level of management, and most of the time, applicants for middle management see the promotable employees as equal in terms of educational attainment because most firms require educational attainment at first entry. the challenge to look beyond educational credentials such as personality and traits or job satisfaction is one of the many things that is difficult to view on paper. just like the judgment for the first line management, respondents were also consistent in their decision-making with regards to the ranking of criteria with a consistency ratio of 0.01. table 5. aggregate pairwise comparison of the main criteria for the top management. goal c1 c2 c3 c4 c5 priority vector c1 1.00 0.83 0.36 0.17 0.55 0.08 c2 1.21 1.00 0.73 0.18 0.66 0.10 c3 2.76 1.37 1.00 0.45 0.70 0.17 c4 5.77 5.50 2.22 1.00 2.70 0.47 c5 1.83 1.51 1.43 0.37 1.00 0.18 cr = 0.015 table 5 shows the weight vector for the criteria set with experience and skills as arguably the topmost consideration in promoting personnel to top management. the top management is responsible in driving the firm towards the realization of its managerial goals, corporate mission and its long-term vision. therefore at this stage, high conceptual skill is extremely needed in order to properly and correctly manage the firm. results show that experience and skills are the paramount considerations in promoting towards top management as well as an employee must have top level business skills and significantly sufficient experience not only with the firm but as well as the familiarization with the industry that the business is in. with weight equal to 0.47, top management personnel needs to have the experience and the skills in order to make sound decisions and fair judgment especially during times when they are highly needed. int. j. prod. manag. eng. (2017) 5(1), 11-22creative commons attribution-noncommercial-noderivatives 4.0 international appropriate criteria set for personnel promotion across organizational levels using analytic hierarchy process (ahp) 17 http://creativecommons.org/licenses/by-nc-nd/4.0/ it is also important to note that the rest of the criteria have weights close to each other at 0.18, 0.17 and 0.08. this indicates that to be a top management personnel, one has to have equal or sufficient amount of job satisfaction, social capital, personality and traits and educational attainment. the aggregate judgment of the respondents is consistent with a cr value of 0.015. table 6 shows the final weights of each sub-criterion with respect to the weights of the main criteria for the first line management. table 6. final weights of the sub-criteria of the first line management. sub-criteria final weights rank young generation 0.04 11 older generation 0.03 12 educational credentialing 0.02 16 trustworthiness 0.13 1 emotional intelligence 0.08 5 interpersonal traits 0.07 7 dominant traits 0.04 10 quantity of school ties 0.01 17 quality of school ties 0.02 15 strategic entrepreneurial connections 0.03 14 family of businesses 0.03 13 content specific knowledge and skills 0.05 8 creative and innovative skills 0.10 3 strategic and organizational skills 0.05 9 job involvement 0.11 2 employee engagement 0.08 6 organizational commitment 0.10 4 for the first line management, managers tend to promote employees that show trustworthiness, job involvement, and creative and innovative skills. in this level, customer and supplier interaction is observed. first line managers are usually those that are in the grass roots, are those who communicate with the clients and suppliers, and know the situation of the processes, methods, and procedures. they are also responsible in overseeing first line technical employees. with this, results show that trustworthiness is one of the most coveted trait that are being looked in promoting to first level of management. companies need to have a set of managers who are trustworthy for the employees and customers alike. employees promoted to this level need to exhibit job involvement or on to how he identifies himself with his job. it might take months, or even years, to promote a technical employee into a managerial post but that length of time is sufficient in order to determine if the employee fully identifies himself with his job. with this, first level managers would know the nature of the work and tasks given to technical employees and will be able to manage or supervise them effectively. contrary to the literature which states that first line managers should exhibit content-specific knowledge and skills to supervise employees better, the results show that creative and innovative skills are one of the top criterion that an employee must possess in order to get the first level management post. more often than not, first level managers master already the technical skills that sometimes, human skills are needed to manage people under the manager properly. table 7 shows the final weights of each sub-criterion with respect to the weights of the main criteria for the middle management. table 7. final weights of the sub-criteria of the middle management. sub-criteria final weights rank young generation 0.02 16 older generation 0.06 7 educational credentialing 0.04 11 trustworthiness 0.07 6 emotional intelligence 0.03 13 interpersonal traits 0.03 12 dominant traits 0.02 15 quantity of school ties 0.01 17 quality of school ties 0.02 14 strategic entrepreneurial connections 0.04 10 family of businesses 0.04 9 content specific knowledge and skills 0.04 8 creative and innovative skills 0.10 3 strategic and organizational skills 0.16 1 job involvement 0.08 5 employee engagement 0.09 4 organizational commitment 0.14 2 for the middle management, results show that strategic and organizational skills are a must for first line managers who will be promoted to the next level of management, the middle management. middle managers are the set of people that serve as the bridge of the first line technical employees and managers, and the top management. these are the people who conduct departmental or division-level decision-making. with that, results show that these managers should possess strategic and organizational skills which are needed to fill-in the gap between the first line management and the top management. int. j. prod. manag. eng. (2017) 5(1), 11-22 creative commons attribution-noncommercial-noderivatives 4.0 international castillo, c. n., degamo, f. k., gitgano, f. t., loo, l. a., pacaanas, s. m., toroy, n., ocampo, l., ocampo c. o., sia l. 18 http://creativecommons.org/licenses/by-nc-nd/4.0/ closely important with strategic and organizational skills, organizational commitment must also be present in first line managers who would be promoted to middle management. this would entail a significant number of years into the company, significant amount of corporate targets achieved, and a positive feedback from evaluations. and not far from the organizational commitment is creative and innovative skill. middle management must have a balance of technical, human and conceptual skills in order to perform the role properly. with this mix, innovating the role as middle manager is required in order to effectively and efficiently do the tasks. the criteria are closely related in terms of the weight because of the nature of the job. it is important to note the complexity of the job of middle managers who need not only show technical skills, but human and strategic skills as well. topping this level is strategic and organizational skill which is ‘conceptual in nature’, organizational commitment which is a must for every manager in the middle level, creative and innovative skills which is ‘technical’ nature and is required for a first level manager as above mentioned. table 8 shows the final weights of each sub-criterion with respect to the weights of the main criteria for the top management. table 8. final weights of the sub-criteria of the top management. sub-criteria final weights rank young generation 0.01 15 older generation 0.03 10 educational credentialing 0.04 9 trustworthiness 0.04 8 emotional intelligence 0.03 12 interpersonal traits 0.02 14 dominant traits 0.01 16 quantity of school ties 0.01 17 quality of school ties 0.02 13 strategic entrepreneurial connections 0.07 5 family of businesses 0.06 6 content specific knowledge and skills 0.08 4 creative and innovative skills 0.09 3 strategic and organizational skills 0.29 1 job involvement 0.05 7 employee engagement 0.03 11 organizational commitment 0.09 2 consistent with the literature, top management requires heavily the conceptual skills and strategic thinking from a manager. in this level of management, the strategic direction and intent of the company in the short and long term is formulated. reports from different departments and/or division are being utilized to come up with a strategy needed to stir the company to success and away from irrelevance and obsolescence. compared to middle management, the need for a top level manager to have strategic and organizational skills is substantially important. at a significant rise from 16% (middle management) to 29%, the requirement of a strategic mindset to promote to a post in top level management is thus required. figure 2 shows the disparity in the required level of strategic and organizational skills between middle and top management. figure 2. disparity of the weights in strategic and organizational skills. as mentioned from the literature, top level managers need to have high conceptual skills as what the results also display. organizational commitment and creative and innovative skills are two criteria that are equally important for top level management, following the need for strategic thinking. it can also be observed that other criteria (including the organizational commitment and creative and innovative skills) have weights that are below 10%. the results show that other criteria are already expected from a top level manager (as proven by how he has managed to be promoted at this level) which explains why the percentages are closely equal and the corresponding criteria closely important. moreover, the results show the high need to think strategically which illustrates the overly high percentage that strategic and organizational skills got. int. j. prod. manag. eng. (2017) 5(1), 11-22creative commons attribution-noncommercial-noderivatives 4.0 international appropriate criteria set for personnel promotion across organizational levels using analytic hierarchy process (ahp) 19 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5. conclusion this paper explored the various criteria and subcriteria that must be considered for personnel promotion for each management level in an organization. with the use of the ahp, results of this work confirm that different mix of criteria make up for the criteria set specific for a particular management level. results show that for first line management, the specific criteria set are: trustworthiness, job involvement, creative and innovative skills, younger generation, and family of businesses. for the middle management, the specific criteria set are: job involvement, creative and innovative skills, trustworthiness, family of business, and education credentialing. and for the top management, the specific criteria set are: strategic and organizational skills, organizational commitment, creative and innovative skills, content specific knowledge and skills, and strategic entrepreneurial connections. in general, this work shows that personality and traits, job satisfaction and experience and skills are more critical rather than social capital across different management levels. this implied that selection and promotion committees must focus more on selecting employees that have the right capabilities for the job rather than those that are sponsored by social connections. the insights of this study would help aid human resource managers in general and promotion committees in particular in their promotion decision processes. furthermore, results of this study are beneficial for human resource practitioners in developing trainings and other relevant support infrastructures for each management level in enhancing required skills that are required for promotion. references ahrens, j., landmann, a., woywoode, m. (2015). gender preferences in the ceo successions of family firms: family characteristics and human capital of the successor. journal of family business strategy, 6(2), 86-103. https://doi.org/10.1016/j.jfbs.2015.02.002 anis, a., islam, r. (2015). the application of analytic hierarchy process in higher-learning institutions: a literature review. journal for international business and entrepreneurship development, 8(2), 166-182. https://doi.org/10.1504/jibed.2015.070446 arasli, h., tumer, m. (2008). nepotism, favouritism and cronyism: a study of their effects on job stress and job satisfaction in the banking industry of north cyprus. social behavior and personality: an international journal, 36(9), 1237-1250. https://doi.org/10.2224/ sbp.2008.36.9.1237 baker, d. (2011). forward and backward, horizontal and vertical: transformation of occupational credentialing in the schooled society. research in social stratification and mobility, 29(1), 5-29. https://doi.org/10.1016/j.rssm.2011.01.001 bali, o., dagdeviren, m., gumus, s. (2015). an integrated dynamic intuitionistic fuzzy madm approach for personnel promotion problem. kybernetes, 44(10), 1422-1436. https://doi.org/10.1108/k-07-2014-0142 bandura, a. (1986). social foundations of thought and action: a social cognitive theory. englewood cliffs, nj: prentice-hall. barker, t.j., zabinsky, z.b. (2011). a multicriteria decision-making model for reverse logistics using analytical hierarchy process. omega, 39(5), 558-573. https://doi.org/10.1016/j.omega.2010.12.002 bills, d. (2003). credentials, signals, and screens: explaining the relationship between schooling and job assignment. review of educational research, 73(4), 441-449. https://doi.org/10.3102/00346543073004441 bills, d., brown, d. (2011). new directions in educational credentialism. research in social stratification and mobility, 29(1), 1-4. https:// doi.org/10.1016/j.rssm.2011.01.004 brown, p., hesketh, a., williams, s. (2003). employability in a knowledge-driven economy. journal of education and work, 16(2), 107-126. castello-branco, d., rathmann, r., borba, b.s., de lucena, a.f. (2012). a multicriteria approach for measuring the carbon-risk of oil companies. energy strategy reviews, 1(2), 122-129. https://doi.org/10.1016/j.esr.2012.06.002 chandio, i.a., matori, a.n.b., wanyusof, k.b., talpur, m.a.h., balogun, a.l., lawal, d.u. (2013). gis-based analytic hierarchy process as a multicriteria decision analysis instrument: a review. arabian journal of geosciences, 6(8), 3059-3066. https://doi.org/10.1007/ s12517-012-0568-8 chowdhury, s., schulz, e., milner, m., van de voort, d. (2014). core employee based human capital and revenue productivity in small firms: an empirical investigation. journal of business research, 67(11), 2473-2479. https://doi.org/10.1016/j.jbusres.2014.03.007 cohen, d.j. (2015). hr past, present and future: a call for consistent practices and a focus on competencies. human resource management review, 25(2), 205-215. https://doi.org/10.1016/j.hrmr.2015.01.006 dahl, g. b. (2008). the demand for sons. the review of economic studies, 75(4), 1085-1120. https://doi.org/10.1111/j.1467937x.2008.00514.x int. j. prod. manag. eng. (2017) 5(1), 11-22 creative commons attribution-noncommercial-noderivatives 4.0 international castillo, c. n., degamo, f. k., gitgano, f. t., loo, l. a., pacaanas, s. m., toroy, n., ocampo, l., ocampo c. o., sia l. 20 https://doi.org/10.1016/j.jfbs.2015.02.002 https://doi.org/10.1504/jibed.2015.070446 https://doi.org/10.2224/sbp.2008.36.9.1237 https://doi.org/10.2224/sbp.2008.36.9.1237 https://doi.org/10.1016/j.rssm.2011.01.001 https://doi.org/10.1108/k-07-2014-0142 https://doi.org/10.1016/j.omega.2010.12.002 https://doi.org/10.3102/00346543073004441 https://doi.org/10.1016/j.rssm.2011.01.004 https://doi.org/10.1016/j.rssm.2011.01.004 https://doi.org/10.1016/j.esr.2012.06.002 https://doi.org/10.1007/s12517-012-0568-8 https://doi.org/10.1007/s12517-012-0568-8 https://doi.org/10.1016/j.jbusres.2014.03.007 https://doi.org/10.1016/j.hrmr.2015.01.006 https://doi.org/10.1111/j.1467-937x.2008.00514.x https://doi.org/10.1111/j.1467-937x.2008.00514.x http://creativecommons.org/licenses/by-nc-nd/4.0/ delbari, s.a., ng, s.i., aziz, y.a., ho, j.a. (2016). an investigation of key competitiveness indicators and drivers of full service airlines using delphi and ahp techniques. journal of air transport management, 52, 23-34. https://doi.org/10.1016/j.jairtraman.2015.12.004 dion, k., berscheid, e., walster, e. (1972). what is beautiful is good. journal of personality and social psychology, 24(3), 285-290. https:// doi.org/10.1037/h0033731 dirks, k.t., ferrin, d.l. (2002). trust in leadership: meta-analytic findings and implications for research and practice. journal of applied psychology, 87(4), 611-628. https://doi.org/10.1037/0021-9010.87.4.611 dolan, j.g., isselhardt, b.j., cappuccio, j.d. (1989). the analytic hierarchy process in medical decision making: a tutorial. medical decision making, 9(1), 40-50. https://doi.org/10.1177/0272989x8900900108 dumitrescu, m. (1995). introduction to management and general management. eurounion house, oradea, 205. fernandez, r.m., abraham, m. (2011). glass ceilings and glass doors? internal and external hiring in an organizational hierarchy. mit sloan research paper no. 4895-11. https://doi.org/10.2139/ssrn.1804896 fleishman, e.a., mumford, m.d., zaccaro, s.j., levin, k.y., korotkin, a.l., hein, m.b. (1991). taxonomic efforts in the description of leader behavior: a synthesis and functional interpretation. leadership quarterly, 2(4), 245-287. https://doi.org/10.1016/10489843(91)90016-u garg, c.p. (2016). a robust hybrid decision model for evaluation and selection of the strategic alliance partner in the airline industry. journal of air transport management, 52, 55-66. https://doi.org/10.1016/j.jairtraman.2015.12.009 godinho, p., costa, j.p., fialho, j., afonso, r. (2011). some issues about the application of the analytic hierarchy process to r&d project selection. global business and economics review, 13(1), 26–41. https://doi.org/10.1504/gber.2011.039188 hackman, j.r. (1976). motivation through the design of work: test of theory. organizational behavior and human performance, 16(2), 250279. https://doi.org/10.1016/0030-5073(76)90016-7 harter, j.k., schmidt, f.l., hayes, t.l. (2002). business-unit-level relationship between employee satisfaction, employee engagement, and business outcomes: a meta-analysis. journal of applied psychology, 87(2), 268-279. https://doi.org/10.1037/0021-9010.87.2.268 herva, m., roca, e. (2013). review of combined approaches and multi-criteria analysis for corporate environmental evaluation. journal of cleaner production, 39, 355-371. https://doi.org/10.1016/j.jclepro.2012.07.058 ho, w. (2008). integrated analytic hierarchy process and its applications – a literature review. european journal of operational research, 186(1), 211-228. https://doi.org/10.1016/j.ejor.2007.01.004 humburg, m., van der velden, r. (2015). skills and the graduate recruitment process: evidence from two discrete choice experiments. economics of education review, 49, 24-41. https://doi.org/10.1016/j.econedurev.2015.07.001 ishizaka, a., labib, a. (2011). review of the main developments in the analytic hierarchy process. expert systems with applications, 38(11), 14336-14345. https://doi.org/10.1016/j.eswa.2011.04.143 islam, r., rasad, s.b.m. (2006). employee performance evaluation by the ahp: a case study. asia pacific management review, 11(3), 163-176. kim, y., canella, a.a. (2008). social capital among corporate upper echelons and its impacts. journal of world business, 43(1), 85-96. https://doi.org/10.1016/j.jwb.2007.10.009 kwon, i., meyersson milgrom, e.m.   (2014). the significance of firm and occupation specific human capital for hiring and promotions. labour economics, 31, 162-173. https://doi.org/10.1016/j.labeco.2014.07.003 lu, h., barriball, k.l., zhang, x., while, a.e. (2012). job satisfaction among hospital nurses revisited: a systematic review. international journal of nursing studies, 49(8), 1017-1038. https://doi.org/10.1016/j.ijnurstu.2011.11.009 maleki, h., zahir, s. (2013). a comprehensive literature review of the rank reversal phenomenon in the analytic hierarchy process. journal of multi-criteria decision analysis, 20(3-4), 141-155. https://doi.org/10.1002/mcda.1479 mittal, k.c., goel, a.k., mohindru, p. (2009). performance evaluation of employees using analytical hierarchical process: a case study of indian it industry. asia-pacific journal of management research and innovation, 5(4), 119-127. https://doi.org/10.1177/097324700900500410 mumford, m.d. (1986). leadership in the organizational context: conceptual approach and its application. journal of applied social psychology, 16, 212–226. https://doi.org/10.1111/j.1559-1816.1986.tb01156.x mumford, m.d., schultz, r.a., osburn, h.k. (2002). planning in organizations: performance as a multi-level phenomenon. in francis j. yammarino, fred dansereau (ed.) the many faces of multi-level issues (research in multi-level issues, volume 1), emerald group publishing limited, pp. 3-65. https://doi.org/10.1016/s1475-9144(02)01026-3 nichols, a.l., cottrell, c.a. (2014). what do people desire in their leaders? the role of leadership level on trait desirability. the leadership quarterly, 25(4), 711-729. https://doi.org/10.1016/j.leaqua.2014.04.001 ocampo, l. (2015). a hierarchical framework for index computation in sustainable manufacturing. advanced in production engineering and management, 10(1), 40-50. https://doi.org/10.14743/apem2015.1.191 ocampo, l., clark, e. (2015). an analytic hierarchy process (ahp) approach in the selection of sustainable manufacturing initiatives: a case in a semiconductor manufacturing firm in the philippines. international journal of analytic hierarchy process, 7(1), 32-49. https://doi. org/10.13033/ijahp.v7i1.223 int. j. prod. manag. eng. (2017) 5(1), 11-22creative commons attribution-noncommercial-noderivatives 4.0 international appropriate criteria set for personnel promotion across organizational levels using analytic hierarchy process (ahp) 21 https://doi.org/10.1016/j.jairtraman.2015.12.004 https://doi.org/10.1037/h0033731 https://doi.org/10.1037/h0033731 https://doi.org/10.1037/0021-9010.87.4.611 http://mdm.sagepub.com/search?author1=james+g.+dolan&sortspec=date&submit=submit http://mdm.sagepub.com/search?author1=bernard+j.+isselhardt&sortspec=date&submit=submit http://mdm.sagepub.com/search?author1=joseph+d.+cappuccio&sortspec=date&submit=submit https://doi.org/10.1177/0272989x8900900108 https://doi.org/10.2139/ssrn.1804896 https://doi.org/10.1016/1048-9843(91)90016-u https://doi.org/10.1016/1048-9843(91)90016-u https://doi.org/10.1016/j.jairtraman.2015.12.009 https://doi.org/10.1504/gber.2011.039188 https://doi.org/10.1016/0030-5073(76)90016-7 https://doi.org/10.1037/0021-9010.87.2.268 https://doi.org/10.1016/j.jclepro.2012.07.058 https://doi.org/10.1016/j.ejor.2007.01.004 https://doi.org/10.1016/j.econedurev.2015.07.001 http://expert systems with applications https://doi.org/10.1016/j.eswa.2011.04.143 https://doi.org/10.1016/j.jwb.2007.10.009 https://doi.org/10.1016/j.labeco.2014.07.003 https://doi.org/10.1016/j.ijnurstu.2011.11.009 https://doi.org/10.1002/mcda.1479 https://doi.org/10.1177/097324700900500410 https://doi.org/10.1111/j.1559-1816.1986.tb01156.x https://doi.org/10.1016/s1475-9144(02)01026-3 https://doi.org/10.1016/j.leaqua.2014.04.001 https://doi.org/10.14743/apem2015.1.191 http://creativecommons.org/licenses/by-nc-nd/4.0/ ocampo, l.a., clark, e.e., promentilla, m.a.b. (2016). computing sustainable manufacturing index with fuzzy analytic hierarchy process. international journal of sustainable engineering, 9(5), 305-314. https://doi.org/10.1080/19397038.2016.1144828 ocampo, l., vergara, v.g.n., impas sr., c.g., tordillo, j.a.s., pastoril, j.s. (2015). identifying critical indicators in sustainable manufacturing using analytic hierarchy process (ahp). journal of manufacturing and industrial engineering, 14(3-4), 1-8. https://doi.org/10.1080/1 9397038.2016.1144828 promentilla, m.a.b., furuichi, t., ishii, k., tanikawa, n. (2006). evaluation of remedial countermeasures using the analytic network process. waste management, 26(12), 1410–1421. https://doi.org/10.1016/j.wasman.2005.11.020 robbins, s.p., judge, t.a. (2013). organizational behavior 15th ed. prentice hall: pearson education inc. saaty, t. l. (1980). the analytic hierarchy process. mcgraw-hill, new york. saaty, t.l., ramanujam, v. (1983). an objective approach to faculty promotion and tenure by the analytic hierarchy process. research in higher education, 18(3), 311-331. https://doi.org/10.1007/bf00979603 schmidt, k., aumann, i., hollander, i., damm, k., von der schulenburg, j.m.g. (2015). applying the analytic hierarchy process in healthcare research: a systematic literature review and evaluation of reporting. bmc medical informatics and decision making, 15(112), 1-27. https://doi.org/10.1186/s12911-015-0234-7 sipahi, s., timor, m. (2010). the analytic hierarchy process and analytic network process: an overview of applications. management decision, 48(5), 775-808. https://doi.org/10.1108/00251741011043920 subramanian, n., ramanathan, r. (2012). a review of applications of analytic hierarchy process in operations management. international journal of production economics, 138(2), 215-241. https://doi.org/10.1016/j.ijpe.2012.03.036 taylor iii, f.a., ketcham, a.f., hoffman, d. (2006). personnel evaluation with ahp. management decision, 36(10), 679-685. https://doi. org/10.1108/00251749810245336 tinkler, j.e., whittington, k.b., ku, m.c., davies, a.r. (2015). gender and venture capital decision-making: the effects of technical background and social capital on entrepreneurial evaluations. social science research, 51, 1-16. https://doi.org/10.1016/j. ssresearch.2014.12.008 umi narimawati, s.e. (2007). the influence of work satisfaction, organizational commitment and turnover intention towards the performance of lecturers at west java’s private higher education institution. journal of applied sciences research, 3(7), 549-557. vaidya, o.s., kumar, s. (2006). analytic hierarchy process: an overview of applications. european journal of operational research, 169(1), 1-29. https://doi.org/10.1016/j.ejor.2004.04.028 yuen, k.k.f. (2014). the primitive cognitive network process in healthcare and medical decision making: comparisons with the analytic hierarchy process. applied soft computing, 14, 109-119. https://doi.org/10.1016/j.asoc.2013.06.028 zahedi, f. (1986). the analytic hierarchy process: a survey of the method and its applications. interfaces, 16(4), 96-108. https://doi. org/10.1287/inte.16.4.96 int. j. prod. manag. eng. (2017) 5(1), 11-22 creative commons attribution-noncommercial-noderivatives 4.0 international castillo, c. n., degamo, f. k., gitgano, f. t., loo, l. a., pacaanas, s. m., toroy, n., ocampo, l., ocampo c. o., sia l. 22 https://doi.org/10.1080/19397038.2016.1144828 https://doi.org/10.1080/19397038.2016.1144828 https://doi.org/10.1080/19397038.2016.1144828 https://doi.org/10.1016/j.wasman.2005.11.020 https://doi.org/10.1108/00251741011043920 https://doi.org/10.1016/j.ijpe.2012.03.036 https://doi.org/10.1108/00251749810245336 https://doi.org/10.1108/00251749810245336 https://doi.org/10.1016/j.ssresearch.2014.12.008 https://doi.org/10.1016/j.ssresearch.2014.12.008 https://doi.org/10.1016/j.ejor.2004.04.028 https://doi.org/10.1016/j.asoc.2013.06.028 https://doi.org/10.1287/inte.16.4.96 https://doi.org/10.1287/inte.16.4.96 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2018.8771 received 2017-10-17 accepted: 2017-12-18 designing an environmental zone in a mediterranean city to support city logistics lorenzo ros-mcdonnella1, maría victoria de-la-fuente-aragona2*, diego ros-mcdonnellb and manuel cardós carbonerasc a research goup “industrial engineering”, technical university of cartagena, spain b research goup “project+city”, technical university of cartagena, spain c dpto. organización de empresas, universidad politécnica de valencia, spain a1 lorenzo.ros@upct.es, a2* marivi.fuente@upct.es, b diego.ros@upct.es, b mcardos@doe.upv.es abstract: european cities are facing enormous challenges in accessibility and livability terms due to several european directives, which are compulsory in the mid/long term, traffic congestion levels are still increasing, and air pollution and noise disturbs citizens’ lives. this work presents the study carried out in a mediterranean city to define an environmental zone with traffic restrictions for vehicles in the historical centre of the city of cartagena (spain) by exploring different urban logistics measures to tackle main problems caused by freight deliveries and pickups in the city centre. these solutions aimed to enhance the efficiency of vehicles, and to reduce both traffic congestion the environmental impacts caused by freight delivery in the city in order to improve urban sustainability. key words: environmental zone, urban freight transport, sustainable urban mobility, city logistics. 1. introduction the main priority of authorities is to constantly search for feasible economical solutions in urban logistics problems, such as traffic congestion and parking difficulties, inadequate public transport, environmental impacts and energy use, accidents and safety, freight distribution, automobile dependency, etc. (bulkeley & betsill, 2005). several european directives, which are compulsory in the mid/long term, have led many european cities to propose solutions in accessibility and livability terms because congestion levels are still increasing, and air pollution and noise disturb citizens’ lives. (newman & kenworthy, 1999; southworth, 2003; eu-com, 2011). a clear trend towards increasing pedestrian areas in city centres is the common solution for local authorities (appleyard, 1981; engwicht, 1999; crawford, 2000). although city centre pedestrianisation has led to certain problems appearing, such as limited parking areas, traffic access limitations, difficulties to deliver operations, as well as difficult access for the neighbours who live in these areas, urban transport issues have become more important to better support life for people, and towards a better environment in urban areas (dablanc, 2008). therefore, balancing smart economic activities and cleaner, quieter, and safer communities are needed to create more sustainable and liveable cities (taniguchi, 2014). to cite this article: ros-mcdonnell, l., de-la-fuente-aragon, m.v., ros-mcdonnell, d., cardós carboneras, m. (2018). designing an environmental zone in a mediterranean city to support city logistics. international journal of production management and engineering, 6(1), 1-9. https://doi.org/10.4995/ijpme.2018.8771 int. j. prod. manag. eng. (2018) 6(1), 1-9creative commons attribution-noncommercial-noderivatives 4.0 international 1 mailto:lorenzo.ros@upct.es mailto:marivi.fuente@upct.es mailto:diego.ros@upct.es mailto:mcardos@doe.upv.es http://creativecommons.org/licenses/by-nc-nd/4.0/ different european projects, like civitas, together with the eu commission transport 2050 strategy, set goals for different transport modes, including co2-free city logistics in major urban centres by 2030 (eu-com, 2011; van roijen & quak, 2014; dotter, 2015). our research group is exploring different urban logistics measures to tackle the main problems caused by freight deliveries and pickups in the city centre. these solutions aim to enhance the efficiency of vehicles, and to reduce both traffic congestion and the environmental impacts due to freight delivery in the city in order to improve urban sustainability. this work is arranged as follows: sections 1 and 2 present a brief literature review about environmental zones (ez), european projects on urban logistics and different outlines developed in european cities; section 3 defines the proposal of an ez, its characteristics and the objectives to pursue with it, as well as the work methodology followed to study and implement it. section 4 presents the main actions that the local authority must take for the ez, the expected results after the first year of the ez in operation, and the initial measures taken by the city council. finally, this paper end with the conclusions drawn the research done on the developed ez. 2. environmental zones in eu an environmental zone (ez) is a defined geographical area that can be entered only by vehicles that meet specified emissions criteria (allen & wild, 2008). the purpose of an ez is to either restrict or charge the most polluting vehicles if they enter the ez when their emissions are over the set level. ezs are implemented into locations in which air pollution can (has reached) reach levels that are dangerous to public health. by introducing an ez it is hoped that air quality improves and this will reduce the health problems and fatalities associated with poor air quality. given this health hazard, the european union, and many countries around the world, have set air quality targets to be met, and low emission zones are being implemented (allen & wild, 2008; browne et al., 2012; dotter, 2015). this can be applied to different types of vehicles (only goods vehicles, a selection of motor vehicles, all motor vehicles). an ez, therefore, differs from the following types of access restrictions that can be placed on goods vehicles in urban areas: weight, length, time restrictions, loading capacity, etc.). however, the types of access restrictions can be implemented in addition to an ez. environmental zone schemes can take many forms according to their objectives, the geographical area they cover, the times at which the ez comes into force, the vehicle emissions standards required for vehicles to enter zones, the types of vehicles that need to comply with the ez, and the implementation and enforcement approaches used (allen & wild, 2008; russo & comi, 2012). environmental zones have been successfully implemented and run for several years in scandinavian countries, and are widely considered by other european cities (london, bologna, madrid), and countries (austria, italy, germany, the netherlands, denmark). several schemes have already been implemented, according to the problems and specific characteristics of the city or region. the implementation of an environmental zone usually requires cooperation between the national government and local authorities to help ensure a common system within a city, and further in a region and in the country (geroliminis & daganzo, 2005; witkowsky & kiba-janiak, 2014). given current eu policies and the link among the use of infrastructure, air quality and noise problems, ezs are relevant from a eu perspective in the road traffic restrictions context (geerling & stead, 2003). 3. proposal of an environmental zone in cartagena 3.1. methodology the paper presents a study carried out to define an environmental zone, with traffic restrictions for vehicles, in the historical centre of the spanish mediterranean city of cartagena, and the stages for its implementation, which will be properly performed by local authorities. the present research was conducted according to not only the recommendations of the work group in best urban freight solutions ii (allen & brown, 2008; allen & wild, 2008), but also to the taniguchi methodology (taniguchi et al., 1999), which is based on three fundamental aspects of an urban functional area: collecting data that allow to characterise and quantify transport in the study area; determining and describing any existing problems inside it; solving int. j. prod. manag. eng. (2018) 6(1), 1-9 creative commons attribution-noncommercial-noderivatives 4.0 international ros-mcdonnell, l., de-la-fuente-aragon, m.v., ros-mcdonnell, d. and cardós carboneras, m. 2 http://creativecommons.org/licenses/by-nc-nd/4.0/ the considered problems by defining alternatives for modelling the studied urban functional area. the main objectives of establishing this ez are to limit road traffic, and to promote walking and cycling, with the following benefits: reducing pollutants (atmospheric, noise, etc.), and supporting the tourist activities in the mediterranean city, and without detriment to commercial activities and supplying businesses located inside the ez. the steps of the work methodology followed to define this ez and its operation are: step 1: defining the geographic boundaries of the ez, fixing the entries and exits of the ez, and closed streets during its operation (figure 1). step 2: access control system and communication plan to citizenship. step 3: evaluating the parking areas inside the ez and a proposing deterrent parking areas outside the ez. step 4: establishing entry requirements for the ez (types of vehicles, schedules, special permissions, etc.). step 5: progressively implementing the ez. analysing the results of the various defined and taken actions. 3.2. definition of an environmental zone the proposed area to implement an ez covers the entire historical centre of cartagena, but does not include most of the later city expansion. this area covers approximately 1.55 km2, and is delimited by the large avenues and the city’s promenades. this is the most representative area of the city, where an urban plot is conditioned by the topography, narrow streets are irregular in width and of medium or short length, and most have been recently pedestrianised (peopch, 2005). the old town houses numerous historical buildings, whose use is usually administrative, educational, cultural, military or services, and they generate high flows of people in the city. the described set of elements is the basis of an urban fabric characteristic of a mediterranean city model, whose social life is cohesive with a system of neighbourhood diversity, public spaces, and historical and cultural buildings. figure 1 shows the different aspects that define the environmental zone (ez): the border of the ez: denoted by a blue dotted line. the entry and exit points of the ez (denoted by a double red arrow). the research group located four different entries and exits along the delimited perimeter. the current traffic directions figure 1. definition of the boundary of the environmental zone, entry and exit streets. int. j. prod. manag. eng. (2018) 6(1), 1-9creative commons attribution-noncommercial-noderivatives 4.0 international designing an environmental zone in a mediterranean city to support city logistics 3 http://creativecommons.org/licenses/by-nc-nd/4.0/ on the streets inside the ez were not changed for the area to better operate. the streets (marked by black strokes) are closed by heavy elements and mobile devices (e.g. flowerpot stands, planters, etc.) and are easily removable for future changes and activities in the delimited area (figures 2.1 and 2.2). figure 2.1. streets closed by removable elements. figure 2.2. streets closed by removable elements. 3.3. objectives of the environmental zone the creation of the ez in the city centre intends to reduce pollutants in the defined area (emissions of air pollutants, noise pollution, etc.), to encourage walking and cycling in the urban core, and to support the city’s touristic activity (southworth, 2005; eucom, 2011; browne et al., 2012; childers et al., 2014). these actions will improve air quality and cut road traffic, which will thus reduce respiratory diseases and the number of traffic accidents, which will have indirect effects on the city’s economy (southworth, 2003; lindholm, 2012; quak, 2015)). reducing commuting, encouraging cycling or simply walking increase the commercial activity in former heavy traffic zones (russo & comi, 2012; pulawska & starowicz, 2014). nor should we forgot the city’s important tourism aspect, which will benefit from the ez as tourists, many of whom come on cruiser trips, will enjoy a more complete experience of the city thanks to streets being decongested of motor vehicles. finally, it is stressed that the creation of the ez in the city centre of cartagena continues with the roadmap to further improve both the quality of life for neighbours and the sensations lived by tourists in the city. this process began two decades ago, and cartagena has gone from being one of the most polluted spanish cities to being a pole of tourist attraction, thanks to the different actions performed by the local authority to achieve its reconversion, and to combine powerful industry (outside the city) and booming tourism. 3.4. concentration areas inside the ez the work team analysed the population distribution and the current points where vehicles concentrate inside the ez. the population inside the ez is homogenously distributed in several sectors that compose the ez, except in two very specific sectors, which are being used as non-official parking areas. given the high concentration of vehicles in these two sectors, the entry into force of the ez will affect many current users, so the cartagena city council must propose alternatives in these cases that are to be considered (deterrent parking, shuttle buses, etc.). these nuclei where vehicles concentrate will be considered at the time the ez of cartagena is planned: sector 1: upct (campus “muralla del mar”): this sector houses two university faculties (etsii & etsit), which together contain around 4,000 students, many of whom need to use own vehicles. the car park is located in front of the two faculties and runs peripherally to their façades (see figure 3). sector 2: montesacro. this is a car park on the land around the mount sacro, where a large number of vehicles are park daily by the workers in the area and by the neighbours who live in this sector (see figure 4). int. j. prod. manag. eng. (2018) 6(1), 1-9 creative commons attribution-noncommercial-noderivatives 4.0 international ros-mcdonnell, l., de-la-fuente-aragon, m.v., ros-mcdonnell, d. and cardós carboneras, m. 4 http://creativecommons.org/licenses/by-nc-nd/4.0/ figure 3. car park around upct buildings. figure 4. car park around mount sacro. 4. actions to implement the environmental zone the ez in cartagena will be an open space supervised by cameras, with no access restriction elements, but with clearly identified entry and exit points. it will be a similar model to that which is currently operating in the city of london (geroliminis & daganzo, 2005; allen & wild, 2008; van rooijen & quak, 2014; witkowski & kiba-janiak, 2014), but it will be adapted to the characteristics of a medium-sized mediterranean city. to implement the ez, an integrated database will be created, formed by all the vehicle number plates whose users have indicated (via web) their intention to access the ez as they work, study or live inside the ez. to raise the awareness of the affected groups, the local council will open a 6-month adaptation period, during which the performance of the ez will be publicised by mass media (radio, tv, the council’s website, etc.). in addition, notices of non-compliance will be issued by the authorities, and will attach the sanction that applies to them if visitors to the ez commit the infraction again once the awareness campaign has ended. 4.1. general requirements to access the ez entry requirements for the different groups that claim access to the ez will be based on emission levels of airborne and/or malicious life pollutants (nox, co, hc, pm, hncm), type of engine (petrol or diesel), heavy or light commercial vehicles, trucks and buses, and types euro 0, i, ii, iii, iv, v, and vi. for electric vehicles, the total exemption of costs is proposed, while a 50% bonus of the amount envisaged to access the ez is proposed for hybrid vehicles. 4.1.1. entry requirements for residents and workers in the ez in spite of the benefits and improvements of the quality of life inside the ez, the implementation of an ez can be disrupt the daily life of the populations inside it: 1. entry requirements based on the hazardous potential of vehicles will not be applied immediately to residents, but will develop following this schedule: euro i january 2018, euro 2 january 2020, euro 3 – january 2022. in this way, the local authority ensures an acceptable deadline to replace the vehicles that do not comply with emission regulations. 2. for the workers whose job is located inside the ez, they will be guaranteed access provided they meet the general entry conditions and employers pay the access fee. access procedures will be carried out by employers on the website enabled for this purpose. int. j. prod. manag. eng. (2018) 6(1), 1-9creative commons attribution-noncommercial-noderivatives 4.0 international designing an environmental zone in a mediterranean city to support city logistics 5 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3. for students, whose educational or research centre is located inside the ez, they will have unlimited entries during the academic course. access procedures will be processed during the academic enrolment period. 4.1.2. entry requirements for delivery vehicles to the ez every company that undertakes its economic activity (e.g. product deliveries) inside the ez must comply with the general requirements to be able enter the ez. in the same way, the company will follow the administrative process to register in the database all the vehicles that will enter the ez. time slots shall be established for loading/unloading operations, for a maximum time of 30 minutes to carry out the operation and to leave the ez, during two time windows in the daytime: from 09.00h to 11:00h (in the morning) and from 16:00h to 18:00h (in the afternoon). in this way, each vehicle can access the ez only once a day, but can enter and exit up to a maximum of 3 times if its time inside the ez takes less than 45 min (optimal solution for fast messaging companies). 4.1.3. special access licenses to the ez this section presents exceptional situations that require access to the ez (not covered in the previous sections), and they will be attended to by a system of temporary access permits (from 3 days to a maximum of 15 days): licenses for temporary access granted to a vehicle that visits a resident, access licences for hospital stayings or other medical services, access licences for tourists, hotels and vacation apartments in the ez, other special licences. 4.2. control of access to the ez and sanctions an open character access to the ez has been chosen, and no physical elements will be installed that regulate entries (lifting barriers or pivots). the access control will be completely automated through a system of cameras located at the ez entry and exit points, a completely trustworthy and reliable system that ensures compliance (as in the london ez). the artificial vision system will recognise number plates and will determine if the access of this vehicle is allowed inside the ez. entry permits granted to residents, workers and students will be valid for one year, after which the these permits will no longer be valid and must be renewed following the same process (via the website, council office, etc.) as when it was first obtained. those vehicles that have accessed the ez without authorisation will be automatically sanctioned by attaching to the sanction a photograph taken while the infraction was committed. economic sanctions are foreseen for those vehicles that access or circulate inside the ez without a valid access license. this sanction will be imposed only on the same vehicle once a day. generally, all penalties may be claimed during the first 15 days after receiving them should any error have been made while issuing the penalty. 4.3. first actions developed to start the ez at the end of 2016, the department of sustainable development (cartagena city council) began to take different actions to progressively eliminate vehicles in the historic centre. one of the first proposed solutions was to create parkand-ride areas outside the ez, whose construction can lead to better ez performance since they combine the flexibility of a car and the efficiency of public transport. it will be necessary to devise a plan that defines the type of more adequate parking areas according to the ez, its performance and its functional environment. last christmas the local authorities launched two new free bus lines that connected two existing park-and-ride areas with the city centre (figures 5 and 6). the results were not those expected (only 7% of people visiting the historical centre came by bus) int. j. prod. manag. eng. (2018) 6(1), 1-9 creative commons attribution-noncommercial-noderivatives 4.0 international ros-mcdonnell, l., de-la-fuente-aragon, m.v., ros-mcdonnell, d. and cardós carboneras, m. 6 http://creativecommons.org/licenses/by-nc-nd/4.0/ due to lack of publicity about the new bus lines, their routes and stops, and the temporary duration of this campaign. the campaign has been repeated during two other festive periods, (easter and the “carthaginians and romans” festival last september). on both occasions, the use of free buses has progressively increased (from 7% to 15% and 20%, respectively), and the city council is considering making this service permanent with the mid-term implementation of the ez because making these bus lines permanent would encourage their use (pemc, 2017). finally, in order to ensure citizen security in cartagena given the large influx of tourists in summer and autumn, and as a result of the latest terrorist actions in some european cities, the local police have taken several actions that the research team considers are an example of bad practices (see figures 7.1 & 7.2). the local police have made it difficult to enter the main street using police vehicles. rather than a dissuasive measure, it can cause difficulties for disabled people to access and problems if large groups of people are present. figure 7.1. street closed by a police car on the main street of cartagena. figure 7.2. street closed by a police car on the main street of cartagena. figure 5. location of possible park-and-ride areas outside the environmental zone. figure 6. new bus lines between the ez and the park-andride areas. int. j. prod. manag. eng. (2018) 6(1), 1-9creative commons attribution-noncommercial-noderivatives 4.0 international designing an environmental zone in a mediterranean city to support city logistics 7 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.4. benefits of implementing an ez in cartagena as previously mentioned in section 3.2, cartagena is a medium-sized city. the area selected as the ez is characterised by having narrow streets of irregular widths and most have been recently pedestrianised. this means that as the city’s tourist centre, the daily urban logistics for the different establishments inside the ez to operate is becoming increasingly complex. a study recently conducted by the research group (de la fuente et al., 2015) establishes almost 6,500 motorised vehicle entries into an ez on one working day morning. of these, 15% were freight distribution vehicles, and 85% were public and private vehicles used to transport people. bicycle use was minimum (0.01%) compared to motorised vehicles. the heavy vehicle flow, along with the proximity to a large industrial and port area, implies that air quality levels (pm10, nox, co, etc.) are not in line with those expected for a medium-sized coastal city. the values recorded last month (november 2017) fell within the range [65-75%] of values allowed by the eu, but the local authorities hope to reduce them to go below 50%, which is the reason why they have collaborated to develop the ez in cartagena. in general, depending on the stringency of the ez standards the potential socio-economic effects of the ez in a city (based on the feasibility study of the london ez) are to become possible future benefits (geroliminis & daganzo, 2005; browne et al., 2012; van roijen & quak, 2014; dotter, 2015): benefits improved air quality –all pollutants (not just no2 and pm10) progress towards eu air quality limit values health benefits – lost time at work, nhs costs slight reduction in noise more attractive environment for companies and people safety benefits of newer vehicles dissadvantages higher potential business costs for companies (which could negatively affect attractiveness) stronger relative impact on road haulage, wholesale, trade, manufacturing sectors, building companies stronger relative impact on smaller companies given the results expected for the city of cartagena, and after implementing the ez and it operating for 1 year, it has been estimated that: more than 20-30% of motorised traffic inside the ez will reduce. the use of private vehicles (15-25%) will be replaced with non-motorised vehicles (bicycles, e-bikes, etc). nox emissions from heavy vehicles inside the zone will lower by 10% emissions of particulates will lower by 30% sales of the establishments inside the ez will increase by 10-20%. more shops will open (approx. 7.5%) there will be more health benefits and fewer hospital admissions, by 20% of present rates. 5. conclusions european cities are facing enormous challenges in accessibility and livability terms. congestion levels are still increasing. air pollution and noise disturb many people who live in city centres. most european cities are seen as living laboratories for conducting innovative urban sustainability measures. after studying the behaviour and performance model for the selected functional urban area (historical centre), in this paper the research group presents the ez defined at the city centre of a mediterranean city (cartagena), where it will either restrict or charge the most polluting vehicles that enter this geographical area. the research group is planning a set of possible solutions: entry permits for certain types of delivery vehicles and restrictions for other types. bundling and guided routes for goods deliveries. use of clean vehicle technologies, etc. the local administration will be in charge of executing the feasible proposed solutions, but it will require the collaboration and conviction of the population that lives and works in the ez. acknowledgements the research that has led to these results forms part of the work conducted by the students of the subject “industrial logistics” in the master of industrial engineering during academic year 2016-2017 at the technical university of cartagena. int. j. prod. manag. eng. (2018) 6(1), 1-9 creative commons attribution-noncommercial-noderivatives 4.0 international ros-mcdonnell, l., de-la-fuente-aragon, m.v., ros-mcdonnell, d. and cardós carboneras, m. 8 http://creativecommons.org/licenses/by-nc-nd/4.0/ references allen, j., brown, m. (2008). review of survey techniques used in urban freight studies. london: transport studies group, university of westminster. allen, j., wild, d. (2008). environmental zones in european cities. in d1.4 bestuffs policy and research recommendations iv, best urban freight solutions ii, co-ordination action, priority 1.6.2. sustainable surface transport. appleyard, d. (1981). livable streets. berkeley, ca: univ. of california press. browne, m., allen, j., nemoto, t., patier, d., visser, j. (2012) reducing social and environmental impacts of urban freight transport: a review of some major cities. procedia – social and behavioral sciences, 39, 19-33. https://doi.org/10.1016/j.sbspro.2012.03.088 bulkeley, h., betsill, m.m. (2005). rethink sustainable cities: multilevel governance and the urban politics of climate change. environmental politics, 14(1), 42-63. https://doi.org/10.1080/0964401042000310178 childers, d.l., pickett, s.t.a., grove, j.m, ogden, l., whitmer, a. (2014). advancing urban sustainability theory and action: challenges and opportunities. landscape and urban planning, 125, 320-328. https://doi.org/10.1016/j.landurbplan.2014.01.022 crawford, j. (2000). carfree cities. utrecht, the netherlands: international books. dablanc, l. (2008). urban goods movement and air quality policy and regulation issues in european cities. journal of environmental law, 20(2), 245-266. https://doi.org/10.1093/jel/eqn005 de la fuente, m.v., ros-mcdonnell, d., nyerges, l., bajor, p., ros-mcdonnell, l. (2015). analysis of logistics flows in an urban functional area. application to cartagena. 9th international conference on industrial engineering and industrial management, aveiro (portugal). dotter, f. (2015). cleaner, safer and more efficient freight transport in cities. civitas insight, 3, november 2015. engwicht, d. (1999). street reclaiming: creating livable streets and vibrant communities. gabriola island, british columbia: new society publishers. eu-com (2011). roadmap to a single european transport area – towards a competitive and resource efficient transport system. white paper. european commission, brussels. geerling, h., stead, d. (2003). the integration of land use planning, transport and environment in european policy and research. transport policy, 10, 187-196. https://doi.org/10.1016/s0967-070x(03)00020-9 geroliminis, n, & daganzo c. (2005). a review of green logistics schemes used in cities around the world. working paper ucb-itsvwp-2005-5. escholarship. university of california. lindholm, m. (2012). how local authority decision makers address freight transport in the urban area. procedia – social and behavioral sciences, 39, 134-145. https://doi.org/10.1016/j.sbspro.2012.03.096 newman, p., kenworthy, j. (1999). sustainability and cities: overcoming automobile dependence. washington, d.c.: island press. peopch (2005). plan especial de ordenación y protección del conjunto histórico. memoria justificativa y explicativa. ayuntamiento de cartagena. pemc (2017) plan estratégico m17 del municipio de cartagena. ayuntamiento de cartagena, mayo 2017. pulawska, s., starowicz, w. (2014). ecological urban logistics in the historical centers of cities. procedia – social and behavioral sciences, 151, 282-294. https://doi.org/10.1016/j.sbspro.2014.10.026 quak, h.j. (2015). access restrictions and local authorities’ city logistics regulation in urban areas. in: city logistics, mapping the future. ed. crc press. russo, f., comi, a. (2012) city characteristics and urban goods movements: a way to environmental transportation system in a sustainable city. procedia – social and behavioral sciences, 39, 61-73. https://doi.org/10.1016/j.sbspro.2012.03.091 southworth, m. (2003). measuring the livable city. built. environ., 29(4), 3343-3354. https://doi.org/10.2148/benv.29.4.343.54293 southworth, m. (2005). designing the walkable city. j. urban plann. dev.,131(4), 246-257. https://doi.org/10.1061/(asce)07339488(2005)131:4(246) taniguchi, e., thompson, r.g., yamada, t. (1999). modelling city logistics. ed. institute of systems science research, kyoto. taniguchi , e. (2014). concepts of city logistics for sustainable and liveable cities. procedia – social and behavioral sciences, 151, 310-317. https://doi.org/10.1016/j.sbspro.2014.10.029 van rooijen, t., quak, h. (2014). city logistics in the european civitas initiative. procedia – social and behavioral sciences, 125, 312-325. https://doi.org/10.1016/j.sbspro.2014.01.1476 witkowski, j., kiba-janiak, m. (2014). the role of local governments in the development of city logistics. procedia – social and behavioral sciences, 125, 373-385. https://doi.org/10.1016/j.sbspro.2014.01.1481 int. j. prod. manag. eng. (2018) 6(1), 1-9creative commons attribution-noncommercial-noderivatives 4.0 international designing an environmental zone in a mediterranean city to support city logistics 9 https://doi.org/10.1016/j.sbspro.2012.03.088 https://doi.org/10.1080/0964401042000310178 https://doi.org/10.1016/j.landurbplan.2014.01.022 https://doi.org/10.1093/jel/eqn005 https://doi.org/10.1016/s0967-070x(03)00020-9 https://doi.org/10.1016/j.sbspro.2012.03.096 https://doi.org/10.1016/j.sbspro.2014.10.026 https://doi.org/10.1016/j.sbspro.2012.03.091 https://doi.org/10.2148/benv.29.4.343.54293 https://doi.org/10.1061/(asce)0733-9488(2005)131:4(246) https://doi.org/10.1061/(asce)0733-9488(2005)131:4(246) https://doi.org/10.1016/j.sbspro.2014.10.029 https://doi.org/10.1016/j.sbspro.2014.01.1476 https://doi.org/10.1016/j.sbspro.2014.01.1481 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2017.7423 received 2016-12-09 accepted: 2017-06-21 aligning organizational pathologies and organizational resilience indicators manuel morales allendea1, cristina ruiz-martina2*, adolfo lopez-paredesa3 and jose manuel perez ríosa4 ainsisoc. universidad de valladolid. spain a1 manuelmoralesallende@gmail.com, a2 cruiz@eii.uva.es, a3 aparedes@eii.uva.es, a4 rios@uva.es abstract: developing resilient individuals, organizations and communities is a hot topic in the research agenda in management, ecology, psychology or engineering. despite the number of works that focus on resilience is increasing, there is not completely agreed definition of resilience, neither an entirely formal and accepted framework. the cause may be the spread of research among different fields. in this paper, we focus on the study of organizational resilience with the aim of improving the level of resilience in organizations. we review the relation between viable and resilient organizations and their common properties. based on these common properties, we defend the application of the viable system model (vsm) to design resilient organizations. we also identify the organizational pathologies defined applying the vsm through resilience indicators. we conclude that an organization with any organizational pathology is not likely to be resilient because it does not fulfill the requirements of viable organizations. key words: organizational resilience, viable system model, organizational pathologies. 1. introduction building resilience is a key topic in many research fields such as management, engineering, psychology or ecology. governments are investing resources to develop resilient institutions, communities, organizations and individuals. for example, the australian government developed a tool to assess organizational resilience. the proposal is a questionnaire based on thirteen indicators to measure organizational resilience potential. they classify the indicators on three factors: leadership and culture, networks and partnerships and change readiness. the us department of homeland security relates resilience to three main concepts: adaptation to changing conditions, withstand disruptions and ensure rapid recovery. the european commission (2017) for the horizon 2020 program has identified “state and societal resilience” among the top five priorities of the european union’s external actions. the european union action plan for resilience (european commission 2016) outlines three priorities in the area of resilience: (1) support the development and implementation of national resilience capacity, (2) promote innovation and learning capacities to advocate resilience and (3) develop tools and methodologies to improve and measure resilience. being broadly and widely used the term resilience, there is not entire consensus about its definition. some authors state that resilience was first introduce in psychology (coutu, 2002). other authors (henry to cite this article: morales allende, m., ruiz-martin, c., lopez-paredes, a., perez ríos, j. m. (2017). aligning organizational pathologies and organizational resilience indicators. international journal of production management and engineering, 5(2), 107-116. https://doi.org/10.4995/raet.2017.7423 int. j. prod. manag. eng. (2017) 5(2), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international 107 mailto:manuelmoralesallende@gmail.com mailto:cruiz@eii.uva.es mailto:aparedes@eii.uva.es mailto:rios@uva.es http://creativecommons.org/licenses/by-nc-nd/4.0/ and ramirez-marquez, 2010; annarelli and nonino, 2016) say that the concept resilience was popularized after holling (1973), “resilience and stability of ecological systems”. psychology and ecology are not the only research fields where researches have focused their attention on resilience. it has also been studied in other fields including management (disaster management, organization management or community management), sociology or engineering. the dispersion of the research among so many fields may be the cause of a lack of widely accepted definition of resilience. despite the research in different fields seems to be isolated from each other, they are not. to build resilient organizations we need resilient individuals (mallak, 1997; doe, 1994; biggs et al., 2012), resilient supply chains (sheffi, 2007) and resilient infrastructure (bell, 2002; erol et al., 2009). to build resilient organizations we also need to apply resilient engineering principles (righi et al., 2015). resilient communities (kendra and wachtendorf, 2003; lee et al., 2013) and resilient territories (gilly et al., 2014) are built over resilient organizations. in this work, we will focus on organizational resilience. our contribution is to establish a relation between organizational resilience and other theories in management. specifically, we relate the organizational pathologies identified applying the viable system model (vsm) and organizational resilience indicators. grounded in the common properties that resilient and viable organization share, we support that an organization with any organizational pathology cannot be considered resilient because it is not viable. our final objective is to improve the level of organizational resilience. the identification of the pathologies will give the organization the path to improve its viability and therefore its resilience. the rest of the paper is organized as followed. in section 2, the concept of organizational resilience, the factors that contribute to organizational resilience and how it is measured are reviewed. in section 3, the application of the vsm to identify organizational pathologies is explained. in section 4, the application of the vsm and the identification of organizational pathologies to design resilient organizations are defended. in section 5, the results of the research are presented. some organizational pathologies and the scores obtained in different resilience indicators are aligned. finally, in section 6, the conclusions of this work are presented. 2. organizational resilience despite there is not a widely and accepted definition of organizational resilience, we agree with the authors that consider organizational resilience as an ability, capacity or capability to deal with disruptive events (mallak, 1997; hamel and valikangas, 2003; starr et al., 2003; sheffi and rice jr., 2005; jackson, 2007; tillement et al., 2009; hollnagel, 2010; lengnickhall et al., 2011; manyena, 2006; annarelli and nonino, 2016; linnenluecke et al., 2012). to deal with disruptive events organizations need a set of abilities or capabilities. we have reviewed the works that identify those abilities and we have found that there is not a consensus. in ruiz-martin et al. (2017a), the authors identified the most common and repeated characteristics of resilient organizations in the literature. these characteristics include the capability of: building situation awareness, managing organization’s vulnerabilities, having resources, having improvisation capacity, being able to anticipate to events, being agile, having learning capacity, collaboration, having resilient individuals and being flexible and redundant. regarding to the measurement of organizational resilience there are two main streams: the measurement of organizational resilience potential and the measurement of resilience after a disruptive event has occurred (ruiz-martin et al., 2017b). the assessment of organizational resilience potential is usually based on evaluating the characteristics or abilities an organization possess. despite there is a broad literature about organizational resilience indicators (horne iii and orr, 1998; bhamidipaty et al., 2007; somers, 2009; sanchis and poler, 2013; lee et al., 2013; seville 2009; whitman et al., 2013), we have chosen the indicators proposed by lee et al. (2013) because they provide a complete benchmark tool to do the evaluation process. they evaluate the organizational resilience potential based on a questionnaire with 53 items. each item is evaluated based on an eight-point scale. the lower score is achieved if the organization strongly disagrees with the statement in the item. the higher score is obtained when the organization strongly agree with the statement. the 53 items are classified in to 13 int. j. prod. manag. eng. (2017) 5(2), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international morales allende, m., ruiz-martin, c., lopez-paredes, a. & perez ríos, j. m 108 http://creativecommons.org/licenses/by-nc-nd/4.0/ indicators. the indicators are grouped in two factors: adaptive capacity and planning. the indicators included in adaptive capacity are: minimization of silos. it is related to the minimization of barriers in the organization, especially those ones related to communication. internal resources. the organization has enough resources to conduct its business as usual, but it is also able to provide extra resources when needed. staff engagement and involvement. the staff understands the link between their work, the resilience of the organization and its success. moreover, the staff is encouraged to use their problem solving capabilities. information and knowledge. the information is available when needed and stored in different locations. the staff is flexible, so different people can fill key roles. leadership. there is a good leader in the organization. the organization strategies and programs are continuously reviewed. innovation and creativity. the use of novel ways to solve problems is encouraged and rewarded in the organization. decision making. people have the authority to make decision based on their skills. during crisis, authority is delegated to be able to respond to the crisis. situation monitoring and reporting. staff is encouraged and rewarded for performing monitoring activities. early warning signals are rapidly reported to organizational leaders. the indicators included in planning are: planning strategies. there are plans to manage organization vulnerabilities. participation in exercise. there are simulacrums in the organization to practice and evaluate the plans. proactive posture. the organization is prepared to respond to early warning signals. external resource. the organization has a plan to access resources from other organizations when needed. recovery priorities. the priorities are set and the organization understands its minimum operating requirements. analyzing the 13 indicators, we get an idea of how the organization is prepared to respond to a crisis. however, the resilience of the organization will depend on the kind of risk it is dealing with. the level of resilience the organization has exhibited cannot be measured until the risk has occurred. to measure the level of resilience after a disruptive event, henry and ramirez-marquez (2010) propose to evaluate the level of recovery of the organization against its loses. erol, henry and sauser (2010); erol, henry, sauser, et al. (2010) also include the recovery time, the initial vulnerability and the potential loss averted. 3. the viable system model (vsm) and organizational pathologies the vsm was developed by beer (1981). it is a scientific framework, based on organizational cybernetics, which establishes the necessary and sufficient conditions for the viability of a system. according to the vsm a viable system must have the capacities of self-regulation, learning, adaptation, and evolution in the management field, it is applied to the design and study of organizations and its processes (pérez ríos, 2012). the aim is to design viable organizations. it means to design organizations able to survive regardless the changes in its environment. according to beer, the viability of a system is based on the existence of a set of functional systems in the organization and a set of relationships among these functional systems and the environment. the systems and the relations among them are represented in figure 1. an important aspect of the viability of a system is its recursive property: all viable systems contain viable systems and are themselves contained in viable systems. each of the systems defined by beer has a specific functionality on the organization. system 1 is in charge of the production and delivery. it produces the goods or services in the organization. it is composed by several operational units. they can be suborganizations or different divisions in the company. system 2 has to guarantee a harmonic and stable functioning of the organizational units in system 1. the main role of system 3 is to optimize the functioning of the whole set of operational units that compose system. it is responsible for the “here and now” of the organization. system 4 has to int. j. prod. manag. eng. (2017) 5(2), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international aligning organizational pathologies and organizational resilience indicators 109 http://creativecommons.org/licenses/by-nc-nd/4.0/ monitor the organization environment, focusing on the “outside and then” of the organization. its aim is to be prepared for changes. system 5 is responsible for defining the ethos, the vision and the identity of the organization. it takes care of the normative decisions. according to pérez ríos (2012), any shortage on these systems or in their communication mechanisms is translated into different organizational pathologies. any organizational pathology may cause the disappearance of the organization, at least as an independent entity. pérez ríos (2012) classifies the organizational pathologies into three main groups: structural pathologies, functional pathologies and information pathologies. structural pathologies are related to how the organization is designed and how it copes with environmental variability. there are four structural pathologies: non-existence of vertical unfolding, lack of first recursion levels, lack of middle recursion levels and entangled vertical unfolding with interrelated memberships. functional pathologies are related to the adequacy of the organization’s systems to the prescriptions made by the vsm. functional pathologies are classified based on the system they affect and those ones that affect the whole organization. functional pathologies related to system 5 are: ill-defined identity, institutional schizophrenia, lack of metasystem (i.e. collapse of system 5 in system 3) and inadequate representation in higher levels. functional pathologies related to system 4 are headless chicken (i.e. the organization does not monitor the environment and is not able to adapt to changes) and dissociation between system 4 and 3. functional pathologies related to system 3 are: inadequate management style, schizophrenic system 3, week connection between system 3 and 1 and hypertrophy of system 3. the functional pathology related to system 3* is the lack or insufficient development this system. system 2 can present two pathologies: disjointed behavior within system 1 and authoritarian system 2. the pathology related to system 1 are “autopoietic beasts” (i.e organizations that only focus on individual goals and do not take into account the whole) and dominance figure 1. viable system model, adapted from beer, 1981 (pérez ríos, 2012). used with author permission. int. j. prod. manag. eng. (2017) 5(2), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international morales allende, m., ruiz-martin, c., lopez-paredes, a. & perez ríos, j. m 110 http://creativecommons.org/licenses/by-nc-nd/4.0/ of system 1. the pathologies related to the whole organization are organizational autopoietic beasts and lack of metasystem. information system and communication channel pathologies are related to the malfunctioning of the communication and information system. information pathologies are the lack of information systems, the fragmentation of information systems and insufficient or lack off key communication and algedonic channels. 4. applying the viable system model to design resilient organizations in ruiz-martin et al. (2017), the authors discuss the relation between resilient and viable organizations. they highlight that one of the aims of resilient organizations is to recover from challenges or disruptive events. taking into account the definition of a viable organization (i.e. those organizations able to survive despite changes in the environment), a resilient organization needs to be also viable. moreover, they also found that resilient organization should have, at least, the capacities of viable organizations (self-regulation, learning, adaptation, and evolution). they conclude that resilient organizations fit the vsm principles and therefore it is a valid framework to design resilient organizations. more specifically, they propose the use of the methodological framework introduced by pérez ríos (2010) to design a resilient organization. the methodological framework is structured into four steps. the first step is to recognize the identity and the purpose of the organization. the second step is to identify how the organization deals with environmental complexity and to design the vertical structure of suborganizations. the third step is to analyse the proposed structure and check that all the needed elements for the viability of the organization are represent. the final step is to ensure the coherence among the suborganizations identities and purposes and to check the suborganizations connections. in this work, we go a step forward and analyze the organization following both the organizational resilience approach and the vsm framework. we determine the level of organizational resilience potential based on resilience indicators. for this purpose, we use the benchmark tool designed by lee et al. (2013) because they provide a complete questionnaire to do the assessment. we identify organizational pathologies using the score obtained in the resilience questionnaire. we support that the identification and handling of organizational pathologies is the path to improve organizational resilience potential. in section 5, we explain the relation between the indicators proposed by lee et al. (2013) to evaluate organizational resilience potential and the organizational pathologies identified by pérez ríos (2012) applying the vsm. we defend that the identification of the pathologies will give the organization the path to improve its viability and therefore its resilience. 5. identification of organizational pathologies using resilience indicators in table 1, we present the relation between the indicators proposed by lee et al. (2013) and the organizational pathologies introduced by pérez ríos (2012). in the table, we represent with an “x” the set of indicators that will be used to diagnose each pathology. the organization is likely to suffer the pathology if the score on the highlighted indicators is low. to relate the resilience indicators with the organizational pathologies, we have analyzed the questions proposed by lee et al. (2013) to measure each indicator. if a low score in any of the proposed questions to evaluate an indicator can be used as evidence of an organizational pathology, we related the indicator with the pathology with an “x” in table 1. the statements we make along this section to defend the relation between the indicator and the pathology are the authors’ hypothesis based on the organizational properties evaluated by the indicators and the literature regarding the organizational pathologies introduced by pérez ríos (2012). identification of structural pathologies the structural pathologies can be identified by a low score in the following indicators: leadership, decision making and planning strategies. int. j. prod. manag. eng. (2017) 5(2), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international aligning organizational pathologies and organizational resilience indicators 111 http://creativecommons.org/licenses/by-nc-nd/4.0/ a low score on leadership reveals a lack of a good organization structure. a well-designed organization should have clear leadership positions. if the score on decision-making is low, it will probably mean that the decisions are not taken at the right levels. not having a good score on planning strategies would mean that there are no plans to deal with environmental variability, that the plans does not fit the organization purposes or that there is a lack of planning reviews to keep plans up to date. despite we cannot detect which specific structural pathology the organization has, these indicators reveal a lack of organization design adequacy. identification of functional pathologies as we mentioned in section 3, functional pathologies are related to the adequacy of the organization’s systems to the prescriptions made by the vsm. they are classified based on which system the pathologies affect. functional pathologies related to system 5 ill-defined identity and institutional schizophrenia is noticed by a low score in the following indicators: minimization of silos, leadership, decision-making, planning strategies and recovery priorities. if the identity (vision and mission) of the organization is not well defined, there will be undesirable behaviors, a lack of leadership according to the organizational objectives and a lack of proper decision-making strategies and role assignment. moreover, it is not possible to define the organization strategy and recovery priorities if the vision, mission and objectives of the company are not defined. if the organization objectives are not defined, it is very luckily that system 5 constantly change its opinion based on what is more convenient at each moment without thinking long term. lack of metasystem is identified through a low score in the following indicators: leadership and decision-making. the decision are not taken by the right person and the leader of the “here and now” in the organization does not have the information and power to take the decisions. inadequate representation in higher levels is observed by a low score in leadership and information and knowledge. the knowledge is not properly distributed among different organizational levels. the leaders cannot perform well in their job because of the disconnection and lack of information. functional pathologies related to system 4 headless chicken (i.e. the organization does not monitor the environment and is not able to adapt to changes) is noticed by a low score in innovation and creativity, situation monitoring and reporting, planning strategies and proactive posture. a lack of a well design system 4 carries a mismanaging of the “outside and then” in the organization. it is translated in a lack of innovative and creative solutions to adapt to the environment, an insufficient monitoring of the environment and therefore a lack of information to develop sound strategies. dissociation between system 4 and 3 is observed by a low score in internal resources, information and knowledge, situation monitoring and reporting, planning strategies and recovery priorities. the resources would not be used properly as system 3 and 4 would apply their own criteria. the incoordination between the two systems would carry lack of information sharing; the strategies and the recovery priorities may be incoherent. moreover, system 3 would not have information about the evolution of the environment to adapt for the future. functional pathologies related to system 3 inadequate management style and hypertrophy of system 3 are noticed by a low score in information and knowledge because the information is not properly transmitted among the system. system 3 does not absorb enough environment variability. the score in leadership, decision making and planning strategies would also be low due to the over involvement in task that are not among system 3 competencies. schizophrenic system 3 is manifested by a low score in internal resources, information and knowledge and proactive posture. the low score is due to the constant changes made by system 3 in the decisions taken. there would be incongruences in the resource assignments, in the information transmitted or in the criteria to detect changes in the environment based on early warning signals. weak connection between system 3 and 1 is perceived by a low score in internal resources, information and knowledge, leadership, situation monitoring and reporting and recovery priorities. there would be internal competing for resources in system 1, each int. j. prod. manag. eng. (2017) 5(2), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international morales allende, m., ruiz-martin, c., lopez-paredes, a. & perez ríos, j. m 112 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 1. relation of lee et al. (2013) organizational resilience indicators and perez ríos (2012) organizational pathologies. indicators pathologies adaptive capacity planning m in im iz at io n of s ilo s in te rn al re so ur ce s st af f e ng ag em en t a nd in vo lv em en t in fo rm at io n an d kn ow le dg e l ea de rs hi p in no va tio n an d cr ea tiv ity d ec is io n m ak in g si tu at io n m on ito ri ng a nd re po rt in g pl an ni ng s tr at eg ie s pa rt ic ip at io n in e xe rc is e pr oa ct iv e po st ur e e xt er na l r es ou rc e r ec ov er y p ri or iti es structural pathologies non-existence of vertical unfolding x x x x lack of first recursion levels x x x x lack of middle recursion levels x x x x entangled vertical unfolding with interrelated memberships x x x x functional pathologies related to system 5 ill-defined identity x x x x x institutional schizophrenia x x x x x lack of metasystem x x inadequate representation in higher levels x x functional pathologies related to system 4 headless chicken x x x x dissociation between system 4 and 3 x x x x x functional pathologies related to system 3 inadequate management style x x x x schizophrenic system 3 x x x week connection between system 3 and 1 x x x x x hypertrophy of system 3 x x x x functional pathologies related to system 3* lack or insufficient development of system 3* x x x x x x functional pathologies related to system 2 disjointed behavior within system 1 x x x x x authoritarian system 2 x x functional pathologies related to system 1 “autopoietic beasts” x x x x x x dominance of system 1 x x x x x x functional pathologies related to the whole organization organizational autopoietic beasts x x x x lack of metasystem x x x x x x x x x x x x x information system and communication channel pathologies lack of information systems x x x x fragmentation of information systems x x x x lack of key communication channels x x x x lack of or insufficient algedonic channels x x x x int. j. prod. manag. eng. (2017) 5(2), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international aligning organizational pathologies and organizational resilience indicators 113 http://creativecommons.org/licenses/by-nc-nd/4.0/ operation unit would set their own recovery priorities, the information flow and knowledge sharing among system 1 components would be tense due to the lack of authority, and the employees in system 1 would perceive a lack of leadership. additionally, system 3 would not be able to monitor the well function of system 1 and set new guidelines with needed. functional pathologies related to system 3* the lack or insufficient development of system 3* is observed by a low score in minimization of silos, staff engagement and involvement, leadership, planning strategies, participation in exercises and recovery priorities. the malfunctioning of audits would difficult the homogenization process in organizational behaviors and will show the lack of commitment of the staff with the work methodologies set in the organization. moreover, this pathology may be caused by a weak leadership or lack of planning strategies to develop the audit system. if there is no audit there is no guarantee that the recovery priorities in system 1 are well implemented and that the participation in exercises is done. functional pathologies related to system 2 the disjointed behavior within system 1 is manifested by a low score in internal resources, information and knowledge, decision making, situation monitoring and reporting and recovery priorities. if the system 2 does not transmit the decision made in system 3, operational units in system 1 will make their own decision, set their own recovery priorities and will fight for resources. an authoritarian system 2 is noticed by a low score in staff engagement and involvement and leadership. the staff is not motivated and they perceive that the organization is too bureaucratic. the staff does not understand the purpose of the tasks they are assigned. they perceive them as a waste of time. the units in system 1 may feel that the organization lacks from a leader. functional pathologies related to system 1 the “autopoietic beasts and dominance of system 1 pathologies are identified by a low score in minimization of silos, internal resources, information and knowledge, leadership, planning strategies and recovery priorities. each operational unit has its own goals and they do not care about the whole organization. the same occurs when the system 1 dominates the whole organization. the low score in the indicators would be due to the different ways of working, the fights for internal resources, the lack of leadership that controls system 1, the lack of information shearing among operational units or organization system components, the lack of coherence in planning strategies and the establishment if individual recovery priories. functional pathologies related to the whole organization organizational autopoietic beasts are revealed by a low score in staff engagement and involvement, leadership, decision making and planning strategies. the low score is due to the existence of systems that does not cooperate with each other and they just focus on their individual objectives. this behavior would be translated in low staff engagement, leaders that are overlapping other staff responsibilities and decision would not be taken at the right level. additionally the hypertrophic system would have too much power when developing the organization strategies. lack of metasystem is mainly recognized by a low score in planning strategies. if the organization lacks from metasystem there is nobody in charge of setting organizations strategies and priorities. we consider that an organization without metasystem is luckily to have a low score in every indicator. identification of information system and communication channel pathologies information system and communication channel pathologies are detected by a low score in the following indicators: minimization of silos, internal resources, information and knowledge and innovation and creativity. the lack of well-defined communication structures and communication channels would carry the inability to track staff behaviors and to correct undesirable ones. therefore, the score in the minimization of silos indicator would be low. there would also be conflict to get the common resources (i.e. a low score in internal resources). the important information and instructions would not be transmitted on time and it would be more difficult to share the knowledge across the organization (i.e. low score in information and knowledge). the score in innovation and creativity would also be low as these processes require the involvement of different int. j. prod. manag. eng. (2017) 5(2), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international morales allende, m., ruiz-martin, c., lopez-paredes, a. & perez ríos, j. m 114 http://creativecommons.org/licenses/by-nc-nd/4.0/ organization departments and good communication is a key factor for their success. as we explained in structural pathologies, the indicators are not enough to specify the type of information pathology, but they provide evidences of a malfunctioning of the organization in this area. 6. conclusion in this work, we have studied the alignment of organizational resilience assessment through indicators and some organizational pathologies already identified in the management literature. we have presented how we can identify a path to improve organizational resilience combining an analysis based in resilience indicators and the vsm methodological framework. the vsm provides a formal methodological framework that will support the organizations to be more resilient. though a questionnaire to the organization, we obtain a score in several indicators that reflect the organizational resilience potential. the score in these indicators is used to identify the pathologies an organization can suffer according to the vsm. we defend that treating these pathologies, the organization would improve its viability and therefore its level of resilience. future research lines will aim to validate the proposed relation between the resilience indicators and the pathologies of the organization. we will provide the questionnaire to different organizations to analyze the relation between the scores they obtain and the pathologies they present. we will also use the empirical study to establish score thresholds for the different pathologies. the “x” in table 1, would be translated in a score range. acknowledgements this research has been partially supported by banco santander and universidad de valladolid. references annarelli, a., nonino, f. (2016). strategic and operational management of organizational resilience: current state of research and future directions. omega, 62: 1–18. https://doi.org/10.1016/j.omega.2015.08.004 australian government, organizational resilience. available at: https://www.organisationalresilience.gov.au/pages/default.aspx [accessed july 12, 2017]. beer, s. (1981). brain of the firm: the managerial cybernetics of organization, j. wiley new york. bell, m. (2002). the five principles of organizational resilience. gartner research. [online], 2–4. bhamidipaty, a., lotlikar, r., banavar, g. (2007). rmi: a framework for modeling and evaluating the resiliency maturity of it service organizations. in ieee international conference on services computing (scc 2007). salt lake city, ut, 300–307. https://doi. org/10.1109/scc.2007.94 biggs, d., hall, c.m., stoeckl, n. (2012). the resilience of formal and informal tourism enterprises to disasters: reef tourism in phuket, thailand. journal of sustainable tourism, 20(5): 645–665. https://doi.org/10.1080/09669582.2011.630080 coutu, d.l. (2002). how resilience works. harvard business review, 80(5), 46-56. doe, p.j. (1994). creating a resilient organization. canadian business review, 21: 22. erol, o., henry, d., sauser, b., manosouri, m. (2010). perspectives on measuring enterprise resilience. in 2010 ieee international systems conference proceedings, syscon 2010: 587-592. https://doi.org/10.1109/systems.2010.5482333 erol, o., henry, d., sauser, b. (2010). exploring resilience measurement methodologies. in 20th annual international symposium of the international council on systems engineering, incose 2010: 302–322. https://doi.org/10.1002/j.2334-5837.2010.tb01072.x erol, o., mansouri, m., sauser, b. (2009). a framework for enterprise resilience using service oriented architecture approach. in 2009 ieee international systems conference proceedings: 127–132. https://doi.org/10.1109/systems.2009.4815785 european commission (2016). building resilience : the eu's approach. available at: http://ec.europa.eu/echo/files/aid/countries/factsheets/ thematic/resilience_en.pdf [accessed july 12, 2017]. european commission (2017). resilience as a strategic priority of the external action of the eu. roadmap ares(2017)1137007. available at: https://ec.europa.eu/info/law/better-regulation/initiatives/ares-2017-1137007_en [accessed july 12, 2017]. gilly, j.p., kechidi, m., talbot, d. (2014). resilience of organisations and territories: the role of pivot firms. european management journal, 32(4): 596–602. https://doi.org/10.1016/j.emj.2013.09.004 hamel, g., valikangas, l. (2003). the quest for resilience. harvard business review, 81(9): 52–65. int. j. prod. manag. eng. (2017) 5(2), 107-116creative commons attribution-noncommercial-noderivatives 4.0 international aligning organizational pathologies and organizational resilience indicators 115 https://doi.org/10.1016/j.omega.2015.08.004 https://www.organisationalresilience.gov.au/pages/default.aspx https://doi.org/10.1109/scc.2007.94 https://doi.org/10.1109/scc.2007.94 https://doi.org/10.1080/09669582.2011.630080 https://doi.org/10.1109/systems.2010.5482333 https://doi.org/10.1002/j.2334-5837.2010.tb01072.x https://doi.org/10.1109/systems.2009.4815785 http://ec.europa.eu/echo/files/aid/countries/factsheets/thematic/resilience_en.pdf http://ec.europa.eu/echo/files/aid/countries/factsheets/thematic/resilience_en.pdf https://ec.europa.eu/info/law/better-regulation/initiatives/ares-2017-1137007_en https://doi.org/10.1016/j.emj.2013.09.004 http://creativecommons.org/licenses/by-nc-nd/4.0/ henry, d., ramirez-marquez, j.e. (2010). a generic quantitative approach to resilience: a proposal. in 20th annual international symposium of the international council on systems engineering, incose 2010: 291–301. https://doi.org/10.1002/j.2334-5837.2010.tb01071.x holling, c.s. (1973). resilience and stability of ecological systems. annual review of ecology and systematics: 1–23. https://doi. org/10.1146/annurev.es.04.110173.000245 hollnagel, e. (2010). how resilient is your organisation? an introduction to the resilience analysis grid (rag). sustainable transformation: building a resilient organization. horne iii, j.f., orr, j.e. (1998). assessing behaviors that create resilient organizations. employment relations today, 24(4): 29–39. jackson, s. (2007). a multidisciplinary framework for resilience to disasters and disruptions. journal of integrated design and process science, 11(2): 91–108. kendra, j., wachtendorf, t. (2003). elements of resilience after the world trade center disaster: reconstituting new york city’s emergency operations centre. disasters, 27(1): 37–53. https://doi.org/10.1111/1467-7717.00218 lee, a.v, vargo, j., seville, e. (2013). developing a tool to measure and compare organizations’ resilience. natural hazards review, 14: 29–41. https://doi.org/10.1061/(asce)nh.1527-6996.0000075 lengnick-hall, c.a., beck, t.e., lengnick-hall, m.l. (2011). developing a capacity for organizational resilience through strategic human resource management. human resource management review, 21(3): 243–255. https://doi.org/10.1016/j.hrmr.2010.07.001 linnenluecke, m.k., griffiths, a., winn, m. (2012). extreme weather events and the critical importance of anticipatory adaptation and organizational resilience in responding to impacts. business strategy and the environment, 21(1): 17–32. https://doi.org/10.1002/ bse.708 mallak, l.a. (1997). how to build a resilient organization. in proceedings of the industrial engineering solutions 1997 conference. miami, 170–177. manyena, s.b. (2006). the concept of resilience revisited. disasters, 30(4): 433–450. https://doi.org/10.1111/j.0361-3666.2006.00331.x pérez ríos, j. (2012). design and diagnosis for sustainable organizations: the viable system method, springer heidelberg, new york, dordrecht, london. pérez ríos, j. (2010). models of organizational cybernetics for diagnosis and design. kybernetes, 39(9/10): 1529-1550. https://doi. org/10.1108/03684921011081150 righi, a.w., saurin, t.a., wachs, p. (2015). a systematic literature review of resilience engineering: research areas and a research agenda proposal. reliability engineering & system safety, 141: 142-152. https://doi.org/10.1016/j.ress.2015.03.007 ruiz-martin, c., pérez rios, j.m., et  al. (2017). the application of the viable system model to enhance organizational resilience. in advances in management engineering: 88–100. https://doi.org/10.1007/978-3-319-55889-9_5 ruiz-martin, c., lopez-paredes, a., wainer, g. (2017). organizational resilience: conceptualization and measurement. submitted. sanchis, r., poler, r. (2013). definition of a framework to support strategic decisions to improve enterprise resilience. in ifac proceedings volumes (ifac-papersonline): 700–705. https://doi.org/10.3182/20130619-3-ru-3018.00600 seville, e. (2009). resilience: great concept... but what does it mean for organizations? tephra, july: 9–14. sheffi, y. (2007). building a resilient organization. the bridge national academy of engineering, 37(1): 30–36. sheffi, y., rice jr., j.b. (2005). a supply chain view of the resilient enterprise. mit sloan management review, 47(1): 41–48+94. somers, s. (2009). measuring resilience potential: an adaptive strategy for organizational crisis planning. journal of contingencies and crisis management, 17(1): 12–23. https://doi.org/10.1111/j.1468-5973.2009.00558.x starr, r., newfrock, j., delurey, m. (2003). enterprise resilience: managing risk in the networked economy. strategy and business, 30: 70–79. tillement, s., cholez, c., reverdy, t. (2009). assessing organizational resilience: an interactionist approach. management, 12(4): 230–264. https://doi.org/10.3917/mana.124.0230 us department of homeland security, resilience. available at: https://www.dhs.gov/topic/resilience [accessed july 12, 2017]. whitman, z.r. et al. (2013). short-form version of the benchmark resilience tool (brt-53). measuring business excellence, 17(3): 3–14. https://doi.org/10.1108/mbe-05-2012-0030 int. j. prod. manag. eng. (2017) 5(2), 107-116 creative commons attribution-noncommercial-noderivatives 4.0 international morales allende, m., ruiz-martin, c., lopez-paredes, a. & perez ríos, j. m 116 https://doi.org/10.1002/j.2334-5837.2010.tb01071.x https://doi.org/10.1146/annurev.es.04.110173.000245 https://doi.org/10.1146/annurev.es.04.110173.000245 https://doi.org/10.1111/1467-7717.00218 https://doi.org/10.1061/(asce)nh.1527-6996.0000075 https://doi.org/10.1016/j.hrmr.2010.07.001 https://doi.org/10.1002/bse.708 https://doi.org/10.1002/bse.708 https://doi.org/10.1111/j.0361-3666.2006.00331.x https://doi.org/10.1108/03684921011081150 https://doi.org/10.1108/03684921011081150 https://doi.org/10.1016/j.ress.2015.03.007 https://doi.org/10.1007/978-3-319-55889-9_5 https://doi.org/10.3182/20130619-3-ru-3018.00600 https://doi.org/10.1111/j.1468-5973.2009.00558.x https://doi.org/10.3917/mana.124.0230 https://www.dhs.gov/topic/resilience https://doi.org/10.1108/mbe-05-2012-0030 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2017.7013 received 2016-12-09 accepted: 2017-06-21 approach of the two-way influence between lean and green manufacturing and its connection to related organisational areas rodrigo salvadora1, casiano moro piekarskia2 and antonio carlos de franciscoa3 adept. of industrial engineering. federal university of technology of paraná, ponta grossa 84016-210, brazil a1 salvador.rodrigors@gmail.com, a2 piekarski@utfpr.edu.br, a3 acfrancisco@utfpr.edu.br abstract: initiatives toward lean and green manufacturing are given mainly due to organisational response to current market’s economic and environmental pressures. this paper, therefore, aims to present a brief discussion based on a literature review of the potential two-way influence between lean and green manufacturing and its role on the main organisational areas with a closer relationship to such approaches, which were observed to be more extensively discussed in the literature. naturally lean practises seem more likely to deploy into green outcomes, though the other way around can also occur. there is some blur on the factual integration of both themes, as some authors suggest. notwhithstanding, they certainly present certain synergy. thereupon, further research is needed to unveil the real ties, overlaps and gaps between these approaches. key words: lean & green, manufacturing, lean, green, sustainability. 1. introduction the lean & green approach has been considered in the search for environmental and economic enhancement of production systems (kainuma and tawara, 2006) and this topic has been gaining special interest since 2010 (garza-reyes, 2015). the economic characteristics of the market have been forcing organisations to tighten their belts and lean their processes in order to save resources and drop waste production. so then, lean manufacturing starting point was marked by the need for maximising the use of resources and minimising waste (sundar, balaji and satheeshkumar, 2014) and later, as it developed, it started to grow and ramify, deriving similar approaches on diverse spheres, such as green manufacturing. besides market pressure, governments also play a certain role in an organisation’s performance, regulating industrial activities (salvador et al., 2014), aiming to prevent them to thrive on either the nature’s or the population’s expenses, environmental and economy-wise. for those reasons, an integrated approach seems to be an easier and better option in the path to leaner and greener practises. therefore, this paper aims to present a brief discussion based on a literature review of the potential two-way influence between lean and green manufacturing and its role on the main organisational areas with a closer relationship to such approaches, which were observed to be more extensively discussed in the literature. such organisational areas were product planning and design, supply chain and quality management, organisational culture and performance and logistics. the novelty of this piece of research lies exactly on the identification of the main areas within an organisation where the connection between the to cite this article: salvador, r., piekarski, c.m., de francisco, a.c. (2017). approach of the two-way influence between lean and green manufacturing and its connection to related organisational areas. international journal of production management and engineering, 5(2), 73-83. https://doi.org/10.4995/raet.2017.7013 int. j. prod. manag. eng. (2017) 5(2), 73-83creative commons attribution-noncommercial-noderivatives 4.0 international 73 http://creativecommons.org/licenses/by-nc-nd/4.0/ discussed approaches can be observed more often and densely, though there might be other areas with a weaker relationship, which are not addressed. this paper, thus, is structured as follows: firstly, the concepts of lean manufacturing and green manufacturing are presented. in the sequence, the main considerations about their influences on each other are briefly discussed. then, their influence on each of the aforementioned organisational areas is pointed out, followed by some considerations from controversial literature. lastly, the final considerations with regard to this piece of research are drawn and the references used are listed. 2. lean manufacturing developed in japan (rohani and zahraee, 2015), this business strategy has been discussed since the 1950s. there is no one definite lean manufacturing definition, as it has been constantly developing (jabbour et al., 2013). pioneers (womack, jones and roos, 1990) consider it, though, as “making more with less”. it originated from total productive maintenance (tpm) practises, in the automobile industry (roosen and pons, 2013). the main objectives of lean practises are reduction of costs and defects (ohno, 1998) and customer focus (jabbour et al., 2013). its principles lie on minimising various forms of waste throughout the entire value chain, reducing variability (shah and ward, 2007). the 7 forms of waste considered by womack, jones and roos (1990) are: transport, inventory, motion, waiting, over-processing, overproduction and defects. it can be considered, thereafter, as a starting point to similar production approaches, such as the green manufacturing, as presented in the next section. moreover, the lean manufacturing presents five principles in order to achieve the reduction of those 7 forms of waste, being them: value, value stream, flow, pull and perfection (womack and jones, 1996). such principles aim to specify value, through the customers’ perspective, identify the value flow, make the value flow with no interruptions and allow costumers pull the value along the product’s value stream (halllam and contreras, 2016). 3. green manufacturing also referred to as sustainable manufacturing, green manufacturing has as a major concern the consequences of production activities and the entire product life cycle on the environment. even though it is hard to find a “one” definition for green manufacturing (paul, bhole and chaudhari, 2014) the concept has been widening (silva, silva and ometto, 2015) and the convergent idea is that it is the set of strategies, practises and behaviours, which aims to prevent, reduce and/or eliminate negative impacts on the environment. the main approaches found in the literature used to cope with environmental issues are control, prevention and product stewardship (hallam and contreras, 2016). the first approach considers “end-of-pipe” treatment to emissions of all sorts, concerning with the problem after it has been created (rusinko, 2007). the second one, prevention, focus on eliminating pollution/emissions (hart, 1995). product stewardship, though, involves other parties and stakeholders (such as internal [research and development, designers] and external [such as suppliers]) and their influence in the search for improvement of the environmental profile (rusinko, 2007). moreover, there are tools to assist the application of the concept and keep practises and activities on a green track. the main tool is the life cycle assessment (lca) (dües, tam and lin, 2013; piekarksi et al., 2013), which is considered the most complete and complex tool for environmental assessment (bocken et al., 2012), allowing to identify and manage the impacts of any considered system. 4. two-way influence: lean ↔ green the alongside adoption of green and lean manufacturing, besides appearing to present a positive effect on all dimensions of organisational performance (garza-reyes, 2015), can lead to an increase in market share and organisational profit (carvalho et al., 2011). greinacher et al. (2015) imply that from a lean approach, the reduction of inputs’ consumption int. j. prod. manag. eng. (2017) 5(2), 73-83 creative commons attribution-noncommercial-noderivatives 4.0 international salvador, r., piekarski, c.m. & de francisco, a.c. 74 http://creativecommons.org/licenses/by-nc-nd/4.0/ presents a logical continuation to reach “lean + green”, getting maximum efficiency and enhancing competitiveness. larson and greenwood (2004) consider lean manufacturing and green manufacturing parallel universes, whereas dües, tam and lin (2013) affirm that the adoption of lean practises works as a catalyst for better green practises. additionally, overlapping environmental and operational practises get leveraged (yang et al., 2010; carvalho and cruz-machado, 2009). galeazzo, furlan and vinelli (2014) corroborate the idea suggesting that, based on previous studies, lean and green manufacturing are complementary, as it can be observed in figure 1. thus, environmental practises would be able to help an organisation to become leaner and lean practises would help to greening the organisation (ng, low and song, 2015). womack and jones (1996), in turn, state that lean production is environmentally friendly by nature, since it leads to reduction of energy consumption, wastes and by-products, which, according to yang et al. (2010) might reflect on the organisational performance in a comprehensive way. furthermore, epa (2006) affirms that environmental wastes, somehow are related to or embedded in ohno’s seven wastes, hence, tying once again lean and green. hallam and contreras (2016) conclude that lean practises tend to lead to greener outcomes, whereas evidence that green pulls leaner outcumes is lacking, so what the latter does is to support existing lean practises. nevertheless, it can be observed that they work together (upadhye et al., 2010) and have a synergistic relationship (galeazzo et al., 2014; yang et al., 2011). product planning & design supply chain management quality management & performance organisational culture logistics focus: cost reduction and flexibility focus: cost reduction and flexibility customer: driven by costs, satisfied by cost and lead time reduction customer: driven by costs, satisfied by cost and lead time reduction waste: 7 wastes waste: 7 wastes product design: maximise performance, minimise cost product design: maximise performance, minimise cost practise: increase replenishment frequency practise: increase replenishment frequency manufacturing: high average utilisation, jit manufacturing: high average utilisation, jit end-of-life: no concern for impact of product user of product use or end-of-life recovery end-of-life: no concern for impact of product user of product use or end-of-life recovery kpi: costkpi: cost dominant cost: physical cost dominant cost: physical cost principal tool: lean value stream mapping principal tool: lean value stream mapping leanlean lean & green overlap lean & green overlap focus: sustainable development and ecological impact focus: sustainable development and ecological impact customer: conscience-driven, satisfied by helping them to be more environmentally friendly customer: conscience-driven, satisfied by helping them to be more environmentally friendly waste: inefficient use of resources, non-product output (scrap and emissions) waste: inefficient use of resources, non-product output (scrap and emissions) product design: life-cycle assessment product design: life-cycle assessmentpractise: reduce replenishment frequency practise: reduce replenishment frequency manufacturing: remanufacturing capabilities manufacturing: remanufacturing capabilities end-of-life: consideration of impact of product use and end-of-life recovery in form of re-use or recycling end-of-life: consideration of impact of product use and end-of-life recovery in form of re-use or recycling kpi: co2kpi: co2 dominant cost: cost for future generations dominant cost: cost for future generations principal tool: life-cycle assessment principal tool: life-cycle assessment greengreen focus: waste reduction focus: waste reduction waste reduction tecniques waste reduction tecniques people & organisation people & organisation lead time reduction lead time reduction supply chain relationship supply chain relationship kpi: service level kpi: service level tools/ practises tools/ practises figure 1. overlap of lean & green paradigms and its influence on related organisational areas (source: adapted from dües, tam and lin (2013)). int. j. prod. manag. eng. (2017) 5(2), 73-83creative commons attribution-noncommercial-noderivatives 4.0 international approach of the two-way influence between lean and green manufacturing and its connection to related organisational areas 75 http://creativecommons.org/licenses/by-nc-nd/4.0/ the two aforementioned and discussed approaches seem to overlap in various areas within the organization, since early steps such as planning and design, as in processes and products, as presented hereafter. figure 2 shows a set of pieces of research which approach the integration of lean and green in the areas approached in figure 1. the filling colours of the bubbles indicate which organisational area they relate to and the connecting arrows point to the focus of the work, having been identified four main focusses within the papers analysed. it can be seen that much attention has been given to the improvement of processes, which tends to lead to performance improvement, as well as to the development of new frameworks, which help to systematise and better understand the relations of lean and green within the organisation. furthermore, chart 1 shows the overall purpose of the papers presented in figure 2. the papers presented in chart 1 approach a few implications of the alongside adoption (by diverse meand) of lean and green. such papers, together with the information provided in figure 2, show potential research trends in organisational areas, which were mentioned earlier and are briefly addressed hereafter. 4.1. product planning and design many organisations, as easily perceived, adopt lean practises and worry about environmental issues of its processes only after certain period of activity. it might be due to lack of resources, knowledge, technical expertise and/or lack of interest. on this regard, simons and mason (2003) claim that such strategies should be adopted since the conception of the project of the business/product/system, thus, preventing potential environmental and/or economic damages. in this sense, kainuma and tawara (2006) affirm that in product planning and design the lean approach aims the reduction of costs and maximisation of performance, while green manufacturing, through the principles of life cycle assessment (lca) aims to prevent impacts in each of the product life cycle phases. hence, naturally, use of materials and diverse wastes are reduced, applying principles of lean manufacturing, evidencing, hereby, the interrelation between these two themes. anand and kodali (2008) yet, point that product development plays an important role in an organisation’s path to become lean & green. anand and kodali (2008) duarte and machado (2017) carvalho et al (2017) kainuma and tawara (2006) galeazzo, furlan and vinelli (2014) hajmohammad et al (2013) mittal et al (2017) kumar et al (2017) thani, govindan and thakkar (2016) fercoq, lamouri and carbone (2016) pampanelli, found and bernardes (2014) verrier et al (2014) león and calvoamodio (2017) salleh, kasolang and jaffar (2012) esmer, çetin and tuna (2010) process/performance improvement framework development decision making support partnership management product planning and design supply chain management quality management and performance organisational culture logistics legend figure. 2 main focus of research papers integrating lean and green and related organisational areas. int. j. prod. manag. eng. (2017) 5(2), 73-83 creative commons attribution-noncommercial-noderivatives 4.0 international salvador, r., piekarski, c.m. & de francisco, a.c. 76 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.2. supply chain management lean & green can also assist supply chain management. kainuma and tawara (2006) analysed different interaction models between the parties in the supply chain in order to identify the information flows, as showed in figure 3, and the roles of these parties in management strategies. the elements within the dotted line belong to a typical supply chain, whereas the figure 3 as a whole represent a lean & green supply chain management scheme. by such interactions, the chain encompasses a broader set of collaborators and in order to maintain its activities aligned with the pillars of both lean and green manufacturing there is an extension of the aspects considered to the management of that supply chain, facilitating integration and collaboration of different areas and sectors. logistics strategies are to be considered seeking minimum motion, lower emissions, and a more chart. 1 overall research purpose of papers addressing lean and green integration. reference overall research purpose anand and kodali (2008) development of a framework to reduce various forms of waste in new product development (npd) process duarte and machado (2017) development of a conceptual framework to asses the implementation of lean and green using some key-criterion to identify green and lean initiatives carvalho et al. (2017) propose a model to support decision making and to help identifying the best set of green and lean supply chain management practises to improve eco-efficiency kainuma and tawara (2006) extend the range of the supply chain to include re-use and recycling throughout the life cycle of products and services galeazzo, furlan and vinelli (2014) using a case study methodology, the piece of research aims to discern how lean and green interact and how they yield maximum synergy in improving both operational and environmental performance hajmohammad et al. (2013) proposition of a conceptual model to help understand the roles of lean and supply management in regards to improving the firm’s environmental performance. the model proposes that the magnitude of environmental practices mediates the relationship between lean and supply management with environmental performance mittal et al. (2017) identification and ranking of enablers of lean and green in the adoption of leen-greenagile manufacturing systems kumar et al. (2017) literature review seeking to investigate the green impacts of green on lean, six sigma and lean six sigma. discussing the issues of energy management, global warming, pollution and usage of resources thani, govindan and thakkar (2016) applying ahp method, it seeks to investigate the impact of select lean and green practices on performance benefits, and to evaluate the influence of lean and green paradigms on overall performance of smes fercoq, lamouri and carbone (2016) quantitative study of lean/green integration focusses on waste reduction techniques in manufacturing processes. the study claims that the 3r hierarchy must be preferred to a deadly waste approach on defining a waste minimisation program pampanelli, found and bernardes (2014) propose a model to integrate environmental sustainability into pure lean thinking, to improve mass and energy flows in manufacturing environments that already possess the necessary deployment level to apply lean thinking verrier et al. (2014) assess how lean and green actions could be enhanced when used together, proposing a framework for lean and green management, including lean indicators, green performance indicators and green intentions indicators león and calvo-amodio (2017) identify the building blocks of lean for sustainability (environmental, social and economic spheres) salleh, kasolang and jaffar (2012) integrate environmental information management practices are in a lean total quality management (tqm) framework in order to establish a green lean tqm system esmer, çetin and tuna (2010) analyse a turkish container termimal performance both in terms of lean and green dimensions int. j. prod. manag. eng. (2017) 5(2), 73-83creative commons attribution-noncommercial-noderivatives 4.0 international approach of the two-way influence between lean and green manufacturing and its connection to related organisational areas 77 http://creativecommons.org/licenses/by-nc-nd/4.0/ responsive system (dües, tam and lin, 2013), as well as reduced costs and lead-time (esmer, çetin and tuna, 2010). additionally, as collaboration is of utmost importance within a supply chain (dües, tam and lin, 2013), narrow, highly reliable and responsive relationships are necessary in order for the production activities to be conducted in time, with the required resources and within the required environmental-wise behaviour. 4.3. quality management and organisational performance the lean manufacturing supports the total quality management (tqm), whereas green manufacturing practises support the environmental management system as a whole (salleh, kasolang and jaffar, 2012). regarding process’ characteristics, reduction of variability will be supported by the rigor on the use of materials (chugani et al., 2017), there not being wastes of such, for instance energy and time for conducting low value-added activities, aside from a higher accuracy on the production activities, preventing eventual rework (pampanelli, found and bernardes, 2014). with regard to environmental characteristics, change begins with the reduction of wastes, through the lean practises, being supported by a more rigorous selection of partnerships within the supply chain, requiring partners to bear the same concerns in mind and to prove it by obtaining the pertinent environmental certifications, hence, extending the environmental commitment. with this regard, hallam and contreras (2016), claiming that there are no models to actually integrate these themes, proposed one model with such purpose, which is showed in figure 4. the authors do not address the mathematical formulation of the elements’ relationships, though. as it can be seen in figure 4, green actions can avail themselves of lean ones assisting on the achievement of performance objectives. it all reflects on the organisational performance and will persuade clients’ and partners’ standpoint toward the organisation. as galeazzo, furlan and vinelli (2014) state, there can be both simultaneous and sequential benefits from the alongside implementation/practise of lean and green manufacturing, which should, preferably, be applied simultaneously. figure. 3 the range of the lean & green supply chain (source: kainuma and tawara (2006)). int. j. prod. manag. eng. (2017) 5(2), 73-83 creative commons attribution-noncommercial-noderivatives 4.0 international salvador, r., piekarski, c.m. & de francisco, a.c. 78 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.4. organisational culture and performance as dües, tam and lin (2013) imply, the integration between green and lean manufacturing to the organisation’s culture can bring about change to its core, adapting mission, vision and values, even revolutionising production practises, depending on the gap between the current and the desired (future) states. moreover, seeking cleaner products, the importance of this relation becomes evident, once the characteristics of the activities performed by suppliers have direct influence on the products and services offered by the core organisation (hajmohammad et al., 2013). hence, there can often be imposition of requisites and environmental certifications, mainly by large corporate networks, to their suppliers and partners, thus guaranteeing continuity of the commitment put on to their activities and products. 4.5. logistics to verrier et al. (2014), environment and logistics are antagonic. a proper management of the environmental characteristics of logistics activities is, as well as a legal requirement, a users requirement (esmer, çetin and tuna, 2010); besides, more than transportation and distribution channels, many places of congruent negotiations of loads (such as ports, distribution centres) are also business centres and value-adding spots. in what regards the characteristic emissions of the activities and cost management, such places need to deploy lean and green manufacturing strategies (esmer, çetin and tuna, 2010). furthermore, jointly, leand and green manufacturing in logistics operations may result in minimum motion systems, so as to achieve lower emission levels, as well as reducing costs, besides bringing down waiting time and creating more responsive systems (dües, tam and lin, 2013). it is possible to drop the emissions levels by giving preference, where possible, to electricity driven equipment, replacing combustion motors, whereas simulation methods, from process modelling, can guide reducions of operations’ time and costs (esmer, çetin and tuna, 2010). moreover, as showed by piekarski et al. (2017), replacing suppliers, choosing the ones located closer to the industrial facility may reduce greenhouse gas figure 4. proposed management model integrating lean and green with firm performance (source: hallam & contreras (2016)). int. j. prod. manag. eng. (2017) 5(2), 73-83creative commons attribution-noncommercial-noderivatives 4.0 international approach of the two-way influence between lean and green manufacturing and its connection to related organisational areas 79 http://creativecommons.org/licenses/by-nc-nd/4.0/ emissions, on the green side, and at the same time, contribute to a leaner system, diminishing delivery time and cost. showing, this, a case of green pulling lean, opposite to what most of current literature claims. furthermore, cabrita et al. (2016) affirm that logistics is one of the key areas to seek collaboration in order to green a system. 5. controversial literature even though lean and green manufacturing hold certain harmony, they differ in a few aspects, such as: main focus, process structure, construct value, performance measurements, type of consumers, definition of waste e techniques utilized (johansson and sundin, 2014). under this light, kleindorfer et al. (2005) state that lean and green manufacturing practises are distinct and, therefore, they have different impacts on organisational performance. evidences of integrated approaches using lean and green can be extensively found in the literature. the links and the evidences proposed to successfully integrating them, however, are suggested by hallm and contreras (2016) to be largely weak. many authors corroborate the idea that lean and green manufacturing are approaches that go alongside and have positive influence on each other. johansson and sundin (2014), however, affirm not to be possible to say that actions toward one directly lead to consequences on the other, being this area yet incipient (garza-reyes, 2015). also, potential relations might have been neglected, being this investigation inconclusive (jabbour et al., 2013). one example of a contradictory outcome out of the approached relationship is that lean practises may impact negatively on green ones by the use of a just-in-time delivery process (martínez-jurado and moyano-fuentes, 2014), due to the enlargement of the emissions of greenhouse gases in transportation. on the one hand, according to hallam and contreras (2016), the majority of studies found in the literature corroborate a strong link and positive outcomes from the integration of both themes. on the other hand, the authors state that the area still lacks operating models of the firms to relate them. with few attempts to, actually, integrate them both and that, according to the results of biggs (2009), both themes have developed independently. for johansson and sundin (2014), one of the main overlaps between lean and green manufacturing is that they are both driven by the increase in competitiveness and, even though their constructs are different, these are not conflicting. 6. final considerations it is possible to observe that the adoption and the joint use of lean and green manufacturing can result in benefits and influences on various areas within an organisation, such influences might occur simultaneous or sequentially. activities in a certain area can influence on another and, likewise, they can be potentialized, so then multiple single actions can result in a non-linear increase on the organisational performance (1+1 ≠ 2). the organisational areas in the overlap region of figure 1, product planning and design, supply chain management, quality management and performance, organisational culture and logistics, are the ones where current research is being conducted on the influence of lean and green on each other. so it can be noted that they have been being studied because of the perceived value of the joint use of the mentioned approaches on such areas. therefore, future research lines should be drawn from those areas/fields to further explore and better assess the practical implications of actions and decisions addressing lean and green approaches on an organisation’s outcomes. notwithstanding, this piece of research is not exempt from limitations. this review is not, by any means, exhaustive and it was limited to the particular topics approached. the authors believe, however, that it allowed to identify representative trends of the growing body of literature on lean and green. lean and green efforts seem to be made independently on an organisation. it does not seem to be common agreement that they surely are or can be integrated, since their constructs might differ, however, it appears to be undeniable that they have certain synergy and have sharing features. nevertheless, further research is needed to broadly investigate their relationship, unveil the real ties, overlaps and gaps between them. int. j. prod. manag. eng. (2017) 5(2), 73-83 creative commons attribution-noncommercial-noderivatives 4.0 international salvador, r., piekarski, c.m. & de francisco, a.c. 80 http://creativecommons.org/licenses/by-nc-nd/4.0/ acknowledgements the authors would like to express their gratitude for the support provided by the federal university of technology – paraná. also, the authors would like to thank the reviewers for their thoughtful and helpful comments on the earlier version of this manuscript. references anand, g., kodali, r. (2008). development of a conceptual framework for lean new product development process. international journal of product development, 6(2):190-224. https://doi.org/10.1504/ijpd.2008.019240 biggs, c. (2009). exploration of the integration of lean and environmental improvement. phd thesis. cranfield university, uk bocken, n.m.p., alwood, j.m., willey, a.r. et al. (2012). development of a tool rapidly assessing the implementation difficulty and emissions benefits of innovations. technovation, 32:19-31. https://doi.org/10.1016/j.technovation.2011.09.005 carvalho, h., azavedo, s., cruz-machado, v. (2014). trade-offs among lean, agile, resilient and green paradigms in supply chain management: a case study approach. in: proceedings of the seventh international conference on management science and engineering management lecture notes in electrical engineering, 242: 953-968. https://doi.org/10.1007/978-3-642-40081-0_81 carvalho, h., azevedo, s.g., cruz-machado, v. (2010). supply chain performance management: lean and green paradigms. int. j. bus. perform. supply chain model, 2(3/4): 304-333. https://doi.org/10.1504/ijbpscm.2010.036204 carvalho, h., duarte, s., cruz-machado, v. (2011). lean, agile, resilient and green: divergences and synergies. int. j. lean six sigma, 2(2): 151-179. https://doi.org/10.1108/20401461111135037 carvalho, h., govinddan, k., azevedo, s.g., cruz-machado, v. (2017). modelling green and lean supply chains: an eco-efficiency perspective. resources, conservation and recycling, 120: 75-87. https://doi.org/10.1016/j.resconrec.2016.09.025 chugani, n., kumar, v., garza-reyes, j.a., rocha-lona, l., upadhyay, a. (2017). investigating the green impact of lean, six sigma and lean six sigma: a systematic literature review. international journal of lean six sigma, 8(1):7-32. https://doi.org/10.1108/ ijlss-11-2015-0043 diaz-elsayed, n., jondral, a., greinacher, s. et  al. (2013). assessment of lean and green strategies by simulation of manufacturing systems in discrete production environments. cirp annals manufacturing technology, 62: 475–478. https://doi.org/10.1016/j. cirp.2013.03.066 duarte, s., machado, v.c. (2017). green and lean implementation: an assessment in the automotive industry. international journal of lean six sigma, 8(1): 65-88. https://doi.org/10.1108/ijlss-11-2015-0041 dües, c.m., tan, k.m., lim, m. (2013). green as the new lean: how to use lean practices as a catalyst to greening your supply chain. journal of cleaner production, 40: 93-100. https://doi.org/10.1016/j.jclepro.2011.12.023 esmer, s., çetin, i.b., tuna, o. (2010). a simulation for optimum terminal truck number in a turkish port based on lean and green concept. the asian journal of shipping and logistics, 26(2): 277-296. https://doi.org/10.1016/s2092-5212(10)80006-9 environmental protection agency (epa) (2006). lean and environment training module 2 – lean and environment toolkit. environmental protection agency, washington, dc. espadinha-cruz, p., grilo, a., puga-leal, r. et al. (2011). a model for evaluating lean, agile, resilient and green practices interoperability in supply chains. in: proceedings of the ieee international conference on industrial engineering and engineering management (ieem), singapore, 1: 1209-1213. https://doi.org/10.1109/ieem.2011.6118107 fercoq, a., lamouri, s., carbone, v. (2016). lean/green integration focused on waste reduction techniques. journal of cleaner production, 137: 567-578. https://doi.org/10.1016/j.jclepro.2016.07.107 folinas, d., aidonis, d., triantafillou, d., malindretos, g. (2013) exploring the greening of the food supply chain with lean thinking techniques. procedia technology, 8: 416-424. https://doi.org/10.1016/j.protcy.2013.11.054 galeazzo, a., furlan, a., vinelli, a. (2014). lean and green in action: interdependencies and performance of pollution prevention projects. journal of cleaner production, 85: 191-200. https://doi.org/10.1016/j.jclepro.2013.10.015 garza-reyes, j.a. (2015). green lean and the need for six sigma. int. j. lean six sigma, 6(3): 226-248. https://doi.org/10.1108/ ijlss-04-2014-0010 greinacher, s., moser, s., hermann, h., lanza, g. (2015). simulation based assessment of lean and green strategies in manufacturing systems. procedia cirp, 29: 86-91. https://doi.org/10.1016/j.procir.2015.02.053 hajmohammad, s., vachon, s., klassen, r., gavronski, i. (2013). lean management and supply management: their role in green practices and performance. journal of cleaner production, 39: 312-320. https://doi.org/10.1016/j.jclepro.2012.07.028 int. j. prod. manag. eng. (2017) 5(2), 73-83creative commons attribution-noncommercial-noderivatives 4.0 international approach of the two-way influence between lean and green manufacturing and its connection to related organisational areas 81 https://doi.org/10.1504/ijpd.2008.019240 https://doi.org/10.1016/j.technovation.2011.09.005 https://doi.org/10.1007/978-3-642-40081-0_81 https://doi.org/10.1504/ijbpscm.2010.036204 https://doi.org/10.1108/20401461111135037 https://doi.org/10.1016/j.resconrec.2016.09.025 https://doi.org/10.1108/ijlss-11-2015-0043 https://doi.org/10.1108/ijlss-11-2015-0043 https://doi.org/10.1016/j.cirp.2013.03.066 https://doi.org/10.1016/j.cirp.2013.03.066 https://doi.org/10.1108/ijlss-11-2015-0041 https://doi.org/10.1016/j.jclepro.2011.12.023 https://doi.org/10.1016/s2092-5212(10)80006-9 https://doi.org/10.1109/ieem.2011.6118107 https://doi.org/10.1016/j.jclepro.2016.07.107 https://doi.org/10.1016/j.protcy.2013.11.054 https://doi.org/10.1016/j.jclepro.2013.10.015 https://doi.org/10.1108/ijlss-04-2014-0010 https://doi.org/10.1108/ijlss-04-2014-0010 https://doi.org/10.1016/j.procir.2015.02.053 https://doi.org/10.1016/j.jclepro.2012.07.028 http://creativecommons.org/licenses/by-nc-nd/4.0/ hallam, c., contreras, c. (2016). integrating lean and green management. management decision, 54(9): 2157-2187. https://doi. org/10.1108/md-04-2016-0259 hart, s.l. (1995). a natural-resource-based view of the firm. academy of management review, 20(4): 986-1014. jabbour, a.b.l.s., jabbour, c.j.c., freitas, w.r.s. et al. (2013). lean and green? empirical evidence from the brazilian automotive industry. gestão da produção, 20(3):653-665. https://doi.org/10.1590/s0104-530x2013000300011 johansson, g., sundin, e. (2014). lean and green product development: two sides of the same coin? journal of cleaner production, 85: 104-121. https://doi.org/10.1016/j.jclepro.2014.04.005 kainuma, y., tawara, n. (2006). a multiple attribute utility theory approach to lean and green supply chain management. international journal of production economics, 101: 99-108. https://doi.org/10.1016/j.ijpe.2005.05.010 kleindorfer, p.r., singhal, k., van wassenhove, l.n. (2005). sustainable operations management. prod. oper. manag., 14(4): 482-492. https://doi.org/10.1111/j.1937-5956.2005.tb00235.x kurdve, m., zackrisson, m., wiktorsson, m. et al. (2014). lean and green integration into production system models experiences from swedish industry. journal of cleaner production, 85: 180-190. https://doi.org/10.1016/j.jclepro.2014.04.013 león, h.c.m., calvo-amodio, j. (2017). towards lean for sustainability: understanding the interrelationships between lean and sustainability from a systems thinking perspective. journal of cleaner production, 142: 4384-4402. https://doi.org/10.1016/j.jclepro.2016.11.132 martínez-jurado, p.j., moyano-fuentes, j. (2014) lean management, supply chain management and sustainability: a literature review. journal of cleaner production, 85: 134-150. https://doi.org/10.1016/j.jclepro.2013.09.042 mittal, v.k., sindhwani, r., kalsariya, v., salroo, f., sangwan, k.s., singh, p.l. (2017). adoption of integrated lean-green-agile strategies for modern manufacturing systems. procedia cirp, 61: 463-468. https://doi.org/10.1016/j.procir.2016.11.189 ng, r., low, j.s.c., song, b. (2015). integrating and implementing lean and green practices based on proposition of carbon-value efficiency metric. journal of cleaner production, 95: 242-255. https://doi.org/10.1016/j.jclepro.2015.02.043 ohno, t. (1988). toyota production system: beyond large scale production. cambrigde: productivity press pampanelli, a.b., found, p., bernardes, a.m. (2014). a lean & green model for a production cell. journal of cleaner production, 85: 19-30. https://doi.org/10.1016/j.jclepro.2013.06.014 paul, i.d., bhole, g.p., chaudhari, j.r. (2014). a review on green manufacturing: it’s important, methodology and its application. procedia materials science, 6: 1644-1649. https://doi.org/10.1016/j.mspro.2014.07.149 piekarski, c.m., francisco, a.c., luz, l.m., kovaleski, j.l., silva, d.a.l. (2017). life cycle assessment of medium-density fiberboard (mdf) manufacturing process in brazil. science of the total environment, 575: 103-111. https://doi.org/10.1016/j.scitotenv.2016.10.007 piekarski, c.m., luz, l.m., zocche, l., francisco, a.c. (2013). life cycle assessment as entrepreneurial tool for business management and green innovations. journal of technology management and innovation, 8(1): 44:53. https://doi.org/10.4067/s071827242013000100005 rohani, j.m., zahraee, s.m. (2015). production line analysis via value stream mapping: a lean manufacturing process of color industry. procedia manufacturing, 2: 6-10. https://doi.org/10.1016/j.promfg.2015.07.002 roosen, t., pons, d. (2013). environmentally lean production: the development and incorporation of an environmental impact index into value stream mapping. journal of industrial engineering, 2013: 1-17. https://doi.org/10.1155/2013/298103 rusinko, c.a. (2007). green manufacturing: an evaluation of environmentally sustainable manufacturing practices and their impact on competitive outcomes. ieee transactions on engineering management, 54(3): 445-454. https://doi.org/10.1109/tem.2007.900806 salleh, n.a.m., kasolang, s., jaffar, a. (2012). green lean total quality information management in malaysian automotive companies. procedia engineering, 41: 1708-1713. https://doi.org/10.1016/j.proeng.2012.07.372 silva, d.a.l., silva, e.j., ometto, a.r. (2015). green manufacturing: uma análise da produção científica e de tendências para o future. production, 26(3). https://doi.org/10.1590/0103-6513.032513 simons, d., mason, r. (2003). lean and green: doing more with less. ecr journal, 3(1): 84-91. salvador, r., francisco, a.c., piekarski, c.m. et al. (2014). life cycle assessment (lca) as a tool for business strategy. independent journal of management & production, 5(3): 733-751. https://doi.org/10.14807/ijmp.v5i3.186 shah, r., ward, p.t. (2003). lean manufacturing: context, practice bundles, and performance. journal of operations management, 21(2): 129-149. https://doi.org/10.1016/s0272-6963(02)00108-0 sundar, r., balaji, a.n., satheeshkumar, r.m. (2014). a review on lean manufacturing implementation techniques. procedia engineering, 97: 1875-1885. https://doi.org/10.1016/j.proeng.2014.12.341 upadhye, n., deshmukh, s.g., garg, s. (2010). lean manufacturing for sustainable development. global business and management research: an international journal, 2(1): 125. verrier, b., rose, b., caillaud, e. et al. (2014). combining organizational performance with sustainable development issues: the lean and green project benchmarking repository. journal of cleaner production, 85: 83-93. https://doi.org/10.1016/j.jclepro.2013.12.023 womack, j.p., jones, d.t. (1996). lean thinking banish waste and create wealth in your corporation, new york: free press. int. j. prod. manag. eng. (2017) 5(2), 73-83 creative commons attribution-noncommercial-noderivatives 4.0 international salvador, r., piekarski, c.m. & de francisco, a.c. 82 https://doi.org/10.1590/s0104-530x2013000300011 https://doi.org/10.1016/j.jclepro.2014.04.005 https://doi.org/10.1016/j.ijpe.2005.05.010 https://doi.org/10.1111/j.1937-5956.2005.tb00235.x https://doi.org/10.1016/j.jclepro.2014.04.013 https://doi.org/10.1016/j.jclepro.2016.11.132 https://doi.org/10.1016/j.jclepro.2013.09.042 https://doi.org/10.1016/j.procir.2016.11.189 https://doi.org/10.1016/j.jclepro.2015.02.043 https://doi.org/10.1016/j.jclepro.2013.06.014 https://doi.org/10.1016/j.mspro.2014.07.149 https://doi.org/10.1016/j.scitotenv.2016.10.007 https://doi.org/10.4067/s0718-27242013000100005 https://doi.org/10.4067/s0718-27242013000100005 https://doi.org/10.1016/j.promfg.2015.07.002 https://doi.org/10.1155/2013/298103 https://doi.org/10.1109/tem.2007.900806 https://doi.org/10.1016/j.proeng.2012.07.372 https://doi.org/10.1590/0103-6513.032513 https://doi.org/10.14807/ijmp.v5i3.186 https://doi.org/10.1016/s0272-6963(02)00108-0 https://doi.org/10.1016/j.proeng.2014.12.341 https://doi.org/10.1016/j.jclepro.2013.12.023 http://creativecommons.org/licenses/by-nc-nd/4.0/ womack, j.p., jones, d.t., roos, d. (1990). the machine that changed the world. new york: rawson associates, 1990. yang, c., lin, s., chan, y. et al. (2010). mediated effect of environmental management on manufacturing competitiveness: an empirical study. international journal of production economics, 123: 210-220. https://doi.org/10.1016/j.ijpe.2009.08.017 yang, m.g.m., hong, p., modi, s.b. (2011). impact of lean manufacturing and environmental management on business performance: an empirical study of manufacturing firms. international journal of production economics, 129(2): 251-261. https://doi.org/10.1016/j. ijpe.2010.10.017 int. j. prod. manag. eng. (2017) 5(2), 73-83creative commons attribution-noncommercial-noderivatives 4.0 international approach of the two-way influence between lean and green manufacturing and its connection to related organisational areas 83 https://doi.org/10.1016/j.ijpe.2009.08.017 https://doi.org/10.1016/j.ijpe.2010.10.017 https://doi.org/10.1016/j.ijpe.2010.10.017 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2017.5916 received 2016-06-11 accepted: 2017-06-09 a column generation for the heterogeneous fixed fleet open vehicle routing problem majid yousefikhoshbakht a and azam dolatnejad b* adepartment of mathematics, faculty of science, bu-ali sina university, hamedan, iran. byoung researchers & elite club, tehran north branch, islamic azad university, tehran, iran. * a_dolatnejad@aut.ac.ir abstract: this paper addressed the heterogeneous fixed fleet open vehicle routing problem (hffovrp), in which the vehicles are not required to return to the depot after completing a service. in this new problem, the demands of customers are fulfilled by a heterogeneous fixed fleet of vehicles having various capacities, fixed costs and variable costs. this problem is an important variant of the open vehicle routing problem (ovrp) and can cover more practical situations in transportation and logistics. since this problem belongs to np-hard problems, an approach based on column generation (cg) is applied to solve the hffovrp. a tight integer programming model is presented and the linear programming relaxation of which is solved by the cg technique. since there have been no existing benchmarks, this study generated 19 test problems and the results of the proposed cg algorithm is compared to the results of exact algorithm. computational experience confirms that the proposed algorithm can provide better solutions within a comparatively shorter period of time. key words: open vehicle routing problem, heterogeneous fixed fleet, np-hard problems, column generation. 1. introduction while the travelling salesman problem (tsp) is perhaps the most famous single-vehicle problem, the vehicle routing problem (vrp) is an important variant of tsp which has many applications in industrial and service firms (yousefikhoshbakht and khorram, 2012). this problem plays a vital role in supply chains where in the first transportation step to collect agricultural products, for instance or in the final distribution phase to deliver goods to customers. the vrp was first defined by dantzig and ramser more than 50 years ago (1959) and can be defined on an undirected complete graph with one depot (node 0) and n customers indexed from 1 to n (if the graph is not complete, we can instead lack of each arc with the arc that has infinite size) (saadati eskandari and yousefikhoshbakht, 2012). a fleet of same vehicles is based at the depot in which each vehicle has a fixed capacity and perhaps a route-length restriction which limits the maximum distance it can travel. furthermore, each customer has a known demand and a defined distance is associated with each edge. this problem involves routing a fleet of vehicles that start to move simultaneously from the depot and come back to the depot after visiting customers. in other words, each route in the vrp is a hamiltonian cycle over the subset of customers visited on the route. the objective is to design a set of minimum cost routes to serve all customers so that: the load on a vehicle is below vehicle capacity at each point on the route. to cite this article: yousefikhoshbakht, m., dolatnejad, a. (2017). a column generation for the heterogeneous fixed fleet open vehicle routing problem. international journal of production management and engineering, 5(2), 55-71. https://doi.org/10.4995/raet.2017.5916 int. j. prod. manag. eng. (2017) 5(2), 55-71creative commons attribution-noncommercial-noderivatives 4.0 international 55 http://creativecommons.org/licenses/by-nc-nd/4.0/ each customer is serviced by only one visit of a single vehicle, i.e., split deliveries are not allowed. the minimum number of vehicles is also required to service all customers. in other words, the number of trips or vehicles used is not imposed, it is a decision variable. the vrp is not very realistic. to make vrp models more realistic and applicable, there are many varieties of the vrp obtained by adding constraints to the basic model. for example, several variations and specializations of the vrp are the load along each route must not exceed the service of the customers and must occur within given time windows (vehicle routing problem with time windows, vrptw) (wang and chen, 2012), the customer has pick-up and delivery demand (vehicle routing problem with pickup and delivery, vrppd) (çatay, 2010), the customer demands may not be completely known in advance (stochastic vehicle routing problem, svrp) (lei, laporte and guo, 2011), the service of a customer may be split among different vehicles (split delivery vehicle routing problem, sdvrp) (aleman and hill, 2010), precedence relations may exist between the customers (vehicle routing problem with backhauls, vrpb) (anbuudayasankar et al., 2012), and the demands or the travel times may vary dynamically (dynamic vehicle routing problem, dvrp) (gendreau et al., 2006). another version of the vrp is heterogeneous fleet vehicle routing problem (hfvrp) in which the fleet may contain heterogeneous vehicles like most of the companies in the real-life context (li, tian and aneja, 2010). in this problem, the fleet is composed of various vehicle types. each type k is defined by a capacity qk, a fixed cost fk and a cost per distance unit αk often called variable cost. it should be noted that a cycle of length l done by a vehicle of type k has a cost fk+ αk·l. as in the vrp, each customer must be visited by one vehicle only, each vehicle must start and finish its travel at the depot and the capacity of a vehicle and maybe the maximum length of a route must not be violated. the objective of the hfvrp is to compute a set of cycles and to assign vehicles to cycle to minimize total cost which includes both the vehicle variable and fixed costs. the idea is not only to consider the routing of the vehicles, but also the composition of the vehicle fleet (gendreau et al., 1999). using a heterogeneous fleet of vehicles has several advantages such as the scheduler can revise the fleet composition to better suit customer needs and vehicles of different carrying capacities give the flexibility to assign capacity according to the customer’s varying demand by deploying the suitable vehicle types to areas with the analogous concentration of customers. it is also possible to facility customers requiring small vehicles because of accessibility restrictions in urban areas, environmental concerns or physical restrictions on the vehicle size and weight. if we assume that the number of vehicles of each type is unlimited, we get a type of problem known as the fleet size and mix vehicle routing problem (fsmvrp) (brandão, 2009). furthermore, the heterogeneous fixed fleet vehicle routing problem (hffvrp) is a variant of the fsmvrp in which there is a limited number of homogeneous and heterogeneous vehicles respectively. in other words, in contrast to the fsmvrp, the number of vehicles of each type is limited. although the fsmvrp and hffvrp are very similar, these two types of problems are used in rather different situations. the fsmvrp is more appropriate for tactical decisions (selecting the acquired number of vehicles) and operational ones (computing the routes and the vehicles assigned to them). in these strategic decisions, a company wants to buy a vehicle fleet and needs to define its size and composition, but the hffvrp represents better the operational decisions of defining the vehicles that should be used in order to serve the customers among those available. the hffvrp intensifies from a practical perspective when the vehicle fleet is hired, that is, vehicles do not constitute company assets. in such cases, effective planning is a critical success factor for the operational efficiency and the resulting service level, since non-company resources are responsible for the physical interface with the final customer. the open vehicle routing operational framework is faced by a company which either does not own a vehicle fleet at all, or its fleet is inappropriate or inadequate to satisfy the demand of its customers. thus, the company has to contract all or part of its distribution activities to external carriers. these contractors have their own vehicles, they pay their own vehicle costs (e.g. capital, operating, maintenance and depreciation), and they usually consider a compensation model based on mileage. whenever the company does not need the contractor or the vehicle back at the depot, the paths followed by the vehicles must not include the vehicle trip after the last delivery (i.e. the return trip to the depot) that will add extra mileage to the compensation model. this problem, called heterogeneous fixed fleet int. j. prod. manag. eng. (2017) 5(2), 55-71 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakht, m. and dolatnejad, a. 56 http://creativecommons.org/licenses/by-nc-nd/4.0/ open vehicle routing problem (hffovrp) (li, leung and tian, 2012). this reference addressed this problem for the first one. besides, a multistart adaptive memory programming metaheuristic with modified tabu search algorithm was proposed to solve this new version of vrp. finally, the efficiency and effectiveness of the proposed algorithm are experimentally evaluated on a set of generated instances. figure 1 presents a feasible solution for the hffovrp. the hffovrp compared to hffvrp has a unique character in that the vehicles are not required to come back to the depot after completing a service. the hffovrp is utilized in practice in delivering packages and newspapers to homes. in this important variant of the ovrp, contractors who are not employees of the delivery company use their own vehicles and do not return to the depot. furthermore, companies which use contractors to deliver newspapers to residential customers do not require the contractors and their vehicles to return to the depot. as a result, interest of researchers in the ovrp and its variations has increased dramatically and a wide variety of new algorithms have been developed over the last ten years to solve the problem. figure 1. feasible solution for the hffovrp. however, to the best of our knowledge, a first done work to address hffovrp, which is a relaxation of the standard vrp, was the paper of li, leung and tian (2012). this problem is more practical in distribution management and in which the fleet consists of vehicles with different attributes and the number of available vehicles is fixed like hffvrp. in contrast to hffvrp, the vehicles do not return to the starting point after completing the service of the last customers. as we know, most works on ovrp has aimed at minimizing the number of vehicles, but this objective was not considered in the hffovrp proposed in this work. to solve the problem, a multistart adaptive memory programming metaheuristic (mamp) was proposed as a new memory-based algorithm. then, the parameters tuning is conducted systematically by an automatic procedure and the advantage of considering a multi-start strategy in the mamp is verified with the obtained best configuration. finally, the results are reported on the generated instances. after that, penna et al. presented a heuristic algorithm based on the iterated local search metaheuristic and on the randomized variable neighborhood descent (subramanian et al., 2010) for solving the hffovrp. this work is an extension of the one proposed by themselves for the hfvrp (penna, subramanian and ochi, 2013). this paper dealt with the ovrp and the hffovr which often arises in distribution management and transportation. they solved both variants by a multi-start algorithm based on the iterated local search metaheuristic. the proposed algorithm uses a variable neighborhood descent procedure, with random neighborhood, ordering in the local search phase. yousefikhoshbakht et al. (2014) proposed an efficient adaptive memory based algorithm equipped with diversification and intensification to solve the hffovrp. this algorithm called brmts is a bone route algorithm combined with a modified tabu search (brmts) which is effective for solving hffovrp problems in a reasonable computing time. furthermore, the proposed algorithm directly produces a new solution from a component of the other solutions while using new diversification and intensification mechanisms. the brmts employs the generalized route for constructive algorithm which was first presented in (tarantilis and kiranoudis, 2014) for generating initial diversified solutions and modified ts improvement procedure. computational results on some test-generated instances demonstrate that the proposed brmts finds high quality solutions within a reasonable time. besides, these authors proposed a column generation technique and an elite ant system for the hffovrp. this efficient hybrid heuristic used versions algorithms, including sweep, insert, swap, and 2-opt moves in order to solve generated hamiltonian paths with some modifications to generate feasible columns efficiently. these modifications lead to improve both the performance of the algorithm and the quality of the solutions (yousefikhoshbakht et al., 2016). the results show that the hybrid algorithm explores int. j. prod. manag. eng. (2017) 5(2), 55-71creative commons attribution-noncommercial-noderivatives 4.0 international a column generation for the heterogeneous fixed fleet open vehicle routing problem 57 http://creativecommons.org/licenses/by-nc-nd/4.0/ different parts of the solution space and try to be not trapped at the local optimum points. the hffovrp is an np-hard combinatorial problem, since it reduces to the ovrp if the number of types of vehicles is just one, i.e., the vehicle fleet is homogeneous, and the number of vehicles is unlimited. from the point of view of graph theory, the difference between the ovrp and the vrp is that a solution of the former consists of a set of hamiltonian paths rather than hamiltonian cycles. the problem of finding the best hamiltonian path for each set of customers assigned to a vehicle is np-hard (syslo, deo and kowalik, 1983). hence hffovrp is also np-hard. therefore, most of the practical examples of this problem cannot be solved by exact algorithms to optimality within reasonable time. furthermore, because there is no known polynomial algorithm that will find the optimal solution in every instance, the use of heuristics and metaheuristic are considered as a reasonable approach in finding solutions. in this paper, we have proposed a heuristic column generation algorithm (cg) in order to improve both the performance of the algorithm and the quality of the solutions of the exact algorithm. the algorithm explores different parts of the solution space, and the search method is not trapped at the local optimum. the experimental results have shown that the proposed algorithm is to be very efficient and competitive in terms of solution quality compared to exact algorithm. the structure of the remainder of the paper is as follows. in section 2, a proposed mixed integer model is described and in section 3, the proposed idea based on column generation is explained in great detail. in section 4, the proposed algorithm is compared with an exact algorithm on generated or standard problems belonging to hffovrp library. finally, some concluding remarks are given in the section 5. 2. problem description and formulation from a graph theoretical point of view, we can define the hffovrp as follows. let g = (v, e) be an undirected connected graph with v={0,1,…,n} as the set of vertexes and the set of arcs ( , ): ,e i j i j n0# #= " , (if the graph is not complete, we can instead lack of each arc with the arc that has infinite size). node 0 is the depot and the customer set c consists of n customers, i.e., , ,...,c n1 2= " , . a nonnegative cost cij (cii= 0, 0 ≤ i ≤ n) associated with each arc ( , )i j e! and each vertex i c! is a customer with a non-negative demand pi . the available fleet consists of k different type vehicles located at the depot and the number of available vehicles of each type is fixed and equal to nk . a capacity qk , a fixed cost fk and a variable cost αk are associated with each type of vehicle k in which αk is cost per unit of distance corresponding to each vehicle type k. hence, c cijk ij k#a= represents the cost of the travel from customer i to j with a vehicle of type k. the hffovrp deals with finding the minimum total transportation cost, including the fixed and variable cost for a fleet of vehicles which start and end at the depot so that the following constraints are taken into account: the total load of each vehicle cannot exceed the capacity of the corresponding vehicle type. the used number of vehicles of type k cannot exceed nk. the demand of each customer is satisfied by exactly one vehicle in only one visit. we present following mathematical formulation for hffovrp using variables and yij where, xij k take the value 1 if a vehicle of type k travels directly from customer i to customer j, and 0 otherwise; denotes the route. the flow variables yij specify the quantity of goods that a vehicle k is carrying when leaves customer i to service customer j. min f x c xk jk j n k k ij k j n i n k k ij k 0 11 001 + == === || ||| (1) subject to , ,...,x j n1 1 2ijk i n k k 01 6 == == || (2) , ,...,x i n1 1 2ijk j n k k 11 6# = == || (3) , , , , , , x x j n k k 0 1 1 2 1 2 ij k i n ji k i n 1 1 6 f 6 f # #= = = = | | (4) , , ,…,x n k k1 2jk j n k0 1 6# = = | (5) , ,..., y y p j n1 2 ij k ji k j k k i n i n k k 1 0 01 6 = = = === | ||| (6) int. j. prod. manag. eng. (2017) 5(2), 55-71 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakht, m. and dolatnejad, a. 58 http://creativecommons.org/licenses/by-nc-nd/4.0/ ( ) , , ,..., , , , ,..., p x y q p x i j n i j k k 0 1 1 2 j ij k ij k k i ij k 6 6 !# # = = (7) , , ,...,x k k0 1 2ik j n 0 1 6= = = | (8) , , , ,..., , , , ,..., x i j n i j k k 0 1 0 1 1 2 ij k 6 6 !! = = " , (9) , , ,..., , , ,..., y i j n k k 0 0 1 1 2 ij k 6 6 $ = = (10) the objective function (1) gives the sum of the total fixed cost of the vehicles used plus the total variable routing cost. constraints (2) mean that only one arc can be entered for each customer; however, constraints (3) show that almost one arc can be exited from each customer. constraints (4) states that if a vehicle visits a customer, it can remain there or depart from it. the maximum number of vehicles available for each vehicle type is guaranteed by constraints (5). equality equations (6) insure that the demands of all customers are fully satisfied. constraints (7) state that the vehicle capacity is never exceeded. constraints (8) guarantee that there is not any arc from each customer to the depot. constraints (9) describe that each arc in the network has the value 1 if it is used and 0 otherwise. finally, restrictions (10) force the flow to remain non-negative. 3. the proposed algorithm a route-vehicle pair in open vehicle routing, (r,k), is feasible when route r starts at the depot and the sum of demand of customers on route r is not larger than the capacity of the vehicle type k. the reformulation of the hffovrp is required the following additional notation: k the set of vehicle type, { , ,..., }k k1 2! . rk the set of feasible route r for vehicle type k. r the set of all of feasible route, r r! , r rk k k = ! ' crk cost of rout r for vehicle type k, (r, k) rk! , c c ( , ) r k ij i j r = ! | airk binary parameter of equal to 1 if the customer i is serviced by route r ∈ r and vehicle type k ∈ k, 0 otherwise. xr k binary variable of equal to 1 if route r ∈ rk is used in the solution, 0 otherwise. nk the number of vehicle type k ∈ k. now, hffovrp can be formulated as follows: min c xrk rk r rk k k!! || (11) subject to ,…,a x i n1 1irk rk r rk k k 6= = !! || (12) x n k krk k r rk 6# ! ! | (13) , ,x k k r r0 1rk k6! ! !" , (14) in this formulation, the objective function (11) minimizes the total costs. constraints (12) indicate that each customer is assigned to once and only once feasible route-vehicle pair. constraints (13) ensure that that maximum of the number of used vehicles of each type does not exceed the number of available vehicles of same type. finally, constraints (14) describe that each arc in the network has the value 1 if it is used and 0 otherwise. in this type of formulation, each column corresponds to one routevehicle pair. as number of columns is extremely large, this problem cannot be solved directly, so we use a column generation approach to solve this problem. 3.1. column generation approach column generation approach deals with two problems, the master-problem (mp) and the pricing sub-problem (psp). the mp is linear programming problem that the goal of this problem is to find the minimal cost and to produce the shadow prices of the temporary optimal solution to be used in the pricing sub-problem. since the number of columns of mp is arbitrarily large, a restricted mp is used to initiate the computations. the goal of the pricing sub-problem is to generate additional columns for mp. in this section, the proposed sweep-based algorithm is first described and then the formulations of the master problem and pricing sub-problem are presented. int. j. prod. manag. eng. (2017) 5(2), 55-71creative commons attribution-noncommercial-noderivatives 4.0 international a column generation for the heterogeneous fixed fleet open vehicle routing problem 59 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.1.1. the proposed sweep-based algorithm in this section, the sweep-based algorithm is described. this heuristic belongs to the class of heuristic called cluster first-route second in which a set of initial feasible open routes is generated. to apply the heuristic, we assume that the location of each customer is known in terms of an (x,y) coordinate. we compute the polar coordinates of each customer with respect to the depot and then order the customers by increasing polar angle and generate a list of customer. note that if customer i and customer j have same polar angle, we put the customer that has lower distance respect to the depot earlier in the list. then do the following steps: step 1: sort the type of vehicles randomly. step 2: for each of the type of vehicles (k), perform the following step, until all of the customers to be swept or all vehicles of this type are loaded: step 2.1: for the unused vehicle, we gradually add customers in vehicle route according to list of customers until the capacity constraint is attained. now, we add pair (r,k) to the set of initial feasible routes. step 3: delete the first customer in the list and add it to the end of list. step 4: if updated list was the initial list, stop and the algorithm is terminated, otherwise go to step 1. in figure 2, we illustrate the heuristic described by means of an example involving 7 customers and 3 vehicles. the customer demands and the vehicle capacity are given below. figure 2. sweep algorithm. 3.1.2. master-problem the mp can be formulated as a set covering formulation. this formulation is linear programming problem that each customer is assigned to at least one feasible route–vehicle pair, since the arc costs satisfy the triangle inequality and each customer will be visited exactly once to minimize the costs. also in constraint (14), integrality of variable xr k is eliminated. ( )mp min c xrk rk r rk k k!! || (15) subject to ,...,xa i n1 1rkirk r rk k k 6$ = !! || (16) k n k krk k r rk 6# ! ! | (17) x k r0 1rk k6# # ! (18) creating all feasible routes is considered as a nphard problem. the main idea of column generation approach is to use only a small number of feasible routes in order to find the optimal solution out of a large set of possible feasible routes and additional routes are added only when needed. therefore, at first, we considered the restricted master problem (rmp) containing only the routes that have been generated by sweep algorithm. let generated routes for vehicle type k indicated by r rk k1l , then the rmp created by replaced rk in the mp by r'k. the rmp is solved from this solution and we are able to obtain shadow prices for each of the constraints in the rmp. this information is then utilized in the objective function of the pricing sub-problem. each time the pricing sub-problem is being solved, new routes are being generated as needed and inserted in the set r rk k k = ! l l' . we solve the rmp by solver cplex 12.1 and vector π*=(π*i1,…, π * in, π * 1,…, π * m) is the values of the dual variables (shadow price) corresponding to constraints (16) and (17). pricing sub-problem the pricing sub-problem (psp) is a new problem created to identify a new route–vehicle pair with the minimum reduced cost. the psp uses the shadow price information from rmp to generate promising a route–vehicle pair that is also feasible. we formulated a pricing sub-problem for vehicle type k (psp(k)) and consequently have m problems. we int. j. prod. manag. eng. (2017) 5(2), 55-71 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakht, m. and dolatnejad, a. 60 http://creativecommons.org/licenses/by-nc-nd/4.0/ consider vehicle type k and r ∈ rk. this open route is as follows: r = (i0,i1,…,ih ,ih+1); i0=, ih+1≠ 0 the reduced cost cr k of r ∈ rk is defined as: ( ) ( ) c c c c a f f * * , * * , * r k r k i ir k i i k i i k h h k i h h k i i k i k h h 1 1 1 1 h h h h h h 1 1 r r r r r = = + = = + ! = = = + r | | | | (19) where π*i0=π * k . therefore, reduced cost cr k is equal to sum of the fixed cost and the cost of route r on the directed sub-graph gk=(v, ak) for vehicle type k with arc modified cost defined as follows: c c c i i 0 0 * * *ij k ij k j k ij k j ! r r r = -= r * (20) now, the psp formulation can be written as follows: ( ): _psp k min c f c xrk k ijk ijk j n i n 00 = + == r r|| ( 21) subject to , ,…,x j n1 1 2ijk i n 0 6# = = | (22) ,…,x i n1 1ijk j n 1 6# = = | (23) ≤ ≤ , ,…,x x j n0 1 1 2ijk i n ji k i n 0 0 6= = = | | (24) x 0ik i n 0 1 = = | (25) , ,…,j ny y q x 1 2ijk i n ji k i n j ij k i n 0 0 0 6= = = = = | | | (26) j ≤ ≤ ( ) , , , ,…, ≠q x y q q x i j ni j0 1 2ijk ijk k i ijk 6= (27) ≥ , , , ,…,y i j n0 0 1 2ijk 6 = (28) , , , , ,…, , ≠x i j n i j0 1 0 1 2ijk 6! =" , (29) therefore, the result of psp(k) will be one route for vehicle type k, if the optimal objective value of psp(k) is negative, the route–vehicle pair having the minimum reduced cost cr k will be added to the set r'k in rmp. then the rmp will be solved again and obtain shadow prices for each of the constraints in rmp. for vehicle type k, arc modified cost cij k is obtained and the process is repeated until no routes with negative reduced cost are identified. the cg approach will be described as follows: 1. find initial sets r' of routes by sweep algorithm for the mp. 2. solve the mp by using solver cplex 12.1 and obtain the shadow prices of the optimal solution. 3. for vehicle type k, produce the modified costs cij k, as equation (20). 4. for vehicle type k, solve psp(k) by using solver cplex 12.1 and find feasible route–vehicle pairs with negative reduced costs and add them to the sets r'k. 5. if no new route–vehicle pair was found go to step 6, otherwise continue with step 2. 6. if solution obtained from the last rmp is integer, the approach terminates, otherwise replace rk in hffovrp formulation by r'k , then solve it and the approach terminates. figure 3 presents the column generation approach flowchart. start stop find initial solution for the mp by sweep algorithm solve the restricted master problem produce modified costs for vehicle type k, solve psp(k) hffovrp formulation is integer solution? add route-vehicle pair to rmp route-vehicle pairs with reduced cost <0? yes yes no no figure 3. flowchart of the column generation approach for the hffovrp. int. j. prod. manag. eng. (2017) 5(2), 55-71creative commons attribution-noncommercial-noderivatives 4.0 international a column generation for the heterogeneous fixed fleet open vehicle routing problem 61 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. results the proposed algorithm was coded in aimms with solver gurobi 4.5 and all the experiments were implemented on a pc with pentium 4 at 2.3ghz and 4gb ram running windows xp home basic operating system. aimms is an advanced development environment to build advanced planning systems and optimizing the problems in applied research studies. the numerical experiment is performed using two sets of problem instances. these problems are built of 10-85 nodes including the depot that all randomly located over a square with no service time. they have fixed fleet with capacity restrictions, without route length. euclidean distances are used in the all problems. the first one consists 14 generated by author instances from 10 to 85 customers and the second one consists 5 test problems available in the literature from the wellknown taillard’s benchmark for heterogeneous fixed fleet vehicle routing problem (hffvrp) provided by taillard (1999). it is noted that the first data set is also derived from taillard’s benchmark and the second set consists of instances 9, 10, 11, 12 and 17 with sizes ranging 50, 50, 50, 50 and 75 respectively without the depot. for more information regarding the provided examples visit: http://mistic.heig-vd.ch/taillard/problemes.dir/vrp. dir/vrp.html in this section, we first introduce the benchmark problems and then the detailed computational results obtained. the specifications of these twenty two problems are reported in table 1. table 1. data for problems. instance number of customers number of different type vehicles capacity fixed cost variable cost available number of kind kth of vehicle 1 10 1 20 20 1 1 2 30 40 1.1 1 3 40 70 1.3 2 4 70 200 1.7 1 2 15 1 30 60 1 1 2 60 100 1.1 1 3 80 250 1.5 1 4 150 300 2 1 3 20 1 20 70 1 1 2 35 120 1.1 2 3 50 200 1.2 2 4 120 250 2 3 4 25 1 25 50 1 2 2 35 80 1.1 2 3 50 200 1.2 3 4 120 250 1.7 3 5 30 1 25 35 1 3 2 35 50 1.1 2 3 50 75 1.2 4 4 120 150 1.7 4 6 35 1 50 60 0.7 1 2 120 75 1 2 3 160 200 1.1 3 7 40 1 60 20 1 3 2 140 50 1.7 2 3 200 120 2 1 8 45 1 50 20 1 2 2 150 35 1.4 1 3 200 50 1.4 1 4 300 120 1.7 1 9 50 1 20 20 1 4 2 30 35 1.1 2 3 40 50 1.2 4 4 70 120 1.7 4 5 120 225 2.5 2 6 200 400 3.2 1 int. j. prod. manag. eng. (2017) 5(2), 55-71 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakht, m. and dolatnejad, a. 62 http://mistic.heig-vd.ch/taillard/problemes.dir/vrp.dir/vrp.html http://mistic.heig-vd.ch/taillard/problemes.dir/vrp.dir/vrp.html http://creativecommons.org/licenses/by-nc-nd/4.0/ the proposed compound heuristic algorithm is compared to the solver cplex 12.1 in aimms as an exact algorithm in the table 2. in this table, the first column includes the instance name, the second column shows the number of customers n, and the third and fourth columns present the results obtained by exact algorithm and its cpu time. it is noted that in these results shown in the third column, the exact algorithm continues until the software finishes the search. number of generated pair (r,k) for each instance number of customers number of different type vehicles capacity fixed cost variable cost available number of kind kth of vehicle 10 50 1 120 100 1 4 2 160 1500 1.1 2 3 300 3500 1.4 1 11 50 1 50 100 1 4 2 100 250 1.6 3 3 160 450 2 2 12 50 1 40 100 1 2 2 80 200 1.6 4 3 140 400 2.1 3 13 55 1 20 10 1 2 2 50 35 1.3 2 3 100 100 1.9 2 4 150 180 2.4 1 5 250 400 2.9 1 6 400 800 3.2 1 14 60 1 20 10 1 2 2 50 35 1.3 2 3 100 100 1.9 2 4 150 180 2.4 1 5 250 400 2.9 1 6 400 800 3.2 1 15 65 1 20 10 1 1 2 50 35 1.3 2 3 100 100 1.9 2 4 150 180 2.4 2 5 250 400 2.9 1 6 400 800 3.2 1 16 70 1 20 10 1 3 2 50 35 1.1 3 3 100 100 1.2 2 4 150 180 1.4 2 5 250 400 1.9 1 6 400 800 2.2 1 17 75 1 20 10 1 4 2 50 35 1.3 4 3 100 100 1.9 2 4 150 180 2.4 2 5 250 400 2.9 1 6 400 800 3.2 1 18 80 1 20 10 1 5 2 50 35 1.3 6 3 100 100 1.9 2 4 150 180 2.4 2 5 250 400 2.9 1 6 400 800 3.2 1 19 85 1 20 10 1 5 2 50 35 1.3 6 3 100 100 1.9 3 4 150 180 2.4 2 5 250 400 2.9 1 6 400 800 3.2 1 int. j. prod. manag. eng. (2017) 5(2), 55-71creative commons attribution-noncommercial-noderivatives 4.0 international a column generation for the heterogeneous fixed fleet open vehicle routing problem 63 http://creativecommons.org/licenses/by-nc-nd/4.0/ instance is presented in fifth column. besides, the sixth and seventh columns of table 2 show the results of the proposed algorithm and running time of the algorithm. the best results of the exact algorithm for the presented time in the seventh column are shown in the eighth column. in other word, the exact algorithm continues until the presented time shown in seventh column is over. the best solutions found by sweep algorithms is displayed in the ninth column and finally, to show the method’s performance more clearly, we present the best known solutions (bks) in the last column of table 2. a simple criterion to measure the efficiency of two algorithms cg and sweep algorithm is to compute the relation percentage deviation of their solutions on specific benchmark instances. these values are calculated by below formula. ( ) 100gap valueof thesweepalgorithm valueof thecgalgorithm valueof thecgalgorithm #= figure 4 presents the gap between these two algorithms for the all the instances. in this figure, the horizontal axis shows the name of instances and the vertical axis indicates the percentage of sweep algorithm compared to the cg. form this figure, we conclude that the sweep algorithm is high quality algorithm in order to produce an initial solution because this algorithm able to find very good solution with average almost 41% gap in comparison with the cg. moreover, the lower gap in all instances is 1 with almost 17% gap and the upper gap is instance 19 with almost 70% gap. finally, we see that there is not any relationship between size of the instance and value of gap in this figure. generally, the exact algorithm fails to find optimal solutions for most of the problems especially instances with more than 35 customers and is not able to be used as an efficient and applicant algorithm. as a result, the performance comparison of results between cplex and the cg shows that the proposed algorithm clearly yields better cpu time than the other algorithms in table 2. in more detail, the cg not only can find optimal solution for four inctance including 1, 2, 3 and 4 but also this algorithm capale to find near optimal in two other instance 5 and 6. the ratio of cpu time exact algorithm to cg until instances with 35 customers is shown in figure 5. in this figure, the horizontal axis shows the customer’s number of instance 1 to instance 6 and the vertical axis indicates the ratio of these algorithms. as mention above, the result of cg and exact algorithm in the same cpu time are shown in the column 6 and 8 respectively in table 2. column 7 in table 2 shows this cpu time in second. it is found in table 2 that cplex can reach an optimal solution for only one small-scale instance out of 5 instances. furthermore, we find that in other instances the exact algorithm lower bound is far away from its best solution table 2. comparison results between the proposed algorithm and aimms. instance n cplex12.3 time (sec) #(r,k) cg time (sec) cplex12.3 sweep bks 1 10 191.10 1.09 107 191.10 0.36 191.10 224.22 191.10 2 15 282 171.02 168 282 4.28 525.56 345.14 282 3 20 379.63 16.55 277 379.63 5.44 390.25 465.01 379.63 4 25 437.79 22.78 397 437.79 9.78 532.18 534.09 437.79 5 30 472.76 61.78 585 473.31 16.46 497.48 615.69 472.76 6 35 346.26 55567.75 788 349.95 370.08 na 487.61 346.26 7 40 na 629 600.99 767.42 na 997.13 600.99 8 45 na 787 676.04 2408.40 na 950.46 676.04 9 50 na 1263 907.3 256.72 na 1308.23 907.3 10 50 na 867 507.58 2253.26 na 737.18 507.58 11 50 na 855 826.19 840.18 na 1061.61 826.19 12 50 na 886 947.81 698.96 na 1261.77 947.81 13 55 na 1110 1074.91 1940.82 na 1801.45 1074.91 14 60 na 1274 1937.03 8807.25 na 2239.13 1937.03 15 65 na 1363 1563.33 4203.14 na 2047.70 1563.33 16 70 na 1638 962.57 6357.70 na 1452.27 962.57 17 75 na 1873 1356.67 6021.97 na 2274.05 1356.67 18 80 na 2222 1285.74 8331.64 na 2162.70 1285.74 19 85 na 2470 1295.58 13731.35 na 2199.34 1295.58 int. j. prod. manag. eng. (2017) 5(2), 55-71 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakht, m. and dolatnejad, a. 64 http://creativecommons.org/licenses/by-nc-nd/4.0/ result of the bks. besides, the cg has been able to find the best solutions in four instances including 1, 2, 3 and 4 of the 5 examples. therefore, the cg has been able to find the better solutions than exact algorithm in four out of five examples including 2, 3, 4 and 5. in other words, in 4 examples except instance 1 among remaining 5 examples, the proposed algorithm has been capable of improving the solutions gained from exact algorithm. furthermore, cg has failed in improving the solutions in only instance 1 and has come up with solutions similar to the ones found by exact algorithm. hence, it can be concluded that cg is more efficient than exact algorithm in finding good solutions at the same cpu time. 5. conclusion and future works this paper investigates hffovrp in the transportation system and presents a new combined heuristic algorithm base on cg, which has allowed us to obtain high quality solutions which cannot obtain by exact algorithm. the hffovrp is a variant of the classical ovrp in which customers are served by a given hired heterogeneous fixed fleet of vehicles with various capacities and variable costs. this problem has significant applications in the transportation system, especially when a company used hired vehicles to serve customers. computational results generally have shown that the proposed cg gives better results compared to the exact algorithm in terms of the solution quality in the same cpu time. it seems that using effective heuristic algorithms can lead to gain better initial solutions than the sweep algorithm. furthermore, powerful metaheuristic can be used for solving this problem and other versions of hffovrp. future projects will focus on working on such ideas and making them operational. references aleman, r. e., hill, r. r. (2010). a tabu search with vocabulary building approach for the vehicle routing problem with split demands, international journal of metaheuristics, 1(1), 55-80. https://doi.org/10.1504/ijmheur.2010.033123 anbuudayasankar s.p., ganesh b, k., lenny koh s.c., ducq, y. (2012). modified savings heuristics and genetic algorithm for bi-objective vehicle routing problem with forced backhauls. expert systems with applications, 39, 2296-2305. https://doi.org/10.1016/j. eswa.2011.08.009 brandão j. (2009). a deterministic tabu search algorithm for the fleet size and mix vehicle routing problem. european journal of operational research, 195, 716-728. https://doi.org/10.1016/j.ejor.2007.05.059 çatay, b. (2010). a new saving-based ant algorithm for the vehicle routing problem with simultaneous pickup and delivery. expert systems with applications, 37(10), 6809-6817. https://doi.org/10.1016/j.eswa.2010.03.045 dantzig, g., ramser, j. (1959). the truck dispatching problem. management science, 6, 80-91. https://doi.org/10.1287/mnsc.6.1.80 gendreau, m., guertin, f., potvin, j. y., séguin, r. (2006). neighborhood search heuristics for a dynamic vehicle dispatching problem with pick-ups and deliveries. transportation research part c, 14, 157-174. https://doi.org/10.1016/j.trc.2006.03.002 gendreau, m., laporte, g., musaraganyi, c., taillard, e. d. (1999). a tabu search heuristic for the heterogeneous fleet vehicle routing problem. computers and operations research, 26, 1153-1173. https://doi.org/10.1016/s0305-0548(98)00100-2 lei, h., laporte, g., guo, b., (2011). the capacitated vehicle routing problem with stochastic demands and time windows. computers & operations research, 38(12), 1775-1783. https://doi.org/10.1016/j.cor.2011.02.007 figure 4. gap between cg and sweep algorithms. figure 5. cpu comparison between cg and cplex. int. j. prod. manag. eng. (2017) 5(2), 55-71creative commons attribution-noncommercial-noderivatives 4.0 international a column generation for the heterogeneous fixed fleet open vehicle routing problem 65 https://doi.org/10.1504/ijmheur.2010.033123 https://doi.org/10.1016/j.eswa.2011.08.009 https://doi.org/10.1016/j.eswa.2011.08.009 https://doi.org/10.1016/j.ejor.2007.05.059 https://doi.org/10.1016/j.eswa.2010.03.045 https://doi.org/10.1287/mnsc.6.1.80 https://doi.org/10.1016/j.trc.2006.03.002 https://doi.org/10.1016/s0305-0548(98)00100-2 https://doi.org/10.1016/j.cor.2011.02.007 http://creativecommons.org/licenses/by-nc-nd/4.0/ li, x., leung, s. c. h., tian, p. (2012). a multi start adaptive memory-based tabu search algorithm for the heterogeneous fixed fleet open vehicle routing problem. expert systems with applications, 39, 365-374. https://doi.org/10.1016/j.eswa.2011.07.025 li, x., tian, p., aneja, y. p. (2010). an adaptive memory programming metaheuristic for the heterogeneous fixed fleet vehicle routing problem. transportation research part e, 46, 1111-1127. https://doi.org/10.1016/j.tre.2010.02.004 penna, p. h. v., subramanian, a., ochi, l. s. (2013). an iterated local search heuristic for the heterogeneous fleet vehicle routing problem. journal of heuristics, 19(2), 201-232. https://doi.org/10.1007/s10732-011-9186-y saadati eskandari, z., yousefikhoshbakht, m. (2012). solving the vehicle routing problem by an effective reactive bone route algorithm, transportation research journal, 1(2), 51-69. subramanian, a., drummond, l., bentes, c., ochi, l.s., farias, r. (2010). a parallel heuristic for the vehicle routing problem with simultaneous pickup and delivery. computers & operations research, 37(11), 1899-1911. https://doi.org/10.1016/j.cor.2009.10.011 syslo, m., deo, n., kowalik, j. (1983). discrete optimization algorithms with pascal programs, prentice hall. taillard, e. d. (1999). a heuristic column generation method for the heterogeneous fleet vrp, rairo operations research, 33, 1-14. https:// doi.org/10.1051/ro:1999101 tarantilis, c. d., kiranoudis, c. t. (2007). a flexible adaptive memory-based algorithm for real-life transportation operations: two case studies from dairy and construction sector. european journal of operational research, 179, 806-822. https://doi.org/10.1016/j. ejor.2005.03.059 wang, h. f., chen, y. y. (2012). a genetic algorithm for the simultaneous delivery and pickup problems with time window. computers & industrial engineering, 62, 84-95. https://doi.org/10.1016/j.cie.2011.08.018 yousefikhoshbakht, m., didehvar, f. and rahmati, f. (2014). solving the heterogeneous fixed fleet open vehicle routing problem by a combined meta-heuristic algorithm. international journal of production research, 52(9), 2565-2575. https://doi.org/10.1080/0020 7543.2013.855337 yousefikhoshbakht, m., dolatnejad, a., didehvar, f., rahmati, f. (2016). a modified column generation to solve the heterogeneous fixed fleet open vehicle routing problem, journal of engineering, https://doi.org/10.1155/2016/5692792 yousefikhoshbakht, m. and khorram, e. (2012). solving the vehicle routing problem by a hybrid meta-heuristic algorithm. journal of industrial engineering international, 8(11), 1-9. https://doi.org/10.1186/2251-712x-8-11 int. j. prod. manag. eng. (2017) 5(2), 55-71 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakht, m. and dolatnejad, a. 66 https://doi.org/10.1016/j.eswa.2011.07.025 https://doi.org/10.1016/j.tre.2010.02.004 https://doi.org/10.1007/s10732-011-9186-y https://doi.org/10.1016/j.cor.2009.10.011 https://doi.org/10.1051/ro:1999101 https://doi.org/10.1051/ro:1999101 https://doi.org/10.1016/j.ejor.2005.03.059 https://doi.org/10.1016/j.ejor.2005.03.059 https://doi.org/10.1016/j.cie.2011.08.018 https://doi.org/10.1080/00207543.2013.855337 https://doi.org/10.1080/00207543.2013.855337 https://doi.org/10.1155/2016/5692792 https://doi.org/10.1186/2251-712x-8-11 http://creativecommons.org/licenses/by-nc-nd/4.0/ appendix a. best solutions obtained by cg the best solutions found by the proposed algorithm for the problems are presented in tables 1-19. all the calculations have been performed with a precision of 64 bits and the total solution cost is presented with three or four decimal places. instance 1 routes kind of vehicle cost 1 1-7-8 3 36.262 2 1-6-10-11 3 50.119 3 1-5 1 14.866 4 1-3-4 4 61.745 5 1-2-9 2 28.110 sum 191.102 instance 2 routes kind of vehicle cost 1 1-10-11-16 1 50.93977564 2 1-13-6-12-3-4 4 108.7786929 3 1-7-15-5-14 3 78.72884492 4 1-2-9-8 2 43.54901829 sum 281.9963 instance 3 routes kind of vehicle cost 1 1-3-2 3 34.94452747 2 1-17-4-19 3 53.63437012 3 1-7 1 9.219544457 4 1-14-21 2 48.33244534 5 1-6-16 2 36.36848352 6 1-18-13-10-11 4 86.07287612 7 1-8-9-20-5-12 4 96.91824471 8 1-5 4 14.14213562 sum 379.6326 instance 4 routes kind of vehicle cost 1 1-6-22 3 45.64608577 2 1-11 2 28.600 3 1-3 2 16.01624176 4 1-14-16-21 4 53.19762955 5 1-17-24 1 32.22273631 6 1-18-13-10-26-19-25 3 105.6486204 7 1-7-2 3 42.41666743 8 1-8-9-20-15-12 4 82.380508 9 1-3 1 19.6468827 10 1-5 4 12.02081528 sum 437.7962 instance 5 routes kind of vehicle cost 1 1-4-17-24 2 45.5984443 2 1-28-14 1 23.22656223 3 1-7-3-29-22 4 63.41043657 4 1-2-23 1 35.30470192 5 1-8-9-20-15-12 4 82.380508 6 1-5-31-30-6-16-21 4 86.41338606 7 1-27-11 3 31.3292963 8 1-18-13-10-26-19-25 4 105.6486204 sum 473.312 int. j. prod. manag. eng. (2017) 5(2), 55-71creative commons attribution-noncommercial-noderivatives 4.0 international a column generation for the heterogeneous fixed fleet open vehicle routing problem 67 http://creativecommons.org/licenses/by-nc-nd/4.0/ instance 6 routes kind of vehicle cost 1 1-31-22 1 19.14935811 2 1-3-29-23-2-24-25 2 71.99992496 3 1-27-8-36-20-15-12 3 52.62606364 4 1-5-35-9-14-28-30-6-16-21 3 82.03318409 5 1-7-34-17-4-33-10-26-19 3 79.53758332 6 1-18-13-11-32 2 44.60737999 sum 349.9535 instance 7 routes kind of vehicle cost 1 1-41-23-3-16 1 52.1982838 2 1-14-7-6-18-17-38-15-39 2 139.0471265 3 1-19-9-8-12-20-37 1 73.52931824 4 1-28-32-11-33-31-21-10-36 2 125.7796298 5 1-2-34-4-25-30-35 1 69.13258122 6 1-29-13-27-22-5-26-40-24 3 141.302948 sum 600.9899 instance 8 routes kind of vehicle cost 1 1-37-8-20-12 1 80.24315107 2 1-3-16-44-43-14-15-38-39-45-17-7 3 273.1300937 3 1-2-34-4-30-25 1 51.43887377 4 1-29-13-27-41-22-5-26-40-24-23-42 4 153.3682033 5 1-28-32-11-33-31-21-10-36-35 2 117.8604791 sum 676.0408 instance 9 routes kind of vehicle cost 1 1-15 1 27.80287755 2 1-39 1 26.92582404 3 1-47 2 12.29837388 4 1-25 1 33.13608305 5 1-12 1 29.15475947 6 1-4-19-26 3 56.6872941 7 1-8-36-20 3 31.70669462 8 1-7-34-2-44-42-43 5 103.267770 9 1-18-41-45-33-51 5 91.76199689 2 1-31-49-22 4 51.56637333 10 1-5-46-30-6-48-37-38-21 6 168.6069753 11 1-28-14-16 3 41.07187468 12 1-35-9 3 20.4852813 13 1-27-11-32 4 67.2543306 14 1-13-40-10 4 45.05302896 15 1-3-29-23 4 57.96477825 16 1-17-50-24 2 42.55449871 sum 907.2988 instance 10 routes kind of vehicle cost 1 1-21-46-30-38-6-37 1 94.59151952 2 1-47-35-14-28-16-5 1 65.94248802 3 1-18-41-33-45-4-17-50-25 3 62.37450693 4 1-13-40-10-26-51-19 1 51.04362313 5 1-7-34-2-44-43-42-24 1 52.97935262 6 1-27-8-36-9-20-15-12-39-11-32 2 114.3707895 7 1-3-31-49-48-22-29-23 3 66.27947061 sum 507.5818 int. j. prod. manag. eng. (2017) 5(2), 55-71 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakht, m. and dolatnejad, a. 68 http://creativecommons.org/licenses/by-nc-nd/4.0/ instance 11 routes kind of vehicle cost 1 1-49-27-28 1 49.18808315 2 1-9-32-29 1 37.56681532 3 1-15-25-7-44-24-8 2 136.4414249 4 1-11-6-34-46-16-13 2 150.2146344 5 1-12-39-50-10-17-51-35-31-40 3 143.1569457 6 1-48-19-26-14-42-20-41 3 133.7873043 7 1-5-18-38-45-43 1 47.49518318 8 1-33-2-23-4-21-36-37 2 90.11914556 9 1-47-3-30-22 1 38.22563859 sum 826.1952 instance 12 routes kind of vehicle cost 1 1-44-24-8-49-27 2 125.1892307 2 1-6-34-46-16-13-38-45-18-43 3 221.9402168 3 1-14-19-26-15-25-7 3 180.3042078 4 1-39-10-22-30 1 43.75622479 5 1-50-11-40 1 39.24908076 6 1-47-12-3-17-51-35-31 3 105.7905219 7 1-28-9-32-294 2 75.30656257 8 1-48-5-42-20-41 2 74.12480107 9 1-33-2-23-21-36-37 2 82.1455313 sum 947.8064 instance 13 routes kind of vehicle cost 1 1-14-16 6 33.561 2 1-27-13-40-10-26-56 3 90.871 3 1-5-31-49-6-30-46-28-53-35-47-9-36-54-15-12-39-11-32 1 390.379 4 1-8-20-55 5 45.383 5 1-52-24-50-25 5 64.040 6 1-7-34-2-44-42-43-23-29-22-48-37-38-21 2 284.046 7 1-18-41-4-45-33-51-19 4 112.274 8 1-3 4 34.945 9 1-17 6 19.416 sum 1074.915 instance 14 routes kind of vehicle cost 1 1-20-9-47-55-35-53-14-28-58-16-5-21-46-30 2 529.739 2 1-38-6-61 5 74.677 3 1-4-45-19-18-51-33-41-56-26-10-40-13-32-11-59-27-39-12-54 1 830.922 4 1-15 6 27.802 5 1-57-24-17-52-50-25 4 168.043 6 1-22 6 27.294 7 1-7-34-2-44-42-43-23-29 3 148.414 8 1-8-36-60 5 51.558 9 1-3-31-49-48-37 4 78.582 sum 1937.031 instance 15 routes kind of vehicle cost 1 1-10-40-13-32-11-59-27-39-66-12-54-8-60-15-36-20-9-47-55 1 727.202 2 1-64-57-24 5 55.975 3 1-46-30-49-48-37-61 3 93.873 4 1-35-53-28-14-58-16-6-38-21 4 143.206 5 1-52-50-25-19-51-26-56 3 136.353 6 1-7-63-23 5 41.080 7 1-5-31-3-29-22-62 4 128.298 8 1-18-41-33-45-4-17-34-2-44-42-43-65 2 236.353 sum 1562.34 int. j. prod. manag. eng. (2017) 5(2), 55-71creative commons attribution-noncommercial-noderivatives 4.0 international a column generation for the heterogeneous fixed fleet open vehicle routing problem 69 http://creativecommons.org/licenses/by-nc-nd/4.0/ instance 16 routes kind of vehicle cost 1 1-18 6 8.062 2 1-34-64-57-24-17 4 68.869 3 1-5 6 7.071 4 1-63-23-65-43-2-44-7-42 3 168.074 5 1-32 6 37.108 6 1-52-50-25-19-51-26-56 4 86.117 7 1-27-13-41-4-45-33-10-40-59-11-39-66-67-12-54-15-60 1 257.205 8 1-68-35-47-9-53-28-46-30-6-38-21-71-61 2 133.779 9 1-31-49 5 23.528 10 1-55-14-58-16 5 52.377 11 1-69-3-29-62-22-48-37-70 3 91.320 12 1-8-36-20 5 29.064 sum 962.574 instance 17 routes kind of vehicle cost 1 1-18-51 2 38.60423502 2 1-32 1 37.10795063 3 1-5 2 9.192388155 4 1-22 1 27.29468813 5 1-48-37-72-62 2 82.44826844 6 1-7-34-74-2-44-42-43-65 4 126.5496831 7 1-27-23-41-4-45-33-10-40-73-59-11-39-66-67-12-54-15-60 6 374.1160165 8 1-76-69-3-31-75-29-63-23 4 118.3894944 9 1-64-24-57 2 48.63452897 10 1-68-35-47-53-28-46-30-49-6-38-21-71-61 5 203.3042731 11 1-52-17-50-25-19-26-56 3 132.1757352 12 1-8-36-20-55-14-58-16 3 105.6807833 13 1-9 1 15.8113883 14 1-70 1 37.36308338 sum 1356.673 instance 18 routes kind of vehicle cost 1 1-41 6 14.142 2 1-31 6 14.318 3 1-75-70 6 38.868 4 1-28-14-55 5 40.348 5 1-52-17-50-25-24-57 4 108.747 6 1-74-23 6 31.661 7 1-27-13-40-10-33-51-19 3 117.933 8 1-59-73-32 5 53.940 9 1-8-36-60 5 51.559 10 1-7-34-64-2-44-42-43-65 3 135.777 11 1-53-81-58-16 5 44.297 12 1-4-45-26-56 5 62.867 13 1-76-5-46-30-49-6-38-21-71-61-72-37-48-22 2 237.586 14 1-69-3-63-29-62 4 75.047 15 1-68-77-35-47-78-9-80-20-79-15-54-12-67-66-39-11 1 250.591 16 1-18 6 8.062 sum 1285.743 int. j. prod. manag. eng. (2017) 5(2), 55-71 creative commons attribution-noncommercial-noderivatives 4.0 international yousefikhoshbakht, m. and dolatnejad, a. 70 http://creativecommons.org/licenses/by-nc-nd/4.0/ instance 19 routes kind of vehicle cost 1 1-51 6 29.68 2 1-27 6 6.083 3 1-49 6 20.616 4 1-3 6 14.560 5 1-29 6 24.515 6 1-83-19-26-56 5 74.505 7 1-28-14-55 5 40.348 8 1-8-36-20-79-15-60 4 87.922 9 1-53-81-58-16 5 44.297 10 1-59-73-32 5 53.940 11 1-13-41-40-10-33-45-4 3 111.872 12 1-68-77-35-47-78-9-80-82-54-12-67-66-39-11 1 212.519 13 1-76-69-7-34-74-2-86-44-42-43-65 2 169.280 14 1-31-75-22-48-37-70 4 93.151 15 1-64-85-24-57 5 52.321 16 1-18-52-17-84-50-25 4 74.836 17 1-63-23-62 5 55.418 18 1-5-46-30-6-38-21-71-61-72 3 129.714 sum 1295.577 int. j. prod. manag. eng. (2017) 5(2), 55-71creative commons attribution-noncommercial-noderivatives 4.0 international a column generation for the heterogeneous fixed fleet open vehicle routing problem 71 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2019.9469 received 2017-10-15 accepted: 2018-10-12 sales and operations planning: a comparison between the demand-driven and traditional approaches bozutti, d.f.a1, b and espôsto, k.f.a2 auniversidade de são paulo, av. trabalhador são-carlense, 400, centro, são carlos, são paulo, brazil. buniversidade de ribeirão preto, av. costábile romano, 2.201, ribeirânia, ribeirão preto, são paulo, brazil. a1 daniel.bozutti@gmail.com, a2 kleberesposto@usp.br abstract: supply chain management (scm) is an important concept to establish links among companies. with the aim to reach the scm goals, companies must define processes that links the decisions areas. in this context, a process to be dealt is the sales and operations planning (s&op). the s&op is a tactical planning process, executed on monthly-basis and led by senior management with the aim to balance demand, production, distribution, procurement and finance, to ensure the plans and performance are aligned to support the business strategic plan. in this sense, a literature review was presented in this paper in order to compare the traditional approach and the demand-driven approach for the s&op. as expected, because of the more complex environment to be dealt in a demand-driven environment, the s&op evolved to be able to be executed properly. however, further studies in this area should be developed in order to obtain a final framework for the demand-driven s&op, to analyse applications in industries, to understand performance implications and to develop a performance framework for the demand-driven s&op. key words: sales and operations planning, s&op, demand-driven environment, demand-driven sales and operations planning, supply chain management, literature review. 1. introduction the companies cannot be considered isolated entities in the competitive environment. this scenario is also motivated by the fact that the companies are tending to focus on their core activities (novaes, 2007), increasing the demand of partnering in the market and creating the necessity of the supply chain management (scm) to fulfil the consumer expectations. the importance of scm increases because of the current reality of the market place (i.e., short lead times, more customization, reduction of working capital and internet-based transactions) that increases the possibility of disruption in the supply chain (smith and ptak, 2011). lambert and cooper (2000) argue that the great change in the modern management paradigm is that the companies cannot be dealt as unique and without relations in the supply chain. they complete saying that the new competition will not be among companies, but among supply chains. thus, the scm is an important concept to establish links among companies. the study of scm subject has been gaining prominence since 1990’s. however, after about thirty years of studies, many definitions were created, leading to a difficulty of comprehension and studies of this subject. stock and boyer (2009) conducted a study considering 173 different definitions about scm, from papers to cite this article: bozutti, d.f. and espôsto, k.f. (2019). sales and operations planning: a comparison between the demand-driven and traditional approaches. international journal of production management and engineering, 7(1), 23-38. https://doi.org/10.4995/ijpme.2019.9469 int. j. prod. manag. eng. (2019) 7(1), 23-38creative commons attribution-noncommercial-noderivatives 4.0 international 23 http://creativecommons.org/licenses/by-nc-nd/4.0/ and books. thereby, they proposed the following definition to scm (stock and boyer, 2009:706): “the management of a network of relationships within a firm and between interdependent organizations and business units consisting of material suppliers, purchasing, production facilities, logistics, marketing, and related systems that facilitate the forward and reverse flow of materials, services, finances and information from the original producer to final customer with the benefits of adding value, maximizing profitability through efficiencies, and achieving customer satisfaction”. in this sense, the scm is a competitive model for the companies (pires, 2004), being characterized as an approach that links the manufacturing processes, market, purchasing, sales, financial and all distribution network in order to satisfy the consumer expectations (arnold and chapman, 2004; bozutti et al., 2010). with the aim to reach the scm goals, companies must define processes that links the areas of decision. currently, most of the companies use the enterprise resource planning (erp) systems (ihme and stratton, 2015) whose origin was substantiated by the materials requirement planning (mrp) and manufacturing resource planning (mrp-ii). the structure of mrp-ii and its relationship with erp can be represented as shown in figure 1 (apics, 2006a, 2006b, 2006c, 2006d; chase et al., 2006; corrêa et al., 2007; fernandes and godinho, 2010). it is possible to understand from the figure 1 that the core of the erp system is the mrp (fernandes and godinho, 2010). when mrp evolves to mrp-ii, new features appear in the model, that is, the operations strategy, sales and operations planning (s&op), master production scheduling (mps), shop floor control (sfc) and the capacity planning analysis of each level. the integration among others company areas is achieved with the erp. however, smith and ptak (2011) cite that mrp rules are old to the current reality, leading to a necessity of a new approach. in this sense, companies need an operating model, metrics and communication approaches that lead to visibility creation and efficient sharing of information and risks (smith et al., 2017). the companies need to understand the market and be demand-driven; a first step is to evolve from the “older” mrp to the new demand-driven mrp, which strategically defines buffers in order to decouple direct connections to minimize the total system variability (smith, 2015). to support the demand-driven mrp, the s&op takes an important role and shall also be modified to support this new environment. this study covered the sales and operations planning (s&op) process. this is an important figure 1. mrp-ii structure and the erp (source: adapted from apics (2006a, 2006b, 2006c, 2006d); chase et al. (2006); corrêa et al. (2007) and fernandes and godinho (2010)). int. j. prod. manag. eng. (2019) 7(1), 23-38 creative commons attribution-noncommercial-noderivatives 4.0 international bozutti, d.f. and espôsto, k.f. 24 http://creativecommons.org/licenses/by-nc-nd/4.0/ process that integrates manufacturing, sales and marketing, financials and research and development (r&d) areas with the aim to guarantee the company strategic decision be executed in operations level (corrêa et al., 2007). the integration objective inherent to s&op is in accordance with the scm goals and helps the company to reach good results in internal performance and to fulfil consumer requirements (thomé et al., 2012). considering the new demand economy, companies cannot deal with their traditional approach to plan and control their operations in the same way done prior. it is necessary and fundamental to understand their costumer and this action must begin in the companies’ strategy definition (burrows, 2012). thus, to integrate and coordinate the supply chain operations with the aim to match the demand with supply chain requirements (mendes et al., 2016) is an essential prerequisite to survive and to be competitive in the market (dreyer et al., 2010). currently, a new understanding comes to companies to perceive and fulfil the customer needs. this understanding is called demand-driven approach. to be demand-driven is necessary to establish pattern of responses (chatzopoulos et al., 2012) to respond quickly and efficiently to customers’ orders in accordance of their needs of time, price, quality and quantity (mendes et al., 2016). in accordance with this new scenario, gollamudi (2013) cites that companies must become demanddriven because (i) markets are volatile, (ii) demand fluctuates, (iii) products are specialized, (iv) products has higher variety, (v) necessity of low-cost facilities and (vi) external focus. some elements must be integrated and considered to understand the costumer and achieve the market goals in this new supply chain environment (ambe and badenhorst-weiss, 2011; bjartnes et al., 2008; dreyer et al., 2010; verdouw et al., 2010): control processes defined; integrated decision support tools and methods integrated; roles clear and defined; collaboration models applied; performance measurement; enabled information and communication technology. thus, which are the main differences between the traditional s&op approach and the demand-driven s&op approach? the objective of this paper is to compare the traditional s&op approach and the demand-driven s&op approach. this is an important topic to be dealt, because bower (2016a) bets for the next ten year in s&op that: (i) academic research on s&op will explode, (ii) independent standards will develop, (iii) s&op will become much more virtual, (iv) s&op will focus more on supply-side volatility, (v) supporting technology will improve, (vi) s&op will continue to move inside out, (vii) “assessing risk” will become an s&op catchphrase, (viii) s&op will become more range based, (ix) s&op benefit streams will become more apparent and (x) s&op will begin to proliferate throughout the service sector. there are many benefits of a well-executed s&op, for instance, improved forecast accuracy, reduced inventory, better plant efficiency, fewer schedule cuts, greater profitability, and so on (bower, 2016a). besides these benefits, the same author cites others benefits, which are more intangibles: (i) calm – with the s&op process functional areas work more for results than to fight among them, (ii) rhythm – s&op establishes a rhythm for operations, because it proposes a well-structured process to plan, (iii) control – plans and forecasts tend to become more accurate and the operations knowledge increases; (iv) unknowns – problems can occur during the operations (for instance, a production line failure, an unplanned customer or promotion that spikes demand), but these problems can be anticipated and alternatives, to these problems, can be proposed, (v) team and engagementworking within the s&op process builds teamwork among, thus personnel involved becomes more engaged with their work. murray (2016) cites that s&op plays an important role to the company stakeholders because it enables companies to leverage strategy deployment, financial planning, active ownership and engagement by the corporate team. the same author completes saying that s&op provides the companies with the aid to demonstrate that have a strong team, clear vison and a market-winning strategy. thereby, this paper aims to contribute with the evolution of the s&op processes to be more adherent with the business environment, once its int. j. prod. manag. eng. (2019) 7(1), 23-38creative commons attribution-noncommercial-noderivatives 4.0 international sales and operations planning: a comparison between the demand-driven and traditional approaches 25 http://creativecommons.org/licenses/by-nc-nd/4.0/ importance has already been pointed and researches are necessary. the structure of this paper is: section 1: a contextualization was provided, the objectives were defined and the research importance and relevance were described; section 2: the methodology and the research steps were defined; section 3: a theoretical background was provided, considering the s&op traditional and demanddriven approaches; section 4: the comparison between the traditional and demand-driven approaches was performed; section 5: concluding comments were presented. 2. methodology the sales and operations planning traditional approach was described considering a traditional literature review. once these are established concepts, the classical authors, newest papers and grey literature papers were considered during the development. a current and updated view of the theme were provided, considering both theoretical and practical sides. to achieve the better results for the research about sales and operations planning in a demanddriven environment, a systematic literature review was conducted. this method provided a consistent approach to locate, evaluate and analyse data to conclude what is known or not about the theme (denyer and tranfield, 2009). the systematic review is an essential step to summarize existing knowledge and to find gaps for further researches (kitchenham, 2004). the chosen steps to conduct the systematic literature review were defined considering the best approaches of many authors (brereton et al., 2007; conforto et al., 2011; cook et al., 1997; de souza et al., 2010; denyer et al., 2009; kitchenham, 2004) and are detailed as follow: 1. to define the research problem and the objectives; 2. to select the databases; 3. to define the keywords to conduct the research; 4. to select the exclusion criteria for studies; 5. to review the selected abstracts; and 6. to review the full text of the selected articles emphasizing the analysis of the sales and operations planning process in a demand driven environment. the chosen databases to conduct the research were emerald and sciencedirect, which are databases that comprise operations, organizational management, and social sciences researches. scopus, which is one of largest abstract and citation available database of peer-reviewed literature and considers the main operations and management publishers indexed (for example, emerald, elsevier, springer, oxford university press, ieee, and others), was also used to guarantee a full research. to complete the references bases, it was also considered articles of sales and operation planning from apics (american production and inventory control society), which is one of the most important worldwide association of operations management with a strong link with the most important worldwide companies. other scientific grey literature considered were newsletters, reports, theses and conference papers (weintraub, 2000). the following phrases were adopted in the search engine: “demand-driven sales and operations planning” or “demand driven sales and operations planning” or “demand-driven s&op” or “demand driven s&op” or “dds&op” not “s op”. the period of publication was limited to 2000-2017 to obtain a current view. a parallel search was conducted to obtain the main concepts of the demand driven approach (considering the same 2000-2017 period). to guarantee the state-of-the-art presentation, other research was also conducted considering the following phrases adopted in the search engine: “sales and operations planning” or “s&op” or not “s op” (considering the same 2000-2017 period). for this research were considered both quantitative and qualitative papers. the papers´ frequency along the years is illustrated in graphic 1 (ebsco host, 1 1 2 3 5 4 5 6 3 11 10 14 14 15 15 18 14 10 0 2 4 6 8 10 12 14 16 18 20 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 graphic 1. s&op papers´ frequency along the years. source: ebsco host graphic 1. s&op papers´ frequency along the years. source: ebsco host. int. j. prod. manag. eng. (2019) 7(1), 23-38 creative commons attribution-noncommercial-noderivatives 4.0 international bozutti, d.f. and espôsto, k.f. 26 http://creativecommons.org/licenses/by-nc-nd/4.0/ total of 151 papers), graphic 2 (emerald, total of 66 papers) and graphic 3 (sciencedirect, total of 62 papers). 1 4 1 1 3 3 4 4 2 5 3 4 2 7 5 5 12 0 2 4 6 8 10 12 14 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 graphic 2. s&op papers´ frequency along the years. source: emerald 3. graphic 2. s&op papers´ frequency along the years. source: emerald. 1 1 2 1 4 3 6 2 11 5 6 9 7 4 0 2 4 6 8 10 12 2001 2002 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 graphic 3. s&op papers´ frequency along the years. source: sciencedirect graphic 3. s&op papers´ frequency along the years. source: science-direct. with the found papers, a theoretical background was written and a comparison about the approaches were provided. finally, this research considered the principles of the systematic review as suggested by denyer and tranfield (2009), i.e., replicable, exclusive and aggregative. 3. theoretical background 3.1. s&op traditional approach 3.1.1. introduction to the s&op traditional approach the s&op is a tactical planning process, executed on monthly-basis and led by senior management (pedroso and silva, 2015) with the aim to balance demand, production, distribution, procurement and finance to ensure, the plans and performance are aligned to support the business strategic plan (apics, 2016; feng et al., 2008). it could be considered the “steering wheel” for a company’s business, thus planners must driving down to the strategic road to match the supply-demand plans (lapide, 2009). in this sense, the main role of s&op is to maintain the balance between supply and demand and, in case of misbalancing, to provide proper warnings for contingency plans (vollmann et al., 2005) while help the organization to overcome the silo effect, i.e., departments with individuals goals (swaim et al., 2016). however other s&op objectives can be found (pedroso and silva, 2015): to elaborate operational plans and the company performance; to evaluate the company´s performance on a continuous base; to align the company goals with the operational performance; to contribute with the strategic planning of the company; to guarantee the changes are done correctly; to promote a consistent customer service; to elevate workgroup; to link different plans and function present in the company; to offer consensus in decisions. the main outputs of s&op are (corrêa and corrêa, 2017:472): establishment of revenues monthly goals; billing projections; inventory projections; cash flow projections; budget for material purchase projections and expenses; definition of the tolerance limits of the mps variability; definition of monthly quantities to be manufactured during the demand time fence. until 1950´s, s&op was not known by its denomination. companies used to plan with the aggregate production planning (app). by mid-1980´s the evolution was to manufacturing resources planning (mrp-ii). currently, s&op concept considers a business process with the aim to align sales, operations, development and financial (thomé et al., 2012). int. j. prod. manag. eng. (2019) 7(1), 23-38creative commons attribution-noncommercial-noderivatives 4.0 international sales and operations planning: a comparison between the demand-driven and traditional approaches 27 http://creativecommons.org/licenses/by-nc-nd/4.0/ although, in practice, it is possible to find many definitions to s&op and it is important to have a well-defined process (mccollum, 2011). in this paper was adopted the american production and inventory control society (apics) definition for s&op; this definition considers a synthesis of the main definitions of this process (considering both academicals and practitioners’ side). apics is a global and professional association for operations and scm and a provider of research, education and certification programs, that involves academics and industry practitioners of scm. according to apics dictionary, 15th edition (apics, 2016:164), the definition of s&op is: “a process to develop tactical plans that provide management the ability to strategically direct its business to achieve competitive advantage on a continuous basis by integrating customer-focused marketing plans for new and existing products with the management of supply chain. the process brings together all the plans for the business (sales, marketing, development, manufacturing, sourcing and financial) into one integrated set of plans.” s&op shall be a collaborative process (bower, 2015b) and considering the definition prior described, it is noticed the importance of integrating sales/ marketing and manufacturing. this is not a new issue. shapiro (1977) addresses this issue considering the different point of views existing in these two areas. the author pointed the market perception of lack of production customer-orientation focus and the manufacturing perception of lack of marketing costs, profit and operations orientation. he concluded that a good beginning to minimize this conflict is to have a development and promulgation of clear corporate policies. apics (2006d) affirms that the business planning process provides s&op with the companies and market needs. the functional areas, normally, has no integration and no cooperation, thus the s&op process should make the plans come together. 3.1.2. the s&op traditional approach process as described prior, s&op is a process, executed on monthly-basis and led by senior management with the aim to balance demand, production, distribution, procurement and finance. one of the first difficulties in this process is the participation; scheduling time of all senior management could be arduous, but it is essential to the success of the process (schneider, 2013b). schneider (2013a) also completes citing that the meeting will take from one to two hours (depending on the phase) and must happens at least monthly. prior to begin the s&op process is a pre-requisite to have all leaders of functional areas involved. the people should not have the perception that the s&op preparation work and meetings are an “extra-work”, but it is a fundamental process to achieve customer needs (schneider, 2013b). the s&op process could be divided in five steps (apics, 2006c; corrêa and corrêa, 2017; schneider, 2013b; wallace and stahl, 2008). the first step is to run the sales forecast reports. normally this activity occurs at the end of the month and is executed in the information system department. it consists of three elements (apics, 2006c): updating the file with data from the ended month (inventories, production, backlog, etc.); generating information for sales and marketing personnel to support the forecasting development; divulgating the information to the appropriate people. it is important to notice that is not good if the sales be superior than the planned, because production would not have capacity to produce this excess or future sales orders could be compromised. for someone, it is difficult to understand this, because the company must sell. the sales quantity must be next to the planned quantity, if this not occurs the root cause of the deviation must be analyzed (corrêa and corrêa, 2017). the second step is the demand planning phase. in this phase, sales and marketing personnel review the information received from the first step. the aim of the second step is to update the existing forecast or generate new forecast. the forecast must include all product families and any product life-cycle phase changes. in this step is generated a sales plan. once the available resources are limited, it is important to prioritize markets, thus the sales plan shall reflect the strategical positioning of the company in the market (corrêa and corrêa, 2017; wallace and stahl, 2008). schneider (2013a) proposes that the two first steps lasts about seven days. the first two days are spent int. j. prod. manag. eng. (2019) 7(1), 23-38 creative commons attribution-noncommercial-noderivatives 4.0 international bozutti, d.f. and espôsto, k.f. 28 http://creativecommons.org/licenses/by-nc-nd/4.0/ on data preparation. on the third day team members gather field intelligence, for that is important that the sales people be closer to the customer. on the seventh day, a demand meeting takes place to analyze the demand forecasts and analysis; this meeting must have the participation of the higher-level managers from sales. it is important to notice that the output of this meeting is an unconstrained forecast. in the third step the operations personnel review the output from the second step. in this phase plans must be validated or changed against the availability of supply resources, using the resource planning. the supply resource shall be understood as the resources necessary to produce and distribute the goods and includes materials, personnel, machines and distribution modals. in this steps is proposed a projection of inventory/production that satisfies the sales plan and production capacity restrictions (corrêa and corrêa, 2017; wallace and stahl, 2008). this third step lasts from the eleventh day through the fifteenth day. the master scheduler normally is the facilitator of this step. as prior described, this process identifies any potential constraints with machinery, people or suppliers and develop a supply plan and countermeasures (schneider, 2013b). the objectives of the fourth step, this is the pre-sales and operations plan meeting, are (apics, 2006c): decide about the balancing of supply and demand; resolve problems and differences among the functional areas; identify the areas where agreement cannot be reached; develop scenarios to propose actions alternatives to solve a given problem. decision-makers of all functional areas (sales, marketing, product development, finance and operations) must participate in this step. their job is to check resources constraints, establish priorities and review the demand and supply plans of all product families (wallace and stahl, 2008). the comparison of the actual versus planned performance shall be done. the outputs of this step are (wallace and stahl, 2008): financial plan updated; action plan for each product family; plans for new product introduction; resource plan changes recommendations; scenarios and impacts for areas where agreement could not be reached; recommendations changes to demand and supply strategies; agenda for the executive s&op meeting. this meeting normally occurs on the sixteenth day of the s&op process. this meeting should last from two to three hours and must be faced as a working session. the leaders’ participation is essential and they must feel ownership of the process, by collecting feedback and to make the gathering more effective (schneider, 2013b). the fifth step is considered the main event of the monthly s&op process, whose objectives are (apics, 2006c): to make decisions on each product family, considering the outputs of the fourth step; to authorize changes in the rates of production and procurement; to compare the production plans to the business plan in financial terms; to make decisions where there is not agreement, as expected in the fourth step; to review business key performance indicators and follow up on areas where is having lack of performance. the output of this fifth step is an authorized companywide plan which has the meeting minutes, summary of the decisions, summarized action plan with due dates and responsibilities and the complete production plan for each product family (apics, 2006c). at the end of the all five-step s&op process, operations will have (apics, 2006c): the production plan (for manufactured product families); the purchase plan (for purchased product families); the inventory plan (for make-to-stock product families); and the backlog plan (form make-to-order families). schneider (2013a) and bower (2015b) show the importance of the participation of the president or ceo of the organization to lead this last meeting. the int. j. prod. manag. eng. (2019) 7(1), 23-38creative commons attribution-noncommercial-noderivatives 4.0 international sales and operations planning: a comparison between the demand-driven and traditional approaches 29 http://creativecommons.org/licenses/by-nc-nd/4.0/ duration of this meeting is from one to two hours and occurs normally on the eighteenth day. in addition, (schneider, 2013a) lists six rules that should never been broken to have success in a s&op process: the president or chief executive officer (ceo) should own the s&op process; all participants must take meeting attendance seriously; promote a team environment, but encourage healthy debate; supply does not change the forecast; perform a kaizen event on the s&op process at least twice a year; use an action item list to keep people accountable and ensure what gets measured gets done. to the entire process be well-succeed is necessary to establish an agenda that is followed by all people involved. considering mainly the fourth and fifth steps, the following topics should be covered during the meetings (bower, 2016b): agenda review; review of open items; key performance indicators (kpi) analysis; plan assumptions; plan conversation; plan approval; other relevant topics; executive sponsor comments and questions; next steps and action items; and process recap. 3.1.3. the path to the s&op demand-driven approach the traditional s&op configuration (that of running periodic, multi-functional planning meetings) should not change in the demand-driven environment, but should consider the increased supply chain uncertainties (lapide, 2009). the same author describes five target points to take in consideration for the new s&op scenario: supply-side planners must improve their communication with sales and market; rely more on downstream information, to detect more effectively changes in product consumption; better understanding on the economy impacts on demand; focus on products and markets segments to minimize the demand uncertainties; utilize more formal risk management strategies as the supply-demand risks increase. to identify if the s&op process is with problems and not running properly, schneider (2014) proposes a brief checklist, as follow: top management supports the process in name only; the financial forecast is not derived from the s&op forecast; attendance is lacking or discussion during the meetings is stagnant; metrics are not improving despite the efforts. if problems are found, it is a justification to change to the demand-driven approach. 3.2. demand-driven s&op approach 3.2.1. introduction to the demand-driven s&op approach in the new supply chain management environment, the planners should not solely react to the market, but they need to plan effectively (lapide, 2009), thus to keep competitiveness, customers must see value in what the companies are offering (burrows, 2012). to fulfill customer’s requirements, companies must define their performance objectives, considering on time, reducing lead times, reducing workingin-process (wip), reducing the cost of goods sales, and so on (miclo et al., 2016). in this scenario, to configure business processes, it is necessary a clear reflection of the specific demand requirements to achieve the customer’s need (verdouw et al., 2010) and to look for insulation from market fluctuation has become essential (gollamudi, 2013). the demand-driven environment requires companies to be flexible. hadaya and cassivi (2007) considers five flexibility types to be considered in this new environment: volume flexibility: the ability to adjust production to capacity; launch flexibility: the ability to introduce new products in a rapid and effective manner; access flexibility: the ability to cover the distribution network; int. j. prod. manag. eng. (2019) 7(1), 23-38 creative commons attribution-noncommercial-noderivatives 4.0 international bozutti, d.f. and espôsto, k.f. 30 http://creativecommons.org/licenses/by-nc-nd/4.0/ product flexibility: the ability to produce different products with different characteristics; and responsiveness to target market(s): the ability to respond the needs of the target markets. budd et al. (2012) considers the importance of the information flow in the demand driven environment, thus, they define a demand driven supply chain as the supply chain that has the capacity to share real-time information with the supply chain’s participants. this ability creates to the supply chain’s participants the capacity to react rapidly and effectively to the unexpected changes. winig (2016) conducted a case research in which a bank in south africa gained competitive advantage by developing a customer relationship considering a data-driven approach, i.e., working effectively with big data and information flows (internally and externally), proving the importance of the information flow in this new demand environment. budd et al. (2012) also consider four key pillars for the companies within a demand driven environment: visibility: demand and inventory level information must be transparent in the supply chain; infrastructure: robust infrastructure allows companies to respond effectively to market changes; coordination: coordination promotes flawlessly and cost-effectively execution; and optimization: not solely cost reduction, but configuration to best fulfil the customer requirements. the same authors cite some benefits of the demanddriven supply chain: (i) reduced inventory, (ii) decreased working capital, (iii) improved forecasting accuracy, (iv) reduced transportation costs, (v) optimized infrastructure, (vi) decreased orderexpediting costs, (vii) reduced operating costs, (viii) reduced head count, (ix) decreased sales-planning and operations planning time, (x) reduced lost sales, (xi) improved customer sell-through and satisfaction. considering the methodology described in section 2, it was not found publications that deals with s&op in a demand-driven environment. two papers were found (grey literature) and one book that deals with the theme. thus, demand-driven s&op approach is primarily based on the book of burrows (2012) called “the market-driven supply chain: a revolutionary model for sales and operations planning in the new on-demand economy”, the paper of cecere et al. (2009) called “sales and operations planning: transformation from tradition” and the paper of gollamudi (2013) called “demand driven s&op – maximizing output to match demand variation”. 3.2.2. cecere et al. (2009) demand-driven s&op approach in the the cecere et al. (2009) demand-driven s&op approach, the s&op process evolved from the five-step-s&op process to the nine-step-s&op process, with the aim o be more sensitive and respond correctly to the demand and market complexity. the first step, to collect sales and market inputs, has the aim to collect data, in a collaborative manner, from sales and marketing. the data is historical data and bias shall be evaluated. the second step, to develop a demand plan, has the aim to build multi-period forecasting plan, considering many sources of demand and information from the first step. the third step, demand consensus and refinement, has the aim to find exceptions, understand and deal with them through a comparison between the statistically multi-period forecast and the collective sales forecast. the fourth step, shape demand based on what-if analysis on demand for supply, has the aim to develop a demand plan based on quantity and financial. it can be used marketing intelligence to take advantage from the competitors. at this point the demand shall be shaped. the fifth step, develop a constrained plan by supply, has the aim to identify manufacturing constraints and capacity opportunities for the consensus meeting review. the output from the third step shall be used and options shall be provided considering return on assets, profitability, revenue, customer service and working capital. the sixth step, conduct a what-if analysis by supply to determine trade-offs on the measurements and identify demand-shaping opportunities, has the aim to perform an evaluation of the fourth-step whatif demand shaping based on profitability, revenue, customer service and working capital. the constrains int. j. prod. manag. eng. (2019) 7(1), 23-38creative commons attribution-noncommercial-noderivatives 4.0 international sales and operations planning: a comparison between the demand-driven and traditional approaches 31 http://creativecommons.org/licenses/by-nc-nd/4.0/ of demand shortfalls and capacities opportunities shall be clearly identified. the seventh step, review and gain agreement through a consensus meeting, has the aim to review scenarios and have a consensus on an operating plan based on pricing, operational and functional tradeoffs. the eighth step, publish the constrained plan, has the aim to communicate the plan to the global operational and financial teams for execution. the ninth step, measure and communicate the plan, has the aim to evaluate how the decision prior taken are being performed during the month. the goal is to have a learning cycle to be used in the next demanddriven s&op monthly cycle. there are four steps of maturity to achieve the demand-driven s&op (cecere et al., 2009): step 1: reacting. at this step the goal is to develop an operations planning. the main metrics are: order fill rate, asset utilization and inventory levels; step 2: anticipating. at this step the goal is demand and supply matching. the main metrics are: order fill rate, forecasts errors, inventory turns and functional costs; step 3: collaborating. at this step the goal is profitability. the main metrics are demand error, customer service, working capital and total costs; step 4: orchestrating. at this step the goal is demand sensing to drive an optimized demand response. the main metrics are: demand risk, customer service, cash flow, market share and profit. 3.2.3. burrows (2012) demand-driven s&op approach burrows (2012) cites that one fundamental point to have success in the demand-driven s&op implementation is education. it is necessary to through people, not solely with them. the main idea is to develop people to be able to work in the entire process and not just in part of process that is on a “prepackaged form”. burrows (2012) completes affirming that demanddriven s&op shall be designing to implement the company’s strategy. for that, two other points, besides education, should be handles carefully: design s&op to run the business: as s&op brings all functions together, it becomes the way the business is run. through the s&op meeting, handled in a collaborative manner, the team determines how to interpret the information and makes decision with consensus. the responsibility for results belongs to entire team; design to enable customer centricity: the demanddriven s&op shall be centered on customer and this is possible with a strategy focused on market. the market feedback should be handled in the s&op’s meeting and the team must have the functions well-defined and briefly described. the metrics and strategical goals are different compared with the traditional process, i.e., not solely considering internal goals but measuring enabling concepts to fulfill customer’s need. for that metrics should consider complexity, flexibility, customization and strategic alignment. burrows (2012) considers twenty-six horizontal planning processes that should be aligned during the demand-drive s&op process. the process is illustrated in the figure 2. the numbers presented in the figure 2 may lead to a sequence of steps to be dealt. nonetheless, it differs from the process proposed by cecere et al. (2009) and from the traditional s&op process. the strategic integration has the aim to come back to the company’s strategy in order to find gaps between what is being executed and what was planned. it is important to understand such gaps to find errors in the strategic planning or to make corrections in what is being executed during the month. the demand planning receives the information obtained during the strategic integration to plan how explore the market and customers’ segments. at this point, it is not solely a forecasting, but it is a fundamental process that links the company’s strategy to the market exploration. the rated-based planning has the aim to create a production plan that fulfills the demand at a minimal inventory accumulation based on an agreed service level. the planning in cross-functional coordinating families may be considered an enable process that int. j. prod. manag. eng. (2019) 7(1), 23-38 creative commons attribution-noncommercial-noderivatives 4.0 international bozutti, d.f. and espôsto, k.f. 32 http://creativecommons.org/licenses/by-nc-nd/4.0/ lead to the success of the s&op process. companies works with a large amount of items, when these items are aggregate in families, planning and forecast errors are minimized and the decisionmaking process becomes easier. when these families are planned in a cross-functional way, many areas has the visibility of the concerning issues of the families, thus opportunities can be find easily and contingencies plans can be done faster. the network design has the aim to understand the supply chain issues and opportunities, to evaluate which plant would produce which items, to define distribution strategy (inventory’s position in the supply chain and modals of transport), to understand leadtimes and to make visible the constraints to reach the market. the inventory simulation has the aim to evaluate and define the quantity and location of the inventory in the supply chain. on the one hand inventory is costly but on the other hand inventory defines the service level. by using simulation (specialized software may be used), it would help the decision makers to define the better strategy concerning inventory issues in the supply chain. the cost-to-serve analysis evaluates all costs incurred to serve the customer segments considering the network design and inventory positions and quantities. the view of forecasts should not be limited to statistical forecast analysis. multiple views should make part of this process, i.e., sales person forecasts, customer forecasts, trade association forecasts, marketing forecast of promotions, customers’ segments forecasts, etc. all these forecasts views shall be displayed on a chart, in order to have a great visibility of different points of view. the risk/opportunity analysis has the aim analyse the agreed prior decisions and look for the risks of them and opportunities. this process is important because issues may occur, depending on made decisions and opportunities may not be foreseen. the contingency planning deals with the information of the risk/opportunity analysis to create system resilience. issues may occur and, based on the risks prior evaluated, a contingency plan shall be defined in order to keep the plan on the rails. the sales-gap closure is the last foreseen process for sales planning in which sales persons shall present the final considerations about market and customers and based on the value-chain program a final sales planning is defined in order to continue the entire demand-driven s&op process. the financial report brings to financial terms all prior decisions. not solely costs, but a complete report considering the financial data. scheduling value-chain planning sales planning financial planning marketing planning strategic planning senior management sales-gap closure 11 tilt meetings 15 competitors analysis 19 strategic integration 1 process governance 26 contingency planning 10 annual operating planning 14 supplier alignment 18 new product planning 22 meeting customer metrics 25 risk / opportunity analysis 9 integrated projections 13 promotional alignment 17 generating free cash 24 views of the forecast 8 financial reporting 12 demand sensing 16 monthly s&op meeting 20 meeting performance metrics 23 rate-based planning 4 cost to serve analysis 7 planning in crossfunctional coordinating families 3 inventory simulation 6 demand planning 2 network design 5 customer alignment 21 figure 2. burrows (2012) demand-driven s&op approach (source: adapted from burrows figure 2. burrows (2012) demand-driven s&op approach (source: adapted from burrows (2012)). int. j. prod. manag. eng. (2019) 7(1), 23-38creative commons attribution-noncommercial-noderivatives 4.0 international sales and operations planning: a comparison between the demand-driven and traditional approaches 33 http://creativecommons.org/licenses/by-nc-nd/4.0/ the integrated projections considers the financial reports to be formulated. this projections considers the financial impacts in the future in all evolved areas in the process, that’s why, it is called integrated. the important aspect of this process is the holistic approach, because some areas, sometimes, need to have a poor result in order that the whole company have a positive result. the annual operating planning is not an exclusive process of the s&op, but it inputs information in order to make financial decisions and the s&op process support the creation/review of the annual operating planning. the annual operating planning has the importance to evaluate the company’s financial healthy. the tilt meetings has the aim to formalize all financial decisions made during the s&op process. these meetings also brings the opportunity to find mistakes, to look for opportunities, to evaluate and mitigate financial risks and to make financial information available to the s&op team. the demand sensing is one of the key process of the demand-driven s&op, that’s why, it is a process that is also in evidence at the cecere et al. (2009) proposal. in this process is necessary to identify few and important leading indicators, and not try to go into the terabytes of company’s system information. it is important to analyse, besides the internal perspective of the company, the external perspective, i.e., customer trends, point-of-sale data, economy trends, populational changes, habits change, and so on. the idea is to be prepared and aware with the environment the company is inserted. forecasting may be used to support the decisions of this process. the promotional alignment has the aim to make clear all marketing efforts to promote products. this is an important point to be dealt, because promotions affects the demand and, thus, the necessity of using the manufacturing capacity. the supplier alignment has the aim to communicate the suppliers with the decisions provided by demand sensing and promotional alignment. suppliers should not be aside of the company’s decisions, because, if this happens, raw materials could not be available for production. strong relationships and enhanced information shared shall be better with key suppliers. the competitors analysis brings the opportunity to understand and analyze the competitors movement and actions in the market. it is important to have competitiveness advantages with internal capabilities, but it is also necessary to analyze the competitors. the monthly s&op meeting is a cross-functional meeting in which the decisions must be taken considering the taken considerations foreseen in the previous steps. the areas’ leaders must have a participation in this meeting. the customer alignment has the aim to understand the restrictions of customer’s segment and to define and make clear the service level of each customer’s segment. all areas involved in the process shall know the service level of each customer’s segment with the aim to provide with the correct fulfillment strategy. the new product planning brings to all involved the strategy of new products launch. the new product planning is important because action of all areas could be taken, for instance, manufacturing capacity utilization, warehouse occupation, sales-force efforts, marketing efforts, financial assumptions, etc. the performance metrics meeting brings the view of how the company is performing, considering the operational side. capacity utilization, transportation performance, warehouse occupation and breakdowns analysis are issues that could be analyzed. the metrics shall be shown in a dashboard in order to make the visualization and analysis of each defined key performance indicator easier. the generating free cash is an important process because to be demand-driven companies may face financial issues to perform the proposed service level. this process evaluates financial issues and compares with the agreed service level. the customer metrics meeting is a meeting performed after performance metrics meeting and generating free cash. the aim is to understand and make clear to all involved how the company’s customer segments is being fulfilled. lessons learned may be used to the next monthly cycle and the metrics shall be shown in a dashboard in order to make the visualization and analysis of each defined key performance indicator easier. the process governance is a core process in which all issues must be shared in order to improve the process as a whole. lessons learned of previous monthly cycles shall be tracked to check if they are being applied. improvements actions shall also be set to the next cycles. int. j. prod. manag. eng. (2019) 7(1), 23-38 creative commons attribution-noncommercial-noderivatives 4.0 international bozutti, d.f. and espôsto, k.f. 34 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. comparison among s&op approaches three approaches were found with the conduct research. the first approach was the traditional s&op process, with main contributions from apics (2006d), corrêa and corrêa (2017) and wallace and stahl (2008) which are the classic authors of the theme. the second approach, which is a demanddriven approach, was from cecere et al. (2009). the third approach, also a demand-driven approach, was from burrows (2012). at this point, it is important to find the main differences between the presented approaches. this comparison among the approaches has the aim to make clear the differences and support the further frameworks development. fourteen dimensions were defined considering the information and concepts presented in the researched papers. the goal was to structure the information presented in these papers to provide a better and a clear comparison. the defined dimensions are as follow: number of foreseen processes during the s&op cycles; involved areas and participants; duration of the s&op cycle; level and/or techniques of production aggregation; demand characteristic; level of product variety that can be dealt efficiently; customers need fulfilment approach; level of cooperation among the areas during the s&op cycle; organization structure; operations’ behaviour to fulfil the demand; financial approach; response to the market approach; risk analysis and mitigation; number of meeting during the s&op cycle. the numbers of processes, or steps, during the s&op cycle vary among the approaches. the traditional approach has five main processes to be dealt, while the cecere et al. (2009) approach deals with nine main processes and burrows (2012) approach deals with twenty-six main processes. because of market, manufacturing, logistics, financial and product complexity, more process becomes necessary, the level of details to be dealt increases and the demand-driven approach becomes easier to be properly applied. the involved areas and participants during the s&op cycle do not vary among the approaches, i.e., manufacturing, sales, marketing, financial, research and development, logistics are foreseen to participate during the s&op cycle and meetings. the difference, as it will be after describe, is the level and way of cooperation among these areas. the duration of the s&op cycle also does not vary among the approaches. the recommended cycle is one month. the level and/or techniques of production aggregation is the same for the traditional approach and for the cecere et al. (2009) approach, that is, the planning is done for product families. on the other hands, burrows (2012) proposes a cross-functional coordinating families, in which the definition and coordination of the families shall evolve deeply all areas of the s&op process. the demand characteristic in which the traditional s&op process works better is for predictable demand. in the demand-driven environment the demand is unpredictable and might change rapidly, that’s why, cecere et al. (2009) and burrows (2012) considers in their approach the possibility to work with unpredictable demand. the level of product variety that can be dealt efficiently by the traditional s&op process is from low to medium variety. cecere et al. (2009) approach deals from medium to high product variety and burrows (2012) approach deals with high product variety. the customers need fulfillment approach for the traditional s&op process takes in consideration customers’ demographics, while, for both demanddriven approaches the value proposal is taken in consideration. the level of cooperation among the areas during the s&op cycle for the traditional s&op process is from low to medium. for cecere et al. (2009) approach is considered medium to high collaboration and for burrows (2012) approach is considered high level of collaboration. the organization structure for the traditional s&op approach is functional, that is, the areas are divided in silos. for cecere et al. (2009) is expected to have collaboration through well-defined horizontal process and for burrows (2012) is expected to have int. j. prod. manag. eng. (2019) 7(1), 23-38creative commons attribution-noncommercial-noderivatives 4.0 international sales and operations planning: a comparison between the demand-driven and traditional approaches 35 http://creativecommons.org/licenses/by-nc-nd/4.0/ high level of collaboration through well-defined horizontal process. the operations’ behavior to fulfill the demand for traditional s&op process is reactive to demand, that’s why, it is possible to face stock-outs during the order fulfillment process. for the demand-driven approach is expected to use simulation to have a better understand of the demand, that’s why, it can be considered proactive. the financial approach in the traditional s&op process focuses on minimize loss and increase profit, blinding the customer’s need. for the demanddriven approach, the focus is on balance-sheet, so the service level is considered, customers retention is considered and, of course, profit is also considered. the response to the market approach for the traditional s&op process is based on inventory accumulation to fulfill the customer’s orders, that’s why, stock-outs might happen. on the other side, for the demand-driven approach, demand sensing is used in order to know better customers’ needs and behaviors, thus, the level of stock-outs tends to be minimized. the risk analysis and mitigation is foreseen in all s&op approaches considered in this text. the number of meetings during the cycle for the traditional s&op process is two. for cecere et al. (2009) is three and for burrows (2012) is seven. with more meetings, the issues can be easily shared, the decisions are taken in consensus and the probability to be more adherent with the customer’s expectations becomes higher. the comparison, considering the fifteen dimensions, is summarized in the chart 1. 5. concluding comments s&op has an important role in the companies´ planning, because it generates an integrated plan for all areas. nowadays, with the marketplace complexity, new approaches shall be developed in order to make companies able to have competitive advantage. this paper compared three approaches for the s&op, one being the traditional approach and two being the demand-driven approach. it has been noted that the demand-driven approach is a reality that companies will face, thus this paper contributed with this new approach. further studies might come from this paper: final framework development for the demanddriven s&op; case research of applications of the demanddriven presented approaches; performance implications of the demand-driven s&op; and development of a performance framework for the demand-driven s&op. chart 1. s&op approaches comparison. int. j. prod. manag. eng. (2019) 7(1), 23-38 creative commons attribution-noncommercial-noderivatives 4.0 international bozutti, d.f. and espôsto, k.f. 36 http://creativecommons.org/licenses/by-nc-nd/4.0/ references ambe, i.m., badenhorst-weiss, j.a. (2011). an automotive supply chain model for a demand-driven environment. journal of transport and supply chain management, 1(1), 1–22. https://doi.org/10.4102/jtscm.v5i1.18 apics. (2006a). detailed scheduling and planning (3th ed.). virginia: apics. apics. (2006b). execution and control of operations (3th ed.). virginia: apics. apics. (2006c). master planning of resources. (t. f. cox, m. r. gartner, t. p. geraghty, g. l. kilty, c. v. nemer, & m. ross, eds.) (3th ed.). virginia: apics. apics. (2006d). strategic resources management (3th ed.). virginia: apics. apics. (2016). apics dictionary. (p. h. pittman & j. b. atwater, eds.) (15th ed.). chicago: apics. arnold, j.r.t., chapman, s.n. (2004). introduction to materials management (5th ed.). new jersey: pearson. bjartnes, r., strandhagen, j.o., dreyer, h., solem, k. (2008). intelligent and demand driven manufacturing network control concepts. in third world conference on production and operations management. bower, p. (2015a). party crasher? apics magazine, 25(september/october). bower, p. (2015b). the s&op elevator pitch. apics magazine, 25(july/august). bower, p. (2016a). intangibles. apics magazine, 26(july/august). bower, p. (2016b). planning to plan. apics magazine, 26(september/october). bower, p. (2016c). s&op prophecies. apics magazine, 26(january/february). bozutti, d.f., bueno-da costa, m.a., ruggeri, r. (2010). logística: visão global e picking (série apon). são carlos: edufscar. brereton, p., kitchenham, b.a., budgen, d., turner, m., khalil, m. (2007). lessons from applying the systematic literature review process within the software engineering domain. journal of systems and software, 80(4), 571-583. https://doi.org/10.1016/j.jss.2006.07.009 budd, j., knizek, c., tevelson, b. (2012). the demand-driven supply chain; making it work and delivering results. bcg perspectives. burrows, r.p. (2012). the market-driven supply chain: a revolutionary model for sales and operations planning in the new on-demand economy. new york: amacom american management association. cecere, l., barrett, j., mooraj, h. (2009). sales and operations planning: transformation from tradition. industry value chain strategies, (may), 9. chase, r.b., jacobs, f.r., aquilano, n.j. (2006). administração da produção para a vantagem competitiva. (r. b. taylor, ed.). são paulo: bookman. chatzopoulos, c., gračanin, d., veža, i., tsigkas, a. (2012). demand-driven supply chain using lean & agile principles: a culture for business excellence. in proceedings of international conference for entrepreneurship, innovation and regional development. https://doi.org/10.13140/2.1.1422.8488 conforto, e.c., amaral, d.c., silva, s.l. da. (2011). roteiro para revisão bibliográfica sistemática : aplicação no desenvolvimento de produtos e gerenciamento de projetos. 8° congresso brasileiro de gestão de desenvolviemnto de produto cngdp 2011, (1998), 1–12. retrieved from http://www.ufrgs.br/cbgdp2011/downloads/9149.pdf cook, d.j., mulrow, c.d., haynes, r.b. (1997). systematic reviews: synthesis of best evidence for clinical decisions. in annals of internal medicine (vol. 126, 376-380). https://doi.org/10.7326/0003-4819-126-5-199703010-00006 corrêa, h.l., corrêa, c.a. (2017). administração de produção e operações manufatura e serviços: uma abordagem estratégica (4th ed.). são paulo: atlas. corrêa, h.l., gianesi, i.g.n., caon, m. (2007). planejamento, programação e controle da produção: mrp ii/erp : conceitos, uso e implantação (5a ed.). são paulo: atlas. de souza, m.t., carvalho, r. de. (2010). revisão integrativa: o que é e como fazer. einsten, 8, 102-107. https://doi.org/10.1590/s010407072008000400018 denyer, d., tranfield, d. (2009). producing a systematc review. in d. a. buchanan & a. bryman (eds.), the sage handbook of organizational research methods (pp. 671-689). thousand oaks, ca: sage publications inc. dreyer, h.c., strandhagen, j.o., romsdal, a., hoff, a. (2010). principles for real-time, integrated supply chain control: an example from distribution of pharmaceuticals. ifip advances in information and communication technology, 338 aict, 187-194. https://doi. org/10.1007/978-3-642-16358-6_24 feng, y., d’amours, s., beauregard, r. (2008). the value of sales and operations planning in oriented strand board industry with make-toorder manufacturing system: cross functional integration under deterministic demand and spot market recourse. international journal of production economics, 115(1), 189-209. https://doi.org/10.1016/j.ijpe.2008.06.002 fernandes, f.c.f., godinho, m. (2010). planejamento e controle da produção: dos fundamentos ao essencial. são paulo: atlas. gollamudi, r. (2013). demand driven s&op – maximizing output to match demand variation. bristlecone: your supply chain optimized, 22. int. j. prod. manag. eng. (2019) 7(1), 23-38creative commons attribution-noncommercial-noderivatives 4.0 international sales and operations planning: a comparison between the demand-driven and traditional approaches 37 https://doi.org/10.4102/jtscm.v5i1.18 https://doi.org/10.1016/j.jss.2006.07.009 https://doi.org/10.13140/2.1.1422.8488 http://www.ufrgs.br/cbgdp2011/downloads/9149.pdf https://doi.org/10.7326/0003-4819-126-5-199703010-00006 https://doi.org/10.1590/s0104-07072008000400018 https://doi.org/10.1590/s0104-07072008000400018 https://doi.org/10.1007/978-3-642-16358-6_24 https://doi.org/10.1007/978-3-642-16358-6_24 https://doi.org/10.1016/j.ijpe.2008.06.002 http://creativecommons.org/licenses/by-nc-nd/4.0/ hadaya, p., cassivi, l. (2007). the role of joint collaboration planning actions in a demand-driven supply chain. industrial management & data systems, 107(7), 954-978. https://doi.org/10.1108/02635570710816694 ihme, m., stratton, r. (2015). evaluating demand driven mrp: a case based simulated study. in international conference of the european operations management association (pp. 1–10). neuchatel. kitchenham, b. (2004). procedures for performing systematic reviews. keele, uk, keele university, 33, 1-26. lambert, d.m., cooper, m.c. (2000). issues in supply chain management. industrial marketing management, 29(1), 65-83. https://doi.org/10.1016/s0019-8501(99)00113-3 lapide, b.y.l. (2009). the s&op rudder. supply chain management review, 13(2), 4-6. mccollum, b. (2011). know your s&op. apics magazine, 21(january/february). mendes, p., leal, j.e., thomé, a.m.t. (2016). a maturity model for demand-driven supply chains in the consumer product goods industry. international journal of production economics, 179, 153-165. https://doi.org/10.1016/j.ijpe.2016.06.004 miclo, r., fontanili, f., lauras, m., lamothe, j., milian, b. (2016). an empirical study of demand-driven mrp. in 6th international conference on information systems, logistics and supply chain ils conference 2016 (pp. 1–7). murray, p. (2016). warding off activist investors. apics magazine, 26(march/april). novaes, a.g. (2007). logística e gerenciamento da cadeia de distribuição (6a ed.). rio de janeiro: elsevier. pedroso, c.b., silva, a.l. (2015). dinâmica de implantação do sales and operations planning: principais desafios. gestão & produção, 22(3), 662-677. https://doi.org/10.1590/0104-530x1754-14 pires, s.r.i. (2004). gestão da cadeia de suprimentos: conceitos, estratégias, práticas e casos. são paulo: atlas. schneider, j. (2013a). getting tough with s&op. apics magazine, 23(november/december). schneider, j. (2013b). s&op step by step. apics magazine, 23(september/october). schneider, j. (2014). seeing your s&op. apics magazine, 24(may/june). shapiro, b.p. (1977). can marketing and manufacturing coexist? harvard business review, 55(5), 104-114. smith, c. (2015). the power of decoupling: unlocking the secret to demand driven material requirements planning (ddmrp). demand driven institute. smith, c., ptak, c. (2011). topic : demand driven mrp 21st century supply chains require new demand driven rules and tools. demand driven institute, (april). smith, c., ptak, c., ling, d. (2017). the demand driven adaptive enterprise model: the case for a new system of enterprise management. demand driven institute. stock, j.r., boyer, s.l. (2009). developing a consensus definition of supply chain management: a qualitative study. international journal of physical distribution & logistics management, 39(8), 690-711. https://doi.org/10.1108/09600030910996323 swaim, j.a., maloni, m., bower, p., mello, j. (2016). antecedents to effective sales and operations planning. industrial management & data systems, 116(6), 1279-1294. https://doi.org/10.1108/imds-11-2015-0461 thomé, a.m.t., scavarda, l.f., fernandez, n.s., scavarda, a.j. (2012). sales and operations planning and the firm performance. international journal of productivity and performance management, 61(4), 359-381. https://doi.org/10.1108/17410401211212643 verdouw, c.n., beulens, a.j.m., trienekens, j.h., wolfert, j. (2010). process modelling in demand-driven supply chains: a reference model for the fruit industry. computers and electronics in agriculture, 73(2), 174-187. https://doi.org/10.1016/j.compag.2010.05.005 vollmann, t.e., berry, w.l., whybark, d.c., jacobs, f.r. (2005). manufacturing planning and control systems for supply chain. new york: mcgraw-hill. wallace, t.f., stahl, r.a. (2008). sales & operations planning: the how-to handbook (3rd ed.). cincinnati, ohio: t.f. wallace & co. weintraub, i. (2000). the impact of alternative presses on scientific communication. international journal of grey literature, 1(2), 54-59. https://doi.org/10.1108/14666180010327195 winig, l. (2016). a data-driven approach to customer relationships: a case study of nedbank’s data practices in south africa. in mit sloan management review (p. 16). int. j. prod. manag. eng. (2019) 7(1), 23-38 creative commons attribution-noncommercial-noderivatives 4.0 international bozutti, d.f. and espôsto, k.f. 38 https://doi.org/10.1108/02635570710816694 https://doi.org/10.1016/s0019-8501(99)00113-3 https://doi.org/10.1016/j.ijpe.2016.06.004 https://doi.org/10.1590/0104-530x1754-14 https://doi.org/10.1108/09600030910996323 https://doi.org/10.1108/imds-11-2015-0461 https://doi.org/10.1108/17410401211212643 https://doi.org/10.1016/j.compag.2010.05.005 https://doi.org/10.1108/14666180010327195 http://creativecommons.org/licenses/by-nc-nd/4.0/ pme i j http://polipapers.upv.es/index.php/ijpme international journal of production management and engineering https://doi.org/10.4995/ijpme.2017.6633 received 2016-09-25 accepted: 2017-06-29 stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain kaveh khalili-damghania*, ayda shahrokhb and alireza pakgoharc adepartment of industrial engineering, south tehran branch, islamic azad university, tehran, iran. bdepartment of system and industrial engineering, industrial management institute, tehran, iran. cdepartment of finance, accounting & business systems, sheffield business school, sheffield hallam university, uk * kaveh.khalili@gmail.com abstract: in this paper a multi-period multi-product multi-objective aggregate production planning (app) model is proposed for an uncertain multi-echelon supply chain considering financial risk, customer satisfaction, and human resource training. three conflictive objective functions and several sets of real constraints are considered concurrently in the proposed app model. some parameters of the proposed model are assumed to be uncertain and handled through a two-stage stochastic programming (tssp) approach. the proposed tssp is solved using three multi-objective solution procedures, i.e., the goal attainment technique, the modified ε-constraint method, and stem method. the whole procedure is applied in an automotive resin and oil supply chain as a real case study wherein the efficacy and applicability of the proposed approaches are illustrated in comparison with existing experimental production planning method. key words: uncertain aggregate production planning, supply chain management, automotive industry. 1. introduction aggregate production planning (app) is lowresolution and high-level plan over a medium or long period of time (leung et al., 2003). the app problems are addressed in the scope of tactical and operational levels in supply chains. several methods and approaches including heuristics, mathematical models, and experimental methods were proposed to handle apps (mirzapour al-ehashem, 2012). apps usually involve several type of uncertainties. stochastic programming is a general way of incorporating probabilistic uncertainty into optimization problems. in this paper, a two-stage stochastic programming model is proposed to deal with a new multi-product multi-period multi-objective aggregate production planning (app) problem in a supply chain in presence of uncertainty. the proposed problem is modeled using multi-objective mixed-integer mathematical programming. three objective functions including minimizing expected total costs of supply chain, maximizing expected customer satisfaction level, and minimizing expected supply chain downside risk are considered concurrently. several constraints such as the work force levels, available time, inventory levels, quantity of production, machine capacity, and quantity of raw material purchased, quantity of to cite this article: khalili-damghania, k., shahrokh, a., pakgohar, a. (2017). stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain. international journal of production management and engineering, 5(2), 85-106. https://doi.org/10.4995/raet.2017.6633 int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international 85 http://creativecommons.org/licenses/by-nc-nd/4.0/ products sold, backordering level, and financial risk of supply chain are also considered. some parameters of the proposed model such as demand values, supply capacities, transportation costs, and shortage costs are assumed to be uncertain and handled through a two-stage stochastic programming (tssp) approach. then, three solution procedures, i.e., (i) goal attainment technique, (ii) modified ε-constraint method, and (iii) stem method, are proposed to solve the proposed tssp, distinctively. the proposed approaches are applied on an automotive paint supply chain as a real case study. the efficacy and applicability of the proposed approach is illustrated in the case study. the main contributions of this study are to: (i) proposing a new multi-product multi-period multiobjective app problem in multi-echelon supply chain through mixed-integer multi-objective mathematical programming; (ii) developing a two-stage stochastic programming approach to solve the multi-echelon supply chain problem considering the supply chain downside risk, (iii) adapting three multi-objective solution procedures, including goal attainment method, modified ε-constraint technique, and stem method, to solve the problem, (iv) applying the proposed problem and solution procedures in a real case study, and (v) comparing the results of the three solution procedures. the next parts of this paper are organized as follows. in section 2, the literature of past works is presented. the proposed mathematical model for multi-objective multi-period multi-product aggregate production planning in supply chain is developed in section 3. the solution procedures are also presented in section 3. the real case study and results are presented in section 4. section 5 is allocated to summarize the conclusion remarks and the recommendations for future research. 2. literature of past works in general, app is defined as one of the major production planning categories (giannoccaro and pontrandolfo, 2001; mula et al., 2006). since classic model proposed by holt et al., (1955) and holt et al., (1961) the app problem has been studied extensively (leung and wu, 2004). nam and logendran, (1992) classified app models. gunasekaran et al., (1998) developed a mathematical model to determine the optimum lot-sizes for a set of products and the capacity required to produce them in a multi-stage production system. leung et al. (2003) addressed the problem of aggregate production planning (app) for a multinational lingerie company in hong kong. the multi-site production planning problem considered the production loading plans among manufacturing factories subject to certain restrictions, such as production capacity, workforce level, storage space and resource conditions of the factories. leung et al. (2003) developed a multi-objective model to solve the associated production planning problem, in which the profit was maximized but production penalties resulting from going over/under quotas and the change in workforce level were minimized. sha and che, (2006) proposed a novel multi-phase mathematical approach for the design of a complex supply chain network. the proposed approach was based on the genetic algorithm (ga), the analytical hierarchy process (ahp), and the multi-attribute utility theory (maut). kogan and portugal, (2006) focused on the control decisions in the area of multiperiod, aggregate production planning. the goal was to minimize the expected total costs including productivity, overtime as well as overand underproduction costs. liang (2007) developed an interactive possibilistic linear programming (i-plp) approach to solve multiproduct and multi-time period app problems with multiple imprecise objectives and cost coefficients by triangular possibility distributions in uncertain environments. the imprecise multi-objective app model designed tried to minimise total production costs and changes in work-force level with reference to imprecise demand, cost coefficients, available resources and capacity. rizk et al., (2008) proposed a tight mixed integer programming model for integrated planning of production and distribution in the network. they also proposed a sequential solution approach, based on the independent, but synchronized, solutions of the production and distribution sub-problems. techawiboonwong and yenradee (2010) presented the aggregate production planning for multiple product types where the worker resource could be transferred among the production lines. a mathematical model was formulated in spreadsheet format. then the spreadsheet-solver technique was used as a tool to solve the model. a real situation of a manufacturing company was selected as a case study. the actual data was used to test and validate the proposed model. int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 86 http://creativecommons.org/licenses/by-nc-nd/4.0/ sakallı et al. (2010) discussed a possibilistic aggregate production planning (app) model for blending problem in a brass factory. the main problem was about computing optimal amounts of raw materials for the total production of several types of brass in a planning period. the model basically had a multi-blend model formulation in which demand quantities, percentages of the ingredient in some raw materials, cost coefficients, minimum and maximum procurement amounts were all imprecise and had triangular possibility distributions. a mathematical model and a solution algorithm were proposed to solve the model. mirzapour al-e-hashem et al., (2011) solved the multisite, multi-period and multi-product app problem under uncertainty for a supply chain consisting of multiple-supplier, multiple-producer and multiplecustomer. they also considered the costs related to supply chain and demands as the uncertain parameters. mirzapour al-e-hashem et al., (2012) developed a multi-site, multi-period, multi-product, and multiobjective robust app with regard to conflicts among total costs of supply chain, customer service level, and productivity of workers during medium-term planning horizon in an uncertain environment. corominas et al., (2012) discussed the joint aggregate planning of a production system with manufacturing new units and remanufacturing. karmarkar and rajaram, (2012) discussed a competition version of app model with capacity constraints. toptal et al., (2012) proposed manufacturer’s planning problem to schedule order production and transportation to respective destinations. han et al., (2013) proposed a linear programming model for a hybrid remanufacturing and manufacturing system for production planning problems with deterministic returns. wang and zheng, (2013) developed a responsive and flexible production planning system to cope with uncertain manufacturing factors. fortunately, the app models can deal with details of real-world problems while often efficient algorithms are proposed in order to solve them. as identified by many researchers (bushuev, 2014), the app cost function is convex and piecewise. bushuev, (2014) proposed a new convex optimization approach for solving the app problem. yan et al., (2014) modeled an integrated optimization production planning and scheduling problem through a non-linear mixed integer programming formulation. yan et al., (2014) developed an iterative genetic algorithm to solve the problem. several researches have been dedicated to the field of production planning, logistics and supply chain in recent years. khalili-damghani and shahrokh (2014) proposed a multi-period multi-objective multiproduct aggregate production planning problem. three objective functions, including minimizing total cost, maximizing customer services level, and maximizing the quality of end product, were considered, simultaneously. several constraints were also considered by khalili-damghani and shahrokh (2014). the proposed problem was solved using fuzzy goal programming (fgp) approach (khalilidamghani and shahrokh, 2014). khalili-damghani et al., (2017) proposed a customized genetic algorithm to solve multi-period cross docking truck scheduling problem. tavana et al., (2017) compared drone shipping versus truck delivery in a cross-docking system with multiple fleets and products. tahmasebi et al. (2017) developed a model for the problem of location-routing in post offices. tahmasebi et al. (2017) defined a bi-objective location-routing problem for locating town post office and routing parcels. the problem was modeled through mixedinteger mathematical programming. hafezolkotob et al., (2016) proposed a multi-objective multiperiod multi-product supply chain network problem. the problem was modeled using a multi-objective mixed integer mathematical programming. the objectives were maximizing the total profit of logistics, maximizing service level, and minimizing inconsistency of operations. khalili-damghani and ghasemi (2016) proposed an uncertain decentralized decision making approach through coordination mechanism for a multi-product supply chain planning problem. rezaeisaray et al., (2016) proposed a hybrid multi-criteria decision making approach based on decision making trial and evaluation (dematle), fuzzy analytic network process (fanp) and ordinal/ cardinal data envelopment analysis (dea) model to select ousourcing suppliers. khalili-damghani and tajik khaveh (2015) proposed a logistic planning and design problem in a multi-echelon supply chain consisting of suppliers, manufacturers, and distribution centers. a multi-objective mixed integer mathematical programming model for both decreasing several logistics costs and increasing the service level in supply chain was proposed. khalilidamghani et al. (2015) proposed a new bi-objective mixed integer mathematical programming to reduce the total cost of the supply chain and to balance the workload of distribution centers while the due dates of delivery of perishable product were met, concurrently. hafezolkotob and khalili-damghani (2015) proposed a multi-objective multi-period int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 87 http://creativecommons.org/licenses/by-nc-nd/4.0/ supply chain design and planning problem. the problem tried to minimise logistic costs and maximise service level in a three-echelon multi-product supply chain considering back orders. the layers of chain included suppliers, manufacturers and distribution centres. the parts of logistic costs were discussed and modelled while service level was also interpreted as low level of backorder and shortening the delivery time of products to customers. khalili-damghani et al., (2014) proposed a fuzzy bi-objective mixedinteger programming method for solving supply chain network design problems under ambiguous and vague conditions. khalilidamghani and naderi (2014) proposed a mathematical location-routing model of repair centres and ammunition depots in order to support soldiers in civil wars. the real world problems usually include some types of uncertainty. stochastic programming (sp) is a well-known approach to handle uncertainties in optimization problems. in a two-stage stochastic optimization approach the uncertain model parameters are considered random variables with an associated probability distribution and the decision variables are classified into two stages. the firststage variables correspond to those decisions that need to be made “here-and-now”, prior to the realization of the uncertainty. the second-stage or recourse variables correspond to those decisions made after the uncertainty is revealed and are usually referred to as “wait-and-see” decisions. after the first-stage decisions are taken and the random events realized, the second-stage decisions are made subject to the restrictions imposed by the second-stage problem. due to the stochastic nature of the performance associated with the second-stage decisions, the objective function consists of the sum of the first-stage performance measure and the expected second-stage performance (barbarosoglu and arda, 2004; guillen et al., 2005; azaron et al., 2008). uncertainty can be handled through different paradigms. in stochastic programming approach, two different methodologies can be applied. the first methodology is the distribution-based approach (petkov and maranas, 1998). the second methodology is the scenario-based approach. in this approach the uncertainty is described by a set of discrete scenarios. each scenario is associated with a probability level (poojari et al., 2008). the advantage of this methodology is that there is no limitation for the number of considered uncertain parameters (mirzapour al-e-hashem et al., 2012). stochastic programming can be a useful choice for modeling supply chain planning (scp) with uncertain parameters. kira et al., (1997) formulated a hierarchical production planning (hpp) model under uncertain demand. a stochastic linear programming model (slp) was proposed to better reflect the reality. dupačová, (2002) discussed the applications of stochastic programming. leung et al., (2006) presented a stochastic programming approach for multi-site aggregate production planning with uncertain demand data. leung and wu, (2004) developed a robust optimization model and applied it to the stochastic aggregate production planning. leung et al., (2006) addressed the production planning problem with additional constraints, such as production plant preference selection. they proposed a stochastic programming approach to determine optimal medium-term production loading plans under an uncertain environment. leung et al., (2007) developed a robust optimization model to solve production planning problems for perishable products in an uncertain environment in which the setup costs, production costs, labour costs, inventory costs, and workforce costs were minimized. karabuk, (2008) developed a stochastic programming model in order to address the yarn production planning problem. the proposed model explicitly included uncertainty in the form of discrete demand scenarios. azaron et al., (2008) proposed a two-stage stochastic model, in order to take into account, the effects of the uncertainty in the production scenario for multiobjective supply chain design. they used the goal attainment technique to solve the multi-objective problem. in this method, the preferred solution is sensitive to the goal vector and the weighting vector given by the decision-maker; the same as the goal programming technique. the main drawback of the goal attainment technique to solve the problem is that the preferred solution extremely depends on the goals and weights. to overcome this drawback, azaron et al., (2008) developed a multi-objective stochastic programming model to design robust supply chain configuration networks. they used stem and swt methods, which are two main interactive multi-objective techniques, to solve the multi-objective model. mirzapour al-e-hashem et al., (2012) presented a multi-objective model to deal with a multi-period multi-product multi-site app problem under uncertainty and used an efficient algorithm that is a combination of a modified ε-constraint method and genetic algorithm to solve their problem. mirzapour al-e-hashem et al., (2011) developed a multi-objective two stage stochastic programming model to deal with a multi-period int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 88 http://creativecommons.org/licenses/by-nc-nd/4.0/ multi-product multi-site production-distribution planning problem and applied a hybrid algorithm that is a combination of monte carlo sampling method, modified ε-constraint method and l-shaped method. kazemi-zanjani et al., (2011) investigated multiperiod, multi-product (mpmp) production planning in a manufacturing environment with non-homogeneous raw materials, and consequently random process yields. a two-stage stochastic program with recourse was proposed to address the problem. abdelaziz, (2012) presented various solution approaches for multi-objective stochastic problems. mirzapour al-e-hashem et al., (2013) developed a stochastic app approach in a green supply chain. kazemi-zanjani et al., (2013) proposed a stochastic multi-period, multi-product sawmill production planning problem. the proposed model considered two important issues: (i) randomness in yield and in demand; and (ii) set-up constraints. uncertainties were modelled by a scenario tree in a multi-stage environment. due to literature of past works, there is no unique research addressing the multi-period multi-product multi-objective app in an uncertain multi-echelon supply chain considering financial risk, customer satisfaction and training. in this paper, we are going to address this problem and apply it to a real case study. 3. stochastic aggregate production planning problem in supply chain in this section the proposed stochastic aggregate production planning problem is developed. first the problem and assumptions are defined. the multiobjective aggregate production planning is proposed. then, the classic two-stage stochastic programming model is revisited and the stochastic aggregate production planning is developed. 3.1. problem definition and assumptions the proposed multi-period, multi-objective, multiproduct app problem in supply chains is described based on the following assumptions. it is notable that the assumptions are formed based on real-case study of the research. in this problem, we are going to an aggregate production plan for a three-echelon supply chain incorporating supplier, manufacturers, and customers. this plan will be involved in determination of purchase, production, and delivery decision simultaneously. there are s supplier, one manufacturing plant, and c customers in the supply chain. planning is accomplished in a horizon consists of t time periods (t=1,…,t). batch production system with the capability of producing several kinds of the product is considered. the producer can produce i types of products to response market demand. demand can be either satisfied or backordered. no subcontracting is allowed for products. two working shifts are considered in a day (q ∈ {1,2}). regular time production (q=1) and overtime production (q=2). the holding cost of inventory of products are predetermined and known in advance. several skill levels (k-levels) are considered for workforce. during the planning horizon, training courses are accomplished and the skill level of workforce is improved. there are also several (i.e., l=1, 2,…, l) types of training courses. the first type of training (l=1) enhances the workers from level k=1 to level k=2. the second type of training (l=2) enhances the workers from level k=2 to level k=3 and so on. salary of workers is independent of unit production cost. the safety stock is considered in the quantity of production in each period. according to demand of market, hiring and firing of manpower are eligible with restricted limits. the nominal capacity of production is usually decreased by a fixed failure percentage into actual capacity. if an unexpected failure occurs during a shift the repair process is accomplished in the next shift. in each period of planning, the shortage of production is recovered by overtime production. due to inflation and low holding costs, keeping finished products is economic. int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 89 http://www.sciencedirect.com/science/article/pii/s0377221711002700 http://creativecommons.org/licenses/by-nc-nd/4.0/ the transportation costs between factory and each customer’s location are known in advance. three objective functions are considered as expected total costs, expected customer satisfaction level, and expected supply chain downside risk. future economic scenarios (i.e., the nature status) will fit into one of four possible scenarios; (i) boom, (ii) good, (iii) fair, and (iv) poor, with associated probabilities of p1, p2, p3 and p4, respectively. the uncertainty of parameters in real case study are modeled using two-stage stochastic programming. stochastic programming methods usually do not provide any control on the solution’s variability throughout the different scenarios. therefore, the downside risk (or the risk of loss) is considered as a risk measure and incorporated into the twostage stochastic programming model (azaron et al., 2008; azaron et al., 2010). goal attainment technique, modified ε-constraint method, and stem method used to solve the problem. 3.2. notations: parameters, indices, and decision variables notations used in the proposed multi-objective mathematical programming are summarized in following tables. the notations of objective functions are presented in table 1. it is notable that z2 and z3 which are qualitative objective functions are measured through equations (18)-(20). table 2 shows the notation of sets and associated indices. table 3 present the notations used for parameters and decision variables are illustrated in table 4. table 1. notation of objective functions. definition objective function expected total costs of supply chainz1 expected customer satisfaction levelz2 expected supply chain downside riskz3 table 2. sets and indices. definitionnotation number of periods in the planning horizon; t=1,…,t t number of product types; i=1,…, ii raw material type; m=1,…,mm types of shifts q ∈ {1,2}. (q=1; regular time, and q=2; overtime) q skill levels of workers; k=1, 2, …, kk types of training; l=1, 2, …, ll suppliers; s=1, 2, …, ss customers; c=1, 2, …, cc number of objective functions; j=1, 2, 3.j types of scenarios; n=1, 2, …, nn 3.3. framework of two-stage stochastic programming model the general form of the two-stage stochastic programming model is briefly revisited (dantzig, 1955; kall and wallace, 2003; ruszczynski and shapiro; 2003): ( , )minc x e xt1 { ~+ ~" , (1) subject to: ax=b (2) 1nx r +∈ (3) where φ(x,ω)=min qt y (4) subject to: t(ω)x+wy=h(ω) (5) 2ny r +∈ (6) where e :~" , is the expected value function, ω is the random vector, h(ω), and t(ω) represent a particular sample from a multivariate probability space ω. equations (1)–(3) and equations (4)–(6) refer to the first and second stages, respectively. the vector of x is first-stage decision variables. the optimal value of x is not conditional on the realization of the uncertain parameters. the parameter c1 is the vector of cost coefficients at the first stage. a is the first-stage coefficient matrix and b is the corresponding righthand side vectors. also, the vector of y is secondstage (recourse) decision variables, q is the vector int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 90 http://creativecommons.org/licenses/by-nc-nd/4.0/ table 3. notations for parameters. unitdefinitionparameter dollarproduction cost per hour for product i in shift q in period t coiqt hourprocess time of product i in period t ait man/hourminimum workforce available hr dollarcost workforce of level k in period t clkt dollarhiring cost workforce of level k in period t chkt dollarfiring cost workforce of level k in period t cfkt dollartraining cost of type l in period t ctlt man/hourfraction of the workforce variation allowed in period t αt machine/hourrequired machine hours to produce unit of product i in period t mtit machine/hourpercentage of machine capacity that is lost due to interruption in period t εt machine/hourpercentage of machine capacity that is lost due to repairs in period t μt machine/hourmaximum of machine capacity that is available for product i in shift q in period t mciqt houravailable regular time of product type i in both shifts in period t atiqt kilo-gramunits of raw material type m required to produce one unit of product i γim kilo-gramsafety stock of raw material type m ssrm dollarholding cost for raw material type m in period t in factory crmt kilo-gramavailable capacity of factory for storage of raw material type m in period t vrmt dollarholding cost of unit of product i in period t in factory cpit kilo-gramavailable capacity of factory for storage of finished-product i in period t vpit kilo-gramsafety stock of product isspi daydue date of product i in customer location c duic datlead time required for shipping products from factory to customer center cltcc daylead time required for shipping raw material from supplier s to factoryltss dollartransportation cost from supplier s to factory in period tt1st dollartransportation cost from factory to customer center c in period t in scenario nt2nict machine/hourpercentage of machines capacity that is available for overtime.δ man/hourpercentage of workforce that is available for overtime.ol man/hourminimum percentage of workers that is available for training.τ dollartotal cost when the scenario n is realizedtcn dollaravailable budget in scenario nωn -occurrence probability of scenario n pn quantitydemand of product i in customer center c in period t in scenario ndenict quantityallowable shortage of product i in period t θit quantityallowable sale of product iζi kilo-gramallowable purchase of raw material mπm dollarbackordering cost of product i in period t in customer center c in scenario ncbnict dollarcost of raw material m purchased from supplier s in period t cmsmt table 4. notation for decision variables. unitdefinitiondecision variables quantitynumber of product i produced in shift q of period txiqt man/hournumber of available workers of level k in period txlkt man/hournumber of hired workers of level k in period txhkt man/hournumber of fired workers of level k in period txfkt man/hournumber of workers trained course level l in period txtlt kilo-graminventory level of raw material type m at the end of period t in factoryxrmt kilo-gramnumber of units of raw material m purchased from supplier s to factoryxmsmt quantityinventory level of finished-product i in period t in factoryxpit -supply chain downside risk of scenario ndrn daydelivery time of scenario ndn quantitynumber of units of product i produced by factory for customer c in period t in scenario nynict quantitybackorder level of product i in period t in customer location c in scenario nbnict int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 91 http://creativecommons.org/licenses/by-nc-nd/4.0/ of cost (recourse) coefficient for the second stage, w is the second-stage (recourse) coefficient matrix, and h(ω) is the corresponding right-hand side vector. t(ω) is the matrix that ties the two stages together. in the second-stage model, the random constraint defined in (5), h(ω)-t(ω)x, is the goal constraint. the violation of this constraint are allowed, but the associated penalty cost, qt y, will influence the choice of x. the function φ(x,ω) is the recourse penalty cost or second-stage value function, and the notation eω{φ(x,ω)} denotes the expected value of recourse penalty cost (second-stage value function) with respect to the random vector ω. assuming the distribution of ω is discrete, i.e. the random parameter takes one of a finite set of values (scenarios) {ω1,…,ωn} having probabilities{p1,…,pn}, the two-stage model can be re-formulated as follows (ahmed, 2010): min p c x q yn t n nt n1 +_ i| (7) subject to: axn=b , …,n n16 = (8) , , …,t x w y h n n1n n n n n 6~ ~+ = =^ ^h h (9) 1 1,...,nn n nx r + ∀ =∈ (10) 2 1,...,nn n ny r + ∀ =∈ (11) 1 2 ... nx x x= = = (12) note that duplicates of the first-stage variable have been presented for each scenario. the last constraint, known as the non-anticipatively constraint, guarantees that the first-stage variables are identical across the different scenarios. 3.4. multi-objective stochastic app model formulations 3.4.1. objective functions three objective functions are considered for the proposed model of this research as follows: 3.4.1.1. expected total costs of supply chain minz p tcn n n n 1 = ! | (13) the objective function (13) is total expected cost of supply chain. the total costs of supply chain includes eleven terms as follow: (i) production costs per unit, (ii) holding costs of products, (iii) holding costs of raw materials, (iv) the cost of purchased materials, (v) costs of salary of workers, (vi) costs of hiring, (vii) costs of firing, (viii) costs of training, (ix) transportation costs of raw materials, (x) transportation costs of end-products, and (xi) backordering costs, respectively. the sum of these eleven terms is called as total cost (tcn) when the scenario n is realized and defined in (14). tcn = i∈i ∑ q∈ 1,2{ } ∑ t∈t ∑coiqtxiqt + i∈i ∑ t∈t ∑cpitxpit + + m∈m ∑ t∈t ∑crmtxrmt + cmsmtxmsmt t∈t ∑ m∈m ∑ s∈s ∑ + + k∈k ∑ t∈t ∑clktxlkt + k∈k ∑ t∈t ∑chktxhkt + + k∈k ∑ t∈t ∑cfktxfkt + l∈l ∑ t∈t ∑ctltxtlt + t1st xmsmt t∈t ∑ m∈m ∑ s∈s ∑ + + i∈i ∑ t∈t ∑t2ictn yictn + i∈i ∑ t∈t ∑cbictn bictn c∈c ∑ c∈c ∑ , ∀ n (14) based on two-stage stochastic programming models, it can be classified into first and second-stage variables as follows: t n n n n n n n n n tc p tc x p q y ∈ ∈   = +     ∑ ∑ (15) equation (15) includes two terms as: (i) the term “ n n n n p c x ∈        ∑ ” is the first-stage variables (fsv), which is represented by equation (16), (ii) the term “ t n n n n n p q y ∈ ∑ ” is the second-stage variables (ssv) that is represented by equation (17), fsv = i∈i ∑ q∈ 1,2{ } ∑ t∈t ∑coiqtxiqt + i∈i ∑ t∈t ∑cpitxpit + + m∈m ∑ t∈t ∑crmtxrmt + cmsmtxmsmt t ∑ m ∑ s ∑ + k∈k ∑ t∈t ∑clktxlkt + k∈k ∑ t∈t ∑chktxhkt + k∈k ∑ t∈t ∑cfktxfkt + l∈l ∑ t∈t ∑ctltxtlt + t1st xmsmt t ∑ m ∑ s ∑ (16) int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 92 http://creativecommons.org/licenses/by-nc-nd/4.0/ expected customer satisfaction level ssv = pn n∈n ∑ i∈i ∑ c∈c ∑ t∈t ∑t2ictn yictn ⎛ ⎝ ⎜ ⎜ ⎞ ⎠ ⎟ ⎟ + + pn n∈n ∑ i∈i ∑ c∈c ∑ t∈t ∑cbictn bictn ⎛ ⎝ ⎜ ⎜ ⎞ ⎠ ⎟ ⎟ , ∀ n (17) 2 n n n mi pz dn = ∑ (18) where, , i n n ictt ic i i c c t t a duy nd ∈ ∈ ∈ − ∀= ∑∑∑ (19) in this research, on time delivery is considered as a measure to assess the customer satisfaction level. this objective function minimizes the difference between delivery time and the due date for all type of products. as the probabilities of different scenarios are considered in objective function (18), so it calculates the expected customer satisfaction level. delivery of product to customers earlier than due date, which we called “earliness”, is not suitable while delivery of product to customers later then due date, which is called “tardiness”, is not also desirable. so, the equation (19) minimizes both earliness and tardiness, simultaneously. this goal includes two terms as: (i) actual delivery time of product i to customer c, and (ii) due date of delivery of products i to customer c according to contract. where pn represent the occurrence probability of the n-th scenario. 3.4.1.2. expected supply chain downside risk 3 n n n pmi z drn = ∑ (20) the objective function (20) minimizes the downside risk or the expected total loss. where pn and drn represent the probability of occurrence of the n-th scenario and the downside risk of scenario n, respectively. 3.4.2. constraints the workforce level constraints are considered using (21)-(25). hr ≤ k∈k ∑xlkt, ∀t (21) xl kt = xl k t −1( ) + xhkt − xfkt + xtlt , ∀ k, ∀t,t > 1 (22) xhkt +xfkt ≤ α(t−1) . xlk(t−1),∀ k, ∀ t, t > 1 (23) τ . hr ≤ t∈t ∑xtlt, ∀l (24) l∈l ∑xtlt ≤ xlk(t−1), ∀ k, ∀ t, t > 1 (25) set of constraints (21), which are written for all periods of planning, assures that a minimum number of workers should be utilized in a period of planning. set of constraints (22) is a balance equation for workforce level and ensures that the available workforce with skill level k in a certain period are equal to the workforce with the same skill level k in previous period plus the change of workforce level in the current period. set of constraint (23) assures that the change in workforce level in each period of planning does not exceed a predetermined proportion of workers in each period. set of constraints (24) assures that in all periods of planning, the trained workforce should be greater than or equal to a predetermined percentage of the minimum available workers during all periods. set of constraint (25) assures that the number of workforces with skill level k who are trained for upper skill levels in current period cannot exceed the available workforce with skill level k in previous period. the available time limit of working shifts is presented using constraints (26)-(27). ,. , . 1it iqt iqt kt i i i i k k a x a t t qx l ∈ ∈ ∈ ∀≤ =∑ ∑∑ (26) . . 2. , , it iqt iqt kt i i i i k k a x a t ol t qx l ∈ ∈ ∈ ≤ ∀ =∑ ∑∑ (27) set of constraints (26)-(27) assures that the required production time for all periods of planning and in each working shifts are less than or equal to available regular production time and overtime, respectively. int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 93 http://creativecommons.org/licenses/by-nc-nd/4.0/ the inventory level situations are demonstrated using constraints (28)-(29). the limitations of inventory level of raw material and the production limitations in each period of planning are presented using constraints (30)-(31). xpit i∈i ∑ = xpi(t−1) i∈i ∑ + xiqt q∈{1,2} ∑ i∈i ∑ − yic(t−ltcc ) n c∈c ∑ i∈i ∑ , ∀ n, ∀ t, t > 1 (28) xrmt = xrm(t−1) + xmsm(t−ltss ) s∈s ∑ − xiq(t−1) ⋅γ im q∈ 1,2{ } ∑ i∈i ∑ , ∀ m, ∀ t, t > 1 (29) ssrm ≤ xrmt , ∀ m, ∀ t (30) sspi ≤ q∈q ∑xiqt, ∀i, ∀t (31) set of constraints (28), which are written for all products, assures that the amount of inventory of finished products in period t in factory is equal to the amount of inventory of finished products in period t-1 in factory plus the amount of produced finishedgoods in period t in both working shifts minus the amount of all products produced by factory for all customers without considering the lead time. set of constraint (29) assures the balance of raw materials. set of constraints (30) assures the satisfaction of safety stock of raw materials. set of constraints (31), which is written for all product types and all periods of planning, guarantee the satisfaction of safety stock of finished-products in working shifts. the capacity of machines for each planning periods for both working shifts are presented using constrains (32)(33). i∈i ∑mtit .xiqt ≤ mciqt i∈i ∑ − εt. mciqt i∈i ∑ , ∀t,q =1 (32) i∈i ∑mtit . xiqt ≤ δ . mciqt i∈i ∑ − µt. δ . mciqt i∈i ∑ , ∀ t, q = 2 (33) set of constraints (32)-(33) assures the satisfaction of the maximum available capacity of machines in regular time and overtime, respectively. the limitation of number of units of raw material m purchased from all suppliers to factory in each period of planning are presented using constraint (34). , , ssr xm m tm m smt s s $ 6 6#r ! | (34) set of constraints (34), assures that the number of units of raw material m purchased from all suppliers are more than or equal to the certain percentage of the safety stock of raw material m. this determines the lower limit of purchase of raw material m. generally, the purchase orders of raw materials are seasonal. for instance, equation (34) is written for t=3,6 in a planning horizon consists of six periods. for each scenario, the maximum amount of all products that provided by factory for all customers is presented using constraints (35). for each scenario, the minimum amount of product i that provided through factory for customer c in the period t is presented using constraints (36).   yic(t−ltcc ) n t∈t ∑ c∈c ∑ i∈i ∑ ≤ deictn +bic(t−1)n( ) t∈t ∑ c∈c ∑ i∈i ∑ ,∀ n (35)   ζ i ⋅ deict n ≤ yict n , ∀ n,∀i,∀c,∀t (36) set of constraints (35) assures that the amount of all products produced by factory for all customers without considering the lead time in each scenario, is less than or equal to the amount of backorder of all products for all customers in all previous periods and the demand of all products for all customers in all periods. set of constraints (36) assures that the amount of sales or the number of units of product i produced by factory for customer c in period t in scenario n, are more than or equal to the certain percentage of the demand which determines the lower limit of sales. the backorders are accepted and the associated constraints are presented as (37)-(38). ≤ , , , ,b de n c t t tictn it ictn i ii i 6 6 6 !i !! || (37)   , , ,0 ,nict nb i c t t∀ ∀ ∀ ∀ == (38) set of constraints (37) assures that the backorder level at the end of period t cannot exceed the certain percentage of the demand. this determines the upper limit of shortage. set of constraints (38) assures that there is no possibility for backordering at the last of planning period. the financial risk of supply chain is presented using constraint (39). ,n n ntc dr n− ω ≤ ∀ (39) where total costs of supply chain (tcn) was defined in (14). also ωn and drn are available budget of scenario n and the supply chain downside risk for int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 94 http://creativecommons.org/licenses/by-nc-nd/4.0/ scenario n, respectively. non-negativity of decision variables are presented in (40)-(41).   xiqt , xrmt , xm smt , xpit , bict n ,yict n ≥ 0, ∀n,∀i , ∀q, ∀t ,∀c,∀s, ∀m (40)   xlkt , xhkt , xfkt , xtlt ≥ 0, ∀k , ∀l , ∀t (41) 4. solution procedures many real-world optimization problems are involved with more than a single objective function. in multi-objective mathematical programming (momp), there is no single optimal solution that simultaneously optimizes all the objective functions. therefore, the decision makers are looking for the non-dominated solutions, instead of a single optimal solution. the methods for solving momp problems can be classified into three major categories: the a priori methods, the a posteriori methods and the interactive methods (hwang and masud, 1980; masud and hwang, 1980; miettinen, 1998; mavrotas, 2009). the strengths and weaknesses of momp methods are presented in table 5. table 5. the strengths and weaknesses of the methods for solving momp problems. methods advantages shortcomings a-priori easy, low computational burden, available software need unrealistically precise information, need extensive sensitivity analysis a-posteriori the expression of preference follows the optimization phase, can produce subsets of efficient solutions computational burden, not widely available software interactive iterative, the decision maker guides the search, the decision maker “learns” about the problem need extensive interaction with the decision maker , the decision maker decides based on samples in this paper, three momp methods, called: (i) goal attainment method, (ii) modified ε-constraint method, and (iii) stem method, are used to solve the problem (13)-(41). the goal attainment method is one of the a priori methods, the modified ε-constraint is a posteriori method and presents a “comprehensive framework” to obtain the optimal pareto solutions for the multi-objective optimization problem (haimes et al., 1971; mirzapour al-ehashem et al., 2012). the stem method is an interactive method which presents “good” solutions and the relative importance of the objectives for decision maker. 4.1. goal attainment method the goal attainment which was introduced by gembicki and haimes, (1975) is presented here: minimize υ (42) subject to: , r x! !y }+ (43) ( ) , ...,f x w f j n1j j j#y=) (44) where, ψ is the feasible region, the term υ is the attainment element, the weigh vector, wj, is used to express the relative importance of objectives j, fj * is the goal of objective function j, and fj(x) the objective function j. the problem (13)-(41) is transformed into the optimization problem (45)-(47) using the goal attainment approach stated in (42)-(44): minimize υ (45) subject to: constraints (21)-(41) (46) , , ,z w b j 1 2 3j j j 6#y= (47) where, zj, j=1,2,3 are objective functions, respectively. also, are goals of objective functions, and are weights of objective functions, respectively. the solution is sensitive to value of goals and weights. the weights relate the relative under-attainment of the goals and a smaller w is associated with the more important objectives. using goal attainment method, the best solution can be determined by the nearest pareto-optimal solution from bj, j=1,2,3. the weights are generally normalized so that 3 1 1j j w = =∑ . int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 95 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4.2. modified ε -constraint method the modified ε-constraint method which was first introduced by haimes et al., (1971) is also adapted to solve the proposed multi-objective problem. step 1. select one of the objective functions as the main objective function and convert other objective functions into constraint. then create the payoff table by the individual optimization of each objective functions separately. find the range of each objective function. step 2. determine the grid points. then, divide the range of each objective function to λ equal intervals using λ-1 intermediate equidistant grid points; that are used to vary parametrically the right-hand side of objective function. step 3. solve the resultant model for each value of right-hand side. so an momp with minimization objective functions is converted to model (48)-(50) using ɛ-constraint method: ( )min z x r s j u u u j { i= ! ' 1| (48) subject to: ( ) ,z x s u ju u u 6 !f+ = (49) ,x x s ru! ! + (50) where θ is a small number usually between 10-6 and 10-3 su is a slack variable, ru is the range of objective function zu, εu is right-hand side which is parametrically determine in it associated range, and x is the feasible region of the primary model. the first objective function (i.e., expected total costs of supply chain) is selected as the main objective function and the other objective functions (i.e., expected customer satisfaction level, and expected supply chain downside risk) are converted into constraint. therefore the following optimization model (51)-(54) is considered to solve the problem (13)-(41). min z r s r s 1 2 2 3 3{ i= +a k% / (51) subject to: constraints (21)-(41) (52) ,,z s u 2 3u u u 6f+ == (53) , rs s2 3 ! + (54) where, r2 and r3 are the ranges of z2 and z3 , respectively. we used pay-off table to obtain the range of each objective function. 4.3. stem method stem method which was first proposed by benayoun et al., 1971 to interactively solve an momp, is revisited here briefly. step 1. construct a pay-off table using single objective optimization problems. therefore zj *, j=1,…,n are the optimal objective value of single objective models (55)-(56). minimize zj(x), j=1,…,n. (55) subject to: x ∈ ψ (feasible region of problem) (56) step 2. at the , -th cycle, the optimum solution of the model (57)-(60) is sought, which is the nearest to the ideal solution zj *: minimize γ (57) subject to: ( ( ) ) , , ...,z x z j n1*j j j #t c= (58) x x∈  (59) 0$c (60) where x  includes ψ plus any constraint added in the previous 1− cycles, ρj gives the relative importance of the distances to the optimal solution, and γ is the associated mini-max variable. if fj max and fj min be the maximum and minimum values of objective j; then ρj can be determined using (61). , , ..., z z z z z z j n1 max max min max max min j j j j j j j j n 1 t = = = | (61) step 3. the compromise solution x  is presented to the decision maker (dm). if some of the objectives are satisfactory and others are not, the dm relaxes a satisfactory objective jz  enough to allow an int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 96 http://creativecommons.org/licenses/by-nc-nd/4.0/ improvement of the unsatisfactory objectives, called the set u, in the next iterative cycle (azaron et al., 2008). the dm gives ∆zj as the amount of acceptable relaxation. for the next cycle the feasible region is modified using (62). , , , ..., , x x z x z x z j u z x z x j n j u1 l l j j l j j j l 1 " # ! # d= + = + ^ ^ ^ ^ h h h h* (62) the weight , , ..., ,j n j u1j "t = is set to zero and the calculation phase of cycle 1+ begins. the application of stem method to solve the problem (13)-(41) results in model (63)-(68). minimize γ (63) subject to: constrains (21)-(41) (64) , , ,z z j 1 2 3≤*j j j 6t c=_ i (65) , , , ,j u j1 2 3j u u j 6 6 !t b b b = + = (66) , ,, jz z z 1 2 3max max min j j j j 6b = = (67) γ ≥ 0 (68) z1 *,z2 *, and z3 * are achieved using pay-off table. in order to gather relative importance of objective function the procedure by barzilai, 1997 is used. 5. model implementation the proposed model (13)-(41) is coded and implemented in lingo software. we used the goal attainment technique (45)-(47), modified ε-constraint (51)-(54) and stem methods (63)-(68) to solve the proposed model (13)-(41), separately. results of the three methods are compared. moreover, the experimental model that was used in the factory was also considered. 5.1. a real-world industrial case study in order to illustrate the applicability and efficacy of proposed methodology, the proposed model is applied in teiph-saipa company. teiph-saipa paint and resin industries company was established in 1967 under the license from denmark dyrup and under the name of dyrup iran. with around fifty years’ experience in manufacturing different kinds of paint an resin for automotive industry, industrial and construction paints and coatings, and having the world’s most recent technologies and formulations, this company is working as one of the subsidiaries of saipa group. the products of teiph-saipa company are mainly distributed throughout iran and middleeast. this company owns two customer centers located in two different cities. the main customer of this company is saipa as the second largest iranian automotive manufacturer. raw materials are supplied from three suppliers located in italia, korea, and germany. recently, the company has faced with several issues and problems such as decrease of customer satisfaction levels, financial problems, and high total costs. we are going to apply the proposed models in teiph-saipa company. the following assumptions are considered for planning in teiph-saipa company. the planning horizon consists of six months. there are three suppliers and two customers. there are two family groups of products. aggregate unit of production is ton. demand can be either satisfied or backordered. two working shifts are considered in a day. regular production time is 8 hours per shift and overtime production is approximately 3 hours per shift. to produce the products, 24 types of raw materials are required. four skill levels are considered for workers as low (k=1), medium (k=2), good (k=3) and high (k=4). there are 3 types of training (l=1,2,3.). repairs are done just in shift 2 (i.e., overtime). since the filters of reservoirs should occasionally be replaced, inevitable stops are usually occurred during shift 1 (regular times). if the demand of one period is higher than production capacity in regular times and on hand inventory levels also unable to satisfy this demand, the production is continued in overtime. in this planning horizon, the planned purchase order of raw material m from suppliers have been organized for t=3 and t=6. the holding cost of inventory is low during the planning horizon. five operators are working in each site. maximum available budget is 696970 (dollar) over the planning horizon. based on historical data of sale department, the future economic scenarios is estimated as; (i) boom, (ii) good, (iii) fair, and (iv) poor, with associated probabilities of 0.40, 0.30, 0.20 and 0.10, respectively. int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 97 http://creativecommons.org/licenses/by-nc-nd/4.0/ in this case, allowable purchase (πm) for all raw materials and allowable sale (ζi) for all products are constant values 0.001 and 0.01, respectively. also, other deterministic parameters are set as δ=0.125, τ=0.3, ol=0.5, and hr=10. according to the different economic scenarios, the available budgets for each scenario are shown in table 6. table 6. available budgets data (dollar). parameter scenario boom; n=1 good; n=2 fair; n=3 poor; n=4 ωn 696970 666700 636400 606000 the backordering costs for different products in each customer center are shown in table 7. the transportation costs from factory to each customer center and demand data are shown in table 8. 5.2. computational results in this sub-section, the results of the application of the proposed model and solution procedures are presented for the case study. after the data collection, essential information are summarized and reported table 7. backordering costs data; cbnict (dollar/unit). period item customer scenario 1 2 3 4 5 6 1 1 boom 272.7 272.7 278.8 227 454.5 281.8 good 257.6 257.6 263.6 212 439 266.6 fair 242.4 242.4 248.5 197 424 251.5 poor 227 227 233 182 409 236 2 boom 182 182 172.7 227 69.7 169.7 good 167 167 157.5 212 54.5 154.5 fair 151.5 151.5 142 197 39 139 poor 136 136 127 182 24 124 2 1 boom 303 203 69.7 151.5 100 109 good 288 188 54.5 136 84.8 94 fair 272.7 172.7 39.5 121 69.7 78.7 poor 257.6 157.5 24 106 54.5 63.6 2 boom 69.7 100 233 151.5 203 194 good 54.5 84.8 218 136 187.8 178.7 fair 39.4 69.7 203 121 172.7 163.6 poor 24 54.5 187.8 106 157.5 148.5 table 8. transportation costs; t2nict (dollar/unit), and demand data; de n ict (ton). item customer scenario t2nict period de n ict period 1 2 3 4 5 6 1 2 3 4 5 6 1 1 boom 151.5 454.5 454.5 545 454.5 606 6 6 8 11 9 10 good 121 424 424 515 424 575.7 5 5 7 10 8 9 fair 91 394 394 484.8 394 545 4 4 6 9 7 8 poor 60.6 363.6 363.6 454.5 363.6 515 3 3 5 8 6 7 2 boom 212 424 424 636 212 848 4 4 5 11 3 6 good 181.8 394 394 606 181.8 818 3 3 4 10 2 5 fair 151.5 363.6 363.6 575.7 151.5 787.8 2 2 3 9 1 4 poor 121 333 333 545 121 757.6 1 1 2 8 0 3 2 1 boom 151.5 454.5 454.5 545 454.5 606 4 10 3 7 3 4 good 121 424 424 515 424 575.7 3 9 2 6 2 3 fair 91 394 394 484.8 394 545 2 8 1 5 1 2 poor 60.6 363.6 363.6 454.5 1.2×107 515 1 7 0 4 0 1 2 boom 212 424 424 636 212 848 3 5 7 7 4 7 good 181.8 394 394 606 181.8 818 2 4 6 6 3 6 fair 151.5 363.6 363.6 575.7 151.5 787.8 1 3 5 5 2 5 poor 121 333 333 545 121 757.6 0 2 4 4 1 4 int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 98 http://creativecommons.org/licenses/by-nc-nd/4.0/ in the form of tables 6-8. then, proposed model is developed using these data. therefore, a multiobjective stochastic mathematical model is formed and is solved using three solution approaches, separately. also, the amount of the first-stage and the second-stage variables of the two-stage stochastic programming model was determined using three solution approaches. the proposed model was analyzed by all three methods (goal attainment, modified ε-constraint, and stem) in four different scenarios (boom, good, fair and poor) and its results were discussed. finally, stochastic programming model are analyzed using the results of the best solution procedure and the results are presented. also, the methods are prioritized due to the results and the most effective method is selected to solve the problem. lingo software is used to code and solve the proposed stochastic app problem. for all three methods, the first objective function is considered as the main objective function. the preference of dm on priority of objective functions are asked in form of weights of goals: w1=0.66, w2=0.14, and w3=0.20. in the beginning, we make the pay-off table, which is shown in table 9. according to payoff table, minimum and maximum values for objectives are determined as z1∈ [576367.27, 697000], z2∈ [0, 1428], and z3∈ [0, 1515.15]. the total run time and iteration of stochastic mathematical model are 1146(s) and 47077, respectively. the results of goal attainment technique (ga), modified ε-constraint method (mε-c ), and stem method are presented in table 10, table 11, and table 12, respectively. the amount of the first-stage variables (fsv) and second-stage variables (ssv) of the two-stage stochastic programming model for the app problem are shown in table 13. the second-stage variables in goal attainment method have the lowest amount in comparison with other methods. the total cost (z1), the deviation of delivery time and due dates (z2), and the downside risk (z3) of all scenarios for all solution methods are presented in table 14. also, the total computational time and iterations to solve the problem are shown in table 14. table 9. the payoff table of stochastic programming problem. optimal solution z1 (dollar) z2 (hours) z3 (dollar) z1 * 576367.27 718.4 1515.15 z2 * 633167.57 0 151.28 z3 * 697000 1428 0 goal value (gv) 576367.27 0 0 table 10. the results of goal attainment technique (ga). weights objective values γ cpu time (second)w1 w2 w3 z1 z2 z3 0.66 0.14 0.20 576367.27 288.21 2.03e-2 3353.67 1053 table 11. the results of modified ε-constraint method (mε-c) right-hand sides slack variables objective values θ φ cpu time (second)ε2 ε3 s2 s3 z1 z2 z3 1428 4.1123×10-8 1312.83 0 576367.27 115.16 4.11×10-8 10-6 578551.81 1049 table 12. the results of stem method. relative importance objective values γ cpu time (second)ρ1 ρ2 ρ3 z1 z2 z3 0.08 0.46 0.46 576367.27 288.21 4.29×10-10 188.95 1063 table 13. first and second-stage variables of stochastic programming. methods fsv ssv goal attainment 575697.88 699.34 modified ε-constraint 575697.88 2853.94 stem 575697.88 699.37 int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 99 http://creativecommons.org/licenses/by-nc-nd/4.0/ since the problem has multiple objectives, the appropriate solutions are selected by decision makers regarding to the different scenarios and the importance of each goal in desired scenario, the amount of objective functions in each method and each scenario are shown in table 14. it can be concluded from table 14 that the results of stem method and modified ε-constraint method during boom scenario is the same, approximately. but in other scenarios, the results for each method are different. so, decision makers select the effective solution with regard to his/her attitude toward nature states. also, the run time, total number of iterations, total number of variables, and total number of iterations are approximately equal for all solution approaches. the resultant model size for each solution approach is presented in table 15. 5.3. comparing the results we compare the results of proposed methods for all scenarios in figure 1. with regard to the future economic scenarios and the results that obtained in table 14 and figure 1, decisions are made as following scenarios based on attitude of decision maker toward status of the nature. in figure 1, the objective function values of solution approaches for different scenarios are presented, separately. according to the figure 1, for all scenarios; (i) the best solution for minimizing the expected total cost of supply chain is the goal attainment method, (ii) the best solution for minimizing the deviation of delivery time and due dates is the modified ε-constraint method, and (iii) the best solutions for minimizing the supply chain downside risk are modified ε-constraint and stem methods. therefore, adaptation between the results of solutions and company’s strategic is very important for decision makers. the best solutions found using all approaches are presented compared with existing empirical model (em) and goal value (gv) in table 16. they are also depicted in figure 2. in table 16, the results of all solution procedures are compared. since the stochastic app problem is a multi-objective problem with conflicting objectives, the results of the value of objective functions for each method are also proposed. as previously table 14. the results of solution procedures for stochastic programming model. method scenarios goal attainment modified ε-constraint stem z1 boom 575990.90 576430.30 576430.30 good 575898.48 580733.03 576365.75 fair 575819.39 579582.72 576304.54 poor 575752.72 578432.42 576244.54 z2 boom 115.16 115.16 115.16 good 86.47 0 86.47 fair 57.69 0 57.69 poor 28.88 0 28.88 z3 boom 4.94×10-2 0 0 good 1.16×10-3 5.60×10-8 9.87×10-10 fair 4.64×10-4 9.54×10-8 7.14×10-10 poor 9.28×10-4 5.22×10-8 7.24×10-10 run time (s) 1053 1049 1063 iterations 39414 39380 41079 table 15. model size for the case study problem. goal attainment modified ε-constraint stem number of variables total 907 929 907 nonlinear 188 184 190 integers 114 114 114 number of non-zeros total 6873 8133 7735 nonlinear 474 468 474 number of constraints total 1125 1327 1124 nonlinear 13 7 10 int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 100 http://creativecommons.org/licenses/by-nc-nd/4.0/ stated, the objective value of each solution approach is different; (i) in goal attainment method, the attainment element value (the term υ) is 3353.67, (ii) the value of objective function in modified ε-constraint method (φ) is 578551.81, (iii) the value of the associated mini-max variable (γ) in the stem method is 188.95. these are shown in column “objective value” in table 16. the value of z1, z2, and z3 are the objective functions of original supply chain network design problem. it can be concluded form content of table 16 that the achieved solution by proposed methods dominates the empirical model currently are used in the company. as mentioned, the proposed methods, empirical model, and goal value are also compared in figure 2. figure 2 shows that the total costs of supply chain, the deviation of delivery time and due dates, and the supply chain downside risk are all better in the solutions proposed by the goal attainment (ga), modified ε-constraint (mε-c), and stem methods. given the figure 2, for the first objective (total costs of supply chain) the results of goal attainment (ga) 572000   574000   576000   578000   580000   582000   boom  good   fair   poor   goal  a5ainment   modified  ε-­‐constraint   stem   0   50   100   150   boom  good   fair   poor   goal  a1ainment   modified  ε-­‐constraint   stem   0,00e+00   1,00e-­‐03   2,00e-­‐03   3,00e-­‐03   4,00e-­‐03   5,00e-­‐03   boom  good   fair   poor   goal  a8ainment   modified  ε-­‐constraint   stem   (a) total costs of supply chain (dollar) (b) deviation of delivery time and due dates (hours) (c) supply chain downside risk (dollar) figure 1. comparing the objective function values for different scenarios. table 16. general comparison of results. methods solution procedure objective functions objective value z1 z2 z3 proposed model goal attainment (ga) 3353.67 576367.27 288.21 2.03e-2 modified ε-constraint (mε-c) 578551.81 578551.81 115.16 4.11e-8 stem 188.95 576367.27 288.21 4.29e-10 the empirical model (em) 668585.75 409.26 987 goal value (gv) 576367.27 0 0 575000   575500   576000   576500   577000   577500   578000   578500   579000   ga   m εc     st em     gv   0   50   100   150   200   250   300   ga   m εc     st em   gv   4,00e-­‐15   5,00e-­‐11   1,00e-­‐10   1,50e-­‐10   2,00e-­‐10   2,50e-­‐10   3,00e-­‐10   3,50e-­‐10   4,00e-­‐10   g a   m εc     st em   g v   (a) total costs of supply chain (dollar) (b) deviation of delivery time and due dates (hours) (c) supply chain downside risk (dollar) figure 2. comparing the achieved objective functions. int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 101 http://creativecommons.org/licenses/by-nc-nd/4.0/ and stem methods are recommended. for the second objective (deviation of delivery time and due dates) the results of modified ε-constraint (mε-c) method are suggested. also, for the third objective (supply chain downside risk) all methods have good results, but we propose stem, mε-c, and goal attainment (ga), respectively due to increasing order of third objective function reported in table 16. the solution approaches are prioritized due to the obtained results for multi-objective stochastic programming model in table 17. for each objective functions, methods are prioritized. the stem method provides a production plan with the lowest total production cost, and lowest down-side risk among the other approaches. also, the best satisfaction level is achieved by modified ε-constraint method. in order to achieve an appropriate solution approach, the decision-maker needs to see this range of outcomes, to be able to trade-off one goal against the other in terms of the results. so, by solving the proposed stochastic app problem, it is concluded that the relationship between total costs of supply chain, deviation of delivery time and due dates, and supply chain downside risk is not clear and it is not possible to easily define a utility function. that is why three multi-objective approaches (i.e., ga, mε-c, and stem) were adopted to this stochastic app problem. indeed, these methods were selected from three major categories (the a priori, the a posteriori, and the interactive methods) to better understand the problem and to be able to cover the broader area of solutions for decision makers. it seems that by increasing scenarios, the results of stem and modified ε-constraint methods are to be similar, approximately. also, by increasing the goal for each objective, a wider space for other objectives can be achieved. it is notable that although several solution procedures were applied on the proposed problem and several solutions were achieved, but the final decision is highly related to the attitude of decision maker toward state of the nature. 5.4. research findings optimistic manager. an optimistic decision maker expects for the boom scenario. under the boom scenario; and based on relative importance of objective functions in our case study (w1=0.66, w2=0.14, and w3=0.20) the total cost of supply chain has the best value in the proposed goal attainment method. so, the production plan of the supply chain is aligned with the outputs of goal attainment method in this case. it is notable that, under the boom scenario, the results of the second objective function, for all methods are similar. the third objective function of goal attainment method, under the boom scenario, is also tolerable in comparison with the other methods. finally, it can be conclude that, under boom scenario, the goal attainment method is preferred. neutral manager. a neutral decision maker gives expects for the fair scenario. again, under the fair scenario; and based on relative importance of objective functions in our case study (w1=0.66, w2=0.14, and w3=0.20) the total cost of supply chain has the best value in the proposed goal attainment method. although, it is notable that, for the second objective function under the aforementioned situation, the modified modified ε-constraint method presents the best results. it is notable that, under the fair scenario, the results of the third objective function, for all methods are approximately similar. finally, it can be concluded that, under fair scenario, and considering the great relative importance of first objective function, the goal attainment suggests the most suitable production plan. pessimistic manager. a neutral decision maker gives expects for the poor scenario. again, under the poor scenario; and based on relative importance of objective functions in our case study (w1=0.66, w2=0.14, and w3=0.20) the total cost of supply chain has the best value in the proposed goal attainment method. although, it is notable that, for the second objective function under the aforementioned situation, again the modified modified ε-constraint method presents the best results. it is notable that, under the poor scenario, the results of the third objective function, for all methods are approximately similar. finally, it can be concluded that, under poor scenario, and considering the great relative importance of first objective function, the goal attainment suggests the most suitable production plan again. the following points are achieved based on implementation of proposed approach in real case study. all production plans, provided by proposed methods of this study, dominate the existing experimental method of planning the case study. int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 102 http://creativecommons.org/licenses/by-nc-nd/4.0/ no hiring and firing is suggested for the next six periods. the overtime-shift is just used when the regular time cannot satisfy the demands, or handling the training programs. the model suggests training for some periods of planning. the cross-functional quality teams are formed in order to learn new techniques. the performance and experiments of workers are shared during implementation phase. the employee engagement with the new production plan is measured frequently to feedback the success of plan. all workers are trained based on their performance records in order to enhance the learning curves which influences the production time and quality. the customer satisfaction level is measured frequently through a survey questionnaire about quality of delivery, quality of product, and time of delivery. the above issues will help the execution team to check whether the plan is implemented correctly or a deviation is occurred. the main results include: (1) the huge costs of hiring and firing were reduced through a proper plan proposed by the approach of this study, (2) although the training courses imposed costs to company, but its positive effects on the timely delivery of products and learning curves of workers was illustrated, (3) the total cost of the system was increasingly reduced in comparison with the experimental production plan, (4) the satisfaction degree of customers was increased amazingly, (4) the down-side risk of supply chain was decreased through proposed approach. 6. conclusions remarks in this paper a new aggregate production planning problem in supply chain considering financial risk, customer satisfaction, and training was proposed. the expected total cost as the main objective function, the customer satisfaction level, and the downside risk of supply chain were considered as objective function, simultaneously. several constraints regarding workforce level, inventory level, production capacity, backorder, and risk were also considered. the proposed problem was modeled using mixed integer multi-objective mathematical model. then, a new multi-objective stochastic optimization approach was developed to handle this problem. the uncertainty of real-world problem was represented by a set of discrete scenarios with given probability of occurrence. a priori method, a posteriori method, and an interactive method were developed and adapted to solve the proposed multiobjective stochastic mathematical model. a real case study of color and resin company called teiph-saipa, which produced several products was selected as a practical environment in order to test the suitability and applicability of proposed model and solution approaches. the proposed models were coded in lingo software. the results of three methods were compared and analyzed. the results of all approaches were compared with those of experimental method in this case study. all methods dominate the existing experimental method. moreover, some decision making scenarios based on decision maker’s attitude toward uncertainty in nature was developed and discussed. the proposed model of this study can be improved by adding constraints on skill level of workers and their learning curves. the outsourcing option can be considered in the modelling procedure, since many companies may outsource some of the operations. the effects of inflation and discounts may be considered the longterm planning. the failure of facilities and machines, and the rate of products returned from customers can also be considered in future research works. references abdelaziz, f.b. (2012). solution approaches for the multi-objective stochastic programming. journal of the operational research society, 216(1): 1–16. https://doi.org/10.1016/j.ejor.2011.03.033 azaron, a.k.n., brown, s.a., modarres, m. (2008). a multi-objective stochastic programming approach for supply chain design considering risk. international journal of production economics, 116(1): 129-138. https://doi.org/10.1016/j.ijpe.2008.08.002 int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 103 https://doi.org/10.1016/j.ejor.2011.03.033 https://doi.org/10.1016/j.ijpe.2008.08.002 http://creativecommons.org/licenses/by-nc-nd/4.0/ azaron, a., furmans, k., modarres, m. (2010). new developments in multiple objective and goal programming lecture notes in economics and mathematical systems; multi-objective stochastic programming approaches for supply chain management. springer berlin heidelberg, 638: 1-14, doi: 10.1007/978-3-642-10354-4_1. https://doi.org/10.1007/978-3-642-10354-4_1 ahmed, s. (2010). two-stage stochastic integer programming: a brief introduction. wiley encyclopedia of operations research and management science. 14 january, https://doi.org/10.1002/9780470400531.eorms0092 benayoun, r., de montgolfier, j., tergny, j., larichev, o. (1971). linear programming with multiple objective functions: step method (stem). mathematical programming, 1(2): 366-375. https://doi.org/10.1007/bf01584098 bushuev, m. (2014). convex optimisation for aggregate production planning. international journal of production research, 52(4): 10501058. https://doi.org/10.1080/00207543.2013.831998 barbarosoglu, g., arda, y. (2004). a two-stage stochastic programming framework for transportation planning in disaster response. journal of the operational research society, 55: 43–53. https://doi.org/10.1057/palgrave.jors.2601652 barzilai,.j. (1997). deriving weights from pairwise comparison matrices. journal of the operational research society, 48(12): 1226-1232. https://doi.org/10.1057/palgrave.jors.2600474 corominas, a., lusa, a., olivella, j. (2012). a manufacturing and remanufacturing aggregate planning model considering a non-linear supply function of recovered products. production planning & control: the management of operations, 23(2–3): 194-204. https://doi. org/10.1080/09537287.2011.591651 chankong, v., haimes, y.y. (1983). multi-objective decision making: theory and methodology. new york: north-holand. dantzig, g.b. (1955). linear programming under uncertainty. management science, 1(3-4): 197-206. https://doi.org/10.1287/mnsc.1.34.197 dupačová, j. (2002). applications of stochastic programming: achievements and questions. european journal of operational research, 140(2): 281-290. https://doi.org/10.1016/s0377-2217(02)00070-x giannoccaro, i., pontrandolfo, p. (2001). models for supply chain management: a taxonomy. proceedings of the production and operations management. conference poms mastery in the new millennium, orlando, fl, usa. gunasekaran, a., goyal, s.k., martikainen, t., yli-olli, p. (1998). production capacity planning and control in multi-stage manufacturing. journal of the operational research society, 49: 625-634. https://doi.org/10.1057/palgrave.jors.2600580 gembicki, f.w., haimes, y.y. (1975). approach to performance and sensitivity of multi-objective optimization: the goal attainment method. automatic control, ieee transactions, 20(6): 769–771. https://doi.org/10.1109/tac.1975.1101105 hafezalkotob, a., khalili-damghani, k. (2015). development of a multi-period model to minimise logistic costs and maximise service level in a three-echelon multi-product supply chain considering back orders. international journal of applied decision sciences, 8(2): 145-163. https://doi.org/10.1504/ijads.2015.069599 hafezalkotob, a., khalili-damghani, k., ghashami, s.s. (2017). a three-echelon multi-objective multi-period multi-product supply chain network design problem: a goal programming approach. journal of optimization in industrial engineering, 10(21): 67-78. https://doi. org/10.22094/joie.2016.262 han, s.h., dong, m.y., lu, s.x., leung, s.c.h., lim, m.k. (2013). production planning for hybrid remanufacturing and manufacturing system with component recovery. journal of the operational research society, 64: 1447–1460. https://doi.org/10.1057/jors.2012.106 haimes, y.y., lasdon, l.s., wismer, d.a. (1971). on a bicriterion formulation of the problems of integrated system identification and system optimization. ieee transactions on systems, man, and cybernetics, 1(3): 296–297. holt, c.c., modigliani, f., simon, h.a. (1955). a linear decision rule for production and employment scheduling. management science, 2(1): 1–30. https://doi.org/10.1287/mnsc.2.1.1 holt, c.c., modigliani, f., muth, j.f., simon, h.a. (1961). planning production inventories and workforce. the american economic review, 51(4): 697-699. hwang, c.l., masud, a.s.m. (1979). multiple objectives decision making methods and applications. springer-verlag, new york. https:// doi.org/10.1007/978-3-642-45511-7 kazemi zanjani, m., nourelfath, m., ait-kadi, d. (2013). a scenario decomposition approach for stochastic production planning in sawmills. journal of the operational research society, 64: 48-59. https://doi.org/10.1057/jors.2012.28 kazemi zanjani, m., nourelfath, m., ait-kadi, d. (2011). production planning with uncertainty in the quality of raw materials: a case in sawmills. journal of the operational research society, 62: 1334-1343. https://doi.org/10.1057/jors.2010.30 karabuk, s. (2008). production planning under uncertainty in textile manufacturing. journal of the operational research society, 59: 510520. https://doi.org/10.1057/palgrave.jors.2602370 kira, d., kusy, m., rakita, i. (1997). a stochastic linear programming approach to hierarchical production planning. journal of the operational research society, 48: 207–211. https://doi.org/10.1057/palgrave.jors.2600340 kogan, k., portougal, v. (2006). multi-period aggregate production planning in a news-vendor framework. journal of the operational research society, 57: 423-433. https://doi.org/10.1057/palgrave.jors.2602002 kall, p., wallace, s.w. (2003). stochastic programming (second edition). john wiley and sons: chichester. int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 104 https://doi.org/10.1007/978-3-642-10354-4_1 https://doi.org/10.1002/9780470400531.eorms0092 https://doi.org/10.1007/bf01584098 https://doi.org/10.1080/00207543.2013.831998 https://doi.org/10.1057/palgrave.jors.2601652 https://doi.org/10.1057/palgrave.jors.2600474 https://doi.org/10.1080/09537287.2011.591651 https://doi.org/10.1080/09537287.2011.591651 https://doi.org/10.1287/mnsc.1.3-4.197 https://doi.org/10.1287/mnsc.1.3-4.197 https://doi.org/10.1016/s0377-2217(02)00070-x https://doi.org/10.1057/palgrave.jors.2600580 https://doi.org/10.1109/tac.1975.1101105 https://doi.org/10.1504/ijads.2015.069599 https://doi.org/10.22094/joie.2016.262 https://doi.org/10.22094/joie.2016.262 https://doi.org/10.1057/jors.2012.106 https://doi.org/10.1287/mnsc.2.1.1 https://doi.org/10.1007/978-3-642-45511-7 https://doi.org/10.1007/978-3-642-45511-7 https://doi.org/10.1057/jors.2012.28 https://doi.org/10.1057/jors.2010.30 https://doi.org/10.1057/palgrave.jors.2602370 https://doi.org/10.1057/palgrave.jors.2600340 https://doi.org/10.1057/palgrave.jors.2602002 http://creativecommons.org/licenses/by-nc-nd/4.0/ karmarkar, u.s., rajaram, k. (2012). aggregate production planning for process industries under oligopolistic competition. european journal of operational research, 223: 680–689. https://doi.org/10.1016/j.ejor.2012.07.007 khalili-damghani, k., shahrokh, a. (2014). solving a new multi-period multi-objective multi-product aggregate production planning problem using fuzzy goal programming. industrial engineering & management systems, 13(4): 369-382. https://doi.org/10.7232/ iems.2014.13.4.369 khalili-damghani, k., tavana, m., santos-arteagad, f. j., ghanbarzad-dashti, m. (2017). a customized genetic algorithm for solving multiperiod cross-dock truck scheduling problems. measurement, 108: 101–118. https://doi.org/10.1016/j.measurement.2017.05.027 khalili-damghani, k., ghasemi, p. (2016). uncertain centralized/decentralized production-distribution planning problem in multi-product supply chains. industrial engineering & management systems, 15(2): 156-172. https://doi.org/10.7232/iems.2016.15.2.156 khalili-damghani, k., tajik-khaveh, m. (2015). solving a multi-objective multi-echelon supply chain logistic design and planning problem by a goal programming approach. international journal of management science and engineering management, 10(4): 242-252. https:// doi.org/10.1080/17509653.2014.975166 khalili-damghani, k., abtahi, a.r., ghasemi, a. (2015). a new bi-objective location-routing problem for distribution of perishable products: evolutionary computation approach. journal of mathematical modelling and algorithms in operations research, 14(3): 287-312. https://doi.org/10.1007/s10852-015-9274-3 khalili-damghani, k., tavana, m., amirkhan, m. (2014). a fuzzy bi-objective mixed-integer programming method for solving supply chain network design problems under ambiguous and vague conditions. the international journal of advanced manufacturing technology, 73(9-12): 1567-1595. https://doi.org/10.1007/s00170-014-5891-7 khalili-damghani, k., naderi, h. (2014). a mathematical location-routing model of repair centres and ammunition depots in order to support soldiers in civil wars. international journal of management and decision making, 13(4): 422-450. https://doi.org/10.1504/ ijmdm.2014.065360 leung, s.c.h., wu, y., lai, k. (2003). multi-site aggregate production planning with multiple objectives: a goal programming approach. production planning and control: the management of operations, 14(5): 425–436. https://doi.org/10.1080/0953728031000154264 leung, s.c.h., wu, y. (2004). a robust optimization model for stochastic aggregate production planning. production planning and control: the management of operations, 15(5): 502–514. https://doi.org/10.1080/09537280410001724287 leung, s.c.h., wu, y., lai, k.k. (2006). a stochastic programming approach for multi-site aggregate production planning. journal of the operational research society, 57: 123-132.s. https://doi.org/10.1057/palgrave.jors.2601988 leung, c.h., lai, k.k., ng, w.l., wu, y. (2007). a robust optimization model for production planning of perishable products. journal of the operational research society, 58: 413-422. https://doi.org/10.1057/palgrave.jors.2602159 liang, t.f. (2007). application of interactive possibilistic linear programming to aggregate production planning with multiple imprecise objectives. production planning and control: the management of operations, 18(7): 548-560. https://doi.org/10.1080/09537280701530033 masud, a.s.m., hwang, c.l. (1980). aggregate production planning model and application of three multiple objective decision methods. international journal of production research ,18(6): 741–752. https://doi.org/10.1080/00207548008919703 mirzapour al-e-hashem, s.m.j., malekly, h., aryanezhad, m.b. (2011). a multi-objective robust optimization model for multi-product multisite aggregate production planning in a supply chain under uncertainty. international journal of production economics, 134(1): 28-42. https://doi.org/10.1016/j.ijpe.2011.01.027 mirzapour al-e-hashem, s.m.j., aryanezhad, m.b., sadjadi, s.j. (2012). an efficient algorithm to solve a multi-objective robust aggregate production planning in an uncertain environment. international journal of advanced manufacturing technology, 58: 765–782. https://doi.org/10.1007/s00170-011-3396-1 mirzapour al-e-hashem, s.m.j., baboli, a., sazvar, z. (2013). a stochastic aggregate production planning model in a green supply chain: considering flexible lead times, nonlinear purchase and shortage cost functions. european journal of operational research, 230: 26–41. https://doi.org/10.1016/j.ejor.2013.03.033 mula, j., poler, r., garcía-sabater, j.p., lario, f.c. (2006). models for production planning under uncertainty: a review. international journal of production economics, 103: 271–285. https://doi.org/10.1016/j.ijpe.2005.09.001 miettinen, k.m. (1998). nonlinear multi-objective optimization. international series in operations research & management science. kluwer academic publishers, boston. https://doi.org/10.1007/978-1-4615-5563-6 mavrotas, g. (2009). effective implementation of the ε-constraint method in multi-objective mathematical programming problems. applied mathematics and computation, 213: 455-465. https://doi.org/10.1016/j.amc.2009.03.037 nam, s.j., logendran, r. (1992). aggregate production planning a survey of models and methodologies. european journal of operational research, 61: 255-272. https://doi.org/10.1016/0377-2217(92)90356-e poojari, c.a., lucas, c., mitra, g. (2008). robust solutions and risk measures for a supply chain planning problem under uncertainty. journal of the operational research society, 59(2–12). https://doi.org/10.1057/palgrave.jors.2602381 petkov, s.b., maranas, c.d. (1998). design of single product campaign batch plants under demand uncertainty. aiche journal, 44(4): 896911. https://doi.org/10.1002/aic.690440415 int. j. prod. manag. eng. (2017) 5(2), 85-106creative commons attribution-noncommercial-noderivatives 4.0 international stochastic multi-period multi-product multi-objective aggregate production planning model in multi-echelon supply chain 105 https://doi.org/10.1016/j.ejor.2012.07.007 https://doi.org/10.7232/iems.2014.13.4.369 https://doi.org/10.7232/iems.2014.13.4.369 https://doi.org/10.1016/j.measurement.2017.05.027 https://doi.org/10.7232/iems.2016.15.2.156 https://doi.org/10.1080/17509653.2014.975166 https://doi.org/10.1080/17509653.2014.975166 https://doi.org/10.1007/s10852-015-9274-3 https://doi.org/10.1007/s00170-014-5891-7 https://doi.org/10.1504/ijmdm.2014.065360 https://doi.org/10.1504/ijmdm.2014.065360 https://doi.org/10.1080/0953728031000154264 https://doi.org/10.1080/09537280410001724287 https://doi.org/10.1057/palgrave.jors.2601988 https://doi.org/10.1057/palgrave.jors.2602159 https://doi.org/10.1080/09537280701530033 https://doi.org/10.1080/00207548008919703 https://doi.org/10.1016/j.ijpe.2011.01.027 https://doi.org/10.1007/s00170-011-3396-1 https://doi.org/10.1016/j.ejor.2013.03.033 https://doi.org/10.1016/j.ijpe.2005.09.001 https://doi.org/10.1007/978-1-4615-5563-6 https://doi.org/10.1016/j.amc.2009.03.037 https://doi.org/10.1016/0377-2217(92)90356-e https://doi.org/10.1057/palgrave.jors.2602381 https://doi.org/10.1002/aic.690440415 http://creativecommons.org/licenses/by-nc-nd/4.0/ rezaeisaray, m., ebrahimnejad, s., khalili-damghani, k. (2016). a novel hybrid mcdm approach for outsourcing supplier selection: a case study in pipe and fittings manufacturing. journal of modelling in management, 11(2): 536-559. https://doi.org/10.1108/jm2-062014-0045 rizk, n., martel, a., d’amours, s. (2008). synchronized production–distribution planning in a single-plant multi-destination network. journal of the operational research society, 59: 90-104. https://doi.org/10.1057/palgrave.jors.2602316 ruszczynski, a., shapiro, a. (2003). stochastic programming; handbooks in operations research and management science. elsevier: amsterdam. sakallı, ü.s.,  baykoç, ö.f., birgören, b. (2010). a possibilistic aggregate production planning model for brass casting industry. production planning & control: the management of operations, 21(3): 319-338. https://doi.org/10.1080/09537280903449438 sha, d.y., che, z.h. (2006). supply chain network design: partner selection and production/distribution planning using a systematic model. journal of the operational research society, 57: 52-62. https://doi.org/10.1057/palgrave.jors.2601949 tavana, m., khalili-damghani, k., santos-arteaga, f.j., zandi, m.h. (2017). drone shipping versus truck delivery in a cross-docking system with multiple fleets and products. expert systems with applications, 72(15): 93–107. https://doi.org/10.1016/j.eswa.2016.12.014 tahmasebi, m., khalili-damghani, k., ghezavati, v. (2017). a customized bi-objective location-routing problem for locating post offices and delivery of post parcels. journal of industrial and systems engineering, 10: 1-17. techawiboonwong, a., yenradee, p. (2010). aggregate  production  planning  with workforce transferring  plan  for multiple product types. production planning & control: the management of operations, 14(5): 447-458. https://doi.org/10.1080/09537280310001597343 toptal, a., koc, u., sabuncuoglu, i. (2012). a joint production and transportation planning problem with heterogeneous vehicles. journal of the operational research society, 65: 180-196. https://doi.org/10.1057/jors.2012.184 wang, h.f., zheng, k.w. (2013). application of fuzzy linear programming to aggregate production plan of a refinery industry in taiwan. journal of the operational research society, 64: 169-184. https://doi.org/10.1057/jors.2012.12 yan, h.s., wan, x.q., xiong, f.l. (2014). integrated production planning and scheduling for a mixed batch job-shop based on alternant iterative genetic algorithm. journal of the operational research society, 66(8): 1250-1258. https://doi.org/10.1057/jors.2014.88 int. j. prod. manag. eng. (2017) 5(2), 85-106 creative commons attribution-noncommercial-noderivatives 4.0 international khalili-damghania, k., shahrokh, a. & pakgohar, a. 106 https://doi.org/10.1108/jm2-06-2014-0045 https://doi.org/10.1108/jm2-06-2014-0045 https://doi.org/10.1057/palgrave.jors.2602316 https://doi.org/10.1080/09537280903449438 https://doi.org/10.1057/palgrave.jors.2601949 https://doi.org/10.1016/j.eswa.2016.12.014 https://doi.org/10.1080/09537280310001597343 https://doi.org/10.1057/jors.2012.184 https://doi.org/10.1057/jors.2012.12 https://doi.org/10.1057/jors.2014.88 http://creativecommons.org/licenses/by-nc-nd/4.0/