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), Scholz- Reiter 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 (Scholz- Reiter, 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 load- orientation 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 set- up, 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 queue- length 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/S0007- 8506(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:10.1007/978-3-540- 76862-3_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, 317- 320. 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-to- order 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:10.1007/978-3-540- 77046-6_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/