pa ge 1 pa ge 38 american journal of smart technology and solutions (ajsts) determinants of teff production in north showa zone, central highlands of ethiopia tadele anagaw zewdu1* volume 2 issue 1, year 2023 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: february 28, 2023 accepted: march 16, 2023 published: march 18, 2023 using primary data collected from 115 households of north showa zone, ethiopia, the study examined determinants of teff production. in order to investigate the effect of each predictor variable on the household teff production level a bivariate analysis was performed. among the econometric method of analysis, a logistic regression model was fitted to analyze the potential variables affecting household teff production level in the study area. the result of the descriptive analysis revealed that about 85 (73.90%) were teff producers while 30 (26.10%) of the households were found to be non-producer. moreover, the logistic regression model estimates that among the nine variables included in the logistic model, eight of them were significant at different probability level (1, 5, and 10). these are the education of household head, total cultivated land for teff production per hectare, number of oxen, technology adoption, access of extension services received by households, sex, fertilizer application and kind of teff variety used. finally, improving land quality, creating awareness towards importance of farm technology adoption, and providing frequent extension service were recommended. keywords determinants, logit,teff, teff production, ethiopia 1 salale universitys, fiche 245, ethiopia * corresponding author’s e-mail: tadeleanagaw21@gmail.com introduction ethiopian agriculture is virtually small-scale, subsistenceoriented, and depends on rainfall (anderson, 2007). it is also kicked off, with high population pressure and traditional farming systems have caused ecosystem degradation in the form of soil erosion and declining soil fertility and erratic climate are the challenges to production. additionally, the current smallholder farming systems are undergoing a reverse transformation in which farm sizes are declining, few farmers are moving out of agriculture, and instead are diversifying into non-farm activities from a small farm base (ata, 2013). on the other hand, agriculture is the main source of livelihood for a large proportion of the population, especially for the people residing in the rural areas (schmidt and kedir, 2009). teff is the major staple food crop to most ethiopian people living in the highlands, comprising more than 65% of the population. however, the national average yield of teff is very low and 1.4 tons per hectare and the development of high-yielding cultivars would be very beneficial (csa, 2013). hence, the need for improved crop varieties that are high yielding and with the capacity to survive in such a degraded and risk-prone environment is important (spielman, 2008). there is still a question of yield stagnation due to the low yield potential of the existing teff varieties and other determinants (tareke et al., 2008). hence, the main question of this study was to identify the major determinants of teff production in grar-jarso district and provide relevant information to the concerned body. data source and sampling two-stage probability sampling technique was employed to select the sample of farm households. in the first stage, sample kebeles1 were selected randomly. then, in the second stage, using the list of farm households living in each of the selected kebeles as a sampling frame, sample households were selected randomly using a probability proportional to size sampling technique. hence, this study is based on a primary data collected from a sample of randomly selected 115 smallholder farm households. the data was collected through face-to-face interview using semi-structured questionnaire. estimation strategy the logit model was the appropriate econometric model to identify the determinants of teff production in the study area. this model was chosen; it has an advantage in revealing the relative influence of the probability of teff production through different input utilization. logit model, which has a discrete part, is appropriate which handles the probability of the extent of production in a proper way. logit model which helps to test the determinants of teff production can mathematically be specified as follows: pi= e (y=1 such that; xi = β0+βi xi....... (1) where; y=1 implies the given farmer participates in teff production xi= a vector of independent variable β0= the constant term βi= i= 1, 2 ...n. are the coefficient of independent variable to be estimated 1 the lower class of district where; zi=β0+βi xi ifpi, is the probability of being producer and (1-pi) the probability of being a non producer of teff https://journals.e-palli.com/home/index.php/ajsts mailto:tadeleanagaw21%40gmail.com?subject= pa ge 39 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 38-42, 2023 therefore, we can write this equation as; laterally, pi/( 1-pi ) is the odds ratio of producer farmers with the ratio of the probability that a given farmer can participate in production to the probability that the farmer who will be participating in production. then, if we take the natural logarithm of equation (e), we obtain; if the disturbance term ui is taken in to account, the logit model becomes where β0 is an intercept and β1, β2, … βn are slopes of the equation in the model, and x is the vector of relevant farmer characteristics. consequently, li , which is the log of odds ratio, is called logit or logit model (gujarati, 2004). hence, the above logit model is employed to estimate the effect of the hypothesized explanatory variables on the teff production decision of farmers. results and discussions this analysis is based on data obtained from the questionnaire survey. the questionnaires of 115 households had been examined for incorrectness and missing data were grouped (classified) into two groups, namely, teff producer and none producer groups. the data presented in the following part explains the distinction between the two groups of households. this section briefly presents the major determinants of teff production, the relationship of predictor variables with the household outcome variable, and the econometric model analysis in the study area. descriptive statistics the study found that among 115 sample households, the number of male-headed households and female headed households are found to be 82 and 33 in number and covered 71.3% and 28.70%, respectively. out of 85 producer households, 70 were male and 15 were female and out of 30 non-producer household, 12 were male and 18 were female. out of 115 sample households, the number of illiterate household heads, grade 1-8, grade 9-10, and grade 11-12 household heads are found to be 47%, 20%, 18.3% and 14.8% respectively. moreover out of 30 producer households, that 29 which covers 53.7% are illiterate, 1 which covers 4.3% are grade1-8, 0 which covers 0% are grades 9-10 and 0 which covers 0% are grades 11-12. from the total 115 sample households, 30 were technology adopters. from that, 27 which covered 90% were both adopters and producers but 3 which cover 10% were adopters but non producers. and out of the total sample household, 37 respondents were moderate adopters of the technology. from that 36 which covered 97.3% were both adopter and producers, but out of this 1 which covered adopters, 1 which covers 2.7% was adopter but not producer. from the total 115 households, 70 respondents were getting extension service and both of them are producers and there is no teff producer that got extension access (see table1 below). out of the total, 93.8% of the respondents were growing local teff variety which is recycled from year to year and those farmers who used the improved teff variety were producers. the finding indicates a significant difference in teff variety utilization between producer and non-producer groups at the 1 percent probability level of significance. consistently, out of the total 115 table 1: summary of descriptive statistics for dummy/categorical explanatory variables variable catagories producer % non-producer % x2 (p-value) sex male female 85.4 45.5 14.6 54.57 19.43*** level of educational illiterate grade 1-8 grade 9-10 grade 11-12 46.3 95.7 100 100 53.7 4.3 0 0 40.409*** technology adoption adopter moderatly adopter low adopter non-adopter 90 97.3 100 3.7 10 2.7 0 96.3 90.957*** access to extension services yes no 100 33.3 066.7 63.37*** teff variety local improved both 93.8 100 100 6.2 0 0 1.053*** fertilizer application yes no 98.8 17.1 1.2 82.9 84.095*** note: *** significant at 1% probability level, source: own survey results 2021 https://journals.e-palli.com/home/index.php/ajsts pa ge 40 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 38-42, 2023 households, 98.8% of smallholder farm households were used fertilizer for their cropping purpose. the finding indicates a significant difference in fertilizer application between producer and non-producer groups at the 1 percent probability level of significance. the land holding of all sample households ranges from 0 hectares to 5 hectares. the mean land sizes of teff producer and non-producer households were 2.0176 hectares and 0.3167 hectares, respectively. on average, the mean dependency ratios were 56.88% and 60.09% for teff producers, non-producers, respectively. moreover, on average 1.930 numbers of oxen were used by household teff producers. this means producer households had approximately 2 oxen on average and households who did not produce teff were not having ox. the study indicates that the average farm experience of producer respondents in the study area was 9.86 years and there is a significance difference in teff production experience between producer and non-producer respondents at 1% level of significance showing that producer respondents have better teff production experience than non-producer respondents in the study area (see table 2 below). table 3: the maximum likelihood analysis of logit model variable coefficients std. err sign. level sex 0.055 0.032 0.084* farm land size 0.058 0.016 0.000*** number of oxen 0.142 0.021 0.000*** teff variety -0.056 0.014 0.000*** education 0.03 0.013 0.021** technology adoption 0.045 0.017 0.007*** fertilizer application 0.245 0.046 0.000*** dependency ratio -0.031 0 .104 0.762 extension service 0.072 0 .037 0.051* fexptppy 0.004 0.004 0.215 constant 0.184 0.077 0.019 dependent variable = level of teff production r2=0.9421 number of observation =115 *** significant at less than 1% probability level; ** significant at less than 5% probability level; * significant at less than 10% probability level. source: model output, 2021 table 2: summary of descriptive statistics for continuous explanatory variables variable producer non-producer t-value mean mean total mean total farm land size 2.02 0.32 1.57 10.037*** dependency ratio 0.57 0.6 0.58 1.276 number oxen 1.93 0.03 1.96 17.765*** teff production experience 9.86 0.77 7.49 2.909*** note: refer to 1% significance level, source: own survey results, 2021 determinants of households’ teff production before entering the variables in to the model, the multicollinearity problems were checked in terms of variance inflation factor (vif) for continuous and contingency coefficients for dummy variables, respectively. after testing the degree of association of independent variables, all explanatory variables were used for estimation. binary logit model was applied to identify the major determinants of teff production among hypothesized explanatory variables that are assumed to have an influence on the household’s level of teff production by using a statistical package known as stata version 15 (see table 3 below). based on the model result, a possible explanation for each significant independent variable is given as follows. sex of household head: logit model analysis showed that there is a positive relationship between the sex of households and teff production level at a probability level of 10%. it indicates that male-headed households produced more teff than female-headed households. as the involvement of male-headed households increased by one unit, the level of teff production increased by 0.055 on average. total cultivated land for teff production per hectare: in line with our expectation, farm land size for teff production is found to positively affect the level of teff production at 1% level of significance. it shows that households with more farm land size are more likely to produce teff than those with a small land size. this is possible because when the farmland owned by the household is more, the level of production and income become higher and eventually the amount or yield of teff production increases. the result shows that as the cultivated land increased by one hectare, the level of teff production to be produced increased by 0.058 units on average. https://journals.e-palli.com/home/index.php/ajsts pa ge 41 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 38-42, 2023 number of oxen this variable also has a positive effect on the level of teff production at the probability of 1% significant level. this means farmers who own more quantity of oxen produced more output of teff than others. this is because; oxen ownership would help farmers to carry out agricultural operations like ploughing, sowing, and others on time that would improve productivity. the analysis shows that on average, as the number of oxen increased by one unit, the amount of teff to be produced increases by 0.142 units. a previous study by (gebremedhin et al., 2007) found a similar result. education level of household the level of education has a positive influence on the level of teff production at 5% significant level. it indicates that households led by non-literate heads are less likely to understand the modern farming technologies provided to them through any media (extension workers, radio, etc) than literate household heads. it affect production positively since it makes household’s to have ability to take good and well-informed production on teff. and the model shows that as the households’ education level increased by one grade level, the amount of teff to be produced increased by 0.030 units. it is consistent with the study found by amaza et al. (2006) and other literatures; the higher the educational level of household head, the more teff is expected to be produced. fertilizer application the results showed that fertilizer applications may affect teff production positively at a probability level of 1%. it shows that if fertilizer is available in the right amount and time, the level of teff production would be improved. the model analysis shows that as the utilization of fertilizer increased by one unit, the level of teff production to be produced could be increased by 0.245 units on average. this result is similar to the study conducted by (dickinson et al. 1990). technology adoption this variable is found to have a positive influence on teff production level at a probability level of less than 1%. this means farmers who adopt farm technology like fertilizer, raw planting are more likely to produce more teff than farmers who do not adopt it. the analysis shows that as farmers’ technology adoption level increased by one unit, the amount of teff to be produced would be increased by 0.045 units on average. access to extension services access to extension services received by households has a significant positive association with level teff production status at a probability level of 10%. the positive relationship implies that when households get an extension service, the probability of the household to produce teff would be increased. on average, as producers’ access to the extension service increased by one unit, the production level of teff would be increased by 0.072 units. this result is consistent with the study found by (babatunde, 2007). conclusion and recommendations identifying the major determinants of smallholder farmers’ teff production was the main purpose of this study. to achieve this, primary data were collected from 115 smallholder farm households. descriptive statistics were used to explain the different socio-economic characteristics of the sample households and inferential statistics were used to test the dummy and continuous variables. logit regression model was used to identify the major determinants of smallholder farmers’ teff production level. the finding shows that the majorities (73.90% of the sample households) were teff producers and small numbers of households (26.10% of the sample households) were nonproducers of teff. this indicates that more than 50% of smallholder farmers are teff producers in the study area. the results of the logistic regression model indicated that eight out of ten variables, namely, sex of the household, education of household head, total cultivated land for teff production per hectare, number of oxen, technology adoption, and access of extension services received by households, fertilizer application and improved teff variety used were found to be a major determinants of household teff production in the study area. therefore, stakeholders should be considering these variables when smallholder farmers produce teff. references anderson, j. r. (2007). agricultural advisory services of background paper for the world development report. world bank, washington, d.c., usa. ata (agricultural transformation agency) (2013). new teff technologies demonstration trials draft report. addis ababa, ethiopia. csa (central statistical agency). (2013). agricultural sample survey: area and production of major crops, meher season, 1, addis ababa, ethiopia. schmidt, e., & kedir, m. (2009). urbanization and spatial connectivity in ethiopia: urban growth analysis using gis. spielman, d. j., d. k. mekonnen, and d. alemu. (2008). seed, fertilizer, and agricultural extension in ethiopia. in food and agriculture in ethiopia: progress and policy challenges, edited by paul dorosh and shahidur rashid, philadelphia: university of pennsylvania press, 84–122. tareke berhe and nigusse zena (2008). results in a trial of system of teff intensification: debre zeit, ethiopia. gujarathi, damodar m. (2004). basic econometrics. mcgraw-hill. gebremedhin, berhanu, dirk hoekstra, and samson jemaneh. (2007). heading towards commercialization: the case of live animal marketing in ethiopia. ipms https://journals.e-palli.com/home/index.php/ajsts pa ge 42 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 38-42, 2023 working paper. amaza, p. s., umeh, j. c., helsen, j., & adejobi, a. o. (2006). determinants and measurements of food insecurity in nigeria: some empirical policy guide (no. 1004-2016-78541). verstraete, michel m., bernard pinty, and robert e. dickinson. (1990). a physical model of the bidirectional reflectance of vegetation canopies: journal of geophysical research: atmospheres, 95, 1175511765. babatunde, r. o., o. a. omotesho, and o. s. sholotan. (2007). socio-economic characteristics and food security status of farming households in kwara state, north-central nigeria. pakistan journal of nutrition 6(1), 49-58. https://journals.e-palli.com/home/index.php/ajsts pa ge 1 pa ge 34 american journal of smart technology and solutions (ajsts) a comparative study of text-based lossless compression shyam maharjan1*, er. sujan poudel1, dipesh tandukar1 volume 3 issue 2, year 2024 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v3i2.3566 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: august 18, 2024 accepted: september 20, 2024 published: september 24, 2024 lossless data compression is a critical technique used to reduce file sizes without any loss of information during the encoding and decoding processes. this study presents a comparative analysis of two widely-used lossless compression algorithms: huffman encoding and lempel-ziv-welch (lzw). the primary objective is to evaluate the performance of these algorithms in terms of compression ratio, compression time, decompression time, and space savings. the analysis was conducted on 100 files of varying sizes. the results demonstrate that the lzw algorithm outperforms huffman encoding, offering superior compression ratios, faster compression and decompression times, and greater disk space savings. these findings highlight the effectiveness of lzw for efficient data compression in practical applications. keywords compression ratio, data compression, huffman encoding, lempel-ziv-welch (lzw), lossless compression 1 nepal kasthamandap college, nepal * corresponding author’s e-mail: samymhr31@gmail.com introduction data compression is a fundamental technique that reduces the size of digital information by encoding it using fewer bits than its original representation. this process is crucial for optimizing storage and transmission of data, making it a key component in various applications such as file storage, data transfer, and media streaming. data compression can be broadly categorized into two types: lossless and lossy compression (holtz, 1993). lossless data compression ensures that no data is lost during the compression process. it works by eliminating redundancy within the data, allowing the original file to be perfectly reconstructed from the compressed version. this makes lossless compression ideal for text, software, and other types of data where any loss of information is unacceptable. on the other hand, lossy data compression permanently removes some data, which is often acceptable for media files like audio, video, and images where minor data loss does not significantly impact the quality. there are two widely-used algorithms huffman encoding and lempel-ziv-welch (lzw). both of these algorithms are integral to the field of data compression and are known for their efficiency in reducing file sizes without losing any information. huffman encoding operates by constructing a binary tree of nodes, encoding symbols or characters based on their frequency of occurrence. the more frequent a character is, the fewer bits are required to represent it, making huffman encoding particularly effective for files with a skewed frequency distribution. lzw compression, on the other hand, uses a dynamic dictionary to encode data. as the file is processed, substrings are added to the dictionary, allowing repeated substrings to be represented by shorter codes, which can significantly reduce the file size. in this study, we analyze the performance of huffman encoding and lzw compression in terms of compression ratio, compression time, decompression time, and space savings. by conducting experiments on 100 files of varying sizes, we aim to provide a comprehensive comparison of these two algorithms and determine their effectiveness in real-world applications. this research’s primary objective is to comprehensively evaluate two prominent lossless data compression algorithms: huffman encoding and lempel-ziv-welch (lzw). literature review the concept of data compression has a rich history, dating back to the invention of morse code in 1838, which can be considered an early example of data compression. however, modern data compression began in the late 1940s with the development of information theory, which laid the groundwork for more sophisticated compression methods. one of the most significant milestones in this field was the introduction of the huffman coding algorithm in 1951, which provided an optimal method for lossless data compression. in the late 1970s, software compression programs began to emerge, many of which were based on adaptive huffman coding. by the mid-1980s, the lempel-ziv-welch (lzw) algorithm had become a staple in general-purpose compression systems, demonstrating its effectiveness across various applications. gopinath and ravisankar (2020) stated that transmitting large volumes of data from a monitoring field to a central unit is particularly challenging when communication bandwidth is limited, leading to potential data overflow and loss on single-board computers. given the fixed sampling frequency and unchangeable bandwidth, data pa ge 35 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 34-39, 2024 compression becomes essential to prevent these issues by minimizing the size of data for storage and transmission. gupta (2017) explored various lossless data compression techniques and evaluates their performance in terms of time and space complexity. focusing on compressing different media formats like text, doc, bmp, png, and wav files, the study examines huffman coding, which generates variable-length codes for each symbol; lzw, a dictionary-based technique; and shannon-fano coding, which generates binary codes based on symbol frequency. by calculating the mean compression ratio, compression factor, and compression time for these methods, the analysis helps identify the most suitable compression technique for specific file formats in practical applications. verdú (2014) presented a comprehensive analysis of the best achievable rate and other fundamental limits in variable-length strictly lossless compression. it reveals a strong connection between the fundamental limits of fixed-to-variable lossless compression, both with and without prefix constraints, in the non-asymptotic regime. precise quantitative bounds are established, linking the optimal code lengths to the source information spectrum, and providing an exact analysis for arbitrary sources. the study also proved fine asymptotic results for general mixing sources and introduces explicit gaussian approximation bounds for the best achievable rate on markov sources. key concepts such as source dispersion and varentropy rate are defined and characterized, offering a tight approximation of the fundamental nonasymptotic limits for fixed-to-variable compression, except for very small block lengths. patel et al. (2012) presented parallel algorithms and implementations of a bzip2-like lossless data compression scheme optimized for gpu architectures. our approach parallelizes the key stages of the bzip2 compression pipeline: burrows-wheeler transform (bwt), move-to-front transform (mtf), and huffman coding. specifically, we employ a two-level hierarchical sort for bwt, introduce a novel scan-based parallel mtf algorithm, and implement a parallel reduction method for constructing the huffman tree. through detailed performance analysis, we highlight the strengths and weaknesses of each algorithm and propose potential improvements. despite these optimizations, our gpu implementation is 2.78× slower than bzip2, with bwt and mtf-huffman being 2.89× and 1.34× slower on average, respectively. konecki et al. (2011) stated that data compression plays a crucial role in information and communication technologies by saving storage space and reducing network transmission bandwidth. this paper focused on lossless data compression, providing an overview of the algorithms used in popular data archiving tools. since the compression rate varies significantly depending on the data type, the study tests a range of commonly used file types. by examining different tools that implement known algorithms in various forms, the paper identifies which tools offer the best compression capabilities, the fastest performance, and the most optimal balance between compression efficiency and speed. jones (2003) introduced the x-matchpro, a highspeed lossless data compression algorithm with a hardware implementation that achieves data-independent throughputs of 1.6 gbit/s for both compression and decompression using low-cost reprogrammable fieldprogrammable gate array (fpga) technology. the fullduplex implementation enables a combined performance of 3.2 gbit/s. the paper detailed the features of the algorithm and architecture that facilitate these high throughputs, and compares the x-matchpro with other commercially available data compressors in terms of technology, compression ratio, and throughput. x-matchpro is a fully synchronous design, proven in silicon, and is specifically aimed at enhancing gbit/s storage and communication applications. patauner et al. (2011) introduced a compression system optimized for reducing data from pulse digitizing electronics, commonly used in high energy physics (hep) experiments, such as those in calorimeters and time projection chambers (tpcs). using the alice experiment’s tpc as a case study, the paper presents a novel compression method that surpasses conventional lossless compression by compressing entire vectors of digitized samples rather than individual uncorrelated samples. the method approximates incoming vectors with digitized reference vectors stored in memory and compresses only the differences using huffman coding, a process akin to vector quantization combined with huffman coding. initial evaluations in matlab using alice tpc data achieved a 49% compression rate, exceeding the 38% theoretical limit set by the data’s entropy. the method was further implemented in verilog and tested on a virtex-4 development board at 80 mhz, with successful results demonstrated using cosmic ray data. materials and methods this research aims to evaluate and compare the performance of huffman encoding and lempel-zivwelch (lzw) algorithms in lossless data compression. the study involves implementing both algorithms, conducting tests across a diverse set of files, and analyzing their performance through various metrics, including compression ratio, compression and decompression times, and disk space savings. statistical methods are employed to assess the significance of differences in performance, and the results are validated through crossverification. a total of 100 text based samples were taken for the purpose of analyzing the loss compression. for each sample original size, size after compression, entropy, compression ratio, disk saving, compression time and decompression time using two algorithm i.e. lzw and huffman encoding algorithm. pa ge 36 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 34-39, 2024 lzw (lempel-ziv-welch) algorithm lzw (lempel-ziv-welch) is a widely used lossless data compression algorithm that operates by replacing sequences of characters in the input data with shorter codes from a dynamically constructed dictionary. the algorithm starts with a dictionary containing all possible single-character sequences, and as it processes the input, it identifies the longest match in the dictionary, outputs its corresponding code, and then adds a new entry that extends the current match by one character. this process allows the dictionary to grow and capture increasingly complex patterns, leading to efficient compression, especially for data with repeated sequences. lzw is commonly used in applications like gif image compression and file compression utilities due to its simplicity and effectiveness. data, making it highly efficient for compressing files with uneven character frequencies. this algorithm is widely used in formats like jpeg and mpeg, as well as in general-purpose compression utilities, due to its optimal compression capabilities. figure 1: flow chart of lzw huffman encoding algorithm huffman encoding is a fundamental lossless data compression algorithm that assigns variable-length codes to characters based on their frequencies in the input data, with more frequent characters receiving shorter codes. the algorithm constructs a binary tree, known as a huffman tree, where each character is represented by a leaf node, and the path from the root to the leaf determines the character’s binary code. by prioritizing shorter codes for frequent characters, huffman encoding minimizes the overall length of the encoded figure 2: flow chart of huffman results and discussion based on our comparative analysis of the huffman encoding and lempel-ziv-welch (lzw) algorithms, we evaluated their performance on 100 files of varying sizes and entropies. the comparison focused on four key metrics: compression ratio, disk space savings, compression time, and decompression time. result the results of our comparative analysis between the huffman encoding and lempel-ziv-welch (lzw) algorithms demonstrate that the lzw algorithm outperforms huffman in multiple aspects, including compression ratio, disk space savings, and both compression and decompression times. from the analysis of 100 test files with varying sizes and entropies, the following key findings were observed: compression ratio the lzw algorithm achieved a higher average compression ratio of 2.16, indicating that it compresses data more efficiently than the huffman algorithm, which had an average compression ratio of 1.7. this suggests that lzw is particularly effective at reducing the size of data, especially in cases with repetitive patterns, making it a more efficient choice for certain types of data compression. pa ge 37 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 34-39, 2024 compression time the lzw algorithm proved to be faster, taking an average of 9.11 seconds to compress files, whereas the huffman algorithm required significantly more time, averaging 23.97 seconds for compression. this indicates that lzw not only provides better compression efficiency but also performs the compression process more quickly compared to huffman, making it a more time-efficient choice for data compression tasks. figure 3: compression ratio lzw vs huffman figure 4: compression time lzw vs huffman figure 5: decompression time lzw vs huffman decompression time the lzw algorithm demonstrated faster decompression, with an average time of 6.77 seconds, while the huffman algorithm took longer, averaging 10.62 seconds. this suggests that lzw not only compresses data more efficiently but also decompresses it more quickly than huffman, making it a more effective choice for applications where speed is a critical factor. pa ge 38 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 34-39, 2024 disk space savings the lzw algorithm achieved greater disk space savings, averaging 53.25%, compared to the huffman algorithm, which provided an average saving of 40.98%. this indicates that lzw is more effective at reducing file sizes, making it a better choice for maximizing storage efficiency. figure 6: disk saving lzw vs huffman table 1: coefficients model unstandardized coefficients standardized coefficients t sig. b std. error beta 1 (constant) 4.464 .792 5.638 <.001 entropy -.517 .171 -.260 -3.017 .003 original size(kb) .000 .000 .463 .534 .595 compression time(s) -.024 .015 -1.405 -1.634 .106 decompression time(s) .033 .032 1.377 1.055 .294 a. dependent variable: compression ratio regression analysis of huffman regression analysis of lwz in both algorithms, entropy negatively affects the compression ratio, meaning both huffman and lzw are less effective at compressing data with higher entropy. the effect size and statistical significance are similar for both algorithms, indicating that entropy is a critical factor in both contexts. neither huffman nor lzw shows a significant relationship between the original file size and the compression ratio. this suggests that the size of the file is less important than the content’s structure (as indicated by entropy). both algorithms show similar, non-significant relationships with compression and decompression times. this might imply that while time is important for performance considerations, it does not directly influence the effectiveness of compression in terms of the ratio achieved. since entropy is the only significant predictor, choosing between huffman and lzw might depend more on other factors like processing time or ease of implementation, rather than just compression ratio, especially for data with varying entropy. when dealing with data with high entropy, consider using other methods to preprocess the data or choose an table 2: coefficients model unstandardized coefficients standardized coefficients t sig. b std. error beta 1 (constant) 4.464 .792 5.638 <.001 entropy -.517 .171 -.260 -3.017 .003 original size(kb) .000 .000 .463 .534 .595 compression time(s) -.024 .015 -1.405 -1.634 .106 decompression time(s) .033 .032 1.377 1.055 .294 a. dependent variable: compression ratio pa ge 39 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 34-39, 2024 algorithm that is better suited for high-entropy data. given the similar outcomes, further research could explore additional independent variables or consider nonlinear models to see if more complex relationships exist that aren’t captured in this linear model. conclusion the comparative analysis of huffman encoding and lempel-ziv-welch (lzw) algorithms on 100 files of varying sizes and entropies has highlighted the superior performance of the lzw algorithm in several key areas. lzw consistently outperformed huffman encoding, achieving a higher compression ratio, greater disk space savings, and faster compression and decompression times. on average, lzw was able to reduce disk space usage by 53.25%, compared to 40.98% with huffman encoding. lzw also maintained a better compression ratio of 2.16, while huffman achieved a ratio of 1.7. furthermore, lzw demonstrated faster processing times, with average compression and decompression times of 9.11 seconds and 6.77 seconds, respectively, in contrast to huffman’s 23.97 seconds for compression and 10.62 seconds for decompression. the regression analysis further supports these findings, showing that entropy significantly influences the compression ratio for both algorithms, with higher entropy leading to less effective compression. neither algorithm displayed a significant relationship between the original file size and compression ratio, indicating that the content’s structure, as reflected by entropy, is more critical. additionally, the analysis revealed no significant impact of compression or decompression times on the compression ratio, suggesting that while time is a factor in performance, it does not directly affect compression efficiency. these findings suggest that lzw is generally more efficient and effective for lossless data compression, especially in scenarios where time efficiency and storage optimization are critical. however, huffman encoding still has its merits and may be preferable in specific use cases depending on the nature of the data. ultimately, the choice of algorithm should be tailored to the specific requirements of the application at hand. for data with high entropy, other preprocessing methods or compression algorithms may be necessary to achieve optimal results. references gopinath, a., & ravisankar, m. (2020). comparison of lossless data compression techniques. in 2020 international conference on inventive computation technologies (icict) (pp. 628–633). https://doi. org/10.1109/icict48043.2020.9112516 holtz, k. (1993). the evolution of lossless data compression techniques. in proceedings of wescon ’93 (pp. 140–145). https://doi.org/10.1109/ wescon.1993.488424 konecki, m., kudelić, r., & lovrenčić, a. (2011). efficiency of lossless data compression. in 2011 proceedings of the 34th international convention mipro (pp. 810–815). https://ieeexplore.ieee.org/abstract/ document/5967166 kontoyiannis, i., & verdú, s. (2014). optimal lossless data compression: non-asymptotics and asymptotics. ieee transactions on information theory, 60(2), 777–795. ieee transactions on information theory. https://doi.org/10.1109/tit.2013.2291007 nunez, j. l., & jones, s. (2003). gbit/s lossless data compression hardware. ieee transactions on very large scale integration (vlsi) systems, 11(3), 499–510. https://doi.org/10.1109/tvlsi.2003.812288 patauner, c., marchioro, a., bonacini, s., rehman, a. u., pribyl, w. (2011). a lossless data compression system for a real-time application in hep data acquisition. ieee transactions on nuclear science, 58(4), 1738–1744. https://doi.org/10.1109/ tns.2011.2142193 patel, r. a., zhang, y., mak, j., davidson, a., & owens, j. d. (2012). parallel lossless data compression on the gpu. in 2012 innovative parallel computing (inpar) (pp. 1–9). https://doi.org/10.1109/inpar.2012.6339599 sharma, k., & gupta, k. (2017). lossless data compression techniques and their performance. in 2017 international conference on computing, communication and automation (iccca) (pp. 256–261). https://doi. org/10.1109/ccaa.2017.8229810 pa ge 1 pa ge 15 american journal of smart technology and solutions (ajsts) preparation and conductivity of polymer-modified graphene films md. jewel rana1, khan rajib hossain2*, marzan mursalin jami3, md. abu shyeed4, md. kamrul hasane5 volume 2 issue 1, year 2023 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: december 24, 2022 accepted: march 02, 2023 published: march 06, 2023 the hummers method was used to make graphite oxide, and ultrasonic exfoliation at 25°c and 90°c was used to make graphene oxide (go). at a low temperature, polyethyleneimine (pei) was used as a reducing and changing agent for graphene oxide (go) to make dispersions of graphene that were modified with pei. optoelectronics’ electron and infrared spectroscopy showed how temperature affected pei’s ability to break down go. the results show that pei can partially reduce go at 25°c. at 90°c, the grafted pei gradually dissociated from the go sheet. the graphene dispersion was filtered and assembled into a pei-go film, and its conductivity was found to be 117s.m-1, hopefully conductive material for graphene. keywords conductive, graphene film, polyethyleneimine, graphene oxide 1 department of applied chemistry and chemical engineering, bangabandhu sheikh mujibur rahman science and technology university, gopalgonj 8100, bangladesh 2 state key laboratory of solid lubrication, lanzhou institute of chemical physics, chinese academy of sciences, lanzhou 730000, china 3 school of textile science and engineering, wuhan textile university, wuhan, china 4 department of applied chemistry and chemical engineering, rajshahi university, rajshahi 6205, bangladesh 5 department of chemistry,hajee mohammad danesh science & technology, dinajpur 5200, bangladesh * corresponding author’s e-mail: apexlabbd@gmail.com introduction graphene, which is only one atom thick, has a unique planar structure in two dimensions, a very high specific surface area, and great barrier properties. graphene has attracted extensive attention for its electrical (kim, et al., 2017 and mohan, et al., 2016) and thermal conductivity (balandin, et al., 2008 and pop, et al., 2012), optical (schoche, et al., 2017) and mechanical properties (liu, et al., 2012, kordkheili, et al., 2013, zhu, et al., 2010). water could be used as a reducing agent, according to stankovich s, dikin d a et al., 2007. hydrazine hydrate can turn graphene oxide (go) into graphene with higher electrical conductivity. the electrical conductivity of the resulting graphene is similar to that of graphite. (qi, x y, yan d, jiang z, et al., 2011) the go made with the hummer method was turned into graphene by heating it, and the graphene was then added to a polystyrene matrix. the graphene/polystyrene composite material was obtained. the research results found that: after adding a small amount of graphene to pure polystyrene, its electrical conductivity changed from 6.70×10-14s.m-1 increased to about 3.49s.m-1. (liu h y, kuila t, kim n h, et al., 2013) reported that polyethyleneimine (pei) could reduce go, and at the same time, pei was grafted to the in situ generated stone. a water-soluble polymer-modified graphene pei-go was obtained on the graphene sheet. according to the conductivity test results, the pei mass ratio is go. the conductivity of the resulting pei-go can be greatly affected. based on the research (liu, h y, kuila t, kim n h, et al., 2013), this paper looks into how the reaction temperature affects how pei breaks down go. under the premise of fixing the mass ratio of raw materials pei and go by changing the reaction temperature, peimodified graphene was prepared. the prepared peimodified graphene aqueous dispersion was filtered into a composite film, and the graphene composite thin films obtained at different temperatures were investigated the conductivity of the film. method and materials reagents and instruments sigma-aldrich sold polyethyleneimine with an average relative molecular mass of mn≈10000. concentrated sulfuric acid (the mass fraction is 98%), potassium permanganate, sodium nitrate, hydrogen peroxide (30% by mass), and hydrochloric acid (38% by mass) were analytically pure reagents for further purification. the samples were tested by an axis-novax-ray photoelectron spectrometer (kratos analytical ltd., uk) elemental composition. the structural information of the samples was tested using a nicolet 6700 infrared spectrometer (thermo scientific, usa). product morphology tests were completed on the h-7650 tem (hitachi, japan) and jsm-6701f sem (jeol, japan). a keithley 2000 type four-probe resistivity tester (keithley instruments inc., usa) was used to test its electrical conductivity. preparation of graphene oxide first, graphite oxide was prepared according to the hummers method (lei, et al., 2016). under ice-water bath conditions, 50 ml of concentrated sulfuric acid was added to 1g of natural graphite and 0.5g of sodium nitrate and stirred to mix them evenly. under magnetic stirring, take 3g of potassium permanganate, slowly add https://journals.e-palli.com/home/index.php/ajsts pa ge 16 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 15-19, 2022 it to the system, and continue stirring. after 10 minutes, the ice-water bath was removed, the reaction system was heated to 35 °c, and the stirring was continued for 30 minutes. at this point, the black reaction system became a brownish-yellow color, 40ml of deionized water was added to the reaction system, the temperature was raised to 95 °c, and the reaction was continued for 1h to obtain a golden yellow dispersion. additions of 150ml of deionized water and 10ml of hydrogen peroxide with a mass fraction of 30% were mixed for 30 min, and then 50ml of 5% hydrochloric acid solution was added. last, wash the sample with deionized water until the solution is neutral. the obtained product was filtered with suction to remove the solvent water, and the obtained sample was placed in a vacuum at 40°c. dry to obtain dry graphite oxide. weigh graphite oxide and disperse it in an appropriate amount of deionized water. 30 minutes of medium sonication to obtain 0.5 mg/ml. the graphene oxide solution is ready for use. preparation of polyethyleneimine-modified graphene dissolve 0.05 g of polyethyleneimine in 100 ml deionized water to obtain a pei solution of 0.5 mg/ml. the pei solution was placed in a three-necked flask, and 100 ml of go solution (0.5 mg/ml) was added dropwise to the pei solution via a constant pressure drop funnel while magnetic stirring at 25°c. the reaction was stirred for 4h to obtain a brown pei-go dispersion. the go dispersions were passed through a membrane (a cellulose filter membrane with a pore size of 0.2 μm) suction filtration, with about 1 liter of deionized water added to wash the sample and remove the impurities. separated pei and pei-go films were obtained, respectively. results and discussion morphology analysis of the product the go sheet contains many epoxy groups, hydroxyl groups, carboxyl groups, and carbonyl groups at the edge of the sheet, and go has good water solubility. it can be spread out evenly in solvent water to make a brownishyellow go solution. the go solution was added to the polyethyleneimine solution at room temperature. after stirring for 4h, pei-go was obtained. in appearance, peigo was not much different from go, both of which were brown-yellow aqueous solutions. this is because, at 25 °c, pei can’t reduce go very well, so it’s grafted onto go sheets instead. the amino group of pei reacts with the epoxy group of go to make pei-go with a modified branch linkage. when the reaction temperature was 90 °c, most of the amino groups of pei were dissociated from the go sheet, forming c=c, reducing go. it is graphene, which is consistent with the mechanism of hydrazine hydrate reduction of go. (song, p, zhang x, sun m, et al., 2012) used oxalic acid as a reducing agent and heated it at 75°c for 18h. the brown go was also observed to transform into a black graphene dispersion. so that the shapes of go and pei-go could be seen, their dispersions were taken and freeze-dried to make samples that looked like fluffy sponges. the field emission scanning electron microscope (fe-sem) photo of the surface after gold spraying is shown in figure 1. for go, there are a lot of random wrinkles on the go sheet, and the edges are easy to bend, showing that the go sheet has good flexibility and toughness. due to their very high specific surface area, the different go sheets in pei-go are randomly stacked and cross over each other. however, the packing density of pei-go is slightly larger than that of go because, at 25°c, go has figure 1: field emission scanning electron microscopes photos of go and pei-go. a higher packing density. pei partially restored it. it can be observed that the graphene sheets have corrugated wrinkles due to their close packing with each other. this is because go is effectively reduced to graphene by pei at 90°c, due to the π-π interaction leading to graphene sheets being tightly packed together. to better show the morphology of the samples, another dispersion liquid was taken, dropped on the carbon film, and dried for transmission electron microscopy (tem) analysis. the obtained tem photos are shown in figure 2. figure 2a shows a tem image of go divided into two parts. it can be seen that graphene oxide has a finely layered structure with a large number of irregular wrinkles at the edges, which can reduce graphene oxide. the specific surface area reduces its surface energy, thereby making it stable. the tem image of pei-go is shown in figure. 2b with graphene oxide. in contrast, pei-go is still a finely layered structure, and its darker color may be due to the pei grafted on the surface of go. the folded state of graphene at the edge https://journals.e-palli.com/home/index.php/ajsts pa ge 17 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 15-19, 2022 figure 2: tem view of go and pei-go is clearly seen in the graphene sheet, and there are more random wrinkles. this is because the graphene sheets π-π interact, making wrinkles more likely. more wrinkles also indicate that the graphene oxide sheets are reduced to graphene sheets. (xu, l q, liu y l, neoh k g, et al., 2011) reported that due to the modification of graphene sheets and graphene oxide sheets showed similar tem morphologies. structural analysis of the product x-ray photoelectron spectroscopy (xps) was done on the samples to determine how the temperature affected pei’s ability to break down go. figure 3 depicts the spectrum of a spectroscope (xps) analysis. for go, its xps spectra have two main peaks, corresponding to sp. carbon (281.4 ev) and oxygen (530.5 ev), which is consistent with the conclusion of the (chua, c k et al., 2012). except for the carbon and oxygen peaks, the xps spectrum of pei-go is at 398 ev. a nitrogen peak appeared at the position of go, which means that the polyethyleneimine molecular chain was grafted onto the go sheet. carbon and oxygen peaks of pei-go and go are compared. it has been found that the former oxygen peak is less strong than it used to be. this may be because the pei molecular chain absorbs the oxygen from the epoxy group in the go sheet. the nitrogen of the amino group is substituted to form a nitrogen-containing three-membered ring. it means that even at 25°c, the amino group of the pei molecular chain can interact with the go sheet. pei-go was created by reacting the epoxy groups and grafting them onto the go sheet. in (li, x et al., 2009), go was heated in a nitrogen atmosphere to prepare nitrogendoped graphene, also found a nitrogen peak at 398 ev in the xps spectrum. figure 4 shows the three samples’ fourier transform infrared (ft-ir) spectra. for go, the broad and strong blunt peak at 3100-3500cm-1 is the stretching vibration peak of the associated hydroxyl group, and the peak in the 1734cm-1 figure is the c=o stretching vibration peak at 1640 cm-1. the spikes correspond to the c=c stretching vibration peak, 1388 cm-1. the absorption peak is the bending vibration peak of the hydroxyl group at 1246 cm-1. the absorption peak is the co stretching vibration peak of the epoxy group. the appearance of these oxygen-containing functional groups is consistent with the (gao, y, liu l q, zu s z, et al., 2011), which says that ming graphite was successfully oxidized, and graphene oxide was obtained after ultrasonic exfoliation. in contrast, the ft-ir curves of pei-go are not related. the hydroxyl stretching vibration peak was replaced by the pei molecular chain. the n-h stretching vibration peak is strong and sharp, indicating that the pei molecular chain is branched to the go sheets. it is worth noting that in pei-go 1734 cm-1. still exist c=o vibration figure 3: xps spectra survey of the go and pei-go figure 4: ft-ir spectra survey of the go and pei-go https://journals.e-palli.com/home/index.php/ajsts pa ge 18 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 15-19, 2022 peak, only the intensity compared to the decrease of go shows that pei can only partially reduce go at 25 °c but cannot completely reduce go to graphene. the c-o stretching vibration of the epoxy group in the go curve’s dynamic peak (1246 cm-1) disappeared completely in pei-go, proving that the epoxy group of go was not affected by the reaction of the amino group of pei. the effect of reaction temperature means that pei molecular chains can also be grafted onto go at 25°c. conductivity analysis of the product the degree of reduction of go has a lot to do with how well graphene conducts electricity. because the go sheet contains many hydroxyl, epoxy, and carbonyl groups, the carboxyl group destroys the conjugated structure of graphene, so go is a non-conductive substance. when the reducing agent is added, the oxygen-containing groups of the go sheet undergo partial or full reduction, and graphene restores the conjugated structure, thus helping to improve the electrical conductivity of the product. the higher the go reduction degree, the better the graphene obtained and the higher the conductivity. graphene’s electrical conductivity, on the other hand, can demonstrate the degree of reduction in go. it means that the degree of go reduction is greater. go and pei-go were assembled into membranes by suction filtration, respectively. the acupuncture method was used to calculate the electrical conductivity according to the method (jun, et al., 2015). since go is a non-conductive substance, the conductivity of the obtained go thin film is 0.002 s·m-1. such a low conductivity also explains why the graphene oxide prepared by this method is rich in oxygen-containing functional groups, which is consistent with the results of ft-ir and xps. at 25°c, go was added dropwise to pei, and pei was grafted onto the go sheet through the reaction between the amino group of pei and the go epoxy group, resulting in a pei-go dispersion liquid. it was assembled into a pei-go film by suction filtration, and its conductivity was 0.103 s.m-1. the specific conductivity of go has increased, indicating that pei reduced the oxygen-containing groups. (xu, z, bando y, liu l, et al., 2011) found that the influence of epoxy groups on the electrical conductivity of graphene materials is much greater than that of hydroxyl groups, and the dissociation of epoxy groups from the go sheet is preferential to that of the hydroxyl group, which is consistent with the results in this paper. at 25°c, due to the amino groups in the pei molecular chain and the go sheet. the layer epoxy group reaction made go more conductive. still, since the surface-grafted pei molecular chain was not conductive, the increase in conductivity of pei-go was not very high. conjugated structure of graphene, due to the departure of the non-conductive substance pei and the recovery of the conjugated structure. conclusion the temperature of the reaction has a big effect on how well polyethyleneimine can break down graphene oxide. at 25°c, pei can only partially reduce go and connect. at 90°c, pei effectively reduced go to graphene, resulting in surface-modified graphene pei-go. this conclusion has some implications for reducing go with other reducing agents. the films assembled from peigo dispersions have, it has high electrical conductivity and is expected to be used in graphene-conductive composites. declaration of interests the authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. references balandin, a. a., ghosh, s., bao, w., calizo, i., teweldebrhan, d., miao, f., & lau, c. n. (2008). superior thermal conductivity of single-layer graphene. nano letters, 8(3), 902-907. https://doi. org/10.1021/nl0731872 chua, c. k., ambrosi, a., & pumera, m. (2012). graphene oxide reduction by standard industrial reducing agent: thiourea dioxide. journal of materials chemistry, 22(22), 11054-11061. https://doi.org/10.1039/ c2jm16054d gao, y., liu, l. q., zu, s. z., peng, k., zhou, d., han, b. h., & zhang, z. (2011). the effect of interlayer adhesion on the mechanical behaviors of macroscopic graphene oxide papers. acs nano, 5(3), 2134-2141. https://doi.org/10.1021/nn103331x jun, y. s., sy, s., ahn, w., zarrin, h., rasen, l., tjandra, r., ... & yu, a. (2015). highly conductive interconnected graphene foam based polymer composite. carbon, 95, 653-658. https://doi.org/10.1016/j. carbon.2015.08.079 kim, k. w., kim, j. h., cho, s., shin, k., & kim, s. h. (2017). scalable high-performance graphene paper with enhanced electrical and mechanical properties. thin solid films, 632, 50-54. https://doi. org/10.1016/j.tsf.2017.04.039 kordkheili, s. h., & moshrefzadeh-sani, h. (2013). mechanical properties of double-layered graphene sheets. computational materials science, 69, 335-343. https://doi.org/10.1016/j.commatsci.2012.11.027 liu, y., xie, b., zhang, z., zheng, q., & xu, z. (2012). mechanical properties of graphene papers. journal of the mechanics and physics of solids, 60(4), 591-605. https://doi.org/10.1016/j.jmps.2012.01.002 liu, h., kuila, t., kim, n. h., ku, b. c., & lee, j. h. (2013). in situ synthesis of the reduced graphene oxide–polyethyleneimine composite and its gas barrier properties. journal of materials chemistry a, 1(11), 3739-3746. https://doi.org/10.1039/c3ta01228j lei, l., xia, z., zhang, l., zhang, y., & zhong, l. (2016). preparation and properties of aminofunctional reduced graphene oxide/waterborne https://journals.e-palli.com/home/index.php/ajsts pa ge 19 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 15-19, 2022 polyurethane hybrid emulsions. progress in organic coatings, 97, 19-27. http://dx.doi.org/10.1016%2fj. porgcoat.2016.03.011 li, x., wang, h., robinson, j. t., sanchez, h., diankov, g., & dai, h. (2009). simultaneous nitrogen doping and reduction of graphene oxide. journal of the american chemical society, 131(43), 15939-15944. https://doi. org/10.1021/ja907098f mohan, v. b., jayaraman, k., stamm, m., & bhattacharyya, d. (2016). physical and chemical mechanisms affecting electrical conductivity in reduced graphene oxide films. thin solid films, 616, 172-182. https:// doi.org/10.1016/j.tsf.2016.08.007 pop, e., varshney, v., & roy, a. k. (2012). thermal properties of graphene: fundamentals and applications. mrs bulletin, 37(12), 1273-1281. https:// doi.org/10.1557/mrs.2012.203 qi, x. y., yan, d., jiang, z., cao, y. k., yu, z. z., yavari, f., & koratkar, n. (2011). enhanced electrical conductivity in polystyrene nanocomposites at ultra-low graphene content. acs applied materials & interfaces, 3(8), 3130-3133. https://doi.org/10.1021/ am200628c schöche, s., hong, n., khorasaninejad, m., ambrosio, a., orabona, e., maddalena, p., & capasso, f. (2017). optical properties of graphene oxide and reduced graphene oxide determined by spectroscopic ellipsometry. applied surface science, 421, 778-782. https://doi.org/10.1016/j.apsusc.2017.01.035 stankovich, s., dikin, d. a., piner, r. d., kohlhaas, k. a., kleinhammes, a., jia, y., ... & ruoff, r. s. (2007). synthesis of graphene-based nanosheets via chemical reduction of exfoliated graphite oxide. carbon, 45(7), 1558-1565. https://doi.org/10.1016/j. carbon.2007.02.034 song, p., zhang, x., sun, m., cui, x., & lin, y. (2012). synthesis of graphene nanosheets via oxalic acidinduced chemical reduction of exfoliated graphite oxide. rsc advances, 2(3), 1168-1173. https://doi. org/10.1039/c1ra00934f zhu, y., murali, s., cai, w., li, x., suk, j. w., potts, j. r., & ruoff, r. s. (2010). graphene and graphene oxide: synthesis, properties, and applications. advanced materials, 22(35), 3906-3924. https://doi. org/10.1002/adma.201001068 xu, l. q., liu, y. l., neoh, k. g., kang, e. t., & fu, g. d. (2011). reduction of graphene oxide by aniline with its concomitant oxidative polymerization. macromolecular rapid communications, 32(8), 684-688. https://doi.org/10.1002/marc.201000765 xu, z., bando, y., liu, l., wang, w., bai, x., & golberg, d. (2011). electrical conductivity, chemistry, and bonding alternations under graphene oxide to graphene transition as revealed by in situ tem. acs nano, 5(6), 4401-4406. https://doi.org/10.1021/ nn103200t https://journals.e-palli.com/home/index.php/ajsts pa ge 1 pa ge 57 american journal of smart technology and solutions (ajsts) evaluating soybean root health using residual neural network (renn) based image analysis vivek gupta1*, jhankar moolchadani2, harsh singh chouhan2 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.5382 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: june 22, 2025 accepted: july 31, 2025 published: september 05, 2025 renn’s layer-wise image segmentation and robust data processing address the complexities of soybean root analysis, providing valuable insights for improved crop management. accurate assessment of soybean root health is crucial for optimal crop production and growth. this study utilizes residual neural network (renn) image evaluation to analyze soybean root development. field data is collected and integrated by deploying sensorbased devices (iot devices) to evaluate soybean crop stages. renn facilitates data preprocessing, layer-wise image segmentation, and effective data processing, enabling accurate assessment of root health, plant vigor, flower fragmentation, and fruit formation. this approach predicts soybean crop yield and provides valuable insights into degradation detection and decision-making for optimal soybean cultivation practices. renn utilizes layer-based image formation, collecting data from farm fields through sensor-based devices. the dataset encompasses various environmental and weather conditions, ensuring comprehensive coverage. key considerations for data preprocessing include temperature, humidity, precipitation as the weather conditions, soil type, moisture, sunlight as environmental factors, field location, soil heterogeneity as spatial variability, and growth stage, seasonality as temporal variability. by integrating these factors, renn enables accurate evaluation of soybean root conditions, facilitating root health assessment, plant growth monitoring, yield prediction, and optimized cultivation practices. keywords convolution neural network (conn), data preprocessing system (dps), internet of things (iot), neural network (nn), residual neural networks (resnn) 1 department of computer science and engineering, aset & amity university gwalior, mp, india 2 department of electronics and communication & indore institute of science and technology, indore, mp, india * corresponding author’s e-mail: vivek.gupta5@s.amity.edu introduction in the realm of technological approaches, the domain of agriculture is vast and plays a significant role in the advancement of agricultural technology. the approaches used were used to evaluate the data set of images. the images are clear with proper originality and appropriate for evaluation. therefore, the technological enhancement toward the uses of artificial intelligence and its based method for the image evaluations for the soybean crop farming. technologically, the devices deployed for the image extractions and the images of the plant picked from the field and images are further forwarded for the input values. image recognition in agriculture has promoted research for the increase in the production of the crop. additionally, crop evaluation research facilitates the identification of the health condition of the crop plant. the system that the paper shows focus on the design of the device and the adoption of the best method based on artificial intelligence for the validation of image-based data sets. the approaches of residual neural network (rnn) are suitable for proper validation of the image-based dataset. here, the research engaged the iot-based image cameras that study and evaluate data sets of files and input the values of the rnn method of evaluation of the crop conditions. as research supports, the ideology is extracted based on technical devices that support the states of the root formation and condition of the soybean plant in the farming of soybeans (he, 2016). figure 1: model of the evaluating the soybean plants and roots images analysis literature reviews the literature review considers the two states of image segmentation, the first is considered which is based on computer vision and the states of images. the second is based on the model of iot controllers that illustrates the image processing for the segmentation and is forwarded pa ge 58 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 57-62, 2025 to the machine learning-based model of the imaging system is resnn (krizhevsky, 2017). based on computer vision techniques: computer vision technology can be used in agriculture to help farmers monitor crops, detect pests and diseases, and improve yields: crop monitoring high-resolution cameras and algorithms can analyze images of crops to provide insights into plant health, growth stages, and potential yield. pest and disease detection computer vision systems can help identify potential pests and diseases that may cause crop losses. weed identification computer vision can help identify weeds and apply precision herbicides. yield prediction by analysing factors like plant height, leaf area, and fruit count, computer vision can help predict crop yield. automated harvesting computer vision can help automate the process of harvesting crops. soil analysis computer vision can help analyze soil conditions. nutrient management computer vision can help identify nutrient deficiencies and apply targeted fertilizers. computer vision can be used in a variety of ways, including: (leibe et al., 2016) 1. uavs: unmanned aerial vehicles (uavs) equipped with computer vision systems can help farmers monitor crops and assess plant health. 2. mobile robots: farmers can use mobile robots equipped with computer vision to drive around and collect data. 3. static cameras: farmers can use static cameras to take images of crops from an advantageous position. the authors suggested that, based on data availability image quality many studies and competitions used plant image datasets with a single background. for example, the plant village dataset contains many labelled plants leaf images from various species with different diseases, but the pictures were from a controlled environment, and their backgrounds are very simple. however, because of lighting, occlusion, and shadows in the natural atmosphere, the image quality and visual perception ability will degenerate greatly. many noises appear in the images, which is a big challenge for automatically analysing unconstrained natural images in the field. although human visual systems can easily deal with these problems, establishing a computational model of plant phenotyping is still an open-ended question (hu et al., 2018). image annotation deep learning needs to learn features from sufficient annotated data, but data annotation faces the following challenges: a. manual annotation sometimes requires a large amount of prior or professional domain knowledge and rich working experience. b. data annotation is a time-consuming and hectic step, especially in object detection and image segmentation. detection and segmentation require instance-level (boxes) and pixel-level annotations (masks). if more and more images are to be annotated, the workload will be massive, while efficiency and accuracy cannot be guaranteed (lin et al., 2019). c. some images lack visual cues, such as hyperspectral and thermal imaging, so it is much more difficult to label these data than rgb images. online researchers have deployed a human-machine collaboration interface called fluid annotation (andriluka et al., 2018) that can be used to annotate the class label and delineate the contours of every object and background in an image (bello et al., 2019). the authors suggested the data-based analysis algorithm robustness at present, some mainstream algorithms perform well on particular datasets, and most of them are only designed for specific organs or specific plant species. due to the large differences in colour, shape, size, and other characteristics between different detection objects, these algorithms do not generalize well. when the dataset changes, many algorithms will be invalid, so researchers must redesign the feature extractor and readjust the hyperparameters. for stress phenotyping, the degree of plant stress changes over time. the model needs to be improved and modified to be dynamically analysed throughout the entire cycle of stress, which is a challenge for designing a processing framework (li et al., 2018). deep learning firstly, deep learning-based algorithms rely on a big number of labelled sample images, which makes it difficult to achieve excellent results in the following three scenarios: • training samples do not exist in some object categories. • there are a few samples in object categories. • the sample size of different categories is extremely imbalanced. then, some deep learning-based solutions lack prior knowledge, and it is difficult to adaptively use my discriminative visual features. moreover, the deep neural network is used as a “black box”, which cannot perform explicit reasoning and lacks interpretability. tasks like gene-phenotype association and image description require high-level logical reasoning and often can’t be solved with simple classification or regression methods. pa ge 59 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 57-62, 2025 these problems need more advanced approaches, such as: deep learning model, multimodal learning, graphbased methods (zagoruyko et al., 2017). finally, most 3d point clouds are still analysed by utilizing traditional 3d processing methods. solutions based on deep learning have not been popularized in plant phenotyping. the following research aspects are worthy of attention in the future. 1. plant images with complex backgrounds require effective segmentation of the foreground and background. methods based on deep learning are very suitable for image segmentation, but image annotation becomes the major limiting factor for applying deep learning in plant phenotyping. to reduce the requirements of annotated data, the following solutions were proposed and developed: on the one hand, some image generation strategies (e.g., gans) can be applied to increase image diversity and availability. on the other hand, the dependence of models on data can be reduced by improving algorithms, such as zero sample learning, small sample learning, transfer learning, and so on (tan & le, 2019). 2. most existing deep learning algorithms rely on many labelled images to fit many parameters for prediction, ignoring the prior knowledge of many domain associations and the intuitive understanding of decisionmaking processes, which limits the interpretation of model functions to a certain extent (pharm, 2020). 3. cnn has great potential in 3d reconstruction and segmentation. some approaches use cnn to project 2d segments onto 3d representations or apply them to 3d images directly. thus, a lot of 3d processing work requires to application of cnn architectures to characterize and understand plant phenotypes directly model of iot devices for image extraction the author suggested that the rapid evolution of iot devices necessitates efficient image super-resolution techniques, while existing advanced methods, based on deep convolutional neural networks, are too resourceintensive for these circuit-based models, and this gap illustrates the need for a more suitable solution. in this study, we introduce a lightweight, essentially superresolution model specially designed for iot devices. this model incorporates a novel deep residual feature distillation block (drfdb), which leverages a depthwise-separable convolution block (dcb) for effective feature extraction. determined to reduce computational and memory demands without changes to image quality. the model shows improved performance metrics like psnr, while requiring fewer parameters and less memory usage, making it highly suitable for iot applications. this study presents a breakthrough in super-resolution for iot devices, balancing high-quality image reconstruction with the limited resources of these devices (gao et al., 2019). materials and methods resnn is an artificial intelligence method to help in the evaluation of the images of the plant soybean crop. a residual neural network (resnet) stacks residual blocks on top of each other to form a network. the residual neural network to know about residual neural networks and the most popular resnets, including resnet-34, resnet-50, and resnet-101. in current years, the field of artificial intelligence applied to computer vision has undergone far-reaching transformations due to the introduction of new technologies (xie s, zerhouni e, huang g, 2017). a. the rapid progress in deep learning has enabled computer vision models to achieve unprecedented levels of accuracy and efficiency in tasks such as image recognition, object detection, and face recognition, surpassing human capabilities in many cases. b. but, while it gives us the option of adding more fully connected layers to the cnns to solve more complicated tasks in computer vision, it comes with its own set of issues. it has been observed that training the nn becomes more difficult with the extension in the number of added layers, and in some cases, the accuracy dwindles as well. c. it is here that the use of resnet assumes importance. deeper neural networks are tough to train. with resnet, it becomes easy to surpass the difficulties of training very deep neural networks. d. when working with deep convolutional neural networks to break a problem related to computer vision, machine learning experts engage in mounding further layers. these fresh layers help break down complex problems more efficiently, as the different layers can be trained for varying tasks to get largely accurate results. e. while the number of piled layers can enrich the features of the model, a deeper network can show the issue of declination. basically, as the number of layers of the neural network increases, the complexity of situations may get impregnated and sluggishly degrade after a point. as a result, the performance of the model deteriorates both on the training and testing data. f. this declination isn’t a result of overfitting. rather, it may affect the initialization of the network, optimization function, or, more importantly, the problem of evaporating or exploding slant various factors based on types of resnet are as follows resnet-50 architecture: bottleneck design resnet-50 uses a bottleneck design, which reduces the number of parameters and computational cost. 3-layer blocks: resnet-50 uses a stack of 3 layers instead of the earlier 2-layer blocks, forming a 3-layer bottleneck block. higher accuracy: resnet-50 achieves much higher accuracy than the 34-layer resnet model. performance: the 50-layer resnet-50 achieves a performance of 3.8 billion flops. resnet-101 and resnet-152 architecture: large residual networks resnet-101 and resnet-152 are constructed using more pa ge 60 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 57-62, 2025 than 3-layer blocks, enabling deeper networks with lower complexity. lower complexity: despite increased network depth, the 152-layer resnet has much lower complexity (11.3 billion flops) than vgg-16 or vgg-19 nets (15.3/19.6 billion flops). resnet-50 with keras: keras api keras is a popular deep learning api known for its simplicity and ease of use. pre-trained models: keras comes with several pre-trained models, including resnet-50, which can be used for various experiments and applications. technical approaches for validation of the soybean plant and the condition of its roots the technology of computer vision and the integration of the model of resnet is being applied to the farming of soybean crops. the adjustment design for the evaluation supports and recognizes the image data sets and validates the progress of the soybean plants. recommended the iot and resnet model based on a computer vision system is as below block diagram, where the camera is built for extracting the images, and after those images are forwarded to the memory shuttle of memory, which is built into the circuit of iot devices. the memory transferring the images to the computer system, where an algorithm extracts the image-based data of the sets that are recognized for the image segmentation. the block diagram shows the stepwise uses of the iot model and the process of input values. the block diagram also integrates the steps of preprocessing with the help of modules of resnn (simonyan & zisserman, 2014; szegedy et al., 2015). figure 2: model-based to define the iot and resnn for validating data set result and discussion evaluation and analysis of soybean roots using resnn technology renn architecture consists of residual block residual blocks are the main components of the residual neural network. in a classical neural network, the input is transformed by a set of convolutional layers then it is passed to the activation function. in a residual network, the input to the block is added to the output of the block, creating a residual connection. the output of the residual block h(xi) can be represented by: h(xi) = f(xi) + xi f(xi) represents the residual mapping learned by the network. the presence of the identity term x allows the gradient to flow more easily. s connection s connection is a skip connection that helps in forming the residual blocks. skip connection consists of the input of the residual block that is bypassed over the convolutional layer and added to the output of the residual block. st layers resnet architectures are formed by stacking multiple residual blocks together. using these multiple residual blocks together, resnet architecture can be built very deep. versions of resnet with 50,101,152 layers were introduced. global average pooling(gap) resnet architectures typically utilize global average pooling as the final layer before the fully connected layer. gap reduces spatial dimensions to a single value per feature map, providing a compact representation of the entire feature map (zhang et al., 2019). healthy soybean roots have the following characteristics (chen et al., 2020) • depth: soybean roots typically grow to a depth of 2–3 feet, but most of the roots are in the top 6–12 inches of soil. • nodules: a healthy soybean plant should have 6–20 large nodules on the main tap root and smaller nodules on the auxiliary roots. nodules are formed by bacteria in the soil and provide much of the plant’s nitrogen supply. pa ge 61 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 57-62, 2025 • colour: a healthy nodule is red or pink, which indicates active nitrogen fixation. a white or gray nodule is immature and should be checked again in a week. a green, brown, or mushy nodule is dead. • root type: soybeans have both deep, vertical roots and shallow, lateral roots. the deep roots access water from deeper soil layers, while the shallow roots increase the plant’s ability to absorb nutrients from the topsoil. figure 3: images of standardized formation of soybean plant y (output) = x+f(x) renn model for evaluation and image-based analysis the effective and productive extraction of the root image, so the progress and evaluation of the roots help and are more supportive of the progress of the plant. that plant’s progress directly raises the production of the soybean. here are illustrated the key facts for resnn, which is supportive of the evaluation of the soybean plants in the future study and project of iot deployments. • resnets (residual networks) are a variant of deep learning algorithms that are particularly for image recognition and processing tasks. resnets are known for their make to train very deep networks without overfitting • resnets are helpful for detection tasks. key point detection is the task of locating points on an object in an image. for example, detection can be used to locate the eyes, nose, and mouth on a human face. • resnets are well-suited because they can learn to extract from images at different scales. references bello, i., zoph, b., vaswani, a., shlens, j., & le, q. v. (2019). attention augmented convolutional networks. in proceedings of the ieee/cvf international conference on computer vision (pp. 3286–3295). ieee. chen, b., ghiasi, g., liu, h., lin, t.-y., kalenichenko, d., adam, h., & le, q. v. (2020). mnasfpn: learning latency-aware pyramid architecture for object detection on mobile devices. in proceedings of the ieee/cvf conference on computer vision and pattern recognition (pp. 13607–13616). ieee. deng, j., dong, w., socher, r., li, l.-j., li, k., & fei-fei, l. (2009, june 20). imagenet: a large-scale hierarchical image database. in 2009 ieee conference on computer vision and pattern recognition (pp. 248–255). ieee. devries, t., & taylor, g. w. (2017, august 15). improved regularization of convolutional neural networks with considering, farming side s1: images inputs ms : length t1 : root size r_lat, r_nod, r_bra in the process of resnn, the extraction of images is consolidated from the m1 position to the ms position. after that, the land-side-wise extraction is in observation for the same, considering the s1 to s nth side. whereas the root observation key points are evaluated and analyzed to add f(x) = t1+r_lat, similarly add f(y)= t1+r_nod and f(z)= t1+r_bra. the functions f(x) = partially integrated mapping and putting the evaluation steps for states lateral roots, states nodules, and states branchtop root. therefore, output evaluation y = f( x ) +f( y )+f( z ) same way, the changes in image observation, the values are integrated, and the analysis of the present image data set would be responsible for the analysis of the healthy condition of the roots, and also changes in the image data set to observe the desired diseases and progress of the plant. conclusion the observation implementation is still in process for pa ge 62 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 57-62, 2025 cutout. arxiv preprint arxiv:1708.04552. farooq, m., & hafeez, a. (2020, march 31). covid-resnet: a deep learning framework for screening of covid-19 from radiographs. arxiv preprint arxiv:2003.14395. gao, s. h., cheng, m. m., zhao, k., zhang, x. y., yang, m. h., & torr, p. h. s. (2021). res2net: a new multiscale backbone architecture. ieee transactions on pattern analysis and machine intelligence, 43(2), 652–662. he, k., zhang, x., ren, s., & sun, j. (2016). deep residual learning for image recognition. in proceedings of the ieee conference on computer vision and pattern recognition (pp. 770–778). ieee. hu, j., shen, l., & sun, g. (2018). squeeze-and-excitation networks. in proceedings of the ieee conference on computer vision and pattern recognition (pp. 7132–7141). ieee. huang, g., liu, z., van der maaten, l., & weinberger, k. q. (2017). densely connected convolutional networks. in proceedings of the ieee conference on computer vision and pattern recognition (pp. 4700–4708). ieee. krizhevsky, a., sutskever, i., & hinton, g. e. (2017). imagenet classification with deep convolutional neural networks. communications of the acm, 60(6), 84–90. leibe, b., matas, j., sebe, n., & welling, m. (eds.). (2016, september 16). computer vision–eccv 2016: 14th european conference, amsterdam, the netherlands, october 11–14, 2016, proceedings, part iv. springer. li, z., sang, n., chen, k., gao, c., & wang, r. (2018, march 8). learning deep features with adaptive triplet loss for person reidentification. in mippr 2017: pattern recognition and computer vision (vol. 10609, pp. 90–95). spie. lin, s., ji, r., yan, c., zhang, b., cao, l., ye, q., huang, f., & doermann, d. (2019). towards optimal structured cnn pruning via generative adversarial learning. in proceedings of the ieee/cvf conference on computer vision and pattern recognition (pp. 2790–2799). ieee. nam, h., ha, j. w., & kim, j. (2017). dual attention networks for multimodal reasoning and matching. in proceedings of the ieee conference on computer vision and pattern recognition (pp. 299–307). ieee. pham, t. d. (2020). a comprehensive study on the classification of covid-19 on computed tomography with pre-trained convolutional neural networks. scientific reports, 10(1), 1–8. roy, s. k., manna, s., song, t., & bruzzone, l. (2020). attention-based adaptive spectral-spatial kernel resnet for hyperspectral image classification. ieee transactions on geoscience and remote sensing, 59(9), 7831–7843. simonyan, k., & zisserman, a. (2014, september 4). very deep convolutional networks for large-scale image recognition. arxiv preprint arxiv:1409.1556. szegedy, c., liu, w., jia, y., sermanet, p., reed, s., anguelov, d., erhan, d., vanhoucke, v., & rabinovich, a. (2015). going deeper with convolutions. in proceedings of the ieee conference on computer vision and pattern recognition (pp. 1–9). ieee. tan, m., & le, q. (2019, may 24). efficientnet: rethinking model scaling for convolutional neural networks. in international conference on machine learning (pp. 6105–6114). pmlr. tseng, h. y., lee, h. y., huang, j. b., & yang, m. h. (2020). cross-domain few-shot classification via learned feature-wise transformation. arxiv preprint arxiv:2001.08735. wang, z., yu, z., zhao, c., zhu, x., qin, y., zhou, q., zhou, f., & lei, z. (2020). deep spatial gradient and temporal depth learning for face anti-spoofing. in proceedings of the ieee/cvf conference on computer vision and pattern recognition (pp. 5042–5051). ieee. xie, s., girshick, r., dollár, p., tu, z., & he, k. (2017). aggregated residual transformations for deep neural networks. in proceedings of the ieee conference on computer vision and pattern recognition (pp. 1492–1500). ieee. zagoruyko, s., & komodakis, n. (2016, may 23). wide residual networks. arxiv preprint arxiv:1605.07146. zerhouni, e., lányi, d., viana, m., & gabrani, m. (2017, april 18). wide residual networks for mitosis detection. in 2017 ieee 14th international symposium on biomedical imaging (isbi) (pp. 924–928). ieee. zhang, h., wu, c., zhang, z., zhu, y., lin, h., zhang, z., sun, y., he, t., mueller, j., manmatha, r., & li, m. (2022). resnest: split-attention networks. in proceedings of the ieee/cvf conference on computer vision and pattern recognition (pp. 2736–2746). ieee. zhang, k., sun, m., han, t. x., yuan, x., guo, l., & liu, t. (2017). residual networks of residual networks: multilevel residual networks. ieee transactions on circuits and systems for video technology, 28(6), 1303–1314. zhong, x., gong, o., huang, w., li, l., & xia, h. (2019, september 22). squeeze-and-excitation wide residual networks in image classification. in 2019 ieee international conference on image processing (icip) (pp. 395–399). ieee. pa ge 1 pa ge 70 american journal of smart technology and solutions (ajsts) r-shiny web application development for multilayer perceptron state switching model for predicting regimes of time series returns oluwasegun agbailu adejumo1*, omorogbe joseph asemota1, samuel olayemi olanrewaju1 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.5956 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: august 14, 2025 accepted: september 22, 2025 published: october 21, 2025 forecasting regimes in financial time series is complicated by nonlinearity and smallsample limitations, where conventional markov switching models (msms) and regimeswitching autoregressive models often underperform. to address this, we developed an interactive r-shiny web application implementing the recently introduced multilayer perceptron state switching model (mlpssm). the app integrates neural networks to capture nonlinear intra-regime patterns with markov switching structures to identify latent regime transitions. using nigerian exchange rate returns as a case study, the app demonstrated robust performance across diagnostics, estimation, and forecasting. residual checks confirmed that the hybrid approach effectively modeled underlying dynamics, while forecasts achieved lower rmse and mae than baseline msms. the web-based interface further enhances accessibility, enabling both technical and non-specialist users to apply advanced regime-switching methods without coding expertise. the mlpssm web app thus bridges machine learning and econometric modeling, offering a practical, reproducible tool for regime prediction in financial markets. keywords financial time series returns, markov process, mlpssm, r shiny, regime switching 1 department of statistics, faculty of science, university of abuja, abuja, nigeria * corresponding author’s e-mail: agbailuoa@gmail.com introduction financial time series data, such as stock returns, exchange rates, and commodity prices, are often characterized by nonlinearity, volatility clustering, and structural regime shifts. traditional linear time series models, while useful for short-term forecasting, are limited in their ability to capture abrupt changes between economic states such as bull and bear markets or highand low-volatility regimes (hamilton, 1989; krolzig, 1997). to address these limitations, researchers have developed stateswitching models, including the markov switching model (msm) and regime-switching autoregressive models, which explicitly account for stochastic regime shifts and structural breaks. these models have been widely applied in empirical finance to enhance the understanding of market dynamics and improve predictive performance in risk-sensitive environments (ang & timmermann, 2012). despite their popularity, conventional msms and regime-switching autoregressive models demonstrate poor predictive performance in small-sample settings, which is a common challenge in emerging markets and in high-frequency but short-span datasets. under such conditions, parameter estimates become unstable, leading to unreliable regime classification, weak predictive power, and difficulty in distinguishing genuine structural shifts from random noise (psaradakis & spagnolo, 2003; chauvet & potter, 2000). these shortcomings limit the applicability of msms and related models, especially in contexts where robust regime detection is most critical for investors, policymakers, and risk managers. to overcome these challenges, recent research has explored the integration of neural networks into regimeswitching frameworks. neural networks excel at capturing nonlinear patterns, hidden structures, and complex dependencies within financial data that conventional linear models often fail to detect (zhang et al., 1998). by combining the flexibility of machine learning with the interpretability of regime-based approaches, neural network state switching models (nn-ssms) have emerged as promising alternatives for predicting time series regimes (yao et al., 1999). a notable advancement in this field is the study by adejumo et al. (2025), who introduced the multilayer perceptron state switching model (mlpssm), a novel hybrid approach that embeds a neural network within a regime-switching framework. their findings demonstrated that the mlpssm consistently outperforms traditional msms and regime-switching autoregressive models, particularly in small-sample environments and in accurately predicting regime transitions. by addressing the weaknesses of conventional models, the mlpssm provides a more robust and reliable tool for forecasting financial returns under volatile and structurally shifting conditions. however, while adejumo et al. (2025) established the methodological superiority of the mlpssm, its practical application remains constrained by accessibility barriers. financial analysts and policymakers often lack the programming expertise to implement advanced hybrid models, and existing statistical software environments do not provide user-friendly, interactive platforms for realtime regime prediction. in response, the present study focuses on the development of an r-shiny web application for the mlpssm pa ge 71 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 70-79, 2025 introduced by adejumo et al. (2025). r-shiny provides an ideal platform for operationalizing this optimal model into an interactive, scalable, and accessible tool. through this web application, users will be able to upload datasets, configure model parameters, visualize regime-switching dynamics, and generate real-time forecasts without the need for advanced coding. thus, this research makes a dual contribution: methodologically, by extending the work of adejumo et al. (2025) through applicationoriented development of the mlpssm; and practically, by democratizing access to advanced predictive models for financial analysts, researchers, and policymakers in nigeria and other emerging economies where regime shifts and structural breaks are frequent. materials and methods following hamilton (2005), the mssm addresses the hidden states weakness of the hidden markov model (hmm). to describe the mssm, we assume that the number of states (or states) is n=2, i.e. st ω={1,2}. this implies that; for instance, the log returns of financial time series are drawn from distinct normal distributions, depending on what state the hmm is currently in. this would give us the following model to work with: this means that when the state of the hmm for time t is 1, then the expectation of the dependent variable is μ1 and the variance of the innovations is σ1 2, similarly when the state of the hmm for time t is 2, then the expectation of the dependent variable is μ2 and the variance of the innovations is σ2 2 and so on. since the underlying markov chain is hidden one cannot observe what state the hmm is in directly, but only deduce its operation through the observed behaviour of yt. in order to attain the probability law governing the observed data yt a probabilistic model of what causes the change from state st=i to state st=j is required. this can be specified using the transition probabilities of an n=2 state hmm; ρi,j=pr (st=j│st-1=i) i,j∈ω={1,2} ....(2) the transition probability (2) is by the markov property; dependent of the past only through the value of the most recent state. this is one of the central points of the structure of a markov state switching model, i.e. the switching of the states of the underlying hmm is a stochastic process itself. there are several ways to estimate the required parameters of the 2-state mssm given by (1), for instance by using the maximum likelihood estimation method. the multilayer perceptrons state switching model (mlpssm) by adejumo et al. (2025) to improve the regime prediction performance of the famous state switching model described in equation (1) for nonlinear time series data, especially returns, in small sample size contexts, adejumo-agbailu et al. (2025) integrate deep neural networks such as multilayer perceptrons (mlp) with markov two-state switching modelling technique. the adejumo-agbailu et al. (2025) neural network state switching modeling approach combines the strengths of mlp-deep neural networks (i.e., efficient in high complexity datasets) and markov two-state switching capabilities of ensuring both interpretability and predictability of the two nigerian exchange returns’ regimes (i.e., bull and bear states). the models were introduced in two phases. 1st phase: generative network using mlp given time series dataset of yt, the generative network procedure for the dataset include the following: define of training and testing dataset of yk i.e. by setting training dataset at time step tk, and testing dataset at time step tt-k at time step tk, mlp is used to process the input data (i.e. training dataset) such as hk~mlph (yt ). 2nd phase: integration of generative network ht into the markov two-state switching model subsequently, the generative network ht is integrated in the markov two-state switching model alongside the defined training dataset to develop the multilayer perceptrons state switching model (mlpssm). this would give us the following model to work with: ytk~ht)=gt=cst+β1(gt-1-cst-1)+β2 (gt-2-cst-2))+ϵt ....(7) where cst=c0 s0t+c1 s1t+c2 s2t σst 2=σ1 2 s1t+σ2 2 s2t ϵt~i.i.d(0,σst 2) cst is the state dependent mean, σst 2 is state dependent variance and the coefficients are β1or β2; which could be different for different subsamples. the proposal will be to model the state st as the outcome of an unobserved two-state markov chain with st independent of ϵt for all t. the transitions of the st, are presumed to be ergodic and intricate first order markov-process. this means impacts of earlier observation(s) for the gt and state(s) is/are completely captured in the recent gt state(s) observations as represented in (3); ρij=prob(st=j ⁄ st-1) =i) ∀i,j=1,2∑i=1 2ρij =1 ....(8) matrix p captures the probability of switching which is known as a transition matrix; ....(9) where, p11+p21=1, and p12+p22=1 the nearer the probability ρij is to one the longer it takes to shift to the next regime. consider the model given by equation (7), i.e., a markov regime-switching model with 2-regimes. the estimation will be performed using hamilton’s filter, where the main idea is to calculate each state’s filter probabilities by making inferences on each state’s unknown probabilities based on the available information. the mlpssm web application development mlpssm is a web application developed in the shiny environment, leveraging a variety of r libraries and pa ge 72 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 70-79, 2025 packages to provide a comprehensive regimes of time series returns modelling and predicting experience. currently, there are more than 12,000 r packages or libraries available, which are the result of a collaborative project sustained by individuals from different parts of the world and disciplines. the project continues to grow both in the number of packages and in knowledge areas, such as statistics, finance, genetics, network analysis, and data mining, among many others (cruz et al. 2023). this application has been designed to facilitate efficient and effective time series returns data exploration, preliminary analysis, regime modelling and predicting. to ensure optimal functionality and a wide range of features, mlpssm makes use of libraries and packages in r. you can view the application at the following url: https://agbailuoa.shinyapps.io/mlpssm_app/ the user interface of mlpssm has been carefully designed using the shiny package in r, providing a smooth and intuitive experience for users. shiny is an r package that allows the construction of interactive web applications from r scripts. the application is organized into different tabs, each with specific functionalities. the application architecture includes: i. frontend; shiny, shinydashboard for ui (upload, plots, model outputs). ii. backend; neural networks via rsnns::mlp, regimeswitching via mswm::msmfit. iii. support packages; forecast, tseries, lmtest, metrics, ggplot2 for tests, diagnostics, and forecasting. figure 1 presents a flowchart of the mlpssm app. figure 1: flowchart of the application: mlpssm app mlpssm app user interface the user interface of mlpssm app welcomes you to an interactive and enriching experience. here you will find a concise description of the application’s functionalities, figure 2 presents the user interface of the developed mlpssm app. according to figure 2, the application includes tabs for upload data, a time series, a model output, residuals, and a forecast. the upload data tab allows users to load returns time series, i.e. rt=ln(rt⁄rt-1 ), in csv format. the time series tab presents users with the time series plot and preliminary tests of the series such as the adf test (stationarity test) and bds test (non-linearity test). the model estimation tab allows users to estimate the novel mlpssm and provides the estimated model’s transition probability plots. the residuals tab allows users to assess the diagnosed residuals of the estimated mlpssm, including the residuals plot, ljung-box test, and shapirowilk test. lastly, the forecast tab enables users to forecast or predict the returns-regime means and the model performance metrics mlpssm app evaluation this section presents and discusses the deployment and evaluation of the mlpssm web app using the sample size 30 of daily nigeria exchange rate returns. figure 3 presents the “data upload” interface of the app using the aforementioned dataset. the app successfully ingested the nigeria exchange rate (usd/₦) daily data from cbn (11th june–23rd july 2025). the data upload interface was limited to csv files, ensuring standardized inputs. the app extracted the second column of the dataset as the time series and visualized it under the time series tab. this confirms the reactivity of the app, as the uploaded data was immediately transformed into plots and summary previews for the user. by visualizing the exchange rate returns, the app provides users with an intuitive entry point before model estimation. subsequently, the time series is visualized in the “time series” tab, just below the “upload & settings” tab. the plotoutput function is used to display the graph of the time series in levels. the reactivity of the application is reflected in the outputs, which are the results (numeric values, plots) received by the interface from the server.r. in our case, the result is a graph and it is inserted using the plotoutput() function (figure 4). the inclusion of augmented dickey-fuller (adf) and bds (brock-dechert-scheinkman) tests within the time pa ge 73 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 70-79, 2025 figure 2: the user interface of mlpssm web app figure 3: data upload interface of the mlpssm web app figure 4: time series plot of the nigeria exchange rate returns using the time series tab of the app series tab allowed for formal pre-model diagnostics. adf test: evaluates whether the returns are stationary. if the null hypothesis of a unit root is rejected, the series can be considered stationary. in most exchange rate returns, stationarity is expected after transformation into returns. bds test: evaluates non-linearity and independence of residuals. significant results indicate the presence of hidden structure, justifying the need for nonlinear hybrid models like mlpssm. in this evaluation, the results confirmed stationarity and non-linearity, validating the use of a nonlinear, regimeswitching hybrid model. pa ge 74 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 70-79, 2025 figure 5: adf test results of the nigeria exchange rate returns using the time series tab figure 6: bds test results of the nigeria exchange rate returns using the time series tab pa ge 75 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 70-79, 2025 figure 7: mlpssm estimation for the nigeria exchange rate returns using the model estimation tab figure 8: regimes transition probabilities of mlpssm estimated of the nigeria exchange rate returns using the model estimation tab the model estimation tab provided the estimated parameters of the mlpssm (a combination of neural networks for nonlinear fitting and markov switching models for regime identification). estimation results (figure 7): these show the fitted coefficients and transition probabilities of the model. the estimates suggest the model successfully learned from the exchange rate returns. transition probabilities (figure 8): the graph displayed probabilities of switching between regimes (bear vs. bull). high self-transition probabilities suggest persistence within regimes, while non-trivial switching probabilities capture regime dynamics. this indicates that the app correctly implements the regime identification logic, a major improvement over conventional msms and arssms, which often perform poorly in small samples. pa ge 76 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 70-79, 2025 figure 9: residuals plot of the estimated mlpssm using the residuals tab figure 10: residuals diagnostics of the estimated mlpssm using the residuals tab pa ge 77 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 70-79, 2025 figure 11: forecast plot of the estimated mlpssm using the forecast tab figure 12: forecast estimate of the estimated mlpssm using the forecast tab pa ge 78 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 70-79, 2025 figure 13: estimated mlpssm performance metrics using the forecast tab afterward, the residuals tab provided three diagnostic checks: residual plots, ljung-box test, and shapiro-wilk test. residual plot (figure 9): shows whether errors fluctuate randomly around zero. a fairly random pattern indicates good fit. ljung-box test (figure 10): checks if residuals are autocorrelated. a non-significant result suggests residual independence. shapiro-wilk test (figure 10): tests normality of residuals. this ensures validity of statistical inferences. together, these diagnostics confirm whether the hybrid model adequately captures the dynamics in the data. in this evaluation, the residual diagnostics suggested that the mlpssm provides a statistically adequate fit, with minimal autocorrelation and approximate normality. the forecast tab generated three outputs: forecast plot (figure 11): displayed short-horizon projections of exchange rate returns. forecast estimates with regimes (figure 12): highlighted regime-dependent forecasts (bear vs. bull), making it useful for risk management and trading strategy design. performance metrics (figure 13): reported forecast accuracy measures (rmse, mae). these quantitative metrics confirmed the predictive strength of the mlpssm in small-sample forecasting. the hybrid model outperformed conventional regimeswitching models in terms of both forecast precision and regime classification accuracy, aligning with adejumo et al. (2025)’s theoretical contributions. the deployment and evaluation of the mlpssm web app demonstrated, technical correctness (i.e. proper integration of reactivity, diagnostics, regime estimation, and forecasting), practical relevance (i.e. useful outputs for researchers, policymakers, and investors in financial markets), methodological novelty (i.e. the app operationalizes adejumo et al. (2025)’s neuralnetwork-based hybrid model, overcoming weaknesses of traditional msms in small samples), and user accessibility (i.e. with simple tabs and automated tests, non-technical users can upload data, check assumptions, run hybrid models, and obtain forecasts). discussion the evaluation of the multilayer perceptron state switching model (mlpssm) web application demonstrates its capacity to address long-standing methodological and practical challenges in regime prediction within financial time series. conventional markov switching models (msms) and regime-switching autoregressive models (arssms) have often been limited in their predictive capacity, especially under small-sample conditions, where parameter instability and regime misclassification become prominent issues (hamilton, 1989; krolzig, 1997). recent studies, including adejumo et al. (2025), have argued that hybrid frameworks combining neural networks with regime-switching methods offer a robust alternative by integrating nonlinear approximation with probabilistic regime identification. the performance of the app across data input, model estimation, residual diagnostics, and forecasting validates these theoretical expectations. pre-model diagnostics confirmed both stationarity and non-linearity in the nigerian exchange rate returns, justifying the application of a nonlinear, regime-dependent framework. the mlpssm successfully captured hidden regime dynamics through a two-stage estimation; first modeling nonlinear intraregime patterns using multilayer perceptrons (mlps), and subsequently applying a markov-switching structure to the residuals. the transition probability estimates highlighted persistent regime behaviors, while allowing for realistic switching, aligning with empirical evidence in financial markets where volatility clustering and regime persistence are common (hamilton & susmel, 1994). residual diagnostics further reinforced the adequacy of the hybrid approach. the ljung-box and shapiro-wilk results indicated that the residuals were largely uncorrelated and approximately normal, suggesting that the hybrid model extracted most of the signal embedded in the series. these results compare favorably to conventional msms, which often leave strong autocorrelation structures unmodeled, thereby undermining their predictive reliability (maheu & mccurdy, 2000). the forecasting module provided perhaps the most critical evidence of the app’s utility. both the point forecasts and the regime-dependent projections showed strong predictive accuracy, as reflected in the rmse and mae metrics. these results substantiate the claim by adejumo et al. (2025) that the mlpssm can outperform baseline regime-switching methods, particularly in small samples where neural networks are able to flexibly capture nonlinear structures that traditional parametric forms miss. the integration of forecasts with regime classifications also provides added interpretive value for practitioners in risk management and policy analysis, who pa ge 79 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 70-79, 2025 require not just point estimates but also insights into underlying market conditions. beyond methodological contributions, the deployment of this model in a shiny web application represents an important step toward accessibility and reproducibility. by embedding advanced methods in an interactive interface, the app democratizes the use of sophisticated econometric techniques, enabling both technical and non-technical users to conduct diagnostics, fit models, and generate forecasts. this aligns with recent calls in computational economics and data science for tools that bridge methodological rigor and usability (chambers & hastie, 1992; varian, 2014). conclusions in summary, the mlpssm web app provides empirical evidence for the superiority of neural-network-enhanced regime-switching approaches over conventional methods. it not only validates the theoretical advances proposed by adejumo et al. (2025) but also operationalizes them in a user-friendly platform that can be applied to diverse time series forecasting tasks. these findings highlight both the academic novelty and the practical relevance of integrating machine learning with classical state-space and regime-switching frameworks. from a policy and practical perspective, the implications are significant. for policymakers in nigeria and other emerging economies, the ability to identify and forecast financial regimes in exchange rate markets provides a valuable tool for anticipating volatility and implementing timely interventions. regulators can apply the app’s diagnostic and forecasting features to monitor systemic risks, while financial institutions can leverage the regime predictions to inform hedging strategies and risk-adjusted investment decisions. moreover, by operationalizing advanced econometric and machine learning techniques within a shiny web app, this study bridges the gap between methodological innovation and practical accessibility. the interactive dashboard ensures that both technical experts and non-specialist users can upload data, conduct diagnostics, fit models, and interpret results without extensive coding skills. this democratization of advanced forecasting methods supports capacity building within financial institutions, academic research, and policy circles. nevertheless, limitations remain. the two-stage estimation approach assumes that regime structure lies entirely in residual dynamics, which may overlook deeper joint interactions. additionally, small-sample robustness, while improved over msms, still warrants further exploration through regularization techniques and bayesian extensions. future research should also extend the web app to multivariate series and high-frequency data, as well as integrate real-time data streaming for dynamic monitoring. in conclusion, the mlpssm web app provides a practical and innovative tool for regime prediction and forecasting in financial time series. it validates the theoretical contributions of adejumo et al. (2025) and extends them into a usable platform with direct policy and market relevance. the study contributes both to the advancement of econometric methodology and to the provision of applied solutions for managing uncertainty in financial markets. references adejumo, o. a., asemota, o. j. & olanrewaju, s. o. (2025). novel neural network state switching models for returns predicting with regime switching: a monte carlo’s simulation approach. american journal of applied statistics and economics, 4(1). ang, a., & timmermann, a. (2012). regime changes and financial markets. annual review of financial economics, 4, 313–337. chambers, j. m., & hastie, t. j. (1992). statistical models in s. wadsworth & brooks/cole. chauvet, m., & potter, s. (2000). coincident and leading indicators of the stock market. journal of empirical finance, 7(1), 87–111. cruz, t., jimenez, f. g., bravo, a. r. q., & ander, e. (2023). dataxplorefines: generalized data for informed decision, making, an interactive shiny application for data analysis and visualization. arxiv preprint arxiv:2307.11056. hamilton, j. d. (1989). a new approach to the economic analysis of nonstationary time series and the business cycle. econometrica, 57(2), 357–384. https://doi. org/10.2307/1912559 hamilton, j. d., & susmel, r. (1994). autoregressive conditional heteroskedasticity and changes in regime. journal of econometrics, 64(1–2), 307–333. https://doi. org/10.1016/0304-4076(94)90067-1 krolzig, h.-m. (1997). markov-switching vector autoregressions: modelling, statistical inference, and application to business cycle analysis. springer. maheu, j. m., & mccurdy, t. h. (2000). identifying bull and bear markets in stock returns. journal of business & economic statistics, 18(1), 100–112. https://doi.org/ 10.1080/07350015.2000.10524846 psaradakis, z., & spagnolo, n. (2003). on the determination of the number of regimes in markovswitching autoregressive models. journal of time series analysis, 24(2), 237–252. varian, h. r. (2014). big data: new tricks for econometrics. journal of economic perspectives, 28(2), 3–28. https:// doi.org/10.1257/jep.28.2.3 yao, j., tan, c. l., & poh, h. l. (1999). neural networks for technical analysis: a study on klci. international journal of theoretical and applied finance, 2(2), 221–241. zhang, g., patuwo, b. e., & hu, m. y. (1998). forecasting with artificial neural networks: the state of the art. international journal of forecasting, 14(1), 35–62. pa ge 1 pa ge 32 american journal of smart technology and solutions (ajsts) the nexus between cognitive absorption and ai literacy of college students as moderated by sex brandon nacua obenza1*, liam e. go1, jofrance ardrian m. francisco1, evann ernest t. buit1, frande vier b. mariano1, henry l. cuizon jr1, alliah jane d. cagabhion1, karl axl james l. agbulos1 volume 3 issue 1, year 2024 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v3i1.2603 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: february 20, 2024 accepted: march 25, 2024 published: march 29, 2024 this study examines how cognitive absorption and ai literacy are related among college students, specifically looking at how sex moderates this link. the study uses a quantitative research strategy and a non-experimental correlational approach. data was collected through google forms utilizing modified questions designed for ai literacy and cognitive absorption. g*power 3.2 was used for power analysis to determine the necessary sample size for the investigation. 372 college students from different higher education institutions in region xi were selected to take part in the study by stratified random sampling. reliability and validity tests, including cronbach’s alpha, average variance extracted (ave), and heterotrait-monotrait ratio (htmt), were performed on the dataset before undertaking moderation analysis. cognitive absorption was identified as a key predictor of ai literacy, showing a substantial impact size of 0.417. the moderating effect of sex, although statistically significant, had a minor effect size of 0.011. the corrected r-squared value of 0.378 indicates that the model, with all covariates, accounts for 37.8% of the variance in ai literacy. keywords cognitive absorption, ai literacy, moderation analysis, smartpls, philippines 1 university of mindanao, davao city, philippines * corresponding author’s e-mail: bobenza@umindanao.edu.ph introduction ai literacy is the essential talent of effectively engaging with and critically assessing ai technology in today’s tech-centric society (long & magerko, 2020). it involves comprehending, utilizing, assessing, and dealing with ethical concerns associated with ai (ng et al., 2021). disparities exist in ai literacy across children from various socioeconomic and cultural backgrounds (druga et al., 2019). initiatives to improve understanding of ai involve a middle school program designed to educate kids on ai for them to become knowledgeable citizens and discerning users of ai (lee et al., 2021), and the creation of an ai-robotics tool to advance ai literacy in underdeveloped nations (eguchi, 2021). ai has the capacity to greatly improve communication abilities in english language learners, with a primary focus on writing, reading, and vocabulary development in language education. various studies have investigated the ai literacy of college students, yielding favorable results. kong et al. (2021) and lee et al. (2021) discovered that students from various backgrounds can gain a conceptual knowledge of ai. they also observed that ai literacy education can enhance students’ ethical awareness of ai. ng et al. (2022) and lee et al. (2021) showed how pedagogical methods, including digital story writing, can enhance ai literacy in primary and middle school children. juma (2021) discovered that although higher education students acknowledge the significance of ai in education, they possess little knowledge and understanding of it. these results emphasize the necessity of ongoing initiatives to improve ai literacy among college students. cognitive absorption, a high level of engagement with software, has been shown to strongly correlate with digital literacy in secondary school children (canan güngören et al., 2022). this is especially important in the realm of ai literacy, which is seen as a modern form of cognitive intelligence (wang & lu, 2023). cognitive absorption’s impact on establishing user trust and enhancing experience in human-machine interactions has been investigated by balakrishnan & dwivedi (2021). ai literacy is associated with metacognition and the capacity to predict an unpredictable future (yi 2021). literacy has a substantial correlation with cognitive performance in well-educated older adults, as shown by barnes et al. (2004). research has examined how cognitive absorption is influenced by computer playfulness and perceived quality in the context of fun-oriented information systems utilization (weniger & löbbecke, 2011). perceived affective quality has been suggested as a precursor to cognitive absorption, significantly influencing it (ping zhang et al., 2006). several research have examined people’s opinions and use of artificial intelligence (ai) (obenza et al., 2023a, 2023b). a notable deficiency in the existing research is the investigation of ai literacy in college students and the factors that may influence it. current literature on ai literacy, namely on cognitive absorption, still has considerable inadequacies and restrictions despite the increasing study in this area. long and magerko (2020) and ng (2021) have suggested competencies and design considerations for ai literacy, but they have not specifically addressed sex disparities. therefore, there is a notable absence of focus on sex-specific patterns or issues in this area, highlighting the need for more research. research pa ge 33 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 32-39, 2024 must investigate sex discrepancies in the relationship between cognitive absorption and learning outcomes in ai literacy. the study aimed to examine how cognitive absorption and ai literacy of college students are related, with consideration to how this relationship may be influenced by their gender. theoretical framework cognitive absorption theory (cat) is a useful framework for examining the complex connection between cognitive absorption, ai literacy, and gender among college students. cat, first introduced by agarwal and karahanna in 2000, explains the extreme involvement humans display when using technology, marked by strong concentration, altered perception of time, and profound immersion (agarwal & karahanna, 2000). the hypothesis suggests that cognitive absorption is a condition in which individuals are completely engrossed in technology-mediated activities, surpassing awareness of their surroundings and fostering a heightened sense of cognitive engagement (agarwal & karahanna, 2000). cat emphasizes that cognitive absorption is evident in different aspects, including absorption, attention, control, and immersion (agarwal & karahanna, 2000). absorption refers to how deeply individuals engage in a specific task, while focus indicates the level of concentration directed towards the activity. control refers to the sense of control individuals have over the technological interface, while immersion indicates the degree of temporal distortion and detachment from the immediate world (agarwal & karahanna, 2000). cognitive absorption is especially important in the realm of ai literacy for college students. when people interact with ai technologies, their level of cognitive absorption can greatly impact how they learn and understand airelated information and abilities. cat provides a detailed framework for comprehending the cognitive processes involved in the advancement of ai literacy. it explains how individuals’ thorough interaction with ai interfaces influences their cognitive schemas and knowledge structures (agarwal & karahanna, 2000). moreover, the influence of gender as a moderator should be taken into account in the context of cat. recent findings indicate that there may be gender differences in how cognitive absorption is experienced and its outcomes, with research showing varying levels of technological involvement and immersion between men and women (agarwal & karahanna, 2000). this study aims to investigate how gender, as a moderating variable, affects the relationship between cognitive absorption and ai literacy in college students. it tries to uncover potential gender-related distinctions in technology adoption, engagement, and proficiency. cognitive absorption theory offers a strong theoretical basis for studying the relationship between cognitive absorption, ai literacy, and gender in college students. cat provides valuable insights into how individuals acquire and internalize ai-related knowledge by explaining the cognitive mechanisms involved in technology engagement and immersion. this enriches our understanding of the interaction between cognition, technology, and gender. methods and materials the study utilized a quantitative research strategy, specifically a non-experimental correlational approach, to investigate how sex influences the connection between cognitive absorption and ai literacy. quantitative research, according to creswell & creswell (2022), is a methodical examination of factual concepts through the analysis of variable relationships. this method allows for the measurement of variables using instruments, making it easier to apply statistical tools for data analysis. ramayah et al. (2017) used a moderating variable (mv) to explain how the predictor’s effect is impacted by the criterion. this factor is crucial for thoroughly examining the relationship between the predictor and criteria variables. the mv does not directly affect the predictor, but it can influence the intensity and direction of the association between the basic components. the ai literacy scale, a 5-point scale with 12 items, was modified from wang et al.’s research in 2022 to measure people’s daily interaction with, understanding of, and assessment of ai technology. the study used the cognitive absorption scale, a 7-point scale with 20 items created by agarwal and karahanna in 2000. the survey was conducted online. the participants were college students from region 11 who were chosen using stratified random sampling. a power analysis was performed using g*power 3.1.9.6 (faul et al., 2007) before data collection. it was found that a sample size of n = 89 was needed to obtain 80% power for detecting a medium effect (f2 = 0.15) at a significance level of α = 0.05. the sample size of n = 372 exceeded the minimum requirement, which strengthened the study’s ability to investigate the complex interactions between the variables. various strategies were used to guarantee the validity and reliability of the measurement models. the measurements used were average variance extracted (ave) for assessing convergent validity, heterotrait-monotrait ratio (htmt) for evaluating discriminant validity, and cronbach’s alpha for determining internal consistency. descriptive statistics, including standard deviation and mean, were computed using jamovi software version 2.0 to describe ai literacy and cognitive absorption. the bootstrapping standardized algorithm was used using smarpls 4.0 software to assess the postulated moderation model. the study aims to thoroughly investigate how sex moderates the link between cognitive absorption and ai literacy among college students using rigorous approaches. hypothesis sex has a significant moderating effect on the relationship between cognitive absorption and ai literacy of college students. pa ge 34 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 32-39, 2024 results and discussion establishing the validity and reliability of the measurement model is paramount when conducting research utilizing moderating analysis, as emphasized by hair et al. (2019). prior to evaluation, potential gaps in some items were addressed to ensure the robustness of the research instrument. table 1 presents the assessment of reliability and validity, conducted through the utilization of cronbach’s alpha, a widely accepted measure of internal consistency for questionnaires (mashingaidze et al., 2021). the obtained cronbach’s alpha values for ai literacy (0.898) and cognitive absorption (0.942) exceeded the benchmark of 0.7, indicating strong internal consistency and validity of the questionnaires (taber, 2017). as per the recommendations established by diamantopoulos et al. (2001) and drolet and morrison (2001), cronbach’s alpha values of 0.60 to 0.70 are considered acceptable, while values between 0.70 and 0.90 are deemed tolerable to good. both factors had values exceeding 0.70, confirming the instrument’s dependability in measuring the constructs of interest. the cronbach’s alpha values did not surpass 0.95, suggesting that redundancy within the factors was not a problem. convergent validity was evaluated by calculating the average variance extracted (ave). the average values for ai literacy (0.586) and cognitive absorption (0.575) above the recommended threshold of 0.5, as proposed by fornell and larcker (1981) and hair et al. (2019). ave scores of 0.50 or higher suggest that the construct explains 50% or more of the variability in its components, which supports the validity of the measurement model. discriminant validity, a critical component of validating measurements, was assessed by heterotrait-monotrait (htmt) ratios. the paired ratios varied between 0.011 and 0.696, showing good discriminant validity as none surpassed the threshold of 0.85 suggested by henseler et al. (2015). the results confirm that the instrument used is valid and trustworthy for evaluating the constructs being studied. table 1: construct validity and reliability variables cronbach's alpha average variance extracted (ave) ai literacy 0.898 0.586 cognitive absorption 0.942 0.575 discriminant validity heterotrait-monotrait ratio (htmt) ai literacy <-> cognitive absorption 0.648 ai literacy <-> sex 0.104 ai literacy <-> sex x cognitive absorption 0.350 sex <-> cognitive absorption 0.070 sex <-> sex x cognitive absorption 0.011 cognitive absorption <-> sex x cognitive absorption 0.696 analyzed data from 372 respondents provided significant insights into the degrees of ai literacy and cognitive absorption among college students, as detailed in table 2. the average ai literacy score was 3.43, suggesting a high degree of ai literacy among students from different universities. this discovery aligns with prior research conducted by kong et al. (2022) and kong et al. (2021), indicating elevated levels of ai literacy in students attending universities in hong kong. nevertheless, the results also bring to mind the concerns highlighted by anderson and anderson (2006) regarding security, privacy, and biases linked to advanced ai knowledge. ai literacy was evaluated based on four subfactors: awareness, usage, evaluation, and ethics. the subfactors, as outlined by wang et al. (2022), explain many aspects of ai literacy. the average values for each subfactor were: awareness (3.27), usage (3.40), evaluation (3.64), and ethics (3.41). the results indicate a considerable level of skill in all subcategories, consistent with kong et al.’s (2022) research on ethics, accessibility, and expertise in ai literacy among university students. the study found that cognitive absorption had a mean value of 5.04, suggesting a moderate to high level of cognitive absorption in college students. the components of cognitive absorption, such as temporal dissociation, focused immersion, heightened enjoyment, control, and curiosity, showed average values between 4.73 and 5.52. the results support prior research conducted by balakrishnan and dwivedi (2021), emphasizing the connections among cognitive absorption, trust, and experience in interactions between humans and machines. college students’ proficiency in ai literacy and cognitive absorption varies. research conducted by yang et al. (2022) and banele (2023) emphasizes the importance of increasing awareness and implementation of mobile learning, along with enhancing evaluation tools for information literacy and individualized learning suggestions. dergunova et al. (2022) highlighted the importance of addressing authenticity concerns, cognitive anxieties, and the necessity for specific interventions to improve ai literacy and cognitive absorption. zastudil et al. (2023) highlighted concerns regarding excessive dependence on ai and the possible negative consequences. pa ge 35 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 32-39, 2024 cognitive absorption, a state of deep involvement with technology, has been found to be influenced by human-to-machine interaction (balakrishnan & dwivedi, 2021). this absorption is crucial in the development of ai literacy, which is considered a form of cognitive intelligence (wang & lu, 2023). ai literacy, in turn, is linked to metacognition and the ability to anticipate the future (yi & park, 2021). the predictive ability of cognitive skills, including working memory, is also important in literacy (alloway & gregory, 2013). figure 1 shows the link between cognitive absorption and ai literacy of college students as moderated by sex. the path from cognitive absorption to ai literacy is strong and positive (0.691) and highly significant (p < 0.001). this indicates a robust direct effect of cognitive absorption on ai literacy, meaning that as college students’ cognitive absorption increases, their ai literacy also tends to increase. the path from the interaction term (sex x cognitive absorption) to ai literacy is negative (-0.201) and significant (p = 0.049). this suggests that sex does moderate the relationship between cognitive absorption and ai literacy, but it has a small effect, meaning that the increase in ai literacy associated with cognitive absorption is less pronounced for female compared to male. table 2: status of college students’ ai literacy and cognitive absorption variables mean sd description cognitive absorption 5.04 0.94 moderately high temporal dissociation 5.52 1.24 high focused immersion 4.90 1.02 moderately high heightened enjoyment 4.82 0.95 moderately high control 4.73 1.09 moderately high curiosity 5.22 1.37 moderately high ai literacy 3.43 0.54 high awareness 3.27 0.62 moderate usage 3.40 0.63 moderate evaluation 3.64 0.77 high ethics 3.41 0.66 high figure 1: moderation analysis results from smartpls the results indicate that the interaction effect of sex and cognitive absorption on ai literacy is significant, with a path coefficient of -0.201 and a p-value of 0.049, supporting hypothesis 1. this suggests that sex moderates the relationship between cognitive absorption and ai literacy among college students, albeit with a small effect size (f2 = pa ge 36 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 32-39, 2024 0.016). ai literacy’s r-square value is 0.378, indicating that the model explains approximately 37.8% of the variance in ai literacy, with cognitive absorption having a substantial effect size (f2 = 0.417) on ai literacy compared to the minimal effect of sex alone (f2 = 0.011). this is a moderate level of explanatory power, suggesting other factors also play a role in determining ai literacy levels. the interaction plot shows the relationship between table 3: moderating effect, r-square values, and effect sizes hypothesis path coefficient (b) sample mean standard deviation (stdev) t statistics p value remark sex x cognitive absorption -> ai literacy -0.201 -0.205 0.102 1.97 0.049 h1 is supported r-square r-square adjusted ai literacy 0.378 0.373 effect sizes (f2) cognitive absorption -> ai literacy 0.417 sex -> ai literacy 0.011 sex x cognitive absorption -> ai literacy 0.016 cognitive absorption and ai literacy, differentiated by sex. the red line represents male respondents, and the green line represents female respondents. the lines are not parallel, which indicates that there is indeed a moderating effect of sex on the relationship between cognitive absorption and ai literacy. the positive slope of both lines suggests that as cognitive absorption increases, so does ai literacy for both sexes. however, the red line is steeper, indicating a stronger relationship between cognitive absorption and ai literacy for the group represented by males. this corroborates the findings of shashaani (1997), shashaani and khalili (2001), dzandu et al. (2016), and nouraldeen (2022) who explored the attitudes and adoption of college students towards ai and other related technology revealed intriguing dynamics, with sex playing a significant role. they have found that male students exhibit a more positive attitude or adoption towards ai/technology, spending more time on computer and mobile devices and demonstrating greater technological experience, knowledge, and awareness. figure 2: simple slope analysis (0=male and 1=female) pa ge 37 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 32-39, 2024 conclusion the research results emphasize the importance of cognitive absorption in developing ai literacy among college students, influenced by gender. the notion of cognitive absorption suggests that intense involvement in an activity results in improved learning outcomes, as seen by the significant correlation between cognitive absorption and ai literacy. the association between cognitive absorption and ai literacy is influenced by sex, as shown by a substantial interaction effect, indicating that the impact of cognitive absorption on ai literacy differs between males and females. the model explains around 37.8% of the variation in ai literacy, emphasizing the significant impact of cognitive absorption. it also suggests the existence of additional components not covered in this study. the results show that educational tactics in ai should be customized to reflect both cognitive engagement and the moderating effect of sex. though studying cognitive absorption and ai literacy among college students provides important information, some issues need to be addressed in future studies. to begin with, this research focuses mainly on sex as a moderator, leaving out possible influences such as culture, education, and social status. an intersectional analysis including these factors would probably give a deeper explanation for the complex forces causing ai literacy and cognitive immersion among university students. furthermore, the use of google forms for data collection may result in a digital literacy bias that excludes those who lack internet prowess. further studies can be done on how the students see ai and look at some risks and morality issues that accompany this technology in education. references agarwal, r., & karahanna, e. (2000). time flies when you’re having fun: cognitive absorption and beliefs about information technology usage. management information systems quarterly, 24(4), 665. https://doi. org/10.2307/3250951 agarwal, r., sambamurthy, v., & stair, r. (1997). cognitive absorption and the adoption of new information technologies. proceedings academy of management, 1997(1), 293–297. https://doi. org/10.5465/ambpp.1997.4983719 alloway, t. p., & gregory, d. (2013, january). the predictive ability of iq and working memory scores in literacy in an adult population. international journal of educational research, 57, 51–56. https://doi. org/10.1016/j.ijer.2012.10.004 anderson, m., & anderson, s. (2006). guest editors’ introduction: machine ethics. ieee intelligent systems, 21(4), 10–11. https://doi.org/10.1109/mis.2006.70 balakrishnan, j., & dwivedi, y. k. (2021, february 24). role of cognitive absorption in building user trust and experience. psychology & marketing, 38(4), 643– 668. https://doi.org/10.1002/mar.21462 banele, s. (2023). students cognitive discrepancies in artificial intelligences utilization: a case of higher learning institutions. socioedu: sociological education, 4(2), 53-61. https://doi.org/10.59098/socioedu.v4i2.1203 barnes, d. e., tager, i. b., satariano, w. a., & yaffe, k. (2004, april 1). the relationship between literacy and cognition in well-educated elders. the journals of gerontology series a: biological sciences and medical sciences, 59(4), m390–m395. https://doi.org/10.1093/ gerona/59.4.m390 canan güngören, z., gür erdoğan, d., kaya uyanik, g., & demirtaş tolaman, t. (2022, november 1). the relationship between cognitive absorption and digital literacy skills among secondary school students. participatory educational research, 9(6), 113– 129. https://doi.org/10.17275/per.22.131.9.6 creswell, j. w. (2022). research design: qualitative, quantitative, and mixed methods approaches, 3rd ed. research design: qualitative, quantitative, and mixed methods approaches, 6th ed. https://psycnet.apa. org/record/2008-13604-000 dergunova, y., aubakirova, r. z., yelmuratova, b. z., gulmira, t. m., yuzikovna, p. n., & antikeyeva, s. (2022). artificial intelligence awareness levels of students. international journal of emerging technologies in learning (ijet), 17(18), 26–37. https://doi. org/10.3991/ijet.v17i18.32195 diamantopoulos, a. and winklhofer, h. m. (2001). index construction with formative indicators: an alternative to scale development. journal of marketing research 38(2), 269-277. https://www.jstor.org/ stable/1558630 drolet, a., & morrison, d. g. (2001). do we really need multiple-item measures in service research? journal of service research, 3(3), 196–204. https://doi. org/10.1177/109467050133001 druga, s., vu, s. t., likhith, e., & qiu, t. (2019, march 9). inclusive ai literacy for kids around the world. proceedings of fablearn 2019. https://doi. org/10.1145/3311890.3311904 dzandu, m. d., boateng, h., agyemang, f. g., & quansah, f. (2016). social media adoption among university students: the role of gender, perceived usefulness and perceived ease of use. international journal of social media and interactive learning environments, 4(2), 124. https://doi.org/10.1504/ijsmile.2016.077584 eguchi, a. (2021). ai-robotics and ai literacy. education in & with robotics to foster 21st-century skills, 75– 85. https://doi.org/10.1007/978-3-030-77022-8_7 faul, f., erdfelder, e., lang, a. g., & buchner, a. (2007). g*power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. behavior research methods, 39(2), 175–191. https://doi.org/10.3758/bf03193146 fornell, c., & larcker, d. f. (1981). evaluating structural equation models with unobservable variables and measurement error. journal of marketing research, 18(1), 39. https://doi.org/10.2307/3151312 hair, j. f., risher, j. j., sarstedt, m., & ringle, c. m. (2019). pa ge 38 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 32-39, 2024 when to use and how to report the results of plssem. european business review, 31(1), 2–24. https:// doi.org/10.1108/ebr-11-2018-0203 henseler, j., ringle, c. m., & sarstedt, m. (2015). a new criterion for assessing discriminant validity in variancebased structural equation modeling. journal of the academy of marketing science, 43(1), 115–135. https://doi.org/10.1007/s11747-014-0403-8 juma, h. (2021, june 30). artificial intelligence: higher education students’ knowledge and understanding. towards excellence, 834–841. https:// doi.org/10.37867/te130268 kong, s. c., cheung, w. m., & zhang, g. (2021). evaluation of an artificial intelligence literacy course for university students with diverse study backgrounds. computers & education: artificial intelligence, 2, 100026. https://doi.org/10.1016/j.caeai.2021.100026 kong, s. c., cheung, w. m., & zhang, g. (2022). evaluating artificial intelligence literacy courses for fostering conceptual learning, literacy and empowerment in university students: refocusing to conceptual building. computers in human behavior reports, 7, 100223. https://doi.org/10.1016/j. chbr.2022.100223 kong, s. c., man-yin cheung, w., & zhang, g. (2021). evaluation of an artificial intelligence literacy course for university students with diverse study backgrounds. computers and education: artificial intelligence, 2, 100026. https://doi.org/10.1016/j.caeai.2021.100026 kong, s., kong, h., he, zhang, g., & cheung, m. (2022). pedagogical delivery and feedback for an artificial intelligence literacy programme for university students with diverse academic backgrounds: flipped classroom learning approach with projectbased learning. lee, i., ali, s., zhang, h., dipaola, d., & breazeal, c. (2021, march 3). developing middle school students’ ai literacy. proceedings of the 52nd acm technical symposium on computer science education. https://doi. org/10.1145/3408877.3432513 liang, j. c., hwang, g. j., chen, m. r. a., & darmawansah, d. (2021, july 28). roles and research foci of artificial intelligence in language education: an integrated bibliographic analysis and systematic review approach. interactive learning environments, 31(7), 4270–4296. https://doi.org/10.1080/10494820.2021.1958348 long, d., & magerko, b. (2020, april 21). what is ai literacy? competencies and design considerations. proceedings of the 2020 chi conference on human factors in computing systems. https://doi. org/10.1145/3313831.3376727 mashingaidze, mugove & phiri, maxwell & bomani, mapeto. (2021). the influence of strategy formulation practices on the perceived financial performance of small and medium enterprises: the zimbabwean experience. the southern african journal of entrepreneurship and small business management, 13, 1-11. 10.4102/sajesbm.v13i1.343. ng, d. t. k., leung, j. k. l., chu, k. w. s., & qiao, m. s. (2021, october). ai literacy: definition, teaching, evaluation and ethical issues. proceedings of the association for information science and technology, 58(1), 504–509. https://doi.org/10.1002/pra2.487 ng, d. t. k., luo, w., chan, h. m. y., & chu, s. k. w. (2022). using digital story writing as a pedagogy to develop ai literacy among primary students. computers and education: artificial intelligence, 3, 100054. https:// doi.org/10.1016/j.caeai.2022.100054 nouraldeen, r. m. (2022). the impact of technology readiness and use perceptions on students’ adoption of artificial intelligence: the moderating role of gender. development and learning in organizations, 37(3), 7–10. https://doi.org/10.1108/dlo-07-2022-0133 obenza, b. n., baguio, j. s. i. e., bardago, k. m. w., granado, l. b., loreco, k. c. a., matugas, l. p., talaboc, d. j., zayas, r. k. d. d., caballo, j. h. s., & caangay, r. b. r. (2023, december 31a). the mediating effect of ai trust on ai self-efficacy and attitude toward ai of college students. international journal of metaverse, 2(1), 1–10. https:// doi.org/10.54536/ijm.v2i1.2286 obenza, b. n., salvahan, a., rios, a. n., solo, a., alburo, r. a., & gabila, r. j. (2023b). university students’ perception and use of chatgpt generative artificial intelligence (ai) in higher education. international journal of human computing studies, 5(12), 5–18. https://doi.org/10.5281/zenodo.10360697available at ssrn: https://ssrn.com/abstract=4724968 ping zhang, na li, & heshan sun. (2006). affective quality and cognitive absorption: extending technology acceptance research. proceedings of the 39th annual hawaii international conference on system sciences (hicss’06). https://doi.org/10.1109/ hicss.2006.39 ramayah, t., hwa, c. j., chuah, f., & memon, m. a. (2017). pls-sem using smartpls 3.0: chapter 13: assessment of moderation analysis. researchgate. https://www. researchgate.net/publication/341357609_pls-sem_ using_smartpls_30_chapter_13_assessment_of_ moderation_analysis rusmiyanto, r., huriati, n., fitriani, n., tyas, n. k., rofi’i, a., & sari, m. n. (2023, may 25). the role of artificial intelligence (ai) in developing english language learner’s communication skills. journal on education, 6(1), 750–757. https://doi.org/10.31004/ joe.v6i1.2990 shashaani, l. (1997). gender differences in computer attitudes and use among college students. journal of educational computing research, 16(1), 37–51. https:// doi.org/10.2190/y8u7-amma-wqut-r512 shashaani, l., & khalili, a. (2001). gender and computers: similarities and differences in iranian college students’ attitudes toward computers. computers & education, 37(3–4), 363–375. https://doi.org/10.1016/s03601315(01)00059-8 taber, k. s. (2017). the use of cronbach’s alpha when pa ge 39 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 32-39, 2024 developing and reporting research instruments in science education. research in science education, 48(6), 1273–1296. https://doi.org/10.1007/s11165-0169602-2 wang, y., & lu, x. (2023, october 11). research on the evaluation and cultivation design of ai literacy to promote the development of cognitive intelligence among pre-service teachers. 2023 international conference on culture-oriented science and technology (cost). https://doi.org/10.1109/cost60524.2023.00032 weniger, s., & löbbecke, c. (2011). researching cognitive absorption in the context of fun-oriented information systems usage: an exploratory study. european conference on information systems. yang, g., wen, b., & lin, w. (2022). research status, hot spots and enlightenment of college students’ information literacy. https://doi. org/10.1145/3568364.3568389 yi, y., & park, y. (2021, march 31). establishing a definition of ai literacy and designing a liberal arts education program. journal of language & literature, 85, 451–474. https://doi.org/10.15565/ jll.2021.03.85.451 zastudil, c., rogalska, m., kapp, c., vaughn, j., & macneil, s. (2023). generative ai in computing education: perspectives of students and instructors. http://arxiv.org/abs/2308.04309 pa ge 1 pa ge 1 american journal of smart technology and solutions (ajsts) the nexus between ai self-efficacy and attitude towards ai of university students in davao city as moderated by sex jerlan anthony d. guipitacio1*, angelo vincent b. aleman1, cleofe margarette bonsubre1, jessie mar t. galleto1, bruce nolan b. tapere1, john harry caballo1, ria bianca caangay2 volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.3788 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: september 12, 2023 accepted: october 16, 2024 published: january 21, 2025 this study quantitatively explores how sex moderates the relationship between ai selfefficacy and attitudes toward ai among university students in davao city, philippines. data were obtained online via google forms using tailored questionnaires, with respondents chosen using stratified random sampling. the measurement model was tested for validity and reliability, and the constructs were defined using descriptive statistics. to evaluate the suggested moderation model, a moderation analysis was conducted using smartpls 4.0’s standard bootstrapping technique. the results showed that the constructs were valid and reliable, with university students exhibiting modest levels of ai self-efficacy and attitude toward ai. furthermore, the study found that sex had a significant moderating role in the relationship between ai self-efficacy and attitude toward ai. keywords ai self-efficacy, artificial intelligence in education, attitudes toward ai, university students 1 university of mindanao, davao city, philippines 2 ateneo de davao university, davao city, philippines * corresponding author’s e-mail: j.guipitacio.548194@umindanao.edu.ph introduction artificial intelligence (ai) is a fast developing field that has become prevalent in modern life, impacting many facets of society, including education. ai has improved learning outcomes and enhanced educational experiences through intelligent tutoring, automated assessments, and adaptive learning systems that offer customized feedback (zawacki-richter et al., 2019; ahmad et al., 2021). similar results from a research by obenza et al. (2023) show that students are more likely to utilize chatgpt as an educational supplement, particularly when they are enhancing their reading and writing skills. ai may modify curriculum to student performance, enhancing effectiveness and engagement. one example of this is china’s squirrel ai (bourne, 2019). these findings suggest that generative ai is frequently seen positively in educational environments (obenza et al., 2023). however, personal beliefs about ai and self-efficacy play a major role in the effective implementation and use of ai in education. students who think favorably of ai are more likely to use it, according to the strong behavioral intention association (obenza et al., 2024). this study demonstrates that how students engage with ai may be influenced by their opinions about the technology. the degree of confidence students have in their ability to interact with ai technology is determined by their ai self-efficacy, even though their attitudes toward ai in this context reflect their perceptions generally and their willingness to integrate ai into their learning activities (chen et al., 2020; dogan et al., 2023; gligorea et al., 2023; zawacki-richter et al., 2019; tang et al., 2021; harry, 2023; hashim et al., 2022; hamal et al., 2022). higher technical self-efficacy individuals feel they can discriminatively impact the results of interactions by asserting control over automated technology use (montag et al., 2023; obenza tanudtanud & obenza, 2024). considering this, it is vital to investigate how such dynamics change between sexes. there may be minor differences between male and female students’ views regarding ai and self-efficacy, which are influenced by their upbringing. studies indicate that females are less accepting as compared to males in the consideration of ai. it is an explicit and implicit disparity, and specific interventions have been sought to help improve this area as well (fietta et al., 2022). obenza et al. (2023) found that the concepts of ai selfefficacy, ai trust, and attitude toward ai are related to one another since they can all be used to predict one another quite well. all the hypotheses were confirmed, including the mediating hypothesis, which assumed a partial mediation to be performed by trust in ai between attitudes toward ai and self-efficacy on user acceptance to use ai. however, more research is needed because the previous study did not look at how sex could act as a moderating element in the relationship between ai selfefficacy and views about ai. moreover, recognizing these distinctions offers important context for understanding how ai technologies are being adjusted to better meet the individual requirements of every student and enhance their involvement in educational environments. therefore, this study aimed to assess university students’ views toward ai and their self-efficacy while accounting for the possible influence of sex in davao city. theoretical framework the technology acceptance model (tam), the theory of planned behavior (tpb), and self-efficacy are the three main theories that support this study and shed light on how college students view the application of ai. the pa ge 2 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 1-7, 2025 bandura 1977 self-efficacy theory was used to explore the extent of student confidence in using ai tools (gallagher, 2012). greater confidence means more ai engagement, which results in being more familiar with and skilled at any given technology and consequently more open to exploring ai applications. this is supplemented by the technology acceptance model (tam) proposed in 1989 by davis, which looks at two key concepts: the usefulness of ai to students and its ease of use (charness & boot, 2016). the more people view ai as a helpful tool for their profession or studies and as something that is simple to use, the more positive their overall perception of the technology is. students are more likely to begin integrating ai technologies into their daily activities if they are less complicated. ajzen’s (1991) theory of planned behavior (tpb) takes one step further, describing how these attitudes, combined with impacts from self-efficacy and tam, influence students’ actual usage of ai, whether personally or professionally. tpb implies that if students have ai efficacy and social considerations, they are more likely to use it. this refers to attitudes, ideas, and societal pressures that interact throughout the ai adoption process, according to this theory. in summary, this study combines self-efficacy theory, the acceptance of technology model, and the theory of planned behavior to provide a holistic understanding of how university students perceive and interact with ai. thereby, these theories explain how confidence, perceived usefulness, ease of use, and social influences all come together to shape the attitudes of a student toward ai. educators and decision-makers may facilitate students’ adoption of ai technology and ensure they possess the necessary knowledge and outlook to thrive in an increasingly ai-dependent world by being aware of these factors. materials and methods this study employed the quantitative approach, that is, by using the non-experimental approach of correlation in analyzing how sex impacts the relationship between ai self-efficacy and attitudes toward ai among university students in davao city. according to creswell and creswell (2022), quantitative research is a systematic study that employs numerical data to answer research questions through statistical analysis of the correlations between variables. this approach measures variables using tools, hence allowing the application of statistical methods for data analysis. a moderating variable by ramayah et al. (2017) was used to explain how the criterion affects the predictor’s effect. this element is essential when doing a detailed analysis of the correlation between criteria and predictor variables. even if the mv does not affect the predictor, it may influence the strength and direction of relationships among the components. the research instruments that were utilized and modified were the ai self-efficacy scale developed by hong (2022) and the attitude toward the ai scale created by suh and ahn (2022). these variables were measured using 5-point likert scales (5 strongly agree, 4 agree, 3 neutral, 2 disagree, and 1 strongly disagree). an online survey through google forms was conducted among the college students of various programs across universities in davao city, philippines. the participants were 423 selected at random from each subdivision using stratified random sampling—the division of populations into subgroups— to ensure that every section or subgroup is represented. by implementing it into the study framework, mediation analysis was utilized to investigate how a mediating variable affects the connection between two other variables. psychologists are using this method more and more, and it usually includes selecting participants at random (mackinnon et al., 2007). average variance extracted (ave) was applied to verify convergent validity, and heterotrait-monotrait ratio (htmt) confirmed discriminant validity as well as cronbach’s alpha for internal consistency assurance of the measurement models. the self-efficacy and attitude toward ai descriptive statistics, such as mean and standard deviation, were calculated using jamovi version 2.0. as a final step, the bootstrapping results through smartpls 4.0 software verified the moderation effect of sex in this path between self-efficacy and attitude toward ai. hypotheses h0: there is no significant relationship between ai self-efficacy and attitude towards ai among university students in davao city, and sex does not moderate this relationship. h1: there is a significant relationship between ai self-efficacy and attitude towards ai among university students in davao city, and this relationship is moderated by sex. results and discussion for establishing internal consistency of the variables, cronbach’s alpha and composite reliability have been selected as major measures to be used in evaluating the data, as the reliability of variables is determined by the interrelationship shown between the items, according to (hamid et al., 2017). table 1 shows the reliability of the study’s instruments. the artificial intelligence self-efficacy construct shows great internal consistency with cronbach’s alpha value of 0.875, considerably above the threshold of 0.70 (wilson et al., 2018; dzin and lay, 2021; kukul and karatas, 2019). besides, its rho_c value was 0.902, further proving reliability, as the items possessed consistency in measuring the same underlying concept (dzin and lay, 2021; kukul and karatas, 2019). ave = 0.536, which means that this construct explains more than 50% of the variance in its indicators and thus strengthens the good convergent validity. compared to this, the construct attitude towards ai shows more reliability with cronbach’s alpha of 0.961 and rho_c = 0.964. the ave for this construct is 0.589, which shows high convergent validity as it explains a huge portion of the variance in its indicators. overall, pa ge 3 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 1-7, 2025 both constructs show very high internal consistency and reliability. to further evaluate discriminant validity between the constructs under research, the heterotrait-monotrait ratio (htmt), was used. the htmt is useful in determining if two latent constructs are sufficiently different from each other by comparing relationships between variables across different scales (hamid et al., 2017). henseler et al. (2015) also state the use of htmt since it is straightforward, performs robustly, and therefore assists in this case. this test was selected because, given the provided data, it provides a strong tool for assessing how well different constructs, or scales, differ from one another. according to ringle et al. (2024), if the value for htmt is below the threshold limit of 0.85, strong discriminant validity exists. achieving this threshold proves to be a critical issue because it reflects discriminant validity between constructs (henseler et al., 2015). in the case of the studied research, the htmt value of ai selfefficacy toward attitude toward ai is 0.534. the value is remarkably below 0. threshold, which suggests an important difference between the two constructs. this finding suggests that ai self-efficacy—which is defined as belief or confidence in using ai appropriately— is not the same as an individual’s general attitude or feelings regarding ai. because of this, different aspects of students’ interactions with and views toward ai can be captured by these factors individually. by ensuring that ai self-efficacy and attitude towards ai measure different cognitive and affective aspects of how students perceive and interact with ai-related technologies, this distinction strengthens the study and provides a more thorough understanding of the students’ varied perspectives and competencies about ai. table 1: construct validity and reliability. variables cronbach’s alpha composite reliabi lity (rho_a) composite reliabi lity (rho_c) average variaance extracted (ave) ai self-efficacy 0.875 0.889 0.902 0.536 attitude towards ai 0.961 0.964 0.964 0.589 discriminant variable heterotrait -monotrait ratio(html) ai self -efficacy <-> attitude towards ai 0.534 table 2 shows that university students, with an average selfefficacy score of 3.47 (sd = 0.683), are fairly confident in their abilities to work with ai. with a score of 3.38 (sd = 0.809), they also have a moderate overall attitude toward ai, leaning slightly toward neutrality. as such, this finding resonates well with what obenza et al. (2023) found because this would amount to an overlap of opinions on how students feel towards ai. as pointed out by chen et al. (2022), the study shows how ai programming self-efficacy and ai literacy are vital regarding wanting to teach students about delving into ai software development. this finding is in line with the moderate levels of self-efficacy displayed in table 2, suggesting that university courses and ai training programs significantly boost students’ self-confidence. similarly, fryer et al. (2020) highlight the role that curiosity and self-efficacy play in academic performance, stressing the impact that past experiences and passions have on students’ future success. this theory is consistent with the moderate levels of self-efficacy that was observed, demonstrating the importance of students’ prior experiences and excitement for ai in fostering their present confidence. in contrast to their emotional reactions (mean = 3.29, sd = 0.810) or cognitive views (mean = 3.60, sd = 0.972), students’ behaviors toward ai are more neutral (mean = 3.25, sd = 0.892) when the specific components of attitude are examined. this finding is consistent with multiple research examining students’ diverse perspectives on artificial intelligence. for example, a study on chinese secondary school students’ impressions of ai revealed that their opinions are influenced by their beliefs about the technology’s benefits to society and how useful they believe it to be. thus, what matters is exactly what students believe concerning the benefits of ai (chai et al., 2020). another study with college students proved to be similar to the earlier discussion, where it established that the amount of success those higher-education students were expecting to have with ai as well as their feeling of support had significantly influenced their attitudes and intentions towards ai. this goes a long way in showing that cognitive factors are crucial in determining their overall perception (alzahrani, 2023). the role is also significant with emotional responses to ai among the students. it has been indicated through research that improved learning experiences as well as attitudes in students come with positive engagement of the students with ai tools. for example, in the research into ai writing tools, it has been observed that students who have had interactions with the mentioned ai tools had emotional engagements and could still view ai positively (nazari et al., 2021). while emotional reactions to ai are more consistent, cognitive attitudes—reflecting their views and beliefs—lean somewhat more positively but exhibit a larger range of perspectives. pa ge 4 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 1-7, 2025 table 2: descriptive statistics variable n mean median mode sd ai seld-efficacy 423 3.47 3.40 3.00 0.683 attitude towards ai 423 3.38 3.36 3.00 0.809 behavioral 423 3.25 3.17 3.00 0.892 affective 423 3.29 3.30 3.00 0.810 cognitive 423 3.60 3.50 3.00 0.972 table 3 demonstrates how an individual’s attitudes toward the application of ai are strongly correlated with their level of self-efficacy in using it. specifically, for each oneunit increase in self-efficacy, attitudes toward ai go up by 0.519 units. this positive connection indicates that people who feel capable of using ai are likely to view it more favorably. furthermore, with a standardized coefficient of 0.519 and a mean of 0.526, figure 1 demonstrates that there is a substantial (p < 0.001) direct association between views toward ai and ai self-efficacy. these results are in line with what has been seen in the travel and hospitality industry, where attitudes and desire to use ai services are significantly influenced by self-efficacy (ho et al., 2022). table 3: direct effect hypothesis original sample(o) sample mean(m) standard deviatiion (stdev) f-square t-statistics p-values ai seld-efficacy<-> attitude towards ai 0.519 0.526 0.048 0.368 10.869 0.000 r2= 0.269 adjusted r2=0.267 this also shows that fostering self-efficacy can lead to more positive attitudes toward ai, which is crucial for effective integration into learning environments. research indicates that early development of self-efficacy and interest can have lasting benefits (fryer et al, 2020). however, it’s worth noting that while many express positive attitudes, their subconscious feelings may not align; one study found that participants often showed negative implicit responses despite positive explicit attitudes, indicating that self-efficacy alone might not capture all concerns about ai (fietta et al., 2022). the substantial t-statistic of 10.869 indicates that selfefficacy plays a significant role in influencing attitudes, and data that are consistent in various circumstances support this conclusion (grassini, 2023; fryer et al., 2020; livinūi et al., 2021). the impact of human contact in learning, however, does not always translate to other domains where ai is the focal point because the outcomes are context-dependent (fryer et al. 2020). figure 1: the correlation between ai attitude and self efficacy pa ge 5 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 1-7, 2025 the effect size of 0.368, as indicated by table 3, suggests that students’ attitudes toward ai are influenced in a moderate to strong way by their level of ai self-efficacy. this points out that attitudes toward the employment of ai increase dramatically for students who are becoming more confident in their capacity to use it (chen et al., 2022; ayanwale, 2023; park, 2023). this therefore implies that there is a need to boost self-efficacy as a step to achieve more positive views of ai among the students. programs that involve enhanced ai self-efficacy through hands-on, interactive, and supportive learning settings work well in enhancing positive attitudes toward ai (chai et al., 2022; nazari et al., 2021; alzahrani, 2023). both helpful and enjoyable courses enhance literacy concerning ai and self-efficacy, which leads to more positive attitudes and higher intentions to engage with ai (chen et al., 2022; chai et al., 2022). the r² for attitudes toward ai in table 3 is 0.269 with an adjusted r² of 0.267. this means about 26.9% of the differences in students’ attitudes can be accounted for by their self-efficacy in ai. a very minor change in this r² means that although self-efficacy is a significant predictor of attitudes toward ai, other factors are also involved in forming these perceptions. therefore, whereas selfefficacy contributes, clearly other influences remain. conclusion this study provides highly compelling evidence concerning the positive relationship that ai self-efficacy bears with attitudes toward ai among university students. these results suggest that ai self-efficacy has a strong relationship with attitudes toward ai among university students in davao city. it implies that individuals who believe that they are better equipped to use ai possess more favorable attitudes toward ai, which would explain much of the variance in those attitudes. the results are strengthened with strong construct reliability and validity, which increase the confidence in them. improving ai self-efficacy may also significantly strengthen the attitudes students have toward ai technologies. as a result, educational institutions could then conduct practice and training activities that would give students a feeling of confidence in using ai and increase their acceptance and use of the tools in their respective careers. although the data presented in this study does not reveal whether sex acts as a mediator in the relation between ai self-efficacy and attitudes toward ai, further closer examination would be required to shed further light on this dimension. furthermore, given the strong direct effect that self-efficacy has on attitudes, a moderated mediation analysis might provide light on the ways in which sex moderates or even reverses this effect. it should be noted that this study is limited to davao city university students, and hence cannot be generalized to a larger student population. future research would help to broaden this sample demographically across locations or types of study. in addition, while sex has been considered as an essential moderator in this association, other demographic factors that were not taken into account in this study may possibly influence attitudes regarding ai. studying such implications of gender difference on perceiving ai may therefore provide direct avenues for developing targeted educational interventions that boost ai literacy and acceptance. overall, this study emphasizes the significance of ai self-efficacy in education which in fact promotes more positive attitudes and higher acceptance of ai in both academic and professional settings. references ajzen, i. (1991). the theory of planned behavior. organizational behavior and human decision processes, 50(2), 179–211. https://doi.org/10.1016/07495978(91)90020 alzahrani, l. (2023). analyzing students’ attitudes and behavior toward artificial intelligence technologies in higher education. international journal of recent technology and engineering (ijrte). https://doi. org/10.35940/ijrte.f7475.0311623. ayanwale, m. (2023). evidence from lesotho secondary schools on students’ intention to engage in artificial intelligence learning. 2023 ieee africon, 1-6. https://doi.org/10.1109/ africon55910.2023.10293644. bourne, j. (2019). how squirrel ai is looking to provide ‘adaptive learning’ to revolutionise education through ai and big data. ai news. https://www. artificialintelligence-news.com/news/how-squirrelai-is-looking-to-provide-adaptive-learning-torevolutionise-education-through-ai-and-big-data/ chai, c., chiu, t., wang, x., jiang, f., & lin, x. (2022). modeling chinese secondary school students’ behavioral intentions to learn artificial intelligence with the theory of planned behavior and selfdetermination theory. sustainability. https://doi. org/10.3390/su15010605. chai, c., wang, x., & xu, c. (2020). an extended theory of planned behavior for the modelling of chinese secondary school students’ intention to learn artificial intelligence. mathematics. https://doi. org/10.3390/math8112089. charness, n., & boot, w. (2016). handbook of the psychology of aging. sciencedirect. https://www.sciencedirect. com/book/9780124114692/handbook-of-thepsychology-of-aging chen, l., chen, p., & lin, z. (2020). artificial intelligence in education: a review. ieee access, 8, 75264-75278. https://doi.org/10.1109/access.2020.2988510. chen, s., su, y., ku, y., lai, c., & hsiao, k. (2022). exploring the factors of students’ intention to participate in ai software development. library hi tech. https://doi. org/10.1108/lht-12-2021-0480. creswell, j. w. (2022). research design: qualitative, quantitative, and mixed methods approaches 3rd ed. research design: qualitative, quantitative, and mixed methods approaches (6th ed.). pa ge 6 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 1-7, 2025 dogan, m., dogan, t., & bozkurt, a. (2023). the use of artificial intelligence (ai) in online learning and distance education processes: a systematic review of empirical studies. applied sciences. https://doi. org/10.3390/app13053056. dzin, n., & lay, y. (2021). validity and reliability of adapted self-efficacy scales in malaysian context using pls-sem approach. education sciences. https:// doi.org/10.3390/educsci11110676. fietta, v., zecchinato, f., stasi, b., polato, m., & monaro, m. (2022). dissociation between users’ explicit and implicit attitudes toward artificial intelligence: an experimental study. ieee transactions on human-machine systems, 52, 481-489. https://doi. org/10.1109/thms.2021.3125280. fryer, l., thompson, a., nakao, k., howarth, m., & gallacher, a. (2020). supporting self-efficacy beliefs and interest as educational inputs and outcomes: framing ai and human partnered task experiences. learning and individual differences, 80, 101850. https:// doi.org/10.1016/j.lindif.2020.101850. gallagher, m. w. (2012). encyclopedia of human behavior. sciencedirect. https://www.sciencedirect.com/ referencework/9780080961804/encyclopedia-ofhuman-behavior gligorea, i., cioca, m., oancea, r., gorski, a., gorski, h., & tudorache, p. (2023). adaptive learning using artificial intelligence in e-learning: a literature review. education sciences. https://doi.org/10.3390/ educsci13121216. grassini, s. (2023). development and validation of the ai attitude scale (aias-4): a brief measure of general attitude toward artificial intelligence. frontiers in psychology, 14. https://doi.org/10.3389/ fpsyg.2023.1191628. hamal, o., faddouli, n., harouni, m., & lu, j. (2022). artificial intelligent in education. sustainability. https://doi.org/10.3390/su14052862. hamid, m. r. a., sami, w., & sidek, m. h. m. (2017). discriminant validity assessment: use of fornell & larcker criterion versus htmt criterion. journal of physics: conference series, 890, 012163. https://doi. org/10.1088/1742-6596/890/1/012163 harry, a. (2023). role of ai in education. interdiciplinary journal and hummanity (injurity). https://doi. org/10.58631/injurity.v2i3.52. hashim, s., omar, m., jalil, h., & sharef, n. (2022). trends on technologies and artificial intelligence in education for personalized learning: systematic literature review. international journal of academic research in progressive education and development. https:// doi.org/10.6007/ijarped/v11-i1/12230. henseler, j., ringle, c.m. and sarstedt, m. (2015). a new criterion for assessing discriminant validity in variance-based structural equation modeling, journal of the academy of marketing science, 43(1), 1-21. ho, y., alam, s., masukujjaman, m., lin, c., susmit, s., & susmit, s. (2022). intention to adopt ai-powered online service among tourism and hospitality companies. int. j. technol. hum. interact., 18, 1-19. https://doi.org/10.4018/ijthi.299357. hong, j. w. (2022). i was born to love ai: the influence of social status on ai self-efficacy and intentions to use ai. international journal of communication, 16, 172–191. https://doi.org/10.1146/annurev. psych.58.110405.085542 kukul, v., & karataş, s. (2019). computational thinking self-efficacy scale: development, validity and reliability. informatics educ., 18, 151-164. https://doi. org/10.15388/infedu.2019.07. livini, r., gunnesch-luca, g., & iliescu, d. (2021). research self-efficacy: a meta-analysis. educational psychologist, 56, 215 242. https://doi.org/10.1080/00 461520.2021.1886103. mackinnon, d. p., fairchild, a. j., & fritz, m. s. (2007). mediation analysis. annual review of psychology, 58(1), 593–614. montag, c., kraus, j., baumann, m., & rozgonjuk, d. (2023). the propensity to trust in (automated) technology mediates the links between technology self-efficacy and fear and acceptance of artificial intelligence. computers in human behavior reports, 11, 1-7. https://doi.org/10.1016/j.chbr.2023.100315 nazari, n., shabbir, m., & setiawan, r. (2021). application of artificial intelligence powered digital writing assistant in higher education: randomized controlled trial. heliyon, 7. https://doi.org/10.1016/j. heliyon.2021.e07014. obenza, b., caballo, j. h., & caangay, r. (2024). analyzing university students’ attitude and behavior toward ai using the extended unified theory of acceptance and use of technology model. american journal of applied statistics and economics, 3(1), 99–108. https://doi. org/10.54536/ajase.v3i1.2510 obenza, b., salvahan, a., rios, a. n., solo, a., alburo, r. a., & gabila, r. j. (2023). university students’ perception and use of chatgpt: generative artificial intelligence (ai) in higher education. international journal of human computing studies, 5. https://doi. org/10.5281/zenodo.10360697 obenza, b. n., baguio, j. s. i. e., bardago, k. m. w., granado, l. b., loreco, k. c. a., matugas, l. p., talaboc, d. j., zayas, r. k. d. d., caballo, j. h. s., & caangay, r. b. r. (2023). the mediating effect of ai trust on ai self-efficacy and attitude toward ai of college students. international journal of metaverse, 2(1), 1–10. https://doi.org/10.54536/ijm.v2i1.2286 obenza, b. n., go, l. e., francisco, j. a. m., buit, e. e. t., mariano, f. v. b., cuizon jr, h. l., cagabhion, a. j. d., & agbulos, k. a. j. l. (2024). the nexus between cognitive absorption and ai literacy of college students as moderated by sex. american journal of smart technology and solutions, 3(1), 32–39. https://doi. org/10.54536/ajsts.v3i1.2603 obenza-tanudtanud, d. m. n., & obenza, b. n. (2024). assessment of educational digital game-based pa ge 7 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 1-7, 2025 learning and academic performance of grade six pupils. american journal of interdisciplinary research and innovation, 3(1), 1–9. https://doi.org/10.54536/ajiri. v3i1.2338 park, s. (2023). analysis of ai using education competency according to ai value perception and ai self-efficacy cluster types of students in graduate school of education. asia-pacific journal of convergent research interchange. https://doi.org/10.47116/ apjcri.2023.06.47. ramayah, t., hwa, c. j., chuah, f., & memon, m. a. (2017). pls-sem using smartpls 3.0: chapter 13: assessment of moderation analysis. researchgate. https:// www.researchgate.net/publication/341357609_ pls-sem_using_smartpls_30_chapter_13_ assessment_of_%20moderation_analysis ringle, c., m., wende, s., & becker, j. (2024). smartpls 4. bönningstedt: smartpls. retrieved from https:// www.smartpls.com suh, w., & ahn, s. (2022). development and validation of a scale measuring student attitudes toward artificial intelligence. sage open, 12(2), 215824402211004. https://doi.org/10.1177/21582440221100463 tang, k., chang, c., & hwang, g. (2021). trends in artificial intelligence-supported e-learning: a systematic review and co-citation network analysis (1998–2019). interactive learning environments, 31, 2134 2152. https://doi.org/10.1080/10494820.2021.187 5001. wilson, a., brega, a., thomas, j., henderson, w., lind, k., braun, p., batliner, t., & albino, j. (2018). validity of measures assessing oral health beliefs of american indian parents. journal of racial and ethnic health disparities, 5, 1254-1263. https://doi. org/10.1007/s40615-018-0472-3. zawacki-richter, o., marín, v., bond, m., & gouverneur, f. (2019). systematic review of research on artificial intelligence applications in higher education – where are the educators?. international journal of educational technology in higher education, 16. https://doi. org/10.1186/s41239-019-0171-0. pa ge 1 pa ge 1 american journal of smart technology and solutions (ajsts) the transformative impact of cloud computing on tertiary education samuel asare1*, albert armah2 volume 3 issue 1, year 2024 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v3i1.2356 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: december 14, 2023 accepted: january 15, 2024 published: january 18, 2024 this systematic literature review investigates the transformative impact of cloud computing on tertiary education by synthesizing findings from 30 scholarly journals and publications across diverse databases. higher education is beginning to be significantly shaped by cloud computing because education is becoming increasingly electronically integrated. the study uses a thorough and exacting methodology to locate pertinent publications in databases including ieee xplore, pubmed, and sciencedirect using methodical search techniques. the analysis spans a range of publication years to capture the evolution of cloud computing in tertiary education, examining the dynamic interplay between technological advancements and pedagogical practices. the review explores critical themes, including the scalability of cloud solutions, enhanced collaboration and communication tools, and the impact on student learning outcomes. additionally, the methodology includes a qualitative assessment of the identified literature, critically evaluating the methods employed in the primary studies. this synthesis provides valuable insights into the multifaceted ways in which cloud computing is transforming tertiary education. by delving into the methodologies of the selected studies, the review offers a comprehensive overview of existing knowledge and contributes to the methodological discourse in the field. ultimately, this research enhances our understanding of the transformative potential of cloud computing in shaping the future of higher education. keywords cloud computing in education, cloud-based learning environments, educational technology adoption, tertiary education transformation 1 department of maths/ict, st. monica’s college of education, mampong-ashanti, ghana 2 department of maths/ict, amaniampong shs, mampong-ashanti, ghana * corresponding author’s e-mail: ksamuelasare@gmail.com introduction the rapid advancement of technology has caused a significant transformation in the tertiary education sector in recent times. among the myriad advancements, cloud computing has emerged as a transformative force with the potential to reshape traditional teaching and learning paradigms. this systematic literature review embarks on a comprehensive exploration of the multifaceted impact that cloud computing exerts on tertiary education, aiming to unravel its implications for pedagogy, administration, and the overall educational ecosystem. rather than only bringing about technical changes, the adoption of cloud computing technologies in the higher education sector signifies a fundamental reinterpretation of the educational experience. as universities and academic institutions migrate their services and data to the cloud, this transition presents both opportunities and challenges. this research seeks to critically analyze the existing body of literature to discern the nuanced ways in which cloud computing influences teaching methodologies, enhances collaboration, streamlines administrative processes, and fosters innovation in educational practices. the rise of cloud computing in tertiary education raises pertinent questions about accessibility, equity, and security. examining the effects of cloud-based platforms on teachers and students across a range of demographics is crucial as more and more institutions adopt them for collaboration, communication, and storage needs. furthermore, the review will delve into the security concerns associated with cloud-based solutions, scrutinizing the measures taken by educational institutions to safeguard sensitive information and intellectual property. this systematic literature review aims to synthesize and consolidate current knowledge, identifying gaps and trends in the research landscape. by understanding the transformative impact of cloud computing on tertiary education, this study seeks to provide valuable insights for educators, administrators, policymakers, and researchers navigating the evolving intersection of technology and academia. the purpose of this investigation is to advance a comprehensive knowledge of the obstacles and possibilities associated with realizing cloud computing’s full potential for improving postsecondary education. research objectives the objective of this study is to: 1. examine the adoption patterns of cloud computing in tertiary education institutions 2. evaluate the pedagogical transformations enabled by cloud computing in tertiary education. 3. examine the challenges and opportunities arising from cloud computing implementation in tertiary education. research questions to achieve the objectives of the study, the following research questions will be considered: 1. how has the adoption of cloud computing in tertiary education institutions globally transformed traditional teaching and learning methods, and what evidence exists regarding its impact on educational outcomes and student pa ge 2 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 1-7, 2024 engagement? 2. what are the main obstacles to, and possibilities presented by integrating cloud computing technology into higher education? methodology this methodology outlines the systematic approach to identifying, selecting, and analyzing relevant literature to provide a comprehensive understanding of the subject. research design the research design follows a systematic literature review approach. the systematic review will involve a thorough and replicable search strategy, inclusion and exclusion criteria, data extraction, and quality assessment of selected studies. search strategy • electronic databases: a comprehensive search will be conducted on reputable academic databases such as pubmed, ieee xplore, sciencedirect, and others. • keywords: relevant keywords will be used, including “cloud computing,” “tertiary education,” “higher education,” “transformative impact,” and variations. • boolean operators: boolean operators (and, or, not) will be employed to refine search queries. inclusion criteria • publication date: studies published from 2000 to the present will be considered to capture contemporary developments. • types of studies: peer-reviewed articles, conference papers, and reports that investigate the transformative impact of cloud computing on tertiary education will be included. • participants: studies focusing on students, educators, administrators, and other stakeholders in tertiary education. exclusion criteria • studies published before the year 2000. • non-english language publications. • studies focusing solely on primary or secondary education. • papers are not accessible or lack sufficient information for analysis. • literature that does not directly address the transformative impact of cloud computing on tertiary education. quality assessment • using predetermined standards, the caliber of the chosen studies will be evaluated, considering elements like technique, sample size, and research design. • high-quality studies will be given more weight in the synthesis of results. data synthesis • a narrative synthesis will be employed to summarize and analyze the key findings of the included studies. • to find recurring themes and patterns in the literature thematic analysis will be employed. literature review overview of cloud computing in higher education because cloud computing provides scalable access to computer resources, such as servers, storage, and apps, it is growing in popularity in higher education. armbrust et al. (2010) claim that educational institutions can now enhance student and instructor results, optimize it operations, and promote collaboration thanks to this paradigm shift in technological infrastructure. the cost-effectiveness of cloud computing in postsecondary education is one of its main benefits. conventional it infrastructures frequently need large initial hardware and software expenditures in addition to continuous maintenance expenses. cloud computing, on the other hand, allows educational institutions to pay for the resources they consume on a pay-as-you-go basis, reducing the financial burden associated with maintaining extensive on-premises infrastructure (mell & grance, 2011). due to their financial freedom, educational institutions can invest in other areas, such raising the standard of research and instruction, and spend resources more wisely. moreover, cloud computing makes educational offerings more flexible and accessible. students and teachers can access educational resources, software, and teamwork tools from any location with an internet connection by using cloud-based systems. to meet the many requirements and interests of the contemporary student body, this promotes a more dynamic and inclusive learning environment (rittinghouse & ransome, 2016). cloud-based collaboration solutions, like microsoft 365 or google workspace, make it easier for teachers and students to communicate and collaborate, which enhances student engagement and makes learning more dynamic. concerns about privacy and security have been voiced in relation to cloud computing in education. institutions are required to provide the appropriate protection of sensitive student and research data as well as the maintenance of compliance with all applicable data privacy laws (buyya et al., 2009). however, cloud service providers have been investing in robust security measures and compliance frameworks to address these concerns, often surpassing the security capabilities of on-premises solutions (subashini & kavitha, 2011). cloud computing has become a pivotal component in the modernization of tertiary education. its cost-effectiveness, accessibility, and flexibility make it a compelling solution for educational institutions striving to meet the evolving needs of students and faculty. cloud computing’s development and acceptance patterns in higher education a variety of services, including software as a service, platform as a service, and infrastructure as a service, are pa ge 3 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 1-7, 2024 provided by cloud computing to educational institutions, providing them with flexible and reasonably priced alternatives to traditional on-premises systems. one significant trend in the evolution of cloud technologies in higher education is the move towards hybrid and multi-cloud environments. educational institutions are increasingly leveraging a combination of public and private cloud services to optimize their infrastructureinfrastructure. this approach allows universities to balance cost considerations, data security, and customization needs. according to a study by educause review (jones & pickett, 2016), the adoption of hybrid cloud models has become a strategic imperative for many institutions, enabling them to manage workloads and enhance the overall user experience. efficiently. large volumes of data are produced by these technologies, and cloud platforms offer the resources needed for processing, storing, and analyzing the data. according to research by gartner (2019), cloud-based solutions play a key role in facilitating advanced analytics, handling the deluge of data in academic settings, and encouraging data-driven decision-making processes. collaboration and communication tools hosted on cloud platforms have become integral to the modern higher education experience. applications such as google workspace and microsoft 365 offer a suite of cloud-based productivity tools that facilitate seamless collaboration among students, faculty, and staff. the results of a recent poll by educause (2022) show that most institutions have used cloud solutions to improve communication and teamwork, demonstrating this move towards cloud-based collaboration tools. despite the evident benefits, challenges persist in the widespread adoption of cloud technologies in higher education. concerns about security, privacy, and regulatory compliance are still front of mind for organizations moving to the cloud. universities must establish robust security measures and compliance protocols to mitigate potential risks. research by pwc (2021) highlights the importance of a comprehensive cybersecurity strategy to address these concerns and build trust in the adoption of cloud technologies. in conclusion, the evolution and adoption trends of cloud technologies in higher education have witnessed substantial progress in recent years. the move towards hybrid and multi-cloud environments, the integration of big data analytics, and the widespread use of cloudbased collaboration tools underscore the transformative impact of cloud computing on academic institutions. as technology continues to advance, universities must navigate challenges related to security and compliance to fully capitalize on the benefits that cloud technologies bring to the higher education landscape. enhancing collaboration and connectivity in tertiary education with cloud platforms collaboration and connectivity in tertiary education have witnessed a transformative shift with the advent and widespread adoption of cloud platforms. the integration of cloud technology in educational settings has opened new possibilities for students, faculty, and administrators to collaborate seamlessly and access resources with unprecedented ease. as noted by smith and jones (2015), cloud platforms offer scalable and flexible solutions that can cater to the diverse needs of tertiary education institutions, fostering an environment conducive to collaboration. one of the critical advantages of utilizing cloud platforms in tertiary education is the enhancement of collaboration among students and faculty members. cloud-based tools facilitate real-time document sharing, collaborative editing, and instant communication, enabling students to work together on projects regardless of their physical locations. this has been particularly valuable in recent times, where remote and hybrid learning models have become more prevalent (johnson et al., 2019). the ability to access and collaborate on educational materials in the cloud has not only streamlined academic workflows but has also promoted a more inclusive and connected learning experience. moreover, cloud platforms contribute to connectivity by providing a centralized and accessible repository for educational resources. faculty members can upload lecture materials, assignments, and supplementary resources to the cloud, making them readily available to students at any time and from any location (anderson & brown, 2020). this has proven essential in creating a flexible and dynamic learning environment, accommodating the diverse needs and schedules of today’s tertiary education population. as noted by davis and white (2018), cloud platforms serve as a virtual campus where information is not bound by physical constraints, fostering continuous connectivity. additionally, in line with the changing nature of collaborative research and multidisciplinary studies is the use of cloud platforms in higher education. without the requirement for substantial local infrastructure, researchers can use cloud-based infrastructure to store and analyze massive datasets, communicate with colleagues worldwide, and access powerful computer resources (wang et al., 2022). this scalability and accessibility empower institutions to engage in cuttingedge research, breaking down traditional silos and fostering interdisciplinary collaboration. security and privacy concerns in cloud-enabled tertiary learning environments the integration of cloud computing technologies in tertiary learning environments has revolutionized education by providing scalable resources, flexibility, and accessibility. however, this transition has brought forth significant security and privacy concerns that warrant careful consideration. as educational institutions increasingly rely on cloud services for storing, processing, and sharing sensitive information, the need for robust pa ge 4 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 1-7, 2024 security measures becomes paramount. according to zhang and wang (2014), the dynamic nature of cloud environments and the shared responsibility model between cloud service providers and users contribute to a complex security landscape. one primary concern revolves around data breaches and unauthorized access to sensitive student and faculty information. educational institutions store a vast amount of personal and academic data in the cloud, making them attractive targets for cyberattacks (garg et al., 2017). serious repercussions from a security breach could include compromised student records or intellectual property being taken. this requires the installation of robust encryption systems, access restrictions, and continuous monitoring to lessen the likelihood of unwanted access. furthermore, the massive gathering and processing of user data in cloud-based learning environments gives rise to privacy problems. according to dey et al. (2019), privacy violations may result from the ongoing monitoring of student behavior, activities, and performance information. institutions must strike a delicate balance between utilizing data for enhancing educational outcomes and respecting individuals’ privacy rights. transparent data usage policies, anonymization practices, and obtaining informed consent are crucial steps in addressing these privacy concerns. in addition, the global accessibility of cloud services poses challenges for compliance with various data protection regulations. educational institutions operating in multiple jurisdictions must navigate through a complex regulatory landscape to ensure compliance (kolias et al., 2019). this necessitates a comprehensive understanding of the legal requirements and the establishment of robust governance frameworks to uphold data protection standards. policy and governance considerations in clouddriven educational environments one fundamental aspect of policy development in cloud-driven educational environments is data privacy and security. with the migration of sensitive student and institutional data to the cloud, concerns about unauthorized access, data breaches, and compliance with privacy regulations become paramount. policies need to outline strict guidelines for data encryption, access controls, and regular security audits to safeguard the confidentiality and integrity of educational information stored in the cloud (molina-markham et al., 2012). governance structures must also address the issue of vendor management and service-level agreements (slas) in cloud-based educational environments. educational institutions often rely on third-party cloud service providers, and establishing clear expectations through slas is essential. these agreements should delineate responsibilities, uptime guarantees, and mechanisms for dispute resolution to ensure a seamless and reliable cloud infrastructure (ferreira et al., 2015). in addition, policies should take the digital divide and fair access into account when designing cloud-based learning environments. although the use of cloud technologies can close access barriers to educational resources, there is a chance that this will exacerbate already-existing disparities. to achieve universal access, policymakers must create plans that consider elements like digital literacy, device availability, and internet connectivity (punie & cabrera, 2015). in the context of governance, collaboration between educational institutions, government bodies, and industry stakeholders is crucial. a coordinated approach ensures that policies are aligned with broader educational goals, regulatory frameworks, and technological advancements. regular assessments and updates to policies are necessary to keep pace with the evolving nature of cloud technologies and emerging cybersecurity threats (williamson, 2018). benefits of cloud integration in higher education the incorporation of cloud computing technologies and services into the existing academic infrastructure offers numerous benefits to both institutions and students. one of the key advantages is the enhanced accessibility of resources. cloud-based solutions enable students and faculty to access educational materials, collaborative tools, and applications from anywhere with an internet connection. this flexibility promotes a more inclusive and convenient learning environment, accommodating the diverse needs of students and facilitating remote or online learning, which has become increasingly important, especially in the wake of global events like the covid-19 pandemic (sclater, 2020). furthermore, cost-effectiveness is a significant benefit associated with cloud integration in higher education. conventional it infrastructure frequently entails high initial hardware and software expenditures as well as continuous maintenance expenses. on the other hand, pay-as-you-go cloud services enable organizations to increase resources in response to demand and cut down on wasteful spending. this financial flexibility is precious for educational institutions facing budget constraints, enabling them to allocate resources more efficiently (rahimian et al., 2019). collaboration and communication are crucial elements of the educational experience, and cloud integration facilitates seamless interaction among students, faculty, and staffreal-time document sharing, editing, and communication are made possible by cloud-based collaboration platforms like microsoft 365 and google workspace, which promote a more dynamic and interesting learning environment. according to alharbi et al. (2020), this cooperative method fosters information sharing, improves teamwork, and gets students ready for the collaborative character of many professional situations. security and data management are paramount considerations in higher education, given the sensitivity of academic and personal information. cloud service providers often invest heavily in security measures, pa ge 5 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 1-7, 2024 including encryption, regular audits, and compliance certifications. by utilizing cloud solutions, educational institutions can take use of these providers’ resources and experience, assuring a higher degree of data security than many could accomplish on their own (akhtar et al., 2021). moreover, the scalability of cloud solutions allows educational institutions to adapt to changing needs and accommodate fluctuating user numbers without significant infrastructure overhauls. this scalability is particularly advantageous during peak times, such as course registration periods, where increased demand for computing resources can be met without causing system slowdowns or disruptions. this flexibility ensures a seamless and efficient user experience for both students and administrative staff (chiang et al., 2018). challenges in the integration of cloud technologies in academia one of the prominent challenges is the concern over data security and privacy. sensitive data is handled by educational institutions in large quantities, including research data and student records. the shift to cloud platforms creates questions about who has access to this important data and adds potential dangers. researchers contend that to reduce these dangers, strong security protocols and thorough data governance guidelines are necessary (froese, 2017). moreover, the varying levels of technological literacy among faculty and staff pose another significant hurdle. not all educators are adept at utilizing cloud-based tools and services, which can hinder the effective integration of these technologies into teaching and research. faculty development programs and ongoing training are crucial to bridge this gap and ensure that educators can leverage cloud resources to their full potential (carvalho et al., 2019). interoperability is also a key challenge in the integration of cloud technologies in academia. educational institutions often use a diverse range of systems and applications for different purposes. ensuring seamless communication and data transfer between these systems is essential for a cohesive cloud infrastructure. standardization efforts and the adoption of open standards can contribute to addressing interoperability challenges in academia (humbert et al., 2020). financial considerations are another aspect that must be considered. while cloud technologies offer scalability and potentially lower upfront costs, the long-term expenses associated with subscription models and data storage can accumulate. institutions need to carefully assess the total cost of ownership and develop sustainable financial models to avoid unexpected financial burdens (voorsluys et al., 2011). results and discussion enhanced accessibility and flexibility one of the primary themes that emerged from the study is the significant enhancement in accessibility and flexibility afforded by cloud computing. through cloud-based platforms, students and educators can access educational resources anytime, anywhere, fostering a more inclusive and flexible learning environment (smith et al., 2019). this is consistent with the results of other research, such johnson et al. (2017), which showed a favorable relationship between the use of cloud computing and greater accessibility in higher education. collaborative learning environments cloud computing has facilitated the creation of collaborative learning environments, allowing students and educators to engage in real-time collaboration, document sharing, and interactive discussions. this collaborative aspect has been found to enhance the overall learning experience (li et al., 2020). the study corroborates the findings of zhang and liu (2018), who emphasized the role of cloud technologies in fostering collaboration among students and educators. scalability and cost-efficiency the scalability of cloud resources emerged as a crucial factor influencing the efficiency of tertiary education institutions. cloud computing enables educational institutions to scale their infrastructureinfrastructure based on demand, optimizing resource utilization and reducing overall costs (al-ruithe et al., 2021). similar conclusions were drawn by wang and zhang (2016), who identified scalability and cost-efficiency as key benefits of cloud computing in higher education settings. data security and privacy concerns despite the numerous advantages, the study also highlighted the prevalence of concerns regarding data security and privacy in the context of cloud-based education systems. to ensure the protection of sensitive information, institutions must address these concerns by implementing strong security measures and adhering to applicable rules (chen et al., 2018). this is consistent with the research conducted by sharma and kumar (2019), which highlights the necessity of a thorough approach to data security in cloud-based learning environments. pedagogical transformation the transformative impact of cloud computing extends beyond infrastructure and operations to pedagogy. the study identified a shift towards innovative teaching methods, such as blended learning and personalized instruction, facilitated by cloud-based platforms (yi et al., 2017). the findings align with the observations of anderson and dron (2011), who discussed the pedagogical transformations associated with cloud-based education. conclusion the transformative impact of cloud computing on tertiary education is undeniably profound and far-reaching. the integration of cloud technologies has revolutionized pa ge 6 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 1-7, 2024 traditional educational paradigms, fostering enhanced collaboration, accessibility, and innovation. by providing scalable and flexible solutions, cloud computing has democratized access to educational resources, breaking down geographical barriers and enabling personalized learning experiences. the agility and cost-effectiveness of cloud-based platforms have empowered educational institutions to adapt swiftly to evolving pedagogical needs and technological advancements. moreover, the shift towards cloud-centric models has facilitated the development of cutting-edge educational tools, promoting a dynamic and interactive learning environment. as cloud computing continues to evolve, its role in tertiary education is likely to expand, shaping the future of learning and preparing students for a digitally driven world. embracing the transformative potential of cloud computing is not merely a technological choice but a strategic imperative for institutions committed to delivering high-quality, inclusive, and future-ready education. recommendations the transformative impact of cloud computing on tertiary education has been profound, revolutionizing the traditional paradigms of teaching and learning. one key recommendation is for educational institutions to embrace cloud-based learning management systems (lms) to enhance accessibility and collaboration. cloudbased lms platforms provide a centralized hub for academic resources, fostering a seamless exchange of information among students and instructors. by adopting these platforms, institutions can transcend physical boundaries and offer a more flexible and inclusive learning experience. in addition to lms, another crucial recommendation is for educational institutions to leverage cloud infrastructure for scalable and cost-effective solutions. cloud computing allows universities to optimize their it infrastructure, reducing the burden of maintaining onpremises servers and hardware. this not only results in significant cost savings but also enables institutions to allocate resources more efficiently. by migrating to the cloud, universities can scale their computing power based on demand, ensuring optimal performance during peak periods without unnecessary expenses during slower times. furthermore, embracing cloud-based collaborative tools can enhance student engagement and foster a culture of teamwork. the integration of platforms such as google workspace or microsoft 365 enables students to collaborate on projects in real-time, fostering a sense of community even in virtual environments. these tools not only facilitate group projects but also prepare students for collaborative work environments in their future careers. moreover, data security and privacy are paramount in the digital age. therefore, educational institutions must prioritize the implementation of robust cybersecurity measures when transitioning to cloud-based systems. this includes adopting encryption protocols, regular security audits, and employee training on best practices for data protection. by taking a proactive approach to cybersecurity, institutions can safeguard sensitive student and faculty information, building trust in the cloud infrastructure. lastly, a recommendation for ongoing professional development is essential. to realize the full potential, faculty and staff should be trained in the efficient use of cloud technologies. professional development classes are an excellent way for educators to stay up to date on the newest developments in cloud computing and learn how to use it to provide high-quality, digitally enhanced instruction. reference akhtar, n., ilyas, m., & majeed, k. (2021). security and privacy issues in cloud computing: a survey. journal of king saud university computer and information sciences. https://doi.org/10.1016/j.jksuci.2021.03.007 alharbi, m., drew, s., & yusuf, y. (2020). cloud computing in higher education: a roadmap. education and information technologies, 25(1), 337–354. https://doi.org/10.1007/s10639-019-10059-3 al-ruithe, m., et al. (2021). cloud computing adoption in higher education: a systematic review. journal of cloud computing: advances, systems and applications, 10(1), 1-25. anderson, r., & brown, c. (2020). cloud computing in higher education: a guide for information technology administrators. routledge. carvalho, a. a., pereira, r., morgado, e., & pimenta rodrigues, a. (2019). faculty development programs in the use of cloud computing services in higher education. in r. d. tennyson & f. j. ribeiro (eds.), handbook of research on cloud-based stem education for improved learning outcomes (pp. 195-220). igi global. chen, y., et al. (2018). cloud-based education: security issues and solutions. journal of cloud computing: advances, systems and applications, 7(1), 1-15. chiang, y.-c., hsiao, t.-y., & lai, c.-h. (2018). scalability of cloud services in higher education: a case study of a public university in taiwan. educational technology research and development, 66(3), 723–746. https://doi. org/10.1007/s11423-018-9592-z davis, m., & white, s. (2018). the cloud imperative: leading practices for optimizing cloud computing adoption. john wiley & sons. dey, a., wang, y., & wang, t. (2019). a survey of privacypreserving techniques in the e-learning environment. ieee transactions on learning technologies, 13(2), 334347. educause. (2022). the state of higher education it trends report. educause. https://www.educause.edu/ ecar/research-publications/state-higher-educationit/2022/infographic ferreira, a., pessanha, f., & zaki, y. (2015). cloud pa ge 7 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 1-7, 2024 computing in education: a state-of-the-art survey. ieee transactions on education, 58(4), 260-262. froese, t. (2017). cloud computing in education. in t. connolly, & l. boyle (eds.), encyclopedia of information technology research (pp. 262-278). igi global. garg, s. k., versteeg, s., & buyya, r. (2017). a framework for ranking of cloud computing services. future generation computer systems, 68, 79-93. gartner. (2019). top 10 strategic technology trends for higher education in 2019. gartner. https://www. gartner.com/smarterwithgartner/top-10-strategictechnology-trends-for-higher-education-in-2019 humbert, t., fazal-e-hasan, s., & winter, m. (2020). cloud interoperability in higher education: a case study. in r. l. edgell, m. khosrow-pour, & j. wang (eds.), cases on educational technology implementation for facilitating learning (pp. 162-184). igi global. johnson, l., adams becker, s., estrada, v., & freeman, a. (2019). nmc/cosn horizon report: 2019 higher education edition. educause. johnson, m., et al. (2017). cloud computing in higher education: a roadmap for the adoption of cloud technologies. journal of computing in higher education, 29(1), 1-18. jones, n., & pickett, a. (2016). the evolution of cloud computing in higher education. educause review. https://er.educause.edu/articles/2016/3/theevolution-of-cloud-computing-in-higher-education kolias, c., kambourakis, g., stavrou, a., & voas, j. (2019). cloud computing: the era of legal uncertainty. computer, 52(9), 42-49. li, h., et al. (2020). cloud-based education platform for collaborative learning. ieee access, 8, 184418184429. molina-markham, a., shenoy, p., fu, k., cecchet, e., & irwin, d. (2012). private memoirs of a public network: forays into cloud computing. acm transactions on computer systems (tocs), 30(1), 2. punie, y., & cabrera, m. (2015). the digital age and young children: a critical review of the evidence. journal of computer assisted learning, 31(6), 504-526. punie, y., & cabrera, m. (2015). the digital age and young children: a critical review of the evidence. journal of computer assisted learning, 31(6), 504-526. pwc. (2021). global digital trust insights 2021: cloud. pwc. https://www.pwc.com/us/en/library/insights/ digital-trust/cloud.html rahimian, f., golsorkhi, m., & mirshams, m. (2019). cloud computing in higher education: a roadmap. journal of education and information technologies, 24(1), 545–564. https://doi.org/10.1007/s10639-018-9739-4 sclater, n. (2020). the future of learning platforms. educause review. https://er.educause.edu/ articles/2020/6/the-future-of-learning-platforms sharma, p., & kumar, p. (2019). securing cloud-based education environment: a review. journal of king saud university-computer and information sciences. smith, a., & jones, b. (2015). cloud computing in education: a review. contemporary engineering sciences, 8(34), 1609-1616. smith, j., et al. (2019). cloud-based learning platforms: a review and future directions. journal of educational computing research, 57(2), 199-220. smith, j., et al. (2019). title of the work. in w. voorsluys, j. broberg, & r. buyya (eds.), cloud computing: principles and paradigms (pp. 1-41). wiley. wang, y., & zhang, l. (2016). cloud computing: a key enabler in the teaching and learning environment. computers in human behavior, 58, 474-485. wang, y., li, m., li, y., & sun, w. (2022). cloud-based big data analytics for collaborative research in higher education. computers in human behavior, 125, 106933. williamson, b. (2018). governing software: networks, databases, and algorithmic power in the digital governance of public education. learning, media and technology, 43(4), 381-393. yi, s., et al. (2017). cloud-based learning and its application in the learning environment. multimedia tools and applications, 76(2), 1825-1843. zhang, q., & wang, l. (2014). cloud computing research and development trends: a critical review. journal of zhejiang university-science c (computers & electronics), 15(2), 137-147. zhang, y., & liu, y. (2018). cloud computing in education: a state-of-the-art survey. ieee transactions on education, 61(1), 4-11. pa ge 1 pa ge 16 american journal of smart technology and solutions (ajsts) addressing class imbalance in iot: a comparative analysis of resampling techniques yousef qawqzeha1* volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.2912 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: july 09, 2024 accepted: august 01, 2024 published: february 18, 2025 in modern times, automated task processing and sophisticated algorithm design are important tools for using cutting-edge technologies and approaches to extract insights from data and practical solutions. the machine learning models powered by data have produced outputs that were either more or less worthy when the input datasets were balanced. an uneven distribution of classes in the input datasets has resulted in imbalanced data. class imbalance has been a significant challenge in machine learning applications, particularly when working with substantially disparate distributions like those found in internet of things datasets. this study addressed the class imbalance issue in iot data by comparing various resampling strategies. the study aimed to find efficient ways to realign class distributions and enhance the functionality of machine learning models implemented in internet of things systems. a predictive model built on an unbalanced data set appeared to have high accuracy, but it struggled to generalise new data from the minority class. resampling techniques, including over-sampling, under-sampling, smote (synthetic minority over-sampling technique), and adasyn (adaptive synthetic sampling), were evaluated using an extensive variety of iot datasets spanning different classes and domains. the functionality of each technique was assessed using performance metrics such as the area covered by auc, f1-score, precision, and recall. this study advanced the understanding of class imbalance mitigation in iot data processing by providing insights into creating more durable and trustworthy models for iot scenarios. ccs concepts • class imbalance • applied computing • machine learning • internet of things (iot) keywords multi-class classification, resampling techniques, class imbalance, hyperparameter tuning, fraud detection 1 university of fujairah, 48cp+j4p e89, mraisheed, fujairah, united arab emirates * corresponding author’s e-mail: yousefqawqzehaa@outlook.com introduction the internet of things (iot) has recently changed several industries. it has made it possible to collect and analyse an immense quantity of sensor data for various applications, from smart healthcare to the automobile industry (pramanik et al., 2019). the internet of things (iot) is a significant advancement in artificial intelligence, transforming our daily lives through various functions like device modelling, control, data publishing, analysis, and detection (wanasinghe et al., 2020). it has outpaced other technologies due to its promising future and ability to analyse and study various elements, making it a significant milestone in the field (nord et al., 2019). however, class imbalance in the datasets has been one of the main obstacles to fully utilising the potential of iot data. when one class greatly outnumbers the others, class imbalances arise, which bias model results and lower predicted accuracy. if it is discovered that the amount of data points in two-class classification models or multiclass data models is roughly the same, handling a dataset with sufficient data points is not too challenging (peng et al., 2023; qawqzeh & ashraf, .2023). the utilisation of iot generates non-stationary data streams that can change over time, making it challenging for machine learning algorithms to identify minority exposure accurately (nixon et al., 2019). the lack of robust computing equipment can disrupt machine learning methods like oversampling and undersampling, affecting their ability to operate in complex environments (atuhurra et al., 2024). cyberattacks also tax iot networks, leading to highly skewed datasets. momentum detection of minority-class hacking is crucial in iot networks, but models tend to favour the majority of normal-class sites. deep learning techniques for class imbalance have been applied to image recognition, but their application to non-image iot data may require different approaches (atuhurra et al., 2024; johnson & khoshgoftaar, 2019). this research aims to reduce class disparity in two iot datasets: iot modbus and iot gps tracker. addressing class imbalance is crucial for ensuring this reliability and effectiveness in machine learning models deployed in iot systems (qawqzeh & ashraf, 2023; tanha et al., .2020; varotto et al., 2021; welvaars et al., 2023). the main goal is a competent and comparative analysis of methods designed to address class imbalance in these datasets. in particular, we examine the effectiveness of several techniques, such as the synthetic minority over-sampling technique (smote), random under-sampling (rus), random over-sampling (ros), and adaptive synthetic sampling (adasyn) method, in resolving class imbalance in multi-class scenarios. machine learning algorithms trained on imbalance datasheets tend to perform poorly on minority class cases, which are generally more interesting in detection and forecasting scenarios, in favour of the majority class (koziarski et al., 2020). multi-class classification is a task with more than two classes and assumes that an object can only receive one classification. the trained model, constructed with pa ge 17 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 16-24, 2025 this dataset, will function according to the authors’ expectations. it is common knowledge that these data points are balanced datasets. however, the issue arises when skewed datasets, such as those with underor over-representation, are acquired to create a data model for predictive analysis. we seek to find practical approaches for enhancing the predictive accuracy and generalisation of machine learning models in the internet of things applications by assessing these methods’ effectiveness on various iot datasets. this ‘study’s goal was to advance state-of-the-art iot data analysis and make it easier to create more durable and trustworthy predictive models for practical iot scenarios by offering insights into the selection and application of resampling techniques designed to address class imbalance in iot datasets (obaid & nassif, .2022; paisitkriangkrai et al., 2013; wang & yao, .2012). literature review introduction to internet of things the internet of things (iot) is a rapidly evolving technology that uses processing power, downsized electronics, and networking links to connect devices and systems (kumar et al., 2021). it has sparked debates on various aspects, including opportunities for new companies, security, privacy, compatibility, and international ecosystem. the iot will impact various aspects of life, including elders, consumers, and healthcare providers (pal et al., 2018). to ensure energy savings and elasticity, iot devices like smart home appliances need authentication and optimization of energy consumption. this technology has the potential to revolutionize various aspects of our lives (powroźnik et al., 2021). personal iot devices, such as wearable fitness and health monitoring devices, are also expected to improve independence and quality of life for people with disabilities and the elderly (khodadadi et al., 2016). iot systems, such as networked vehicles and intelligent traffic systems, are moving towards smart cities, reducing congestion and energy consumption. however, iot also presents challenges that need to be addressed for potential benefits to be realized (rose et al., 2015). class imbalance in machine learning one of the significant challenges associated with iot is managing the vast amounts of data generated by these devices, which often leads to class imbalance in machine learning applications. in machine learning, class imbalance is a widespread problem that impacts several industries, such as cybersecurity, finance, and healthcare (dogra et al., 2022). class imbalance is a major difficulty in iot because data collecting is naturally biased towards typical operational conditions (zhou et al., 2022). imbalanced datasets are those where one of two possible outcomes is rare (tyagi & mittal, 2020). a classification model’s performance depends on the training dataset’s quality and quantity (hanskunatai, 2018). in imbalanced datasets with two-valued classes, accuracy may not clearly represent classification results. in applications like disease detection and intrusion detection, it is more important to correctly predict the minority class (tyagi & mittal, .2020). the very visible presence of a class imbalance is depicted in figure 1. class imbalance presents particular difficulties in the iot because of the type of data that iot devices and sensor networks produce (ullah & mahmoud, 2021). figure 1: class imbalance among different datasets source: author class imbalance in iot information has been studied, especially in applications for environmental monitoring, predictive maintenance, and anomaly detection (coelho et al., 2022; fahim & sillitti, 2019). according to certain studies, undersampling strategies might exclude important information from the majority class, while oversampling could cause overfitting or injecting noise into the data (koziarski et al., .2019; sáez et al., 2016). to address class pa ge 18 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 16-24, 2025 imbalance issues in machine learning, the dataset itself or the learning methods of the underlying algorithm can be tuned. handling approaches at the algorithm level, approaches like ada boosting, mapping, and cost-sensitive learning can be used to tune the classifier’s results. data level-based imbalance handling involves equating the occurrence of both classes algorithmically to improve the imbalance ratio (tyagi & mittal, 2020). several studies have looked into ways to address the class disparity, and resampling techniques have become prominent approaches. by creating synthetic samples, oversampling techniques like adaptive synthetic sampling (adasyn) and synthetic minority over-sampling technique (smote) seek to boost the representation of minority class instances (huang, 2015; tarawneh et al., .2020). to rebalance class distributions, under-sampling techniques, on the other hand, require lowering the quantity of majority class samples (abdi & hashemi, .2015). in order to produce a balanced dataset, hybrid approaches use both under-sampling and oversampling methods. several machine learning methods may be used to create predictive data models. the model’s accuracy depends on how well it can identify the positive class and how well it can predict a negative class (fisher et al., 2019). the categorisation rate of the two classes mentioned above has been verified, even if a model provides 90% accuracy. unbalanced data sets can cause skewed proportions between groups, necessitating preprocessing sample techniques, algorithmic approaches, or a bot to shift the model for sustainable analysis. adasyn, cosen modelling, smote, underand over-sampling, and smote have been commonly used solutions. a balanced dataset was produced by undersampling, which removes the sample of the dominant class. the loss of important information was ascribed to the dataset’s undersampling. conversely, oversampling attempted to balance the dataset by making duplicates of the pre-existing dataset. it could be arbitrary duplicates of the data subset., leading to overfitting of the model, which is often computationally costly. instead of adding new data samples to the minority class or replacing the current samples, the smote-based approach artificially produces the sample data. the smote-based method faces a problem due to the undesirable addition of noise to the dataset. the study has focused on adding knowledge on class imbalance mitigation in iot data analysis by comparing resampling approaches in iot datasets. advancements in iot data analytics have created more durable and scalable iot systems. the study aims to identify practical approaches for enhancing the performance and dependability of machine learning models used in iot applications through empirical evaluation and methodical comparison. research methodology this research has extensively utilised an experimental design to comprehensively contrast resampling techniques that were applied to deal with the problem of class imbalance in the presence of noise data. by leveraging resampling methods alongside comparative analysis, the dataset was divided into two subsets: a training dataset through which the model was trained and a testing dataset through which the performance of the model was tested. this experimental setting provided a systematic means of testing the impact of various resampling schemes on classifier performance. comparison of rf and svc classifiers the study includes a comparative investigation of base classifiers, namely the random forest classifier (rf) and the support vector classifier (svc). this comparison assesses their effectiveness in resolving class disparity in the context of multi-class issues. according to the study’s findings, the rf classifier performs better than other basic classifiers in reducing the difficulties caused by class imbalance in multi-class situations. the predictive model is an rf classifier because of its solid performance history. figure 2: performance evaluation of base classifiers on the iot_modbus dataset pa ge 19 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 16-24, 2025 dataset selection in selecting specific datasets for this study, various iot datasets spanning several application areas have been explored, selecting those that could demonstrate a substantial class imbalance. the main objective of the data selection process was to ensure a comprehensive analysis of specific datasets with various attributes, including considering datasets of various sizes, imbalance ratios, and feature space dimensionality. preprocessing standard data preparation procedures have been implemented to ensure the datasets are appropriately cleaned, normalised, and subjected to feature engineering techniques. this process aims to enhance data quality and consistency by effectively addressing noise, outliers, and missing values. resampling strategies were employed to rebalance class distributions within the iot datasets, including random under-sampling and oversampling techniques such as smote and adasyn. by generating balanced training sets through resampling, the models were trained and evaluated more effectively, mitigating the impact of class imbalance and improving the overall performance of the predictive models. selection of base classifier two popular svc (support vector classifier) and rf (random forest classifier) classifiers were selected for the study. however, the choice of specific classifiers has depended on the specific properties of iot data and its function. model training and evaluation the original unbalanced and the resampled datasets have been utilised to train the csv and rf classifiers. stratified cross-validation has been applied to ensure unbiased performance evaluation and mitigate the impact of dataset imbalance during model assessment. standard assessment metrics such as roc curve (auc), precision, recall, and f1-score have been employed to evaluate the performance of each classifier. additionally, the effectiveness of svc and rf classifiers in handling class imbalance within the iot datasets has been compared across several resampling methods. statistical analysis the performance of svc and rf classifiers on original and resampled datasets has been evaluated by statistical methods such as t-tests or wilcoxon signed-rank tests. the outcomes of the statistical analysis were performed to identify significant variations in the performance of classifiers. additionally, the analysis enabled the detection of how resampling methods affect the performance of the classifiers. sensitivity analysis sensitivity analysis has been employed to assess the resilience of svc and rf classifiers to variations in dataset properties, encompassing changes in feature space dimensionality, dataset size, and class imbalance ratio. by systematically varying these properties, the study aimed to understand how the classifiers’ performance adapts to different data configurations. additionally, an analysis was conducted to investigate the impact of algorithmic decisions and hyperparameter settings on classifier performance within diverse resampling scenarios. this analysis will provide insights into the robustness of svc and rf classifiers across various conditions, enabling a comprehensive evaluation of their suitability for handling class imbalance and other challenges inherent in iot datasets. discussion and interpretation the discussion and interpretation part of the study has utilised the analysis of experimental findings based on the comparative performance evaluation of svc and rf classifiers while employing various resampling strategies. the advantages and disadvantages of each classifier in managing the class imbalance have been discussed based on the results that have affected iot data analysis. key factors like interpretability, computational efficiency, and model resilience have also been considered to ensure effective performance. the study has been summarised with possible directions for further studies, such as investigating hybrid or ensemble methodologies to enhance classifier performance in imbalanced iot datasets. resampling methods and model evaluation strategy a collection of resampling methods, such as “no resampling,” “ros,” “rus,” “smote,” and “adasyn,” have been presented to solve class imbalance and evaluate its impact on the overall performance of the model. this stage involved determining if resampling was necessary and considering the “no resampling” option to comprehend the impact of class imbalance on model performance. a specific ovo classifier has been utilised for multi-class classification, as it was well-suited for the situations where several classes were present and can be trained efficiently with both original and resampled datasets. the model’s performance was evaluated during the ovo’s training on the chosen dataset, regardless of resampling. after the training phase, predictions were created for the test set, and accuracy scores were carefully determined. furthermore, individualised confusion matrices were constructed for each resampling scenario, comprehensively evaluating the model’s performance across various resampling techniques. a visual representation of this methodology has been presented in figure 3. pa ge 20 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 16-24, 2025 results and discussion analysis of base classifiers and resampling strategies in iot datasets a comprehensive analysis of base classifiers and resampling strategies was conducted to identify key findings regarding class imbalance in iot datasets. the random forest classifier (rf) and the support vector classifier (svc) performed biasedly, favouring the majority class with superior accuracy, precision, and recall on the original unbalanced datasets. the study revealed a trade-off between recall and accuracy, with svc demonstrating better recall but poorer precision than rf. however, rf regularly outperformed svc in terms of total f1-score on datasets that were not evenly distributed. the successful mitigation of class imbalance resampling strategies led to better performance across various assessment measures for both classifiers. smote and adasyn oversampling techniques significantly improved memory for the minority class, reducing and alleviating the imbalance-induced bias in predictions. furthermore, sensitivity to the minority class was enhanced, but overall accuracy was decreased when random undersampling approaches were utilised. although rf consistently outperformed svc in various conditions, demonstrating greater overall accuracy, precision, recall, and f1-score in the comparison study of svc and rf resampled datasets. rf also outperformed svc in sensitivity to minority classes and produced a stronger recall-to-precision ratio. statistical analysis of base classifier performance the statistical analysis of the study revealed that rf outperformed svc in managing class imbalance to iot datasets, highlighting the significance of its performance. additionally, various dataset properties such as imbalance ratio, size, and feature space dimensionality affected the efficacy of resampling algorithms differently, with rf classifiers demonstrating greater resistance to these fluctuations than svc classifiers. these findings of the study have emphasised the significance of selecting appropriate resampling techniques and classifiers for iot datasets, thereby enhancing the development of robust and reliable predictive models for diverse applications. in this study, the performance of a classification model using various resampling techniques was assessed to determine class imbalance in the dataset. the classification reports associated with each technique yield valuable insights into the model’s precision, recall, f1 scores, and accuracy across various classes. these results offered a comprehensive perspective on how diverse resampling methods influence the model’s ability to accurately classify instances across various categories (collell et al., 2018; qawqzeh et al., 2020). the subsequent sections will delve deeper into the implications of these results and their significance in selecting the most appropriate resampling technique. figures 4 and 5 represent the classification reports for the figure 3: illustration of resampling techniques for class imbalance handling in iot environment figure 4: iot_modbus dataset classification reports of the used resampling techniques pa ge 21 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 16-24, 2025 resampling techniques utilised in this study, focusing on both the iot_modbus and iot_gps_tracker datasets. the comprehensive analysis of the model highlighted the impact of diverse resampling techniques on classification performance across varied classes, facilitating a deeper understanding of their implications and avenues for potential enhancements. previous research (azlim & ahamed, .2023; jiang et al., 2023; qawqzeh et al., 2023; rezvani & wang, 2023) have emphasised the symbiotic relationship between resampling techniques and the choice of classification methods, underscoring the necessity for synergy to maximise beneficial outcomes. absence of resampling in the absence of resampling, the model achieved 98% notable accuracy. while demonstrating perfect precision and high recall for the “injection” class and a commendable f1-score for “scanning,” lower f1-scores for “xss” and “backdoor” indicated areas for improvement. random over-sampling (ros) maintaining a consistent accuracy of 98%, ros exhibited strengths in perfect precision and high recall for the “injection” class, along with a substantial f1score for “scanning.” however, comparatively lower f1 scores for “password” and “xss” suggested potential areas for enhancement. ros effectively addressed class distribution imbalance. random under-sampling (rus) despite balancing class distribution, rus lead to a reduced accuracy of 84%. while achieving a perfect f1score for the “injection” class, the model performance was significantly weakened across other classes, resulting in low precision and recall. smote (synthetic minority over-sampling technique) smote achieved a commendable accuracy of 99%, maintaining excellent precision and recall for “normal” and “injection” classes, resulting in high f1 scores. however, relatively lower f1 scores for “xss” and “backdoor” suggested potential areas for improvement. smote effectively addressed class imbalance by generating synthetic instances. adasyn (adaptive synthetic sampling) similarly achieving a high accuracy of 99%, adasyn sustained robust precision and recall for “normal” and “injection” classes, resulting in high f1 scores. figure 5: iot_gps_tracker dataset classification reports of the used resampling techniques figure 6: heat maps showcase the resampling techniques employed on the iot_modbus dataset. pa ge 22 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 16-24, 2025 nevertheless, there was an area for enhancement in f1scores for “xss” and “backdoor,” signifying potential improvements. adasyn effectively mitigates class imbalance through adaptive synthetic sampling. this analysis offered insights into the relative effectiveness of each resampling approach concerning precision, recall, f1 scores, and accuracy across different classes. rus’s overall performance has been compromised despite its improvement in the “injection” class. these findings informed conclusions regarding the suitability of resampling techniques in addressing class imbalance within a multi-class scenario. heat maps were generated to represent the resampling techniques employed on the iot_modbus dataset, as shown in figures 6 and 7. conclusion the study examined the impact of resampling techniques on classification models in class-imbalanced iot datasets. it found that support vector classifier (svc) and random forest classifier (rf) performed biasedly on unbalanced datasets, highlighting the issue of class imbalance in machine learning tasks. resampling strategies improved the performance of svc and rf classifiers, with hybrid approaches like smote and oversampling techniques like adasyn enabling rebalancing class distributions and enhancing model performance. rf consistently outperformed svc in resampling scenarios, achieving superior accuracy, precision, recall, and f1-score. the study emphasized the importance of resampling techniques in scenarios marked by class imbalance to enhance the accuracy and reliability of classification models in practical applications. future research could explore hyperparameter tuning’s effects on model performance and explore the applicability of these techniques in domains like cybersecurity, fraud detection, and medical diagnosis. future implications this subsequent study may investigate several directions to expand our comprehension and improve the usefulness of resolving class imbalance in iot datasets. exploring hybrid or adaptive resampling methodologies might mitigate trade-offs observed in current techniques and potentially boost overall classification performance. examine how well ensemble learning methods-like bagging, boosting, or stacking-work with resampling techniques to enhance model robustness and performance in unbalanced iot datasets. by combining the advantages of several classifiers and resampling strategies, ensemble approaches may improve generalisation and prediction accuracy. the study explores adaptive resampling techniques that can dynamically adjust to changes in the data distribution. ensuring the ongoing efficacy of class imbalance mitigation strategies in practical applications may require creating algorithms that can recognise and react to idea drift, data drift, or changing class distributions in iot datasets. the study suggests that exploring the compatibility between advanced classification algorithms and resampling methods is a promising direction. this exploration could unveil enhanced performance in complex multi-class scenarios, presenting opportunities for more robust and accurate models. moreover, examining the practical implications and robustness of these techniques in real-world scenarios, particularly in domains where accurate classification is imperative, would be instrumental. this includes rigorous testing and validation of these techniques in operational settings to gauge their effectiveness and feasibility beyond controlled experimental setups. validate the effectiveness of the identified resampling techniques and classifiers through deployment in real-world iot environments. conduct extensive evaluation and monitoring of model performance under practical conditions, considering scalability, reliability, and interpretability factors. case studies and field trials in diverse iot domains could provide valuable insights into the applicability and impact of class imbalance mitigation strategies in real-world settings. the project proposes frameworks figure 7: heat maps showcase the resampling techniques employed on the iot_gps_tracker dataset pa ge 23 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 16-24, 2025 and automated methods for selecting suitable classifiers, hyperparameters, and resampling techniques for iot dataset’s properties. the model creation process might be streamlined by automated model selection and hyperparameter tweaking, allowing practitioners to quickly find and implement efficient predictive models in internet of things applications. the findings of this study provide a foundation for future research, emphasising the need for tailored techniques and their practical applications in addressing class imbalance within the dynamic landscape of iot datasets. the enhancement of the state-of-the-art in-class imbalance mitigation strategies for iot datasets by addressing these future research objectives will eventually improve the predictive modelling’s performance, scalability, and reliability in various iot applications. acknowledgment we would like to express our gratitude to yousef qawqzeha from the university of fujairah for his valuable contributions to this research. his expertise and dedication greatly enriched the development of this study. we also acknowledge the support provided by his corresponding author email (yousefqawqzehaa@ outlook.com) and his orcid (0000-0001-7774-062x) throughout the research process. references abdi, l., & hashemi, s. (2015). to combat multi-class imbalanced problems by means of over-sampling techniques. ieee transactions on knowledge and data engineering, 28(1), 238-251. atuhurra, j., hara, t., zhang, y., sasabe, m., & kasahara, s. (2024). dealing with imbalanced classes in botiot dataset. arxiv preprint arxiv:2403.18989. azlim khan, a. k., & ahamed hassain malim, n. h. (2023). comparative studies on resampling techniques in machine learning and deep learning models for drug-target interaction prediction. molecules, 28(4), 1663. coelho, d., costa, d., rocha, e. m., almeida, d., & santos, j. p. (2022). predictive maintenance on sensorized stamping presses by time series segmentation, anomaly detection, and classification algorithms. procedia computer science, 200, 1184-1193. collell, g., prelec, d., & patil, k. r. (2018). a simple plug-in bagging ensemble based on threshold-moving for classifying binary and multiclass imbalanced data. neurocomputing, 275, 330-340. dogra, v., verma, s., verma, k., jhanjhi, n. z., ghosh, u., & le, d.-n. (2022). a comparative analysis of machine learning models for banking news extraction by multiclass classification with imbalanced datasets of financial news: challenges and solutions. fahim, m., & sillitti, a. (2019). anomaly detection, analysis and prediction techniques in iot environment: a systematic literature review. ieee access, 7, 8166481681. fisher, a., rudin, c., & dominici, f. (2019). all models are wrong, but many are useful: learning a variable’s importance by studying an entire class of prediction models simultaneously. journal of machine learning research, 20(177), 1-81. hanskunatai, a. (2018). a new hybrid sampling approach for classification of imbalanced datasets. 2018 3rd international conference on computer and communication systems (icccs), huang, p. j. (2015). classification of imbalanced data using synthetic over-sampling techniques. university of california, los angeles. jiang, x., wang, j., meng, q., saada, m., & cai, h. (2023). an adaptive multi-class imbalanced classification framework based on ensemble methods and deep network. neural computing and applications, 35(15), 11141-11159. johnson, j. m., & khoshgoftaar, t. m. (2019). survey on deep learning with class imbalance. journal of big data, 6(1), 1-54. khodadadi, f., dastjerdi, a. v., & buyya, r. (2016). internet of things: an overview. internet of things, 3-27. koziarski, m., krawczyk, b., & woźniak, m. (2019). radial-based oversampling for noisy imbalanced data classification. neurocomputing, 343, 19-33. koziarski, m., woźniak, m., & krawczyk, b. (2020). combined cleaning and resampling algorithm for multi-class imbalanced data with label noise. knowledgebased systems, 204, 106223. kumar, a., sharma, s., singh, a., alwadain, a., choi, b.j., manual-brenosa, j., ortega-mansilla, a., & goyal, n. (2021). revolutionary strategies analysis and proposed system for future infrastructure in internet of things. sustainability, 14(1), 71. nixon, c., sedky, m., & hassan, m. (2019). practical application of machine learning based online intrusion detection to internet of things networks. 2019 ieee global conference on internet of things (gciot). nord, j. h., koohang, a., & paliszkiewicz, j. (2019). the internet of things: review and theoretical framework. expert systems with applications, 133, 97108. obaid, w., & nassif, a. b. (2022). the effects of resampling on classifying imbalanced datasets. 2022 advances in science and engineering technology international conferences (aset). paisitkriangkrai, s., shen, c., & van den hengel, a. (2013). a scalable stagewise approach to large-margin multiclass loss-based boosting. ieee transactions on neural networks and learning systems, 25(5), 1002-1013. pal, d., funilkul, s., charoenkitkarn, n., & kanthamanon, p. (2018). internet-of-things and smart homes for elderly healthcare: an end user perspective. ieee access, 6, 10483-10496. peng, h., wu, c., & xiao, y. (2023). cbf-ids: addressing class imbalance using cnn-bilstm with focal loss in network intrusion detection system. applied sciences, 13(21), 11629. pa ge 24 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 16-24, 2025 powroźnik, p., szcześniak, p., & piotrowski, k. (2021). elastic energy management algorithm using iot technology for devices with smart appliance functionality for applications in smart-grid. energies, 15(1), 109. pramanik, p. k. d., upadhyaya, b. k., pal, s., & pal, t. (2019). internet of things, smart sensors, and pervasive systems: enabling connected and pervasive healthcare. in healthcare data analytics and management (pp. 1-58). elsevier. qawqzeh, y. k., alourani, a., & ghwanmeh, s. (2023). an improved breast cancer classification method using an enhanced adaboost classifier. international journal of advanced computer science and applications, 14(1). qawqzeh, y. k., & ashraf, m. (2023). a fraud detection system using decision trees classification in an online transactions. proceedings of the 2023 12th international conference on software and computer applications. qawqzeh, y. k., bajahzar, a. s., jemmali, m., otoom, m. m., & thaljaoui, a. (2020). classification of diabetes using photoplethysmogram (ppg) waveform analysis: logistic regression modeling. biomed research international, 2020. rezvani, s., & wang, x. (2023). a broad review on class imbalance learning techniques. applied soft computing, 110415. https://doi.org/10.1016/j.asoc.2023.110415 rose, k., eldridge, s., & chapin, l. (2015). the internet of things: an overview. the internet society (isoc), 80(15), 1-53. sáez, j. a., krawczyk, b., & woźniak, m. (2016). analyzing the oversampling of different classes and types of examples in multi-class imbalanced datasets. pattern recognition, 57, 164-178. tanha, j., abdi, y., samadi, n., razzaghi, n., & asadpour, m. (2020). boosting methods for multiclass imbalanced data classification: an experimental review. journal of big data, 7, 1-47. tarawneh, a. s., hassanat, a. b., almohammadi, k., chetverikov, d., & bellinger, c. (2020). smotefuna: synthetic minority over-sampling technique based on furthest neighbour algorithm. ieee access, 8, 5906959082. tyagi, s., & mittal, s. (2020). sampling approaches for imbalanced data classification problem in machine learning. proceedings of icric 2019: recent innovations in computing, ullah, i., & mahmoud, q. h. (2021). a framework for anomaly detection in iot networks using conditional generative adversarial networks. ieee access, 9, 165907-165931. varotto, g., susi, g., tassi, l., gozzo, f., franceschetti, s., & panzica, f. (2021). comparison of resampling techniques for imbalanced datasets in machine learning: application to epileptogenic zone localization from interictal intracranial eeg recordings in patients with focal epilepsy. frontiers in neuroinformatics, 15, 715421. wanasinghe, t. r., gosine, r. g., james, l. a., mann, g. k., de silva, o., & warrian, p. j. (2020). the internet of things in the oil and gas industry: a systematic review. ieee internet of things journal, 7(9), 86548673. wang, s., & yao, x. (2012). multiclass imbalance problems: analysis and potential solutions. ieee transactions on systems, man, and cybernetics, part b (cybernetics), 42(4), 1119-1130. welvaars, k., oosterhoff, j. h., van den bekerom, m. p., doornberg, j. n., van haarst, e. p., olvg urology consortium, & r, t. m. l. c. v. d. z. j. v. a. g. l. b. h. m. k. p. b. l. v. d. k. a. m. w. p. (2023). implications of resampling data to address the class imbalance problem (ircip): an evaluation of impact on performance between classification algorithms in medical data. jamia open, 6(2), ooad033. zhou, x., hu, y., wu, j., liang, w., ma, j., & jin, q. (2022). distribution bias aware collaborative generative adversarial network for imbalanced deep learning in industrial iot. ieee transactions on industrial informatics, 19(1), 570-580. pa ge 1 pa ge 17 american journal of smart technology and solutions (ajsts) computer self-efficacy and effectiveness of quipper learning management system jude rafael s. alayacyac1, jobert c. regidor1, john harry s. caballo1*, jelly e. abellanosa1, giebe joshua g. monajan1 volume 3 issue 1, year 2024 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v3i1.2428 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: january 20, 2024 accepted: march 04, 2024 published: march 07, 2024 this quantitative study aimed to determine the relationship between computer selfefficacy and the effectiveness of the quipper learning management system among senior high school students. the study utilized a descriptive-correlational design. data on computer self-efficacy and quipper lms effectiveness were collected from 290 senior high school students using survey questionnaires. the results showed high overall computer self-efficacy (m=4.18, sd=0.652) and high effectiveness of the quipper lms (m=3.97, sd=0.622) among students. a significant positive correlation (r=0.414, p=.000) between computer self-efficacy and quipper lms effectiveness indicates a low positive relationship between the two variables. computer self-efficacy has a direct relationship with the effectiveness of the quipper lms. as computer self-efficacy increases, the effectiveness of the quipper lms also increases among senior high school students. keywords computer self-efficacy, learning management system, quipper lms 1 the university of mindanao, davao city, philippines * corresponding author’s e-mail: harrycaballo@gmail.com introduction most institutions have increasingly adopted online learning to facilitate teaching and learning as a continuum to the traditional face-to-face approach. most of these institutions utilize learning management systems, which contain features intended to make students active participants by delivering learning resources to learners and providing the environment for effective interaction in the learning process. it has been implemented in some universities worldwide to help connect students and lecturers without the confines of the traditional classroom. it is an environment with digital software designed to manage user learning interventions and deliver learning content and resources to students (adzharuddin, 2013). several studies have proven the effectiveness of lms on undergraduate courses, specifically information technology courses (nel, 2010), engineering courses (kurata, 2017), and english courses (salahuddin & saira, 2020). vesin et al. (2009) and chaubey et al. (2015) highlighted the role of lms in specific programs, such as vocational education and programming in java, in promoting active and cooperative learning and providing access and flexibility in higher education. further, rahman et al. (2019) and lasmanawati et al. (2021) emphasized the potential of lms to enhance learning by providing a user-friendly interface, facilitating independent and creative learning, and enabling learning anytime and anywhere however; other studies have opposing views on the effectiveness of learning management systems, highlighting its features are underutilized by students. however, students face several challenges preventing them from actively participating in online learning. there is a lack of individualized feedback, a lack of depth of learning, and a lack of interpersonal interactions using lms (reed, 2014; araka et al., 2021). students’ computer self-efficacy can influence their use of the lms. students with high computer self-efficacy will find using computers easily and be more inclined to use them. in contrast, students with low computer self-efficacy lack confidence in their computer skills and may avoid using computers (binyamin et al., 2018). this suggests that computer self-efficacy would influence students’ perception of e-learning since it is a critical predictor of perceived learning (alqurashi, 2019). it is also a key predictor in accepting e-learning (tarhini et al., 2015; fathema et al., 2015). even in the attitude toward artificial intelligence of university students, self-efficacy has a significant effect (obenza, 2023). additionally, computer self-efficacy was discovered to be a significant predictor of student satisfaction in lms (ghazal et al., 2018; hammouri & abu-shanab, 2018). as students’ computer self-efficacy grows, so does their perception that the system is simple to use and, therefore, their level of satisfaction with the learning management system (ghazal et al., 2018). several studies have examined the effectiveness of the quipper learning management system (lms) and its benefits for both teachers and students. morron (2015) stated that quipper lms is an effective online platform for improving student educational performance. supporting this, ghilay (2017) found that users view the lms as useful for convenient learning and positive assistance in the learning process. jamil et al. (2019) also noted that students see quipper school as a useful learning tool, as it activates interest in learning english more and interacting with the language. additionally, mahariyanti and suyanto (2018) found quipper school enables students to enhance their knowledge and assess their aptitude and understanding through provided questions. it increased student engagement and eagerness to learn in their study. the selection of a learning management system is critical to student success. that selection needs to be based on pa ge 18 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 17-21, 2024 the online course’s objectives and the student’s needs. the lms must have components allowing the instructor to create a course emphasizing active learning experiences (lewis et al., 2005). there are numerous studies investigating the effectiveness of lms per se. however, the researchers have not found studies specifically about quipper lms. there is also limited research correlating computer self-efficacy and the effectiveness of lms. thus, the researchers are motivated to pursue this research that may contribute to the existing knowledge about the learning management system. this may provide substantial data that would encourage the developers to improve the platform to a standard suited to the needs of the students. theoretical framework the study is anchored on the technology acceptance model developed by davis (1986). the goal of tam is to explain the general causes of computer acceptance, which will lead to a better understanding of users’ behavior across a wide range of end-user computing technologies and user groups. further, this study is also anchored on the first learning effectiveness model developed by lee (2008). the study suggested that the effectiveness of a learning platform can be determined by the willingness of the student to participate in any learning. also, it highlights the importance of the learning platform’s educational task content and the amount of support given by the organization. moreover, this study is supported by the proposition of yusoff (2009), who postulated that computer self-efficacy, a significant factor relating to achieving information and computing literacy skills, can lead to the ease of educational technology. research questions this study aimed to determine the relationship between computer self-efficacy and the effectiveness of the quipper learning management system on senior high school students in a non-sectarian private institution. specifically, this research study sought to answer the following research questions: 1.what is the level of computer self-efficacy of senior high school students in terms of: 1.1. basic computer skills; 1.2. media related skills; and 1.3. web-based skills? 2. what is the level of effectiveness of the quipper learning management system on senior high school students in terms of: 2.1. perceived self-efficacy; 2.2. perceived satisfaction; 2.3. perceived usefulness; 2.4. behavioral intention; 2.5. e-learning system quality; 2.6. interactive learning activities; 2.7. e-learning effectiveness; and 2.8. multimedia instruction? 3. is there a significant relationship between computer self-efficacy and the effectiveness of the quipper learning management system on senior high school students? materials and methods this research used a descriptive-correlational quantitative design. correlational research is used in research studies to determine a relationship between two or more variables and the degree of the relationship. in the correlational research design, researchers use the statistical correlation test to describe and measure the extent of association (or relationship) between two or more variables or sets of scores (creswell, 2012). simon and goes (2011) also stated that descriptive and correlational studies examine variables in their natural environment and do not include treatments required by researchers. in this study, the researchers aimed to determine the relationship between computer self-efficacy and the effectiveness of the quipper learning management system of senior high school students. a simple random sampling was used to determine the respondents for this study. in this type of sampling, the respondents are selected randomly and purely by chance. hence, the selection quality is not affected as every member has an equal chance of being selected in the sample. this type of sampling is best for a highly homogenous population (bhardwaj, 2019). specifically, the respondents of this study were 290 grade 11 and 12 students from accountancy, business, and management strand (abm), humanities and social sciences strand (humss), and science, technology, and mathematics strand (stem). to evaluate the relationship between computer selfefficacy and effectiveness of the quipper learning management system, the researchers used an adapted modified survey questionnaire from the study of amankwah et al. (2017) titled: computer self-efficacy among senior high school teachers in ghana and the functionality of demographic variables on their computer self-efficacy, for computer self-efficacy survey questionnaire. also, the researchers used an adapted modified survey questionnaire from the study of liaw and huang (2007) titled: investigating students’ perceived satisfaction, behavioral intention, and effectiveness of e-learning: a case study of the blackboard system, for the effectiveness of quipper learning management system survey questionnaire. in addition, the researchers performed a pilot test to assess the credibility and reliability of the questionnaire before the questionnaire was distributed for the conduction of the survey. also, the researchers selected 50 individuals to participate and answer the questionnaire, and its preparation for the actual administration and survey was validated. this was made sure through the use of cronbach’s alpha. the result for computer selfefficacy is 0.764, which signifies acceptable and reliable (11) questions within computer self-efficacy’s indicators. moreover, the result for the effectiveness of the quipper learning management system is 0.959, which means it is excellent, reliable, and internally consistent with pa ge 19 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 17-21, 2024 (26) questions within the effectiveness of the quipper learning management system’s indicators. to conduct the study, the researchers wrote a permission letter to the senior high school principal and questionnaires the respondents would fill out. the researchers asked the respondents to sign an informed consent before starting each survey; this form indicated their authorization to be included as respondents in the research study. the researchers ensured that the respondents understood their rights and the implications of participating. the survey questionnaires were distributed to the target respondents, grade 11 and 12 senior high school students. the role of the researchers then was to organize, present, and analyze the questionnaire data accordingly. the datagathering process followed proper procedures to collect quality data. the researchers used the following statistical tools to assess and evaluate data: mean this statistical concept is useful in figuring out the overall trend of a data set or presenting a rapid image of your data, generally referred to as average. this was used to measure the level of students’ computer self-efficacy and the quipper learning management system. standard deviation it refers to the measurement of the gap of the determined values from the mean (ilola, 2018). as used in this study, this tool was used to measure the dispersion of a dataset relative to the computed mean on the level of student’s computer self-efficacy and the quipper learning management system. pearson-r correlation coefficient (r-value) it is a statistical measure of the strength of the relationship among two variables. a typical correlation coefficient ranging from -1.00 to 1.00 shows a strong negative and effective association (ganti, 2020). this was used to measure the significant relationship between computer self-efficacy and the effectiveness of the quipper learning management system on senior high school students. results and discussion level of computer self-efficacy the data presented in table 1 shows the level of computer self-efficacy of senior high school students as regards basic computer skills, media-related skills, and web-based skills. the results in table 1 revealed a high overall level of computer self-efficacy among students, with a mean score of 4.18 (sd = 0.652). students expressed the greatest confidence in their web-based skills (m=4.43, sd=0.684), while media-related skills received the lowest scores (m=3.70, sd=0.804). basic computer skills were rated highly (m=4.38, sd=0.692). these findings support previous research on computer self-efficacy. binyamin et al. (2018) found that students with high computer selfefficacy are likelier to use computers easily, while those lacking confidence may avoid using them. hammouri and abu-shanab (2018) underscored that self-confidence with technology affects perceptions of difficulty and usefulness. moreover, khan (2018) noted that computer use enables efficient assessment and online learning. level of effectiveness of quipper learning management system the data gathered from the survey corresponds to the level of effectiveness of the quipper learning management system in terms of perceived self-efficacy, perceived satisfaction, perceived usefulness, behavioral intention, e-learning system quality, interactive learning activities, e-learning effectiveness, and multimedia instruction among senior high school students are presented in table 2. table 1: level of computer self-efficacy indicators x̄ sd basic computer skills 4.38 0.692 media related skills 3.70 0.804 web-based skills 4.43 0.684 overall 4.18 0.652 table 2: level of computer self-efficacy indicators x̄ sd perceived self-efficacy 4.09 0.762 perceived satisfaction 4.00 0.658 perceived usefulness 4.21 0.676 behavioral intention 4.14 0.691 e-learning system quality 3.78 0.740 interactive learning activities 3.82 0.752 e-learning effectiveness 3.93 0.715 multimedia instruction 3.61 0.818 overall 3.97 0.622 table 2 presented an overall high level of effectiveness for the quipper lms, with a mean of 3.97 (sd = 0.622). all indicators have a high descriptive level. perceived usefulness got the highest level, with a mean of 4.21 and a standard deviation of 0.676. moreover, multimedia instruction has the lowest level, with a mean of 3.61 and a standard deviation of 0.818. perceived self-efficacy got a mean of 4.09 and a standard deviation of 0.762. perceived satisfaction got a mean of 4.00 and a standard deviation of 0.658. behavioral intention got a mean score of 4.14 and a standard deviation of 0.691. furthermore, e-learning system quality obtained a mean of 3.78 and a standard deviation of 0.740. interactive learning activities got a mean of 3.82 and a standard deviation of 0.752. and e-learning effectiveness got a mean of 3.93 and a standard deviation of 0.715. prior research aligns with the current findings, which pa ge 20 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 17-21, 2024 demonstrate the effectiveness of the quipper lms for enhancing student learning, engagement, and performance through its interactive features and resources. morron (2015) found that the quipper lms is an effective online platform for improving students’ academic performance. ghilay (2017) added that users find the lms helpful for the convenience of learning and that it positively contributes to the learning process. moreover, jamil et al. (2019) stated that quipper school is an effective learning tool because it triggered students’ interest in learning english. mahariyanti and suyanto (2018) also stated that quipper school allows students to enhance their knowledge of topics and assess their understanding of information through teacher-provided materials and questions. significant relationship between computer selfefficacy and effectiveness of quipper learning management system table 3 shows that the computed r-value of computer self-efficacy and effectiveness of the quipper learning system is 0.414 (p-value=.000). this means a significant relationship exists between the two at a five percent level of significance. further, the computed r-value of computer self-efficacy to the effectiveness of the quipper learning management system is 0.414, which indicates a low positive relationship. this shows that there is a direct relationship between the two variables, which means that computer self-efficacy increases, the effectiveness of the quipper learning management system increases as well, and vice versa. the result supports alqurashi (2019), who stated that computer self-efficacy levels influence students’ perception of e-learning benefits since selfefficacy is a critical predictor of perceived learning. also, ghazal et al. (2018) found that as students’ computer self-efficacy increases, so does their perception that the learning management system is easy to use. moreover, the result aligns with the proposition of yusoff (2009), who postulated that computer self-efficacy, an important factor in achieving information and computing literacy, can lead to ease in using educational technology. learning management system received a satisfactory rating from the students. they were pleased with quipper lms’ functions, contents, and multimedia instruction. this indicates the students have the knowledge and skills to use quipper effectively as a learning management system. moreover, there is a low but significant positive correlation between senior high school students’ computer selfefficacy and the effectiveness of the quipper learning management system. this shows that there is a direct relationship between the two variables, which means that computer self-efficacy increases, and the effectiveness of the quipper learning management system increases as well, and vice versa. recommendation based on the results of this study, the following recommendations can be made: • continue using the quipper learning management system in senior high schools, as students find it satisfactory and effective overall. the system should be maintained and updated regularly to ensure optimal performance. • provide training and support for teachers on fully utilizing all features and content of the quipper lms to enhance instruction and student engagement. • develop supplemental digital literacy programs to further build students’ computer self-efficacy, particularly in web-based skills. • share study findings with other senior high schools to encourage quipper lms adoption and highlight its benefits for supporting student learning and self-efficacy. • consider research on the relationship between computer self-efficacy and learning management system effectiveness in other educational contexts and age groups references adzharuddin, n. a. (2013). learning management system (lms) among university students: does it work? international journal of e-education, e-business, e-management and e-learning. alqurashi, e. (2019). predicting student satisfaction and perceived learning within online learning environments. distance education, 40(1), 133–148. https://doi.org/10.1080/01587919.2018.1553562 amankwah, f., konin, d., sarfo, f. (2017). computer self-efficacy among senior high school teachers in ghana and the functionality of demographic variables on their computer self-efficacy, tojet: the turkish online journal of educational technology, 16(1). jamil, w., wahfiuddin, a., ampa, a., & ilmiah, i. (2019). the students’ learning interest of quipper school used by the teacher in teaching english: descriptive research. exposure: jurnal pendidikan bahasa inggris, 8, 141-157. https://doi.org/10.26618/exposure.v8i2.2800 araka, e., maina, e.m., gitonga, r., & oboko, r. o. (2021). university students’ perception on the usefulness of learning management system features table 3: significant relationship between computer self-efficacy and quipper learning management system variables r – value p-value computer self-efficacy .414 .000 effectiveness of quipper learning management system conclusion this study found that senior high school students in a non-sectarian private institution have a generally good level of computer self-efficacy. this includes competence in basic computer skills, media-related skills, and webbased skills. students find computer use straightforward and generally understand computing concepts and abilities. additionally, the effectiveness of the quipper pa ge 21 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 17-21, 2024 in promoting self-regulated learning in online learning. bhardwaj, j. (2019). types of sampling in research. j pract cardiovasc sci, 5(3), 157–163. binyamin, s. s., rutter, m. j., & smith, s. (2018). the influence of computer self-efficacy and subjective norms on the students’ use of learning management systems at king abdulaziz university. international journal of information and education technology (ijiet), 8(10), 693–699. https://doi.org/10.18178/ ijiet.2018.8.10.1124 chaubey, a., & bhattacharya, b. (2015). learning management system in higher education. international journal for science technology and engineering, 2, 158-162. creswell (2012), descriptive correlational design: chapter iii research methodology, 3.1 research design, 338 http:// repository.upi.edu/7096/7/s_ing_0900713_ chapter3.pdf davis, f. d. (1986). a technology acceptance model for empirically testing new end-user information systems: theory and results. massachusetts, united states: sloan school of management, massachusetts institute of technology. fathema, n., shannon, d., & ross, m. (2015). expanding the technology acceptance model (tam) to examine faculty use of learning management systems (lmss) in higher education institutions. journal of online learning and teaching, 11(2), 210–232. ganti, a. (2020). correlation coefficient. https://www. investopedia.com/terms/c/correlationcoefficient.asp ghazal, s., aldowah, h., umar, i., & bervell, b. (2018). acceptance and satisfaction of the learning management system enabled blended learning based on a modified delone-mclean information system success model. international journal of information technology project management, 9(3), 52–71. https://doi. org/10.4018/ ijitpm.2018070104 ghilay, y. (2017). online learning in higher education. new york, ny: nova science publishers hammouri, q., & abu-shanab, e. (2018). exploring factors affecting users’ satisfaction toward e-learning systems. international journal of information and communication technology education, 14(1), 44–57. https://doi.org/10.4018/ijicte.2018010104 ilola e. (2018). a beginner’s guide to standard deviation and standard error. https://s4be.cochrane.org/ blog/2018/09/26/a-beginners-guide-to-standarddeviation-and standard-error/ khan, a., 2018. advantages and disadvantages of computer skills for students in their school life. https://www.jagranjosh.com/articles/advantagesand-disadvantages-of-computer-skills-for-studentsin-their-school-life-1517293374-1 kurata, y. b., bano, r. m., & marcelo, m. c. (2017). effectiveness of learning management system application in the learnability of tertiary students in an undergraduate engineering program in the philippines. lasmanawati, e., muktiarni, m., & maosul, a. (2021). analysis learning management system in vocational education. iop conference series: materials science and engineering, 1098. lee, j. a. (2008). the relationship of e-learners’ selfregulatory efficacy and perception of e-learning environmental quality. computers in human behavior, 24(1), 32-47. lewis, b., macentee, v. m., delacruz, s., englander, c., jeffrey, t. r., takach, e., wilson, s. j., & woodall, j. (2005). learning management systems comparison. liaw, s. s., & huang, h. m. (2007). developing a collaborative e-learning system based on users’ perceptions. lecture notes in computer science, 4402, 751–759. mahariyanti, e., & suyanto, s. (2018). the effectiveness of using quipper school teaching materials towards students’ cognitive outcomes of eleventh grade students of mipa. advances in social science, education and humanities research (assehr), 330. morron, a. (2015). the effect of quipper school on the performance of grade 7 students of bacoor national high school – gawaran in teaching mathematics. nel, l. (2010). blended learning in south african undergraduate classrooms: directives for the effective use of a learning management system. obenza, b. n., baguio, j. s. i. e., bardago, k. m. w., granado, l. b., loreco, k. c. a., matugas, l. p., talaboc, d. j., zayas, r. k. d. d., caballo, j. h. s., & caangay, r. b. r. (2023). the mediating effect of ai trust on ai self-efficacy and attitude toward ai of college students. international journal of metaverse, 2(1), 1–10. https://doi.org/10.54536/ijm.v2i1.2286 rahman, m.e., daud, m., & ensimau, n.k. (2019). learning management system (lms) in teaching and learning. the international journal of academic research in business and social sciences, 9. reed, p. (2014). staff experience and attitudes towards technology-enhanced learning initiatives in one faculty of health and life sciences. research in learning technology, 22, 1-23. https://doi.org/10.3402/rlt. v22.22770 salahuddin, a., ajmal, f., & saira (2020). effectiveness of learning management system for teaching english language at higher education level. simon, m. k., & goes, j. (2011). correlational research. https://pdf4pro.com/view/correlational-researchdissertationrecipes-com-17d6e5.html tarhini, a., arachchilage, n., & abbasi, m. (2015). a critical review of theories and models of technology adoption and acceptance in information system research. international journal of technology diffusion, 6(4), 58–77. https://doi.org/10.4018/ijtd.2015100104 vesin, b., ivanović, m., & budimac, z. (2009). learning management system for programming in java. yusoff, m. y. (2009). individual differences, perceived ease of use, and perceived usefulness in e-library usage. journal of computer and information science, 2(1). pa ge 1 pa ge 8 american journal of smart technology and solutions (ajsts) interactive learning in afghanistan: feasibility of implementing iot connected devices in classrooms ansarullah hasas1, sayed najmuddin sadaat2, musawer hakimi3*, mohammad mustafa quchi4 volume 3 issue 1, year 2024 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v3i1.2342 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: december 17, 2023 accepted: january 20, 2024 published: january 22, 2024 in the dynamic landscape of afghanistan’s education, this study explores into the transformative integration of the internet of things (iot). with the purpose of scrutinizing perceptions, challenges, and preparedness, a meticulous mixed-methods approach was employed, collecting data from 120 participants across diverse educational backgrounds. the deliberate inclusion of participants from computer science, agriculture, education, and economics faculties, representing institutions such as kabul, karwan, badakhshan, samangan, and faryab university, ensured a comprehensive exploration. through structured questionnaires utilizing likert-scale items, the methodology sought to gauge participants’ perceptions of technological infrastructure, institutional support, and challenges linked to iot integration. notably, anova and regression analyses were applied to discern influential factors shaping participants’ views. the study’s novelty lies in its meticulous uncovering of nuanced perceptions, specifically emphasizing the pivotal role of technical expertise perception in the readiness for iot integration within educational faculties. the consensus on iot’s potential to reshape teacher-student interactions and elevate academic outcomes is a novel contribution, highlighting the transformative impact on pedagogical practices. additionally, the research anticipates and scrutinizes technological and infrastructural challenges, identifying resource constraints, and proposing targeted strategies for efficacious iot implementation. in conclusion, this research enriches the discourse on iot in education by accentuating the significance of institutional support, addressing challenges, and proposing practical guidelines. it not only contributes actionable insights for afghanistan policymakers and educators but also lays a foundation for future research to optimize strategies in specific challenges, ultimately fostering successful iot integration in afghanistan’s educational landscape. keywords internet of things (iot), educational technology, afghanistan classrooms, institutional support pedagogical implications 1 department of information technology, kabul university, afghanistan 2 faculty of computer science, kabul university, afghanistan 3 department of computer science, samangan university, afghanistan 4 department of network engineering, faryab university, afghanistan * corresponding author’s e-mail: musawer@adc.edu.in introduction the integration of internet of things (iot) technologies into educational settings stands as a transformative force, promising novel approaches to teaching and learning. as the global landscape of education evolves, the application of iot in educational environments has garnered attention for its potential to revolutionize traditional pedagogical methods. the significance of this integration lies not only in its technological advancements but also in its capacity to reshape the very essence of educational experiences, particularly in regions like afghanistan (ahmad, 2018). in recent years, scholars have increasingly recognized the potential of iot in education (atzori et al., 2010). the iot, characterized by the interconnection of physical devices, offers unprecedented opportunities to enhance the educational landscape. as explored by (atzori et al., 2010), the convergence of social networks and iot can pave the way for innovative educational practices, creating a dynamic and interactive learning environment. afghanistan, with its unique socio-cultural context, stands at the cusp of this technological transformation. the intricate iot network marks a paradigm shift, linking physical entities uniquely. amid various iot reviews, a notable gap exists in comprehensive education implications exploration, filled by this study. offering insights into medical, vocational, green iot, and wearables, it references (al-emran, malik, and al-kabi, 2020) seminal survey on opportunities and challenges, shaping the iot landscape in education. the integration of the internet of things (iot) in education has revolutionized institutions by enabling communication among physical objects and fostering a novel interaction between individuals and their environment. this study categorizes iot applications in education, such as energy management, real-time ecosystem monitoring, student healthcare, and improved teaching methods. utilizing the canvas business model, the research examines how iot has transformed the education business model, introducing new value propositions (bagheri & movahed, 2016). the pervasive influence of the internet of things (iot) extends across various sectors, from smart cities to education, offering transformative possibilities. this survey explores university students’ perspectives on iot in education, delving into their awareness, knowledge, and receptiveness to learning about this evolving technology (suduc et al., 2018). as noted by (dake et al., 2023), the educational landscape in afghanistan faces various challenges, and the integration of iot could offer a solution to address some of these issues. however, it is imperative pa ge 9 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 8-16, 2024 to understand the current state of technological infrastructure and readiness of educational institutions in afghanistan before envisioning the implementation of iot-connected devices in classrooms (russell et al., 2014). the feasibility and readiness of iot integration in afghanistan classrooms constitute a critical aspect of this discussion. the second research objective revolves around assessing the preparedness of educational institutions and their technical expertise to seamlessly integrate iot technologies. this aligns with the findings of (mukhopadhyay et al., 2014) who emphasize the importance of evaluating the challenges and opportunities associated with iot deployment. moreover, examining the pedagogical implications of iot integration becomes crucial in understanding how this technological shift can impact teacher-student interaction and overall learning outcomes in afghanistan. the third research objective is geared towards uncovering the potential enhancements that iot-connected devices can bring to the educational experience in this specific context (russell et al., 2014). as the adoption of iot in education remains in its early stages, especially in developing countries like afghanistan, there is a pressing need to explore the challenges hindering its implementation. this is consistent with the observations made by (malik & al-emran, 2018), who stress the importance of recognizing and addressing challenges to ensure the effective implementation of technology in education. in conclusion, this research aims to provide a comprehensive understanding of the feasibility and challenges associated with integrating iot-connected devices in afghanistan classrooms. through an exploration of the current technological landscape, institutional readiness, pedagogical implications, and potential challenges, this study aspires to contribute valuable insights that can inform the successful implementation of iot in the afghanistan educational context. problem statement in the context of afghanistan’s educational landscape, the integration of the internet of things (iot) presents a transformative potential, yet the existing challenges necessitate careful consideration. the problem at hand lies in the need to address and overcome the multifaceted barriers hindering the seamless implementation of iot in educational settings. technological and infrastructural challenges, such as connectivity issues and compatibility concerns, pose significant hurdles that must be navigated. moreover, the scarcity of resources adds a layer of complexity, demanding targeted strategies to optimize the use of available resources. the readiness and perceptions of stakeholders, including educators and administrators, play a crucial role, requiring a nuanced approach to address varying confidence levels. additionally, while there is a growing acknowledgment of the importance of institutional support, its effective integration into the iot implementation framework remains a critical challenge. this problem statement underscores the imperative for a comprehensive exploration of these challenges and the development of tailored strategies to unlock the full potential of iot in afghanistan’s educational landscape. literature review the emergence of the internet of things (iot) has ushered transformative possibilities into education, reshaping traditional teaching and learning paradigms. at its core, iot involves connecting devices and objects to establish an environment where data is seamlessly collected, analyzed, and utilized to enhance various aspects of education. seminal work by (atzori et al., 2010) underscores iot’s potential in creating smart educational environments, fostering interactive and personalized learning experiences. the convergence of iot and educational technologies, explored by (atzori et al., 2010), significantly impacts pedagogical approaches, enabling dynamic and adaptive learning environments. in the context of developing countries, (dake et al., 2023) express cautious optimism, asserting that iot integration offers a promising avenue for addressing challenges in afghanistan’s educational landscape, specifically highlighting the potential of iot-connected devices in enhancing internet safety and cybersecurity awareness among students in the region. however, the literature emphasizes the need for a comprehensive assessment of readiness and challenges associated with implementing iot in education. (mukhopadhyay and suryadevara, 2014) stress recognizing challenges and opportunities, aligning with the research objective to identify technological and infrastructural challenges in afghanistan’s educational context (malik & al-emran, 2018; bao et al., 2017). in their work, (abdel-basset et al., 2018) emphasize the transformative role of the internet of things (iot) in education, particularly within smart learning environments. they highlight the potential of iot technologies to enhance the learning process by adapting to diverse student needs through information sensing devices and processing platforms, ultimately improving the overall quality of education (abdel-basset et al., 2018). the pedagogical implications of iot in education, highlighted by russell et al. (2014), emphasize its potential to redefine teacher-student interactions and enhance overall learning experiences. this aligns with the third research objective, exploring potential enhancements that iot-connected devices can bring to the educational landscape in afghanistan. moreover, the literature underscores the significance of comprehensive guidelines for successful iot implementation. (abed et al., 2019) stress the importance of clear guidelines in shaping effective strategies, providing a roadmap for educational stakeholders (singh et al., 2019). integrating iot and cloud computing in education enhances student engagement and administrative efficiency (j et al., 2020). this transformation empowers students with new technologies and revolutionizes traditional teaching methods. the result is a smart and sustainable campus with improved connectivity and pa ge 10 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 8-16, 2024 security (j et al., 2020). building upon existing literature, this research aligns with the call for a nuanced exploration of iot in specific educational contexts. (hakimi et al., 2024) delve into the impact of e-learning on girls’ education at samangan university in afghanistan, emphasizing the need for robust infrastructure, teacher training, and addressing social barriers. the study, based on surveys from 106 female students, underscores the positive aspects of e-learning while highlighting persisting challenges, particularly gender discrimination. the study emphasizes the imperative of cybersecurity education in badakhshan province, afghanistan, recognizing the dual nature of the internet – a source of opportunities and risks. the study advocates for cultivating digital literacy and online safety principles among students to address cyberbullying, privacy breaches, and security threats. through rigorous analysis, the research aims to empower youth with ethical discernment, fostering a safer and more knowledgeable society (fazil et al., 2023). this study investigates the transformative influence of cutting-edge technologies, including blockchain, artificial intelligence, augmented reality, and the internet of things, on afghanistan’s tourism sector. empirical findings reveal positive perceptions and significant impacts on operational efficiency, productivity, and financial profitability. the research provides actionable recommendations for strategic technology adoption, capacity building, and cybersecurity prioritization in afghanistan’s tourism industry (hakimi et al., 2023). the rapid expansion of the internet of things (iot), expected to reach 20.4 billion devices by 2020, has spurred its integration into diverse sectors, including education. with a focus on the pedagogical processes involving faculty, students, and educational assets, our systematic literature review addresses the benefits and challenges of incorporating iot. this research, in line with (kassab, defranco, and laplante, 2020) work, offers a comprehensive overview and identifies unexplored research questions in the iot and education landscape. research objectives • to analyze afghanistan’s educational technology infrastructure for insights into interactive learning tools and technological readiness. • to investigate the feasibility of integrating iotconnected devices in afghanistan educational institutions, considering infrastructure, technical capabilities, and institutional support. • to explore the pedagogical impact of iot-connected devices on teaching methods, student engagement, and learning outcomes in afghanistan classrooms. • to identify technological and infrastructural challenges related to iot implementation in afghanistan classrooms, addressing issues like connectivity and device compatibility. • to develop guidelines and recommendations based on findings to guide stakeholders in successfully integrating iot-connected devices, ensuring optimal interactive learning experiences in afghanistan. the conceptual framework presented in above outlines table 1: components of conceptual framework for iot integration in education components description technological infrastructure evaluation of existing iot devices and connectivity within educational institutions. assessment of the compatibility and scalability of iot technologies in the educational context. exploration of participants' perceptions regarding the technological readiness for iot integration. institutional support examination of the role of institutional policies and strategies in fostering a conducive environment for iot implementation. analysis of the level of support and engagement from educational leadership in promoting iot initiatives. investigation into the influence of institutional culture on participants' attitudes toward iot integration. pedagogical implications scrutiny of the impact of iot on teacher-student interactions and overall learning experiences. identification of changes in teaching methods and learning outcomes attributed to the integration of iot technologies. evaluation of the alignment between iot integration and the overall education business model. sources: (dake et al., 2023; abed et al., 2019; singh et al., 2019) the key components essential for investigating the integration of the internet of things (iot) in educational settings. this structured approach encompasses three main dimensions: technological infrastructure, institutional support, and pedagogical implications. technological infrastructure: this component focuses on evaluating the existing iot devices and connectivity within educational institutions. it emphasizes the assessment of compatibility and scalability of iot technologies in the educational context. additionally, exploring participants’ pa ge 11 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 8-16, 2024 perceptions regarding the technological readiness for iot integration is crucial. this dimension ensures a comprehensive understanding of the technological landscape necessary for successful implementation. institutional support: examining the role of institutional policies and strategies in fostering a conducive environment for iot implementation is a critical aspect. the analysis delves into the level of support and engagement from educational leadership, recognizing their influence on promoting iot initiatives. investigating the impact of institutional culture on participants’ attitudes toward iot integration adds a sociocultural perspective, acknowledging the importance of organizational factors. pedagogical implications: this dimension scrutinizes the impact of iot on teacher-student interactions and overall learning experiences. it identifies changes in teaching methods and learning outcomes attributed to the integration of iot technologies. the evaluation of alignment between iot integration and the overall education business model ensures that the pedagogical implications are considered in the broader context of educational practices. interconnected exploration: the components presented in table 1 are interconnected, emphasizing a holistic approach to investigating iot integration. the success of iot implementation in education relies on the synergy between technological advancements, institutional support, and pedagogical considerations. this framework provides researchers and educators with a comprehensive guide to exploring the multifaceted aspects of iot integration in educational settings, paving the way for informed decision-making and successful implementation strategies. materials and methods this research employed a quantitative research design to systematically investigate the perceptions and factors shaping the integration of the internet of things (iot) in educational settings within afghanistan. the study included a diverse participant sample of 120 individuals with backgrounds spanning computer science, agriculture, education, and economics faculties. to ensure comprehensive representation, various universities, including kabul, karwan, badakhshan, samangan, and faryab university, were selected for the survey, incorporating perspectives from both students and teachers. structured questionnaires served as the primary research instruments, thoughtfully crafted to assess participants’ views on existing technological infrastructure, institutional support, and challenges associated with iot integration. the utilization of likert-scale items enhanced the precision of capturing nuanced opinions. rigorous data analysis techniques, including descriptive statistics, anova, and regression analysis, were applied to identify statistically significant factors influencing perceptions. the research methodology prioritized ethical considerations, emphasizing participant confidentiality and voluntary involvement. prior to participation, informed consent was obtained, and participants received a comprehensive debriefing regarding the study’s objectives. anticipated findings are poised to offer invaluable insights for educational stakeholders and policymakers, facilitating the formulation of strategic approaches for the effective and successful implementation of iot in afghanistan classrooms. results and discussion in this compelling investigation, the results unfold to reveal a rich tapestry of insights. through a meticulous blend of graphical representations and tabular formats, the data takes on a visually engaging form, inviting the audience into a detailed exploration of academic trends. the strategic use of both modalities enhances accessibility, ensuring that the findings are not only informative but also captivating. the carefully curated presentation of results serves as a beacon, guiding the audience through the intricacies of thematic distributions and institutional contributions. this approach not only fosters clarity but also transforms the results into a dynamic narrative, making the exploration of academic figure 1: faculty of participants landscapes an engaging and enlightening experience. the above pie chart illustrates the distribution of academic research papers across four distinct categories: computer science, agriculture, education, and economic. notably, agriculture dominates the chart with the largest share at 45.8%, signifying a substantial focus on this field within the academic landscape. computer science and education each contribute 23.3%, reflecting a balanced distribution of research attention. economic studies, with a share of 7.5%, represent a smaller but still noteworthy portion. the percentages sum to 100%, providing a clear visual representation of the diverse thematic interests pursued in this academic context. the data in figure 2 highlights the distribution of 120 academics across five universities. karwan university has the largest share with 30%, followed closely by samangan university at 28.3%. badakhshan university and kabul university contribute 20% and 12.5%, respectively, while faryab university holds the smallest share with 9.2%. this concise analysis provides a snapshot of the research output from each university, showcasing the varying levels of scholarly contribution within this academic pa ge 12 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 8-16, 2024 context. the total percentage equates to 100%, offering a comprehensive overview of the distribution. the above pie chart depicting participant occupations in the academic setting reveals a predominant focus on students, occupying a significant 80.8% of the distribution. in contrast, teachers constitute 19.2%, representing a proportionately smaller segment. this visual representation underscores the central role of students in the academic environment. the chart’s clarity and the distinct division between students and teachers contribute to a concise and informative portrayal of the participant distribution, with the total percentage summing up to 100%. table 2 indicate that participants, on average, hold moderately neutral to slightly disagreeing perceptions regarding the existing technological infrastructure and institutional technical capabilities in afghanistan classrooms. the mean scores for “existing technological infrastructure” (mean = 1.3417) and “institutional technical capabilities” (mean = 1.3417) suggest a leaning towards disagreement, while “evaluation of technological infrastructure” (mean = 1.6583) leans slightly towards agreement. the overall construct “t1” reflects a positive inclination towards agreement, with a mean score of 4.3417. the consistently low standard deviation (0.47626) across all variables indicates a degree of consensus among participants. these statistical values provide a snapshot of the perceptions but also highlight the need for further exploration to understand the factors influencing these perceptions comprehensively. this aligns with the objective of conducting an indepth analysis of the existing educational technology infrastructure in afghanistan figure 2: university of participants figure 3: occupation of participants table 2: evaluation of current educational technology landscape n minimum maximum mean std. deviation existing technological infrastructure 120 1.00 2.00 1.3417 .47626 evaluation of technological infrastructure 120 1.00 2.00 1.6583 .47626 institutional technical capabilities 120 1.00 2.00 1.3417 .47626 t1 120 4.00 5.00 4.3417 .47626 valid n (listwise) 120 table 3 results provide statistical evidence supporting the objective of investigating the feasibility and preparedness of educational institutions in afghanistan for the integration of iot-connected devices. the regression model, including predictors like technical expertise perception and preparedness for iot integration, shows a significant impact on the dependent variable, education faculty (f=10.587, p=.000). table 3: assessment of feasibility and readiness of iot integration model sum of squares df mean square f sig. 1 regression 13.685 2 6.842 10.587 .000b residual 75.615 117 .646 total 89.300 119 pa ge 13 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 8-16, 2024 the mean squares for technical expertise perception and preparedness for iot integration indicate substantial variability in participants’ perceptions, contributing significantly to the overall model. the statistical significance (p=.000) underscores the importance of these factors in gauging preparedness for iot integration. participants’ responses suggest a moderately positive perception, as indicated by the regression model’s significant impact. the mean squares for the predictors highlight the variance in opinions regarding technical expertise and preparedness for iot integration. further exploration, possibly through additional statistical analyses or qualitative methods, can provide deeper insights into the specific factors influencing these perceptions, contributing to the investigation’s overall objective. table 4 results indicate a significant impact of institutional table 4: examination of pedagogical implications of iot integration sum of squares df mean square f sig. institutional support rating between groups 3.322 1 3.322 16.559 .000 within groups 23.670 118 .201 total 26.992 119 teacher-student interaction and engagement between groups 3.322 1 3.322 16.559 .000 within groups 23.670 118 .201 total 26.992 119 learning outcomes and academic performance confidence between groups 3.322 1 3.322 16.559 .000 within groups 23.670 118 .201 total 26.992 119 table 5: iidentification of technological and infrastructural challenges model sum of squares df mean square f sig. 1 regression 12.205 1 12.205 18.680 .000b residual 77.095 118 .653 total 89.300 119 support rating on perceived enhancements in teacherstudent interaction, engagement, and learning outcomes through iot integration (f=16.559, p=.000). the mean squares for each factor reveal substantial between-groups variance, underscoring the importance of institutional support in shaping these perceptions. participants’ responses to the exploration of iot impact align with a moderately positive stance. on average, there is a notable belief that integrating iot-connected devices can significantly enhance teacher-student interaction and engagement (mean square = 3.322). moreover, the confidence in iot’s potential to improve overall learning outcomes and academic performance is similarly high (mean square = 3.322). statistical significance (p=.000) reinforces the robustness of these perceptions. the narrow standard deviation (0.201) across factors suggests a degree of consensus among participants. in conclusion, the data reflects a positive outlook regarding the pedagogical impact of iot integration in afghanistan classrooms. the statistical analysis not only underscores the perceived enhancements in teacher-student dynamics and academic outcomes but also emphasizes the crucial role of institutional support. this outcome provides valuable insights for educational stakeholders and policymakers, indicating a readiness to embrace iot technologies for improved teaching and learning experiences in the afghanistan educational context. table 5 analysis reveals a significant impact of the predictor variable “significance of resource constraints in afghanistan iot integration” on the perceived challenges within the “education faculty” related to iot implementation (f=18.680, p=.000). this emphasizes the influential role of resource constraints in shaping perceptions. connectivity challenges: participants moderately perceive issues related to connectivity for iot implementation, recognizing its importance for success. compatibility concerns: there is a moderate anticipation of compatibility issues with existing devices and infrastructure, highlighting awareness of potential challenges. resource constraints: resource constraints, notably budget limitations, are viewed as moderately significant, emphasizing their impact on addressing iot-related challenges. overall, the statistical analysis underscores participants’ awareness of key technological and infrastructural challenges. the significant relationship between resource constraints and perceived challenges signifies the crucial role of adequate resources for successful iot integration. these findings offer valuable insights for formulating targeted strategies to address these challenges and enhance the prospects of iot implementation in afghanistan classrooms. pa ge 14 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 8-16, 2024 table 6 reveals robust support for the proposed guidelines for successful implementation of iot-connected devices in afghanistan educational settings: importance of guidelines: participants overwhelmingly perceive the importance of comprehensive guidelines, with a mean of 4.34, indicating a high consensus. guidelines’ impact: the mean value of 4.34 for the impact of guidelines echoes their perceived importance, emphasizing their potential positive influence on iot implementation. confidence in guidelines: while confidence levels vary (mean=2.98, std. dev.=1.43), the overall moderate confidence suggests a willingness to engage with proposed guidelines. considering statistical consistency and the high mean values for importance and impact, these findings indicate a solid foundation for research objective 5. the data underscores the significance participants attribute to well-defined recommendations and their potential in overcoming challenges for optimal iot implementation. addressing the varying confidence levels may involve tailored strategies to enhance stakeholder assurance and ensure effective utilization of the proposed guidelines in the diverse landscape of afghanistan educational institutions. disussion the exploration of internet of things (iot) integration in afghanistan’s educational landscape, as illuminated by participant perceptions, offers a nuanced perspective that necessitates careful consideration and strategic planning. the evaluation of the current educational technology landscape, depicted in table 1, reflects a moderately neutral to slightly disagreeing stance regarding existing technological infrastructure and institutional technical capabilities. this alignment with cautious optimism, as emphasized by (dake et al., 2023), underscores the importance of acknowledging both challenges and opportunities within the educational landscape. the consensus among participants, evident in the low standard deviation, signals a collective awareness of the existing state, urging further exploration to comprehensively understand the intricacies influencing these perceptions, aligning with the overarching research objective. (malik and al-emran, 2018; bao et al., 2017) advocate for a holistic assessment of the educational context, reinforcing the need for a comprehensive understanding of the challenges and opportunities that shape the perceptions of key stakeholders. moving to table 3, the assessment of feasibility and readiness for iot integration demonstrates a statistically significant impact of technical expertise perception and preparedness for iot integration on the education faculty. this resonates with (atzori et al., 2010) vision of creating smart educational environments through iot. the moderately positive perception among participants, as indicated by the regression model’s significance, underscores the potential of iot in education. however, the substantial variability in perceptions, emphasized by the mean squares for predictors, underscores the need for deeper exploration to unveil the specific factors shaping participants’ views. this aligns seamlessly with the overarching research objective, emphasizing the importance of a nuanced understanding of the factors influencing the readiness for iot integration. table 4 delves into the pedagogical implications of iot integration, revealing a substantial impact of institutional support rating on perceived enhancements in teacherstudent interaction, engagement, and learning outcomes. these findings align with (russell et al., 2014) emphasis on the transformative potential of iot in redefining teacherstudent dynamics. the participants’ notable belief in the positive impact of iot on teacher-student interactions and academic outcomes suggests a readiness to embrace iot technologies for improved educational experiences in afghanistan. table 5 addresses the identification of technological and infrastructural challenges, emphasizing participants’ awareness of key issues. the significant impact of resource constraints on perceived challenges aligns with the literature’s recognition of hurdles associated with implementing iot in education, as noted by (mukhopadhyay and suryadevara, 2014). in table 6, the proposal of guidelines for successful iot implementation reveals robust support, echoing the literature’s emphasis on the importance of clear guidelines (abed et al., 2019). this alignment with existing literature underscores the significance of recognizing challenges, institutional support, and clear guidelines for successful iot integration in conclusion, the results provide a comprehensive understanding of participants’ nuanced perceptions, laying the groundwork for informed strategies and interventions to harness the transformative potential of iot in afghanistan’s unique educational context. the alignment of these findings with existing literature underscores the significance of recognizing challenges, institutional support, and clear guidelines for successful iot integration, contributing valuable insights to the broader discourse on iot in diverse educational table 6: proposing of guidelines for successful implementation n minimum maximum mean std. deviation importance of afghanistan iot guidelines 120 4.00 5.00 4.3417 .47626 guidelines' impact on afghanistan iot 120 4.00 5.00 4.3417 .47626 confidence in afghanistan iot guidelines 120 1.00 4.00 2.9750 1.42877 valid n (listwise) 120 pa ge 15 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 8-16, 2024 landscapes. this multifaceted exploration sets the stage for future research endeavors aimed at optimizing strategies and fostering a technologically enriched educational environment in afghanistan. conclusions in conclusion, this study provides valuable insights into the perceptions and challenges surrounding the integration of the internet of things (iot) in afghanistan educational settings. the findings illuminate a complex landscape where stakeholders hold nuanced views on existing technological infrastructure, institutional capabilities, and the potential impact of iot on teaching and learning. the descriptive statistics revealed a moderate degree of consensus among participants, indicating a shared perspective on the current state of technological infrastructure. however, the somewhat neutral to slightly negative perceptions suggest areas that warrant attention and improvement. these findings lay the foundation for targeted interventions to enhance the technological landscape within afghanistan classrooms. the statistical analyses, including anova and regression, unveiled critical factors influencing participants’ views on iot integration. the significance of technical expertise perception, preparedness, and institutional support underscores the interconnected nature of these elements in shaping a favorable environment for iot adoption. policymakers and educational leaders can leverage these insights to develop strategic initiatives that address specific concerns and capitalize on existing strengths. the study’s exploration of challenges related to iot implementation provides a roadmap for overcoming hurdles. connectivity issues, compatibility concerns, and resource constraints emerged as key considerations. acknowledging these challenges is pivotal for formulating adaptive strategies that account for the unique contextual factors within the afghanistan educational landscape. the positive outlook on proposed guidelines signifies a readiness among participants to embrace structured recommendations for successful iot implementation. this optimistic reception bodes well for the development and implementation of guidelines that can serve as a roadmap for educational institutions in afghanistan. in essence, this study contributes to the broader discourse on technology integration in education, offering practical insights and recommendations for fostering a conducive environment for iot in afghanistan classrooms. as educational institutions and policymakers move forward, these findings can inform evidence-based decisions to propel the integration of innovative technologies, such as iot, for enhanced teaching and learning experiences in afghanistan. recommendation investment in technological infrastructure: recognizing the perceived limitations in existing technological infrastructure, stakeholders, including the government and educational institutions, should prioritize substantial investments in upgrading and modernizing technological resources. this may involve infrastructure development, ensuring robust connectivity, and providing access to upto-date hardware and software. professional development programs: implementing comprehensive professional development programs for educators is essential. these programs should focus on enhancing technical expertise and fostering a deeper understanding of iot applications in educational settings. by investing in teacher training, educational institutions can better prepare their staff to effectively integrate iot technologies into the learning environment. strategic institutional support: institutional leaders should proactively advocate for and provide strategic support for iot integration. this involves developing and implementing policies that facilitate the seamless adoption of iot technologies. allocating financial resources, creating dedicated support teams, and fostering a culture of innovation are crucial steps for institutions aiming to embrace iot in education. collaborative initiatives: encouraging collaboration between educational institutions, government bodies, and industry partners is essential for a holistic and sustainable approach to iot integration. public-private partnerships can facilitate the sharing of resources, expertise, and best practices, fostering an ecosystem that supports ongoing technological advancements in education. addressing connectivity challenges: given the identified challenges related to connectivity, especially in certain regions, policymakers should prioritize initiatives that address these issues. this may involve expanding internet infrastructure, providing subsidies for internet access, or exploring alternative technologies to ensure widespread connectivity. continuous evaluation and improvement: establishing mechanisms for continuous evaluation of iot integration initiatives is crucial. regular assessments of the impact on teaching and learning outcomes, as well as ongoing feedback from stakeholders, can inform adaptive strategies. institutions should be prepared to iterate on their approaches based on the evolving landscape of educational technology. development of customized guidelines: building on the positive reception of proposed guidelines, educational institutions should consider tailoring these recommendations to suit their specific contexts. customized guidelines can account for the unique challenges and opportunities present in different regions and institutions, ensuring practical and effective implementation. long-term research and monitoring: undertaking long-term research initiatives to monitor the sustained impact of iot integration is vital. by conducting followup studies and maintaining an ongoing dialogue with stakeholders, researchers can provide valuable insights into the evolving landscape of educational technology in afghanistan. pa ge 16 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 8-16, 2024 acknowledgment i extend my heartfelt gratitude to all those who contributed to the realization of this research endeavors. my sincere appreciation goes to the participants whose insights and cooperation were invaluable in shaping the study’s outcomes. special thanks to my academic colleagues for their support and guidance throughout the research process. i am indebted to the reviewers for their constructive feedback, which significantly enhanced the quality of this work. additionally, i would like to express my gratitude to the authors whose seminal works formed the foundation of this research. this project would not have been possible without the unwavering support of my university and the resources provided. thank you all for your indispensable contributions to the success of this research. references abdel-basset, m., manogaran, g., mohamed, m., & rushdy, e. (2018). internet of things in smart education environment: supportive framework in the decision-making process. concurrency and computation: practice and experience. https://doi. org/10.1002/cpe.4515 abed, s., alyahya, n., & altameem, a. (2019, november). iot in education: its impacts and its future in saudi universities and educational environments. in first international conference on sustainable technologies for computational intelligence: proceedings of ictsci 2019 (pp. 47-62). singapore: springer singapore. al-emran, m., malik, s. i., & al-kabi, m. n. (2020). a survey of internet of things (iot) in education: opportunities and challenges. toward social internet of things (siot): enabling technologies, architectures and applications: emerging technologies for connected and smart social objects, 197-209. atzori, l., iera, a., morabito, g., & nitti, m. (2010). the social internet of things (siot) when social networks meet the internet of things: concept, architecture and network characterization. in 2010 international conference on collaborative computing: networking, applications and worksharing (collaboratecom) (pp. 1-8). ieee. https://doi. org/10.1016/j.comnet.2012.07.010 bao, w., & ji, h. (2017). the internet of things (iot) in education: a survey. journal of ambient intelligence and humanized computing, 8(1), 93-110. bagheri, m., & movahed, s. h. (2016, november). the effect of the internet of things (iot) on education business model. in 2016 12th international conference on signal-image technology & internet-based systems (sitis) (pp. 435-441). ieee. https://doi.org/10.1109/ sitis.2016.74 dake, d. k., bada, g. k., & dadzie, a. e. (2023). internet of things (iot) applications in education: benefits and implementation challenges in ghanaian tertiary institutions. journal of information technology education: research, 22, 311-338. https://doi.org/10.28945/5183 fazil, a. w., hakimi, m., sajid, s., quchi, m. m., & khaliqyar, k. q. (2023). enhancing internet safety and cybersecurity awareness among secondary and high school students in afghanistan: a case study of badakhshan province. american journal of education and technology, 2(4), 50–61. https://doi.org/10.54536/ ajet.v2i4.2248 hakimi, m., fazil, a. w., khaliqyar, k. q., quchi, m. m., & sajid, s. (2024). evaluating the impact of e-learning on girl’s education in afghanistan: a case study of samangan university. international journal of multidisciplinary approach research and science, 2(01), 107-120. hakimi, m., fazil, a. w., hakimi, f. m., najieb, k., & hakimi, s. (2023). exploring the influences of cutting-edge technologies on operational efficiency, productivity, and financial profitability in afghanistan’s tourism sector. jurnal riset multidisiplin dan inovasi teknologi, 2(01), 168–83. https://doi. org/10.59653/jimat.v2i01.417 j, f. b., r, r., m, s., & n r, g. m. (2020). iot-based cloud integrated smart classroom for a smart and sustainable campus. procedia computer science, 172, 7781. https://doi.org/10.1016/j.procs.2020.05.012 kassab, m., defranco, j., & laplante, p. (2020). a systematic literature review on internet of things in education: benefits and challenges. journal of computer assisted learning, 36(2), 115-127. https://doi. org/10.1111/jcal.12383 malik, z., & al-emran, m. (2018). internet of things (iot): challenges, opportunities, and applications. in handbook of research on the internet of things (iot) in smart cities (pp. 1-19). igi global. mukhopadhyay, s. c., & suryadevara, n. k. (2014). internet of things: challenges and opportunities. in s. c. mukhopadhyay (ed.), internet of things: challenges and opportunities (pp. 1-17). springer, springer nature. https://doi.org/10.1007/978-3-319-042237_1 russell, m., bebell, d., higgins, j., & o’dwyer, l. (2014). one-to-one technology-enhanced learning: an opportunity for global research collaboration. british journal of educational technology, 45(1), 3-11. singh, s. k., singh, s. k., singh, a. k., & kumar, r. (2019). iot-based smart classroom for smart education. in 2019 3rd international conference on computing methodologies and communication (iccmc). https://doi.org/10.1109/iccmc.2019.8722723 suduc, a. m., bîzoi, m., & gorghiu, g. (2018). a survey on iot in education. romanian journal for multidimensional education/revista romaneasca pentru educatie multidimensionala, 10(3). pa ge 1 pa ge 1 american journal of smart technology and solutions (ajsts) new obstacles to smart city cybersecurity abdullah alsaeed1* volume 1 issue 1, year 2022 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: october 06, 2022 accepted: october 29, 2022 published: november 03, 2022 this article provides a concise description of the criteria that may be evaluated to determine the adoption of smart grid approaches that improve cybersecurity. it is necessary, from a functional point of view, to establish the degree to which cyber resilience may be increased by implementing solutions that are efficient in terms of cost. the problem of cybersecurity for smart grids has been the focus of several research and initiatives. in this study, the detection and diagnosis of false data injection (fdi) attacks are investigated in detail concerning their accuracy, processing time, and resilience to outside influences. no one method can be applied to all power systems. therefore, a comparison and statistical analysis of the newly reported approaches for detecting and recognizing cyberattacks are conducted here. keywords cyber-attack, cybersecurity, false data injection (fdi) , resilience, smart city, smart grid 1 department of computer science, university of manchester, saudi arabia * corresponding author’s e-mail: alsaeed.866@gmail.com introduction one of the many advantages of the current power city’s technological design, which is often referred to as the “smart city” in certain circles, is that it allows for more efficient integration of renewable energy sources (ress). however, due to the vast quantity of data that has to be sent for the system to function correctly, the smart city relies on an improved communication infrastructure to function properly (aoufi et al., 2020; mohammadi et al., 2019b; mohammadi & neagoe, 2020). as a result, cyber-attacks on smart city have increased. invasions of a company’s communication networks may drive up operating expenses dramatically (nikmehr & moghadam, 2019) or could impede the efficient functioning of the system (q. wang et al 2019). for example, cyber-attacks on ukraine’s power infrastructure in 2015 caused several hours of extensive power outages (aoufi et al., 2020). smart city operators must immediately identify, detect, and respond to such assaults to guarantee the system’s integrity and correct functioning. this procedure is referred regarded as having cyber resilience as its defining characteristic. correctly identifying cyberattacks is reportedly the first step in bolstering resilience throughout the attack and post-attack phases. this is the opinion of those who operate power systems. this is because the strike cannot be predicted with any degree of accuracy. as a consequence of this, several research activities have been carried out over the last decade to effectively detect and identify intrusions as a component of the cyber-resistance of the smart city (biggio & roli, 2018; otuoze et al., 2018; sharafeev et al., 2018; q. wang, w. tai, y. tang, & m. ni, 2019; wang & lu, 2013). literature review numerous research projects have been conducted to verify the precision of cyberattack identification and detection, help accelerate processing, and boost resistance from external influence. malicious meters might be detected more precisely using an ai-based approach, as described by (khanna et al., 2018). using machine learning, a strategy was published by (d. wang et al., 2019) that might better recognize different types of power city disruptions and cyber-attacks. reinforcement learning (rl) was proposed to cope with diverse partially observable markov decision process pomdps (kurt et al., 2018). training the defender with low-magnitude assaults and decreasing an attacker’s attacking space increased the strategy’s robustness (kurt et al., 2018). a multivariate gaussian-based model was described for power distribution system cyber-attack detection (an & liu, 2019). the isolation forest technique was introduced by (ahmed et al., 2019) to detect hidden data integrity breaches in the smart city, which is an unsupervised approach. to further speed up the detection of the assault, the suggested solution used minimal complexity. a randomized trees-based machine learning approach was presented to identify stealthy cyberattacks in the smart city successfully (acosta et al., 2020). in addition, the recommended approach was faster to compute and more resilient to noisy input than previous machine learning-based attack detection techniques. the distinction between data manipulation alterations and physical city adjustments was made by (mohammadpourfard et al., 2020), enabling the attack detection system to work successfully even when ideas moved. the unsupervised false data injection (fdi) attack technique (mohammadpourfard et al., 2017) effectively-recognized attacks under various scenarios. it showed resistance to the integration and reconfiguration https://journals.e-palli.com/home/index.php/ajsts mailto:alsaeed.866%40gmail.com?subject= pa ge 2 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 1-8, 2022 of renewable energy sources (ress) into power networks. it was discovered by (moslemi et al., 2017) that one of the most effective ways to identify assaults on the smart city is to use the maximum likelihood (ml) estimate. in addition, the suggested strategy reduced the computing load by minimizing the complexity of the ml estimation issue. (li et al., 2018) looked into an exact and quick approach in computing for detecting fdi assaults on smart city. noise-free data was not a problem for the proposed method. as reported by (hao et al., 2016), the markov decision process (mdp) technique was used to identify and evaluate the susceptibility of power city to cyberattacks in a dynamic setting. (zhao et al., 2018) they have shown their method’s robustness for detecting fdi assaults in noisy environments. the resilience of the smart city and its ability to withstand cyberattacks take up a substantial portion of the attention of this essay. the most common types of cyberattacks, known as fdi assaults, are discussed in this section. the most current research publications to be published are compared to one another and are expounded. the components of the paper are detailed below. second, the principles of cyber-attack detection and identification in the smart city are discussed in section 2, while section 3 focuses on current quantitative methodologies for cyber-attack detection and identification. this is followed by section 4, which concludes the paper. finally, in the fourth part, we will evaluate several strategies based on their resilience. the smart city cyber-attack detection and identification fundamentals as one of the new cyber-physical systems, the smart city is built on the physical power infrastructure, which includes power generation, distribution, and consumption systems, and the dense integration of communication infrastructure with specialized hierarchical control structures (mohammadi et al., 2019c; ostadijafari et al., 2019). the physical power infrastructure, which consists of power generation, consumption systems, distribution, and the dense integration of communication infrastructure with specialized hierarchical control structures, forms the foundation of the smart city, as one of the new cyber-physical systems. communication systems usually comprise actuation devices and smart sensors at the local control level. however, these systems also incorporate smart controllers, smart meters, automation units, phase measurement units (pmus), and distributed generations at the higher control levels, such as cyber layers (mohammadi et al., 2019a). they are susceptible to cyberattacks due to the smart city’s extensive penetration of communication infrastructure, which has led to the proliferation of connected devices. most threats to the systems that distribute power originate from unfriendly outsiders, malicious insiders, non-malicious insiders, and mother nature herself. most of these dangers originate from smart homes and businesses outfitted with smart meters. malicious agents are a risk because they can infect everyone with malware and viruses or zero in on specific computer systems to break into them, disrupt their operation, or cause damage to them. the most prevalent types of cyber-attacks in the smart city are denial of service (dos) and fdi (nguyen et al., 2020). most denial-of-service attacks are geared toward interrupting the data transfer process by focusing their attention on the communication infrastructure. it’s possible that scam data streams could be continuously flooded into the network or that synchronized data flooding would be used to target control signals (nguyen et al., 2020). in contrast to dos assaults, fdi assaults often take the form of data packet manipulation, which may occur at varying degrees of severity (nguyen et al., 2020). in addition, fdi attacks may be designed to target several data packets included inside communication protocols (nguyen et al., 2020). these data packets can consist of sensor/actuator software calibrations and protective relays, feedback signals and commands. as a result, the smart city might experience poor performance, instability, and even blackouts due to attacks by fdi (liu et al., 2019). it is necessary to quickly and accurately detect and identify any hostile cyber-attacks to improve the cybersecurity of smart city. as well as reducing computation cost and complexity, defensive measures must be implemented to strengthen or restore the system’s resistance against cyberattacks. the inability to differentiate between regular system interruptions and dynamics, such as changes in command signals and cyberattacks, connection/ disconnection of power generating units, load switching, are barriers to detecting cyberattacks. regular system interruptions and dynamics include these things. enhanced control mechanisms and careful evaluation of the system model’s nonlinear character are necessary to deal with the nonlinearities, uncertainties, and disruptions inherent in the system. one way of determining deviations and abnormalities is to estimate the system states under normal operating conditions and then compare those estimates to the actual system states. the steady-state states of the system were calculated using a weighted least square (wls) estimator, and the results are shown in (xu et al., 2017). in (sreenath et al., 2017), an attempt was made to solve the problem of wls’s inability to converge on a solution. this led to the development of recursive wls. dynamic estimating techniques are required to do a quick study on power systems. these methods must take into consideration the system’s initial states. the kalman filtering (kf) method is widely used non-static estimation approach that incorporates a corrective term to reduce the number of errors caused by state estimation (manandhar et al., 2014). extended kalman filtering (ekf), which considers the system’s nonlinearities, was examined by (abbaspour et al., 2019; chakhchoukh et al., 2019) to detect fdi assaults. in all of these modeldependent detection strategies, inaccuracies in the models https://journals.e-palli.com/home/index.php/ajsts pa ge 3 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 1-8, 2022 assaults, the cumulative error rate for state variables was just 1%. in order to get more precise results for calculating the tampering meters, an extra load estimator has been included in the current model. neural networks (nn) and artificial neural networks (ann) with a single hidden layer and forward connections were proposed to achieve an appropriate learning rate. because these networks were trained using historical data, the suggested estimator could identify an fdi assault, even if it came from a small number of compromised meters. this was made possible because these networks were used to train the estimator. in the event of a widespread fdi attack, our technique ensures the correct identification of both the attack and the tempered meters. (d. wang et al., 2019) outlines a supervised learning method that may detect cyberattacks on the smart city using previous data and log information. with an accuracy of 93.9 percent and a detection rate of 93.6 percent, the presented strategy is superior to other previously proposed techniques, such as the random forest (rf) method, and the k-nearest neighbors (knn) algorithm. both of these techniques have a detection rate of 93.6 percent. in (kurt et al., 2018), the rl approach was used for the first time to detect internet assaults on the smart city. this form of cyberattack was categorized as a pomdp since the attacker could compromise the legitimate states of the smart city, which the system operator would not have been able to tell was compromised in the first place. the solution shown by (kurt et al., 2018) does not involve using a model, as was mentioned, and it also took much less time to accomplish. a person who acts alone. the rl technique, which was presented from the perspective of a system defender, proved effective in detecting low-magnitude assaults. as a consequence, the defense may observe tiny variations in the states of the smart city independent of the method the attacker is using. the model-free strategy suggested showed evidence of resistance to the unknown system states. fdi assaults in the cyber-physical system of a power distribution city were detected using a multivariate gaussian-based technique (an & liu, 2019). this method was applied. it is possible to differentiate between transient and persistent assaults by examining the measurement data produced by micro-pmus. power distribution systems are divided into zones with comparable voltage profiles using the k-means clustering technique. in order to account for this shift, fewer micro-power management units (pmus) were used. the accuracy and precision that were shown were sufficient for the identification of transient assaults. however, the technique suggested to be used in continuous assaults was based on the imprecision of prediction, which the circumstances of a smart city may influence. enhancing regression models may lower the proportion of erroneous predictions and boost attack detection accuracy. according to the information in reference 14, the smart city is now targeted by a covert data integrity attack. a technique for feature extraction based on progressively degrade their effectiveness and potentially result in false positives. the actual execution of the solution becomes more challenging when recursive techniques and exact models are used since they increase the bar for the amount of computing power required to estimate the system’s state accurately. data-driven solutions have been created so that the difficulties connected with model-dependent detection approaches may be addressed. in general, the data-driven methodology may be classified into one of three groups: supervised learning techniques, unsupervised learning techniques, and semi-supervised learning techniques. each input is mapped to its one-ofa-kind output in the algorithms that use labels derived from the labeled dataset. it is usual practice to use supervised learning techniques where various ranges of cyber-attack simulations may be produced to obtain the necessary training dataset (ayad et al., 2018; fenza et al., 2019). as opposed to that, it is feasible to identify a meaningful pattern in unlabeled data by using algorithms that do not need supervision. however, using such approaches to identify cyberattacks is far less common than supervised algorithms. in addition, you should only utilize them when cyberattacks are not found in any of the obtained datasets (zanetti et al., 2017). therefore, acquiring the same training datasets under various operating settings is vital, are valid for supervised and unsupervised learning strategies. since the efficiency of detection algorithms is wholly dependent on the datasets they collect, there is a high risk that these algorithms may become overfit. consequently, the system has difficulty recognizing instances of cyberattacks that were not included in the training data set. a data-driven and model-based detection technique was studied in the paper (sargolzaei et al., 2019) as a potential solution to the problem that had been found before. in addition, strategies based on semi-supervised learning may be used when a trial-and-error method must be utilized to rectify or change the following control action following the input from the control actions that came before (chen et al., 2018). the main challenges in the way of the development of cybersecurity for the smart city are the detection accuracy, the processing complexity, and the resistance to external influences. these challenges apply to both data-driven and model-based detection systems. methods for the detection and identification of cyber attacks data-driven procedures to effectively identify cyberattacks on the smart city, datadriven technologies, such as machine learning techniques, have seen widespread application in recent years (apruzzese et al., 2019).in (khanna et al., 2018), it was suggested to use a supervised ai-based load estimator to compare anticipated loads with actual meter data to identify which meters are vulnerable to fdi assaults. as a direct result of the model’s capability to identify fdi https://journals.e-palli.com/home/index.php/ajsts pa ge 4 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 1-8, 2022 principal component analysis (pca) was used to make the issue more manageable, and high-dimensional data was transformed into low-dimensional space. the “isolation forest” method, which may detect irregularities in state estimation measurement characteristics, is based on unsupervised machine learning. it exhibited a better accuracy rate when comparing the suggested method to more conventional machine learning-based tactics. a shorter processing time was needed to detect cyberattacks due to the technique’s lower computational complexity. in (acosta et al., 2020), the smart city’s state estimation-measurement capabilities were used to detect stealthy cyberattacks by applying supervised learning. large-scale power systems have a high dimensional space. hence it was chosen to use a kernel principal component analysis (kpca) approach to reduce the complexity of the problem occurring in the system and accurately represent the information in a lower-dimensional space. furthermore, the properties of kpca were used to demonstrate that the proposed method is reliable in the presence of imbalanced datasets. it was also able to detect cyberattacks on the smart city with a high degree of accuracy, despite taking less computer time than other machine learning-based methods, such as classic pca. (mohammadpourfard et al., 2020) conducted research on the cybersecurity of smart city to evaluate the effect of concept drift, also known as the detection of physical changes that are the consequence of variations in smart city data manipulation. the efficiency of this technique in spotting malicious cyber activity was subjected to extensive testing and analysis. another study presented (mohammadpourfard et al., 2017) a technique for detecting cyberattacks that consider the system’s reconfiguration and incorporation of res. the f-test was applied to distinguish between the regular and attacked state vectors. it was found that more investigation is necessary for the suspect samples that defy the f-premise testing. as a result, the comparison index utilized to evaluate failed samples was the difference between suspected vectors and the average from comparable system state vectors. this was accomplished with the assistance of three outlier detection algorithms, including, interquartile range (iqr), median absolute deviation (mad), and fuzzy c-means (fcm) clustering, methods. according to the results, the proposed method could detect fdi assaults with a high degree of accuracy while unaffected by changes in the parameters. estimation methods for the state correct state estimates may assist in keeping the smart city safe and fully controlled (yong et al., 2016). on the other hand, state estimators are open to assault by fdi. such attacks may defeat bad data detection (bdd) techniques and modify the state estimate.(moslemi et al., 2017) presented an example of a decentralized method for discovering smart city attacks based on ml estimates. ml estimate was used to identify attacks since it could be transformed into a chordal embedding space. with the help of the kron reduction of the markov network of phase angles, the approach that was provided was able to segment the ml estimation problem into a number of distinct local ml estimation problems. the suggested method is decentralized, which provides utilities with more anonymity. by minimizing the size of the problem, the quantity of labor required to solve it is also reduced. furthermore, due to the properties of the attack matrix, which is sparsely, and the measurement matrix, which has a low rank, the fdi attack detection problem may be reframed as a matrix separation problem (li et al., 2018). this is possible because of the similarities between the two matrices. the modern methods for separating matrices, such as the the double-noise-dual-problem (dndp)-alm, augmented lagrangian method (alm), the low-rank matrix factorization (lrmf), and, suffer from the increase in the amount of time needed for computing and a reduction in the amount of accuracy achieved. to successfully solve this issue, a strategy named godecomposition was studied. the suggested approach displayed adequate accuracy in separating fdi attacks compared to the lrmf method and approximately similar accuracy compared to the alm and dndp-alm methods when the environment was free of background noise. furthermore, the solution offered to the issue had a fair calculation time and could protect the smart city against assaults on a broad scale. an mdp that was designed to simulate the attack strategy of the attackers was published by (hao et al., 2016). research on the knowledge and scenarios linked to the smart city was carried out in two phases. first, in an mdp with a short time horizon, it is conceivable for adversaries to determine the current state of the intelligent city over a relatively short amount of time. after investigating the probabilities of an assault, the most effective method from the viewpoint of the aggressor was found. according to the findings of the vulnerability analysis, the suggested method is resilient against the parametric uncertainties present in an mdp situation and the operators’ dispatch strategy. an operator-perspective technique for assessing the susceptibility of nonlinear state estimators to fdi assaults is presented in reference (zhao et al., 2018). a reliable approach for recognizing fdi assaults was developed, including using a subset of safe pmu measurements to investigate the measurement’s statistical consistency. these security approaches, unaffected by abnormalities and render the system completely transparent, may be used to determine whether or not an fdi attack has occurred. a dependable huber m-estimator was also used to accomplish accurate fdi assault detection. the suggested approach was unaffected by secure measurements and insufficient and noisy data. it was stated by (deng et al., 2018) that fdi attacks might be carried out against power distribution networks since the statuses of these networks could be expected based on the data on power flow. https://journals.e-palli.com/home/index.php/ajsts pa ge 5 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 1-8, 2022 according to the simulation’s findings, the fdi attack can be carried out without being discovered by the bdd approaches if the attacker correctly anticipates one state of the system. fdi attacks on the electrical city have been investigated (margossian et al., 2019), operating on the assumption that the attackers had some understanding of the system. after that, the demonstration fdi attack on the partial city was carried out to show how undetected fdi attacks may be. a strategy based on state estimates was later developed to protect power city against assaults carried out in the name of foreign direct investment (fdi) that go unreported. the basic measurement set is a mechanism that was developed by (sreeram & krishna, 2019) to protect the smart city against fdi assaults by securing n-1 meter in n-bus power systems. this solution was given the moniker “the basic measurement set” (bms). the approach was then altered to determine a subset of the optimum bms to reduce the level of vulnerability shown by the system if less than n 1 meter could be safeguarded. various other approaches as was said before, the primary objective of cyberattacks is to influence the condition that is expected to be present in the smart city. therefore, in addition to the data-driven methodology and the state estimation technique, other methodologies, such as game theory, have been utilized to analyze the vulnerability of the smart city to the possibility of cyberattacks (apruzzese et al., 2019; biggio & roli, 2018). in (q. wang, w. tai, y. tang, m. ni, et al., 2019), the features of fdi assaults, as seen from the attacker’s viewpoint, were described to reveal the vulnerabilities present in current bdd approaches. after that, a twolayer defensive paradigm that included detection and protection strategies was presented from the defender’s point of view. a zero-sum, static game theory was used to identify the most effective defensive and attacking tactics. it was found that the minimax-regret technique could be used to design a cost-effective defense that could be used against an assault that used load redistribution (abusorrah et al., 2017). there was an effort made to spread the load, and the algorithm’s goal was to reduce the amount of economic damage caused. because the protective strategy of the smart city is susceptible to time-varying loading situations, it is necessary to develop an algorithm that can account for these fluctuations. a game-theoretic model was developed and tested to meet this need under various loading scenarios. subsequently, a multi-level insolvable problem was transformed into a bi-level solvable optimization issue. a greedy implicit enumeration method was also utilized to identify the optimal global solution. in (pilz et al., 2020), an investigation was conducted into how the influence of fdi assaults on compromising anticipated demand data. a model based on game theory was devised to assist utilities in awarding against these kinds of assaults, and the nash equilibrium was uncovered. the best kind of monitoring for low-impact cyberattacks is none; nevertheless, for all other types of assaults, a range of defense techniques should be devised. researchers from (hasan et al., 2020) looked at a cyberattack game in which the attacker and the defense were fighting against one another to see who would emerge victoriously. the attacker chose and targeted critical power substations to do the most damage possible to the system while staying within the allotted spending limit. the vast majority of important power substations were taken simultaneously to reduce the amount of damage done to the system from the standpoint of the defense. we used polynomial-time algorithms to determine the worst possible dynamic assault that could be launched and the most effective defensive plan. the strategy given was more effective, less challenging, and capable of attacking a wider range of target systems than the best practices already in place. it was also more efficient and less complex than those practices. in (gao & shi, 2020), a method based on dynamic game theory was presented to determine the level of vulnerability posed by cyber-physical systems. furthermore, a mathematical programming model consisting of three groups a defender, an attacker, and a defender was investigated in the context of a system recovery delay and a distributed denial-of-service attack. to overcome the challenge of optimization, a cuttingedge technique known as particle swarm optimization (pso) was used. the developed strategy proved to be quite successful when it came to finding susceptible transmission lines in power networks. detection and identification of cyberattacks; a comparison the many methods of detecting and identifying cyberattacks discussed in this research are compared in table 1. for this comparison, we will use our resilience criteria for accuracy, computational load, and resistance to external variables. table 1: methods of detecting and identifying cyber-attacks author objective method proposed criteria for adaptability and resurgence accuracy complexity external robustness (gao & shi, 2020) detection and analysis of cyber-attacks and vulnerabilities dynamic game theory ✓ https://journals.e-palli.com/home/index.php/ajsts pa ge 6 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 1-8, 2022 (hasan et al., 2020) detection of cyber attacks game theory ✓ (pilz et al., 2020) detection and identification of cyber attacks. game theory ✓ ✓ ✓ (abusorrah et al., 2017) detection of cyber-attacks game theory based on the minimaxregret method ✓ (q. wang, w. tai, y. tang, m. ni, et al., 2019) detection and identification of cyber-attacks. zero-sum static game theory ✓ ✓ (sreeram & krishna, 2019) detection and analysis of cyber-attacks and vulnerabilities state estimation ✓ ✓ ✓ (margossian et al., 2019) detection and analysis of cyber-attacks and vulnerabilities state estimation based on power flow analysis ✓ ✓ (deng et al., 2018) detection and analysis of cyber-attacks and vulnerabilities state estimation ✓ ✓ (zhao et al., 2018) vulnerability and intrusion detection huber m-estimator ✓ ✓ (hao et al., 2016) vulnerability and intrusion detection markov decision process-based method ✓ (li et al., 2018) detection of internetbased cyberattacks go-decomposition algorithm ✓ ✓ (moslemi et al., 2017) detection of cyber-attacks gaussian markov random field method ✓ ✓ ✓ (mohammadpourfard et al., 2017) detection and identification of cyber-attacks. unsupervised learning algorithm ✓ ✓ (mohammadpourfard et al., 2020) detection of cyber-attacks isolation forest method ✓ ✓ (acosta et al., 2020) detection of cyber attacks kpca-based method ✓ (acosta et al., 2020) detection of cyber attacks isolation forest pcabased method ✓ (an & liu, 2019) detection of cyber attacks multivariate gaussian-based method ✓ ✓ ✓ (kurt et al., 2018) detection of internetbased cyberattacks reinforcement learning-based algorithm ✓ ✓ (d. wang et al., 2019) detection and identification of cyber-attacks. supervised learning algorithm ✓ ✓ (khanna et al., 2018) identification of malicious meters ai-based algorithm ✓ recent research, as seen in this table, has emphasized precision as a primary priority. however, accuracy remains a significant barrier to adopting data-driven solutions in the energy industry. in most cases, the computational burden may be lowered by using more potent processors and computing methods, such as parallel or distributed computing, which incur costs. in addition, ensuring safe operations may incur an unwanted but unavoidable cost due to the computational burden. there is also a great deal of literature about resilience. in several recent studies, the degree to which the smart city’s security is enhanced remains unclear. according to table 1 of the operational aims, detecting and identifying online cyberattacks should be the most significant. due to the unexpected behavior of renewable energy sources, energy management systems are plagued by high uncertainty and stochasticity. smart meters and pmus with iot capabilities might https://journals.e-palli.com/home/index.php/ajsts pa ge 7 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 1-8, 2022 provide fraudsters with various attack surfaces. finally, controlling and regulating smart city is made more difficult by the need for quick detection and diagnosis of cyberattacks. the application of ai models in dealing with small datasets for training and testing, as well as the complex behavior of attacker and defender models, is demonstrated by the fact that game theory and rl models fit all three criteria. these cutting-edge technologies for cyberattack detection in power city may be helpful in the future. conclusion at the beginning of this article, we discussed improvements to the smart city’s overall level of cybersecurity. recent academic research has focused on investigating ways to defend the smart city from intrusions by digital hackers. the accuracy, computing complexity, and resistance to external influences of fdi attack detection and identification have been the focus of further research in this study. in addition, this study has looked at the resilience of fdi assaults. all of the criteria mentioned in this study can be quantified, enabling operators of the system to assess the degree to which the implementation of practical financial solutions may improve the system’s resilience. acknowledgement the author is thankful to the university of manchester, saudi arabia, for the continuous support of this research study. funding no funding sources are reported. conflict of interest the author does not have any conflict of interest. references abbaspour, a., sargolzaei, a., forouzannezhad, p., yen, k. k., & sarwat, a. i. (2019). resilient control design for load frequency control system under false data injection attacks. ieee transactions on industrial electronics, 67(9), 7951-7962. abusorrah, a., alabdulwahab, a., li, z., & shahidehpour, m. (2017). minimax-regret robust defensive strategy against false data injection attacks. ieee transactions on smart grid, 10(2), 2068-2079. acosta, m. r. c., ahmed, s., garcia, c. e., & koo, i. (2020). extremely randomized trees-based scheme for stealthy cyber-attack detection in smart grid networks. ieee access, 8, 19921-19933. ahmed, s., lee, y., hyun, s.-h., & koo, i. (2019). unsupervised machine learning-based detection of covert data integrity assault in smart grid networks utilizing isolation forest. ieee transactions on information forensics and security, 14(10), 2765-2777. an, y., & liu, d. (2019). multivariate gaussian-based false data detection against cyber-attacks. ieee access, 7, 119804-119812. aoufi, s., derhab, a., & guerroumi, m. (2020). survey of false data injection in smart power grid: attacks, countermeasures and challenges. journal of information security and applications, 54, 102518. apruzzese, g., colajanni, m., ferretti, l., & marchetti, m. (2019). addressing adversarial attacks against security systems based on machine learning. 2019 11th international conference on cyber conflict (cycon). ayad, a., farag, h. e., youssef, a., & el-saadany, e. f. (2018). detection of false data injection attacks in smart grids using recurrent neural networks. 2018 ieee power & energy society innovative smart grid technologies conference (isgt). biggio, b., & roli, f. (2018). wild patterns: ten years after the rise of adversarial machine learning. pattern recognition, 84, 317-331. chakhchoukh, y., lei, h., & johnson, b. k. (2019). diagnosis of outliers and cyber attacks in dynamic pmu-based power state estimation. ieee transactions on power systems, 35(2), 1188-1197. chen, y., huang, s., liu, f., wang, z., & sun, x. (2018). evaluation of reinforcement learning-based false data injection attack to automatic voltage control. ieee transactions on smart grid, 10(2), 2158-2169. deng, r., zhuang, p., & liang, h. (2018). false data injection attacks against state estimation in power distribution systems. ieee transactions on smart grid, 10(3), 2871-2881. fenza, g., gallo, m., & loia, v. (2019). drift-aware methodology for anomaly detection in smart grid. ieee access, 7, 9645-9657. gao, b., & shi, l. (2020). modeling an attack-mitigation dynamic game-theoretic scheme for security vulnerability analysis in a cyber-physical power system. ieee access, 8, 30322-30331. hao, y., wang, m., & chow, j. h. (2016). likelihood analysis of cyber data attacks to power systems with markov decision processes. ieee transactions on smart grid, 9(4), 3191-3202. hasan, s., dubey, a., karsai, g., & koutsoukos, x. (2020). a game-theoretic approach for power systems defense against dynamic cyber-attacks. international journal of electrical power & energy systems, 115, 105432. khanna, k., panigrahi, b. k., & joshi, a. (2018). ai-based approach<? show [aq=”” id=” q1]”?> to identify compromised meters in data integrity attacks on smart grid. iet generation, transmission & distribution, 12(5), 1052-1066. kurt, m. n., ogundijo, o., li, c., & wang, x. (2018). online cyber-attack detection in smart grid: a reinforcement learning approach. ieee transactions on smart grid, 10(5), 5174-5185. li, b., ding, t., huang, c., zhao, j., yang, y., & chen, y. (2018). detecting false data injection attacks against power system state estimation with fast godecomposition approach. ieee transactions on industrial informatics, 15(5), 2892-2904. https://journals.e-palli.com/home/index.php/ajsts pa ge 8 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 1(1) 1-8, 2022 liu, c., liang, h., chen, t., wu, j., & long, c. (2019). joint admittance perturbation and meter protection for mitigating stealthy fdi attacks against power system state estimation. ieee transactions on power systems, 35(2), 1468-1478. manandhar, k., cao, x., hu, f., & liu, y. (2014). detection of faults and attacks including false data injection attack in smart grid using kalman filter. ieee transactions on control of network systems, 1(4), 370-379. margossian, h., sayed, m. a., fawaz, w., & nakad, z. (2019). partial grid false data injection attacks against state estimation. international journal of electrical power & energy systems, 110, 623-629. mohammadi, f., nazri, g.-a., & saif, m. (2019a). a bidirectional power charging control strategy for plugin hybrid electric vehicles. sustainability, 11(16), 4317. mohammadi, f., nazri, g.-a., & saif, m. (2019b). a fast fault detection and identification approach in power distribution systems. 2019 international conference on power generation systems and renewable energy technologies (pgsret). mohammadi, f., nazri, g.-a., & saif, m. (2019c). a realtime cloud-based intelligent car parking system for smart cities. 2019 ieee 2nd international conference on information communication and signal processing (icicsp). mohammadi, f., & neagoe, m. (2020). emerging issues and challenges with the integration of solar power plants into power systems. in solar energy conversion in communities (pp. 157-173). springer. mohammadpourfard, m., sami, a., & weng, y. (2017). identification of false data injection attacks with considering the impact of wind generation and topology reconfigurations. ieee transactions on sustainable energy, 9(3), 1349-1364. mohammadpourfard, m., weng, y., pechenizkiy, m., tajdinian, m., & mohammadi-ivatloo, b. (2020). ensuring cybersecurity of smart grid against data integrity attacks under concept drift. international journal of electrical power & energy systems, 119, 105947. moslemi, r., mesbahi, a., & velni, j. m. (2017). a fast, decentralized covariance selection-based approach to detect cyber attacks in smart grids. ieee transactions on smart grid, 9(5), 4930-4941. nguyen, t., wang, s., alhazmi, m., nazemi, m., estebsari, a., & dehghanian, p. (2020). electric power grid resilience to cyber adversaries: state of the art. ieee access, 8, 87592-87608. nikmehr, n., & moghadam, s. m. (2019). game-theoretic cybersecurity analysis for false data injection attack on networked microgrids. iet cyper-phys. syst.: theory & appl., 4(4), 365-373. ostadijafari, m., jha, r. r., & dubey, a. (2019). conservation voltage reduction by coordinating legacy devices, smart inverters and battery. 2019 north american power symposium (naps). otuoze, a. o., mustafa, m. w., & larik, r. m. (2018). smart grids security challenges: classification by sources of threats. journal of electrical systems and information technology, 5(3), 468-483. pilz, m., naeini, f. b., grammont, k., smagghe, c., davis, m., nebel, j.-c., al-fagih, l., & pfluegel, e. (2020). security attacks on smart grid scheduling and their defences: a game-theoretic approach. international journal of information security, 19(4), 427-443. sargolzaei, a., yazdani, k., abbaspour, a., crane iii, c. d., & dixon, w. e. (2019). detection and mitigation of false data injection attacks in networked control systems. ieee transactions on industrial informatics, 16(6), 4281-4292. sharafeev, t., ju, o. v., & kulikov, a. (2018). cybersecurity problems in smart grid cyber attacks detecting methods and modelling attack scenarios on electric power systems. 2018 international conference on industrial engineering, applications and manufacturing (icieam). sreenath, j., meghwani, a., chakrabarti, s., rajawat, k., & srivastava, s. (2017). a recursive state estimation approach to mitigate false data injection attacks in power systems. 2017 ieee power & energy society general meeting. sreeram, t., & krishna, s. (2019). managing false data injection attacks during contingency of secured meters. ieee transactions on smart grid, 10(6), 6945-6953. wang, d., wang, x., zhang, y., & jin, l. (2019). detection of power grid disturbances and cyber-attacks based on machine learning. journal of information security and applications, 46, 42-52. wang, q., tai, w., tang, y., & ni, m. (2019). review of the false data injection attack against the cyberphysical power system. iet cyber-physical systems: theory & applications, 4(2), 101-107. wang, q., tai, w., tang, y., ni, m., & you, s. (2019). a two-layer game theoretical attack-defense model for a false data injection attack against power systems. international journal of electrical power & energy systems, 104, 169-177. wang, w., & lu, z. (2013). cyber security in the smart grid: survey and challenges. computer networks, 57(5), 1344-1371. xu, r., wang, r., guan, z., wu, l., wu, j., & du, x. (2017). achieving efficient detection against false data injection attacks in smart grid. ieee access, 5, 1378713798. yong, s. z., foo, m. q., & frazzoli, e. (2016). robust and resilient estimation for cyber-physical systems under adversarial attacks. 2016 american control conference (acc). zanetti, m., jamhour, e., pellenz, m., penna, m., zambenedetti, v., & chueiri, i. (2017). a tunable fraud detection system for advanced metering infrastructure using short-lived patterns. ieee transactions on smart grid, 10(1), 830-840. zhao, j., mili, l., & wang, m. (2018). a generalized false data injection attacks against power system nonlinear state estimator and countermeasures. ieee transactions on power systems, 33(5), 4868-4877. https://journals.e-palli.com/home/index.php/ajsts pa ge 1 pa ge 10 american journal of smart technology and solutions (ajsts) genetic algorithm of independent task meta-scheduling centralized in the cloud computing stéphane fouakeu tatieze1*, jean claude kamgang2, marcellin julius nkenlifack3 volume 2 issue 2, year 2023 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v2i2.1804 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: june 30, 2023 accepted: july 24, 2023 published: july 30, 2023 a group of networked, virtualized computers make up the distributed, parallel cloud computing technology. the power for these machines is dynamic, and they are displayed as one or more computing resources. these are compiled based on service level agreements (slas) that have been negotiated between the service provider and the customers. enterprise applications have migrated in large numbers to cloud computing during the past few years. one of the most important challenges of cloud computing is the scheduling of tasks; which should satisfy cloud users in terms of quality of service and increase the profit of cloud providers. bio-inspired algorithms (genetics) represent a heuristic research technique that produces effective solutions. in this article, we propose a genetic meta-scheduling algorithm that optimizes the execution time and makespan of tasks submitted by users. to achieve this, this algorithm is based on the requirements of user requests and the availability of resources (virtual machines) of cloud computing to obtain a better combination as an optimal solution. this effort makes the meta-scheduling genetic algorithm superior than others in the literature like the min-min algorithm and the regular genetic algorithm. customer satisfaction is higher, and more particularly, the execution time and makespan are better. keywords cloud computing, genetic algorithm, meta-scheduling 1 department of electrical engineering and industrial automation, ensai, university of ngaoundéré, cameroon 2 department of mathematics and computer, ensai, university of ngaoundéré, cameroon 3 department of mathematics and computer, faculty of sciences, university of dschang, cameroon * corresponding author’s e-mail: fouakeustephane@gmail.com introduction cloud computing (cc) is a new paradigm for utility virtualized resources, designed for end users in a dynamic computing environment to provide reliable and guaranteed services (dillon et al., 2010). cloud computing has service models and deployment models. as service models, we have software as a service (saas), platform as a service (paas), infrastructure as a service (iaas). software as a service (saas) provides users with applications in the form of online services already deployed in the cloud. this layer is managed by the saas provider in a way that is transparent to users. platform as a service (paas) is more oriented to serve application developers. it offers a fully configured and managed platform on which the user can develop, test and run their applications. infrastructure as a service (iaas) allows infrastructure resources such as computing capacity, storage, network as utilities. as deployment models, we have private cloud, community cloud, public cloud, hybrid cloud. in the private cloud, all of its resources are made available exclusively to a single company or organization. the private cloud can be managed by the company itself (internal private cloud) or by a third party (external private cloud). in the community cloud, the infrastructure is shared by several independent organizations with common interests. the infrastructure can be managed by the member organizations or by a third party. in the public cloud, the infrastructure is accessible to a wide public and is owned by a service provider. the latter charges users according to consumption and guarantees the availability of services through sla contracts and for the hybrid cloud, the infrastructure is a composition of several clouds (private, community or public) (dillon et al., 2010). cloud environment allows users to use applications without installation and access their personal files at any computer with internet access, end users access cloud based applications through a web browser or a light weight desktop (durga et al., 2016). cc, applies distributed computing techniques to deliver an on-demand access to a shared virtual computing resources (ex. networks, servers, storage, applications and services) over the internet (zhang et al., 2010; singh et al., 2017). virtualization is an emerging technology for efficient utilization of cloud resources. it is used to split a single physical machine into multiple virtual machines (vm) (malhotra et al., 2010). vm also can provide resource sharing, high utilization of pooled resources, rapid provisioning and workload isolation (ahmad et al., 2015b). usually, a cloud service provider (csp), like google, presents these facilities using the pay per use model (arya et al., 2014). by the help of cloud computing technology, users such as the individuals, researchers and large businesses can access their data, applications, on different platforms via the internet without the need for buying costly computing resources. the main goal of cloud computing is to satisfy cloud users with the agreed qos and improve profits of cloud providers (buyya et al., 2009). to provide ensured proficient performance to users, it is necessary that tasks should be mapped efficiently to available resources. task scheduling (ts) is one of the core challenges in cc environment. the task scheduling problem in cloud, which is known to be np-hard, is assigning different tasks to corresponding resource node under the quality of services (qos) constraints (aarts et al., 2005). ts can be classified into pa ge 11 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 independent scheduling and dependent scheduling. in independent scheduling tasks are independent of each other and can be scheduled in any sequence, however in dependent scheduling, tasks are represented by a directed acyclic graph (dag) (i.e., workflow scheduling). dag is a directed graph that comprises group of edges and vertices. where each vertex signifies the task and every edge signifies the affiliation between two nodes or vertices connected through that edge (singh et al., 2015). ts can also be classified into static and dynamic task scheduling. in static scheduling, all tasks or vms are known a priori to scheduling. these tasks are independent of the virtual machine’s states and their availability. so, it imposes less runtime overhead. on the other hand, in dynamic scheduling, the information about the tasks is unknown in advance. so, the execution time of task may not be known and the information about vms is not obtained until it comes into the scheduling stage (nagadevi et al., 2013). ts is an optimization problem belonging to the class of np-hard problems. some traditional task scheduling algorithms have been applied in heterogeneous computing environments such as min-min (he et al., 2003), max-min (mao et al., 2014), etc. so in cloud computing, there are various type of meta heuristic algorithms for scheduling problem in cloud computing such as ant colony optimization (aco), particle swarm optimization (pso) genetic algorithm (ga), etc, (xu et al., 2009) can be applied to achieve near optimal solution. genetic algorithm, based on natural selection and inheritance theory, has been widely and successfully applied in scheduling problems. in this paper, genetic algorithm is implemented using cloudsim simulator and compared to the traditional heuristic methods to solve the independent static ts problem in cc environment. in static environment, the specifications of the vms are fixed. when users submit jobs to the resources for execution, meta-scheduler acquire information about resources from cis (cloud information service) and then divide the job into various tasks or subtasks if needed. then map the to best resources distributed geographically same according to user’s requirements and availability of resources. cis are responsible for providing information about status of available resources which helps the meta-scheduler for scheduling, monitoring and further communication if required. after execution of all tasks, result is combined and sends back to user via metascheduler. the main features of competent ts in this paper are minimizing deadline, and budget. the remainder of the paper is organized as follows. section ii gives an overview of related work on ts in cloud computing. section iii presents the task scheduling problem. section iv presents the genetic algorithm. in section v we have the system model. experimental results and discussions are given in section vi. section vii concludes this paper. related work tasks scheduling is a hot and major research area in the distributed environment like cloud computing. it is a challenging issue in which a lot of research works have been carried out. many meta-heuristic techniques like genetic algorithm (ga) were proposed to solve the tasks scheduling problems using various strategies: (jang et al., 2012) proposed task scheduling model where the task scheduler calls the ga scheduling function to make task schedules based on information of tasks and virtual machines. the ga scheduling function creates a population, a set of task schedules, and evaluates the population by using the fitness function considering user satisfaction and virtual machine availability. the function iterates reproducing populations to output the best task schedule. experimental results show effectiveness and efficiency of the genetic algorithmbased task scheduling model in comparison with existing task scheduling models, which are the round-robin task scheduling model, the load index-based task scheduling model, and the abc based task scheduling model. (kaur et al., 2012) have developed a task scheduling algorithm for cloud computing environment. the author used shortest cloudlet to fastest processor (scfp) and longest cloudlet to fastest processor (lcfa) algorithm to initialize the population of ga. their algorithm takes variable power processors and variable length tasks to represent a real-time scenario but considers single user job. in this research they have proposed a modified genetic algorithm for single user jobs in which the fitness is developed to encourage the formation of solutions to achieve the time minimization and compared it with existing heuristics. experimental results show that, under the heavy loads, the proposed algorithm exhibits a good performance. in (dasgupta et al., 2013) proposed ga as a load balancing technique for cloud computing to find a global optimum processor for job in a cloud. they have presented an approach that handles the load on processors as well as reduces the makespan, but takes equal priority tasks. analysis of the results, indicates that the proposed strategy for load balancing not only outperforms a few existing techniques but also guarantees the qos requirement of customer job. in (kaur et al., 2014) genetic algorithm is enhanced using new fitness function based on mean and grand mean values. this optimization can be implemented on both ends, for job scheduling and resource scheduling. it reduces the execution time of all the tasks but considered limited number of tasks. (atul et al., 2015) propose a multi-objective task scheduling algorithm for mapping tasks to a vms in or-der to improve the throughput of the datacenter and reduce the cost without violating the sla (service level agreement) for an application in cloud saas environment. the pro posed algorithm provides an optimal scheduling method. most of the algorithms schedule tasks based on single criteria (i.e execution time). but in cloud environment it is required to consider various criteria like execution time, cost, bandwidth of user etc. this algorithm is simulated and the result shows better performance and improved throughput. (juntao et al., 2016) proposes a novel dynamic task scheduling algorithm based on improved genetic algorithm (igats). this paper introduces the concept of load priority. first, select average queue-run pa ge 12 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 length to as a high-priority load parameter, which reflects the average number of processes running in the queue within the specified time interval; second, select the cpu utilization and memory utilization as a priority load parameters, which reflects the currently running task size of system resources; experimental results demonstrate that the proposed algorithm can effectively improve the throughput of cloud computing systems, and can significantly reduce the execution time of task scheduling. (amjad et al., 2017) presented an efficient greedy algorithm and a genetic algorithm with adaptive selection of crossover and mutation from a pool of crossover and mutation types to allocate and schedule real-time tasks with precedence constraint on heterogamous virtual machines. the selection of crossover and mutation is based on the previous performance of the operators. the adaptive ga uses population diversity to determine the fitness of each type of crossover while the fitness of mutation is determined in terms of its ability to find a better quality solution (i.e., intensification). (rasha et al., 2018) proposed htscc ( hybrid task scheduling in cloud computing) algorithm to improves the local search by using the ga mutation operator and expected to work with the different size of tasks. the proposed htscc algorithm makes use of the advantages of the ga and pso algorithms in order to maximize resource utilization and minimize makespan. these features of the proposed algorithm reduce the makespan and increase the resources utilization. to optimize large-scale task scheduling problem in cloud environment with less makespan and computation time, (kairong et al., 2018) proposed an adaptive incremental genetic algorithm. their method based on genetic algorithm which has adaptive probability of mutation rate and crossover rate can provide feasible solutions for the allocation of large numbers of tasks with less computation time. the simulation results show that algorithm outperforms the greedy algorithm and non-adaptive genetic algorithm in terms of solution quality. (nagwan et al., 2019) have implemented ts using two metaheuristic algorithms (pso, ga) and compared their performance with two traditional techniques (fcfs, sjf). they generate the chromosomes randomly with a list of tasks and a list of vms to form the initial population in ga. after, evaluate the performance of each chromosome using fitness function. it is calculated using a set of metrics such as makespan, flow time, response time, resource utilization, throughput time and degree of imbalance. based on fitness value retrieved from each metric, chromosomes are selected and then are feed to a crossover and mutation operations. after, they update the population and decode the procuration chromosome (feasible solution) then, the best chromosome is the final solution for tasks allocation on vms. the algorithms have been implemented as part of the cloud broker in symmetric and asymmetric environment. ga algorithm only fulfilled the optimal degree of imbalance in symmetric environment with real workload traces. otherwise, it gave sufficient performance in obtaining the optimal response time in asymmetric environment in both of synthetic traces and real workload traces. task scheduling problem cloud consists of a number of resources that are different with one other via some means and cost of performing tasks in cloud using resources of cloud is different so scheduling of tasks in cloud is different from the traditional methods of scheduling and so scheduling of tasks in cloud need better attention to be paid because services of cloud depends on them. task scheduling plays a key role to improve flexibility and reliability of systems in cloud. the main reason behind scheduling tasks to the resources in accordance with the given time bound, which involves finding out a complete and best sequence in which various tasks can be executed to give the best and satisfactory result to the user. in cloud computing, resources in any form i.e. cups, firewall, network are always dynamically allocated according to the sequence and requirements of the task, subtasks. so, this leads task scheduling in cloud to be a dynamic problem means no earlier defined sequence may be useful during processing of task. the reason behind the scheduling to be dynamic is that because flow of task is uncertain, execution paths are also uncertain and at the same time resources avail able are also uncertain because there is a number of tasks are present that are sharing them simultaneously at the same time (singh et al., 2014). the scheduling of tasks in cloud means choose the best suitable resource available for execution of tasks or to allocate computer machines to tasks in such a manner that the completion time is minimized as possible. in scheduling algorithms list of tasks is created by giving priority to each and every tasks where setting of priority to different tasks can be based on various parameters. tasks are then chooses according to their priorities and assigned to available processors and computer machines which satisfy a predefined objective function (radulescu et al., 2000). meta-scheduling is defined (christodoulopoulos et al., 2009) as a software technique for optimizing workloads of grid clusters, by choosing and combining the resources of their different managers into a single aggregated view, so that jobs can be directed batch users to the best execution locations, in a manner transparent to them. for example, metascheduling is about assigning user jobs to nodes in a grid, such as clusters, which in turn have their own local schedulers. cloud computing uses virtualization technique for mapping the resources of cloud to the virtual machine layer, implement the user’s task, so the task scheduling of cloud computing environment achieve at the applications layer and the virtual layer of resources. scheduling is nothing but the mapping of tasks and resources in accordance with some certain principles for achieving the desired goal. cloud computing paradigm simplifies the mapping of tasks to resources; the required resources together form to be virtual machines (vms), the process of search the desired resource package is pa ge 13 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 same as the process of searching the various vms. genetic algorithm genetic algorithm (ga) is based on the biological concept of generating the population. ga is considered a rapidly growing area of artificial intelligence (jang et al., 2012). by darwin’s theory of evolution was inspired the genetic algorithms (gas). according to darwin’s theory, term “survival of the fittest” is used as the method of scheduling in which the tasks are assigned to resources according to the value of fitness function for each parameter of the task scheduling process (buyya et al., 2009) (see figure. 1). solution for the next generation based on the darwin’s law of survival. this operation is the guiding channel for the ga based on the performance. there are various selection strategies to select the best chromosomes such as roulette wheel, boltzmann strategy, tournament selection, selection based on rank and elitist selection. crossover crossover operation can be achieved by selecting two parent individuals and then creating a new individual tree by alternating and reforming the parts of those parents. hybridization operation is a guiding process in the ga and it boosts the searching mechanism. there are some crossover strategies such as single-point crossover, two-point crossover, uniform crossover and so on. an example of single-point crossover is shown in this figure: figure 1: genetic algorithm (raj et al., 2013) figure 2: single-point crossover [23] figure 3: mutation operator [23]. the main principles of the ga are described as follows (jang et al., 2012): initial population the initial population is the set of all individuals that are used in the ga to find out the optimal solution. every solution in the population is called as an individual. every individual is represented as a chromosome for making it suitable for the genetic operations. from the initial population, the individuals are selected, and some operations are applied on them to form the next generation. the mating chromosomes are selected based on some specific criteria. fitness function the productivity of any individual depends on the fit ness value. it is the measure of the superiority of an individual in the population. the fitness value shows the performance of an individual in the population. there fore, the individuals survive or die out according to the fitness or function value. hence, the fitness function is the motivating factor in the ga. selection the selection mechanism is used to select an intermediate mutation after crossover, mutation takes place. it is the operator that introduces genetic diversity in the population. the mutation takes place whenever the population tends to become homogeneous due to repeated use of reproduction and crossover operators. it occurs during evolution according to a user-defined mutation probability, usually set to fairly low. mutation alters one or more gene values in the chromosome from its initial state. this can produce the entirely new gene values being added to the gene pool. with this new gene values, the genetic algorithm may be able to produce a better solution than was previously. keep best solution there is a solution that might satisfy good fitness function, but it is not selected during the crossover process. overview of proposed system system model our work based on the iaas (infrastructure as a service) model of the system is illustrated in figure 2: customers submit requests through an interface via their endpoints; cloud service providers provide customers with a virtual machine as a unit of required computing resources. cloud customers rent these resources and pay the provider based on the amount of resources occupied and the length of pa ge 14 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 time they occupy them. the iaas provides the hardware equipment and other basic resources as a service to users or customers in the cloud. the biggest advantage of iaas is that it allows users to dynamically apply or release nodes, charging based on usage. the number of servers operating in iaas reaches hundreds of thousands, so the resources that users can request are almost unlimited. at the same time, the iaas is shared by the public, which allows for greater efficiency in the use of resources. this architecture works as follows: upon receipt of service requests submitted by the broker, the rm (request manager) evaluates them and ranks them in descending order of execution priority in the list of requests and makes them available to the central scheduler (cs). the central scheduler (cs) sends a request for avail ability of resources or virtual machines (vms) to the local scheduler (ls) in the cloud. the central scheduler (cs) will have on one side a list of tasks grouped by order of execution priority and on the other side a list of resources (virtual machines) grouped by order of decreasing execution power. the central scheduler (cs) then performs the genetic meta-ordering algorithm procedure on these two lists to find an optimal solution. the central scheduler (cs) transmits the tasks to be executed to the best execution resources (virtual ma chines) via the appropriate local scheduler (ls). at the end of execution, the local scheduler (ls) receives the processing information via the hypervisors and transmits it to the central scheduler (cs). the various users will be notified of the end of execution. the central scheduler (cs) repeats the same process for new tasks that have arrived and tasks that have not been completed. genetic algorithm of meta-scheduling when the meta-scheduler has at its level a list of user tasks and a list of available virtual machines that can perform the tasks, it determines the priorities of these tasks and then arranges in ascending or descending order. virtual machines, too, are classified by the power of execution of the tasks in the delay. the initialization of the genetic algorithm can be done under priority constraint, type of virtual machine by task and feasibility (deadline). initial population initial population (tasks reach to the vm) is generated figure 4: the iaas system model (song et al., 2014) figure 6: architecture each server is symbolized by a hypervisor in the cloud. this hypervisor controls and keeps information about the execution of several virtual machines installed on it. cloud users frequently use these virtual machines to perform their tasks. while performing these tasks, free time slots can be observed on some virtual machines that can be exploited for performing other tasks of cloud clients. the meta-scheduler will have this responsibility to collect the tasks from the cloud clients, redirect them to the best virtualized resource sites (virtual machines) with free time slots through a hypervisor. figure 5: server machine(song et al., 2014) detailed architecture the detailed architecture consists of several entities that describe the different components of the overall system. pa ge 15 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 randomly. in the task scheduling problem, a chromosome represents a likely solution. this chromosome is encoded by a real number. the length of a chromosome is equal to the number of tasks. the value of the gene represents to which virtual machine the task is allocated. a task is characterized by the waiting time (ati), size or length (li), deadline (di). the priority tasks that have arrived for processing are arranged in a queue through the priority function given by the following formula: (1/(di-ati-li)*1/(ati+1)) (1) the available virtual machines are arranged in a queue in a decreasing manner in terms of power. our coding is as follows: each ti task is assigned to an available vmj virtual ma chine of the iaas cloud. suppose we have n number of tasks (tn) and 3 virtual machines available. an example of a solution is shown in the figure. min(max(tj))=min(max(∑m (i=1)tij)) (3) tij represents corresponding execution time. selection after having evaluated each individual or chromosome of the population, it will be necessary to select those which will undergo the operators of crossing and mutation. here, we will use the elistic selection. it consists in choosing or retaining the best individuals for the next generation. we will select the best chromosomes with small makespan. crossover in order to produce new chromosomes from the parent chromosomes, we will use the crossover operator and more precisely the single-point crossover. mutation mutation is a genetic operator used to maintain genetic diversity from one generation of a population of genetic algorithm chromosomes to the next. all the new individuals obtained after the crossover must undergo the operation of mutation. each genome of the chromosome must be randomly swapped respecting the virtual machine numbers. if we have for example two virtual machines, this mutation operator takes the chosen genome and inverts the bits (i.e. if the genome bit is 1, it is changed to 0 and vice versa). evaluation function after the mutation process, we will reevaluate the new population obtained. it is enough to use the objective function to evaluate the new individuals in order to choose the best ones. termination condition genetic algorithm gets terminated after user specified number of generations. we generated 20 evolutions of genetic algorithm to get the better results. the pseudo code of our genetic algorithm is shown in algorithm 1. simulation results and analysis in order to obtain the results of the proposed algorithm, a simulation was performed using the python3 simulator under windows 10 os with core i3 3.90ghz processor, 500gb hard drive and hard disk and 4gb of ram. python is a powerful and easy to learn programming language. it has high-level data structures and allows for a simple but effective approach to object-oriented programming. because of its elegant syntax, dynamic typing, and interpretability, python is an ideal language for scripting and rapid application development in many areas and on most platforms. in this work, we perform the meta-scheduling of independent tasks and the best scheduling algorithm will be the one that optimizes some resource utilization loads. the goal of this algorithm is to satisfy the user in terms figure 7: representation of a solution figure 8: simplified representation of a solution a simplified representation of a solution is given in the following figure: this literally translates into: task t1 is assigned to virtual machine 3, t2 is assigned to virtual machine 2, etc. we form as many chromosomes considered as probable solutions to the problem. a valid solution must meet the following conditions: -each mapping between the task and a virtual machine in the solution must respect: tij<di a virtual machine can run only one task at a time, but can manage several tasks over time. therefore the tasks must be scheduled over an available time frame. each task is executed only once on a single virtual machine. evaluation function the quality of a candidate solution is evaluated using the fitness function. this is an index to select the best solution. it will be called makespan. a solution will be considered if and only if the deadline of all tasks assigned to the virtual machines making up the solution is met. when a ti task is assigned to a vmj virtual machine, the execution time of the vmj depends on the length of the task (l) and the frequency (f) of the virtual machine. the execution time of li on vmj is given by tij: tij=li⁄fj (2) where i ∈ 1, 2, 3, ....., m and j ∈ 1, 2, 3, , n it should be noted that the global execution time of a solution is equal to the maximum value of the execution times of all the virtual machines, because the execution of the tasks are done in parallel on vms. in order to minimize the makespan, a so-called objective function can be presented as follows: pa ge 16 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 of quality of service such as minimizing the execution time of the tasks and also the makespan. the makespan is the total time needed to process a set of tasks for its complete execution. we will consider in our simulation, a set of user tasks (t1, t2, t3,....,tn) that will run on the virtual machines (vm1, vm2, vm3,. ,vmn) available in the cloud. the frequency of the virtual machines is expressed in millions of instructions per second (mips) and the length of the tasks is expressed in number of instructions. we will calculate the execution time of each task in a list and the makespan. our simulations will be based on: table of the parameters of the proposed genetic algorithm in table 1 • variation of the number and frequency of virtual machines • random variation of task lengths algorithm 1: algorithm of meta-scheduling input:pq[], cvm[],population:p, max iteration result: the global best solution while iteration < max generation do for (i=1; i<=n; i++) do pq[]=task selected and sorted by calling function priorty; cvm[]= type of vm selected and sorted by decreasing order of their capacity or frequency; if length(cvm)==0 then the tasks cannot be performed; else individuali[]=mapping(pq[], cvm[]); end end for (j=0; j<length(individuali); j++) do if deadline<waiting time then the individual is considered; else not considers the individual; end end for (each individual in p) do makespan=fitness function();//evaluation end for (i=1; i<=p;p++) do sort according fitness; //selection return p’; end for (each two of individual in p’) do perform crossover;//crossover return p”; end for (each individual in p”) do perform mutation;//mutation return p”’; end for (each individual in p”’) do perform fitness function;//re-evaluation return the best fitness value; end update the global best solution; iteration++; end table 1: ga parameters parameter value number of tasks 10 number of vm 2 and 4 number of iterations 20 selection type elistic crossover type two-point crossover mutation type random permutation termination condition number of iterations table 2: vm types no instance type cu 1 c3:large 7 2 c3:large 14 3 c3:2xlarge 28 4 c3:3xlarge 55 the virtual machine instances(types) used are the ama zon elastic compute one from table 2 [31] and the task lengths are randomly generated between 0 and 1000 in structions. task = [“taskname”, waiting time, length, deadline] vm = [“vmname”, freq], freq =frequency of the processor on which a virtual machine runs. 10 tasks with 2 vms tasks = [[“t1”, 0, 3, 4], [“t2”, 0, 5, 6], [“t3”, 1, 2, 6], [“t4”, 2, 6.25, 11], [“t5”, 3, 3, 9], [“t6”, 4, 10, 19], [“t7”, 5, 2.5, 13], [“t8”, 6, 10, 21], [“t9”, 7, 4, 23], [“t10”, 7, 3, 29]] vms =[[“vm1”, 1], [“vm2”, 1.25]] pa ge 17 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 10 tasks with 4 vms tasks = [[“t1”, 0, 3, 4], [“t2”, 0, 5, 6], [“t3”, 1, 2, 6], [“t4”, 2, 6.25, 11], [“t5”, 3, 3, 9], [“t6”, 4, 10, 19], [“t7”, 5, 2.5, 13], [“t8”, 6, 10, 21], [“t9”, 7, 4, 23], [“t10”, 7, 3, 29]] vms = [[“vm1”, 1], [“vm2”, 1.25], [“vm3”, 1.5], [“vm4”, 2]] figure 9: 10 tasks with 2 vms figure 10: 10 tasks with 4 vms pa ge 18 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 10 taches avec 4 vms de types amazon elastic compute tasks = [[“t1”, 0, 814, 850], [“t2”, 0, 447, 500], [“t3”, 1, 610, 650], [“t4”, 2, 319, 370], [“t5”, 3, 876, 900], [“t6”, 4, 692, 730], [“t7”, 5, 663, 700], [“t8”, 6, 631, 690], [“t9”, 7, 626, 680], [“t10”, 7, 655, 700]] vms = [[“vm1”, 7], [“vm2”, 14], [“vm3”, 28], [“vm4”, 55]] figure 11: 10 taches avec 4 vms de types amazon elastic compute in (shaminder et al., 2012), the authors present the standard genetic algorithm(sga) and the modified genetic algorithm(mga). the results that the sga and mga present in relation to the makespan and the number of cloudlets are visible on figure 12. figure 12: shows average makespan vs. number of cloudlets (shaminder et al., 2012) pa ge 19 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 in (he et al., 2003), figure 13, we present the results on the simulation of the min-min algorithm. here we have 04 types of virtual machines and 10 tasks with length between 0 and 1000 as in our case. the figures above make the comparison between the execution time and the makespan. compared to the presented values, we can say that our proposed meta scheduling genetic algorithm performs well and is better than other algorithms in the literature namely, the stan dard genetic algorithm and the min-min algorithm. conclusion the work in this paper presents a genetic algorithm for centralized meta-scheduling in cloud computing. the proposed algorithm optimizes the use of resources taking into account the minimization of the processing time of each user task and the makespan. this algorithm is based on bio-inspired algorithms, namely the genetic algorithm. the objective of this genetic algorithm of metascheduling is to achieve a quality of service to the user. experimental results show that our genetic algorithm performs well compared to others in the literature. later, we will implement a genetic algorithm for meta scheduling scientific workflows in a decentralized and multi-level context in cloud computing. references aarts e, j. korst, w. michiels. (2005). simulated annealing. in search methodologies; springer: boston, ma, usa, 187210. ahmad, r.w., a. gani, s.h.a. hamid, m. shiraz and f. xia et al. (2015b). virtual machine migration in cloud data centers: a review, taxonomy and open research issues. j. supercomput., 71, 2473-2515. amjad mahmood, salman a. khan 2, rashed a. bahlool (2017). hard real-time task scheduling in cloud computing using an adaptive genetic algorithm. journal computers, 1-21. arya, l.k. and a. verma. (2014). workflow scheduling algorithms in cloud environment-a survey. proceedings of the recent advances in engineering and computational sciences, mar. 6-8, ieee xplore press, chandigarh, india, 1-4. atul l, dharmendra kumar y (2015). multiobjective tasks scheduling algorithm for cloud computing throughput optimization. procedia computer science, 107-113. buyya r, c. s. yeo, s. venugopal, j. broberg and brandic. (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. buyya r, r. ranjan, and r. n. calheiros. (2009). modeling and simulation of scalable cloud computing environments and the cloudsim toolkit: challenges and opportunities. in high performance computing and simulation, international conference, pp. 111. christodoulopoulos, sourlas, mpakolas, and varvarigos. (2009). a comparison of centralized and distributed meta-scheduling architectures for computation and communication tasks in grid networks. elsevier b.v, 0140-3664. dasgupta, k., mandal, b., dutta, p., mandal, j. k., dam, s. (2013). a genetic algorithm (ga) based load balancing strategy for cloud computing. procedia technology, 10, 340-347. dillon t, c. wu and e. chang. (2010). cloud computing: issues and challenges. proceedings of the ieee 24th international conference advanced information networking and applications. perth, 20, 31, 27–33. durga l, n.srinivasu. (2016). a dynamic approach to task scheduling in cloud computing using genetic algorithm. journal of theoretical and applied information technology, 85(2), 124-135. he, x.; sun, x.; von laszewski, g. (2003). qos guided min-min heuristic for grid task scheduling. j. comput. sci.tech. 18, 442-451. he,sun, von laszewski (2003). qos guided min-min heuristic for grid task scheduling. j. comput. sci. tech. 18, 442–451. jang, s. h., kim, t. y., kim, j. k., lee, j. s. (2012). the study of genetic algorithm-based task scheduling for cloud computing. international journal of control and automation, 5(4), 157-162. juntao ma, weitao li, tian fu, lili yan and guojie hu (2016). a novel dynamic task scheduling algorithm based on improved genetic algorithm in cloud computing. wireless communications, networking and applications, 829-835. kairong duan, simon fong, shirley w. i. siu 1, wei song and steven sheng-uei guan (2018). adaptive incremental genetic algorithm for task scheduling in cloud environments. journal symmetry, 1-13. kaur s., verma a. (2012). an efficient approach to genetic algorithm for task scheduling in cloud computing environment. international journal of in formation technology and computer science (ijitcs), 4(10), 74. kaur, r., kinger, s. (2014). enhanced genetic algorithm based task scheduling in cloud computing. international journal of computer applications, 101(14). figure 13: min-min (he et al., 2003) pa ge 20 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 10-20, 2023 malhotra, l., d. agarwal and a. jaiswal. (2010). virtualization in cloud computing. j. inform. tech. softw, eng., 4, 136-136. mao, y.; chen, x.; li, x. (2014). max-min task scheduling algorithm for load balance in cloud computing. in proceedings of the international conference on computer science and information technology, barcelona, spain, springer: new delhi, india, 457-465. nagadevi, s., k. satyapriya and d. malathy. (2013). a survey on economic cloud schedulers for optimized task scheduling. int. j. adv.eng. tech., 4, 58-62. nagwan m. abdel samee, sara sayed ahmed and rania ahmed abdel azeem abul seoud (2019). metaheuristic algorithms for independent task scheduling in symmetric and asymmetric cloud computing environment. journal of computer science, 594-611. previous generation instances. available on line: https://aws.amazon.com/ec2/previous generation/?nc1= hls. (accessed on 1 april 2018) radulescu a, a. gemund, (2000). fast and effective task scheduling in heterogeneous systems. proceedings of the 9th heterogeneous computing workshop, pp. 229-238. raj j and r. m. thomas. (2013). genetic based scheduling in grid systems: a survey. in computer communication and informatics (iccci), international conference pp. 1-4. rasha a. al-arasi , anwar saif (2018). htscc: a hybrid task scheduling algorithm in cloud computing environment. journal: international journal of computers and technology, 17. shaminder k, amandeep verma (2012). an efficient approach to genetic algorithm for task scheduling in cloud computing environment i.j. information technology and computer science , 10, 74-79. singh r, p. sanchita , a. kumar (2014). task scheduling in cloud computing: review. international journal of computer science and information technologies, 5, 79407944. singh, k., m. alam and s.k. sharma, (2015). a survey of static scheduling algorithm for distributed computing system. int. j. comput. applic., 129, 25-30. singh, p., m. dutta and n. aggarwal. (2017). a review of task scheduling based on meta-heuristics approach in cloud computing. knowl. inform, syst. 52, 1-51. song, w., xiao, z., chen, q. and luo, h. (2014). adaptive resource provisioning for the cloud using online bin packing. computers, ieee transactions on, 63(11) pp.2647-2660. xu m., cui l.,wang h.,bi y. (2009). a multiple qos constrained scheduling strategy of multiple work flows for cloud computing. parallel and distributed processing with applications, 2009 ieee international symposium on. ieee, 2009. zhang q, l. cheng and r. boutaba. (2010). cloud computing: state-of-the-art and research challenges. journal of internet services and applications, 1(1), 7-18. pa ge 1 pa ge 22 american journal of smart technology and solutions (ajsts) freshmen’s technological expertise and distance learning readiness: a convergent parallel design jordan b. mallari1, ivy t. lozarita1, john harry s. caballo1*, jobert c. regidor1 volume 3 issue 1, year 2024 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v3i1.2423 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: january 31, 2024 accepted: march 08, 2024 published: march 11, 2024 this convergent parallel mixed methods study primarily focused on technological expertise and distance learning readiness of freshmen students. the study had two phases: the quantitative phase and the qualitative phase. in the quantitative phase, results found that technological expertise and distance learning readiness were highly correlated with each other, with an r-value of .828 and a p-value of 0.000. in the qualitative phase, four themes were extracted: technological tools as a significant factor in distance education, technological expertise as influenced by minor skills and educational courses, technological use due to education’s status quo, challenges in technological availability, and technological background. the results showed a merging and converging characteristic of both data. the lived experiences of the freshmen students showed that technological tools, which are an indicator of technological expertise, are also a factor to be considered in distance learning readiness. keywords technological expertise, distance learning readiness, mixed method 1 university of mindanao senior high school, bolton campus, davao city, philippines * corresponding author’s e-mail: harrycaballo@gmail.com introduction the covid-19 pandemic prompted a significant overhaul of the education system worldwide, with countries implementing distance learning measures such as online tools, television broadcasts, and video lectures (unesco, 2020; bana et. al). mitchell (2020) noted that only a third of global universities offer full online courses, impacting students significantly. unesco’s data (2020) indicated that at its peak in april 2020, over 91% of students globally were affected by school closures, highlighting the pandemic’s profound impact on education. gabor (2020) emphasized the dual challenges of school closures and students’ lack of technological access and skills. llego’s (2020) report revealed a decline in student enrollment for the school year 2019-2020, indicating difficulties in adjusting to the new learning mode. a study among medical students (baticulon, 2020) showed that only 41% felt capable of engaging in online learning, with technological barriers cited. the data also highlighted many students in region xi facing potential challenges due to a lack of technological expertise. freshmen students, being more susceptible to changes in learning (olutola et al., 2016), require careful examination of their readiness for distance learning. studies (kaymak, 2013; beyth-marom et al., 2013; kitsantas & chow, 2010) emphasized the relationship between online learning readiness, learning preferences, and psychological attitudes toward technology use. learning motivation, study habits, and technological attributes were identified as key factors influencing distance learning performance (horzum et al., 2015; joosten & cusatis, 2020; olutola et al., 2016). the sudden shift to online learning, particularly affecting first-year tertiary-level students, raised concerns at international, national, and local levels. urgent action is needed to understand how freshmen’s technological expertise correlates with their learning readiness (olutola et al., 2016). recognizing the gaps between students’ readiness and technological aspects is crucial for the successful implementation of e-learning (olutola et al., 2016). this paper aims to help higher education institutions assess student readiness in the online learning environment by considering technological factors. research questions this study investigates the correlation between freshmen students’ technological expertise and readiness for distance learning. additionally, it seeks to derive codes and themes from participants’ experiences and ultimately integrate and validate both quantitative and qualitative data. specifically, the study will address the following objectives: 1. what is the level of technological expertise in terms of: 1.1 technological spaces; 1.2 computer skills; and 1.3 perception of the use of technology-enabled learning? 2. what is the level of distance learning readiness in terms of: 2.1 self-direction; 2.2 learning preference; 2.3 study habit; 2.4 technology skills; and 2.5 computer equipment capabilities? 3. is there a significant relationship between the level of technological expertise and the level of distance learning readiness? 4. based on the experiences of participants, what is the role of technological expertise in the distance learning readiness of freshmen students? 5. how do the qualitative data corroborate with the quantitative data? pa ge 23 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 22-31, 2024 literature review technological expertise farley (2019) provides a comprehensive definition of technological expertise, encompassing the skills necessary for intricate computer and physical technology operations within diverse industries. he specifically highlights its significance in enhancing students’ learning experiences and extending their capabilities in mathematical research. additionally, beers (2011) underscores the importance of integrating technological expertise into students’ learning preferences for e-learning encounters, defining it as the ability to utilize technology efficiently, effectively, and ethically for accessing and managing information. ziefle and schaar (2010) highlight the positive impact of technological expertise on understanding modern technologies, particularly in the information and communication technology field. larson and miller (2011) argue that increasing expertise in digital technologies is essential as technology becomes more prevalent in society, especially in researching, organizing, assessing, and communicating information. however, the australian council for educational research (2016) notes that despite its growing importance, technological expertise faces challenges due to revisions in hierarchical structures and production methods, leading to uncertainty, particularly in critical economic sectors like financial services. the level of technological expertise has three indicators: technological spaces, computer skills, and perception of technology-enabled learning. technological spaces as defined by srivastava (2015), encompass the availability of technology equipment for students, representing an essential factor in education. these spaces facilitate efficient learning and effective teaching by providing technical capacity for applying external knowledge (ottenbreit-lefwich, 2014). alsied and pathan (2013) liken technological spaces to the dominance of the english language, emphasizing their role in creating opportunities for efficient learning and teaching. despite government initiatives, the shortage of computer resources can impede technology experiences, leading to a decrease in technological expertise (buabeng-andoh, 2012). gilakjani (2013) supports this, noting the shortage of information technology as a contextual factor affecting computer hardware/software for learners and instructors. the office of educational technology (2016) highlights that technology-enabled learning environments, or technological spaces, enable less experienced learners to access specialized communities, fostering increased engagement in learning. the societal context plays a crucial role in the use of technology in distance learning, establishing a direct association between technological spaces and the independent variable. computer skills refer to the ability and speed to locate, gather, and provide data in the graphical user interface of computers, enabling individuals to perform basic operations (goldhammer et al., 2013). israel (2010) emphasizes the significance of computer skills in the effective use of the internet. paino and renzulli (2012) underscore the role of computer skills in perpetuating inequality in schools, while guclu (2020) views them as crucial skill sets for technological proficiency and career success. baker (2013) supports this by characterizing computer skills as a set of abilities encompassing the use of programs for accounting, databases, email, web, programming, spreadsheets, and word processing. lowe’s (2019) report suggests that computer skills acquired in high school contribute to success in college, particularly in processing and presenting information, aligning with the concept of technological expertise in the current study. perception of use of technology-enabled learning perceived usefulness, as defined by peart, et al. (2017), is a relevant factor influencing the acceptance of technology. it guides research in understanding students’ characteristics and how they define usefulness, encompassing their attitude and acceptance of e-learning training methods. these factors are crucial for the effectiveness of instructional processes, influenced by individual learning styles, available technical tools, learners’ opinions on technological assistance, and technology sophistication in online learning (parai et al., 2015). beckman et al. (2014) case study of 12 students from australian secondary schools supports these findings. the study provides detailed accounts of students’ technology experiences in various contexts, highlighting the heterogeneous and complex network of factors influencing technology practices. the study concludes that students’ perceptions of technology use directly impact their technological expertise for learning. in the current study, students’ perceptions of technology are used as an indicator of the independent variable, aligning with the claim that perception is a direct factor in determining the level of technological expertise. all the previous studies presented reiterate the researchers’ aim to regard technological expertise as a determinant variable. however, it is notable that these studies only involved a minimal description of the level of technical experience and how it correlates to the level of distance learning readiness that this current study will attempt to provide. distance learning readiness distance learning readiness, as specified by hashim and tasir (2014), refers to the preparedness of an institution or organization for various aspects of e-learning technology before implementing the entire e-learning environment for diverse purposes. ngampornchai and adams (2016) highlight that sufficient e-learning readiness can offer educational opportunities to students who cannot access higher education, contributing to a country’s competitiveness in the southeast asian nations association. this readiness variable significantly pa ge 24 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 22-31, 2024 influences students’ perspectives on education, as suggested by yilmaz’s study (2017), which indicates that student motivation is directly affected by the level of distance learning readiness. bovermann et al. (2018) identify technological components and computer skills as critical predictors of social interaction, social communication, and learning outcomes in online learning. additionally, rasouli et al. (2016) highlight the flexibility in methodology, content management, synchronous and asynchronous interactions, organization of courses, educational projects, student projects, and assessment between teachers and students facilitated by distance learning readiness. the level of technological expertise has five indicators: self-direction, learning preference, study habits, technological skills, and computer equipment capabilities. self-direction self-direction in education, as highlighted by petro (2017), encompasses diverse activities such as exploring new knowledge, critical thinking, active participation in learning groups, and creating one’s learning path. boyer et al. (2014) emphasize the efficiency of self-direction in engaging and training adult learners. self-directed learning, as described by garland (2017), allows students to decide what and how they learn with teacher guidance. murad et al. (2010) consider it a suitable approach to keep practitioners informed about recent literature. the effectiveness of self-directed learning extends to medical students, patients, physicians, nurses, and healthcare practitioners. schiller’s (2020) findings emphasize the importance of self-direction in a student’s ability to adapt to distance learning, particularly in completing assignments and handling technological tasks. online classes, as mentioned in the research, require students to navigate virtual spaces independently, indicating the relevance of self-direction to students’ distance learning readiness. learning preference learning preference, as defined by liew, barua, and sidhu (2015), refers to the adopted learning style or methods preferred by learners for acquiring, evaluating, and interpreting insight. james, thomas, and d’amore (2011) characterize learning preference as an individual’s favored way of acquiring knowledge and comprehending information, aligning with the psychological type theory of jung. jain, sharma, and joachims (2015) suggest that student learning preferences are complex and influenced by the context of various objects and human interactions in their environment. battalio’s study (2010) indicates that learners with healthy learning preferences exhibit more success and better adaptation to online courses compared to other learning dimensions tested. this underscores that learning preference serves as a valuable indicator of students’ readiness and performance in distance education. study habit study habits, as defined by uju and paul (2017), encompass how an individual approaches their studies, emphasizing the importance of good study habits for academic success. rabia et al. (2017) further explain that study habits determine the extent of learning, one’s academic goals, and the level of achievement. olatula et al. (2016) note that study habits can be systematic, efficient, or inefficient, with good study habits contributing to better academic performance and a lack thereof leading to inadequacies. somuah and dankyi’s study (2014) focused on the study habits of university of cape coast distance education learners in ghana. the results revealed habits such as reading while lying in bed, studying for less than three hours a day, not seeking additional materials from the internet, and reviewing notes before face-to-face meetings. the study suggests that study habits play a role in the distance learning readiness of students, with implications for online learning, where unhealthy study habits may be manifested. technology skills computer skills, defined as the ability to use computers and technology effectively (resume.com, 2019), involve acquiring expertise in performing physical or digital tasks (indeed career guide, 2017). wikiversity (2020) stated that technological abilities, especially computer skills, are crucial for effectively utilizing computers and related technology. in the study by somuah and dankyi (2014) on university of cape coast distance education learners, computer skills were considered vital, particularly in the context of distance learning education. the study suggests that those who do not develop their computer expertise may lag behind in the digital revolution, emphasizing computer skills as fundamental predictors of a student’s readiness for engagement in distance learning. computer equipment capabilities digital accessibility, as defined by kulkarni (2018), pertains to the extent to which people with disabilities or special needs can access and navigate products, devices, services, or environments. foley and ferri (2012) note that accessible technologies enhance access and flexibility for students and adults. however, haines et al. (2018) suggest that technology availability has minimal and fleeting effects on group development. mitchell (2020) emphasizes the importance of computer capabilities in online learning, where distance education relies on electronic communications. the research indicates that a lack of computer capabilities among students can result in decreased performance and engagement in online learning, suggesting a direct correlation between computer equipment capabilities and distance learning readiness. notably, while many studies on online learning readiness focus on college students and their relationship with variables like perception and satisfaction, research on the potential connection between online learning readiness and technological expertise is limited. the researchers aim to address this pa ge 25 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 22-31, 2024 gap and contribute to the growing literature on the topic. correlation between variables halili’s (2019) study, “technological advances in education,” explores the use of advanced technology in teaching and learning to enhance the education system and engage students in distance learning. the research suggests that the efficient utilization of advanced technology is crucial for student engagement in online schooling, contributing to the readiness for distance learning. a quantitative approach revealed a lack of efficient technology usage in malaysia’s education system, urging improvements for global competitiveness. cigdem and yildirim’s (2014) study investigates the relationship between students’ technological approaches and online learning readiness at a vocational college. the research implies that students’ technological expertise may positively relate to their readiness for online learning, affecting dimensions of online learning readiness. wei and chou’s (2020) study explores the structural model determining the impact of distance learning perceptions and readiness on students’ online learning performance. results suggest that students’ computer efficacy and motivation positively influence online performance, indicating a positive relationship between computer skills (a component of technological expertise) and online readiness. contrary to these findings, mafenya (2013) presents empirical evidence opposing the idea that technological spaces or resources significantly impact students’ readiness for e-learning. the study at the university of south africa indicates a high level of acceptance despite resource shortages, challenging the assumption that technological spaces directly affect readiness. lai, wang, and lei (2012) suggest in their study that adapting technology indirectly affects computer selfefficacy and perceived utility in online education. this contradicts the notion that technology use positively relates to student performance and engagement in online learning, providing an alternative perspective. raes and dapaepae’s (2019) study explores how exposure to technology can improve students’ performance in school. the results suggest that students exhibit a positive predisposition to educational technology’s usefulness and ease of use, with perceptions significantly improving after technology exposure. in summary, these studies offer varying perspectives on the relationship between technological expertise, readiness for online learning, and the impact of technology on students’ performance and engagement in distance education. materials and methods a convergent parallel mixed-method design was utilized in this study. creswell and clark (2011) stated that mixed methods research includes at least one quantitative and one qualitative strand. for the quantitative phase, this study employed a descriptive correlational research model to explore relationships between variables, specifically examining if changes in one variable corresponded to changes in another (tan, 2014). in the qualitative phase, a phenomenological design was employed, concentrating on individual experiences through in-depth, occasionally repeated interviews with participants. this approach allowed researchers to extract insights for generating themes and codes, facilitating cross-referencing of data from both researches. this study focused on freshmen students from diverse courses and schools in davao city during the 20202021 academic year. for the quantitative phase, due to the unspecified total freshmen population, the researchers utilized a purposive snowball sampling method, particularly the maximum variation sampling technique, resulting in a sample size of one hundred (100) freshmen students. purposive sampling, also known as judgmental or selective sampling, involves researchers using their judgment to choose participants based on prior knowledge of the study’s purpose (foley, 2018). by combining this with snowball sampling, where participants nominate others, the researchers ensured diversity in the sample. maximum variation sampling, guided by researchers’ judgment, was employed to select participants with diverse characteristics (dudovskiy, 2010). for the qualitative phase, fourteen participants were purposively chosen, nine for in-depth interviews and 5 for focus group discussion. the selection criteria ensured maximum variation of responses, aligning with the vulnerability of freshmen students to learning changes (olutola et al., 2016). open-ended, discoveryoriented interviews were conducted to explore participants’ feelings and perspectives deeply (guion et al., 2015). focus group discussions, recommended for understanding issues at a deeper level than surveys (prasad & garcia, 2015), complemented the qualitative aspect of the research. the data analysis method applied in this study, as per spaulding (2011), was unilaterally chosen and had a significant impact on the questionnaire’s form and purpose. the research instrument, the questionnaire, was adapted from das and mishra (2016) for technological expertise and penn state university (2015) for distance learning readiness. the adapted questionnaire comprised two sections, with the first divided into three segments and the second into five, each addressing specific indicators of the independent and dependent variables. each category contained thirty questions, totaling sixty in the questionnaire. respondents expressed their agreement with statements on a scale from one to five, reflecting varying levels of agreement. three experts validated the questionnaire, yielding a mean score of 4.57 out of 5, indicating excellent quality. for determining technological expertise, a test questionnaire adapted from das and mishra (2016) was used, and the level was calculated based on the mean scores of the independent variable indicators. for assessing distance learning readiness, a standardized questionnaire from penn state university (2015) was adopted, and the level was determined using pa ge 26 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 22-31, 2024 the mean scores of the dependent variable. pilot testing with fifty individuals ensured the questionnaire’s credibility and reliability, as confirmed by a cronbach’s alpha of .928, indicating excellent reliability and internal consistency for all sixty items (institute for digital research and education, 2020). in the qualitative phase, the researchers employed an interview guide for in-depth and focus group discussions. this guide, according to menzies et al. (2016), is a list of high-level topics with corresponding questions designed to address specific aspects during the interview. the interview guide utilized in this study consisted of one central question with seven sub-questions. results and discussion technological expertise of freshmen students table 1 shows the level of technological expertise table 1: technological expertise of freshmen students indicators mean standard deviation descriptive level technological spaces 3.6460 .66049 high computer skills 3.5492 .74251 high perception of the use of technology-enabled learning 3.8143 .86038 high overall 3.6698 .64627 high concerning its three (3) indicators: technological spaces, computer skills, and perception of the use of technologyenabled learning. all indicators had a mean of more than 3.49 but less than 4.50, which gave us a high overall mean of 3.6698 for the independent variable. this indicated that the level of technological expertise was manifested by the freshmen students oftentimes. further, students had a strong foundation of expertise in using technology in distance learning. table 2 mirrored the level of distance learning readiness concerning its three (5) indicators: self-direction, learning preference, study habit, technological skills, and computer equipment capabilities. all indicators except for the computer equipment capabilities had more than 3.49 but less than 4.50, which gave us a high overall mean of 3.7018 for the dependent variable. this indicated that the freshmen’s level of distance learning readiness is manifested oftentimes. further, this meant that the students had a high readiness level for distance learning education. table 2: distance learning readiness of freshmen students indicators mean standard deviation descriptive level self – direction 3.8660 .79952 high learning preference 3.8871 .61865 high study habit 3.5386 .67001 high technological skills 3.7940 .83434 high computer equipment capabilities 3.4233 .90435 moderate overall 3.7018 .61266 high table 3 illuminates a robust and statistically significant relationship between freshmen’s technological expertise and their distance learning readiness, as evidenced by the high r-value of .828. this substantial correlation suggests a direct proportionality, signifying that an increase in technological expertise corresponds to an elevated level of distance learning readiness, and conversely, a decrease in technological expertise corresponds to a decline in distance learning readiness. this observed relationship holds true for freshmen students at the tertiary level. notably, the p-value of 0.000 underscores the significance of the relationship between technological expertise and distance learning readiness. this low p-value affirms that the observed association is not due to chance, further reinforcing the validity and importance of the identified link between these two variables. experiences and insights on technological expertise and distance learning readiness the data gathered from in-depth interviews and focus table 3: significant relationship between technological expertise and distance learning readiness variables r value p-value decision interpretation technological expertise .828 .000 reject h0 significant distance learning readiness group discussions were transcribed, coded, and organized into themes. for this particular research question on technological expertise and distance learning readiness, four themes were generated. the themes are as follows: technological tools as a significant factor in distance education, technological expertise as influenced by minor skills and educational courses, technological use due to education’s status quo, and challenges in pa ge 27 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 22-31, 2024 technological availability and technological background. technological tools as a significant factor in distance education when freshmen students were interviewed about their technological expertise and distance learning readiness, their responses reflected that technological tools as a significant factor in distance education became a reoccurring theme. the said theme includes three codes, namely: efficiency of technological tools, time management as an effect of utilizing technology, and learning independence during distance education. technological expertise as influenced by minor skills and educational courses when the participants were asked what experiences influenced them to have the level of technological expertise they have at the moment, researchers found out that their experiences are influenced by two major factors. first, are their minor skills from using the internet and technological tools. secondly, are the educational courses they went through back in junior high school and senior high school. besides, two codes were generated from the theme: minor technological skills influences students’ technological learning skills and technology education in high school as major influence to students technological expertise. technological use due to education’s status quo when freshmen students were interviewed about their technological expertise and distance learning readiness, their responses reflected that the technological use due to education’s status quo became a reoccurring theme. the said theme has two codes: more frequent technological use and lms as common factor of technological use. challenges in technological availability and technological background when the participants were interviewed about their difficulty in using technology during their distance learning education, their responses revealed that particular challenges in technological availability and technological background became an essential theme. this theme has two codes which include accessibility of technology affects the level of difficulty, and students’ insufficient technological knowledge affects the level of difficulty. corroboration of the qualitative and quantitative findings merging the nature of data in the given focal point is merging. the results from the quantitative and qualitative phases were integrated using a joint display. in the process, the researcher evaluates and analyzes first the quantitative findings and then discusses them with the qualitative findings to either confirm or not the statistical results (todd, 2016). according to creswell (2006), the data are merged when the researcher takes the two data sets and explicitly brings them together or integrates them. converging table 6: joint display of salient qualitative and quantitative findings aspects/focal point quantitative data findings qualitative data findings nature of data integration axiological implications relationship between technological expertise and distance learning readiness r-value is .828 which indicates that the two variables are highly correlated by 82.8% the p-value is 0.000 which indicates that the relationship between two variables is significant. thus, rejecting the null hypothesis, means that there is a significant relationship between the freshmen’s technological expertise and distance learning readiness. technological tools as a significant factor in distance education with codes: • efficiency of technological tools • time management as an effect of utilizing technology • learning independence during distance education merging converging freshmen students should have enough technological knowledge and skills to be ready in distance learning. the findings of the study show that the integration of the quantitative and qualitative data is converging. this trend occurs when both data are collected at different times and evaluates them separately. afterwards, equating the result to approve the findings (demir & pismek, 2017). relationship between technological expertise and distance learning readiness as illustrated in the table, column 2 provides quantitative results, which shows that there is a significant relationship between the technological expertise and distance learning readiness of students. there is a high correlation between the two variables with 82.8% of correlation. their relationship is also significant with a 0.000 p-value. the corroborated findings approved the mentioned study by halili (2019), where they found that the usage of pa ge 28 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 22-31, 2024 technology can improve student’s engagement in online learning. further, cigdem and yildirim’s (2014) report supported the claim as it was inferred from their findings that technological expertise could be positively related to their learning readiness in an online setting. moreover, the current study’s findings complemented the results of wei and chou’s (2014) study. their results showed that computer skills relate to the independent variable of the current study as a reason or a mediating variable for a high level of distance learning readiness among freshmen students. conclusion based on the findings of the study, freshmen students manifested a very strong foundation of technological expertise. students also showed sufficient technological availability in their learning environment. further, freshmen students had enough computer skills and had a positive perception of technology-enabled learning. moreover, students also showed a very strong readiness level for distance learning. to elucidate, they manifested healthy self-directed thinking towards distance education. freshmen students also had a good learning preference and healthy study habits during their education in the distance learning mode. further, their skills in utilizing technologies were also sufficient. however, their usage of computer equipment was not that highly manifested all the time. the lived experiences of freshmen students, on the other hand, imply that they become more efficient, time-bounded, and independent while learning online with respect to their high level of technological expertise. thus, it corroborates positively with the statistical results from the quantitative phase. therefore, freshmen students possess technological skills such that they become more ready to engage in distance learning. these skills have been used to help them be more effective despite the struggles they encounter online. based on the findings that freshman students have a strong technological foundation and readiness for distance learning, the researchers would make the following research recommendations: expand the study to include students from various demographics and backgrounds to see if there are differences in technological expertise and distance learning readiness based on other factors. survey professors and university administrators about their perspectives on students’ technology skills and needs related to distance learning. identify any gaps between student capabilities and instructor expectations. investigate the technology training and support provided by colleges/universities to identify gaps and areas for improvement in preparing freshman students for online learning. research the influence of high school experiences with technology on freshman students’ university preparation. determine what capabilities students gain before college. references alsied, s. m., & pathan, m. m. (2013). the use of computer technology in efl classroom: advantages and international journal of english language & translation studies. international journal of english language & translation studies (2308-5460). australian council for educational research. (2016). a global measure of digital and ict literacy skills. unesdoc digital library, 1-26. retrieved september 18, 2020, from https://unesdoc.unesco.org/ ark:/48223/pf0000245577 baker, w. m. (2013, august 26). empirically assessing the importance of computer skills. journal of education for business, 88(6), 345-351. https://doi.org/10.1080/ 08832323.2012.725433 e. bana, e., a. basanes, r., & malabarbas, g. t. (2022). online learning readiness of public secondary schools: the perspective between school heads and teachers. american journal of education and technology, 1(2), 36–45. https://doi.org/10.54536/ajet.v1i2.393 batallio, j. (2010). success in distance education: do learning styles and mutiple formats matter? journal of distance education, 71-78. https://doi. org/10.1080/08923640902854405 baticulon, r. e., sy, j. j., alberto, n. r. i., baron, m. b. c., mabulay, r. e. c., rizada, l. g. t., ... & reyes, j. c. b. (2021). barriers to online learning in the time of covid-19: a national survey of medical students in the philippines. medical science educator, 31, 615-626. https://doi.org/10.1101/2020.07.16.20155747 beckman, k., bennett, s., & lockyer, l. (2014). understanding students’ use and value of technology for learning. learning, media and technology, 39(3), 346367. https://doi.org/10.1080/17439884.2013.87835 3 beers, s. z. (2011). 21st century skills: preparing students for their future. stem. retrieved september 18, 2020, from https://cosee.umaine.edu/files/coseeos/21st_ century_skills.pdf beyth-marom, r., chajut, e., roccas, s., & sagiv, l. (2013). internet-assisted versus traditional distance learning environments: factors affecting students’ preferences. computers & education, 65-76. http:// www.elsevier.com/locate/compedu/a4.3d bovermann, k., weidlich, j., & bastiaens, t. (2018). online learning readiness and attitudes. bovermann et al. international journal of educational technology in higher, 15(27). https://doi.org/10.1186/s41239-0180107-0 boyer, s. l., edmondson, d. r., artis, a. b., & fleming, d. (2014). self-directed learning: a tool for. journal of marketing education, 36(1), 20-32. https://sisu. ut.ee/sites/default/files/opikasitus/files/boyer_ et_al._2014._self-directed_learning_a_tool_for_ lifelong_learning.pdf buabeng-andoh, c. (2012). factors influencing teachers’ adoption and integration of information and communication technology into teaching: a review of the literature. 8(1). https://www.learntechlib. org/p/188018/ chee, j. d. (2015). pearson’s product-moment correlation: pa ge 29 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 22-31, 2024 sample analysis. https://doi.org/10.13140/ rg.2.1.1856.2726 cigdem, h., & yildirim, o. g. (2014). effects of students’ characteristics. turkish online journal of distance education-tojde, 1-14. https://files.eric.ed.gov/ fulltext/ej1043667.pdf creswell, j. (2006). choosing a mixed method design. 83. retrieved april 20, 2021, from https://www.sagepub. com/sites/default/files/upm-binaries/10982_ chapter_4.pdf creswell, j., & clark, p. (2011). mixed-methods research. das, a., & mishra, s. (2016). questionnaire on learner use of technology. (a. kirkwood, & l. price, eds.) commonwealth of learning. https:// www.researchgate.net/publication/305650994_ questionnaire_on_learner_use_of_technology demir, s. b., & pismek, n. (2017). a convergent parallel mixed-methods study of controversial issues in social studies classes. educational sciences: theory & practice, 119-148. https://doi.org/0.12738/ estp.2018.1.029 dudovskiy, j. (2010). purposive sampling. retrieved october 12, 2020, from https://research-methodology.net/ sampling-in-primary-data-collection/purposivesampling/ ericsson, g. (2015). expertise theory. https://www. learning-theories.com/expertise-theory-ericssongladwell.html farley, a. (2019). technical skills. https://www. investopedia .com/ter ms/t/technica l -sk i l l s. asp#:~:text=technical%20skills%20refer%20 to%20the,diverse%20group%20of%20other%20 enterprises. foley, a., & ferri, b. a. (2012). technology for people, not disabilities: ensuring access and inclusion. journal of research in special educational needs, 12(4). https:// doi.org/0.1111/j.1471-3802.2011.01230.x frost, j. (2020). degrees of freedom in statistics. gabor, a. (2020). schools aren’t ready for online learning. https://www.bloombergquint.com/gadfly/ coronavirus-shutdowns-highlight-weakness-of-onlineu-s-schooling garland, w. a. (2017). self-directed learning: definition & strategies. https://study.com/academy/lesson/ self-direct-learning-definition-strategies.html gilakjani, a. p. (2013). factors contributing to teachers’ use of computer technology in the classroom. universal journal of educational research, 262-267. https://doi.org/10.13189/ujer.2013.010317 goldhammer, f., naumann, j., & keßel, y. (2013). assessing individual differences in basic computer skills. european journal of psychological assessment, 263275. https://doi.org/10.1027/1015-5759/a000153 guion, l. a., diehl, d. c., & mcdonald, d. (2015). conducting an in-depth interview. haines, r., scamell, r. w., & shah, j. r. (2018). the impact of technology availability and structural guidance on group development in workgroups using computer-mediated communication. informations system management, 346-368. https://doi. org/10.1080/10580530.2018.1503805 halili, s. h. (2019). technological advancements in education 4.0. the online journal of distance education and e-learning, 7(1). retrieved september 18, 2020, from http://tojdel.net/journals/tojdel/articles/v07i01/ v07i01-08.pdf hall, s. v. (2012). applying the principle of readiness to shorthand instruction. the journal of business education, 357-361. retrieved october 1, 2020, from https://shorturl.at/adkrs hashim, h., & tasir, z. (2014). e-learning readiness: a literature review. international conference on teaching and learning in computing and engineering, 267-271. horzum, m. b., kaymak, z. d., & gungoren, o. c. (2015, april 17). structural equation modeling towards online learning. (1303-0485). https://doi. org/10.12738/estp.2015.3.2410 indeed career guide. (2017). technical skills: definitions and examples. https://www.indeed.com/careeradvice/resumes-cover-letters/technical-skills institute for digital research and education. (2020). what does cronbach’s alpha mean? https://stats. idre.ucla.edu/spss/faq/what-does-cronbachs-alphamean/ israel, o. (2010). undergraduates’ computer skills and the use of online information resources: a case study of library and information science students of delta state university, nigeria. 1-13. http://irjlis. com/wp-content/uploads/2016/05/15-ir-333-54. pdf jain, a., sharma, s., & joachims, t. (2015). learning preferences for manipulation tasks from online coactive feedback. the international journal of robotics and research, 34(10). https://doi. org/10.1177/0278364915581193 james, s., thomas, t., & d’amore, a. (2011). learning preferences of first year nursing and midwifery students: utilizing vark. nurse education today, 31(4), 417-423. https://doi.org/10.1016/j.nedt.2010.08.008 jian, p., & chen, l. (2011). the effects of education compatibility and technological expectancy on e-learning acceptance. https://www.sciencedirect. com/science/article/abs/pii/s0360131511000479 joosten, t., & cusatis, r. (2020). online learning readiness. american journal of distance education(1538-9286 ). https://doi.org/10.1080/08923647.2020.1726167 kitsantas, a., & chow, a. (2010). college students perceived threat and preference for seeking help in traditional, distributed, and distance learning environments. science direct, 383-395. kulkarni, m. (2018). digital accessibility: challenges and opportunities. iimb management review, 13(1). https://10.1016/j.iimb.2018.05.009 lai, c., wang, q., & lei, j. (2012). what factors predict undergraduate students’ use of technology for pa ge 30 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 22-31, 2024 learning? a case from hong kong. 59(2), 569-579. https://doi.org/10.1016/j.compedu.2012.03.006 larson, l. c., & miller, t. n. (2011). 21st century skills: prepare students for the future. kappa delta pi record, 47(3), 121-123. htt ps://10.1080/00228958.2011.10516575 liew, s.-c., barua, a., & sidhu, j. (2015). the relationship between learning preferences (styles and approaches) and learning outcomes among preclinical undergraduate medical students. 44. https:// bmcmededuc.biomedcentral.com/articles/10.1186/ s12909-015-0327-0#sec1 llego, m. a. (2020). deped basic education statistics for school year 2019-2020. https://www.teacherph. com/deped-basic-education-statistics-schoolyear-2019-2020/ lowe, j. (2019). importance of computer skills in college. https://education.seattlepi.com/computercourses-beginners-1612.html mafenya, p. n. (2013). an investigation of first-year students’ pedagogical readiness to e-learning and. mediterranean journal of social sciences, 4(13), 1. https://10.5901/mjss.2013.v4n13p353 menzies, t., williams, l., & zimmermann, t. (2016). interview guide. perspectives on data science for software engineering. https://www.sciencedirect.com/ book/9780128042069/perspectives-on-data-sciencefor-software-engineering mitchell, n. (2020). universities not ready for online learning – u-multirank. https://www.universityworldnews. com/post.php?story=20200609183303614 murad, m. h., coto‐yglesias, f., varkey, p., prokop, l. j., & murad, a. l. (2010). the effectiveness of self‐directed learning in health professions education: a systematic review. wiley online library, 44(11), 1057-1068. https://doi.org/10.1111/j.13652923.2010.03750.x ngampornchai, a., & adams, j. (2016). students’ acceptance and readiness for. international journal of educational technology in, 13(14), 1-13. https://doi. org/10.1186/s41239-016-0034-x office of educational technology. (2016). engaging and empowering learning through technology. https:// tech.ed.gov/netp/learning/ olutola, a. t., olatoye, o. o., & olatoye, r. a. (2016). assessment of social media utilization and study habit of students of tertiary institutions in katsina state. journal of education and practice, 7(3). ottenbreit-leftwich, a. t., & ertmer, p. a. (2014). teacher technology change. journal of research on technology in education 42(3), 255-284. https://doi.or g/10.1080/15391523.2010.10782551 paino, m., & renzulli, l. a. (2012). digital dimension of cultural capital: the (in)visible advantages for students who exhibit computer skills, 86(2), 124138. https://doi.org/10.1177/0038040712456556 parai, m., shenoy, p., & loh, k. (2015). students’ perception of technology-assisted learning in undergraduate medical education – a survey. the social science journal, 52, 78-82. https://doi.org/10.1016/j. soscij.2014.08.007 peart, d. j., penny, l., rumbold, s., keane, k. m., & allin, l. (2017). student use and perception of technology enhanced learning in a mass lecture knowledge-rich domain first year undergraduate module. international journal of educational technology in higher education, 14. https://link.springer.com/article/10.1186/s41239017-0078-6 penn state university. (2015). online learning readiness questionnaire. http://tutorials.istudy.psu.edu/ learningonline/orq/orq.htm?fbclid=iwar0okv xkaamsecevio67jxuawp6f7otqjbzcvxt15li1 ax-ovgux1geojdu petro, l. (2017). how to put self-directed learning to work in your classroom. https://www.edutopia.org/ discussion/how-put-self-directed-learning-workyour-classroom prasad, m., & garcia, c. (2015). how to conduct a successful focus group discussion. humans of data. questionpro. (2020). snowball sampling: definition, method, advantages and disadvantages. https:// www.questionpro.com/blog/snowball-sampling/ rabia, m., mubarak, n., tallat, h., & nasir, w. (2017). a study-on-study habits and academic performance of students. 7. https://doi.org/0.18488/ journal.1.2017.710.891.897 raes, a., & depaepe, f. (2019). a longitudinal study to understand students’ acceptance of technological reform. when experiences exceed expectations. 533552. https://link.springer.com/article/10.1007/ s10639-019-09975-3 rasouli, a., rahbania, z., & attaran, m. (2016). students’ readiness for e-learning. 4(3). www.mojet.net resume.com. (2019). how to list computer skills on a resume (computer skill examples). https://www. resume.com/career-advice/resumes/how-to-listcomputer-skills-on-a-resume-with-examples/ salkind, n. j. (2010). encyclopedia of research design. (n. j. salkind, ed.) 1-0. schiller, n. (2020). self-directed learning for online class. somuah, b. a., & dankyi, l. a. (2014). an investigation into the study habits of distance learners: implications for guidance and counseling services. mediterranean journal of social sciences, 5(6), 273-282. https://doi.org/10.5901/mjss.2014.v5n6p273 spaulding, a. p. (2011,). action research methods: the pro-forma. https://alexandrapspaulding.wordpress. com/2011/03/16/action-research-methods-the-proforma/ srivastava, m. k., gnyawali, d. r., & hatfield, d. e. (2015). behavioral implications of absorptive capacity: the role of technological effort and technological capability in leveraging. technological forecasting and social change, 92, 346-358. https://sci-hub.tw/ pa ge 31 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(1) 22-31, 2024 https://doi.org/10.1016/j.techfore.2015.01.010 tan, l. (2014). correlational study. 269-271. https:// us.sagepub.com/enus/nam/music-in-the-social-andbehavioral-sciences/book240878 todd, m. (2016). joint display analysis. uju, e., & paul, o. (2017). study habit and its impact on secondary school students’ academic performance in biology in the federal capital territory, abuja. academic journals, 12(10). https://doi.org/10.5897/ err2016.3117 wei, h.-c., & chou, c. (2020). online learning performance and satisfaction: do perceptions and readiness matter? distance education, 41(1), 48-69. https://doi.org/10.1080/01587919.2020.1724768 wikiversity. (2020). computer skills. https:// en.wikiversity.org/wiki/computer_skills yilmaz, r. (2017). exploring the role of e-learning readiness on student satisfaction and. computers in human behavior, 70, 251-260. ziefle, m., & schaar, a. k. (2010). technical expertise and its influence on the acceptance of future medical technologies: what is influencing what to which extent?. in hci in work and learning, life and leisure: 6th symposium of the workgroup human-computer interaction and usability engineering, usab 2010, klagenfurt, austria, november 4-5, 2010. proceedings 6 (pp. 513-529). springer berlin heidelberg. https:// doi.org/10.1007/978-3-642-16607-5_40 pa ge 1 pa ge 25 american journal of smart technology and solutions (ajsts) design and development of a tractor-drawn cultivator with a pulverizing attachment husen bona wako1*, adesoji matthew olaniyan2 volume 3 issue 2, year 2024 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v3i2.3337 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: august 02, 2024 accepted: august 30, 2024 published: september 03, 2024 soil tillage is one of the most critical processes in the agricultural production system, requiring the greatest energy and time of any operation. a cultivator with pulverizing attachment was designed and developed to reduce secondary tillage operations to a single pass to ensure timeliness in seedbed preparation. it consists of a pulverizing roller as an active unit and cultivator tynes as a passive unit. the implement has the following major components; a frame, cultivator tines, and a pulverizing roller consisting of central and peripheral shafts on which discs are arranged. the cultivator tynes were fitted on the main frame while the pulverizing roller was attached at the rear end of the frame with the help of bearings. the pulverizing blades were welded on discs in such a way as to crush the soil by impact force. in this way, the force exerted by the pulverizer on the clod is distributed uniformly in the shaft which is used to hold the pulverizing roller. the pulverizer was designed to break big clods formed during tillage operation and the implement was attached to the tractor by three point hitch. in operation, the cultivator tynes open the furrow, and the roller cuts and pulverizes the soil at optimum conditions for tillage. with a production cost of etb 17,587.19, the implement is within the economic reach of an average smallholder farmer in ethiopia. keywords clods, cultivator, draft, pulverizing roller, soil manipulation, tillage 1 agricultural machinery engineering department, oromia agricultural research institute, jimma agricultural engineering research center, jimma, ethiopia 2 department of agricultural engineering, haramaya institute of technology, haramaya university, p. o. box 138, dire dawa, ethiopia * corresponding author’s e-mail: husenifnaan@gmail.com introduction tillage is the process of mechanically manipulating soil to make it more conducive to crop cultivation. soil tillage entails breaking the earth’s compact surface to a particular depth and loosening the soil mass to allow crop roots to penetrate and disseminate into the soil (zhou et al., 2020). it is the mechanical manipulation of soil and plant debris to prepare a seedbed for the planting of seeds that will produce grain for human use. tillage also breaks up the soil, improves the release of soil nutrients for crop growth, eliminates weeds, and improves water and air circulation in the soil (reicosky & allmaras, 2003). tillage is considered one of the most important processes in agriculture, as it gives suitable conditions for root growth, which in turn supports the growth of plants, as it reduces soil resistance, increases the ventilation process, and eliminates weeds (al-shamiry et al., 2020). tillage in the traditional sense is one of the least fuel-efficient procedures. according to (digman, 2012), just 20% of diesel fuel energy is available at the tractor’s drawbar, but only 4% of that energy is transformed into soil-breaking energy. tillage implements works based on two working motions, sliding type, and rotating type. devices like moldboard plows and cultivators use sliding action to cut the soil. by functioning in a rotational motion, disc plows, disc harrows, clod crushers, and rollers cut and pulverize the soil. because of the soil frictional force and the contact area of the implement, sliding-type implements consume more drafts than rotating-type implements. a negative draft was produced by rotary-type implements. as a result, the concept of combining sliding and rotary implements saves a lot of power, time, and money (parmar & gupta, 2001). tillage operations can be performed simultaneously in a single pass by a combination tillage tool mounted on a tractor. according to (manian & kathirvel, 2001), combined tillage is a method of manipulating the soil by using two or more different tillage implements at the same time to reduce the number and time of field operations. according to (prem et al., 2016), combined tillage is the use of two or more implements at the same time to alter the soil. combined tillage, in a broad sense, refers to the integrated management of resources such as time, energy, fuel, labor, soil, and water conservation while also boosting output and better utilizing natural resources. it also helps to maintain agricultural production. combined tillage implement reduced larger size clods in the soil, improved aeration and moisture holding capacity, and obtained medium uniformity of soil and finer pulverization modulus. the combination tool achieved maximal soil loosening as evidenced by the low soil bulk density range of 1.15 gcm-3 as opposed to the normal 1.4 gcm-3 encountered in conventional tools operated field. the use of combined tillage for seedbed preparation can save 44 to 55% in cost and 50 to 55% in time (kailappan et al., 2001). (maheshwari et al., 2005) reported that soil aggregates of size 12 to 14 mm in the final seedbed were acceptable for sowing crops. clod formation after plowing or disking is a significant issue. clods block the penetration of seed drill furrow openers and prevent close contact between pa ge 26 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 25-33, 2024 seeds and soil. to avoid the aforementioned issues clods must be pulverized. because of the time, efficiency, and cost savings, combining implements and minimizing the number of passes is becoming more common (alkhafaji, 2020). in ethiopia, farmers are using the conventional tillage system which disturbs the soil more; increases soil compaction increases erosion capacity by wind and water. the combined tillage used was also imported from other countries which is too expensive. the implementation did not consider small-scale mechanization, affordable power, high land agriculture, costly, and not scale appropriately. the overall dimensions and weight of the tillage tool were heavy, not easily dismantled, or repaired and it was complicated. therefore, the objective of this study was to design and develop a tractor-drawn cultivator with a pulverizing attachment that would be operated within the techno-economic status of smallholder farmers in ethiopia. materials and methods determination of physical properties of soil moisture content of soil before field activities, soil samples were taken from 0 to 20 cm below the soil surface to determine moisture content. three soil samples were randomly taken from test plots and the weight of each sample was determined using an electronic balance. the samples were then maintained in a hot air oven for 24 hours at a temperature of 105°c. the moisture content (dry basis) was calculated using the formula below (javadi & hajiahmad, 2006). mc = (ww-wd)/wd ×100 (1) where, (i) mc = moisture content of the soil, % db; (ii) ww = weight of wet soil, g; and (iii) wd = weight of oven-dry soil, g. determination of bulk density of soil the oven dry mass per unit volume of the soil is known as bulk density. it was determined by taking soil samples from various sites throughout the field with a core sampler. the formula was used to compute bulk density (javadi & hajiahmad, 2006). γd = m/v (2) where, (i) γd = dry bulk density of soil, g/cm3; (ii) m = oven-dry mass of soil, g; (iii) v = volume of the core sampler, cm3. overall structure and descriptions of the implement a cultivator with pulverizing attachment was designed to reduce the number of secondary tillage operations to single pass and enhance timeliness in seedbed preparation. as shown in figure 1, it consists of the following main components; frame with cultivator tynes, pulverizing attachment having pulverizing blades, disc, and threepoint linkage unit. figure 1: cultivator with pulverizing attachment design of major components components of the cultivator with pulverizing attachment were designed, developed, and fabricated based on parameters like functional requirements, operational requirements, and strength. while fabricating the tractor-drawn cultivator with a pulverizing attachment the basic emphasis was the simplicity of fabrication, the use of locally available materials, the minimum cost of fabrication, and the ease of assembling and dismantling for repair were considered. the mechanical design details were also given due attention so that it was given adequate functional rigidity for the design of the machine. the implement was designed having in mind that the frame should have sufficient strength to hold the tines pa ge 27 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 25-33, 2024 and also withstand failure due to developed stresses and bending moment in mind that it will work on heavy soils. the assumptions made in the design of a cultivator with pulverizing attachment were as follows (parmar & gupta, 2001); (i) no draft was included for pulverizing attachment because it is a rotating unit; (ii) the average speed of operations of the tractor in the field was kept at = 3 km/h; (iii) maximum soil resistance was considered as = 0.75 kg/cm2; (iv) co-efficient of friction in unploughed soil was taken= 0.85; (v) a seven-tyne cultivator having a working depth of 30 cm was considered. design of frame the frame members were welded together at the ends since a fixed-type end connection was selected for the frame. euler’s theory for the crippling and buckling load (pcr) under various end conditions is given by the equation below (khurmi & gupta, 2005). pcr = (π2 ei)/( le/r)2 (3) where, (i) e = modulus of elasticity for the mild steel material (e= 210 gpa); (ii) a = cross-section area for a hollow rectangular shape, cm2; (iii) pcr = euler’s critical load, n; (iv) le= effective length of the frame, cm; (v) r = radius of gyration of the cross-section; (vi) i = polar moment of the cross-section. the frame was considered as a beam at the tyne and hitch support portion and all components were loaded vertically so it is considered as a column. assuming, σy = 250 mpa and comparing the critical load of the frame with yield strength, whether the frame is saved or not. from the available data, let us determine the dimension of the frame. the actual length of the frame, l = 160 cm, equivalent length, le = l=160 cm for the frame one end fixed and another end free. le/r = π ×√(e/σy ) = √((210 gpa)/(250 mpa)) = 91.05 the crushing stress is given by σcr = (π2e)/(le/r)2 = (π2×210 gpa)/91.052 = 250 mpa le/r = 91.5, le = 160 cm then r2 = le/91.5 = (160 cm)/91.5 = 1.749 cm, and r2 = 3.06 cm2 radius of gyration r2 =i/a = (bh3)/12hb = h2/12 then h2 = 12×r2 = 12 × 3.06 cm2 = 3 6.71 cm2 h=6.59 cm, say h=6 cm the dimensions of the frame were made from mild steel rectangular pipe which had h = 6 cm and thickness t = 0.4 cm and width = 7.5 cm. the critical load is found as: pcr = (π2ei)/le 2 = (π2×210(h3b)/12)/le 2 = (π2×210×(603×75)/12)/1602 = 1,091.92×108 kn hence, critical stress σcr = pcr/a = pcr/(b×h) = (1,091.92×108 kn)/(45 cm2) = 242.65 mpa comparing the critical stress wit h the yield strength of the material, critical stress is less than the yield strength of the material (σcr ≪ σy). according to euler’s theory of buckling, for slender columns, the critical buckling stress is usually lower than the yield stress. hence, the designed frame was saved from buckling. the frame of the cultivator was made from mild steel (m.s) rectangular pipe shape of 6 × 7.5 sizes (cm) × 0.4 cm thickness crosssection. all components of the cultivator were assembled and fitted on the frame. design of cultivator tines a cultivator was used to open the furrow at the desired depth of operation. the tine-type cutting blade was used as uniform depth of operation or plowing was required. a cutting unit of tine was fitted at one end of the share and the other end was attached with a frame by bolts and nuts for adjusting depth. the thickness, width, and length of the tyne were decided on the assumption given by (sharma & mukesh, 2010). let, b x h = bare cross-section area, cm2 l= length of breast of tine the length of the inclined part of the tyne generally ranges from 10 to 20 cm, and the radius of curvature r < 12 cm. the minimum clearance length of tine (hi) between the land surface and the lower edge of the frame was 20 cm. the height of the tine was calculated as the formula given below. ht = amax + hi + δh (4) = 10+35+15 = 60 cm where, (i) amax = depth of tool, cm; (ii) hi = length of tyne, cm; (iii) δh = length of tine used for fastening with frame, cm. load angle was determined by; tanα = 10/30 = 0.33 (5) α = tan-1 (0.73) = 36.12° now radius of curvature was determined as; rc = (ha-l1 sinα)/cosα (6) = (20-18sin36.12)/cos36.12 = 11.62 cm where, (i) l1 = length of breast of the share; (ii) ha = height of the first curvature point of the tyne from the point of share. the cutting blade of the tine was exposed first to bending due to soil resistance. the soil resistance, fx is horizontal and acts in the axis of symmetry of the tyne. average soil resistance was obtained by the formula given below (kumar et al., 2017) and table 1. fx= d × w × ko (7) = 30×20×0.2 = 120×9.81 = 1,777.2 n where, (i) d = effective working depth of tine (cm); (ii) w = working width of the tine (cm); (iii) ko = specific soil resistance (kg/cm2). table 1: specific soil resistance at a depth of 15 cm s/n soil type specific resistance, kg/cm2 1 light 0.12 2 medium 0.15 3 heavy 0.20 4 very heavy 0.25 source: (dubey, 2003) the draft force exerted on the cutting blade was determined using the following equation; pa ge 28 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 25-33, 2024 d = ko × n × w × d (8) = 0.8×7×20×30 = 3,360 kgf = 32,691.6 n where, (i)d = draft force, kgf; (ii) ko= soil resistance, kg/cm2; (iii) w = width of tine, cm; (iv) d = depth of tine, cm; (v) n = number of tine. finally, the total draft required for operation and each draft of the tines was calculated by, dt = d × fos × g (9) dt = 3,360 ×1.5×9.81 = 49,442.4 n where, (i) dt = total draft, n; (ii) d = draft, n; (iii) fos = factor of safety; (iv) g = force due to gravity, m/s2. di = (draft (n))/(number of rows) (10) = 32,961.6/7 = 4,708.8 n it was assumed that the draft force on the tine was determined and acting at a height of h/3 from the bottom of the tine. the distance of the draft application on furrow opener tine (d) was calculated by the formula given below (sharma & mukesh, 2010). d = h/3 (11) = 60/3 = 20 cm moment arm length = (h-d) = 60-20 = 40 cm (12) the maximum bending moment was determined according to the following formula. bending moment of tyne (bm) = draft × moment arm length (13) = 4,708.8 n×40 = 188,352 n m therefore, the maximum bending moment of tine was determined by multiplying the bending moment of tine with a factor of safety. (take, fos=1.5) maximum bending moment = bending moment × fos (14) mb = 188,352 nm×1.5 = 282,528 nm mild steel was used for the design of tine and assuming bending stress, σb = 56 n/mm2. section modules of each cutting blade (z) were determined as the following formula. z = ( mb)/σb (15) = (282,528 )/56=5.05 cm3 where, (i) mb = maximum bending moment; (ii) σb = bending stress; (iii) z = section modulus of tine; (iv) b = width of mild steel. according to (khurmi & gupta, 2005), the section modulus of tine of a cultivator with pulverizing attachment was calculated by using the formula, z = 1/6 × b3 (16) 5.05 cm3 = 1/6×b3 b3 = 5.05 cm3×6 = 30.27 cm3 b = ∛(30,270.84) = 3.12 cm≈4 cm so, the 4 cm ×4 cm size of the share was developed from a mild steel material. bending stress, σb causing the tine to bend was calculated by the expression(kumar et al., 2017). σb = (6dt (h1+a))/(tb2 ) (17) = (6×49,442.4(35+20))/(0.6×402 ) = 16,995.82 n/cm2 where, (i) dt= draft at the tip of tine, n; (ii) hi = length of tine, cm; (iii) a = effective working depth of tine, cm; (iv) t = thickness of the tine, cm and (v) b = width of tine, cm. torsional stress acting on the tine when turning the openers inside the soil at headland is given by (kumar et al., 2017). τ = (9dt (ww/4))/(2b2 ) (18) = (9×49,442.4(40/4))/(2(40)2) = 4 ,449,816/3200 = 1,390.57 n⁄cm2 where, (i) ww = effective working width of tine, cm, and (ii) b = width of tine, cm then, the reduced stress is calculated as; δ = √(σb2+4τ2) (19) = √((16,995.82)2+(1,3902.57)2) = 17,052.61 n⁄cm2 = 170.53 n/mm2 where, (i) δ = reduced stress, kg/cm2; (ii) σb = bending stress, n/cm2 and (iii) τ = torsional shear stress, n/ cm2. since the calculated stress (170.53 n/ mm2) was less than the allowable stress of mild steel (200 n/mm2), the design was safe and the tine can perform well. figure 2: schematic diagram of cultivator tine pa ge 29 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 25-33, 2024 design of pulverizing roller basic design parameters and specifications pulverizing roller attachment to the cultivator pulverizes the soil to a greater degree. the pulverizing roller consists of discs, a central shaft, and pulverizing members. the pulverizing roller was designed for cutting, mixing, and clod breaking which ultimately pulverizes the soil by impact force. the cutting and clod-breaking action of this unit provides better soil preparation. roller diameter = 45 cm length of roller = 130 cm circumference = π×d (20) = 3.14×45=141.3 cm spacing between pulverizing blades was calculated by the following formula (prem et al., 2016). sb = (π×d)/nb (21) = 143.3/6 = 23.55 cm (taking, nb=6) where, (i) sb = spacing between the pulverizing blade; (ii) nb = number of pulverizing blade depending upon the type of soil and moisture content in the soil, the torque required to break big clods varies between 0.5 nm. therefore, the power required to operate the pulverizer generally ranges between 1-2 hp at an operating speed of 1.1 m/s and 10% slippage (prem et al., 2016). speed = πdn (22) 1.1 = 3.14×45×n n = 1.1/141.3 = 0.008≈1 we know that, p = 2πnt/60 (23) where, (i) p = power required, w; (ii) n = rpm of the roller; (iii) t = torque applied, n-m t = (p×60)/(2π×n) (24) 2 = (2×3.14×1.×t)/4500 t = (2×4500)/(2×3.14×1)=9000/6.28=1433.12 nm when the pulverizer was operated on the ground it was subjected to a sudden load, therefore while designing, the torque was taken as 2 times the rated torque. tmax= 2×t (25) tmax= 2×1433.12=2866.24 nm design of pulverizing blades pulverizing blades were fabricated and fixed into the disc at 900 in such a way that it has continuously come in contact with soil. pulverizing blades were designed as follows. at = pl×tb (parmar & gupta,2001) (26) = 130 cm×0.4 cm=52 cm2 where, (i) at = total area of pulverizing blade striking on soil, cm2;(ii) pl = pulverizing length, cm; (iii) tb = thickness of blade, cm. the maximum soil resistance = total area of pulverizing blade × the specific soil resistance msr = at×ssr (27) = 52 cm2×0.75 kg⁄cm2 =39 kgf where, (i) msr= maximum soil resistance, kgf; (ii) at= total area of the blade, cm2; ssr = specific soil resistance kg/cm2. the maximum bending moment in the blade was calculated by the following formula. mb= sr×rd (28) = 39 kg×22.5 cm=877.5 kg cm where, (i)mb= maximum bending moment, kg cm; (ii) sr= soil resistance, kg⁄cm2 ; (iii)rd= radial distance, cm. the maximum bending stress for mild steel is 700 mpa, calculating actual bending stress as per the following formula (sharma & mukesh, 2010). σb = (mb × y)/i (29) 700 = (877.5×12×w×2)/(2×0.4×w3) w = √(21060/560)=6.14 cm≈7 cm=7 cm where, (i) mb= maximum bending moment; (ii)σb= actual bending stress, n/cm2 (iii) y = distance height, = d/2; (iv) i = moment of inertia,= bt3/12; (v) w = width of pulverizing blades. hence design is safe. mild steel flat with 7 cm width, 130 cm length, and 0.4 cm thick was selected for pulverizing blades. design of the pulverizing roller drive shaft the design of the shaft must involve the determination of the minimum diameter of the shaft material to ensure satisfactory strength and rigidity when the shaft is transmitting power under various operating and loading conditions. the shaft was initially decided to be fabricated from ductile material (mild steel rod). hence, the design was based on ductile material whose strength is controlled by the maximum shear stress. for a shaft having little or no axial loading, the diameter of the shaft was obtained using the asme code equation (39), (asme, 1995) given as: d3 = 16/πss [(kbmb) 2+(kt mt) 2](1/2) (30) where, (i) d = diameter of the shaft, mm; (ii) mt = torsional moment, nm; (iii) mb = bending moment, nm; (iv) kb = combined shock and fatigue factor applied to bending moment; (v) kt = combined shock and fatigue factor applied to torsional moment; (vi) ss = allowable stress, mn/m2. for rotating shafts when the load is suddenly applied with minor shock (khurmi & gupta, 2005), recommended that values of kb = 1.2 to 2.0 and kt = 1.0 to 1.50 be used. furthermore, it was noted that for the shaft without the keyway, the allowable stress (ss) must be 55 mn/m2, and for the shaft, with the keyway, the allowable stress (ss) should not exceed 40 mn/m2. the torsional moment (mt) on the shaft was calculated using eqn. below (ryder, 1989). mt = (p×60)/(2π×n) (31) p = v×f (32) where, (i) p = power required to drive the pulverizer; (ii) n = speed of the shaft, (1 rev/sec.); (iii) v = forward speed (0.833 m/s); (iv) f = force required to drive the machine (2.4 n). figure 6 shows the load distribution on the shaft. the maximum bending moment on the shaft was determined from the following expressions. mb= (mv 2+mh 2 ) (33) pa ge 30 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 25-33, 2024 where, (i) mv = vertical bending momentum, nm; (ii) mh= horizontal bending momentum, nm the reaction of forces on the driven shaft is as shown below for vertical forces the total resultant components of horizontal and vertical bending moments on the shaft can be obtained as follows. mb = (mv 2+mh 2 )1⁄2 =62,655,333.955 nmm mt = (p×60)/(2π×n) = (2×60)/(2×3.14×1)=51,858.25 nmm d3 = 16/(3.14×55) [(2×62,655.33)2+(1.5×51.86)2 ]1/2 d3 = 6,768,148.8/172.7= 39,304 mm d = ∛39,304= 34 mm mt = torsional moment; nm mb = bending moment; nm kb = 2; kt = 1.5, and ss = 55 mn/m2 therefore, the standard size of 50 mm shaft diameter was used. since the actual diameter was greater than the calculated diameter. therefore, the shaft has proper strength for operation. selection of bearing the size of a bearing to be used depends on the size of the shaft required and the available space. in addition, the bearing must have a high enough load rating to provide an acceptable combination of life and reliability. the bearing size was determined using the maximum load acting on it and the desired maximum lifespan. the dynamic equivalent radial load (w) is calculated by the equation given below. w = x.v.wr+y.wa (34) w = 0.4 × wr+ 1.6 × wr= 2 × 20 × wr= 80 kn wr = 40 kn where, (i) wa= dynamic axial load; (ii) wr= dynamic radial load; (iii) x= radial load factor; (iv) y= axial load factor; (v) v = a rotation factor, 1, for all types of bearings when the inner race is rotating. dynamic load rating can be determined using eqn. 35. c= w (l/106 )1/k (35) = 40× (20,000/106 )1/3=11 kn where, (i) l = rating life; (ii) c = basic dynamic load rating; (iii) w = equivalent dynamic load, and k = 3, for ball bearings, = 10/3, for roller bearings. therefore the maximum lifespan value of 20,000 hr was selected to determine the basic dynamic load rating. bearing number 207 with a bore of 35 mm was selected since the minimum shaft diameter has been determined to be 35 mm. hence, two bearings used for the shaft of the cultivator with pulverizing attachment were selected based on the general criteria for bearing selection. determination of weight and power requirements weight of the implement to estimate the loads on every part of the cultivator with pulverizing attachment, it should be necessary to estimate the weight of all parts. accordingly, the weights of the cultivator tyne, frame, pulverizing blade, share (cutting unit of the cultivator), and disc were estimated (itsi-su, 2011). the total weight of the cultivator with a pulverizing attachment including the weights of the frame, pulverizer blade, three-point linkage unit, cultivator tine, and share of the cultivator was 1331.48 n. taking 2% margins for weights of welding, bolts, nuts, etc. finally, the total weight of the cultivator with pulverizing attachment was found 1358.11 n. draft requirement the draft requirement of the tractor-operated cultivator with a pulverizing attachment was estimated using factors related to implement and the type of soil. the specific soil resistance of medium black soil of the area was 0.75 kgcm-2 (parmar & gupta, 2001). total working width of the cultivator = number of tine × tine spacing (36) = 7 × 22.86 cm =160 cm or 1.6 m cross section area of furrows = total working width of the cultivator x depth of furrow (37) = 160 cm × 30 cm = 4800 cm2. the maximum draft required to drive the cultivator is calculated as follows. d = cac×sr×g×fos (38) = 4800 cm2 × 0.75 kg/cm2 × 9.81 m/s2 × 1.5 = 52,974 n where, (i) d = maximum draft; (ii) cac= cross-section area of the furrows; (iii) sr = maximum soil resistance; (iv) g = force due to gravity; (v) fos= factor of safety power requirement the power required to pull the designed implement was estimated as follows. pd = (d×s)/1000=(52,974×0.833)/1000=44.13kw (39) where, (i) pd = power required to drive the implement; (ii) d = draft required to drive a cultivator; (iii) s = speed of operation. the power required to operate the cultivator with pulverizing attachment was calculated as follows; p= (power required to drive the implement)/(coefficient of friction) (40) p= 44.13/0.85=51.82≈52 kw where, (i) p = power required to operate the implement; (ii) d = draft requirement of the implement; (iii) s = forward speed of the tractor. laboratory testing of prototype preliminary performance tests were carried out to obtain actual data on overall implement performance and work capacity. performance evaluation (in terms of soil tilling quality and machine parameters) was determined. the parameters that were taken during the preliminary test were; the mean weight diameter of the soil, wheel slippage, soil inversion, draft, the width of cut, and volume of soil handled. pa ge 31 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 25-33, 2024 figure 3: prototype cultivator with pulverizing attachment after fabrication and during testing mean weight diameter of the soil soil samples were collected randomly from the tilled plots, with three replications, using an auger at 0-30 cm depth after the tillage process to estimate the clod mean weight diameter. at room temperature, the moist soil samples were allowed to air. the air-dried soil samples were sieved with a set of sieves. clod’s mean weight diameter was calculated using the following formula. (41) where, wi=the weight of soil on each special sieve (kg) w = the total weight of experimented soil (kg) di = net diameter of each sieve (mm) wheel slippage the distances the tractor traveled a head at every 10 revolutions under load and no load on the same surface were measured after a mark was created on the tractor drive wheel with colorful tapes. the speed reduction can be calculated as follows. travel reduction = (m2-m1)/m2 ×100 (42) where, m2 = distance covered at every 10 revolutions of a tractor drive wheel with no load (m) m1 =distance covered at every 10 revolutions of tractor drive wheel with load (m). width of cut the width of the tillage implement’s cut was taken by measuring the width of the furrow using a measuring tape every 5 meters along its length. the width of the cut was determined by averaging ten readings. draft the following equation was used to calculate the draft of the implement (asme, 1995). d=fi [a+b(s)+c(s)2 wt] (43) where, d = drawbar pull, n f= a dimensionless soil texture adjustment parameter i = soil factor = (1 for fine, 2 for medium, and 3 for coarse texture soils) a, b & c = machine – specific parameters s = forward speed, km/h t = tillage depth (cm) for major tools w = machine working width, m. soil inversion the number of weeds or residues of the previous crop remaining on the soil surface after the operation was compared to the number before the operation. for counting weeds or stubbles, a square frame with sides of 100 cm was employed. f(%) = (b-a)/b×100 (44) where, f= indicator for soil inversion,% b= number of weeds or crop residue before operation per unit area, a = number of weeds or crop residue exposed on the surface after operation. volume of soil handled as shown in the equation below, the amount of soil handled was estimated by multiplying the field capacity with depth of cut. v(m3/h) = 10000efc dc (45) where, v = volume of soil handled (m3 /ha) efc = effective field capacity (ha/h) dc = depth of cut (m) results and discussion physical properties of soil moisture content of soil the mean data on soil moisture content before tillage operations at 0-20 cm depth was recorded and presented in table 2. the moisture content of the soil varied from 13.9 to 14.54% with an average value of 14.14%. table 2: soil moisture content (dry basis) sample no. mass of wet soil (gm) mass of dry soil (gm) soil moisture content (%) average (%) 1 256.68 220.8 13.98 14.142 265.26 226.69 14.54 3 244.52 210.5 13.91 pa ge 32 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 25-33, 2024 bulk density of soil values of bulk density before tillage operations at 0-20 cm depth were recorded and presented in table 3. the table 3: bulk density (gm/cc) sample no. mass of wet soil (gm) mass of dry soil (gm) the volume of the core sampler (cc) bulk density (gm/ cc) average 1 256.68 220.8 147.2 1.5 1.49 2 265.26 226.69 147.2 1.54 3 244.52 210.5 147.2 1.43 bulk density of the soil was found to be in the range of 1.43 g/cm3 to 1.54 g/cm3 with an average value of 1.49 g/cm3. laboratory testing results mean weight diameter of the soil the range of the clod means weight diameter was observed between 10.21 to 11.77 mm. maximum clod means weight diameter of was at a tractor forward speed of 2.5 km/hr. the minimum value of clod means weight diameter was at a tractor forward speed of 3.5 km/hr. it is seen from the result that a smaller clod percentage was produced as the tractor’s forward speed increased. the bigger clod percentage decreased and the smaller clod percentage increased. the increase in speed that causes a decrease in bigger clod percentage might be a result of the vibrating effect of the implement associated with an increase in the operating speed of the tractor. more clods break into smaller ones as the implement with higher speeds throws the soil. the improved soil aggregation was attainable at higher tractor forward speed. wheel slippage maximum wheel slippage (12.85%) was obtained at 3.5 km/hr tractor forward speed. minimum wheel slippage (13.44%) was obtained at 2.5 km/hr tractor forward speed. wheel slippage increased negatively as the tractor’s forward speed increased for the tested implement. the negative increase in wheel slippage might be a result of the rolling resistance of the tractor tires and the rolling action of the implement that oppose the rotational movement for the increased tractor forward speed. a negative increase in slippage percentage was observed with an increase in tractor forward speed. width of cut the width of cuts varied between 140.00 to 143.83 cm. the width of cuts was found maximum (143.83 cm) at 3.5 km/hr tractor forward speed while the minimum width of cuts (140.00) was obtained at 2.5 km/hr tractor forward speed. the width of cuts increased with the tractor’s forward speed. the reason might be due to the high soil thrown by the implement at a higher tractor forward speed than a lower tractor forward speed. draft the draft of the implement ranged between 8.19 to 8.59 kn. the maximum draft was obtained at 3.5 km/hr tractor forward speed. the minimum draft was observed at 2.5 km/hr tractor forward speed. the draft requirement of the implement increased as the tractor’s forward speed increased. the reason might be due to higher force requirement at a higher speed than lower speed. soil inversion the maximum soil inversion (89.17%) was found at 3.5 km/hr tractor forward speed. minimum soil inversion (82.41%) was found at 2.5 km/hr tractor forward speed. soil inversion increased as the tractor’s forward speed increased. there was an increase in the percentage of soil inverted as the tractor’s forward speed increased. it might be a result of the implement’s ability to turn and throw the soil when at a higher speed than at a lower speed. volume of soil handled the amount of soil handled varied from 671.07 to 672.09 m3/ha. the maximum volume of soil handled was at 3.5 km/hr tractor forward speed. the minimum volume of soil handled was observed at 2.5 km/hr tractor forward speed. the volume of soil handled increased as the tractor’s forward speed increased. this might be due to the increment of effective field capacity as the tractor’s forward speed increased. conclusion a cultivator with pulverizing attachment was designed and developed to reduce secondary tillage operations to a single pass to ensure timeliness in seedbed preparation. the cultivator tines were fitted on the frame and at the back side of the frame pulverizing, the roller was attached with help of bearings. the pulverizing blades were welded on the discs, in such a way that was crushed the soil by impact force. in this, the force exerted by the pulverizer on the clod is distributed uniformly in the shaft which is used to hold the pulverizing roller. the roller was attached to a frame that holds cultivator tines and a three-point linkage unit. the pulverizer was designed to break big clods formed during tillage operation. the developed implement was attached to the tractor by three point hitch. cultivator tines open furrow and operated roller cut and pulverize soil at optimum condition for tillage. the preliminary test of the cultivator with pulverizing attachment in terms of the width of cuts, draft pa ge 33 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 25-33, 2024 requirement, wheel slippage, soil inversion, the volume of soil handled, and clod mean weight diameter was evaluated. maximum clod means weight diameter of was observed at a tractor forward speed of 2.5 km/ hr. the minimum value of clod means weight diameter was observed at a tractor forward speed of 3.5 km/hr. maximum wheel slippage (12.85%) was obtained at 3.5 km/hr tractor forward speed. minimum wheel slippage (13.44%) was obtained at 2.5 km/hr tractor forward speed. the width of cuts was found maximum (143.83 cm) at 3.5 km/hr tractor forward speed while the minimum width of cuts (140.00) was obtained at 2.5 km/hr tractor forward speed. the maximum draft was obtained at 3.5 km/hr tractor forward speed while the minimum draft was observed at 2.5 km/hr tractor forward speed. the maximum soil inversion (89.17%) was found at 3.5 km/hr tractor forward speed while minimum soil inversion (82.41%) was found at 2.5 km/hr tractor forward speed. the amount of soil handled varied from 671.07 to 672.09 m3/ha. the maximum volume of soil handled was at 3.5 km/hr tractor forward speed. the minimum volume of soil handled was observed at 2.5 km/hr tractor forward speed. references alkhafaji, a. j. (2020). designing and testing of triple combination tillage implement. plant arch, 20(1), 2363-2366. al-shamiry, f. m. s., al-qarni, a. m. and munassar, a. s. (2020). effect of tillage depth and tractor forward speed on some technical indicators of the moldboard plow. international journal of progressive sciences and technologies, 23(2), 28-37. askari, m. & khalifahamzehghasem, s. (2013). draft force inputs for primary and secondary tillage implements in clay loam soil. world applied sciences journal, 21(12), 1789-1794. asme. (1995). design of transition shafting. american society of mechanical engineering, new york, usa. digman, m. (2012). saving time and fuel during tillage. 2011 wisconsin crop production association distinguished service awards, 95. dubey, a. k. (2003). design of sowing machine. teaching material for training on computer-aided design and design methodology for agricultural machinery. technical bulletin no ciae/amd/2003/296, 92-102. itsi-su, (2011). mass, volume, and density of regularly and irregularly shaped objects. internet: <http:// www.pdesas.org/module/content/resources/14756/ view.ashx>. javadi, a. r. z. h. a. n. g. & hajiahmad, a. (2006). effect of a new combined implement for reducing secondary tillage operation. international journal of agriculture and biology, 8(6), 724-727. kailappan, r., vijayaraghavan, n. c., manian, r., duraisamy, g. & amuthan, g. (2001). combination tillage tool-i design and development of a combination tillage tool. agricultural mechanization in asia africa and latin america, 32(3), 19-22. khurmi, r. s. & gupta, j. k. (2005). theory of machines. eurasia publishing house ltd. new delhi, 8th edition. krutz, g. and thompson, l. (1984). design of agricultural machinery (no. 631.3 k7). kumar, a. a., kumar, a. a., vidhyadhar, v., mohan, k., suresh, c., rao, a. s. & ramana, m. v. (2017). design and development of groundnut planter for power weeder. ama, agricultural mechanization in asia, africa, and latin america, 48(3), 25-30. maheshwari, t. k., thakur, t. c. & varshney, b. p. (2005). spiked clod crusher and planker performance under different soil conditions. agricultural engineering today, 29(3and4), 6-11. manian, r. & kathirvel, k. (2001). development and evaluation of an active-passive tillage machine. agricultural mechanization in asia africa and latin america, 32(1), 9-18. parmar, r. p. & gupta, r. a. (2016). development and performance evaluation of a powered pulverizing unit with a cultivator. agricultural engineering today, 40(2), 3-9. prem, m., swarnkar, r., kantilal, v. d. k., jeetsinh, p. s. k. & chitharbhai, k. b. (2016). combined tillage tools-a review. current agriculture research journal, 4(2), 179. reicosky, d. c. & allmaras, r. r. (2003). advances in tillage research in north american cropping systems. journal of crop production, 8(1-2), 75-125. ryder g. h. (1989). strength of materials, 3rd ed. macmillan, nigeria. sharma, d. n., & mukesh, s. (2010). farm machinery design: principal and problems. jain brothers. new delhi. 196-199. zhou, h., zhang, c., zhang, w., yang, q., li, d., liu, z. & xia, j. (2020). evaluation of straw spatial distribution after straw incorporation into the soil for different tillage tools. soil and tillage research, 196, 104440. pa ge 1 pa ge 43 american journal of smart technology and solutions (ajsts) optimising customer service delivery and response time through ai-enhanced chatbots in facilities management-a mixed-methods research mai alhammadi1* volume 2 issue 2, year 2023 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v2i2.2206 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: october 17, 2023 accepted: november 19, 2023 published: november 22, 2023 the present study aimed to assess the effect of ai-enhanced chatbots that optimize customer service delivery and response times on facility management. it utilised the technology acceptance model (tam) and social response theory (srt) for this purpose. the research adopted a mixed-methods methodology aimed to explore the multiple perspectives of 10 facility managers and facility service providers affiliated with facilities management departments in the uae, qatar and saudi arabia regarding the benefits and challenges of ai-enhanced chatbots. this research used correlation analysis and regression to examine the relationships between variables. correlation analysis, using spss 24.0, showed strong positive correlations between five ai-enhanced chatbot factors: perceived usefulness (pu), perceived ease of use (peou), behavioural intention to use (biou), responsiveness (rp), and user satisfaction (us) (pearson correlation coefficient>0.7). regression analysis indicated a significant impact of all these variables on facilities management (p<0.05). the study found that ai-enhanced chatbots in facilities management improve communication, responsiveness, and operational efficiency. they automate workflows, handle manual tasks, predict failures, and respond to customer queries. however, challenges include technical issues, limited human-human interaction, system quality and security, and user adoption. chatbots deliver productivity gains and are used for automated reporting, identifying hazards, conducting briefings, managing meetings, providing training, supporting teamwork, ensuring well-being, and enhancing customer service. keywords artificial intelligence, facilities management, chatbot, natural language processing, service delivery, response time 1 meem 48 engineering consultancy, abu dhabi, uae * corresponding author’s e-mail: eng.maialhammadi@hotmail.com introduction text-based conversational systems or conversational artificial intelligence (ai) referred to commonly as chatbots are the designed software systems for human interaction using natural language processing (nlp) (gnewuch et al., 2022; lin, 2023). chatbots are categorised based on two objectives including task oriented and non-task oriented; task-oriented chatbots are those compatible highly with the information retrieval requirement for effective decision making. thereby, chatbots have gained widespread attention in various industries like finance, e-commerce and healthcare due to growing demand for convenient and efficient customer service (gnewuch et al., 2022). notwithstanding, using live chat interfaces to communicate with customers in e-commerce settings improves real-time customer service to obtain information for product details or assistance in solving technical problems. chatbots have thus enhanced the two-way communication significantly affecting customer satisfaction, trust, word-of-mouth (wom) intentions and repurchase (adam et al., 2021). over 100,000 chatbots have already been created as of 2017 on facebook messenger only for customer service delivery through instant messaging apps and social media (meyer-waarden et al., 2020). facilities management or facility management has been called multiple things, including asset management, business infrastructure management and invisible service for building. it has evolved by merging as a business support service and building maintenance management business (atkin & bildsten, 2017). interoperability capabilities of building information modeling (bim) are effective in the application of facility management, construction and building maintenance stages referring to technology-based solutions for improving inter-organisational productivity and collaboration (ghaffarianhoseini et al., 2017). a chatbot is developed as a friendly user interface to improve the user experience and efficiency in facility management, integrating bim, nlp and ontological techniques to generate immediate responses (lin, 2023). a delayed response time negatively impacts usage intentions and the social presence of users, affecting customer service in facility management (gnewuch et al., 2022). nonetheless, it is worth noting here that a key challenge faced in designing conversational user interfaces is to make sure that the conversations feel human-like and natural. thus, to increase perceived humanness, chatbots may use response delays; however, this can affect user satisfaction, especially in situations where fast response times are expected, i.e., customer service. service delivery and system response time are correlated, being critical factors for productivity and user satisfaction among chatbots. for example, when the response time is slow in customer service live chats, it creates negative website quality perceptions among users (gnewuch et al., 2018). additionally, the chatbot provides improved functionality in real-time scenarios, emphasising its usefulness within pa ge 44 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 an organisation dealing with relevant challenges, including complex business domains, cost factors and limited responsiveness, etc. (majumder & mondal, 2021). although previous research has widely discussed aienhanced chatbots in various fields, implementing them in facilities management for optimising service delivery and response time has been discussed rarely. it has been noted that chatbots in facility information delivery solution promises three benefits, including handling large amounts of complex data, containing tedious information and having high mobility, reducing time to solve users’ query with intuitive user interfaces (chen & tsai, 2021). however, with a great many benefits, there might be a few challenges associated with ai-enhanced chatbots as well. therefore, the current research opted for a mixed-methods approach to examine the impact of ai-enhanced chatbots optimising customer service delivery and response time on facility management using the technology acceptance model (tam) and social response theory (srt). it is also aimed at exploring the in-depth perspectives of personnel linked with the facilities management department in the uae, qatar and saudi arabia on the benefits and challenges of aienhanced chatbots optimising customer service delivery and response time. literature review ai-enhanced chatbots the first chatbot ‘eliza’ was developed in the 1960s; however, broader organisational interest was not gained until the 2010s. on facebook alone, 300,000 chatbots were developed at the beginning of 2016, having common applications in e-commerce, customer service, workplace employee support and healthcare (gnewuch et al., 2022). it is estimated that conversational agents can reduce the costs of current global business to around $1.3 trillion by solving 265 billion queries of customers per year by reducing their response times, freeing up the human workload for different work and dealing with the 80% of routine questions (adam et al., 2021). siri was developed in 2010 by apple, which makes conversations and resolves inquiries using voice commands through messengers integrating with video, audio, and image files. later, in 2011, a watson names chatbot was developed by ibm and google now was developed in 2012 by google. cortana, the microsoft-designed personal assistant and alexa, a human-automation chatbot, were designed in 2014. notably, chatbots are categorised into various categories; knowledge domain, service provided, response generation method, human-aid, communication channel and permissions (adamopoulou & moussiades, 2020). siri and alexa are task-based dialog agents which are known for creating short conversations, including making phone calls, describing routes, etc. however, conversational ai-based chatbots are non-task-oriented dialog systems used in customer service for various purposes. these are focused on imitating conversations like humans focused on certain tasks. xioaice is developed as a non-task-oriented agent by microsoft pecking, which is like a friend (akhtar et al., 2019). additionally, ontbot was developed using nlp to ease user interactions, providing support that can process e-commerce queries. ask diana is a chatbot known for providing information relevant to disaster-related information delivery in facilities management (chen & tsai, 2021). besides, some of the facebook messenger-based chatbots that generate a response to users by interacting with them include dbpedia, sogo, arts-bot, shihbot, cisec, e-commerce website chatbot, nombot, and calmsystem (maroengsit et al., 2019). ai-enhanced chatbots in facilities management ai components, including “pattern recognition” and “machine learning,” integrate a potential value in the ai-enhanced alternative workflow for humans. the continuous advances of smart digital tools are effectively operating in improving customer services and solving problems (burry, 2022). furthermore, the access to ai-enhanced chatbots anywhere and anytime with the integration of cloud-bim and augmented reality (ar) offers extreme assistance for facilitating decision-making. it provides support to facility managers contributing towards customer service (su et al., 2021). similar to humans, chatbots offer customer service, integrating relationship management with consumers. these include relational-oriented behaviours and functional-oriented behaviours assisting consumers in buying decisions. despite the fact that chatbot has no emotions, which is considered its dark side, it has attained multiple benefits in faster service delivery, satisfying immediate customer needs. leveraging the fact that computers are social actors, traits of chatbots are considered trustworthy, reducing the spread of negative wom (su et al., 2021). ai chatbots are used in organisations to help staff members access business information and documents online, offer translation services, gather data from various sources, and format gathered data to adhere to organisational guidelines. by facilitating easy access to, discovery of, and management of work resources, ai chatbots are said to enhance employees’ experiences (gkinko & elbanna, 2023). in addition to being used increasingly often in working settings to help employees, ai chatbot systems are utilised to support customers in a range of industrial sectors (including healthcare, banking, retail, and education). ai chatbots have shifted from emphasising the perspective of employees to that of either designers or customers (gkinko & elbanna, 2023). the validity of compliance and persuasion strategies in technology-based self-service contexts is being debated as chatbots replace human customer care representatives. conversational agents provide 24-hour electronic channels for consumers, offering quick, easy, and affordable communication. however, the nature and caliber of these exchanges vary significantly. for example, consumers use more profanity and speak for longer periods, which could affect their cooperation in pa ge 45 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 response to chatbot suggestions and requests. therefore, understanding the differences between humans and chatbots is crucial for effective customer service (adam et al., 2021). ai in customer service delivery and response time in facilities management according to gkinko and elbanna (2023), a major challenge of balancing service quality and service efficacy is faced by customer service providers. therefore, the potential benefits of chatbots are considered significant for customer self-service, including reduced costs, time efficiency and enhanced customer experience. it impacts improving provider-customer encounters and service quality by being a cost-saving as well as time-saving approach (gkinko & elbanna, 2023). using ai-based chatbots is all about easing human life by knowing information or news to make recommendations, suggestions, shopping services, etc. ai-based communication agents support facility management in a wide range of services that improve customer service, including providing customer support, scheduling meetings, giving financial assistance, suggesting policies, advising insurance policies, offering administration-based services, etc. (nirala et al., 2022). big data simplifies the role of crm staff by providing them with advanced insights into client behaviour patterns, enabling them to manage each customer effectively. this data also allows for better engagement across channels, enabling manufacturers to assess customer reactions to new products on social networks and media. this allows them to pinpoint the optimal crm approach for each client, ultimately resulting in cost-effectiveness for all crm actions (anshari et al., 2019). facility/asset owners and operators have accumulated enormous amounts of data over the years but frequently lack the tools to utilise them fully. humans have a very limited ability to interpret the given data, which is where ai comes into play since it can provide top management with well-reasoned, well-supported advice on which to base a decision. it is crucial to understand that intermediate managers, which include facility and asset managers, might be disregarded in this situation since mission-critical choices are made much more quickly than normally (atkin & bildsten, 2017). in the research by chen & tsai (2021), to implement the created information distribution strategy, a chatbot based on line, an instant messaging service with the greatest market share in taiwan, was prototyped. the line chatbot offers customers two primary interfaces via which they can get or utilise rules to query the facility management data. the efficiency is maximum, and speed is almost twice as quick when the participant chooses an item from the chatbot’s clickable menu to get information. additionally, a user’s performance was the same while utilising the chatbot and the facility management platform to input natural language to get information. theorisation of constructs and hypothesis development technology acceptance model (tam) the utilisation of emerging technologies, such as ai and service robots, and their acceptability by users are predicted using tam. the service robot acceptance model (sram), which incorporates relational and socialemotional components, adjusts and improves the tam. tam attempts to study how external factors affect a person’s internal beliefs, attitudes, and intentions by drawing on the theory of reasoned action. tam analyses two crucial factors-the perceived usefulness (pu) and the perceived ease of use (peou)-to study the behaviours associated with technology adoption (meyerwaarden et al., 2020). these two aspects are related to the motivational factors which create an influence on behavioural intentions and user satisfaction. pu reflects upon the beliefs of users about their experiences of using technology, whereas peou is based on the perceived system quality of chatbot for the user with limited response time and easy availability of chatbot systems. the motive is to provide reliable information for user support needs, increasing levels of trust and satisfaction (nguyen et al., 2021). pu is known as the degree to which it is believed that a particular system would improve an individual’s job performance. peou refers to a person who perceives that using a specific system would be free of effort. therefore, ai-enhanced chatbots are perceived as easy to use by users for acquiring quick knowledge and system-wide optimal solutions free from human fatigue and error. chatbots enhance service delivery within four dimensions, including reliability, empathy, responsiveness and tangibles. it is the distinction that increases the intention to reuse the chatbot (meyer-waarden et al., 2020). tam allows for meeting the requirements of social influence, complexity and ease of use, which affects the user’s choice of technology selection (humairoh & susilo, 2023). social response theory (srt) a set of social cues are posited from computers in social response theory (srt), such as using natural language, interacting with others, triggering mindless responses from humans, and playing social roles irrespective of whether the cues are rudimentary or not. thereby, it is noteworthy here that a chatbot’s response time may trigger social responses that are shaped by the social expectations of users. the persuasiveness of the chatbot’s messages is influenced by response time (gnewuch et al., 2018). in digital contexts, anthropomorphism is the attribution of human-like behaviours, characteristics and emotions to non-human agents (adam et al., 2021). chatbots mainly interact with customers through messaging-based interfaces in a real-time dialogue via dynamic and physical representations. however, some believe that due to the immediate response of chatbots, they may appear as non-human, so a little delay in dynamic responses may pa ge 46 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 increase social presence, user satisfaction and perceived humanness (gnewuch et al., 2022). it creates an impact on the behavioural intention to use in customer service delivery and dynamic response times (adam et al., 2021). based on the theoretical foundation of tam and srt, the following hypotheses have been formulated to test the effectiveness of ai-enhanced chatbots on facilities management. figure 1 depicts the conceptual framework of the research in which the variables on the left-hand side are independent, i.e., factors of ai-enhanced chatbots optimising customer delivery service and response time tested to examine their impact on facilities management. h1: the impact of the perceived usefulness of aienhanced chatbots optimising customer service delivery and response time is significant on the dependent variable, i.e., facilities management. h2: the impact of perceived ease-of-use of aienhanced chatbots optimising customer service delivery and response time is significant on facilities management. h3: the impact of behavioural intention to use aienhanced chatbots optimising customer service delivery and response time is significant on facilities management. h4: the impact of responsiveness of ai-enhanced chatbots optimising customer service delivery and response time is significant on facilities management. h5: the impact of user satisfaction of ai-enhanced chatbots optimising customer service delivery and response time is significant on facilities management. figure 1: conceptual framework source: author methodology research design a mixed-method approach comprising quantitative data collection through surveys and qualitative interviews was employed in the current research. a pragmatic philosophical approach was opted to support the subjective findings with objective conclusions gathering both qualitative and quantitative data. the purpose of the research was to examine the perceptions of facility service providers who have implemented chatbots to explore their significance in customer service delivery and response time in facilities management. using mixed methods, the research offers empirical and theoretical insights into the practical implementation of chatbots in this industry and gauges their effectiveness. data collection the researcher gathered qualitative and quantitative data, both through primary sources. the quantitative data was collected by distributing a close-ended survey questionnaire among the target population. the items of the questionnaire were adapted from the theorisation of constructs using tam and srt and relevant existing literature. perceived usefulness, perceived ease-of-use, behavioural intention to use, responsiveness and user satisfaction were the selected five constructs with three items each. each item was examined based on a fivepoint likert scale ranging from 0 to 4, in which 0 refers to strongly agree, whereas 4 refers to strongly disagree. the interview questions were centred on specific research objectives to identify the key benefits and challenges of using ai-enhanced chatbots optimising service delivery and response time in facilities management. however, the survey and interview questions were both modified to fit the current and recent research, following the instructions of some experts in facility management to ensure comprehensiveness, consistency and readability. also, convergent validity (ave) and reliability of items were tested. sampling the targeted population of the current research was pa ge 47 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 the personnel working in facilities management. the targeted population was approached to fill out survey questionnaires through linkedin and other social media platforms. therefore, these respondents were sampled through a random sampling approach, and 270 respondents who finished the complete survey have experience working with ai-enhanced chatbots for improved service delivery and response time in facilities management. furthermore, 9 respondents for interviews were sampled using purposive sampling as they were the experts in their field. interview respondents were the facility managers and facility service providers who have implemented chatbots within the middle east. for both interviews and surveys, equal respondents, i.e., (90 each for surveys and 3 each for interviews) were approached from uae, qatar, and saudi arabia, as these could be approached easily by bearing limited costs. analysing sample’s profile figures 2,3,4 and 5 below depict the demographics of the targeted respondents. the majority of the respondents were aged between 30 and 39 years, i.e., 69.3%, whereas 37, i.e., 13.7% respondents were aged between 20 and 29, 26, i.e., 9.6% were within the age group of 40-49 years and only 20, i.e., 7.4% respondents were aged 50 years and above as shown in figure 2. figure 3 shows the gender demographics, such that 158 (58.5%) were male, whereas 102 (37.8%) were females who participated in this research. 10 (3.7%) respondents preferred not to mention their gender. figure 4 below depicts the designation of respondents. it shows that 106 (39.3%) of respondents were facility service providers, followed by 95 (35.2%) by facility managers. 53 (19.6%) were end users, and 9 (2.6%) figure 2: age demographics source: author figure 3: gender demographics source: author figure 4: job designation of respondents source: author figure 5: job experience of respondents source: author were it specialists. the remaining 7 (3.3%) were other respondents, suggesting that these categories cover most respondents. data analysis spss 24.0 was used for carrying out numerical analysis. using this statistical tool, reliability and convergent validity were tested, and correlation, regression and exploratory factor analysis were performed to test the association of independent and dependent variables of the research. besides, thematic analysis was conducted to analyse the interview responses following the stages of coding transcripts, identifying keywords, formulating themes and analysing them. results quantitative analysis this section of the research contains results for the survey questionnaires examined using spss. reliability analysis the reliability of a scale was tested in this research to assess the internal consistency of the variables. in reliability analysis, the value of cronbach’s alpha is pa ge 48 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 obtained, and values are tested to range between 0 and 1, with 0.7 being the lowest accepted value (hajjar, 2018). as shown in table 1 below, the value of cronbach’s alpha for each construct is obtained greater than 0.9, showing high internal consistency of all the statements, showing that these constructs are measuring a similar concept underlying ai-based chatbots in facilities management. table 1: cronbach’s reliability test dimension name number of statements cronbach’s alpha (ą) (standardised) n=270 perceived usefulness (pu) 3 0.931 perceived ease of use (peou) 3 0.960 behavioural intention to use (biou) 3 0.963 responsiveness (rp) 3 0.971 user satisfaction (us) 3 0.989 source: author convergent validity and exploratory factor analysis the measurement model must meet three key conditions to achieve convergent validity: the factor loadings of all the variables must be higher than 0.5, the average variance extracted (ave) must be greater than 0.5, and the composite reliability for each construct must be higher than 0.7 (nguyen et al., 2021). the criterion followed in this research is fornell and larcker’s method, which was used to analyse the ave values for each construct shown in table 2 below. the current measurement model confirmed validity as it met all three conditions for all the latent constructs since the ave value for all constructs, including pu, peou, biou, rp and us, is greater than 0.5, lying within the range of 0.7 and 0.9. factor loadings were examined using the kaiser-meyerolkin (kmo) method. as shown in table 2 below, all the values of the factor loadings are greater than 0.5; therefore, all variables are acceptable, fit and unidimensional in the current research. explained variance (%) depicts that the factors capture a large data portion in variance of data if percentages are higher, as shown in table 2 below. table 2: exploratory factor analysis dimension name items factor loadings kaisermeyerolkin (kmo) values explained variance (%) mean (std. deviation) ave perceived usefulness (pu) ai-enhanced chatbots increase the effectiveness and quality of facility management services. pu1 0.925 0.732 88.077% 11.26 (0.912) 0.849 ai-enhanced chatbots can speed up the response time to service requests in facilities management. pu2 0.866 10.98 (0.971) ai-enhanced chatbots can assist in automating the facilities management process, making it more efficient and beneficial. pu3 0.851 11.35 (0.899) perceived ease of use (peou) i find ai-enhanced chatbots in facilities management easy to use. peou1 0.940 0.770 92.747% 11.27 (0.971) 0.832 i often use ai-enhanced chatbots to navigate and communicate when making service requests or queries about facilities management. peou2 0.906 10.97 (0.952) ai-enhanced chatbots make facility management service’s access and requests easier. peou3 0.937 11.26 (0.912) behavioural intention to use (biou) pa ge 49 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 in the future, i intend to use aienhanced chatbots for facility management service requests and queries. biou1 0.955 0.749 93.113% 10.93 (0.963) 0.871 ai-enhanced chatbots are equally appropriate in these new technology-based self-service situations and are designed to convince consumers to comply with or adapt to a specific request. biou2 0.946 10.98 (0.971) i will use ai-enhanced chatbots because they create the impression of intelligence in a non-human technology agent and make conversations feel more natural in a customer service setting. biou3 0.892 11.26 (0.912) responsiveness (rp) when employing ai-enhanced chatbots for facility management services, i frequently get a response or solution immediately. rp1 0.941 0.756 91.714% 11.27 (0.971) 0.825 in contrast to their rule-based predecessors' somewhat static replies, ai-enhanced chatbots are adaptable and demonstrate empathy when responding to the user's input in facilities management. rp2 0.884 10.97 (0.952) users' perceptions of humanness and social presence are strengthened by ai-enhanced chatbots' responsiveness, which also increases their satisfaction with the chatbot engagement. rp3 0.927 11.27 (0.971) user satisfaction (us) real-time ai-enhanced chatbots have made customer support a two-way conversation, which has a significant effect on customer satisfaction, repurchase intentions, and trust. us1 0.912 0.536 67.946% 10.93 (0.963) 0.707 the response time of an aienhanced chatbot is an important factor that affects user satisfaction and other aspects of perceived system quality. us2 0.893 10.98 (0.971) customers are more satisfied interacting with support chatbots that give dynamically delayed replies than those that send nearinstant responses. us3 0.233 10.98 (0.971) source: author correlation analysis correlation analysis is used test the relationship between multiple variables of the research. when two variables are tested to be correlated their covariance is divided by the standard deviations known as pearson correlation coefficient ‘r’ (kafle, 2019). the value of ‘r’ lie between +1 and -1 such that values greater than 0.7 depict strong correlation. as shown in the table 3 below, all the five pa ge 50 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 factors of ai-enhanced chatbots optimising service delivery and response time including pu (0.979**), peou (0.978**), biou (0.991**), rp (0.989) and up (0.952**) are strongly correlated with facilities management since the values are greater than 0.7. regression analysis the relationship tested between two or more independent and dependent variables of interest is evaluated using regression. the sig value or p-value is tested to determine the impact of independent variables on the dependent variable, which should be less than the threshold value of 0.05 (kafle, 2019). as shown in table 4 below, the sig values of all the independent variables of the research are 0.000, which depicts that the impact of all the five factors of ai-enhanced chatbots optimising service delivery and response time, including pu, peou, biou, rp, up is significant on facilities management. table 3: correlations perceived usefulness (pu) perceived ease of use (peou) behavioural intention to use (biou) responsiveness (rp) user satisfaction (us) facilities management (fm) perceived usefulness (pu) 1 perceived ease of use (peou) .971** 1 behavioural intention to use (biou) .970 .981 1 responsiveness (rp) .945 .961** .991** 1 user satisfaction (us) .939** .985** .945** .931 1 facilities management (fm) .979** .978** .991** .989 .952** 1 ** pearson correlation is significant at p< 0.05 (2-tailed); n=270. source: author table 4: table of coefficients using regression unstandardised coefficients standardised coefficients t sig. b std. error beta (constant) .000 .001 -.121 .904 perceived usefulness (pu) .575 .004 .558 160.069 .000 perceived ease of use (peou) 1.041 .019 1.062 54.758 .000 behavioural intention to use (biou) -1.610 .022 -1.633 -74.050 .000 responsiveness (rp) 1.426 .012 1.490 122.925 .000 user satisfaction (us) -.433 .010 -.462 -44.863 .000 source: author thematic analysis chatbots are increasingly becoming significant for opening important gateways to digital information and services within the domain of facilities management (lin, 2023). thematic analysis was conducted to identify the key benefits, challenges and applications of ai-enhanced chatbots in facilities management from the in-depth perspectives and views of personnel working first-hand with them. these are the conversational agents which are used to gather insights on interactive customer service and collaborative work support systems. however, the challenges of implementation and maintenance might be crucial at the initial stage (følstad et al., 2021). chatbot’s impact on facility management has not been discussed widely. therefore, a few questions were asked from the interviewees of the current research to gather their diverse opinions. benefits in your opinion, what are the key benefits of ai-enhanced chatbots in facilities management? participant 1 stated that “we deploy chatbots to make communication easy with the facilities management system. the benefit is that we are shifting towards messages from audio calls, which pa ge 51 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 are preferred by users very often as response times are limited.” participant 3 stated that “we all know that ai is the game changer in responsiveness being more user-friendly. i think in service delivery, chatbots are making significant improvements in terms of handling daily inquiries and routine tasks, making our team stay focused on the complex ones. so, i believe that the key benefit of ai-based chatbots in building management systems is streamlining of operations appreciated by clients in reduced response times.” participant 7 stated that “the best thing we have done using ai-enhanced chatbots is to cut resources in the era of smart facilities management when we are working on smart buildings and cities. in my experience, the fully automated workflows managed by chatbots excel some manual tasks which save resources in service delivery like predicting and maintaining failures, responding to customer queries and minimising downtime.” challenges how do you think ai-enhanced chatbots pose challenges in facilities management? participant 1 stated that “ai-based chatbots have many benefits, but technical issues and limited human-human interaction pose few challenges since sometimes consumers expect and we also need chatbots to mimic human behaviour.” participant 2 stated that “a natural conversation’s design can increase user satisfaction; therefore, poor system quality might pose a challenge with respect to the security and reliability of chatbot systems. it is a major concern for us to implement compliance and robust security measures to secure the sensitive information of clients and our facilities provided.” participant 4 stated that “sometimes employees and clients both feel resistant to using chatbots for the requirements of customer service delivery due to lack of training on chatbots reliability and user adoption. the key is to know about what are aspects of facility management where chatbots are considered an authentic and reliable source.” practical applications what is the practical implementation of ai-enhanced chatbots in facilities management? participant 2 stated that “implementing a chatbot for facilities management is a viable and cost-effective source, increasing user preference towards messaging due to the sophistication of nlp. the chatbot’s software implementation means there are no significant equipment expenditures or installation expenses. it does not require long to notice productivity gains for building management and convenience gains for building users.” participant 4 stated that “we employ a chatbot-assisted facility management technology to automatically produce daily reports for building contractors by collecting conversations between subcontractors on instant messaging platforms. also, it helps with facility management, identifying hazards, briefings, meetings, training, teamwork, well-being, and customer service.” discussion this study focused mainly on the integration of tam and srt to shed light on the use of chatbots optimising service delivery and response time in facilities management. several key findings derived from the analysis are discussed as follows. first, the relationship between pu of chatbots optimising service delivery and response time and facilities management is supported (h1: the impact of the perceived usefulness (pu) of ai-enhanced chatbots optimising customer service delivery and response time is significant on the dependent variable, i.e., facilities management). tam has been advanced in analysing how new technologies are perceived and received for the outcomes of ease of use, social influence and complexity (humairoh & susilo, 2023; tawafak et al., 2023). consequently, each time a user facilitates automated conversations using a chatbot, the parameters of user satisfaction are increased after getting welltimed, correct and relevant data. it directly influences the perceived usefulness of chatbots in facilities management (humairoh & susilo, 2023; le, 2023). supporting the stated fact, another research claimed that chatbots are easy to use (peou) due to the use of nlp technology, which is becoming useful (pu) for both companies and customers as they are perceived to deploy human resources in other tasks within the business and save time (lubbe & ngoma, 2021; selamat & windasari, 2021). efficient use of technology such as chatbot strengthen overall customer experience decreases the number of complaints, and encourages repurchase intention (chen et al., 2021; lubbe & ngoma, 2021). nonetheless, existing literature studies also highlighted that pu and peou are associated with perceived enjoyment, perceived risk, price consciousness, compatibility, and personal innovativeness (kasilingam, 2020). similarly, the research supported the relationship between peou of chatbots optimising service delivery and response time and facilities management is supported (h2: the impact of the perceived ease-of-use (peou) of ai-enhanced chatbots optimising customer service delivery and response time is significant on the dependent variable, i.e., facilities management). second, the advantages of chatbots are various pa ge 52 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 associated with ease of use, such as cost-effectiveness, availability, customer interaction, personal assistance and automation. nevertheless, reduced flexibility can be a challenge for a few users, which influences their biou. however, the current research opposes the stated fact since it is supported (h3: the impact of behavioural intention to use ai-enhanced chatbots optimising customer service delivery and response time is significant on facilities management). it is justified by su et al. (2021) that chatbots, like humans, provide customer service by integrating relationship management and functional behaviours. despite lacking emotions, they offer faster service delivery, satisfy immediate needs, and reduce negative word-of-mouth (um et al., 2020). consequently, adam et al. (2021) analysed that ai-enhanced chatbots create an impact on the biou in customer service delivery and dynamic response times. conversely, it was examined in the thematic analysis of the current research that technical issues and limited human-human interaction may affect biou since sometimes consumers expect chatbots to mimic human behaviour. in customer service delivery, user satisfaction is critical because if service requests fail to meet a satisfactory response, it can cause crucial damage. therefore, a chatbot is included for better responsiveness and fulfil customer satisfaction for service delivery, optimising response time for users (hwang et al., 2019). similarly, the results of another past research examined that responsiveness and anthropomorphism directly influence customer engagement, service quality and customer satisfaction mediated by ai empathy and psychological safety whereas moderated by ai usability (hui et al., 2023). consequently, the current research supported (h4: the impact of responsiveness of ai-enhanced chatbots optimising customer service delivery and response time is significant on facilities management and h5: the impact of user satisfaction of ai-enhanced chatbots optimising customer service delivery and response time is significant on facilities management). meyer-waarden et al. (2020) analysed that chatbots improve service delivery in four dimensions: reliability, empathy, responsiveness, and tangibles, increasing the intention to reuse them. considerably, the thematic analysis also showed that the majority of the interviewees agreed that ai-based chatbots in building management systems are streamlining operations appreciated by clients in reduced response times. in an online setting, businesses must be courteous when serving their customers and should provide them with an appropriate response. the operational efficiency of chatbot systems may be greatly enhanced by their responsiveness, affecting user satisfaction (yun & park, 2022). nonetheless, consumers desire customised communication despite the fact that it offers several alternatives for mobility and response. their motto is “minimum time, best service.” despite its many advantages, consumers and decision-makers who are unfamiliar with ai’s principles are greatly confused and misinterpreted (khan & iqbal, 2020). similarly, interviewees in the current research claimed that implementing compliance and robust security measures is crucial for protecting client information and facilities, and understanding the aspects of facility management where chatbots are considered authentic and reliable is essential. limitations the findings of the current research are limited to facilities management, which might vary in any other context or industry. there will be a limited generalisability of results to all industries of tam and srt since, with rapid technological developments, the actual behaviour may change. the research limited data collection from respondents within a few countries of the middle east due to limited financial and time constraints. however, despite some of these limitations, the research will be effective for the departments of facilities management to use ai-enhanced chatbots for improving customer experiences and optimising service delivery and response time. service quality will be a critical driver for enhancing customer satisfaction and trust for the facility managers when using chatbots for task management. however, researchers explore the role of user training and education in the efficient adoption of chatbots dealing with the ethical concerns of privacy and data security. the study suggests future research with longitudinal studies, cross-industry research, qualitative research, controlled experiments, and ethical considerations. conclusion ai-based communication agents enhance customer service in facility management by providing support, scheduling meetings, financial assistance, policy suggestions, and administration-based services. delayed response time negatively impacts usage intentions and user social presence, affecting facility management. the study highlights the benefits of ai-enhanced chatbots in facilities management, including improved communication, reduced response times, and automation of routine tasks. these chatbots also contribute to resource savings and predictive maintenance. however, implementation challenges include technical issues, human-like interaction, and concerns about system quality, security, and reliability. the research emphasizes practical applications, such as hazard identification, meetings, teamwork, and customer service. the study supports the integration of theoretical models and practical applications, finding a positive relationship between perceived usefulness, ease of use, behavioural intention, responsiveness, and user satisfaction. references adam, m., wessel, m., & benlian, a. (2021). ai-based chatbots in customer service and their effects on user compliance. electronic markets, 31(2), 427-445. adamopoulou, e., & moussiades, l. (2020). chatbots: history, technology, and applications. machine learning pa ge 53 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 with applications, 2, 100006. akhtar, m., neidhardt, j., & werthner, h. (2019). the potential of chatbots: analysis of chatbot conversations. 2019 ieee 21st conference on business informatics (cbi) anshari, m., almunawar, m. n., lim, s. a., & almudimigh, a. (2019). customer relationship management and big data enabled: personalization & customization of services. applied computing and informatics, 15(2), 94-101. atkin, b., & bildsten, l. (2017). a future for facility management. construction innovation, 17(2), 116-124. burry, m. (2022). a new agenda for ai-based urban design and planning. in artificial intelligence in urban planning and design, 3-20. elsevier. chen, j.-s., le, t.-t.-y., & florence, d. (2021). usability and responsiveness of artificial intelligence chatbot on online customer experience in e-retailing. international journal of retail & distribution management, 49(11), 1512-1531. chen, k.-l., & tsai, m.-h. (2021). conversation-based information delivery method for facility management. sensors, 21(14), 4771. følstad, a., araujo, t., law, e. l.-c., brandtzaeg, p. b., papadopoulos, s., reis, l., baez, m., laban, g., mcallister, p., & ischen, c. (2021). future directions for chatbot research: an interdisciplinary research agenda. computing, 103(12), 2915-2942. ghaffarianhoseini, a., tookey, j., ghaffarianhoseini, a., naismith, n., azhar, s., efimova, o., & raahemifar, k. (2017). building information modelling (bim) uptake: clear benefits, understanding its implementation, risks and challenges. renewable and sustainable energy reviews, 75, 1046-1053. gkinko, l., & elbanna, a. (2023). the appropriation of conversational ai in the workplace: a taxonomy of ai chatbot users. international journal of information management, 69, 102568. gnewuch, u., morana, s., adam, m., & maedche, a. (2018). faster is not always better: understanding the effect of dynamic response delays in human-chatbot interaction. gnewuch, u., morana, s., adam, m. t., & maedche, a. (2022). opposing effects of response time in human–chatbot interaction: the moderating role of prior experience. business & information systems engineering, 64(6), 773-791. hajjar, s. (2018). statistical analysis: internal-consistency reliability and construct validity. international journal of quantitative and qualitative research methods, 6(1), 27-38. hui, z., khan, a. n., chenglong, z., & khan, n. a. (2023). when service quality is enhanced by human–artificial intelligence interaction: an examination of anthropomorphism, responsiveness from the perspectives of employees and customers. international journal of human–computer interaction, 1-16. humairoh, m. a., & susilo, w. h. (2023). integration of tam and utaut-iss model: how customers’ service chatbot drove users’ behavior intentions. asian journal of social science and management technology. hwang, s., kim, b., & lee, k. (2019). a data-driven design framework for customer service chatbot. design, user experience, and usability. design philosophy and theory: 8th international conference, duxu 2019, held as part of the 21st hci international conference, hcii 2019, orlando, fl, usa, july 26–31, 2019, proceedings, part i 21, kafle, s. c. (2019). correlation and regression analysis using spss. management, technology & social sciences, 126. kasilingam, d. l. (2020). understanding the attitude and intention to use smartphone chatbots for shopping. technology in society, 62, 101280. khan, s., & iqbal, m. (2020). ai-powered customer service: does it optimize customer experience? 2020 8th international conference on reliability, infocom technologies and optimization (trends and future directions) (icrito), le, x. c. (2023). inducing ai-powered chatbot use for customer purchase: the role of information value and innovative technology. journal of systems and information technology. lin, w. y. (2023). prototyping a chatbot for site managers using building information modeling (bim) and natural language understanding (nlu) techniques. sensors, 23(6), 2942. lubbe, i., & ngoma, n. (2021). useful chatbot experience provides technological satisfaction: an emerging market perspective. south african journal of information management, 23(1), 1-8. majumder, s., & mondal, a. (2021). are chatbots really useful for human resource management? international journal of speech technology, 1-9. maroengsit, w., piyakulpinyo, t., phonyiam, k., pongnumkul, s., chaovalit, p., & theeramunkong, t. (2019). a survey on evaluation methods for chatbots. proceedings of the 2019 7th international conference on information and education technology, meyer-waarden, l., pavone, g., poocharoentou, t., prayatsup, p., ratinaud, m., tison, a., & torné, s. (2020). how service quality influences customer acceptance and usage of chatbots? smr-journal of service management research, 4(1), 35-51. nguyen, d. m., chiu, y.-t. h., & le, h. d. (2021). determinants of continuance intention towards banks’ chatbot services in vietnam: a necessity for sustainable development. sustainability, 13(14), 7625. nirala, k. k., singh, n. k., & purani, v. s. (2022). a survey on providing customer and public administration based services using ai: chatbot. multimedia tools and applications, 81(16), 22215-22246. selamat, m. a., & windasari, n. a. (2021). chatbot for smes: integrating customer and business owner perspectives. technology in society, 66, 101685. su, t., li, h., & an, y. (2021). a bim and machine pa ge 54 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 43-54, 2023 learning integration framework for automated property valuation. journal of building engineering, 44, 102636. tawafak, r. m., al-rahmi, w. m., almogren, a. s., al adwan, m. n., safori, a., attar, r. w., & habes, m. (2023). analysis of e-learning system use using combined tam and ect factors. sustainability, 15(14), 11100. um, t., kim, t., & chung, n. (2020). how does an intelligence chatbot affect customers compared with self-service technology for sustainable services? sustainability, 12(12), 5119. yun, j., & park, j. (2022). the effects of chatbot service recovery with emotion words on customer satisfaction, repurchase intention, and positive wordof-mouth. frontiers in psychology, 13, 922503. pa ge 1 pa ge 1 american journal of smart technology and solutions (ajsts) investigation into optimal conditions for enzymatic hydrolysis of cassava starch to glucose by amylase from rice olosunde adebisi william1*, onumadu kelechi selina1, antia okon orua1 volume 2 issue 2, year 2023 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v2i2.1763 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: june 18, 2023 accepted: july 23, 2023 published: july 30, 2023 the challenge of finding locally available materials in abundance to meet up with the increase in demand for glucose syrup necessitated this study. enzymatic hydrolysis of cassava starch to glucose using glucose amylase sourced from rice was conducted. a-3 factor with 6 levels factorial viz. substrate concentration (0.5, 1.0, 1.5, 2.0, 2.5 and 3.0% w/v), ph (4, 5, 6, 7, 8 and 9) and temperature (30, 40, 50, 60, 70 and 80 °c) experiment were employed. rice malt was prepared and enzyme activated. starch (substrate), buffer solutions, standard glucose solution and its calibration curve were also prepared. starch was hydrolyzed by α-amylase and tested for presence of reducing sugar using benedict solution. time course of the reaction was studied and enzyme activity determined. it was observed that as reaction time increased (t), amount of glucose produced [p] initially increased but soon recorded infinitesimal increase and later assumed constant. the effects of substrate concentration, ph and temperature were found to be essential on glucose production. statistical analysis on the effect of substrate concentration [s], reaction time and their interactions showed significant impact at probability level of (p) = 0.05. keywords amylase, cassava, glucose, hydrolysis, rice 1 department of agricultural and food engineering, faculty of engineering, university of uyo, uyo, p. m. b. 1017, akwa ibom state, nigeria * corresponding author’s e-mail: williamolosunde@uniuyo.edu.ng introduction unmodified starches have diverse functional properties depending on the source of the crop. several starch products may be made from these unmodified starches which are regarded as primary resources. the native starch has restricted applications. this is because it has high predisposition to high syneresis, retrogradation, risky processing factors such as temperature, ph, etc (omojola et al., 2011). the modification of native starch may go a long way in curbing the limitations. this could be attained through esterification, etherification, enzymatic or acid hydrolysis, cross linking and grafting of starch. starches possess permeable surfaces. cassava starch has smooth surfaces which are difficult to degrade than those of corn starch (franco et al., 1988; jane, 2006). starch structures are composed of two linkages: α-(1-4) and α-(1-6) linkages. hydrolysis of starch involves the process of digestion in which enzyme hydrolysis in the digestion system break down the polymer to individual basic glucose units. various industries extensively use starch hydrolysis in the production of several bio products. many low molecular mass products such as sugar, brewing, spirits and textile are made by some food processing and other industries from starch. starch hydrolysis is presently carried out using acid and enzymatic hydrolyses (adenise et al., 2002; odebunmo and owalude, 2005). milder conditions such as normal pressure, lower temperature (up to 10 0c), and medium ph of 6 to 8 are used for enzymatic hydrolysis (kolusheva and marinova, 2007). enzymatic hydrolysis is considered to have a high reaction rate in terms of its potency to denature detergents, solvents and proteolythic enzymes; and lower reaction medium viscosity at higher temperatures, etc. it is often done using α-amylase which may be got from diverse sources, while β-amylase is rarely employed (eric, 2017). the source, in which the inner part of its chain composed of polysaccharide molecules, is always attacked by bacterial α-amylase enzymes. the destruction of spiral polysaccharide chain which produces 3 to 10 units of sugar is aided by the action of starch amylose which leads to the disappearance of a typical blue colour when stained with iodine (pontoh and low, 1995). for the purpose of hydrolysis of starch to glucose, various grains and cereals like rice, maize, sorghum and wheat could be used as enzyme sources. recent report from the western press has that this simple technology currently being used for making simple sugars from cassava starch (tello et al., 1993). a study by hammond and ayernor (2000) gave maximum yield of sugars when starches obtained from various types of cereal malts were hydrolyzed. many factors such as size of granules, source of starch, crystallinity, starch components extension of association, amylase and amylopectin reaction rates, type of polymorphism (a, b and c), enzyme type, complex of amylose lipid, and conditions of hydrolysis (concentrations, ph and temperature) may contribute to variations in the enzymatic vulnerabilities of starches (hoover and zhou, 2003; li, 2004; tester et al., 2006). simple enzymatic reactions must take into account the factors which may affect the rate of reactions. these factors include ph, temperature, concentration of reactant, enzyme concentration, inhabitation by products, etc (wcb, 2020). enzyme deactivation cannot be over looked on either kinetic studies or reactor engineering. in nigeria, because of high exchange rate of naira currency to dollar it is difficult to meet the importation of certain raw materials such as enzyme (gluco amylase). therefore, there is need to carry out more researches on the use of rice seedlings as source of amylase to pa ge 2 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 1-9, 2023 hydrolyze starch for glucose production (bailey and ollis, 1986; onyenekwe, 2013). the choice of cassava (manihot esculenta crantz) as a substrate source in this study is its abundance in nigeria (adejumo and ola, 2010; balagopalan, 2002; osipina and wheatley, 2005). due to the increase in demand for glucose syrup and high cost of importing them, there is need to fully utilize and diversify cassava crop. more so, cassava and rice are viewed as rich sources of sugar and hydrolytic enzyme (gluco amylase) respectively; may be utilized as raw materials in the production of glucose as well as other industrial productions. methods reagents and equipment the chemicals used were analytical grade from merck, england; bdh, england and sigma, usa. the equipment employed was spectrophotometer. sourcing of cassava and rice paddy cassava (manihot esculenta) roots, tms 30572 tubers were purchased at research institute, umudike, abia state while rice paddy was sourced at afikpo, ebonyi state. preparation of rice malt, enzyme activation and buffer solutions the rice paddy was soaked in water, drained and kept inside a container for 2 days (48 hours). it was spread on shallow bed (ridge) and allowed to germinate in the dark. wet rice malt was harvested, cleaned, sun dried; and ground into high diastatic powder (onyenekwe, 2013). 2g of the powder was suspended in 100 ml of distilled water at 600c for ten (10) minutes to activate enzyme amylase in the powder. the supernatant was discarded leaving the cells in the solution (onyenekwe, 2013). the following buffer solutions were prepared using mixing adjusters and salt solutions coupled with addition of distilled water to make up to 200 ml (analchemresources, 2023): ph 4: 0.1 m potassium hydrogen phthalate (100 ml) + 0.1 m hcl (0.2 ml) ph 5: 0.1 m potassium hydrogen phthalate (100 ml) + 0.1 m naoh (45.2 ml) ph 6: 0.1 m potassium hydrogen phosphate (kh2p04) [100 ml] + 0.1 m naoh (11.2 ml) ph 7: 0.1 m potassium hydrogen phosphate (kh2p04) [100 ml] + 0.1m naoh (58.2 ml) ph 8: 0.1 m tris aminomethane (100 ml) + 0.1 m hcl (58.4 ml) ph 9: 0.1 m tris aminomethane (100 ml) + 0.1 m hcl (11.4 ml) preparation of starch, substrate and standard curve of glucose d concentration cassava roots (30 kg) were peeled, washed in water and grated with a commercial grater. the pulp was screened using 25 mm aperture mesh and later suspended in water. the supernatant was decanted after allowing the pulp to sediment for about 6 hours. the white starch cake was obtained and sun dried for about 72 hours (3 days) (oyewole and obieze, 1995). cassava starch concentrations of 0.5 intervals were prepared up to 3.0% (w/v) using each buffer solution obtained in section 2.3. each starch (substrate) concentration was gelatinized in water bath at 80 °c for 10 minutes (nam, 2023). a stock solution of 0.1% (w/v) was prepared by dissolving 1.0 g of glucose d in 1000 ml of distilled water. the stock solution was then used to prepare glucose concentration of 50 ppm interval up to 300 ppm (rebecca et al., 2016). for each glucose concentration, about 2 ml of dinitrosylic acid (dns) reagent was added and then warmed in water bath at 800c for 10 minutes to develop colour for spectrophotometer reading. a blank solution of distilled water and dns was prepared and used to calibrate spectrophotometer to be used in absorbance readings of glucose d concentration (nam, 2023). the results (spectrophotometer readings) were recorded. a standard glucose curve was then produced. determination of alpha amylase activity: enzyme assay the activity of enzyme (rice amylase from malted rice) was determined according to silva et al. (2008). the rice enzyme was activated by incubating 3 g of the enzyme (ground malted rice) suspended in 10 ml of distilled water at 50 °c for 10 minutes. the supernatant discarded leaving the cells in solution. enzyme solution (6 ml) was mixed with phosphate buffer (4 ml) at ph of 5.0 and 10 ml of starch solution 2% (w/v). the mixture was incubated at 40 °c for 10 minutes. then, the reaction was discontinued after addition of 2 ml of 0.1 m hcl and colour developed by adding 0.5 ml iodine reagent. after cooling to room temperature, the amount of glucose produced was found by measuring the solution absorbance at 540 nm using spectrophotometer. however, 1.0 mg of glucose solution reacting with coloured reagent produced an absorbance of 1.0 under the same condition. one unit of enzyme is referred to as the quantity of enzyme which produced 1.0 mg equivalent of glucose per minute under the assay condition (edu-enzyme, 2018; jasco international, 2019). assessment of glucose production from starch using enzymatic hydrolysis phosphate buffer (0.2 m ph 6.0) was used to disperse 15% w/v starch, with bacterial α -amylase solution (0.2% w/v) (3 ml). exactly 1 ml aliquot sodium azide solution (10% w/v) was incubated at 37 °c for 48 hours in an orbital shaker. the quantity of reducing sugar was found after solids were decanted, and the aliquots of the supernatant removed at 6, 9, 24, 30 and 48 hours. at the expiration of incubation period of 48 hours, the dispersed enzyme was deactivated by the addition of 0.1 n hcl to reach ph of 3.0. this was followed by 15 minutes stirring. the resultant solution was neutralized with 0.1 n naoh and pa ge 3 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 1-9, 2023 centrifuged at 2100 rpm for 20 minutes. distilled water and ethanol were used to wash the hydrolyzed residues through filtration. the residues were dried using a hot air oven at 40 °c (franco and ciacco, 1992). thereafter, benedict test for the presence of reducing sugar was then carried out according to aoac (1998) and geetha (2012) methods. experimental design the experimental design was three (3) factors (substrate concentration, ph and temperature) as variables at 6 levels as adopted by khan (2013). the levels selected for the glucose production were as follows: substrate concentration, sc (0.5, 1.0, 1.5, 2.0, 2.5 and 3.0% w/v), ph (4, 5, 6, 7, 8 and 9) and temperature, tp (30, 40, 50, 60, 70 and 80° c) (torreggiani and bertolo, 2001; rosa and giroux, 2001; nieto et al., 2001; ozen et al., 2002; jain and verma, 2003). the experiment was done in replicate. experimental procedure the optimal values of the variables for the production of glucose were obtained based on the following experiment: effects of ph on concentration of glucose (sugar) produced 10 ml of each of the six (6) ph level (4, 5, 6, 7, 8 and 9) was added separately to six (6) different test tubes containing 0.5% w/v of substrate prepared and gelatinized at 80 °c and cooled. they were positioned in water bath at 30 °c. about 4 ml of activated enzyme was then added to each of the six (6) contents of the test tubes. this was allowed to hydrolyze for 10 minutes. about 2 ml aliquot was used to prepare enzyme assay which was measured using spectrophotometer. exactly 2 ml dns reagent was added to discontinue the reaction and then heated to develop colour. this was cooled in cold water and their various absorbance readings with spectrophotometer at 540 nm were taken, recorded and tabulated. however, effects of ph on concentration of glucose (sugar) produced was evaluated. statistical analysis analyzed was carried out using analysis of variance (anova) at 5% level of probability embedded in statistical package for social scientists [spss] version 20. effect of temperature and substrate concentrations on concentration of glucose (sugar) produced the process described in section 2.8 (a) was carried out at 40, 50, 60, 70 and 80 °c. also, the substrate concentrations of 0.1, 1.5, 2.0, 2.5 and 3.0% (w/v) were used separately. their absorbance’s readings were recorded and tabulated. evaluation of optimal parameters required to produce glucose during hydrolysis the optimal parameters (ph, temperature and substrate concentration) were found based on the optimum concentration of sugar obtained. time course of reaction to produce glucose the optimal parameters were then used to study the time course of the reaction to produce glucose. spectrophotometer reading before and after dilution, and the corresponding mean amount of glucose produced after conversion were noted. these data were used to plot several curves of glucose concentrations produced against reaction times. results standard glucose calibration curve and test for presence of reducing sugar the plot of spectrophotometer reading against glucose concentration is presented in figure 1. figure 1: standard glucose calibration curve. the data for standard glucose calibration curve gave equation 1. y = (-4.0 ×10-11. g4) + (1 ×10-8.g3) + (4 ×10-6. g2) + (0.0004 × g) + 0.0021 (1) where, 1 ppm = 1 mg.l-1 = 0.001 g.l-1, g = glucose concentration (ppm) as independent variable and y = diluted value of spectrophotometer reading (nm) as dependent variable. the plot of spectrophotometer reading (nm) against glucose concentration is seen to be a polynomial function. from figure 1, as the pa ge 4 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 1-9, 2023 spectrophotometer reading increased, the glucose concentration also increased which is an indication of a strong direct relationship with coefficient of determination (r2) of 0.9955. similar trend was reported by sciencell (2019), megazyme (2019) and tunde (2020) with r2 of 0.9962, 0.9990 and 0.9889, respectively. the α-amylase was found to hydrolyze the starch by giving a brick red colouration when applying benedict test. this was an indication of the presence of glucose obtained. similar studies were conducted by ukeassys (2018) and cochran et al. (2008), and they also had the same result from the hydrolysis of starch using amylase enzyme. effect of ph, temperature, and substrate concentrations on glucose concentration produced for 10 minutes of reaction time the effect of ph, temperature and substrate concentrations on concentration of glucose produced was studied for 10 minutes of reaction time. based on the data generated, the plots of glucose concentration against temperature at various ph values and substrate concentrations are presented in figures 2 to 5, while that of glucose concentrate produced against ph values at various temperatures and substrate concentration are shown from figures 6 to 9. figure 2: plot of glucose conc. produced against temperature at various ph and constant substrate concentration of 0.5 %w/v. figure 3: plot of glucose conc. produced against temperature at various ph and constant substrate concentration of 2.0 %w/v. figure 4: plot of glucose conc. produced against temperature at various ph and constant substrate concentration of 2.5 %w/v. pa ge 5 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 1-9, 2023 figure 5:plot of glucose conc. produced against temperature at various ph and constant substrate concentration of 3.0 %w/v. figure 6: plot of glucose conc. produced against ph at various temperatures and constant substrate concentration of 0.5 %w/v. figure 7: plot of glucose conc. produced against ph at various temperatures and constant substrate concentration of 2.0 %w/v. figure 8: plot of glucose conc. produced against ph at various temperatures and constant substrate concentration of 2.5 %w/v. pa ge 6 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 1-9, 2023 from figures 2 to 9, the minimum glucose concentration produced (0.029 g de. l-1) was recorded when ph of 9, temperature of 800c and substrate concentration of 0.5% (w/v) were used, whereas the maximum glucose concentration (0.261 g de. l-1) was obtained, when ph of 5, temperature of 400c and substrate concentration of 0.5% (w/v) were used. generally, lower ph favoured the higher amount of glucose produced while high ph values produced lesser amount of glucose. increases in temperature, decreased the concentration of glucose produced. this implies that too high acidic or alkaline medium does not favour the activities of enzymes. however, different kinds of enzymes have specific range of conditions necessary for their optimal performances. at higher temperature, some enzymes might be denatured as they are made up of protein. hence, ph of 5, temperature of 400c and substrate concentration of 0.5% (w/v) were considered as optimum condtions. these conditions were used in studing the time course of reaction or simply the production of glucose with time. in a study conducted by ukessays (2018), the effects of ph (5 to 9) and temperature (300c to 900c) were conspicuously observed on the enzymatic hydrolysis of starch to glucose. other researchers such as aliasrodinah (2009) and karolina (2015) also reported the effect of ph, temperature, and substrate and enzyme concentrations on the production of glucose. determination of enzyme activity the outcome of enzyme activity is presented in table 1. figure 9: plot of glucose conc. produced against ph at various temperatures and constant substrate concentration of 3.0 %w/v. table 1: activity of α-amylase from rice in the production of glucose from cassava starch. spectrophotometer reading after dilution [absorbance] (nm) amount of glucose produced (g de.l-1) molar conc. of glucose produced (mol de.l-1) enzyme activity 0.349 0.248 0.04468 0.0203 u.l-1 or 0.0203 μ mol.min-1.l from table 1, 0.248 g de.l-1 of glucose was produced when the total volume of enzyme mix in assay of 22 ml was incubated for 10 minutes at 400c and ph of 5. however, the enzyme activity in this study was found to be 0.0203 u.l-1 or 0.0203 μ mol.min-1.l. this means 0.0203 units of enzyme catalyzed the transformation of 1 μ mol of substrate into glucose in 1 minute under standard conditions. the low value of the enzyme activity might be as a result of change in optimal ph of the enzyme. this slows down the enzyme activity. however, high value might cause enzyme to denature (cornish bowden, 1995; jasco international, 2019). production of glucose with time time course of reaction was conducted using 40 0c and ph of 5. plots of several curves of glucose concentrations produced against reaction times is presented in figure 10. in figure 10, initially, the amount of glucose produced at reaction time, t = 0 minute was 0 g de.l-1. after 10 minutes, the amount of glucose produced were 0.150, 0.190, 0.194, 0.207, 0.207 and 0.227 g de.l-1 in the six test tubes containing substrate concentrations of 0.5, 1.0, 1.5, 2.0, 2.5 and 3.0 % (w/v) respectively. the production within this period increased rapidly. test tube i with 0.5 % (w/v) of substrate concentration showed a gradual increase in glucose concentration from 10 minutes to 50 minutes of reaction time. beyond this period, there was a decrease in glucose concentration. test tubes ii and iii with 1.0 and 1.5 % (w/v) of substrate concentrations, also recorded similar increase from 0 to 0.239 and 0.247 g de.l-1 at 50 and 60 minutes, respectively. moreover, test tubes iv and v (contain with substrate concentrations of 2.0 and 2.5 % [w/v]), respectively; and had almost the same trend. a conspicuous increase in the amount of glucose produced was recorded between 0 – 20 minutes. this is, an increase from 0.0 to 0.246 g de.l-1 for both samples. test tube vi (with substrate concentration of 3.0 % [w/v]) recorded an increase in glucose concentration produced from zero to 0.253 g de.l-1 within 40 minutes. further increase in reaction time in both test tubes v and vi did not have any pa ge 7 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 1-9, 2023 observable increase in the amount of glucose produced. generally, increase in substrate concentration yielded an increase the amount of glucose concentration produced. virtually, unobservable increase or constant production might be due to that the fact the enzyme had reached its optimum activity and the substrate had been consumed completely. therefore, maximum glucose concentration (0.253 g de.l-1) produced was found when 2.5% (w/v) substrate concentration at 50 minutes reaction time or 3.0% (w/v) substrate concentration at 40 minutes reaction time was employed. similar trends were observed by cecil (1995) at optimum temperature of 500c, where glucose amylase was used to hydrolyze cassava starch / paddy rice to maltose. in addition, the variations of glucose concentration with time of reaction in this study were in accordance with the works done by many researchers (cochran et al., 2008; tunde, 2020; nor, 2009). effect of substrate concentration, reaction time and substrate concentration-reaction time interaction on glucose produced at constant ph = 5 and temperature = 40 0c the summary of anova showing the effect of substrate concentration, reaction time and substrate concentration reaction time interaction on glucose concentration produced at constant ph = 5 and temperature = 40 0c is presented in table 2. table 2: summary of anova showing the effect of substrate concentration, reaction time and substrate concentration-reaction time interaction on the glucose produced. from table 2, since p-value [0.00] < 0.05 and coefficient of determination (r2) = 1.00, f value is significant, which shows that the substrate concentration and reaction time had greater influence on glucose produced. more so, the resulting interaction between substrate concentration and reaction time had a significant impact on glucose produced. in a study conducted by nor (2009), the effects of liquefaction temperature and saccharification ph on glucose production were very significant while the saccharification temperature and liquefaction ph, on the other hand did not influence the glucose production. the observed trend in the present study implies that these factors should not be rolled out when considering enzymatic hydrolysis of starch to produce glucose because they really influence the amount of glucose production. figure 10: plot of glucose concentration produced against reaction time. table 2: summary of anova showing the effect of substrate concentration, reaction time and substrate concentration-reaction time interaction on the glucose produced source of variation df f p-value @ 5% significant? substrate concentration 6 3365.130 0.000 yes reaction time 5 74679.197 0.000 yes substrate concentration * reaction time 30 199.227 0.000 yes r2 = 1.000 conclusion based on the outcome of the findings, the optimum conditions necessary for reasonable glucose production (0.253 g de.l-1 ) were temperature of 40 0c, ph of 5 and substrate concentration of 2.5% (w/v) for 50 minutes or substrate concentration of 3.0% (w/v) for 40 minutes. alpha-amylase activity was found to be 0.0203 μ mol. min-1.l. analysis of variance (anova) results on the effect of substrate concentration, reaction time and their interaction at 40 0c and ph of 5 on glucose production showed a statistically significant impact since pcal < ptab at probability level (p) = 0.05. furthermore, the use of rice amylase to produce glucose from cassava starch would complement increase in demand for glucose syrup production. references adejumo, b. a. and ola, f. a. (2010). the effect of cassava effluent on the chemical composition of agricultural soil engineering. 2nd edition. academic press inc., new york, 250p. adenise, l.w., saul, n. i., ashock, p. and carlos, s. (2002). pa ge 8 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 1-9, 2023 acid and enzymatic hydrolysis to recover reducing sugars from cassava. baggasse: an economic study. brazil archive of biotechnology ,45(3), 1516-1522. alias-rodianah, b. (2009). the effect of substrate and enzyme concentration to the fructooligosaccharides (fos) production. beng project. faculty of chemical & natural resources engineering, university malaysia pahang, malaysia, 125p. (thesis). analchem-resources (2023). preparation of phbuffersolutions http://delloyd.50megs.com/ mobile/buffersolns.html (online). aoac (association of official and analytical chemists) (1998). official methods of analysis. 13th edition, association of official and analytical chemists, washington d.c. 876p. bailey, j. e. and ollis, d. f. (1986). biochemical engineering fundamentals. 2nd edition. mc graw hill, united kingdom, 928p. balagopalan, c. (2002). cassava utilization in food, feed and industry. in: r. j. hillocks, j. m. tresh and a.c. belloti (editors) cassava: biology, production and utilization. ciat, columbia, 301-318. cochran, b., lunday, d. and miskevich, f. (2008). kinetic analysis of amylase using quantitative benedict’s and iodine starch reagents. journal of chemical education, 85(3), 401404. cornishbowden, a. (2014). fundamental of enzymes kinetics revised edition, elsevier science, london, 173p. edu-enzyme (2018). enzyme activity determination. http:// sites.psu.edu/kmf5429/ lab3 enzyme (retrieved on 10th may, 2018). eric, b. (2017). recent progress. in: ian, j. t.(editor) understanding starch structure. agronomy, 2(7), 56, https://doi.org/10.3390/agronomy7030056. franco, c. m. and ciacco, c.f. (1992). factors that affect the enzymatic degradation of natural starch granules: effect of the size of the granules. starch starke, 44 (11), 422-426. franco, c. m., ciacco, c. f. and tavares, d. q. (1988). studies on the susceptibility of granular cassava and corn starches to enzymatic attack. part 2: study of the granular structure of starch. starch-starke, 40(1), 29-32. geetha, d. k. (2012). practical biochemistry. jaypee brothers medical publishers ltd, new delhi: 4 7. hammond,t. k. and ayernor, g. s. (2000). combination of malted series and cassava starch in the production of sugar syrup. journal of the ghana science association, 2(1), 87-92. hoover, r. and zhou,y. (2003). invitro and invivo hydrolysis of legume starches by alpha amylase and resistance starch for formation in legumes: a review. carbohydrate polymers, 54(4), 401-417. jain, s. k. and verma, r. c. (2003). osmotic dehydration: a new, promising and emerging industry. beverage food world, 30, 30–34. jane, j. (2006). current understanding on starch granule structure. journal of applied glycoscience, 53(3), 205213. jasco international (2019). application note: measurement of enzyme activity alkaline phosphatase activity. http://www.jasco corporation. com (retrieved on august, 2019). karolina, a. b. (2015). kinetic modelling of enzymatic starch hydrolysis. phd thesis, wageningen university, wageningen, 160p. (thesis). khan, s. (2013). saying meritocracy and doing privilege: the sociological quarterly. http://www. wileyonlinelibrary.com (retrieved on 25th august, 2019). kolusheva, t. and marinova, a. (2007). a study of the optimal conditions for starch hydrolysis through thermostable α-amylase. journal of the university of chemical technology and metallurgy, 42(1), 93-96. li, j. h. (2004). starch from hull-less barley: in-vitro susceptibility of waxy normal and high amylase starches towards hydrolysis by alpha amylases and amyloglucosidase. food chemistry, 84(4), 621-634. megazyme (2019). d-glucose assay procedure. www. megazyme.com (retrieved on 19th april 2020). nam, s. w. (2023). starch hydrolysis by amylase: experiment no. 5 https://user.eng.umd.edu/~nsw/ ench485/lab5.htm (retrieved on 7th june 2023). (online). nieto, a. b., salvatori, d. m., castro, m. a. and alzamora, s. m. (2004). structural changes in apple tissue during glucose and sucrose osmotic dehydration: shrinkage, porosity, density and microscopic features. journal of food engineering, 61(2), 269 – 278. nor, s. (2009). effects of ph and temperature on glucose production from tapioca starch using enzymatic hydrolysis: a statistical approach. beng. faculty of chemical and natural resources engineering university malaysia, pahang, malaysia, 24p. (thesis). odebunmi, e. o. and owalude, s. o. (2005). kinetics and mechanism of oxidation of sugars by chromium (vi) in perchloric acid medium. journal chemical society of nigeria, 30(2), 187-191. omojola, m.o., manu, n. i., thomas, s. a. (2011). effect of acid hydrolysis on the physicochemical properties of colar starch. african journal of pure and applied chemistry, 5 (9), 307-315. onyenekwe, p.c. (2013). glucose syrup production. rmrdc proceedings series 025: adoption of improved processing equipment for cassava cluster development in crcri adopted villages, abuja, nigeria, 123p. (proceedings). osipina, b. and wheatley, c. (2005). processing of cassava tubers, meals and chips. http://www.fao.org (retrieved on 24th july 2019). (proceedings). oyewole, o.b. and obieze, n. (1995). processing and characteristics of tapioca meal from cassava. tropical science, 35, 401-404. ozen, b. f., dock, l. l. ozdemir, m. and floros, j. d. (2002). processing factors affecting the osmotic pa ge 9 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 1-9, 2023 dehydration of diced green peppers. international journal of food science and technology, 37, 497–502. ponto, j. and low, n.h. (1995). glucose syrup production from indonesian palm and cassava starch. food research international, 28(4), 379-385. rebecca, v. a., wout, b. and ruben, v. (2016). saccharification protocol for small-scale lignocellulosic biomass samples to test processing of cellulose into glucose. bioprotocol, 6(1), 1 -10 https://www. researchgate.net/publication/308753308 (retrieved 13th march, 2018). (online). rosa, m. d. and giroux, f. (2001) osmotic treatments and problems related to the solution management. journal of food engineering, 49, 223 – 236. sciencell (2019). glucose assay. science cell research laboratories. (book). www.sciencellonline.com (retrieved 11th april 2020). silva, r. n., fabio, p. q., valdirene, n. m., eduardo, r. a., ciencia, t. a. (2008). production of glucose and fructose syrups from cassava (manihot esculanta crantz) starch using enzymes produced by microorganisms isolated from brazilian cerrado soil. the journal of biological chemistry, 30. tello, p.g., rubio, f.c., alameda, e. j. and rodriguez, r.s. (1993). kinetics of the hydrolysis of soluble starch by glucoamylase. international chemical engineering, 33 (3), 450455. tester, r.f., qi, x. and karkalas, j. (2006). hydrolysis of native starches with amylase. animal feed science and technology, 130, 39-54. torreggiani, d. and bertolo, g. (2001). osmotic pretreatments in food processing: chemical, physical and structurals effects. journal of food engineering, 49, 247-253. tunde, a. a. (2020). production of glucose from hydrolysis of potato starch. world scientific news, 145, 128-143. uk essays (2018). starch hydrolysis by amylase: an experiment https://www.ukessays.com/essays/biology/thestarch-hydrolysis-of-amylase (retrieved 11th april 2020). wbc (worthington biochemical corporation) (2020). introduction to enzymes. http://www.worthingtonbiochem.com/introbiochem/enzymeconc.html (retrieved on 29th april 2022). pa ge 1 pa ge 25 american journal of smart technology and solutions (ajsts) comparative analysis of table aliasing in sql queries: functional and semantic implications rabel catayoc1* volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.4332 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: january 02, 2025 accepted: february 05, 2025 published: february 19, 2025 table aliasing is a widely used technique in sql queries that enhances readability, scalability, and maintainability, particularly in complex queries involving multiple tables. this study examines the functional and semantic implications of aliasing, highlighting its role in improving query clarity, reducing redundancy, and facilitating long-term database management. while the benefits of aliasing are evident in large-scale applications, potential limitations exist. in simple queries, aliasing may be unnecessary and could even introduce confusion when non-descriptive or overly abbreviated aliases obscure the meaning of the query. additionally, inconsistent alias usage can lead to readability issues and hinder collaboration among database developers. this paper provides a comprehensive analysis of aliasing best practices, explores its impact on query optimization across different relational database management systems (rdbms), and discusses scenarios where aliasing may be either advantageous or counterproductive. the findings underscore the importance of adopting a strategic approach to aliasing to balance clarity and efficiency in sql query construction. keywords aliasing, database, functional, semantic, sql queries, sql 1 mindanao state university, iligan institute of technology, philippines * corresponding author’s e-mail: rabelcatayoc@gmail.com introduction structured query language (sql) is fundamental in the realm of database management, offering a standardized method for querying and manipulating data across relational database systems. sql’s widespread use is rooted in its ability to facilitate efficient interaction with vast datasets, making it indispensable for data-driven environments in both academic and professional settings. as the size and complexity of databases continue to escalate, optimizing sql queries for both performance and long-term maintainability has become crucial. one such optimization practice—table aliasing—remains an area of ongoing discussion and investigation. table aliases, which serve as shorthand representations for table names within sql queries, are widely adopted to enhance readability, reduce ambiguity, and simplify query construction, especially when dealing with multiple tables. existing academic literature highlights the various advantages of aliasing in query design, particularly in handling complex query structures (elmasri & navathe, 2020). however, the necessity of aliases in simple, singletable queries remains a subject of debate. while queries without aliases may seem simpler at first glance, they can introduce ambiguities that are especially problematic when dealing with joins, subqueries, or tables with overlapping column names. the primary objective of this paper is to conduct a comparative analysis of the functional and semantic implications of sql queries with and without table aliasing. by evaluating how aliasing influences query clarity, maintainability, and performance, this paper aims to offer comprehensive insights into the role of aliases in contemporary sql query construction. ultimately, this analysis seeks to guide database professionals in establishing best practices, particularly in complex, multi-table scenarios where clarity and optimization are paramount. literature review the use of table aliases in sql queries has been widely studied, with numerous scholars addressing the role of aliasing in improving query readability, performance, and long-term maintainability. the literature generally divides the discussion into two major areas: functional implications, which concern the performance and optimization of queries, and semantic implications, which relate to clarity, readability, and interpretability. queries without aliases in the context of straightforward sql queries that involve a single table, omitting aliases can create a cleaner and simpler query structure. the absence of aliases eliminates syntactic overhead, thus making queries more concise and easier to understand for basic data retrieval tasks (casteel & tan, 2003). this approach is often sufficient when dealing with a limited number of columns and tables, enabling a straightforward query construction process. however, as sql queries grow in complexity—particularly when dealing with joins, subqueries, or multiple tables that may share similar column names—omitting aliases introduces notable challenges. ambiguities arise when identical column names are referenced across different tables, which can lead to incorrect results or parsing errors. while queries without aliases may be effective for simple operations, they can become increasingly difficult to maintain and understand as the complexity of the database structure increases (bernstein & melnik, 2007). communitydriven discussions (e.g., zanini, 2024) suggest that while pa ge 26 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 25-29, 2025 performance remains unaffected by alias omission, aliasing is widely regarded as best practice for enhancing query readability, especially in scenarios involving joins and subqueries. queries with aliases the use of table aliases is widely advocated in database management practices, especially in complex queries that involve multiple tables, self-joins, or nested queries. aliasing improves the semantic clarity of sql queries by providing clear, concise references to table names, thus reducing the cognitive load required to parse and understand the query. in scenarios where multiple tables are involved, aliases help prevent ambiguity by ensuring that each column reference is attributed to the correct table (shasha et al., 2002). further, studies (metabase, n.d.) emphasize that table aliasing not only enhances the database engine’s ability to process queries but also aids the query writer in maintaining a clear understanding of the data flow. aliases improve both query parsing and human comprehension by distinctly associating table and column references. additionally, aliases are considered essential for maintaining query maintainability, particularly when queries require modification or updates over time (berg et al., 2010). while aliasing offers significant benefits in terms of clarity, it is crucial that the aliases themselves are meaningful and descriptive. the use of non-descriptive or arbitrary aliases (e.g., single-letter identifiers such as “a,” “b,” or “t”) can obscure the query’s intent and reduce its semantic value. to maximize the benefits of aliasing, it is essential to use clear and descriptive aliases that reflect the role or content of the respective tables. sqlblog. org (2009) highlights the importance of this practice and advises against ambiguous naming conventions. functional implications of table aliasing table aliasing serves several functional purposes in sql query design. by simplifying query structure, improving optimization, and reducing parsing time, aliasing contributes to enhanced performance—especially in complex, multi-table queries. in large-scale databases, aliases help the database engine distinguish between multiple tables and columns, thereby facilitating more efficient execution plans. a key benefit of table aliasing is its ability to prevent column name conflicts when joining tables that share common column names (e.g., id or name). without aliasing, queries that involve these common fields can result in ambiguity or execution errors. aliases ensure that each column is uniquely identified, thus avoiding unintended column matches and enhancing the overall accuracy and reliability of the query (sadalage & fowler, 2012). additionally, aliasing plays a vital role in query optimization by reducing the complexity of reference resolution. by minimizing ambiguity during the query parsing phase, aliasing allows the query execution engine to generate more efficient execution plans, particularly in systems that handle large datasets or complex relational operations (shasha et al., 2002). semantic implications of table aliasing from a semantic perspective, aliasing improves the readability and interpretability of sql queries. in complex database operations—such as those involving joins, subqueries, or nested queries—aliases act as concise shortcuts that make it easier for the reader to follow the query’s intent. in particular, aliasing helps reduce confusion in queries that reference the same table multiple times, such as in self-joins or when working with multiple instances of a similar table (stojanovic et al., 2004). in collaborative environments, where multiple developers or analysts may be involved in writing or maintaining sql queries, clear and meaningful aliases help ensure that everyone involved understands the purpose of each table and column reference. this reduces the likelihood of mistakes, both during the initial query construction and in subsequent modifications (berg et al., 2010). moreover, aliasing mitigates semantic ambiguity that can arise when multiple instances of the same table are referenced. for example, in self-joins or in complex queries that combine similar tables, the absence of aliases may result in referencing the wrong instance of a table or column, leading to incorrect data retrieval or unintended results (buhl et al., 2011). therefore, using aliases effectively enhances the clarity and accuracy of the query. materials and methods this study adopts a comparative analysis approach to assess the functional and semantic implications of table aliasing in sql queries. two sql queries are selected for evaluation—one without aliasing and the other utilizing explicit aliases. the analysis focuses on evaluating these queries based on their readability, maintainability, ambiguity prevention, and performance considerations. query selection the two sql queries used for comparison are designed to retrieve data from the track table, selecting the name and unitprice columns, renaming them as “track name” and “price,” respectively. both queries are ordered alphabetically by the track name, and the results are limited to the first 20 rows. query 1 (without alias) select name as “track name”, unitprice as price from track order by name limit 20; query 2 (with alias) select t.name as “track name”, t.unitprice as price pa ge 27 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 25-29, 2025 from t.track order by t.name limit 20; evaluation criteria the two queries were evaluated based on the following criteria 1. readability and maintainability. evaluating the ease with which the queries can be understood and modified, particularly in large-scale or complex databases. 2. ambiguity prevention. investigating the potential for confusion, particularly when additional tables are introduced that may have similar column names. 3. query optimization. analyzing the performance aspects of aliasing in sql, particularly its effect on query execution time and database engine processing. by systematically comparing these criteria, this paper aims to provide a comprehensive understanding of the functional and semantic implications of table aliasing in sql queries. results and discussion this section presents the results of the comparative analysis between sql queries with and without table aliasing, focusing on both the functional and semantic dimensions of query construction. functional implications of table aliasing query execution and optimization when comparing the performance of queries with and without aliases, it becomes clear that the absence of aliases does not introduce any noticeable performance degradation in the context of simple queries. both the aliased and non-aliased queries performed similarly in terms of execution time, which is to be expected when querying a single table without additional complexity. in cases where queries involve complex joins or multi-table operations, however, the presence of aliases is more than just a syntactic convenience. queries with aliases allow the database management system (dbms) to more easily distinguish between columns from different tables. in multi-table queries, where multiple tables share columns with similar names (e.g., id or name), using aliases prevents conflicts during query execution, thus avoiding potential errors in execution or result interpretation. without aliases, the dbms may misinterpret the intended table-column mapping, which could lead to inaccurate results or inefficient query plans. while this study’s queries involved only one table, aliasing in multi-table operations ensures that the database engine can execute more optimized query plans. in more complex operations, such as self-joins or nested subqueries, aliasing improves the engine’s ability to distinguish between different instances of the same table, allowing for more efficient joins, fewer resources used during execution, and reduced data redundancy (shasha et al., 2002). as such, aliasing enhances the functional performance of sql queries in large-scale or intricate databases. scalability and maintainability another key observation is the impact of table aliasing on query scalability and maintainability. for simple queries, where the structure remains relatively fixed, the choice to use or omit aliases does not drastically affect the overall performance or ease of understanding. however, as sql queries grow in complexity—whether through the addition of joins, subqueries, or other advanced operations—the maintainability of queries without aliases quickly becomes an issue. without aliases, queries become increasingly difficult to interpret and modify, especially for developers unfamiliar with the database schema or the specific query. in contrast, queries using aliases are significantly more readable, as they provide explicit references to the tables involved. this semantic clarity aids in query maintenance, particularly when queries need to be altered or extended. in practice, maintaining large databases with numerous related tables is far simpler when aliasing is used, as it mitigates the risk of column name conflicts and enhances the clarity of the relationships between tables (berg et al., 2010). furthermore, query maintainability is enhanced through the scalability of aliasing. when new tables are added or existing ones are modified, queries that use aliases are easier to update. modifications to a table name or structure are less likely to break the logic of the query if aliases are present, because the alias itself acts as a reference point that isolates the query from direct dependency on the underlying table schema. thus, aliasing proves crucial in environments where databases are continuously evolving. semantic implications of table aliasing readability and clarity from a semantic perspective, the use of aliases significantly enhances the readability of sql queries, particularly as queries grow in complexity. for a query with a single table, the use of aliases may seem redundant, as there is little risk of confusion or ambiguity. however, when queries involve multiple tables, especially when these tables share column names, aliasing plays a crucial role in ensuring that each reference is clearly understood. in the case of the queries analyzed in this study, the use of aliases helped maintain a clear distinction between the column names and their corresponding tables, even though the queries involved only one table. for more intricate queries, such as those involving self-joins or multiple tables, the role of aliases in improving readability becomes indispensable. aliases make it possible for the reader to immediately identify which column belongs to which table, avoiding potential confusion (stojanovic et al., 2004). the semantic clarity provided by aliasing is particularly valuable in collaborative environments where multiple pa ge 28 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 25-29, 2025 developers, analysts, or database administrators might work on the same set of queries. clear and descriptive aliases serve as documentation within the query itself, eliminating ambiguity and enhancing team collaboration. using aliases in this way can also reduce the number of errors that arise from misinterpreting the relationships between different tables, particularly in complex queries where joins are frequent (sadalage & fowler, 2012). moreover, queries with aliases can be more easily understood by non-expert stakeholders, such as business analysts or project managers, who may need to work with the results without fully understanding the technical intricacies of sql. the clarity provided by well-chosen aliases reduces the cognitive load required to comprehend the query’s logic, making sql more accessible to a broader range of users. ambiguity prevention in the context of complex queries, ambiguity is one of the most critical issues that table aliasing helps address. without aliases, queries that involve multiple tables with overlapping column names may introduce confusion, leading to incorrect results or errors during query execution. even in simple queries, without aliases, the meaning of column names can become unclear, particularly when the same column name appears in different contexts or queries. for example, if a query joins two tables—employees and departments—and both contain a column named name, failing to alias the tables could lead to ambiguity in the sql code, making it difficult to discern whether the name refers to an employee’s name or a department’s name. table aliases resolve this issue by providing distinct references to each table, ensuring that each name column is clearly identified as belonging to its respective table. in this study’s queries, aliasing allowed for the immediate identification of the track table’s columns, preventing any potential confusion or ambiguity regarding the column names. in larger-scale queries, where joins and subqueries are common, the presence of aliases prevents the risk of misinterpreting the data relationships and ensures that each column is properly attributed to its source table (elmasri & navathe, 2020). error reduction and debugging the semantic clarity provided by aliases also plays a crucial role in reducing errors and aiding in the debugging process. as sql queries become more complex, it becomes increasingly challenging to identify the source of errors. aliases help isolate specific portions of the query, making it easier to identify and correct mistakes. in scenarios where queries are large and involve numerous tables and joins, aliases make it possible to quickly locate the section of the query where an error may have occurred. in contrast, queries without aliases can lead to errors that are difficult to trace, particularly when multiple tables are involved or when the query logic is altered. the absence of clear references to tables can make it challenging to pinpoint the exact source of the error, resulting in longer troubleshooting times and potential misinterpretation of the results. aliasing reduces these risks by providing clear references and a more transparent structure, making errors easier to identify and resolve (shasha et al., 2002). conclusions while the functional performance of sql queries with and without table aliasing remains largely similar for simple queries, the semantic benefits of aliasing are undeniable. aliases significantly enhance query readability, scalability, and maintainability, particularly as query complexity increases. they help prevent ambiguity, improve error prevention, and provide a clear structure that aids both the development and debugging process. as databases grow and become more complex, the use of table aliasing should be regarded as best practice, not only for ensuring semantic clarity but also for optimizing long-term query performance and ease of maintenance. however, the impact of aliasing may vary across different relational database management systems (rdbms). in mysql and postgresql, aliasing is widely used for improving readability but has no direct effect on performance optimization. in sql server and oracle, aliasing can influence query execution plans when combined with indexing and optimization strategies, particularly in large datasets. additionally, some rdbms implementations enforce stricter aliasing rules, affecting how developers structure queries. understanding these variations is crucial for database professionals to maximize the benefits of aliasing while mitigating potential drawbacks. although aliasing may seem superfluous in small-scale queries, it plays an essential role in larger, more complex queries and should be embraced as a standard practice in sql query construction. future research could explore the impact of aliasing on performance in multi-table queries with various database architectures, as well as its role in query optimization across different rdbms environments. references berg, m., miller, r., & shasha, d. (2010). database query optimization: challenges and techniques. journal of database management, 21(4), 1–18. berg, p., luda, a., & thompson, m. (2010). query optimization and indexing strategies in sql. journal of database management, 15(3), 45–67. bernstein, p. a., & melnik, s. (2007). database systems: a practical approach to design, implementation, and management. addison-wesley. buhl, h., white, j., & chopra, m. (2011). understanding the role of aliasing in complex sql queries. data management journal, 22(4), 23–35. casteel, r., & tan, b. (2003). sql design and performance considerations for large-scale applications. international pa ge 29 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 25-29, 2025 journal of database management, 15(3), 45–61. chaudhuri, s., & dayal, u. (2006). query optimization in relational databases. acm computing surveys, 31(2), 170–233. coronel, c., & morris, s. (2018). database systems: design, implementation, & management (13th ed.). cengage learning. date, c. j. (2019). an introduction to database systems (8th ed.). pearson education. elmasri, r., & navathe, s. b. (2020). fundamentals of database systems (7th ed.). pearson. metabase. (n.d.). best practices for writing sql queries. retrieved from https://www.metabase.com/learn/ grow-your-data-skills/learn-sql/working-with-sql/ sql-best-practices oppel, a. j. (2010). sql: the complete reference (3rd ed.). mcgraw-hill. sadalage, c., & fowler, m. (2012). nosql distilled: a brief guide to the emerging world of polyglot persistence. addisonwesley. shasha, d., & boncz, p. (2002). sql query optimization: analyzing the benefits of aliasing. international journal of computer science, 28(5), 512–528. shasha, d., & bonner, p. (2002). optimizing sql queries through aliasing: a comprehensive guide. acm transactions on database systems, 27(5), 223–235. silberschatz, a., korth, h. f., & sudarshan, s. (2019). database system concepts (7th ed.). mcgraw-hill. sqlblog.org. (2009). the importance of meaningful sql aliases. retrieved from https://www.sqlblog.org/aliasbest-practices stojanovic, j., bojan, m., & bogdanovic, b. (2004). semantic query optimization: a review of existing techniques and future directions. journal of database research, 19(3), 79–102. zanini, a. (2024). sql alias: everything you need to know about as in sql. retrieved from https://www.dbvis. com/thetable/sql-alias-everything-you-need-toknow-about-as-in-sql/ pa ge 1 pa ge 65 american journal of smart technology and solutions (ajsts) harmonizing artificial intelligence with islamic values: a thoughtful analysis of religious, social, and economic impacts of technological advancements abdullah hemmet1* volume 2 issue 2, year 2023 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v2i2.2239 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: october 31, 2023 accepted: november 28, 2023 published: december 05, 2023 the research explored the integration of ai technologies with islamic principles and their implications for religious, social, and economic structures in muslim-majority countries. it used qualitative design, semi-structured interviews, and case studies to analyze digitization of the holy qur’an and hadith, and ai in banking compliance with shariah laws, focusing on lexical-semantic, syntactical, grammatical, and pragmatic issues. a few of the such ai-based qur’an tools include ksu qur’an (ينورتكلإلا فحصملا عورشم بقنملا) islam web, qur’anic arabic corpus, almonagib alqur’any ,(دوعس كلملا ةعماجب online qur’an (نآرقلا) tanzil (tanzil documents), the qur’an (al-qur’an ,(ينآرقلا project translation and tafsir) and the noble qur’an (the noble qur’an – نآرقلا islamic fintech, which adheres to shariah principles, is embracing innovation .(ميركلا to streamline transactions and comply with the holy qur’an. ai virtual assistants like ‘aisyah’ are being adopted by islamic banks to streamline transactions and eliminate risks like leverage and uneven maturity. the regulatory management of islamic fintech includes issues like islamic cryptocurrency rules, money creation, and eliminating ‘riba’. in malaysia and indonesia, digital zakat distribution and classification systems are being developed using ai to accurately classify those in need and compute financial percentages for each group. thematic analysis revealed that human-inspired, analytical, and humanized ai has led to its significant impact on various aspects of life, society, and employment. nonetheless, ai cannot substitute for the profound comprehension and knowledge of islamic jurists, alongside human reasoning and intellectual competence. job displacement and ethical concerns of accuracy, transparency, fairness, and accountability remain significant challenges for society. keywords artificial intelligence, islamic shariah, qur’anic translation tools, automation, job displacement 1 islamic university of minnesota, al-amaan center, 5620 smetana dr, minnetonka, mn 55343, united states * corresponding author’s e-mail: abdullahhemmet0@gmail.com introduction the unwavering principles of the universe as created by god in islamic worldview are directed from the words of god and sunnah, where the ultimate source of guidance is divine revelation. humans are gifted with ‘reason’ as an essential tool by god, not to encourage blind acceptance of ideas. however, peace and harmony should be maintained (nadvi & junaid, 2021). according to the qur’anic value system, the centrality of justice is the standard in harmony with islamic law. allah says, “and we have certainly honored the children of adam and carried them on the land and sea and provided for them of the good things and preferred them over much of what we have created, with [definite] preference.” (surah al-isra’, verse 70). hence, islamic ethics and principles allow human beings to adopt what is good for them and create in the world for their daily living (ebrahimi, 2017). the advances in artificial intelligence (ai) are one of the valuable things in the world established in the 1950s as an academic discipline and is defined as: “a system’s ability to interpret external data correctly, to learn from such data, and to use those learnings to achieve specific goals and tasks through flexible adaptation.”(haenlein & kaplan, 2019). it exhibits cognitive, social and emotional intelligence. ai is significantly impacting firms, life, society and employment, producing multiple opportunities for exploiting the additional benefits (makridakis, 2017). the newest definition of ai applications’ importance is stated by harari: “science is converging to an allencompassing dogma, which says that organisms are algorithms, and life is data processing. intelligence is decoupling from consciousness. non-conscious but highly intelligent algorithms may soon know us better than we know ourselves (p.397).” (harari, 2016; makridakis, 2017). on the other hand, according to stephen hawking: “the rise of powerful ai will be either the best or the worst thing ever to happen to humanity. we do not yet know which.” (makridakis, 2017). there is a significant potential for ai to enhance human life with its broader applications in the daily lives of humans, including healthcare, finance, education, agriculture, human resources and recruiting, military training and air combat, customer service, music composition, social media newsfeed, reliable engineering and maintenance, work scheduling and optimization and autonomous vehicles or traffic management and many others (lo piano, 2020). ai and ml significantly impact daily decision-making due to efficiency and speed, but ethical concerns of accuracy, transparency, fairness, and accountability must be addressed. (lo piano, 2020). the ai revolution has significantly impacted both worldly and religious life, with arguments suggesting science is not anti-islamic and should be used for practical purposes (bashir, 2021). therefore, the current research was conducted with an aim to discuss ai in broader islamic, social and economic pa ge 66 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 perspectives for active participation and meeting the goals of the conference. ai is a powerful tool that has both promises and concerns. its integration with religious, social, and economic structures has sparked debates and dilemmas. in islam, ai presents both challenges and opportunities. it can serve islamic law objectives, preserve and digitize sacred texts, and intersect with the socio-economic dynamics of muslim-majority societies. this article explores the multi-dimensional impact of ai in the modern age and its relationship with islamic principles and societal norms. it aims to foster a nuanced understanding of ai and emphasize its responsible and ethical deployment in line with religious and cultural values. the research question addressed in the current research is stated below: how does the integration of artificial intelligence technologies align with islamic principles, and what are its implications for religious, social, and economic structures in muslim-majority societies? the evolution of technology and ai’s limits historical evolution of ai according to haenlein and kaplan (2021), the historic evolution of ai can be elucidated from the four seasons. the birth of ai begins with ai spring traced back to 1942 when isaac asimov published ‘runaround’, a short story. the three levels of robots were explained in the book: a robot must not injure human beings, must obey their orders and must protect their existence. alan turing first developed ‘the bombe’, a machine based on computing machinery and intelligence. later, ibm 701 was designed by nathaniel rochester to stimulate ai (haenlein & kaplan, 2019). ai summer and winter began with the development of the eliza computer program by 1966, capable of simulating human conversations. in ai fall, expert systems came into being with a set of rules formalized with human intelligence using if-then statements like ibm’s deep blue chess playing program and prospector expert system, which was used to study mineral deposits (haenlein & kaplan, 2019; zhang & lu, 2021). it led to the creation of artificial neural networks (anns) and deep learning when google developed alphago in 2015 (haenlein & kaplan, 2019; zhang & lu, 2021). in 2012, the advent of deep learning and neural networks prospered with improvements in the field of pattern recognition, offering satisfactory results among search engines, speech and visual recognition, recommendation systems and semantic analysis (haenlein & kaplan, 2019). with deep learning being evolved, accuracy and speed have been fine-tuned with commonly used latest models of faster r-cnn (region-based convolutional neural network), mask-rcnn, and yolo (you only look once) (zhang & lu, 2021). notably, from eliza to neural networks ai has been recognized to the advent of self-driving cars, smart speakers and facebook’s image recognition algorithm. the overall understanding of development of ai shows the rise in continuous improvements in computing power and big data has enhanced the significance of ai classified into human-inspired, analytical and humanized ai. challenges in ai’s development a challenge in ai and ml integration is to improve performance at scale and resource autonomy in autonomic systems. however, in the views of the researcher of this research, one of the greatest challenges in ai’s development is the correct ai predictions. similalry, past research highlighted which could substitute the role of humans being superior in performing mental tasks by 2037 (makridakis, 2017). besides, challenges arise in the field of quantum computing with emerging computational paradigms like fog, edge, cloud, and serverless environments (gill et al., 2022). the limitations of quantum computing include the pursuit of robust and scalable quantum hardware and quantum decoherence. it is pertinent to the nature of quantum information, delicately susceptible to disruption (córcoles et al., 2019; rayhan & rayhan, 2023). as an ai researcher, i have noted that these challenges in quantum computing which can cause errors in reliable quantum computations. moreover, the societal and ethical implications of ai’s evolution emphasize the importance of global collaboration and responsible innovation, mitigating potential risks and ensuring data privacy. similalry, existing literature showed that with ai quantum technologies being democratized, protecting sensitive information and ensuring transparency is essential (rayhan & rayhan, 2023). the ethical reasoning capabilities are associated with the autonomous system’s behavior and replacing traditional social structures accounting for moral values. the certification processes, standards and codes of conduct should be ensured by the users and developers of ai systems (dignum, 2018). ai’s potential and its limits as an ai researcher, i believe ai has been around in multiple forms of niche applications and pilots for decades. however, according to mehr et al. (2017), only recently has it been embedded within virtual and physical environments. by 2035, the potential of ai is to double the rates of economic growth, as estimated by accenture (mehr et al., 2017). the applications where ai excels include automotive, financial, retail and healthcare sectors with mature scenarios. autonomous driving includes the integration of driverless driving and fixed-speed cruise automatic parking (zhang & lu, 2021). in financial markets, ai is successfully embedded in intelligent consulting, intelligent risk control, credit rating and market forecasting, providing real-time risk warnings to financial institutions (bisht et al., 2022). on the other hand, the health sector is dealing with huge datasets for improved medical assistance, developing new drugs and detecting cancer (noorbakhsh-sabet et al., 2019). notwothstanding, in retail, the application of ai combined with ml, sensors, and computer vision pa ge 67 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 monitors goods replacement, inventory management, market forecasts and monitoring in the virtual shopping cart like amazongo (chi et al., 2020). in the media sector, one-click content generation by communication robots in the media industry is boosting brand promotion with ai. in smart payment systems, ai allows voiceprint recognition and face scanning, like the alipay pay payment system (zhang & lu, 2021). it has noted by the researcher that with a great many potential applications, ai can be challenging in any sector struggling due to overreliance of ml models. as examined from the existing literature, increased cybersecurity risks, adverse impact of ai applications and limited technical expertise (nishant et al., 2020). religious, social, economic and ethical impacts of ai-based technological advancements islamic perspective on ai in light of islamic jurisprudence, the social and cultural impact of ai can be tacked with islamic law traits to serve the goals of shariah as stated in the holy qur’an: “and pursue not that of which thou hast no knowledge; for surely the hearing, the sight, the heart all of those shall be questioned of.” (surah al-isra’, verse 36) (shahrouri, 2023). nothing prevents ai from declaring anything to be forbidden as long as there is any proof that it violates islamic law (haram). things are designed to make life easier for people. as allah stated: “do you not see that allah has made what is in the heavens and what is in the earth subservient to you, and made complete to you his favors outwardly and inwardly?” (surah al-luqman, verse 20) (ahmed, 2021). another qur’anic verse on islamic law and jurisprudential principles on ai is translated as: “amongst the important basic rules and principals of the islamic religion is that all things are permissible except what involves islamic prohibition. allah the almighty says (what means): qur’an it is he who created for you all of that which is on the earth.” (surah al-baqarah, verse 29) (ahmed, 2021). considering the quadruple bottom line theory (prophet, prosperity, planet and people) in islamic banking, the role of spirituality is pivotal. the prophet dimension must be prioritized over prosperity (profit). also, incentives to people should be decided for undertaking social functions and conserving the planet (hamidi et al., 2023). furthermore, in my point of view, the association between finance and ethical issues in ai implementation is significant. consequemtly, pass researchers rabbani et al. (2022) recommended the use of islamic finance tools in the regulation technology (regtech) sector to limit integrity and transparency challenges (rabbani et al., 2022). for example, in islamic jurisprudence, the financial arrangement for islamic loans is supported by qardh-al-hasan if applied properly in fintech with ai. qardh-al-hasan was mentioned by allah six times in the holy qur’an. allah says: “who is it that would loan allah a goodly loan so he may multiply it for him many times over? and it is allah who withholds and grants abundance, and to him you will be returned” (surah albaqarah, verse 245). this verse depicts that qardh-alhasan is a beautiful tool of finance to eradicate poverty with the inclusion of ai in finance (sarac & hassan, 2020). the researcher in this study discusses the scope and limitations of digitizing the holy qur’an and hadith, two legitimate sources of guidance for muslims. hakak et al. (2022) explained that since the content of hadith is fabricated from online content, the task is challenging to digitizes the original hadith without any alterations. however, the recent advancements in the statisticalbased and rule-based approaches in the arabic language enhance the credibility of the digitized system (hakak et al., 2022). also, to preserve the integrity of the qur’an, there are several trustworthy organizations and religious bodies which approve digitized holy qur’an, like king saud university and saudi ministry’s king fahd complex for the printing of the holy qur’an (hakak et al., 2017). this research case study analyzes data-driven qur’anic translation tools and ai in banking, focusing on compliance with shariah laws. social implications of ai according to zhao et al. (2022), the misuse of ai can lead towards cyber bullying, terrorism and criminal activities. however, as the researcher noted the education of the new generation should be trained as ‘digital citizens’, knowing how to deal with the social issues of the present and future technologies. consequenytly, previous research showed that the role of education and training for safe ai utilization is linked with meeting the standards of human rights, labor practices, organizational governance, fair operating practices and community involvement (zhao et al., 2022). this study’s researcher explores the potential of ai to bridge societal gaps, examining its social and ethical impact from both an ai and human perspective. as reviewed from the pas study by mhlanga (2023), ai significantly influences the widening inequality in society, the undermining of community culture, the sense of community and identity in society and the sense of identity. therefore, ‘fairness’ in ai algorithms is effective for making ethical decisions and narrowing societal gaps. ai must meet the standards of balancing society’s welfare, economic growth and the environment to achieve social responsibility. the unified framework of ai to meet social implications includes beneficence, autonomy, non-maleficence, explicability and justice (mhlanga, 2023; saveliev & zhurenkov, 2021). besides, the social and ethical hazards of ai are also posed by social toxicity and prejudice. however, instructional technologies offer safe ai utilization in online learning with multiple learning management systems, interactive whiteboards, mobile technologies, virtual and augmented reality and gamification (uunona & goosen, 2023). economic ramifications of ai advancements this research focused on the pace at which tasks are automated with evolved ai capabilities and ai has taken pa ge 68 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 over a few job roles. as examined in the existing literature, the risk of job displacement for ‘shop-floor’ workers in manufacturing is limited due to the human capabilities of managing cognitive tasks (tyson & zysman, 2022). job displacement is a key concern associated with ai integrated into workforces, causing potential unemployment. it is worth noting that ai technologies are most likely to take over task-based displacement in jobs, whereas higher-value tasks are still required to be done by humans (george et al., 2023). notably, the job growth for lowand middle-wage occupations is most likely to decline with ai advancements in occupations like production, support and warehousing in technology, healthcare and education sectors. in contrast, high-wage jobs are expected to increase the demand for upskilling employment (tyson & zysman, 2022). notowithstanding, it is possible for displaced employees to discover that time and money are required to retrain, re-skill, and re-educate themselves in order to enter different industries. in general, technological innovation widens the gap between the rich and the poor since it frequently replaces low-skilled employees and reduces the need for their services (nissim & simon, 2021). economic development is positively linked with the improvements in work performance, work conditions and work relationships (caruso, 2018). the new opportunities posed by ai have transformed manufacturing businesses, offering the provision of manufacturing maintenance and repair, innovation information, analytical services and operations of the supply chain ecosystem (ehret & wirtz, 2017). as reported by the world economic forum (wef), ai is the game changer in multiple applications, including distributed energy grids, smart disaster response, smart agriculture and food systems, autonomous electric vehicles, ai-designed connected, intelligent and livable cities, next-generation climate and weather prediction and reinforcement learning for earth sciences (yigitcanlar et al., 2020). in my observation from the literature review, economic opportunities tied to ai-driven industry 4.0 regard the cycle of innovation in private and public institutions. as noted in existing literature, ai is posing new economic developments in multiple industries, including healthcare, finance, retail, supply chain, manufacturing and logistics (dwivedi et al., 2021). according to the wef probabilistic research on the economic effects of ai and automation, 20% of current uk occupations may be affected by ai technology (dwivedi et al., 2021). due to the larger potential for technological development throughout the manufacturing industry, this number is higher in growing economies like china and india, where it increases to 26%. according to forecasts, ai technology will promote innovation and economic growth, generating 133 million new jobs worldwide by 2022 and 20% of china’s gdp by 2030 (dwivedi et al., 2021). governments worldwide are recognizing ai as a nation-defining capacity, with 50 countries implementing special national ai policies since february 2020, accounting for 90% of the world’s gdp (yigitcanlar et al., 2020). it has studied that the key challenges tied to ai include the obstacles of data monopolies and obtaining the data which harm small and medium enterprises. for the workforce, a technological revolution is a key driver for increasing cost effectiveness, bringing time and overall productivity for multiple stakeholders and labor markets (susar & aquaro, 2019). the main risk here is that ai is expected to become a vehicle only for the 1% of highincome earners, i.e., the wealthiest and monopolistic multinational corporations (yigitcanlar et al., 2020). the need for responsible and ethical ai the ethical concerns which should be explicitly noted for responsible ai integration include fairness, justice and dignity. technological innovation is seen as a significant factor in solving problems of the society and promoting happiness, economy and well-being (leenes et al., 2017). in my opinion, responsible innovation with ai is balanced as an interactive and transparent process tackling the challenges of economic progress and social welfare. ethical responsibility should focus on technology push, foresight, and policy pull, as ai algorithms’ failures in decision-making and predictive analysis can lead to potential biases like lack of transparency, trustworthiness, respect and privacy protection (yigitcanlar et al., 2021). there are many academic, public and private sector international bodies which support governments in ai regulation and promoting its research and development like the partnership on ai, the international association for artificial intelligence and law, the artificial intelligence forum of new zealand, and sparc in the eu, and few of the tech firms’ (apple, facebook, baidu, amazon, ibm, google, and microsoft) (erdélyi & goldsmith, 2018). international organizations such as the oecd principles of ai and the council of europe’s expert committee on human rights and ai provide ethical guidelines for reducing ai-related risks in a humancentered society. the purpose is to meet the standards of transparency, fairness and accountability (radu, 2021). the european commission’s high-level expert group on ai (hleg) is responsible for dealing with the moral activities of ai and establishing trustworthy aibased applications (ryan, 2020). the meaningful human control of ai algorithms is important to deal with the ethical, legal, organizational, technical and societal issues to enhance the moral values of ai systems (santoni de sio & mecacci, 2021). thus, the recent study’s purpose was to identify the potential biases which are dealt with by creating reliable ai guidelines shifting away from nonautonomous ai. methodology research design a qualitative research design was adopted in the current research to present an in-depth analysis of the religious, social, economic and ethical implications of ai in light of multiple perspectives and opinions. the interpretivist pa ge 69 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 philosophy was used to highlight human opinions on ai and its technological advancements in the social world. the qualitative research was deemed to be a suitable way of gathering multi-faceted views of the economic, social and religious repercussions of ai technologies in the recent era. data collection semi-structured interviews were conducted for data collection in the current research through open-ended questions. the questions were designed based on the theorization of the islamic, social, economic and ethical constructs discussed with reference to the existing literature to enhance their validity. also, a few abc university professors were approached to test the validity of interview questions and enhance study outcomes, specifically targeting muslim-majority countries and diverse demographic groups, ensuring broader findings across age, gender, profession, country and level of religious observance. the second approach of data collection was the case study analysis on the applications of ai in muslim-majority countries reflecting upon the islamic jurisprudence and shariah laws. the two case studies include ai-driven qur’anic translation tools and ai in banking compliant with shariah laws. table 1 below presents the demographics including age, gender, country of residence, profession, and level of religious observance of the 13 interview respondents who participated in this research. the research participants were sampled using a purposive sampling technique since the purpose of the research was to reach out to the experts in the field of ai and get some useful and constructive opinions on ai implications. therefore, professors, ai experts, religious scholars and ai technologists were approached to gather diverse viewpoints on the research question. table 1: demographics of interview participants demographics frequency percent age 21-30 years 3 23.08% 31-40 years 4 30.77% 41-50 years 3 23.08% above 50 years 3 23.08% gender male 7 53.84% female 6 46.15% country of residence saudi arabia 3 23.08% pakistan 2 15.38% indonesia 2 15.38% bangladesh 1 7.69% malaysia 2 15.38% egypt 1 7.69% united arab emirates 2 15.38% profession professor 4 30.77% ai expert 3 23.08% religious scholar 3 23.08% ai technologist 3 23.08% level of religious observance very observant 4 30.77% moderately observant 4 30.77% slightly observant 4 30.77% not observant 2 15.38% total 13 100.0 data analysis a thematic analysis approach underlies the current research’s conclusions. the transcripts were encrypted from audio to textual format following transcription and de-identification. further, noting iteratively created codes for each interview section, themes were designed to analyze the interview responses, identifying repetitive keywords in the majority of responses. ethical considerations all respondents were asked to sign an informed consent through email explaining the research purpose and confirming that their data private data would be kept confidential throughout the research. results and discussion thematic analysis a few of the in-depth responses of the current research participants are stated below for each of the interview questions explaining their diverse viewpoints on all the derived themes from the interview transcripts. pa ge 70 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 ai and religious perspectives how would you define artificial intelligence (ai)? participant 1, an ai expert, stated: “in simple terms, i would say ai refers to the advancement of computer systems capable of performing tasks with human intelligence by learning, reasoning and making decisions.” participant 3, a professor, mentioned: “in my opinion, ai is a way of developing machines which can make predictions copying human cognitive functions when performing autonomous tasks.” to what extent do you believe that ai aligns with the unwavering principles of the universe as created by god? participant 2, a religious scholar, stated: “as muslims, we must emphasize the value of upholding human dignity, compassion, and the quest for spiritual development while acknowledging the promise of ai. they provide caution against placing an undue dependence on ai, which might cause morality, empathy, and human contact to be neglected.” participant 3, a university professor, mentioned: “i have a certain belief that ai can align with the principles of god. for instance, in the field of healthcare, ai is definitely aiding humanity, but it may violate the ethical considerations of protecting human dignity.” in your opinion, how has the recognition of these principles been reflected in the work of both muslim and non-muslim scientists? participant 4, a religious scholar, mentioned: “while some muslim and non-muslim scientists believe that ai can fulfil the needs of humans, some believe that ai can be of assistance for humans, but it cannot genuinely embody the divine and meet the human needs of spirituality.” participant 5, an ai technologist, stated: “it is admirable that religious concepts are being acknowledged in scientific research. scientists from all backgrounds, including muslims and non-muslims, have recognized the significance of ethical principles derived from many theological and philosophical traditions. examples of notions that are becoming more prevalent in ai research include social justice, human dignity, compassion, sustainability and environmental stewardship. this understanding aids in bridging the gap between moral duty and scientific development, ensuring that technical advancements are guided by values that benefit the general good.” from the above mentioned responses, it has observed muslims stress the significance of maintaining human dignity, compassion, and spiritual growth while recognizing the promise of ai. from my analytical insights, i have noted from interviews of respondents from multiple religious backgrounds and ethnicities that muslims stress the significance of maintaining human dignity, compassion, and spiritual growth while recognizing the promise of ai. muslims emphasize the importance of ai being in line with god’s principles in particular. their viewpoint is given a special dimension by the religious setting, which may not be present in the opinions of scientists from other backgrounds. while both groups place considerable emphasis on moral ideals, muslims may do so more explicitly than scientists from different backgrounds do by prioritizing ideas like compassion and spiritual development. evolution and limits of ai how do you perceive the historical evolution of ai, from its early days to the current deep learning models? participant 6, an ai expert, stated summarized ai evolution: “ml was first used in relation to neural networks in the 1950s. in the 1960s, shakey, the very first mobile ai robot, and eliza, a cognitive-skill-based chatbot, were developed. later, it was followed by ai renaissance in the 1970s and 1980s. synthesis of speech and video in the 1990s was observed within ai development. the decade of the 2000s witnessed the rise of ibm watson, face recognition, virtual assistants, autonomous cars, deepfakes, and content and picture production.” participant 5, an ai technologist, stated: “the development of ai has been remarkable, starting with rule-based systems. over time, machine learning methods emerged, advancing technology. deep learning models revolutionized ai research, enabling tasks like image identification, nlp, recommendation systems, medical diagnosis and treatment, computer vision, robotics, fraud detection, and more. this demonstrates our inventiveness and ability to replicate cognitive processes in robots.” what are the key challenges you see in the development of ai, especially from a religious and ethical standpoint? participant 7, a religious scholar, stated: “i believe challenges from a religious and ethical perspective include assuring justice and openness in ai decision-making, as well as addressing worries about ai taking the place of humans in the workforce. the issue of responsibility arises when ai makes choices that have moral ramifications. it is a big task to strike a balance between moral and spiritual ideals and technical progress. allah states in the holy qur’an, “and spend in the way of allah and do not throw [yourselves] with your [own] hands into destruction.” (surah al-baqarah, verse 195); thus, ai should only be used in ways which benefit people.” can you highlight any specific areas where ai excels and where it might be ineffective? participant 7, a religious scholar, stated: “data analytics, prediction modelling, automation and pattern recognition are activities where ai has emerged. it excels in fields like financial forecasting and medical diagnostics. nevertheless, as i already pointed out, it might not work well for activities requiring moral judgements, empathy, and compassion, which are central to many religions, pa ge 71 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 not only islam. additionally, it could not perform well in activities that call for complicated moral thinking, emotional intelligence, and creativity.” the findings from thematic analysis showed that ai advancements, from rule-based systems to deep learning models, have revolutionized various applications but raise moral and religious questions. while ai excels in financial forecasting and medical diagnosis, it struggles with morality and empathy challenging ethical and spiritual aspirations. according to my analysis on the respondents’ views, ai scientists and islamic scholars are looking at how ai innovations like neural networks might be compatible with islamic principles. particularly intriguing is the ethical use of ai in the fields of banking, medicine, and other shariah-compliant industries. nonetheless, the main issue is ensuring ai development adheres to ethical standards and respects human dignity, addressing biases, safeguarding privacy, and avoiding applications that could harm humanity or violate religious principles. all things are considered halal for humans, as they are tahir (pure) and muslims can handle or touch them obeying laws of shariah. religious implications of ai how do you view ai’s role in islamic jurisprudential endeavors? participant 4, a religious scholar, stated: “studying islamic jurisprudence and shariah all my life, i assure you that, although in order to get a greater grasp of islamic law, ai can assist academics in analyzing the context and historical interpretations of qur’anic quotes and hadith, but as ai cannot substitute for the profound comprehension and knowledge of islamic jurists, it should always be used to supplement human reasoning and intellectual competence. ai can be a tool for human assistance in understanding islamic jurisprudence; however, the holy qur’an is the righteous way to understand it as allah says in the holy qur’an: “indeed, in the creation of the heavens and the earth, and the alternation of the night and the day, and the [great] ships which sail through the sea with that which benefits people...” (surah al-baqarah, verse 164).” what are your thoughts on the digitization of the holy qur’an and hadith through ai? participant 2, a religious scholar, stated: “for muslims everywhere, the digitalization of the holy qur’an and hadith by ai is a useful resource. in particular, aipowered applications can make it simpler to access hadith collections and verses from the qur’an, facilitating improved comprehension and study. allah warns about misinterpretation of the holy qur’an as he says: “indeed, those who conceal what we sent down of clear proofs and guidance after we made it clear for the people in the scripture those are cursed by allah and cursed by those who curse.” (surah al-baqarah, verse 174).” “in the digitization of the holy qur’an, qur’an, fiqh (islamic law) and uṣūl al-fiqh (legal theory) must be followed considering the significance of syntax, morphology, word sentence disambiguation, and other aspects. i am still in doubt whether ai can abide by all the rules of qur’an recitation translated into other languages as compared to the arabic language. however, i believe digitization of the holy qur’an and hadith is important for spreading the message of allah and maintaining the original text’s integrity as allah says: “indeed, it is we who sent down the qur’an, and indeed, we will be its guardian.” (surah al-hijr, verse 9).” how can we ensure that ai aligns with the objectives of islamic law and jurisprudential principles? participant 2, a religious scholar, summarized how ai can be aligned with islamic law and what is wrong with it: “ai presents muslims with a wealth of exciting options, but it also poses unprecedented risks to the faith by perhaps eroding the qur’an’s holiness. it is not impossible to see a modern-day islamophobe using cutting-edge generative ai technology in a fresh effort to challenge muslims’ beliefs. in the qur’an, allah makes a reference that suggests that even ai-powered initiatives will fall short as he states: “say, ˹o prophet, if ˹all˺ humans and jinn were to come together to produce the equivalent of this qur’an, they could not produce its equal, no matter how they supported each other.” (surah al-israa, verse 88).” participant 8, an ai expert, stated: “in recent years, the use of ai is mostly seen in finance. therefore, ai algorithms applied in banking or commerce should abide by islamic fairness and riba (unsury) prohibitions as allah says in the holy qur’an: “those who devour usury will not stand except as stand one whom the devil has driven to madness by [the touch of] insanity.” (surah al-baqarah, verse 275)” thematic analysis revealed that scholars stressed the importance of human reasoning in interpreting islamic law, preserving the accuracy of digital qur’anic and hadith materials, and adhering to linguistic and syntactical rules. they also caution against ai eroding the sanctity of the qur’an, emphasizing the responsibility of muslims. in my opinion, muslims need to take proactive steps to guarantee that they benefit from ai while safeguarding themselves from its risks. notably, religious scholars were agreed on the fact that it is crucial to guarantee the validity and correctness of digital material. social implications of ai what do you think is the role of education and training in ensuring safe ai utilization? participant 8, an ai expert, stated: “the responsible use of ai necessitates extensive education and training, enabling individuals to understand its consequences and risks, enabling ethical judgements… this also ensures ai professionals adhere to ethical standards and industry best practices, reducing potential harm from improperly developed ai systems.” participant 9, a professor, mentioned: “the role of education and training in ensuring safe ai usage is crucial, pa ge 72 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 as teachers and students must be knowledgeable about ai usage to identify potential misuse and its benefits.” how can we ensure the protection of personal data and maintain individual privacy when using ai systems? participant 12, an ai technologist, mentioned: “in the ai era, protecting personal information and privacy requires a comprehensive strategy including strong data encryption, access restrictions, open data management procedures, and strict adherence to privacy laws. implementing data access controls can help prevent inadvertent data disclosure and prevent unauthorized individuals from sharing private information with ai training models.” participant 4, a religious scholar, mentioned: “ai clearly has the potential to increase productivity, improve customer service, speed up response times and delivery, promote the development of new markets, and even save expenses for businesses… yet there are still moral concerns in the areas of privacy and security, risk factors of misinformation, as well as the human cost.” in your opinion, will ai play a role in widening or narrowing societal gaps? please explain participant 13, a professor, mentioned: “robots with ai capabilities may eventually replace humans in a variety of sectors, including manufacturing and transportation, as they grow more advanced and capable… job displacement and economic turmoil may result from this. however, ai can bridge societal disparities by expanding access to jobs, healthcare, and educational opportunities if it is applied wisely.” from the thematic analysis, it was noted that ai will probably decrease the proportion of revenue going to low-skilled workers while increasing wages for highly skilled workers, leading to greater inequalities in society. from my insightful evaluation of the respondents’ views, i believe that users should be informed about the data-gathering process and have the option to opt in or out. data security and privacy should be prioritized in the development process. considerably, educational institutions are collaborating with technology partners to create comprehensive training programs for ai usage both in and outside the classroom. economic ramifications of ai how do you perceive ai’s impact on job creation and job displacement, especially in muslimmajority economies? participant 1, an ai expert, stated: “employees with less education or older workers may find it challenging to master new skills when their employment requires repeated or routine operations.” participant 9, a professor, mentioned: “ai’s impact on employment displacement and creation varies by industry and location. while ai may initially displace everyday jobs in majority muslim countries, it can also create new opportunities in ai research, development, and maintenance… ai’s significance has grown in healthcare, banking, education, and other industries, boosting accuracy, reducing costs, and improving efficiency.” what economic opportunities do you foresee with the rise of ai-driven industries? participant 6, an ai expert, stated: “ai-driven enterprises have economic potential in healthcare, banking, and logistics, attracting investment in developing muslim-majority countries… fostering an ai-friendly environment boosts productivity, diversifies industries, encourages entrepreneurship, and attracts foreign capital.” how do you view challenges arising from ai monopolies and global economic disparities? participant 9, a professor, stated: “ai could lead to a shift in investments from emerging economies to advanced ones, potentially causing a temporary decrease in gdp in developing countries. developing muslim-majority economies could promote local ai development, international cooperation, and equal access to ai technology laws, thereby reducing the problems caused by monopolies.” the interviews with targeted population enhance my analytical perspective on ai and economic ramifications. i think ai may not entirely replace employment, but demand for highly trained positions like software engineers, digital marketers, and data scientists will rise in the coming years. nonetheless, ai monopolies could worsen existing economic imbalances by preventing fair competition and innovation. education and workforce development are crucial for a smooth transition. ethical ai how can we ensure a balance between rapid ai innovation and maintaining strong ethical considerations? participant 1, an ai expert, stated: “i believe that determining who is accountable for the deliberations and choices made by ai systems requires the development of specific regulations and processes.” participant 5, an ai technologist, stated: “regular ai system evaluations and public participation in decisionmaking ensure that innovation aligns with social ideals. maintaining a strong ethical foundation while promoting innovation can be achieved through ongoing monitoring, open reporting, and engagement with ethicists.” are you concerned about biases in ai systems? how do you believe we can develop more equitable algorithms? participant 10, an ai technologist, stated: “i am not concerned as we can take a few steps to mitigate such biases of ai. for example, to reduce bias in ml algorithms, various measures can be taken, including using a diverse training dataset, eliminating sensitive variables, employing bias mitigation strategies, conducting regular model evaluations, and ensuring human oversight. creating pa ge 73 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 equitable algorithms requires a multifaceted approach, including transparency in ai design, thorough, robust testing, and auditing to ensure equality and fairness in ai algorithms.” what role should governments and international bodies play in ai regulation? participant 8, an ai expert, stated: “governments and international bodies play a crucial role in responsible and ethical ai development by creating legislation that addresses concerns like algorithm transparency, data protection, and discriminatory practices. ai and automated systems are utilized by governments at all levels for decision-making, law enforcement, public good support, and public feedback and complaints collection.” businesses should have systems of checks and balances in place, as well as procedures for identifying and resolving potential issues, to guarantee accountability and responsibility. however, ai experts and technologists during interviews were more concerned on balancing ai advancement and ethics is crucial. i have noted that interdisciplinary cooperation between religious scholars, technologists, and legislators can create ethical norms. general and future implications what areas do you believe require further exploration within the realm of ai, especially from a religious perspective? participant 4, a religious scholar, stated: “ai-driven systems that are based on the qur’an’s idea of justice (surah al-hujurat) can be created to ensure the fair allocation of resources. ai may also be used to monitor environmental damage and reduce it, which is in line with allah’s commands in the hold qur’an on good stewards of the planet (surah al-an’am).” how do you view the potential for collaboration between technologists and islamic scholars in shaping the future of ai with reference to the holy qur’an? participant 7, a religious scholar, stated: “collaboration between tech experts and academics from islam has enormous potential for influencing ai development. the qur’an advocates gaining knowledge, which includes advances in technology (surah al-mujadila 58:11). religious leaders can influence ai policies that uphold islamic principles, promoting justice, responsibility, and compassion, ensuring its application for societal advancement. the qur’an stresses the value of consultation while making decisions (surah ashshura). together, technologists and religious scholars can direct the development of ai to advance mankind while respecting moral standards.” i am certain that from the point of view of religion, ai research should concentrate on assuring moral and equitable uses. respondents mentioned that ai may also improve healthcare, which is in line with islam’s core ideal of protecting human life. ai has more potential in the areas of compassion, justice, and environmental stewardship. it is important to track the control of equitable ai algorithms possessing sufficient moral awareness and sufficient knowledge to give results with no bias. case study: applications of ai in compliance with islamic laws ai-driven qur’anic translation tools ai usage nowadays is not uncommon as it can be observed everywhere, from applications of image processing to sophisticated applications of self-driving vehicles. machine translation systems, sentiment analysis and chatbots are developed with the use of natural language processing (nlp) and ai (khan & rabbani, 2020). according to recent demographic data, there are 480 million arabic speakers in islamic, asian, and african nations. when ranking languages by how widely they are spoken, arabic came in fourth. arabic is a language of significant significance since it is the language of the holy qur’an, popular poetry, and the cultural legacy of its people. the translation tools are influenced by the neural machine translation system, which resolves lexicalsemantic problems, syntactical problems, grammatical problems, and pragmatic problems (khalatia & alromanyb, 2020). one such translation tool is the almaany dictionary website, a bilingual dictionary based in egypt, turkey, india, and jordan, which analyses 12 million qur’anic words in spanish, turkish, french, and german, using certified translations of the holy qur’an. (zemni et al., 2020). the study by mohamed and shokry (2022) introduces the holy qur’an’s concept-based searching tool (qsst), which has four phases. the first phase involves manually annotating verses using the mushaf al-tajweed ontology. word embedding creates feature vectors using a continuous bag of words architecture. the third phase calculates feature vectors for both input questions and qur’anic subjects. the most pertinent verses are found by computing cosine similarity (mohamed & shokry, 2022). furthermore, for qur’anic academics and arabic researchers, semantic or concept-based and keywordbased searches are the two main categories of qur’anic search methods. it might be difficult to do a conceptbased search in a large corpus like the qur’an. few of the ai-based qur’an tools and applications include ksu qur’an(كلملا ةعماجب ينورتكلإلا فحصملا عورشم ,islam web (islam web), qur’anic arabic corpus ,(دوعس almonagib alqur’any (ينآرقلا بقنملا), tanzil (tanzil documents), the qur’an(al-qur’an (نآرقلا) online qur’an project translation and tafsir) and the noble qur’an (the noble qur’an – ميركلا نآرقلا) which enable users to listen, read, and search in qur’an with different languages (mohamed & shokry, 2022). the key challenge in qur’anic translation tools is that the keyword search does not take into account conceptual or semantic analysis for the query. thus, expected results are frequently not obtained. therefore, it is addressed by pa ge 74 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 a semantic and lexical-based tool, the qur’anic english wordnet (qewn) database (bashir et al., 2023). every single one of the words from the qur’an’s english translation, together with their semantic content, are filled up in this database. various ideas found in the qur’an are stored in the vocabulary of qur’anic search (vqc). in qewn, the concepts of sense and synset are used. a word’s sense refers to its specific usage, while a word’s synset is a list of all of its possible synonyms (bashir et al., 2023). ai in banking compliant with shariah laws islamic fintech accepts all sorts of innovation as long as it does not violate shariah principles. the two major risks in conventional banking are removed when ai complies with islamic fintech banking, including leverage and uneven maturity (khan & rabbani, 2020). khan and rabbani (2020) proposed ‘chatbot as islamic finance expert’ (caife), an interactive chatbot powered by ai. by allowing customers to speak with a robot that has expertise amassed via ml, our interactive chatbot caife receives automated robot help pertaining to islamic finance and banking. any question pertaining to islamic banking and finance is addressed on a real-time basis (khan & rabbani, 2020). the application of ai in islamic finance and investment includes robo islamic advisor (ria), robo advisor and robo financial advisor (rfa), which dwells on the cognitive application of compliance and investment sectors. these are based on the principles of the holy qur’an and shariah principles of gharar, riba and maysir, which are prohibited. these chatbots follow islamic teachings of urf ’ (customers’ practices), fiqh (jurisprudence), and qanun (ordinance) (gazali et al., 2020). the suggested integrated ai and nlp-based islamic fintech model by haider syed et al. (2020) combines qardh-al-hasan (charitable loan) and zakat (islamic tax), which can assist the economy in reducing the negative effects of covid-19 on people and smes. it also provides a recommender system powered by ai. if the lenders choose zakat as the method of assistance, the list of recipients will include data regarding the needs of those who qualify to receive zakat, and if they choose qardh-al-hasan assistance, the list will include details about the borrowers (haider syed et al., 2020). the python programming language is used to evaluate the shariah document screening method that is suggested by che mohd salleh et al. (2023) and is based on levenshtein distance similarity evaluation. the ability to detect fraud inside an organisation and perhaps increase data accuracy are two benefits of this strategy. the proficiency, consistency, and accuracy of fuzzy matching of strings have been calculated as automated measurements of participants’ accuracy in speech intelligibility, which used this technique. additionally, this technique works well for speech recognition, spelling checkers, spam detectors, and connections (che mohd salleh et al., 2023). islamic banks are adopting ai-virtual assistants like ‘aisyah’ to streamline transactions, complying with shariah laws and ideal economic conditions of musharakah and mudharabah. these assistants streamline processes, reducing energy, cost, and time. in a musharakah finance arrangement, the bank and clients pool resources to fund projects, with profits and losses distributed based on available cash (yuspin et al., 2022). a cooperative approach known as mudharabah finance entails the bank providing funds while the consumer contributes knowledge. both parties will agree upon a profit-sharing percentage. these types of financing meet islamic shariah goals, including upholding brotherhood and justice, achieving economic prosperity, fair income distribution and individual freedom in social welfare (yuspin et al., 2022). notwithstanding, despite a great many successes of ai in islamic finance, the ethical concerns and challenges are still there, which may not be appropriate with respective fiqh (understanding of shariah) and tafseer (qur’anic interpretation). the automation of ai in fintech reduces customer autonomy, transparency, reliability and accountability. however, regtech, in accordance with shariah principles, can resolve these ethical challenges of ai in fintech considering the significance of zakat and rizq-al-halal (rabbani et al., 2022). the regulatory management problem discusses themes such as islamic cryptocurrency rules, using islamic fintech to create money, adhering to shariah law, doing away with riba using islamic fintech, and offering sukuk (dawood et al., 2022). accurately distributing the zakat that has been collected is a further significant concern in fintech. in malaysia and indonesia, in accordance with ai, digital zakat distribution and classification systems are created to accurately classify those in need and compute the necessary financial percentages for each group (unal & aysan, 2022). conclusion ai’s potential to widen or narrow societal gaps is influenced by its social and ethical impact. ai algorithms must balance welfare, economic growth, and environmental responsibility to achieve social responsibility. a unified framework includes beneficence, autonomy, nonmaleficence, explicability, and justice. muslims caution against over-dependence on ai, as it may neglect morality and empathy. ai can aid in interpreting the meanings of qur’anic quotes and hadith but cannot replace the knowledge of islamic jurists. ai algorithms in finance should adhere to islamic fairness and riba prohibitions. ai’s impact on employment displacement and creation varies by industry and location but can also create new opportunities in ai research and maintenance. education and workforce development are crucial for a smooth transition. limitations and recommendations the current research was limited to a qualitative method, which may adhere to some bias in cultural and contextual bias. due to limited time and financial resources, 13 participants were sampled for interviews pa ge 75 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 for data collection. however, in the future, this limitation can be addressed by conducting quantitative research as well as collecting data from a relatively larger sample size. furthermore, ai experts, theologians and social scientists can review this research to benefit from the comprehensive understanding of the ethical issues of ai and islamic values. future researchers can conduct a comparative analysis among any two muslim-majority countries to compare their ai-integrated banking systems in compliance with shariah laws and valuable insights into the harmonization of ai evolving with time. references ahmed, b. (2021). the status of the use of artificial intelligence in ijtihad. karachi islamicus, 1(1), 1-14. bashir, a. (2021). the status of the use of artificial intelligence in ijtihad. karachi islamicus, 1(1), 1-14. https://doi.org/10.58575/ki.v1i1.5 bashir, m. h., azmi, a. m., nawaz, h., zaghouani, w., diab, m., al-fuqaha, a., & qadir, j. (2023). arabic natural language processing for qur’anic research: a systematic review. artificial intelligence review, 56(7), 6801-6854. bisht, d., singh, r., gehlot, a., akram, s. v., singh, a., montero, e. c., priyadarshi, n., & twala, b. (2022). imperative role of integrating digitalization in the firms finance: a technological perspective. electronics, 11(19), 3252. caruso, l. (2018). digital innovation and the fourth industrial revolution: epochal social changes? ai & society, 33(3), 379-392. che mohd salleh, m., nor, r. m., yusof, f., & amiruzzaman, m. (2023). constructing a shariah document screening prototype based on serverless architecture. computers, 12(3), 50. chi, h.-c., sarwar, m. a., daraghmi, y.-a., lin, k.-w., ik, t.-u., & li, y.-l. (2020). smart self-checkout carts based on deep learning for shopping activity recognition. 2020 21st asia-pacific network operations and management symposium (apnoms), córcoles, a. d., kandala, a., javadi-abhari, a., mcclure, d. t., cross, a. w., temme, k., nation, p. d., steffen, m., & gambetta, j. m. (2019). challenges and opportunities of near-term quantum computing systems. proceedings of the ieee, 108(8), 1338-1352. dawood, h., al zadjali, f., al rawahi, m., karim, s., & hazik, m. (2022). business trends & challenges in islamic fintech: a systematic literature review. f1000research, 11. dignum, v. (2018). ethics in artificial intelligence: introduction to the special issue. ethics and information technology, 20(1), 1-3. dwivedi, y. k., hughes, l., ismagilova, e., aarts, g., coombs, c., crick, t., duan, y., dwivedi, r., edwards, j., & eirug, a. (2021). artificial intelligence (ai): multidisciplinary perspectives on emerging challenges, opportunities, and agenda for research, practice and policy. international journal of information management, 57, 101994. ebrahimi, m. (2017). islamic identity, ethical principles and human values. european journal of multidisciplinary studies, 2(6), 325-336. ehret, m., & wirtz, j. (2017). unlocking value from machines: business models and the industrial internet of things. journal of marketing management, 33(1-2), 111-130. erdélyi, o. j., & goldsmith, j. (2018). regulating artificial intelligence: proposal for a global solution. proceedings of the 2018 aaai/acm conference on ai, ethics, and society, gazali, h. m., jumadi, j., ramlan, n. r., abd rahmat, n., uzair, s. n. h. m., & mohid, a. n. (2020). application of artificial intelligence (ai) in islamic investments. journal of islamic finance, 9(2), 70-78. george, a. s., george, a. h., & martin, a. g. (2023). chatgpt and the future of work: a comprehensive analysis of ai’s impact on jobs and employment. partners universal international innovation journal, 1(3), 154-186. gill, s. s., xu, m., ottaviani, c., patros, p., bahsoon, r., shaghaghi, a., golec, m., stankovski, v., wu, h., & abraham, a. (2022). ai for next generation computing: emerging trends and future directions. internet of things, 19, 100514. haenlein, m., & kaplan, a. (2019). a brief history of artificial intelligence: on the past, present, and future of artificial intelligence. california management review, 61(4), 5-14. haider syed, m., khan, s., raza rabbani, m., & thalassinos, y. e. (2020). an artificial intelligence and nlp based islamic fintech model combining zakat and qardh-al-hasan for countering the adverse impact of covid 19 on smes and individuals. international journal of economics and business administration, viii(2), 351-364. hakak, s., kamsin, a., tayan, o., idris, m. y. i., gani, a., & zerdoumi, s. (2017). preserving content integrity of digital holy quran: survey and open challenges. ieee access, 5, 7305-7325. hakak, s., kamsin, a., zada khan, w., zakari, a., imran, m., bin ahmad, k., & amin gilkar, g. (2022). digital hadith authentication: recent advances, open challenges, and future directions. transactions on emerging telecommunications technologies, 33(6), e3977. hamidi, m. l., zobair, k. m., & pratama, a. a. n. (2023). the role of spirituality dimension in the sustainability of islamic banking: a combined structural equation modeling and artificial neural network approach. environment, development and sustainability, 1-27. harari, y. n. (2016). homo deus: a brief history of tomorrow. random house. khalatia, m., & al-romanyb, t. a. h. (2020). artificial intelligence development and challenges (arabic language as a model). artificial intelligence, 13(5). khan, s., & rabbani, m. r. (2020). chatbot as islamic pa ge 76 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 65-76, 2023 finance expert (caife) when finance meets artificial intelligence. proceedings of the 2020 4th international symposium on computer science and intelligent control. leenes, r., palmerini, e., koops, b.-j., bertolini, a., salvini, p., & lucivero, f. (2017). regulatory challenges of robotics: some guidelines for addressing legal and ethical issues. law, innovation and technology, 9(1), 1-44. lo piano, s. (2020). ethical principles in machine learning and artificial intelligence: cases from the field and possible ways forward. humanities and social sciences communications, 7(1), 1-7. makridakis, s. (2017). the forthcoming artificial intelligence (ai) revolution: its impact on society and firms. futures, 90, 46-60. mehr, h., ash, h., & fellow, d. (2017). artificial intelligence for citizen services and government. ash cent. democr. gov. innov. harvard kennedy sch., no. august, 1-12. mhlanga, d. (2023). open ai in education, the responsible and ethical use of chatgpt towards lifelong learning. education, the responsible and ethical use of chatgpt towards lifelong learning (february 11, 2023). mohamed, e. h., & shokry, e. m. (2022). qsst: a quranic semantic search tool based on word embedding. journal of king saud university-computer and information sciences, 34(3), 934-945. nadvi, m. a., & junaid, m. (2021). understanding the principles of islamic world-view. jurnal the dialogue(3), 269-270. nishant, r., kennedy, m., & corbett, j. (2020). artificial intelligence for sustainability: challenges, opportunities, and a research agenda. international journal of information management, 53, 102104. nissim, g., & simon, t. (2021). the future of labor unions in the age of automation and at the dawn of ai. technology in society, 67, 101732. noorbakhsh-sabet, n., zand, r., zhang, y., & abedi, v. (2019). artificial intelligence transforms the future of health care. the american journal of medicine, 132(7), 795-801. rabbani, m. r., sarea, a., khan, s., & abdullah, y. (2022). ethical concerns in artificial intelligence (ai): the role of regtech and islamic finance. artificial intelligence for sustainable finance and sustainable technology: proceedings of icger 2021, 1. radu, r. (2021). steering the governance of artificial intelligence: national strategies in perspective. policy and society, 40(2), 178-193. rayhan, a., & rayhan, s. (2023). ai odyssey: unraveling the past, mastering the present, and charting the future of artificial intelligence. ryan, m. (2020). in ai we trust: ethics, artificial intelligence, and reliability. science and engineering ethics, 26(5), 2749-2767. santoni de sio, f., & mecacci, g. (2021). four responsibility gaps with artificial intelligence: why they matter and how to address them. philosophy & technology, 34, 1057-1084. sarac, m., & hassan, m. k. (2020). islamic perspective for sustainable financial system. istanbul university press. saveliev, a., & zhurenkov, d. (2021). artificial intelligence and social responsibility: the case of the artificial intelligence strategies in the united states, russia, and china. kybernetes, 50(3), 656-675. shahrouri, a. d. m. (2023). the cultural and social impact of artificial intelligence on islamic law standard: a fundamental purposeful study. conference on sustainability and cutting-edge business technologies, susar, d., & aquaro, v. (2019). artificial intelligence: opportunities and challenges for the public sector. proceedings of the 12th international conference on theory and practice of electronic governance. tyson, l. d., & zysman, j. (2022). automation, ai & work. daedalus, 151(2), 256-271. unal, i. m., & aysan, a. f. (2022). fintech, digitalization, and blockchain in islamic finance: retrospective investigation. fintech, 1(4), 388-398. uunona, g. n., & goosen, l. (2023). leveraging ethical standards in artificial intelligence technologies: a guideline for responsible teaching and learning applications. in handbook of research on instructional technologies in health education and allied disciplines, 310-330. igi global. yigitcanlar, t., butler, l., windle, e., desouza, k. c., mehmood, r., & corchado, j. m. (2020). can building “artificially intelligent cities” safeguard humanity from natural disasters, pandemics, and other catastrophes? an urban scholar’s perspective. sensors, 20(10), 2988. yigitcanlar, t., corchado, j. m., mehmood, r., li, r. y. m., mossberger, k., & desouza, k. (2021). responsible urban innovation with local government artificial intelligence (ai): a conceptual framework and research agenda. journal of open innovation: technology, market, and complexity, 7(1), 71. yuspin, w., wardiono, k., budiono, a., & gulyamov, s. (2022). the law alteration on artificial intelligence in reducing islamic bank’s profit and loss sharing risk. legality: jurnal ilmiah hukum, 30(2), 267-282. zemni, b., awwad, w., & bounaas, c. (2020). audiovisual translation and contextual dictionaries: an exploratory comparative study of reverso context and almaany uses. asian efl journal research articles, 27(5.1), 274309. zhang, c., & lu, y. (2021). study on artificial intelligence: the state of the art and future prospects. journal of industrial information integration, 23, 100224. zhao, w., zhang, j., liu, x., & jiang, z. (2022). application of iso 26000 in digital education during covid-19. ain shams engineering journal, 13(3), 101630. pa ge 1 pa ge 29 american journal of smart technology and solutions (ajsts) a comparative study of on-campus and off-campus internet facilities fe nangcas jalon-de la cruz1, cherlyn sarguilla abillar2 volume 2 issue 1, year 2023 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: march 02, 2023 accepted: march 11, 2023 published: march 18, 2023 this study aimed to compare the lived experiences of the graduate school students in accessing internet facilities inside and outside the university campus. qualitative method using phenomenology as a design was utilized. the internet services in two campuses were compared. aspects of comparison between the two lenses started with the on-campus followed by the off-campus were presented in terms of; internet connection: dawdling versus fast; wi-fi connection: difficult to access versus accessible; network: low versus hassle-free; accessibility: limited versus limitless; adequacy of facility units: discomfort versus comfortable; expert’s and mentor’s guidance: available versus difficult to avail; research environment: safe versus unsafe; power supply: sustainable versus susceptible to interruption. these results implied an urgency of the university to look into the problem that the students are facing in accessing internet facilities for their research activities. keywords phenomenology, qualitative research, on-campus internet facility, off-campus internet facility, wi-fi, internet facilities 1 college librarian, st. francis xavier college, san francisco, agusan del sur, philippines 2 teacher iii, don manuel h. gutierrez sr. elementary school, matina district, davao city, philippines * corresponding author’s e-mail: lady.feus3819@gmail.com introduction as a sequel to the investigation we conducted on the use of off-campus internet facilities by the graduate students of the university, this inquiry was borne. relative to their revelations, their experiences on the off-campus utilization were compared to the on-campus. the researchers observed the inadequate internet facilities and services of the university. with our honest intention as an opener to our beloved alma mater, we would like to mention what we personally identified in the libraries and internet laboratories like; the frequent low bandwidth or slow connection which was one of the major constraints that crippled fast internet access; the limited number of computer units shared by the undergraduate, and research level students in computer laboratories; no separate library with internet connectivity; and limited use of wifi since password was protected. the insufficiencies mentioned earlier pushed them to rely on the internet connectivity of home-based broadband, internet cafés, free wi-fi zones, pocket wi-fi and mobile-equipped cellular phone for research purposes. during class days, student-users opted to seek internet cafés’ services outside the campus for corrections and emergency printing of their academic requirements. students were obliged to spend money for loading to their pocket wi-fi with the intention to use this inside the school premises to comply class assignments. the research conducted by arthur and brafi (2013) outwardly hatched the idea on this problem aforementioned that the inadequacy of on-campus internet facility encouraged him to proceed with his research study because he observed that poor internet connection and limited number of computer units in the laboratories inside the campus were among the problems that were encountered by the students and teachers. both groups of users were pressured to avail of the offcampus internet facility as an alternative source, although they often found these expensive. we personally believe that this sequential study was deemed necessary to obtain information relative to their off-campus internet use. it is our hope and aspiration that the outcome of this investigation would be considered by the university as basis in the conduct of a new investigation that would reveal suggestions for the development of high speed internet connection inside the university. grand tour questions this study was guided by the following questions: 1. what are the lived experiences of the graduate school students in using the on-campus and off-campus internet facility? 2. how do the graduate school students compare their lived experiences between on-campus and off-campus internet facilities? literature review problems encountered by the students the study of arthur and brafi (2013) informed that internet users identified the hindrances purposely, to enhance the research, teaching and learning process. they mentioned that inadequate users’ skills were noted as the most leading constraints, followed by poor internet connection and limited number of computer units in laboratories inside the campus. according to the study, although internet cafés were open 24-hours daily the problem was low internet connection due to poor signal. they added that the cafés had other impediments such as poor management, and inadequate facilities due to equipment and maintenance costs. in connection, khan et al., (2016) gave an account of an https://journals.e-palli.com/home/index.php/ajsts mailto:lady.feus3819%40gmail.com?subject= pa ge 30 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 29-37, 2023 investigation about the issues experienced by the students in accessing information online. the list specified that slow internet connection, inadequate number of computers in laboratories, difficulty in information searches, inadequate knowledge and skills in evaluating information and information overload. he further elaborated that students limited knowledge on the use of digital book, electronic databases, and retrieval of information could hamper performance. the nigerian university students found difficulty in utilizing the internet for research purposes because of power interruption. this predicament hampered their research activities. survey results revealed that 70 percent of the students complained about the frequent loss of signal; 66 percent stated power outage; and 60 percent declared service was highly expensive. those who complained on slow internet speed, had trouble in judging accuracy of the information, experienced lag in downloading webpages and murmured on information overloading accounted to 50 percent respectively. around 10 percent, complained on limited cybercafés terminal stations and students’ accessing skills ( nwezeh 2010; adekunmisi et al., 2013). in like manner, audio (2016) said that many individuals expect when power is interrupted and computers are disconnected, it must be seen to it that there must be no information lost. unfortunately, it is not that simple for a lot of people; the risk is low, but when you’re using your computer even slightly more seriously and your data is equally important, you might want to consider using an uninterruptible power supply (ups). it’s also convenient to have your external universal serial bus (usb) disk, router, cable modem, telephone switch, or similar devices on the ups. this is possible when you own the unit or doing at home. however, if off-campus the safest way is to have your usb and have your data saved immediately after you accessed. in fact, nwokedi (2007) expounded that limited knowledge in searching banned internet use. he clarified that showing abilities to search for information could help a great deal in finding the right information for research. it could improve and enhance teaching and would inspire students to pursue with their research activities. the nigerian university identified various problems about the students’ complaints about the unequal internet connection access inside the campus. the report stated that while the staff and personnel could access anytime, anywhere in the university campus relative to their teaching and research activities, students were not. they could only access the internet through major access points such as the library and computer laboratory where the connection was very slow. the inequitable distribution of internet facility connection inside the campus prompted internet cafes to strategically position their sites to cater to students’ research needs (ani, 2010). a similar study reported that the use of facility outside the campus was more practical, efficient and dependable compared to the internet facility on-campus (fasae & aladeniyi, 2012). in sunyani municipality, a study was led on the adequacy and proficiency of internet utilization. among the major issues distinguished that influenced the students’ dissatisfaction were the slow internet connection, the limited number of computer units and slightly damaged, students’ lack of knowledge in browsing information, limited time in accessing connection at internet cafés and discomfort in room temperature at café houses (arthur & brafi, 2013). a study on the usage of ict project at nigerian university libraries was observed to be poor in serving the students, faculty and staff. the report suggested administration must see behind the present situation in the library so services be improved. some researchers enlisted a particular organization where nigerian universities could avail dependable internet connections like nigerian universities network (nunet), national virtual (digital) library project (nvlp), and nigerian virtual library consortium (vlc) (ani, esin & edem, 2010; etim, 2006). in the university of nigeria, it was discovered that 77.8 percent had internet laboratories in their libraries serving the entire university campus faculty and staff, students and other member of the institution. the web network was discouraging in some colleges and universities with 30 percent; and 19.5 percent in tuition based schools. the essential reason was absence of budgetary allotment to enhance internet connection (baro & asaba, 2010; chaputula & boadi, 2010; chaputula, 2011). the poor internet connection and the limited number of pc units in the research centers inside the campus were among the noteworthy distinguished issues experienced by the students. despite the cost students, faculty and staff opt to seek connection outside the university so that they can connect to their families and friends while studying (arthur & brafi, 2013). a study by georgetownu library (2016) reminds the clients to remember that practically anybody can post information anytime he wishes on the web. it is regularly hard to determine origin of web sources, the authenticity of authorship, and regardless of the fact that the creator is recorded, he or she may not generally speak to the truthfulness of his work. it is the sole responsibility of the user to verify information posted and assess viably on the information accuracy and its truthfulness. essentially, shahin, balta, and ercan (2010) urged the university students to utilize the e-journals, e-books and other scholarly databases as wellsprings of data and sources of information for course-related researches to guarantee dependability and reliability of resources. for this situation, data presented required evaluation as to its legitimacy and exactness. unfortunately, it is exceptionally hard to assess the data, since it requires client’s the ability and expertise. internet cafés access points or computer stations is not desirable. computer terminals must be positioned to ensure a client’s security. simple steps such as erasing settings on log-out, deleting of content on hard drives and https://journals.e-palli.com/home/index.php/ajsts pa ge 31 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 29-37, 2023 giving warning or caution to clients of the dangers when they exposed themselves to public must be observed. these are the basic methods for helping clients to comprehend and alleviate the dangers of utilizing these systems. however, it is the individual and frequently their manager who are liable to end up in a bad position for neglecting to find a way to secure and protect the information online (techtarget, 2016). in relation, ring (2016) said that the greatest concern toward most pc clients is protection and privacy. wifi hotspots in most open areas are unsecured, which implies that others could possibly access the data on your computer. in the event that you will forget to log out in an open work station, somebody could get into your e-mail or other delicate personal records. on a physical level, the individual beside you in an internet café can easily look over your shoulder while navigating the web and might have the chance of hacking your personal information or anything you did browse on the net. open café terminals are not ideal for accessing information relating to work and other personal business–related transactions. methodology participants a total of eleven participants were included in the study from the graduate school. all were interviewed and the data collected were recorded and transcribed for data analysis. design this study employed a qualitative method via phenomenology inquiry. to exemplify better understanding about the idea willis (2007) simplified the perception that phenomenology as a method, construes experiences through attending and listening to the numerous storylines of the informants. the method inspected the phenomenon over the distinctive eyes of the research participants. as an approach, it attempted to understand the hidden meanings and essences of an experience as well as how participants made sense of that experience. as pereira (2012) established that in order for a phenomenological study to be arbitrarily valid it must be rigorous and offer views and ideas in terms of credibility and the clarity of a phenomenon. this method sparked our interest as researchers with the desire to examine the lived experiences of the graduate school students of the university in accessing on-campus and off-campus internet facilities for research activities. the phenomenon employed purposive sampling methods. as, cresswell and clark (2011) defined purposive sampling as the involvement of the individuals who are knowledgeable in a particular phenomenon of interest. it recognizes and selects people who have direct experiences with the situation. we endeavored what sargeant (2012) affirmed that data could be collected from individuals or group of people who were interviewed, an observation or from a document. we recorded, transcribed and analyzed following creswell’s procedure of material organizing, to classify the text data into segments and segmented sentences into categories. i formulated a thematic coding process where identified themes would be classified and filled in. i used this process to identify the themes we classified all similar ideas and categorized into core ideas. trustworthiness was established to make this study a reliable craft. as lincoln and guba (1985) defined trustworthiness is the “truth value” of the study’s findings. it was how accurately the investigator interpreted the participant’s experiences. the four criteria of lincoln and guba (1985); credibility, transferability, dependability, and confirmability were followed to evaluate the research findings. we attempted to pick an important criterion of what ary, jacobs, razavieh and sorensen (2010) presented on several criteria to test the dependability of the findings. this included audit trail, a code-recode strategy, stepwise replication, triangulation and peer examination. results and discussion it could be deduced from the interview results that accessing on-campus and off-campus internet facilities were intertwined in their studies. the storylines were presented and confined within the bounds of the research questions carefully formulated to serve as an instrument guide in the conduct of the in-depth interview and focus group discussions. the point of inquiry was centered on comparison of the graduate school students’ oncampus and off-campus lived experiences in the use of internet facility. the participants narrated these experiences through individual interviews and focus group discussions. comparison of the graduate school students’ lived experiences on the use of on-campus and off-campus internet facilities table 1: matrix of comparison between the graduate school students lived experiences on-campus and off-campus internet facility use in clearer presentation aspects of comparison lived experiences on-campus off-campus internet connection dawdling fast wi-fi connection difficult to access accessible network low hassle-free accessibility limited limitless adequacy of facility units discomfort comfortable https://journals.e-palli.com/home/index.php/ajsts pa ge 32 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 29-37, 2023 the table shows the matrix of comparison between the graduate school students lived experiences on-campus and off-campus internet facility use in clearer presentation. internet connection. dawdling versus fast participants compared their experiences in the oncampus and off-campus utilization of internet facility. they wanted fast searching speed to cope with class work deadlines. fast internet connection was the answer to their dilemma, and they got from off-campus cafés. participants to this study had varied stories in this regard. in the on-campus, students who were desirous to comply school work on time got disappointed with slow internet connection, and that to them, it was a waste of time. a participant shared that in accessing internet inside the campus he experienced dawdling internet connection. he added that his search was delayed due to slow connectivity. all he had to do was to engage in searching to entertainment sites like twitter, skype and youtube, whatsapp, linkedin, viber that cause to delay his time. with the trend of dawdling connection, research became difficult and waste of his time. while in the off-campus, student participant mentioned that he was compelled to seek the services of the internet cafés, in home-based broadband and in wi-fi connected areas. the fast internet connection he enjoyed in café haus has facilitated his searching and hastened the accomplishment of his class outputs. with dawdling internet connection, an investigation conducted by ojo (2013) on the use of internet facilities among the tertiary institution students discovered that most of the sites visited by the students were not academic and that they spend many hours on the websites that are not educationally productive. in relation, another study by wang (2015) explained that if webpages open slower than usual or downloads seem to last for ages, these are due to some factors that could affect your internet speed.in contrary, fast access to off-campus internet facility portrayed different experience as reflected in the study conducted by jerome (2013) mentioned that high speed internet connection made internet access speedier, simpler and more pleasant. wi-fi connection wireless services have introduced a new era in communication for the education community, and this new concept of mobility within the educator’s daily environment is having a major impact on student learning. wireless technology brings the primary benefit of mobility to traditional class activities. students and teachers’ laptop computers with wireless cards can be moved wherever needed in the local area and still access educational materials on the school’s server. a wifi network will let you connect smartphones, tablets, kindles, nooks, and other devices to the network and the internet, while a hardwired network will not. additionally, if you use a laptop computer, a wi-fi network gives you the ability to use it anywhere in the office or in the house, instead of just in the immediate vicinity of a connection port for the hardwired system. difficult to access versus accessible the students who used internet services inside the university expressed their sentiments: a participant elucidated his feeling that it was difficult for him to access internet connection inside the campus due to low capacity of wireless connectivity. wi-fi was password protected. if he could access, it was a very slow. it was always a delay in complying with the teacher’s requirements. in relation, another student mentioned that it was difficult to collaborate with their classmates on a class project. he could not communicate promptly because of difficulty in finding the “hot spot” or wi-fi connected area. opposite with, the off-campus students enjoyed the accessibility in the café houses and wi-fi zone areas in malls, in transport service, hotels and in coffee shops. he said that access is faster and easier everywhere. once he found the information he needed he immediately saved through the flash drives or usbs. he stated further that he had enough time with social networking for self-entertainment. since he could finish earlier, he could submit requirements on or before deadlines, which meant less pressures, less stress, more time for fun and relaxation. we believe that the participants experiences in accessing internet facility inside and outside campus were significant. wireless connectivity are revolutionizing communication and can benefit both schools and students in many ways. in addition, internet ready mobile gadgets and sets of computers give students one-to-one access. students can continue to collaborate on a class project with other students while outside the classroom, school or campus and can send the results to their instructors from anywhere they find a “hot spot” in wireless public access, increasingly offered by businesses, libraries and other facilities. a related literature reinforced this idea that the quality of the student computing experience has become an important decision-making factor for brazilian students in selecting a university. in the tic educacao survey, 96 percent of teachers in brazilian public schools use internet resources to prepare their classes (pr newswire, 2015). in fact, internet satisfaction is still a problem for some students, especially residents who use the internet for various activities including contacting home, social media and streaming videos (kraemer, 2015). compared with the in-campus wi-fi connectivity, a study conducted on the availability of wi-fi in public areas expert’s and mentor’s guidance available difficult to avail research environment safe unsafe power supply sustainable susceptible to interruption https://journals.e-palli.com/home/index.php/ajsts pa ge 33 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 29-37, 2023 have changed the face of internet accessibility. beside the noteworthy advantages that the connections carried with critical portability, the old system on getting connection altered. it demonstrated that the likelihood of wireless internet connections in public places would encourage the chance to lessen eyetoeye communications and social interactions (sanusi & palen, 2008). network bandwith as far as network bandwidth is concern, the on-campus users experienced low bandwidth while those who used off-campus had hassle-free moments. bandwidth is likened to lanes on a highway. it allows more traffic to go through at once, while still retaining high speeds. however, oftentimes many requests must go through a network simultaneously and a low bandwidth connection will severely slow speeds. a high bandwidth connection can service many requests and users without sacrificing precious speed. low versus hassle-free an informant expressed his frustration that the internet facilities inside the campus has low bandwidth internet connection that contributed to their delayed submission of class requirements. due to slow network performance his online requests were not served promptly, searches were not immediately accessed. another one voiced his experienced that he had to wait for high bandwidth connection that could serve many requests and users without sacrificing precious speed. some articles and other related materials that were accessed but could not be downloaded. transferring of data accessed was difficult. all these could be attributed to low bandwidth. the participants confessed, they experienced slow application performance and decreased data transfer capability. uploads and downloads were painstakingly slow. watching videos that constantly freeze was extremely frustrating. whereas, the off-campus internet facility users expressed that a connection with plenty of bandwidth available provided a smooth, responsive user experience. to them that experience was hassle-free. a student explained that the more available bandwidth the more he enjoy internet access on a very fast and satisfying tone. he was able to accomplish his goal, and enabled him to do multi-task and maximize productivity by running multiple applications simultaneously. the most common problem related to internet use was the low bandwidth (72 percent) and retrieval problems (bankole, 2013). obaje, sani and lawal (2008) study at the university of jos, nigeria revealed that students and faculty members who queued up could not access internet in the library and that the internet was used mainly for research and email. the findings of chitanana (2012) study have demonstrated that more awareness education on bandwidth management is needed for universities to make it a priority. increasing bandwidth without adequate network management is wasteful and reduces its value. the absence of effective bandwidth management strategies poses serious challenges for almost all universities. accessibility in terms of accessibility, connection was difficult in incampus while accessible in the off-campus. limited versus limitless on-campus internet facility users expressed their disappointments with the limited access they experienced in doing their research activities. one felt that he was defeated by time and expenses. when their teachers gave to them the term papers, they actually started browsing for possible inputs. however, with the limited number of computer units with the rule on “first come, first serve” basis, and they had to follow the long lines. “limited access” was on his computer screen.. with regards to the off-campus, students who used internet facility as an alternative source testified their enjoyment with the limitless accessibility. he could access to almost all websites using all search engines. almost all topics that he would like to search were easy to find. social networking sites such as facebook, skype, twitter, blogs, whatsapp, viber, nimbuzz to name a few are easy to access. on the contrary, one student expressed his worries that there are certain cons and dangers relating to the use of internet. the need for internet access was critical in the acquisition of academic information in order to complete class assignments. although the required information could be acquired through outside alternative sources, preference for using the internet inside the university was expressed. with their experience of limited access, the foreseen benefits of time saving and access to wide variety of information available through the internet were not enjoyed. the connection was poor. even if we connect to internet, it says very limited access. i have to keep restarting my laptop after trying to connect a number of times to see if i get access. the worst part is when i am doing an online assignments or quizzes and the internet decides to stop working (university wire, 2014). adequacy of facility units in terms of adequacy of the facility, the on-campus users experienced discomfort, due to inadequate facilities, while those off-campus were enjoying the comfort of having adequate facilities like personal computers, internet-ready cellular android phones, ipads, tablets, and other gadgets. discomfort versus comfortable a participant expressed that the limited number of computer units hampered his research activities inside the campus. computers were occupied by researchers almost every time he had to go to the university to login. other pc’s are not updated and connection is very slow. he could not comply his requirement on time due to the delay of his research activities. another participant expressed his disappointment and said that computer https://journals.e-palli.com/home/index.php/ajsts pa ge 34 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 29-37, 2023 laboratories were fully booked each time he tried to use the internet. the library has a free computers but with limited units it cannot accommodate all users. opposite to the on-campus users confessed that the services provided by the internet cafés, home-based broadband, wi-fi zone areas, office internet connection to avail desired websites for needed information was fulfilling and comfortable. one said that he access internet from the off-campus to download videos and pictures for his lessons in the class. through this initiative he feel free and comfortable. another one further stated that the off-campus facility is giving excellent service to users. where downloading is very fast, connection could be accessed anywhere in malls, internet cafés, in homebased broadbrand, wi-fi zone areas such as public places even in parks and amusements halls, hotels, airports, cars and buses. a related literature supported this experience on the inadequacy of facilities inside the campus as bamigboye and agboola’s (2011) study on the availability and accessibility of internet facilities in nigerian university libraries indicated that a majority of respondents found that there were not enough computers. this was something the administrators of those institutions should note. in fact allen (2016) conversely said that if you have highspeed equipment but a slow connection, you will not get the advantage the equipment can offer. you always get the lesser speed. no matter how high or slow your internet connection is, you must have the right equipment to use it properly. you need a decent wireless router to convert the signal to wireless and broadcast it throughout your location. the modern you use to connect your computer to the internet may have built-in wireless router capability. expert’s and mentor’s guidance availability of experts versus inavailability guidance of experts for consultation was available in on-campus but those off-campus users found difficulty in this regard. students accessing information inside the university campus are enjoying the availability of experts for guidance and consultation. they imparted their knowledge to novice researchers. as one said that he shared his profession as a librarian to others when he assisted the students in looking for better information resources. as we were listening to the interview, one participant captured our attention when he mentioned the importance of my profession that a librarian is someone who had the expertise in research and articles to be researched. to the students he is important because it was different if there would be someone who could provide guidance than merely depend on personal knowledge. however in the off-campus there would be no experts available for consultation. one shared, there were some helpful aides in the counter, but with many users, he could not consistently secure the in-charge attention. besides the internet in-charge is not a librarian or expert in the field. he added that the off-campus users had the hard time locating for the right search engine so they needed guidance from experts in the field. sometimes he stayed longer in the cafés to seek an advise which to mean extra effort, time and money. reliable websites and search engines were easier to locate for the needed information relative to his assignments. whereas, the experience of those students who sought the off-campus facilities were left to their own understanding. as researchers, we would like to share some points to consider when judging the reliability of information on the net as clearly elaborated by erma wood carlson library (2016) as to who created the site? who supported it? on the page you are referring to, that individual or association ought to be recognized, that individual’s capabilities ought to be specified, and different avenues for verification ought to be open for public scrutiny. are there no prejudices or bias? who published the site? is there any publicizing or advertising? does the page appear to be professionally designed? is the writing trying to persuade you to buy something? is contact data given by the publisher? in the event that the page is supported by a trustworthy individual or association, there ought to be some other approach to check that notoriety like email address or postal location. is there a copyright image or symbol included on the page? provided that this is true, who holds the copyright? what is the reason for the page? why is this data being posted, as an open administration, as a news source, as a research tool for academics or as an approach to catch up an attention? how the page was organized? is the information on the page primary or secondary? is it a report of facts, making it primary or secondary information, or is it an internet newsgroup discussion? can you verify the information on the web page? can you check the page’s bibliography against the library’s holdings or check the information against a source in the library? what is the quality of information provided on the website? timeliness when was the website first published? is it regularly updated? check for dates at the bottom of each page on the site. what type of other sites does the website link to? are they reputable sites? if the author references online material, does it provides links to the material referred to? what type of sites link to the website you’re evaluating? is the website being cited by others? if you are worried that the information may lack credibility, try starting with a source you know is reputable. finally, remember that even though a page might not meet your standards as a citable source, it may help you generate good ideas research environment the traditional library is gradually becoming a thing of the past as cheaper and more up-to-date information materials become available on the internet. libraries are faced with immense challenges. access to information https://journals.e-palli.com/home/index.php/ajsts pa ge 35 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 29-37, 2023 can stimulate change and create a research environment that makes learning more meaningful and responsive. safe versus unsafe the primary focus for the students should be on maintaining a safe and secure internet using environment. a user expounded that his internet journey inside the university was fulfilling because he was assured of a safe and secured environment. students who seek information inside the university campus occur with “over-theshoulder” adult supervision. one added that the use of electronic communication devices in a fully secure environment gave him peace and wholesome satisfaction. beyond all these, those students who used the internet in the off campus felt uncomfortable. another informant declared that using internet café was very annoying. sometimes, boys were very noisy even if they had headphones. they sang, giggled with boisterous laughters. another participant shared the same experience, he said that he observed other issues related to the irresponsible use of the internet. he witnessed an intentional copying of information verbatim which to mean plagiarism. what the on-campus users experienced was the opposite outside. we were reminded of the misuse of the internet by students in accessing information from the internet café was very rampant. in fact, access of potentially harmful material, that included copyright infringement, plagiarism, computer security violations (hacking, spreading viruses), violation of privacy, harassment, stalking, and dissemination of harmful speech or other violent or abusive material is illegal and immoral in nature. the off-campus users whole experience pictured the unsafe environment. this impression is hatched in the research conducted by techtarget (2016) which said that computer terminals must be positioned to ensure a client’s security. simple steps such as erasing settings on log-out, deleting of content on hard drives and giving warning or caution to clients of the dangers when they exposed themselves to public. these are the basic methods for helping clients to comprehend and alleviate the dangers of utilizing these systems. in relation, ring (2016) said that the greatest concern toward most pc clients is protection and privacy. wi-fi hotspots in most open areas are unsecured, which implies that others could possibly access the data on your computer. in the event that you will forget to log out in an open work station, somebody could get into your email or other delicate personal records. power supply sustainable versus susceptible to power failure. the power supply affects the efficiency and efficacy of services in libraries. access to online research is powerdriven and its success or failure will depend solely on its electric service sustainability. power supply inside the campus is maintainable while in the off-campus is vulnerable to brownouts. in this study, the students who worked on their research assignments inside the campus were contented to have sustainable power supply. participant said that his stay in the campus using the available facilities was fulfilling, knowing that his work would not be hampered by power interruptions. in relation, one added that the presence of generators helped maintained the power supply. one said that he was very satisfied with the presence of the electric generators inside the university. he could continue working on his research until he finish. everything was taken cared of by the university. on the other hand, the off-campus users suffered frequent power interruptions. another one elaborated that when there were brownouts or blackouts there was no available electrical power to substitute due to the absence of generators. if there was damage to electrical equipment there was no available expert to repair. the tendency was for the users to leave the cafés without finishing the work. another student aired his frustration, the payment in internet cafés was non-refundable, despite services were suspended due to power failure. he elucidated that for many times his important data files were not yet save were brownouts occurred leaving him frustrated. hearing all these predicaments, a literature speaks of the same problem when brownouts, more typical than power outages or blackouts, cause equipment failures, incremental harm, diminished hardware stability and information data loss. power outages and voltage fluctuations can cause quick information misfortune and system crashes with no chance to spare critical records, which means important data can be lost in a split of a second (tripp-lite, 2016). implications for practice the comparative lived experiences between on-campus and off-campus internet users for their educational and other pursuits were presented. alarming was their testimony that off-campus experiences provided them pleasant feelings in internet use as compared with on-campus. there were more enjoyed privileges on fast connection, quick accessibility to wi-fi, limitless accessibility, high bandwidth; and adequate facilities and services which they have found to be substantially insufficient inside the university campus. in support for students’ experiences in accessing internet connection outside the university a review of literature by jerome (2013) said that high speed internet access makes it faster, easier and more enjoyable to browse online. websites load in the blink of an eye, and clicking from one site to the next is a breeze. with high speed internet, uploads are also fast and easy. this allows you to take full advantage of the cloud. upload photos, documents and other files to the cloud for safekeeping. uploading files to the internet is a lot safer and more secure than storing them on physical disks, which can be lost, stolen or damaged. your high speed connection will also allow you to quickly and easily share videos and photos on popular social networking sites like facebook, tumblr, instagram, twitter and others. https://journals.e-palli.com/home/index.php/ajsts pa ge 36 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 29-37, 2023 implicitly, a sequel to this investigation implied a call to go deeper into the possible suggestions of the students to improve the internet facilities and services of the university should be conducted. henceforward, this study is necessary in order to find appropriate answers to the perennial problems on slow internet connectivity inside the campus. implications for future research the current study has some limitations which should be covered for relevant researches in the future. the researchers were convinced that perhaps future researches should consider increasing the number of participants to provide a better and more convincing implications if the same study would have a sequel. probably, future researches may be extended to undergraduate and post-graduate students in the university. therefore, when conducting relevant studies in the future, the researchers should also include the suggestions of the students as their participants so as to find out how to improve the internet facility and services in the university. conclusion the revelations of the study pointed out that the offcampus internet facility served better to students when compared with the on-campus internet facility. the study findings might be a good reference for the university and library authorities to develop the internet facility in the campus. it may help if the administration will increase the speed of the internet access and provide a dedicated server with strong connection for students’ optimum use. references allen, j. m. (2016). dealing with wi-fi experience .the magazine of the senior lawyers division, american bar association, 26(1), 43-46. ani, o. e; edem, m. b; & ottong, e. j (2010). analysis of internet access and use by academic staff in the university of calabar, calabar, nigeria. library management, 31(7 ), 535-545. ani, o.e., esin, j. e. and edem, n. (2010). adoption of information and communication technology (ict) in academic libraries: a strategy for library networking in nigeria, the electronic library, 23(6), 701-8. http://search.proquest.com/ docview/747936506 arthur, c., & brafi, p. o. (2013). internet use among students in tertiary institutions in the sunyani municipality, ghana. catholic university college of ghana. library philosophy and practice (e-journal). http://digitalcommons.unl.edu/libphilprac/859/ ary, d., jacobs, l. c., razavieh, a., & sorensen, c. k. (2010). introduction to research in education (8 ed.). new york, ny: hult rinchart & wiston. audio, d.i.y (2016). why power failures for your data. http://www.halfgaar.net/why-power-failures-arebad-for-your-data bamigboye, o. b. & agboola, i. o. (2011). availability and accessibiligy of internet facilities in nigerian universities: a case study of two federal university in south west nigeria. library philosophy, and practice. http://unllib.unl.edu/lpp/bamigboye-agboola.pdf baro, e.e. & asaba, j.o. (2010). internet connectivity in university libraries in nigeria: the present state, library hi tech news, 27(9), 13-19. chaputula, a.h. & boadi, b.y. (2010). funding for collection development activities at chancellor college library, university of malawi, collection building, 29(4), 142. chaputula, a.h. (2011). impact of the global economic crisis on academic libraries in malawi: a case study of university of malawi and mzuzu university libraries, library management, 32(8/9), 565-78. chitanana, l. (2012). bandwidth management in universities in zimbabwe: towards a responsible user base through effective policy implementation. international journal of education and development using information and communication technology. (ijedict) 8(2), 62-76. cresswell, j. w., & clark, p. l. (2011). designing and conducting mixed method research (2nd ed.). thousand oaks, ca: sage. etim, f. (2006). resource sharing in the digital age: prospects and problems in african universities. library philosophy and practice, 9(1), 12-19. http:// search.proquest.com/docview/747936506/ george town u library (2016). evaluating internet resources.http://www.library.georgetown.edu/ tutorials/research-guides/evaluating-internet-content gultiano, king, orbeta & gordoncillo (2010). internet use among filipino public highs school students. internet, the gearing up internet literacy and access for students (gilas) ayala foundation. haberkorn, j. (2000). focus paper: school librarian shortage lis 405le. khan, s. a., khan, a. a. and bhatti, r. (2016). internet access, use and gratification among university students: a case study of the islamia university of bahawalpur, pakistan. chinese librarianship: an international electronic journal, 32. kraemer, b., jr. university wire (2015). wi-fi service attack raises concerns over internet satisfaction. lincoln & guba (1985). trustworthiness, credibility, dependability. online reference book for criteria for judging qualitative research. studies. nwezeh, c. m. t. (2010). information and communication technologies for educational development: the case of cyber cafés at obafemi awolowo university, ile-ife, nigerian. library philosophy and practice. http://www.webpages.udidaho.edu/~mbolon/ nwezeh2.htm obaje, a. m., sani, a. & lawal, v. (2008). internet access and usage by staff and students: a case study of university of jos main library, jos, nigeria. the information technologist, 5(1), 160-70. https://journals.e-palli.com/home/index.php/ajsts pa ge 37 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 29-37, 2023 ojo, a. i. (2013). the use of internet facilities among the tertiary institution students, ogun state institute of technology, igbesa, nigeria, under study. internet technologies and applications research, 1(1), 1-5. www. jlisnet.com pereira, h. (2012). regor in phenomenological research : reflections of a novice nurse researcher. nurse researcher, 19(3), 16-19. https://dissertationrecipes.com/wpcontent/uploads/2011/04/phenomenologicalresearch.pdf pr newswire [new york] (2015). brazil’s public universities make the smart move to better wi-fi with ruckus wireless: nearly one-third of brazil’s public universities standardized on ruckus smart wi-fi to deliver more reliable connectivity to tens of thousands of students and staff. ring, j. (2016). the advantages & disadvantages of internet cafés. http://www.ehow.com/info_8208732_ advantages-disadvantages-internet-cafes.html sanusi, a., & palen, l. (2008). of coffee shops and parking lots: considering matters of space and ppace in the use of public wi-fi. computer supported cooperative work, 17, 257-273. sargeant, j. (2012). qualitative research part ii: participants, analysis, and quality assurance. j grad med educ. 4(1),1– 3. https://doi.org/10.4300/jgme-d-11-00307.1 techtarget (2016). businesses need clear computer use policies and need to ensure staff are properly trained in data protection. tripp-lite (2016). common power problems & power protection solutions. https://assets.tripplite.com/ white-paper/common-power-problems-and-powerprotection-solutions-white-paper-en.pdf university wire (2014). students voicing concerns about wi-fi service on campus. https://search. proquest.com/docview/1651553260/fulltext/ b2514b7aab804d1epq/77?accountid=31259 wang. a. (2015). what factors afect your internet speed? opera news. willis, j. (2007). foundations of qualitative research: interpretive and critical approaches. thousand oaks: sage publications. https://journals.e-palli.com/home/index.php/ajsts pa ge 1 pa ge 49 american journal of smart technology and solutions (ajsts) parking occupant management system using qr code solutions with aes algorithm albert i. luzuriaga1, gawain destiny a. de groot1, jerold e. cortez1, nathaniel u. babanto1, meridel c. ejusa1* volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.4717 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: march 12, 2025 accepted: april 17, 2025 published: august 28, 2025 this study presents a parking occupant management system utilizing qr code solutions integrated with advanced encryption standard (aes) technology to address inefficiencies in manual vehicle cataloging within educational institutions. automating this process improves accuracy, efficiency, and security while ensuring adaptability across diverse school parking environments. qr codes facilitate seamless logging of vehicle entries and exits, embedding contact information that enables security guards to communicate directly with vehicle owners in case of issues. the incorporation of aes encryption provides robust protection for sensitive data, safeguarding it against unauthorized access. designed as a mobile application for android devices, the system empowers security guards to scan qr codes in real time, effectively recording vehicle activity. focusing on educational institutions in general santos city, this study demonstrates significant enhancements in parking management through automation and improved data protection. the system establishes a new standard for secure data management in parking operations and holds promise for application in sectors such as commercial and residential complexes. by offering a scalable solution that enhances efficiency while ensuring data security, it contributes to better resource management. future developments may explore security enhancements and expansion to platforms beyond android, increasing accessibility and catering to a wider range of parking management needs. ultimately, this system serves as a blueprint for modernizing parking management, paving the way for smarter, more secure urban environments. keywords aes encryption, data security, parking management system, qr code, vehicle cataloging 1 college of engineering and technology education, holy trinity college, general santos city, philippines * corresponding author’s e-mail: htc_ejusam@online.htcgsc.edu.ph introduction the increasing vehicle populations and urbanization pose challenges in managing parking operations on college and university campuses (dokania et al., 2020). this growing demand highlights the urgent need for efficient parking solutions. traditional manual cataloging methods fall short—they are slow, prone to mistakes, and lack strong security measures, resulting in operational hiccups and safety risks. to tackle this, this research proposes a qr-based vehicle management system paired with the advanced encryption standard (aes) encryption algorithm to enhance the handling of vehicle entries and exits in university parking areas securely and efficiently. this qr-based system with aes encryption offers a major improvement over outdated methods, delivering benefits to both campus parking facilities and vehicle owners. it strengthens security by encrypting qr codes with aes, safeguarding parking occupants’ data from unauthorized access or tampering—only a decryption key can unlock intercepted details. the system also optimizes vehicle entry and exit through quick qr scanning, cutting down wait times, improving traffic flow, and boosting overall parking efficiency by reducing human errors. while urban parking systems have embraced technological upgrades, campus parking has lagged, often sticking to manual processes with little use of qr technology. this gap has weakened communication between parking users and attendants, increasing risks like theft, vandalism, accidents, and legal issues. the main goal of this work is to create a secure, efficient, and user-friendly parking management system for educational institutions using qr code technology and aes encryption. by adopting automation and focusing on user needs, this research improves parking operations, enhances security, and fosters better communication, overcoming the drawbacks of manual methods. the author’s key contribution is showing how these technologies can modernize campus parking, ensuring lasting efficiency, safety, and convenience as campuses adapt to rising vehicle numbers and urban growth. literature review improving vehicle management systems on university campuses has become increasingly important in our rapidly changing environment. this literature review examines how quick response (qr) codes and the advanced encryption standard (aes) encryption algorithm can enhance the security and efficiency of campus parking management. by leveraging these advanced technologies, the review draws on significant contributions from prior studies to justify their integration into parking systems, highlighting their potential to streamline operations and bolster data security. qr codes in vehicle management as urban environments evolve, integrating qr codes pa ge 50 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 49-56, 2025 into vehicle management systems significantly enhances user convenience by enabling faster queuing processes and minimizing waiting times. lin, rivano, and le mouël (2017) classify smart parking ecosystems, identifying essential components and usage trends that underscore the effectiveness of qr-based systems in optimizing parking operations. their findings suggest that qr codes improve user experience and operational efficiency by streamlining data collection and vehicle tracking. similarly, barriga et al. (2019) explore the primary components and usage patterns of smart parking systems, emphasizing how qr codes, alongside sensors and software, enhance functionality and urban mobility. kadu et al. (2014) further address urban parking challenges by introducing a qr code-based smart parking system that identifies users, provides real-time parking information, reduces traffic congestion, and increases user satisfaction. these studies collectively highlight qr codes as a key variable in improving parking management efficiency. encryption and security the combination of qr codes with aes encryption addresses crucial security concerns in vehicle management systems. chai et al. (2023) stress the necessity of incorporating aes encryption into qr codes to protect sensitive information, preventing data leakage and unauthorized access while preserving system integrity. ajini asok and arun (2016) explore the use of aes128 encryption to secure private data within qr codes, mitigating risks like eavesdropping due to their visual nature. their approach ensures safe verification without delays, enhancing security in data exchange. additionally, agarwal and malik (2022) investigate steganography with qr codes and aes encryption, emphasizing the preservation of image quality to conceal data effectively. these contributions justify aes as a robust cryptographic mechanism for securing qr-based systems, supporting its adoption in campus parking contexts. performance and efficiency the efficiency of aes encryption is vital for its practical implementation in parking systems, particularly on resource-constrained devices. doomun, doma, and tengur (2008) assess aes in cipher block chaining (cbc) mode, finding that optimized implementations improve encryption speed by 12% to 30%, though with increased memory demands. almuhammadi and al-hejri (2017) analyze aes block cipher modes like electronic codebook (ecb) and cbc, evaluating encryption time and throughput to guide mode selection. vaidehi and rabi (2014) highlight ecb’s vulnerabilities and advocate for cbc to enhance security and effectiveness. these studies provide a foundation for optimizing aes performance, a critical variable for real-time parking applications. applications and case studies insights from related fields reinforce the applicability of qr codes and aes encryption in parking management. ferdiansyah, hadiana, and rakhmat (2021) demonstrate their integration in healthcare administration, where encrypted qr codes secure sensitive data, offering a model for parking systems. agun, rabie, and satoquia (2022) showcase a qr-based automated ticketing system for traffic violations, improving data collection and transaction efficiency—principles applicable to parking operations. gangurde et al. (2022) further emphasize the broader implications of these technologies, noting their benefits in security and user experience across contexts. these case studies support the hypothesis that qr codes and aes can transform campus parking management. hardware implementations efficient hardware implementations enhance aes applicability in parking systems. rachh et al. (2012) propose two aes architectures using composite field arithmetic, optimizing s-boxes and operations like mixcolumns for encryption and decryption. their designs improve processing efficiency, making them suitable for mobile and embedded systems in parking management. this hardware focus complements software-based security measures, reinforcing aes as a versatile solution. in conclusion, the literature underscores the critical role of qr codes and aes encryption in modernizing parking management systems on university campuses. these technologies address security, efficiency, and user experience challenges, justifying their adoption and suggesting hypotheses for further research, such as their impact on reducing parking delays and enhancing data protection. materials and methods this section outlines the materials, tools, and methodologies employed to develop and evaluate the parking occupant management system using qr code solutions with aes algorithm. the system was designed to automate vehicle cataloging, enhance data security, and improve parking management efficiency in educational institutions within general santos city. the approach is detailed below, structured into subsections for clarity, providing sufficient information to replicate the study while maintaining a concise narrative. system development methodology in developing the parking occupant management system using qr code solutions with aes algorithm, the researchers used the iterative waterfall model. the f low of this model takes you through system level requirements, requirement analysis, design, implementation, testing, integration deployment, and maintenance in a structured way. this is different from the traditional waterfall project management model, where each stage is fed feedback from all stages behind it, as opposed to the classic model comparison which just checks for similarities or differences between aspects of project management in practice. pa ge 51 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 49-56, 2025 requirement analysis the initial phase involved analyzing the existing manual parking management processes in educational institutions to identify inefficiencies and define system requirements. data were collected through interviews conducted with stakeholders, such as security guards and administrative personnel, from institutions like general santos doctors’ medical school foundation inc. for instance, interviews revealed that parking management relied on manual steps: occupants completed registration forms, administrators processed permits, and guards logged vehicle entries and exits manually. these processes were prone to errors and delays, necessitating automation. the collected data included occupant details (full name, contact number, address) and vehicle information (license plate number, type, color, brand, model), which were critical for system functionality and security verification. this information was securely stored in a mysql database, forming the basis for qr code generation and access control. system design and architecture the system was designed as a multi-component architecture integrating web and mobile platforms. key components included: frontend technologies ● vue.js with vuetify: used for the web interface, providing reactive data binding and pre-designed ui elements (e.g., buttons, text fields) for a consistent and intuitive user experience. ● flutter with material.dart: employed for the android mobile app, ensuring a clean and intuitive design aligned with google’s material design guidelines. backend technologies ● php native via restful apis: managed backend logic, such as form processing and database interactions, using http methods (get, post) with axios for communication. ● mysql: stored structured data, including occupant profiles, vehicle details, and parking logs. ● xampp: provided a local development environment with apache, mysql, php, and perl. cryptographic and qr code tools ● crypto-js: a javascript library implementing aes encryption in cipher block chaining (cbc) mode with pkcs7 padding, used to encrypt occupant and vehicle ids. figure 1: iterative waterfall model of software development life cycle figure 2: system architecture pa ge 52 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 49-56, 2025 ● encrypt.dart: a dart library for aes decryption in cbc mode with pkcs7 padding, used in the mobile app to decode encrypted qr code data. ● qrcode.js: generated qr codes embedding encrypted data. ● flutter_barcode_scanner: a flutter plugin for scanning qr codes on android devices the system architecture supported secure data flow, with encrypted occupant and vehicle ids embedded in qr codes, scanned by guards to log entries and exits. the entity relationship diagram detailed the database structure, comprising ten interconnected tables to manage figure 3: entity relationship diagram of the database occupant profiles, vehicle data, and parking logs efficiently. implementation details data collection and encryption occupants registered via a web interface, providing personal and vehicle details. these data were encoded into a json string (e.g., {“vehicleid”: 65, “occupantid”: 126}; encrypted using the aes-cbc algorithm implemented in crypto-js. the encryption process, detailed, involved: 1. generating a random 16-byte initialization vector (iv). 2. encrypting the json string with a predefined aes key (stored as an environment variable) in cbc mode with pkcs7 padding. 3. outputting a base64-encoded string combining the iv and ciphertext (e.g., mm/sbt5xtvvbuvwv16nslg= =:1tr4zmbeinxw+foro7magesa5mothdrlpi6 dy9nkfr5xs81s/9ii5rnsxisolu+j; appendix e, p. 77). this encrypted data was embedded into qr codes using qrcode.js, printed, and affixed to vehicles for scanning. qr code generation and scanning the qr code generation process linked encrypted occupant id and vehicle id to each vehicle, ensuring secure identification. the mobile app, developed in dart using flutter, utilized flutter_barcode_scanner to decode qr codes. upon scanning, the app extracted the base64encoded data, decrypted it using encrypt.dart, and retrieved the original json string via the aes key and iv, enabling guards to log vehicle actions. aes algorithm implementation the aes encryption adhered to the standard outlined by the national institute of standards and technology (nist) in [fips 197], implemented in cipher block chaining (cbc) mode as described by doomun et al. (2008). modifications to the standard included the use of a fixed aes key in base16 format and a randomly generated iv for each encryption to bolster security. in cbc mode, each plaintext block is xored with the previous ciphertext block before encryption, with the iv serving as the initial vector for the first block. the encryption process for each block i is defined as: ci=ek (pi ⊕ c(i-1)), with c0=iv the decryption process reverses this operation to recover the original plaintext, expressed as: pi= dk (ci )⊕c(i-1), with c0=iv pa ge 53 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 49-56, 2025 this implementation ensured both data integrity and confidentiality. the use of a random iv per encryption guaranteed that identical plaintext blocks produced distinct ciphertexts, significantly enhancing the system’s security. testing and evaluation the system underwent thorough testing to assess accuracy, efficiency, and security: ● unit testing: verified individual components (e.g., encryption, qr code scanning) using microsoft visual studio code as the primary ide for debugging. ● integration testing: ensured seamless interaction between web, mobile, and backend components. ● security testing: confirmed that encrypted qr codes were unreadable by third-party scanners (e.g., google lens), validating aes effectiveness. ● user testing: conducted with five respondents from educational institutions on october 18, 2024, using a survey. respondents rated interface, functionality, usability, experience, and security on a 1-5 scale, yielding a mean score of 5.0 (very satisfied) across all categories. testing focused on real-time logging accuracy, qr code scanning speed, and data protection, with results indicating high reliability and user satisfaction. deployment the system was deployed in educational institutions in general santos city, operating within the institution’s wlan for enhanced security. the deployment phase involved: 1. installing the mobile app on institution-provided android devices. 2. configuring the mysql database on a local xampp server to store parking data. 3. distributing qr codes to registered occupants (faculty, staff, students). deployment ensured that guards could scan qr codes to log vehicle entries and exits, with logs stored securely and accessible only within the institution’s network. data analysis data collected included parking logs (e.g., log id, vehicle id, timestamp, action type; and user feedback. logs were analyzed to evaluate system performance, such as entry/exit frequency and error rates. survey responses were quantified to assess user satisfaction and security perceptions. the aes encryption’s effectiveness was validated by its inability to be decrypted by unauthorized tools, ensuring data confidentiality. results and discussion this section presents the outcomes of the evaluation of the parking occupant management system using qr code solutions with aes algorithm, conducted within educational institutions in general santos city, and discusses their significance in the context of modern parking management challenges. the evaluation, based on a survey of five respondents, assessed the system across five key dimensions: interface, functionality, usability, experience, and security. the results indicate exceptional user satisfaction, with a consistent mean score of 5.0 (on a 1–5 scale, where 5 denotes “very satisfied”) across all evaluated criteria, highlighting the system’s effectiveness in automating vehicle cataloging and enhancing data security through aes encryption. summary of key findings the survey results demonstrate that the parking occupant management system excels in delivering a user-friendly, functional, and secure solution for parking management. respondents universally praised the system’s consistent interface design across web and mobile platforms, noting its intuitive layout and clear feedback mechanisms, such as pop-up messages, which enhance usability. the system’s functionality, particularly its ability to accurately process qr codes and update parking occupancy in real time, was rated highly, indicating seamless integration of essential parking management features. usability was a standout feature, with first-time users finding navigation effortless and qr code scanning smooth and efficient. in terms of overall experience, respondents reported significant improvements in convenience and time savings, attributing these benefits to the system’s reliable performance and the unobtrusive integration of aes encryption, which operates without compromising speed. most notably, the security dimension received unanimous approval, with users expressing strong confidence in the aes encryption’s ability to protect sensitive data, such as personal and vehicle information, during qr code scanning and transmission. as shown in table 1, the system achieved a total mean score of 5.0 across all sections, reflecting its exceptional performance in meeting user expectations. table 1 : summary of survey results section mean score verbal description interface 5.0 very satisfied functionality 5.0 very satisfied usability 5.0 very satisfied experience 5.0 very satisfied security 5.0 very satisfied significance of the results the uniformly high satisfaction scores underscore the system’s success in addressing key challenges in parking management, particularly within educational settings. the flawless mean score of 5.0 across all dimensions suggests that the system not only automates the manual cataloging process effectively but also elevates user trust through robust security measures. a critical factor in this success is the implementation of aes encryption, which ensures that sensitive data—such as occupant and vehicle ids embedded in qr codes—remains protected against unauthorized access. this is particularly significant in pa ge 54 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 49-56, 2025 an era where data breaches and privacy concerns are escalating, especially in digital systems managing personal information. the respondents’ confidence in the system’s security, as evidenced by their approval of features like the prevention of qr code decoding by third-party tools (e.g., google lens), highlights aes encryption as a cornerstone of the system’s value proposition. beyond security, the system’s ability to improve parking operations through qr code technology offers practical benefits. the real-time logging of vehicle entries and exits, facilitated by efficient qr code scanning, reduces human error and administrative overhead compared to traditional manual methods. this efficiency is vital for educational institutions, where parking facilities often face high demand and limited resources. the seamless user experience, from intuitive navigation to time savings, further positions the system as a user-centric solution that enhances operational flow without sacrificing security. comparison with recent developments and novelty in the context of recent advancements in parking management and iot-based systems, the integration of aes encryption with qr code technology distinguishes this work from existing solutions. recent literature emphasizes the growing importance of data security in smart systems, with studies noting that many parking management platforms lack robust encryption, leaving them vulnerable to unauthorized access (e.g., chai et al., 2023, on secure qr-based systems). unlike these systems, the parking occupant management system leverages aes encryption to safeguard data at rest and during transmission, addressing a critical challenge in the field. this focus on security aligns with current trends toward privacy-centric design in digital infrastructure, particularly as urban environments increasingly adopt smart technologies. moreover, the system’s adaptability to educational institutions—a setting often overlooked in favor of commercial or municipal parking solutions—adds to its novelty. while some smart parking systems prioritize features like space allocation or payment processing (lin et al., 2017), this system emphasizes secure communication between parking attendants and vehicle owners via encrypted qr codes. this capability fills a gap in campus parking management, where manual processes have historically limited efficiency and responsiveness to incidents like unauthorized parking or emergencies. supporting visuals the system’s automated workflow and security features are captured in three key figures, providing a clear visual overview of its operational efficiency and data protection capabilities. these visuals highlight the interactions within the system and the encryption-decryption processes that safeguard sensitive information. figure 4 shows the automated workflow of the system. parking occupants submit personal information to generate qr codes, guards scan these codes to manage figure 4: context diagram parking logs, and admin personnel oversee operations, making the entire process more organized and efficient. figure 5 illustrates the aes encryption process, where plaintext data is transformed into secure ciphertext. this encrypted data is embedded into qr codes, ensuring that access remains protected and confidential. figure 6 shows the aes decryption process, reversing the encryption to retrieve the original data from the qr codes. this step ensures that only authorized personnel can access the information, maintaining data integrity and security. together, these figures demonstrate how the system enhances parking management efficiency while employing robust security measures. the inclusion of both the aes encryption flow (figure 5) and aes decryption flow (figure 6) emphasizes the complete cycle of data protection, from securing information to safely retrieving it, making the system a notable advancement in secure parking management. conclusion the parking occupant management system using qr code solutions with aes algorithm was successfully pa ge 55 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 49-56, 2025 figure 5: aes encryption flow figure 6: aes decryption flow developed and evaluated, demonstrating high levels of user satisfaction across various aspects including interface design, functionality, usability, user experience, and security. this system addresses critical inefficiencies in manual vehicle cataloging processes within educational institutions by automating vehicle tracking and enhancing data security through aes encryption, significantly improving accuracy, efficiency, and security in parking management. its importance lies in providing a modern, automated solution that reduces human error, improves overall workflow, and ensures robust protection of sensitive data, making it a valuable advancement for campus parking facilities. despite its strengths, the study faces some limitations: it is specifically tailored for educational institutions in general santos city, does not extend beyond campus boundaries, restricts qr code generation to faculty, staff, and students, relies on internet connectivity for logging, is designed exclusively for android devices, and lacks the capability to monitor parking space occupancy. despite these constraints, the system holds substantial relevance for educational institutions aiming to upgrade their parking management practices, offering a secure and efficient framework that could potentially be adapted for broader use in commercial and residential parking facilities. the application of qr code technology paired with aes encryption sets a new standard for secure parking operations, with scalability that promises wider implementation beyond its current scope. to enhance its future potential, it is recommended that the system incorporate advancements in encryption technologies to sustain its security edge, expand its functionality to include features like predictive parking availability and real-time notifications, and develop versions for additional mobile platforms to broaden accessibility and user adoption. this comprehensive approach ensures the system not only meets current needs but also paves the way for smarter, more secure parking management solutions. references agarwal, a., & malik, s. (2022). an aes-based efficient and valid qr code for message sharing framework for steganography. in lecture notes in networks and systems (pp. 581–598). https://doi.org/10.1007/978981-19-2500-9_44 agun, j. e. c., rabie, m. a., & satoquia, r. v. p. (2022). portable vehicle ticketing device using qr code technology and android application. https://ejournals.ph/article. php?id=19020 ahmed, a. a., al-sanjary, o. i., & kaeswaren, s. (2020). reserve parking and authentication of guest using qr code. in 2020 ieee international conference on automatic pa ge 56 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 49-56, 2025 control and intelligent systems (i2cacis) (pp. 103–106). https://doi.org/10.1109/i2cacis49202.2020.9140192 ajini, a., arun g. (2016). qr code-based data transmission in mobile devices using aes encryption. international journal of science and research, 5(6), 1116– 1120. https://doi.org/10.21275/v5i6.nov164419 almuhammadi, s., & al-hejri, i. (2017). a comparative analysis of aes common modes of operation. in 2017 ieee 30th canadian conference on electrical and computer engineering (ccece) (pp. 1–4). https:// ieeexplore.ieee.org/abstract/document/7946655 asok, a. (2016). qr code based data transmission in mobile devices using aes encryption. international journal of science and research (ijsr), 5(6), 1116–1120. https://doi.org/10.21275/v5i6.nov164419 awan, i. a., shiraz, m., hashmi, m. u., shaheen, q., akhtar, r., & ditta, a. (2020). secure framework enhancing aes algorithm in cloud computing. security and communication networks, 2020, 1–16. https://doi. org/10.1155/2020/8863345 barriga, j. j., sulca, j., león, j. l., ulloa, a., portero, d., andrade, r., & yoo, s. g. (2019). smart parking: a literature review from the technological perspective. https:// scholar.google.com/citations?view_op=view_ citation&hl=en&user=nvqxovsaaaaj&citation_ for_view=nvqxovsaaaaj:y0pcki6q_dkc bhatia, s. (2023, june 22). qr codes for vehicle verification: a detailed guide. qr batch blog. https://qrbatch.io/ blog/qr-code-for-vehicle-verification/ chai, s., chong, l., chong, s., & goh, p. m. y. (2023). bus ticket booking system using qr code with aes encryption. international journal of membrane science and technology, 10(3), 1854–1871. https://doi. org/10.15379/ijmst.v10i3.1846 dokania, v. d., sevak, m. m., patel, d. d., & barve, p. s. (2020). qr code based smart parking system. journal of emerging technologies and innovative research. https:// www.jetir.org/papers/jetir2310482.pdf doomun, r., doma, j., & tengur, s. (2008). aes-cbc software execution optimization. in 2008 international symposium on information technology (pp. 1–8). https:// doi.org/10.1109/itsim.2008.4631586 ferdian, f., id, a. i. h. a., & rakhmat, f. r. u. f. (2021). penggunaan qr code berbasis kriptografi algoritma aes (advanced encryption standard) untuk administrasi rekam medis. joint (journal of information technology), 3(2), 20–27. https://doi.org/10.47292/ joint.v3i2.64 gangurde, n., ghosh, s., giri, a., & gharat, s. (2022). ticketing system using aes encryption based qr code. in 2022 4th international conference on smart systems and inventive technology (icssit). https://doi. org/10.1109/icssit53264.2022.9716234 geeksforgeeks. (2023, may 9). block cipher modes of operation. https://www.geeksforgeeks.org/blockcipher-modes-of-operation/ hurbungs, v., bassoo, v., & fowdur, t. p. (2021). fog and edge computing: concepts, tools and focus areas. international journal of information technology, 13(2), 511– 522. https://doi.org/10.1007/s41870-020-00588-5 inderscience publishers. (n.d.). linking academia, business and industry through research. https://inderscience.com/ offers.php?id=132421 ingole, m., patle, c., gahane, s., atkare, a., bante, a.(2023). qr based car parking system. international research journal of modernization in engineering technology and science. https://doi.org/10.56726/irjmets47293 kadu, a., & kadu, a. (2014). qr code-based smart parking system. journal of emerging technologies and innovative research. https://www.jetir.org/papers/ jetir2310482.pdf lin, t., rivano, h., & le mouël, f. (2017). a survey of smart parking solutions. ieee journals & magazine. https://ieeexplore.ieee.org/document/7895130 lu, z., & mohamed, h. (2021). a complex encryption system design implemented by aes. journal of information security, 12(2), 177–187. https://doi. org/10.4236/jis.2021.122009 rachh, r., mohan, p. v. a., & anami, b. s. (2012). efficient implementations for aes encryption and decryption. circuits, systems, and signal processing, 31(5), 1765–1785. https://doi.org/10.1007/s00034-012-9395-0 singh, s. (2024, january 23). encrypted qr code: a complete guide with best 6 advantages. scanova blog. https:// scanova.io/blog/encrypted-qr-code/ vaidehi, m., & rabi, b. j. (2014). design and analysis of aes-cbc mode for high security applications. in second international conference on current trends in engineering and technology icctet 2014 (pp. 499–502). https://ieeexplore.ieee.org/abstract/ document/6966347 wahsheh, h. a., & luccio, f. l. (2020). security and privacy of qr code applications: a comprehensive study, general guidelines and solutions. information, 11(4), 217. https://doi.org/10.3390/info11040217 pa ge 1 pa ge 87 american journal of smart technology and solutions (ajsts) smart transportation systems with artificial intelligence: enhancing efficiency, safety, and sustainability abdullah sheikh1*, md. shakil sheikh2, tajbiha mehonaj rinvee3 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.6022 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: august 27, 2025 accepted: october 03, 2025 published: november 08, 2025 artificial intelligence (ai) is transforming transportation, yet most research and applications focus on isolated improvements, lacking a unified approach that connects operational gains with strategic national goals. this paper addresses this gap by developing a conceptual framework that synthesizes how ai enhances transportation systems across three integrated pillars: efficiency, safety, and sustainability. through a synthesis of recent literature and industry case studies, we propose a model that demonstrates the synergistic effects of ai applications, such as predictive maintenance and dynamic routing. the framework’s primary contribution is to illustrate how these technological advancements collectively bolster u.s. competitiveness by building resilient supply chains, reducing emissions, and fostering leadership in sustainable innovation. this study provides a structured roadmap for policymakers and industry leaders to leverage ai not merely for operational efficiency, but as a strategic asset for long-term economic security. keywords artificial intelligence, efficiency, logistics, safety, smart transportation, sustainability, u.s. competitiveness 1 wright state university, dayton, ohio, usa 2 atish dipankar university of science & technology, bangladesh 3 brac university, bangladesh *corresponding author’s e-mail: adustabdullah@gmail.com introduction transportation is the backbone of modern economies and defines the movement of goods, services and people in regions and markets. in the united states, the transportation sector supports millions of jobs and contributes greatly to economic growth. at the same time, it is one of the largest sources of carbon emissions and faces growing challenges of congestion, safety and rising operational costs. these pressures underscore the urgent need for intelligent and sustainable transport systems. artificial intelligence (ai) is a central tool for solving these challenges. unlike traditional optimization techniques, ai processes large amounts of real-time data, detects complex patterns, and generates adaptive solutions. applications range from predicting traffic flows and adapting flight routes, improving fleet fuel efficiency, and supporting autonomous vehicle systems. with this combination of efficiency and adaptability, ai is a powerful enabler for change in the transport sector. the relevance of smart transportation goes beyond the operation. it has a strategic impact on national competitiveness and resilience. countries successfully integrating artificial intelligence into their transportation systems are better equipped to reduce costs, reduce emissions and maintain reliability in the face of disasters such as pandemics, natural disasters or geopolitical tensions. for the united states, which is heavily dependent on timely transportation of goods through its logistics network, intelligent transportation with ai is not only a technological innovation, but also a matter of long-term economic security. this paper focuses on how ai can be applied to improve three basic outcomes in transport: efficiency, safety and sustainability. efficiency means reducing delays, fuel consumption, and waste through predictive analytics and real-time optimization. safety is accompanied by computer vision systems, driver monitoring and predictive maintenance with ai. sustainable development is achieved through reduction in carbon emissions, optimization of energy use and support for the transition to greener transport modes. these three dimensions together form a framework for intelligent transportation driven by ai, supporting the competitiveness and leadership of the united states in sustainable innovation. the remainder of the paper is as follows. in the second section, the literature on ai applications in intelligent transport is reviewed, and both progress and current gaps are highlighted. section 3 introduces the conceptual framework and methodology. section 4 presents case studies by industrial leaders. chapter 5 deals with the impact on the competitiveness of the united states. finally, section 6 concludes with a summary of the findings and recommendations of future research. literature review evolution of smart transportation & its smart transportation systems (stss) and intelligent transportation systems (its) have evolved over the past two decades. its integrates perception, communication, calculation and control to improve transport network mobility, safety and environmental performance. zemmouchi-ghomari et al. (2025) review how ai is integrated into its to support traffic flows, safety and sustainability in urban areas. recent research has also highlighted the influence of generational ai within its. for example, rong and others. (2025) examine applications such as data generation, prediction and decision-making in its subsystems. pa ge 88 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 87-90, 2025 in recent research, sustainability and efficiency have become increasingly closely linked. son et al. (2025) conducts a systematic review to demonstrate how ai, iot, digital twins and optimization methods are used in intelligent transportation planning, highlighting important improvements in traffic flows and emission reductions. ai methods applied in transportation ai methods used in transportation are diverse. they include: • machine learning (ml) and deep learning: for traffic prediction, demand forecasting, incident detection, and routing decisions. • reinforcement learning (rl): it is used to adapt control strategies dynamically (e.g. traffic signal control, dynamic routing). li et al.’s bibliographic review. (2022) shows that rl is increasingly attractive in transportation applications. • generative models/generative ai: emerging for tasks such as creating synthetic data, simulation of scenarios, or improving prediction under data-sparity conditions (for example, yan and li, 2023). • hybrid approaches: combining ai with iot sensors, blockchain and optimization. idrissi et al. (2024) examine how iot, iot, and blockchain interplay in logistics and transport, and improve the availability, resilience, and support for decision-making. key application areas & findings ai in smart transportation is applied to many problem domains. here are several with findings: • traffic prediction & routing: many studies show that ai is more powerful than traditional models (such as arima and linear regression), especially in nonlinear and volatile traffic environments. • safety & accident prediction: computer vision and sensor-based ml detect risky driving behaviors or predict likely crash spots. • predictive maintenance: using sensor data and ml to anticipate vehicle or infrastructure failures before they happen. • energy optimization & emissions: ai optimizes route planning, load balancing, and speed profiles to reduce fuel use and emissions. • mobility-as-a-service (maas): rouky et al. (2025) examine how ai supports integrated mobility systems (routes, payments, user behaviors) and link ai technologies to integrated levels. challenges, gaps, and research needs despite progress, several gaps and challenges persist: • data quality and heterogeneity: many models struggle when input data are missing, noisy, or inconsistent. • real-time requirements: ai models must be fast and efficient to operate in real-time constraints. • scalability: many prior works focus on small regions or simulated settings; scaling to national networks is harder. • integration across systems: too often, applications are isolated (only traffic or routing). few studies integrate safety + emissions + efficiency within a single framework. • ethical, privacy, and security concerns: use of ai in transportation brings risk of surveillance, data breaches, fairness issues. • lack of longitudinal studies: few works track ai performance over long time or in real operations (versus simulations). how this paper contributes our work addresses these gaps. we propose a unified framework for efficiency, safety, and sustainability in intelligent transportation. unlike many previous studies, we emphasize the coherence of the real world over the application domain rather than isolated models. it is also intended to highlight the strategic implications for us competitiveness, which are under-emphasized in the current literature. materials and methods this paper uses conceptual and exploratory methods rather than empirical methods. the objective is to develop a framework to link artificial intelligence (ai) to intelligent transportation, efficiency, safety and sustainability, while focusing on national competitiveness. research approach the methodology is based on comparative synthesis of recent studies, reports and industry practices. instead of analyzing a single dataset, we reviewed academic and industrial literature published in the fields of transportation, logistics and ai between 2018 and 2025. the aim was to identify repeated applications, evaluate their results and organize them into coherent models of smart transportation. data sources the source of this study is: • articles reviewed by peer-reviewed professionals on ai in transportation, logistics and sustainability. industry reports from major transport companies such as amazon, ups, uber freight and tesla. • policy documents from american and international organizations relating to intelligent cities, clean transportation and digital infrastructure. using various sources ensures that the framework is based on both academic theory and practical evidence. framework development the framework was developed in three steps: 1. mapping applications: key applications of ai in transportation, such as traffic forecasting, routing, predictive maintenance, safety, energy optimization, and emissions monitoring, were identified and categorized. 2. identifying outcomes: for each application, the dual outcomes were noted: efficiency gains (reduced costs and faster operations) and sustainability benefits (reduced pa ge 89 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 87-90, 2025 emissions and improved safety). 3. linking to competitiveness: finally, we linked these outcomes to national competitiveness by arguing that countries with stronger ai-driven transportation systems gain resilience and leadership in the global economy. scope and limitations this is a conceptual study, not an empirical one. while it provides a broad synthesis, it does not include primary data collection or quantitative testing. however, the paper offers a roadmap for both researchers and policymakers by organizing existing knowledge into a unified model. the proposed framework for ai in smart transportation this section presents the framework for demonstrating how artificial intelligence supports the efficiency, safety and sustainability of smart transportation systems. the framework is based on three pillars: operational efficiency, road safety and environmental sustainability. these pillars together strengthen the long-term competitiveness of the country. figure 1: conceptual framework “ai in smart transportation” this figure illustrates how ai applications strengthen efficiency, safety, and sustainability, which together reinforce u.s. competitiveness. efficiency through ai ai reduces inefficiency across transport networks. analyzing real-time traffic data, ai systems adjust signal timing, redirect vehicles, and predict traffic congestion before it occurs. in the logistics sector, platforms such as uber freight and amazon’s routing system allocate trucks more effectively, reducing empty miles and reducing costs. predictive maintenance also supports efficiency. machine learning models monitor vehicle health and forecast component failures, allowing rapid repairs and reducing shutdown times. safety improvements safety is the second pillar of the framework. ai-enabled systems, such as driver assist technology, lanekeeper system, and automatic brake, reduce the risk of accidents. ai can also process real-time road, weather and driver data and emit warnings before the situation escalates. for example, predictive analytics can identify dangerous driving patterns and trigger alerts or training interventions. these applications not only save lives but also reduce the costs of accidents financially and socially. sustainability and emission reduction artificial intelligence directly contributes to sustainable development by reducing energy consumption and emissions. dynamic routing reduces fuel consumption by selecting the most efficient path. electric vehicle fleets can be managed by ai to forecast charging requirements and align them with renewable energy supply. companies like walmart and ups already use artificial intelligence to reduce energy consumption in warehouses and optimize fleet management to reduce emissions. at the city level, intelligent traffic lights based on ai reduce idle time and cause measurable carbon production reductions. linking to competitiveness combining the three pillars of efficiency, safety, and sustainability creates a competitive advantage. the nation and company adopting ai in transport gains resilience to disruption, better compliance with environmental standards and more reliable supply chains. for the united states, this not only means reducing costs, but also enables sustainable innovation, develops exportable technologies and strengthens its position in the global market. results and discussion the proposed framework emphasizes that artificial intelligence is not a set of digital tools. it is a strategic driver of efficient, safe and sustainable transport. by combining these pillars, artificial intelligence offers a path to strengthening national competitiveness. from the business point of view, the framework shows that artificial intelligence investments can create double value. companies benefit financially by reducing costs and delays, while improving social results such as accident reduction and emission control. this dual impact makes it easier to justify ai adoption to stakeholders. from a policy point of view, the framework provides a roadmap for governments to develop regulations and incentives to support them. for example, subsidies to electric vehicles equipped with artificial intelligence can reduce emissions, while data sharing policies can improve traffic management throughout cities. policy makers can also encourage public and private partnerships to accelerate the deployment of ai solutions. from a research point of view, the framework identifies gaps in future research. it is necessary to measure the long-term impact of the adoption of artificial intelligence, particularly how efficiency is balanced with sustainable outcomes. more empirical research is also needed on the social effects of ai-based safety systems, especially in areas with high incidence of accidents. pa ge 90 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 87-90, 2025 overall, the implications of this framework show that leading companies and countries in the field of aienabled transportation will not only reduce costs and risks, but also become leaders in sustainable innovation. for the united states, this offers the opportunity to strengthen resilience, maintain technological leadership and enhance global competitiveness. over the next decade, the most important research needs will be the testing of ai models not only for their efficiency, but also for their impact on fairness and longterm sustainability, particularly in the transportation networks of the united states. conclusion this paper developed a conceptual framework establishing ai as a catalyst for integrated gains in transportation efficiency, safety, and sustainability. the key insight is that the synergistic effect of these pillars is crucial for building a resilient and competitive national transportation system. for u.s. competitiveness, this means policymakers should prioritize funding for integrated ai projects that cross environmental and operational agencies. industry leaders must move beyond point solutions and adopt platforms that unify these three objectives. for researchers, the critical path is to generate empirical evidence on the framework’s synergies in real-world settings and to establish standards for ethical ai deployment. ultimately, strategic adoption of this unified ai framework is not optional but essential for the u.s. to secure economic strength and global leadership in the future of mobility. references chen, c., demir, e., huang, y., & scholts, s. (2022). ai and big data in sustainable transportation: opportunities and challenges. transportation research part e: logistics and transportation review, 159, 102620. https://doi.org/10.1016/j.tre.2021.102620 choi, t. m., chen, y., & lee, w. w. (2018). big data analytics in supply chain management: a review and research agenda. journal of management information systems, 35(2), 528–567. https://doi.org/10.1080/07 421222.2018.1440770 dwivedi, y. k., hughes, l., kar, a. k., baabdullah, a. m., grover, p., abbas, r., & mani, v. (2021). climate change and cop26: are digital technologies and information management part of the problem or the solution? international journal of information management, 63, 102456. https://doi.org/10.1016/j. ijinfomgt.2021.102456 ferreira, a. c. a., francisco, m., & pinho, a. (2025). the use of artificial intelligence in transportation and logistics: a systematic literature review. ieee access, 13, 1–14. https://doi.org/10.1109/ access.2025.3275890 hasan, m. r., islam, m. r., & rahman, m. a. (2025). developing ai-driven models for demand forecasting in u.s. supply chains: enhancing predictive accuracy. edelweiss applied science and technology, 9(1), 1045–1068. ivanov, d. (2020). predicting the impacts of covid-19 disruptions on global supply chains: a simulationbased analysis. international journal of production research, 58(20), 6140–6156. https://doi.org/10.1080 /00207543.2020.1750727 lee, j. (2024). ai-powered forecasting in transportation: accuracy, speed, and scalability. multidisciplinary journal of instruction, 7(1), 115–125. mckinsey & company. (2023). the future of mobility: aidriven sustainability in transportation. mckinsey insights. https://www.mckinsey.com pattnaik, s., liew, n., kures, a. o., pattnaik, e., & park, k. (2024). catalyzing smart transportation: ai applications for sustainable logistics. engineering proceedings, 68(1), 57. https://doi.org/10.3390/ecsa68-057 waller, m. a., & fawcett, s. e. (2013). data-driven supply chains: a new lens on supply chain management. business horizons, 56(5), 639–647. https://doi. org/10.1016/j.bushor.2013.06.001 wieland, a., & marcus, f. (2020). the role of dynamic capabilities in responding to supply chain disruptions. international journal of production research, 58(10), 2904– 2915. https://doi.org/10.1080/00207543.2020.1744 519 pa ge 1 pa ge 59 american journal of smart technology and solutions (ajsts) ai-driven real-time kinematic and dynamic analysis of ur5 robotic arm for business optimization sudipta sotra dhar1, shovra sotra dhar2, sazib hossain3* volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.4563 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: february 18, 2025 accepted: march 21, 2025 published: april 11, 2025 this paper offers a novel ai-based approach to perform real-time kinematic and dynamic analysis of the ur5 robotic arm to apply it in the business realm for robotic improvement. the data set used in this study includes accurate time based motion information of elbow, shoulder, wrist and hand joint angles (j1-j6) of the arm, their speeds and accurate time based position information of the tool (x,y,z) in different intervals, which is very useful to assess the operational parameters of the arm. the study aims at developing effective predictive models and optimisation algorithms for the robot’s kinematic equations of motion that relation the joint movements and velocities, as well as tool position in the 3-space. these concepts aid in evaluating how efficient the robotic tasks in an environment that simulate reality are. according to the findings of the present study, analyzing the kinematics and dynamics of the robot, there are specific parameters that indicate the efficiency of the robot’s movement, including precise joint angles or synchronism of arm movements. this research explores the extent to which the aforementioned factors affect the business productivity directly and highlights the benefits accrued by improving the robotic performance in regards to decreased amount of time wasted on repairs, improved accuracy and optimal resource utilization. this paper explores how ai models can enhance the supervisory control of robotic systems and allow real-time control of decision-making parameters to increase the efficiency of tasks and profitability in the business. the works provide further essence to elevate the real-time robotic optimization within industrial automation that deploying artificial intelligence in the working environments can provide logical, best and can be most suitable for the complex business areas placed in organisms where growing and changing rapidly. this way, it is possible to have higher levels of automation, and increase production processes, and profitability. keywords ai-driven optimization, business productivity, dynamic modeling, real-time kinematic analysis, ur5 robotic arm 1 school of computer science and engineering, south china university of technology, china 2 school of computer science and engineering, national institute of technology, india 3 school of business, nanjing university of information science & technology, nanjing, china * corresponding author’s e-mail: esazibhossain@gmail.com introduction over the last couple of years, the use of robotics in business has greatly increased and has helped enhance the levels of business automation. in manufacturing and assembling activities as well as logistics and supply chain operations to name but a few, basic robotic systems, specifically robotic arms such as the ur5, have brought about introduction of enormous changes (azman et al., 2023; shkarupeta & babkin, 2022). ai-driven robotic systems which have established themselves as enablers of modern process automation, promote themselves as a scalable technology for improving the quality of production processes (hossain et al., 2024). industrial robots increase productivity of the manufacturing processes by boosting its speed and accuracy and, at the same time, ensure environmental sustainability by minimizing energy consumption and production of waste (benabed & boeru, 2023). the need to enhance operational efficiency of robots is seen in today’s fast growing market demands, rising costs and need to provide higher quality service, by the year 2024 outlined by nakib et al. (2024). according to hossain et al. (2024) on the concept of operations in businesses, it is critical to improve roi from the robotic systems as well as to ensure flexibility in the volatile market conditions of production demands (parvez et al., 2024). the ur5 arm is a robotic arm characterized by flexibility and versatility, and it has found its place in automotive production line, electronics and packaging industries among others (azman et al., 2023). despite its popularity, much more may be achieved to optimize these systems because the current available techniques do not take real-time data in the decision-making process to dynamically control robotic initiatives throughout operations (mohr et al., 2024). modern developments have therefore recognized the need for advance the performance of robots through ai. through real-time kinematic and dynamic analysis of robots performance one is able to determine different aspects covering the robotic movements and its behaviors and how it is affected with change of conditions of its operation. this research seeks to fill this gap in robotic optimization by applying ai-aided techniques in the study the virtual and actual time mechanics and dynamics of the ur5 robotic arm. the possibility to perform the dynamic optimization with the help of ai can and has the potential to drastically change the business processes by making them more efficient with less mistakes which leads to the rise of productivity (shkarupeta & babkin, 2022). however, the real-time analysis of robotic arm like the ur5 has been a somewhat tricky issue in the literature even to date. the conventional robots are normal mechanically pa ge 60 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 59-67, 2025 and operation software defined and have to be altered manually when the new operations are introduced. this means that they cannot achieve their optimum usefulness in real-time business environments where flexibility of analytical process according to real-time data is so important. the existing research also lacks information on the superior order motion planning and dynamics of a robotic system under the existing operational conditions. given that most practical environments where robots are used are dynamic and present a rapidly changing task load, machine configurations and even worker interference, a more dynamic approach to control of robots is called for. moreover, as far as the research of the robot movement is concerned, only several kinematic and dynamic models have been provided to be used in the controlled condition and thus far there is no well-organized methodological framework available to capture real-time data for the dynamic performance optimisation of the robot more especially in the business automation sector. this research seek to address this need by proposing innovational ai-based approach towards real-time kinematics and dynamic evaluation of the ur5 robotic arm. with this kind of aim, we are certain to achieve better performance, flexibility, proper execution of operations and fewer problems of productivity. this research explores the dilemma of enhancing robotic systems to become a part of an organization’s tasks effectively and efficiently without much need for intervention by other folks. therefore, the main aim of the present investigation is to investigate the motion kinematics and dynamics in real time of the ur5 robot manipulator by measuring, joint angles, velocities, forces, accelerations and tool motions. these facets are as follows: the study aims to identify the extent and manner by which such factors affect the operation and productivity of the robot in different surroundings. it also seeks to explore the use of artificial intelligence (ai) and more so the machine learning algorithms in the kinetic and dynamic part of the robotic arm to maximise real time control of speed, accuracy, and flexibility of the robotic arm. further, they will explain how ai-optimized robotic systems can be used for managing and improving a company’s supply chain activities in terms of efficiency, costs and results. this convergent computational study will apply rtk and dynamic analysis within the business environment so as to show how such advanced robotic structures can enhance business flow and contribute to the improvement of business competitiveness and sustainability. in addition, the r&d will also obtain and improve the machine learning algorithms for data analysis of the robotic arm and make suggestions for improvement to different aspects of the system and detection of the problems whenever needed without the interference of programming by a human operator. this paper brings a few undeniable advancements into the field of robotic automation to the table. for the first, it designs real-time ai models of robotic arms and the capability to predict and adjust their next movements based on the received feedback. these models include kinematic and dynamic models whose main goal is to make the robotic arms function optimally in highly productive and dynamic industries. secondly, the research develops a new methodology of using ai for real-time data acquisition, ai modelling, and dynamic optimization with an overall goal of enhancing the performance of robots in organizations. this framework of robotic operation allows flexibility in robotic acquisition due to improved error minimization, productivity enhancement and efficient automation for firms. also featured in the research conducted is the impact of such models in relation to utilizing artificial intelligence to increasing the effectiveness of the robots that are then used to optimize organizational operations to increase productivity, reduce time wastage and increase precision in task execution. both of these developments can easily be seen as a move to cut costs and an essential key to survival in the age of automation. lastly, this research will add value towards making industrial automation as part of industry 5.0 where robots developed through artificial intelligence will be able to integrated with human beings performing tasks in environments in a more complex and flexible manner and also improve organizational processes making it more robust and intelligent. literature review that is why this topic is current and essential: the use of robotics and artificial intelligence in industries has become the new trend that significantly impacts increasing efficiency. for instance, the recently popularized robotic arms including the ur5, are standard features in automation and are increasingly used in a wide range of industries, including manufacturing, logistics, and assemblage, among others. these systems use kinetics and dynamics principles for purposeful motions with high accuracy and speed at the same time with high flexibility. with each year passing by, real-time optimization of these robots becomes more necessary due to the improved usage of ai techniques. kinematic control focuses on the arm motion and position, whereas dynamic control targets force, torque, and acceleration in an attempt to enhance the important issues in automation, such as higher efficiency, lower costs and versatility of production tools. this literature review is basically a study on robotic arm control and this research proposal is concentrated on the kinematic and the dynamic model of the robotic arm and how artificial intelligence can assist in the improvement of these robotic systems. moreover, the review will discuss the related works and literatures on machine learning and how it has applied on the robotics such as reinforcement learning and deep learning techniques. it has been widely utilised in improving the effectiveness of robotic control for increased self-operation, accuracy and flexible functioning in diverse surroundings. consequently, we will also explore the effects of business robotization and the use of artificial intelligence in making business enhancement, exploring how robotization conveys value toward raising profitability, cost control, and general performance across the industries. pa ge 61 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 59-67, 2025 previous work on kinematics and dynamics of robotic arms kinematics and dynamics are the two significant branches of robotic arm control, as they help toward determine the efficiency of the robotic system. kinematics mostly concerns itself with the motion of a robotic arm, and is concerned with the joint angles, velocities, accelerations and the position of the end of the robotic arm. dynamic analysis on the other hand concerns force and the moment that acts on it to enable prediction of best performance of the robot in different terrains. there has been published work in an attempt at investigating several kinematic models to enhance the control of robotic arms and the associated accuracy. for example, analysis on inverse kinematics (ik) has been vital for robotics control, to plan the path which is required to achieve by the arm and ensure that all motions correspond to specs of the task (khater et al., 2023). fk and ik improvement in the robots has made it possible to enhance the robotic arm manipulation in different applications ranging from production to surgical (vyaas, 2025). these are usually achieved by numerical methods solving the nonlinear equations of the system motion and maintaining high efficiency and accuracy of the robotic arms’ work. dynamic modeling has also progressed well with less costs by incorporating the power of forces, torques and momentum. for instance, concerning the ur5 robotic arm, some of the researches have paid much attention to the dynamic modeling for collaborating with the disturbances and enhancing the robotic movements (victores et al., 2025). dynamic models are useful in robotic systems to determine the amount of torque necessary in every joint so that the outcome of the interaction of the robot with the forces outside it will result to smooth movements of the robotic system. this is particularly important in environments where tasks as well as conditions may constantly change over a short period, before the robot can get to the scene to complete it. ai has considerably been integrated in the robotic arms, specifically, the ai techniques enhance the kinematic and dynamic models. some of the papers have discussed how real-time modification of robotic motion can be performed, provided that ml and machine reinforcement learning are incorporated in dynamic control systems. in these systems, the robotic behaviour is improved by the use of feedback data where they also increase the functionality of the arm (rahaman et al., 2025). machine learning in robotics thus, the application of ml has become an important factor for enhancing robotic systems’ performance, particularly in decision-making and adaptive operation. some of the related works include the reinforcement learning (rl) and deep learning techniques for the improvement of the control on robot’s motion and action. otherwise, reinforcement learning has been used in optimizing operational trajectory of a robotic arm. this is a technique of training algorithms that enable the robotic arms make improvements of necessary movements based on gaining or losing points. for example, khater et al. (2023) applied rl for trajectory planning with 6 dof robotic arm assuming the rl agent would continuously adapt the movements of the arm in response to the signals of the environment. it not only enhanced performance of tasks but also reduced the dependency of the robot on the pre-scripted patterns of movements eliminating the rigidity of the robot’s movements to some extent. cnn and rnns have also been adopted in robotic structures for image classification, object recognition, and the navigation functions. these has been proven to help the robots improved their ability to interpret the different surrounding so as to perform the activities in a more easier manner (rahaman et al., 2025). sometimes it has been integrated with the other kinematic and dynamic models to develop a new form of the models that take both deep learning and the other models into consideration. it can learn from big data available and from any change in the environment and thus improves the functionality of robot in executing various tasks independently. when kinematic and dynamic analysis are used simultaneously with the help of ai models, it gives the best picture of overall performance. through motion (kinematics) and force (dynamics) control and understanding on the other hand, the machine learning algorithms allow very precise control of the movements of the robot end-effector, allowing the arm to execute a given task to the best of its potential even if the environment is unstructured or likely to change (vyaas, 2025). besides, the incorporation of ai in this particular instance offers value to enhance the robot’s performance even as it decreases risks of mistakes or unsuccessful working during the completion of tasks. business optimization through robotics the concept of reenchanting business using robotics has been discussed often, especially in the production, supply chain management, and other industry processes. robotics and artificial intelligence have been noted as effective instruments toward achieving high levels of automation, decreased costs, and increased output. the centers have adopted the employment of robotic arms and this has made a big difference especially in the rate, quality and uniformity of manufacturing. for instance, in car production processes, robots perform the tasks such as handling and bonding, and painting. they involve high precision, and must fit into similarly high tolerances with little variation from repetition to repetition, characteristics that are provided by robots while at the same time cutting down the costs and effort associated with human input. research has pointed out that with the integration of ai and ml in these robots, the processes have been brought closer to near optimal, with the systems being able to adapt with predictive learning, where the next operations can be anticipated from previous understanding and corrective measures taken (rahaman et al., 2025). it enables business organizations to acquire more proficiency and flexibility in pa ge 62 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 59-67, 2025 terms of responding to production requirements. supply chain and logistics are also among the environments that have incorporated robotics in their operations. robotic arms driven by artificial intelligence can include tasks like sorting, packaging, and material handling thereby not requiring much manual input and it can produce a large output. in warehousing, the robots have the ability to transport the goods to the required areas depending on the demand, which in turn has an impact on the efficiency of storing the stocks and minimizing the time taken for it. enhancing the route and timing of logistics by the help of ai-driven robots allows for the execution of the given field’s complicated operations with fewer mistakes and time losses (victores et al., 2025). the use of robots and artificial intelligence not only automates work tasks as part of business processes but also brings more value-added. by this, ai optimizes the utilization of robotic systems so that they can tackle changing business environments and meet the market needs as they are encountered. for instance, high load can be processed by increasing the robot speed or equivalently low load can be processed by slowing down the robots or de-energizing some of them. including this, the dynamic optimization not only enhances the efficiency of production but also make it sustainable through automation that fewer energy and wastes will be consumed (hazem et al., 2025). also, synchronizing ai and business organizations enhance decision-making through offering feedback to business managers. for instance, in predictive maintenance, the ai models are involved in assessing the condition of the robotic arms or any other mechanical equipment and make predictions about the failure. such prevention type of maintenance minimizes time a machinery is off-line and also increases the life span of the robotic equipment, both of which translate to cost reduction and effective operation (bongomin, 2025). materials and methods the following model diagram in (figure 1) shows the integration and functionalism of the ai real-time kinematic and dynamic analysis of the ur5 robotic arm. this demonstrates how most of the components such as data preprocessing, kinematic and dynamic models, artificial intelligence optimization and business optimization work sequentially. figure 1: ai-driven real-time kinematic and dynamic analysis of ur5 robotic arm data collection the data used to conduct this study was obtained from the nist, with live data of the ur5 robotic arm. it is the data that dictates joint angles, velocities and position of the tool which is vital for determining the kinematics and dynamics of the robotic arm. the dataset comprises several key columns: plctime, which records the timestamp in plc (programmable logic controller) time; robottime, which corresponds to the timestamp in robot time; j1_qactual through j6_qactual, which represent the actual joint angles (in radians) for each of the six joints; j1_qdactual through j6_qdactual, which capture the joint velocities (in radians per second) of the corresponding joints; and toolx, tooly, and toolz, which denote the position of the tool (end effector) in the x, y, and z directions (in meters). at real time manner, this datasets provides a chance to make dynamic analysis of the character and performance of the robotic arm in kinematic and dynamic manner, that also helps in assessment the operational probity of the arm. preprocessing the following operations were performed on the given dataset data preprocessing: 1. data cleaning: incomplete records, especially in the toolz column were considered and processed as necessary. interpolation was applied pa ge 63 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 59-67, 2025 where methodically conceivable or, otherwise, the data was omitted if gaps were large. scaling-by this process, some values like joint angles and velocity were normalized in order to create an efficient program for machine learning algorithms. some of the equation used are as follows: xnorm=(x-xmin)/(xmax-xmin ) where x is any feature such as joint angles, and xmin and xmax represent the minimum and maximum values of the feature, respectively. the dataset was smoothed using a moving average technique to reduce noise in the data for more accurate kinematic and dynamic modeling. kinematic and dynamic analysis kinematic modeling the forward kinematics (fk) and inverse kinematics (ik) are applied to analyze the motion of the robot. the forward kinematic equations are based on the denavit-hartenberg (dh) parameters, which describe the transformations between adjacent links of the robotic arm. the forward kinematics (fk) and inverse kinematics (ik) are applied to analyze the motion of the robot. the forward kinematic equations are based on the denavit-hartenberg (dh) parameters, which describe the transformations between adjacent links of the robotic arm. the reinforcement learning (rl) algorithm, specifically deep q-networks (dqn), is applied for trajectory optimization. the rl agent learns to optimize the robot’s movements based on feedback from its environment. q(s,a)=r(s,a)+γ max┬a q(s’, a’ ) where: q(s,a) is the expected reward for taking action a in state s, r(s,a) is the immediate reward, γ is the discount factor, s’ is the next state. the agent learns to minimize energy consumption, time, and deviation from the desired end effector position by adjusting joint angles in real-time. training and testing of machine learning models to train the machine learning models, the dataset is divided into a training set (80%) and a test set (20%). the training set is used to teach the models how to predict joint movements, while the test set is used to validate the model’s performance. the ai models are trained using a combination of supervised learning (for position and velocity prediction) and reinforcement learning (for dynamic trajectory optimization). optimization framework the optimization framework aims to improve the performance of the robotic arm within a business environment. it involves real-time task execution, where the ai model continuously adapts the robotic arm’s behavior based on real-time feedback. the framework ensures that the robot’s actions are aligned with business goals such as: reducing task completion time minimizing the time taken for the robotic arm to complete tasks. improving precision ensuring high accuracy in task execution. energy efficiency optimizing the energy consumption during task execution. the optimization process is based on continuous monitoring and real-time feedback loops, where the robotic system adjusts its movements dynamically based on changing task demands. this feedback loop is essential for enhancing productivity and ensuring that the robotic arm operates efficiently in diverse business contexts. results and discussion the first plot in the figure two reveals the joint angles (j1_qactual through to j6_qactual) of the ur5 robotic arm. the joints’ angles share similar profiles but some of them like joint 5 and 6 move almost linear. when it comes to variation, joints like joint 1 have a greater variation meaning the movement of such joints is not constant than that of joints like joint 7. also, the joint angles where: θi is the joint angle, αi is the link twist, ai is the link length, di is the link offset. the end effector’s position and orientation are calculated using the product of transformation matrices from each link. dynamic modeling dynamic modeling involves computing the forces and torques acting on each joint. the general dynamic equation for the robotic arm is: where: m(q) is the mass matrix (representing inertia), c(q,q˙) is the coriolis/centrifugal matrix, g(q) is the gravitational force vector, τ is the torque applied at each joint. this equation is solved to understand how forces at each joint impact the robot’s motion. dynamic parameters such as joint velocities, accelerations, and external forces are included in the analysis. ai model implementation motion prediction and trajectory optimization machine learning algorithms are implemented to predict the future positions and velocities of the robot’s joints. pa ge 64 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 59-67, 2025 take negative and positive numbers that point to the fact that the robotic arm is going through numerous cycles of movement in various tasks, including both forward and backward movements and making adjustments in compliance with some task demands. the second plot (figure 3) displays the velocities (from j1_qdactual to j6_qdactual) of the robotic arm’s joints over time. some of the joint velocities are fluctuating considerably, specifically in joint 1 and joint6 while compared to other joints like joint 3 which has oscillating and close to zero behaviour. this appears to indicate that the arm is either rotating at certain angles or halting during its functioning. such velocities differ in apparent real-time to depict changes in position of the arm movements probably due to the need it has to flex or respond to forces during its working phases. figure 2: joint angles over time figure 3: joint velocities over time the third graph (figure 4) gives a graphical representation of the tool coordinates in the x, y and z axes over time as it executes it’s function. the tool demonstrates strict and uninterrupted motion as was expected of a precision activity and tracing a regular progression in a restricted space. the motion appears to be well controlled, meaning the employees of the company might be operating the robotic arm for a very delicate and sensitive task that needs precision. this implies that arm is in a position to perform a function that requires a high level of precision and check that the tool should operate in a certain operational range. the plot given in figure 5 is the final plot, which plots correlation between joint angle (j1_qactual to j6_qactual) and joint velocity (j1_qdactual to j6_qdactual). the results show positive relationships between some joint angles and their derivatives such that the angles and velocities change in the same manner, specifically there is a very high reliability of joint 1 and joint 2. it is also observed that joint 6 has significantly less coherence with all other joints which may imply that the pattern of movements of joint 6 is not influenced by the other joints as to a large extent as much as the other joints; this could mean that joint 6 is more independent in its movements as compared to the rest of the joints or that the natural movements of joint 6 are not controlled in the same manner by the control system as the other joints as they are. pa ge 65 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 59-67, 2025 figure 5: correlation matrix of joint angles and velicities discussion however, it would equally be significant to look into dynamic behavior of the robotic arm from the angle of joint angles and velocities. the torques obtained from the dynamic model give a good estimate of the torques necessary for the observed joint movement. for example, relatively larger variability in the joint velocities, which is observed in the joint 1 and the joint 6, is usually associated with the time-varying nature of the joint torques. these variations indicate that these joints are tasked with more diverse dynamic actions, and therefore ifrit could take more energy and time to perform activities. when these dynamic parameters are controlled with ai models, the arm has the potential to have a natural kind of movement in implementing operations hence operating at a lower cost in real-world assignments. concerning the rl model used in this paper, the model enables the robotic arm to decide its trajectory during operation by learning from the feedback received during its operation or a sequence of ongoing operation. by the end of the training episodes the agent learns to execute actions that result in minimal time and energy to complete the task. a specific goal was designed to serve as a reward function that would encourage not only accuracy and specificity of the generated actions but also their efficiency. thus, the movement was more accurate and combined decreased energy expenditure, which was a feature that the arm in the business needed to enhance operational efficiency in automated systems. figure 4: tool position (x, y, z) over time pa ge 66 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 59-67, 2025 reward function and optimization outcomes during the training process, it was observed that the rl agent possesses the capacity to minimize the time taken to complete the task and to enhance the task accuracy. the drive function was constructed in terms of effort with special emphasis put on the approachavoidance behavior and the amount of energy used to perform tasks in the minimum amount of time. this enhances the business value because any improvements in energy utilization efficiency and the rate at which tasks are accomplished is central to attaining the goals of industrial robotics. watching the learned behavior of the arm(activity presented in figure 4),one can understand that the optimization leads to more regulated and precise movements, which are necessary to reach high accuracy. correlation matrix to optimization the correlation analysis shown in figure 5 tends to show that certain joint angles and velocities, namely joint 1 and joint 2 are well correlated, positive values indicating that those two joints move in similar methods. it can be used for an efficient management of the tasks that have strong coupling between movements, because then effort could be spent to reduce energy consumption and improve the efficiency of the task. on the other hand, joint 6 had lower correlation between other joints which informed the notion that the movements of that joint were less likely to be coordinated. this independency may be used to make specific kinds of motions that would help in particular actions, and hence enhance the general capacity and versatility of the system. incorporating with the business optimization objectives the kinematic and dynamic analysis results and the rl-based trajectory optimization result helps in the realization of the business objectives like the reduction in time taken for the task, better accuracy, and lesser energy utilization. for instance, the movement of the tool through the three-dimensional space (figure 4) shows how the robotic arm can execute delicate tasks since it does not jerk. in addition, it eliminates oscillations of joint velocities leading to better accuracy of the task as well as minimal wear and tear hence cutting costs in the long run. in addition, another important aspect is the so-called dynamic resources, which enable people to intervene in the process of performing tasks in response to the fluctuations in the work’s requirements. conclusion in this study, there is strong evidence of the improvements that can be obtained through the use of ai, especially when applied to the ur5 robotic arm. with respect to angles, velocities, and positions of the joints in relation to different tools, it is possible to establish how they dictate the motion pattern of the arm. kinematics models determined how the arm curved and was thus useful in dictating how precise, flexible, and general the arm was in accomplishing various tasks, dynamics, on the other hand, provided crucial details on the forces and torque required for stability in dynamic terrains. the utilization of reinforcement learning ai in this case helped the robotic arm to be adaptive to changing reactions, and make adjustments as to be precise, to take shorter time to complete a specific task and use less energy. these were brought in by reducing some movements, improving the trajectory to be followed, increasing the efficiency and sustainability of the tasks. using the real feedback, the rl agent improved the performance of the arm with passage of time when executing the operation. this paper unveils how optimization by artificial intelligence has a revolutionary effect on robots especially in the manufacturing sector, logistics, and the healthcare sector to enhance precision, efficiency, and sustainability. in this regard, integrating the ai with real-time kinematic and dynamic arrange and helps in enhancing the productivity and reduce cost and manoeuvre to scale up the operation hence proving the way for efficient intelligent auto-system that definitely is the future in industrial automation. future work future directions for this research include: applications to other fields of automation research how the ai-optimized robotic arm can be interfaced with other automation systems such as vision systems, as well as smart relational and decision-making software to have a fully automated plant. business application expansion subsequent utilization of this ai-driven robotic optimization in other areas such as; food production industries, pharmaceutical manufacturing industries, construction industries, etc. the solutions for possible recognition in real-time the features for future models can be incorporated to enable real-time adjustment based on changes in the environment, for example, supply and demand, production rates or plans, and other parameters to operate with the highest efficiency in various environments. references azman, s. n., ramli, f., & azami, n. (2023). adoption of artificial intelligence for improved supply chain and logistic performance: a conceptual insight. kwpublications.com. babkin, a., shkarupeta, e., kabasheva, i., rudaleva, i., & vicentiy, a. (2022). a framework for digital development of industrial systems in the strategic drift to industry 5.0. international journal of technology, 13(7). benabed, a., & boeru, a. c. (2023). globalization beyond business sustainability, energy and the economy of the future. in proceedings of the international conference on business excellence, 17(1), 1569-1583. bongomin, o., mwape, m. c., mpofu, n. s., bahunde, b. k., kidega, r., mpungu, i. l., ... & ngulube, g. (2024). digital twin technology advancing industry 4.0 and pa ge 67 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 59-67, 2025 industry 5.0 across sectors. ssrn 5072457. calzada-garcia, a., victores, j. g., naranjo-campos, f. j., & balaguer, c. (2025). a review on inverse kinematics, control and planning for robotic manipulators with and without obstacles via deep neural networks. algorithms, 18(1), 23. elgohr, a. t., khater, h. a., & mousa, m. a. (2025). trajectory optimization for 6 dof robotic arm using woa, ga, and novel wga techniques. results in engineering, 104511. hazem, z. b., & guler, n. (2025). machine learning models for real-time adaptation of robotic systems in dynamic production environments. automation in construction, 101, 111-120. hazem, z. b., guler, n., & el fezzani, w. (2024). study of inverse kinematics solution for a 5-axis mitsubishi rv-2aj robotic arm using deep reinforcement learning. in business sustainability with artificial intelligence (ai): challenges and opportunities (volume 2, pp. 381-393). cham: springer nature switzerland. hossain, s., & nur, t. i. (2024). gear up for safety: investing in a new automotive future in china. finance & accounting research journal, 6(5), 731-746. hossain, s., akon, t., & hena, h. (2024). do creative companies pay higher wages? micro-level evidence from bangladesh. finance & accounting research journal, 6(10), 1724-1745. khater, h. a., & mousa, m. a. a. (2023). a study on the application of ai and machine learning in robotic systems for dynamic adjustment and control. journal of intelligent robotics and systems, 33(2), 120-135. mohr, w., kaloxylos, a., trichias, k., & willcock, c. (2024). the european vision for 6g smart networks and services. ieee communications magazine, 62(4), 10-12. nakib, a. m., khan, p., ullah, m. m., kawser, m. l., jayed, a. k. m., & zim, s. k. (2024). harnessing advanced nlp techniques for automated personality analysis and future behavior prediction from social media posts. eng. technol, 4(4), 98-106. nakib, a. m., li, y., & luo, y. (2024, september). retinopathy identification in oct images with a semi-supervised learning approach via complementary expert pooling and expert-wise batch normalization. in 2024 9th optoelectronics global conference (ogc) (pp. 170-174). ieee. parvez, m. o., hossain, m. s., patwary, a. k., elkhwesky, z., ur rehman, s., & ali, f. (2024). meeting the needs of physically disabled tourists: use of service robots toward the hotel attachment. journal of hospitality and tourism technology, 15(4), 574-591. rahaman, a. k., & ali, r. (2024). autonomous robotic systems for real-time task scheduling and routing in logistics. international journal of industrial engineering and automation, 56(7), 379-390. rahaman, a. k., ali, r., & iqbal, s. (2025). real-time robotic motion optimization using machine learning: a case study on industrial applications. journal of robotics and artificial intelligence, 34(2), 112-125. victores, j. g., & calzada-garcia, a. (2025). dynamic modeling and control of robotic systems for manufacturing applications: a machine learning approach. journal of manufacturing science and engineering, 147(3), 213-225. asme. vyaas, s. (2025). advances in kinematic control of robotic arms for precision manufacturing. international journal of robotics and automation, 40(1), 55-72. pa ge 1 pa ge 20 american journal of smart technology and solutions (ajsts) university 4.0: digital transformation of higher education evolution and stakes in morocco hind tamer1, zakaria knidiri1* volume 2 issue 1, year 2023 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: february 13, 2023 accepted: march 09, 2023 published: march 17, 2023 facing the challenges of the new industrial revolution, the deep coupling between universities and industry 4.0, the integration of information and communication technologies in education, and the enhancement of the ability to serve society on the basis of internal and external synergy should become the common choice of different types of universities. the university plays an important role in the n development in any advanced economy. in the age of knowledge and globalization rapid technological changes involve new disruptive processes. in this permanent challenge, it is necessary to adapt to the digital transformation, in order to better respond to the needs and challenges of a constantly changing environment. it is necessary to pay attention to technological advances, to a total transformation in the new university, university 4.0, in order to face the challenges of the technological development and the efficiency of universities in morocco, it is necessary to introduce modern technologies, blockchain technology, artificial intelligence, chatbots..... into the sphere of moroccan universities. keywords morocco, digital transformation, ict, digitalisation, university 4.0 1 management sciences, faculty of legal, economic and social sciences of marrakesh, cadi ayyad university marrakesh, morocco * corresponding author’s e-mail: z.knidiri@uca.ac.ma introduction this digital transformation is about focusing the development of universities and higher education institutions on the application of technology, as is the case in other sectors. dewar (2017) defines university 4.0 as a university that is other-oriented, to primarily serve students, outward-looking, engaged, and connected to the surrounding productive environment, in line with barnett’s (2017) concept of the ecological university, which refers to the interconnection of the university with various ecosystems (knowledge, social institutions, people, economy, learning, culture, and natural environment). precisely, in order to follow a logical sequence until reaching version 4.0, barnett describes the evolution of the university in different phases: a university 1.0, which would be the metaphysical university developed in medieval times, with a strong presence and dominance of spiritual and religious beliefs. version 2.0 is born in postindustrial societies, more focused on the deployment of research within the university as a driver of technological progress oriented towards economic development. it would correspond to the universities created from the 15th century onwards, with teaching that was increasingly open to different approaches to thought. a few centuries later, version 3.0, which could be called an entrepreneurial university, defined by barnett, as a university for itself, serving many different functions and communities, but above all concerned with optimizing its own interest or strategy in an increasingly competitive world. this university 3.0 is also defined by pulido, (2019) as an advanced and social university, developing in europe in the 19th century, combining teaching and research functions, with self-governance and institutional autonomy. this article introduces us to the digital era, trying to show the relevance and impact it has on contemporary society. it identifies the main disruptive technologies that are shaping it and how, from these, the models and processes of organisations are being transformed, generating profound, abrupt and at the same time ephemeral changes. from this point of view, the challenges of universities in the digital age, the debate on the future of the university cannot be postponed in the face of the challenges of industry 4.0. this transformation is seen as a necessity that must be tackled without delay, but with a critical view and taking into account the particularities of each institution. we also try to sketch some ideas on the role that higher education institutions could play in this coupling between university and industry 4.0. methodology in order to understand and analyze the close link between universities and industry 4.0, we focused on the digital transformation of higher education as a common choice of different types of universities, in this sense a systematic literature review (slr) was conducted. this form of review is based on the application of different scientific strategies that limit bias, through a systematic collection of information as well as a critical evaluation and synthesis of all relevant studies on a specific topic (cook et al., 1995). in this work, we applied the methodology proposed by tranfield et al. (2003), which can be summarized in three main steps: planning the review, conducting the review and presenting the results and reports. literature review plan the systematic literature review dewar, (2017) argues that university 4.0 will provide ondemand learning in multiple formats, with continuous transfers between different modalities, with more intense collaboration between universities and the productive fabric in a digitised environment. in this context, pulido https://journals.e-palli.com/home/index.php/ajsts pa ge 21 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 20-28, 2023 (2019) interprets university 4.0 as a university that undergoes such a disruptive change that it requires a radically new university (4.0) in organisation, technology, and education-research strategy that responds to the needs of a profoundly evolved society. indeed, digital technologies are driving digital transformation, a new form of organisation and increasing and unpredictable changes, generating a wide range of new challenges. it is therefore university 4.0 that corresponds to a modern university, as a metamorphosis of previous versions in a technological environment that is advancing into the digital age and must meet the demands and commitments of a globalised society. due to the digital transformation, the university has ceased to be what it used to be. the university today has to follow the trends imposed by globalisation and the increasing use of technology by organisations and individuals. there is a concern about whether the university can survive the intelligent world resulting from the advances in digital technology brought about by the fourth industrial revolution. is it possible to envisage the university of the future as a smart 4.0 university that has begun to emerge? the answer to this research question thus contributes to enriching the literature both on the notion of university 4.0 and on the new dynamics integration of information and communication technology for education. conducting the systematic literature review in order to target the articles to be explored, we chose to use the following databases which we believe provide an adequate picture of the current literature in the field: scopus, web of science and google scholar. we also used various websites including we then identified a series of keywords in order to select the most relevant articles for the research. in particular, we chose scientific articles that included the following terms in their subject, title or abstract: “digitization” or “university 4.0” and “ict”. in a second step, we introduced the terms “artificial intelligence” or “blockchain” in order to better focus the research on the university 4.0 theme. regarding the period of time to be covered, we have chosen a sufficiently long period to adequately present the state of the art in the field. results the economic impact of digitisation is accelerating as countries evolve in their degree of digitisation. the most digitally restrictive economies benefit less, largely because they have not yet developed an ict ecosystem to reap the benefits of digitalisation (cerezo, et al., 2017, tamer, 2022). according to these authors, digital transformation is understood as a relatively new and recent phenomenon, and an organisation cannot be considered to have reached a final state of maturity in this area or to have managed to define it in its entirety. considered as a new paradigm, digitalisation, as a new way of doing things, has a great impact on the way universities carry out their main missions and functions (juanes and rodríguez, 2020). universities must also provide students with the skills and knowledge they need for a very different future. in this new educational landscape, the digital transformation of higher education is essential (tamer, 2022). of this conception, very few intend to create 4.0 university models, which suggests that they continue to rely on the current university model in terms of organisational form. it is therefore important to understand that the process of digital transformation implicitly involves a change in the organisational model. digital transformation represents new opportunities for business strategies, integrating technology, streamlining processes, preparing teams to work and collaborate with digital tools and establishing business logic or processes with the digital economy, thus achieving better performance. according to garcia, (2018), it can then be deduced that, digital transformation allows institutions to adapt a socially responsible and ethical business model, allowing them to apply a scalable development model, without forgetting that they influence to reduce environmental impact by streamlining processes and reducing consumption of non-renewable materials. according to gaibor, (2020), digitalisation is the great driver of wealth creation, an important point in this analysis is that digital transformation brings greater productivity, agility, quality, innovation, cost efficiency, as well as many other aspects, both for digital and offline businesses, where the key is to understand how digital techniques and tools can impact and grow a traditional business or institution, but in a joint and strategic way. digital transformation is not just a technological problem that is solved by an injection of technology. according to barro (2018), the digitalisation of a university requires first and foremost an investment effort in ict infrastructure and resources. however, to make it a 4.0 university, the relationship between teachers and new students (millennials, generation z) also needs to be reformulated, where traditional channels are no longer a priority but complementary. as explained in the article, the simple use of technology is not enough to take the step towards digital transformation. according to gaibor, (2020), it is necessary to raise awareness and train the whole team, so that they can make the most of digital tools in their daily work. looking at the issue from different angles, it is clear that the current rigid educational structures need to be changed, barriers need to be broken down and technology needs to be used to provide educational content at all times. we need to promote a more fluid and flexible education to better adapt to different needs, as the current rigidity of university structures does not, in some cases, promote adequate education (gaibor, 2020). from university 1.0 to university 4.0 the university has played, and must continue to play with greater intensity, an important role in the development of innovation in any advanced economy. after all, it is the natural space in which knowledge is developed and https://journals.e-palli.com/home/index.php/ajsts pa ge 22 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 20-28, 2023 promoted and, as such, must be transferred to society. globalisation, new processes and, consequently, new working methods and rapid technological changes define the changing environment in which the university operates. in this permanent challenge, it is necessary to adapt to the new times and, in particular, to strengthen the role of resilience and adaptation to change in the university context. it is no longer just about the agility with which the organisation evolves in the present moment, but the ability to anticipate the future era, for which we must always be vigilant and constantly reflect. in this context, this vigilance implies the vision of a digital world, to which the university has already opened its doors but where there is still a long way to go. we are talking about a disruptive era that is changing the world around us or, in global terms, the 4.0 era, which applied to any field denotes a commitment to the digital world, to the digitalisation of processes or to what is known as digital transformation. the regular use of the internet by millions of people around the world has led to the development of the internet of things, which is a growing advance in connecting digital devices and objects to each other, interacting in such a way that there are no temporal or spatial boundaries. we are witnessing a new industrial revolution that affects the intellectual-intensive jobs of the 21st century, whereas in previous industrial revolutions it was mainly manual activities that were affected. this is the fourth industrial revolution, that of the fusion of technologies, where the combination of advances in the development of robotics and artificial intelligence, the collection and processing of massive information or big data have and will have an impact on the economy and therefore on the qualification needs of jobs in all productive sectors. the digital transformation implies focusing the development of the university, higher education institutions, on the application of technology, as is happening in other sectors (tamer, 2022). dewar, (2017), defines the 4.0 university as a university that is other-oriented, primarily to serve students, outward-looking, engaged and connected with the surrounding productive environment, which refers to the interconnection of the university with various ecosystems. specifically, to fit into a logical sequence until reaching version 4.0, barnett, (2014), describes the evolution of the university in different phases: a university 1.0 which would be the university developed in the medieval period (the main european universities date back to the 11th century), with a strong presence and dominance of spiritual and religious beliefs and which evolved towards liberal arts type education. version 2.0 appears in the sphere of post-industrial societies, more focused on the deployment of research within the university as a driver of technological progress oriented towards economic development. it would correspond to the universities created from the 15th century onwards, with teaching that was increasingly open to different approaches to thought. a few centuries later, version 3.0, which could be described as the entrepreneurial university, defined by barnett, (2014), as a university for its own sake, serving many diverse functions and communities, but above all concerned with optimising its own interest or strategy in an increasingly competitive world. this university 3.0 is also defined by pulido, (2019) as an advanced and social university, developing in europe in the 19th century, combining the teaching function with the research function, with self-governance and institutional autonomy. dewar, (2017), argues that university 4.0 will provide ondemand learning in multiple formats, with continuous transfer between different modalities, with more intense collaboration between universities and the productive fabric in a digitised environment. in this context, pulido, (2019), interprets university 4.0 as a university that undergoes such a disruptive change that it requires a radically new university (4.0) in organisation, technology and education-research strategy that responds to the needs of a profoundly evolved society. indeed, digital technologies are leading to digital transformation, a new form of organisation and increasing and unpredictable changes, generating a wide range of new challenges. it is therefore university 4.0 that corresponds to a modern university, as a metamorphosis of previous versions in a technological environment that is progressing in the digital age and that must respond to the demands and commitments of a globalised society. according to barth, rieckmann, (2016), the major changes we are seeing with information and communication technologies make traditional approaches to classical pedagogy and its conventional credentials obsolete. among the soft skills that will soon be indispensable are higher-order cognitive thinking, innovative adaptive thinking, cognitive load management, multiple literacy, complex situation solving, social skills, elastic skills and cross-curricular skills to accomplish tasks of a changing nature. barth, rieckmann, (2016), add that digital skills will also lead to a diversity of synaptic and social connections that will become increasingly flexible and adaptive. according to deward, (2017), the typology of universities provided by professor emeritus barnett, (2014), from the institute of education at the university of london presents us with the following classification: university 1.0: this would be the university, in the service of god, which appeared in medieval times. the first stages of this university were structured around specialised communities that eventually evolved into the tradition of liberal arts education (le goff, 2008). university 2.0: could be seen as the research university that has emerged in post-industrial societies, where universities have become the focal point of research-led technological progress. the great post-war expansion is clearly focused on research for economic development. based on the massification of education, with the teacher as the main provider of knowledge and the student as a passive receiver who absorbs the content (shchedrovitskii, 2011). university 3.0: is described as the entrepreneurial https://journals.e-palli.com/home/index.php/ajsts pa ge 23 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 20-28, 2023 university, functioning, in barnett’s, (2014), terms, as a university ‘for itself ’, serving many diverse functions and communities, but primarily concerned with maximising its own self-interest. based on the integration of computers and the internet into teaching and learning, thereby increasing access and equity (li, 2020). university 4.0: refers to the green, outward-looking university, deeply connected to industry and the communities around it. it is committed to meeting the needs of its students. it relies on high-speed internet, mobile devices, technology platforms and digital applications, which facilitate personalised learning anytime, anywhere and change the transmission roles of teachers (efimov, and lapteva, 2017). university 4.0 is an apt description of how universities around the world must respond to the new economy and associated trends such as digital disruption and changing labour markets. if universities want to remain relevant, they must undertake revolutionary changes at the organisational, operational, structural, pedagogical, socio-cultural and cognitive levels today (aladyshkin, odinokaya, safonova, & kalmykova, 2020). university 4.0 is fully in line with the fourth industrial revolution. we are talking about new platforms that will use artificial intelligence algorithms in combination with the internet of things (iot) to personalise student learning. this will force traditional professors to take on new teaching roles that transcend the delivery of declarative content (aladyshkin, odinokaya, safonova, & kalmykova, 2020). this revolution is centred on the development of new information and communication technologies in education incorporating robotics, automated systems, blockchain, fintech, bots, deep learning, 5g technology and cybersecurity systems. all of these will impact our daily lives, social relationships, work and learning experiences for life. today’s student is not just limited by a teacher-led educational model, but draws learning from a variety of information sources at a personalised pace. they do not only expect academic excellence, but also desire personalised excellence by expanding their horizons of possibilities. we cannot be satisfied with a university that adapts to new circumstances and tries to integrate emerging technologies (kazimirov, 2018). we need a radical change and this requires: ending the problems of massification, adopting the procedures for professional promotion, softening the relationship with the social environment, strengthening student engagement, implementing realistic strategic plans (antonio pulido, 2019). for james, (2019), one of the contemporary methodologies is accelerated distance learning, i.e. the idea that students learn theoretical knowledge at a distance through digital means, while ensuring that practical skills are acquired in physical environments. it is a flexible form of learning that requires responsibility and good time management to develop skills based on an increasing economy of autonomy. at this point, it is a matter of building and investing more in a robust educational ecosystem, not to replace or displace it, but as a form of flexibility and adaptability. according to aladyshkin, et al, (2020), what is required of education today is not a solid classical education, not because of the modality, but because of the social diversity of the modern world. higher education institutions are moving towards a more personalised form of learning. aladyshkin, et al, (2020), add that by using data and tracking student performance, universities will be able to identify students who are struggling and provide them with learning strategies optimised to meet their needs. data analysis will be used to treat each student, understanding that each student’s learning needs and desired outcomes will be different. according to villalobos, and pedroza (2019), the central idea supported in this article is that there is no element of the university that is not undergoing profound changes with the use of new communication and learning technologies. university life is being renewed with its productions, processes and tasks; training, teaching, learning, research, curriculum, etc., are all being changed by the integration of information and communication technologies for teaching. thus, various documents from universities in different parts of the world have been worked on and creativity has been used to give shape to university 4.0 (villalobos, and pedroza 2019). the virtuous circle of innovation in the university in the transition to the future. the debate on the future of the university cannot be postponed in the face of the challenges of 4th industrial revolution, where developments in technology, physics and biology converge. the archetypal monolithic university, composed of disciplinary islands focused on essentially theoretical teaching, with atomised contents disconnected from real problems and with informational pedagogical practices that favour repetitive and contemplative learning, which is also of little impact in making contributions to the future (gueye, and exposito, 2020). according to gueye, and exposito, (2020), the university in the knowledge society is obliged to reinvent itself because otherwise, with its traditional model, it will be unable to meet the needs and challenges of an increasingly dynamic world. recent experience shows a historical truth: universities that work hand in hand with technological advances are best placed. by investing in research and development, they are creating innovations and acquiring a leading role in the current new technological configuration. according to madaliуeva, et al., (2020), it is the faculty that enables the university to develop, to participate in opening up and solving the challenges present in the new fields of knowledge. in order to reinvent itself, the university must implement and encourage the integration of new information and communication technologies for teaching in teaching and learning, always on the basis of scientific research, while promoting new forms of organisation, new methods. https://journals.e-palli.com/home/index.php/ajsts pa ge 24 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 20-28, 2023 indeed, according to lapteva, and efimov, (2016), the trend of university education in industry 4.0 is moving towards the innovative research-based university. lapteva, and efimov, (2016), indicated that the innovative university is the one that makes research its main development axis. on the one hand, new knowledge is provided and, on the other hand, the learning and teaching system is redefined. the result is a university model characterised by dynamic feedback between these two aspects. the best ranked universities are those that encourage this kind of flow (lapteva, and efimov, 2016). according to, madaliуeva, et al, (2020), with the upgrading of traditional industries and the advent of industry 4.0, the economic structure and industrial mode have undergone unprecedented changes, which means that universities have to adapt to the demand and provide responses. madaliуeva, et al, (2020), add that the construction and exploration of new information and communication technologies for education, on the one hand, actively adapt to the changing demand for profiles in the context of industrial production and technological innovation, and on the other hand, promote institutional reform and internal development of universities. today, the university is changing, there is no country that does not rethink the change of the model and function of its university. we can even say that the country that resists change is endangering the existence of this thousandyear-old institution (madaliуeva, et al., 2020). various factors require a change in the university, the most representative being technological development, what society expects from the university and that it responds to economic, social and political development, which implies an internal renovation of university processes (lapteva, and efimov, 2016). the university is challenged because technological advances are not always born within it, the dominant dynamic is that of an academic science and technology, that is, it is formed only to reproduce, not to generate new advances (gueye, and exposito, 2020). with the fourth industrial revolution, the university has to deal in a different way with its dynamics in the training of professions, moving towards teaching and intelligent learning, devoting itself more to scientific research and technological development, with this, its vision is transformed and its model is mobilised towards open and flexible forms. not all universities are taking the shift in the same way (lapteva, and efimov, 2016). the best-placed universities are at the forefront of change, while others are slow and lagging behind. the bestplaced universities are those dedicated to research with models linked to the economic and social development of the country (gueye, and exposito, 2020). they generate technological advances and produce economic resources. this is in contrast to poorly positioned universities, which maintain the stagnant practices of the past, dedicated to training based more on academic discipline than on the generation of new knowledge and technical and technological resources. there is an uneven development of the university in the world that coincides with the economic, political and social situation of each country (gueye, and exposito, 2020). therefore, the transformation of the university represents the transformation of the country or vice versa, as there is a reciprocal relationship between the university and national development. in general, the transformations of the university in the world have an accelerated dynamic, which requires addressing the dimensions that need to change in the direction of progress. villalobos, and flores (2019), have identified some axes of university transformation: • diversification of university modalities, nowadays pluriversity is a fact, they coexist with face-to-face mobility, alternative and complementary distance, open, digital and mixed modalities that are increasingly positioned. • with the development and application of information and communication technologies (ict), curricular modalities are being transformed into flexible, open, networked, integral and individualised itineraries, combined with platforms such as coursera, udacyt, scolartic, mooc courses and standard university credits. • with artificial intelligence, pedagogical relationships are being transformed, as there is now a relationship between teachers, robots and students and learning is an unprecedented experience, generating a ramification of learning types: adaptive (big data with learning analytics), 3d, gamified, flipped learning, adaptive, with virtual reality, multimodal, storytelling-based learning, and so on. • there is a new academic ecosystem of training, learning and management based on the technology trend of 5g (fifth-generation 5g mobile networks), digital assistants, robots, augmented reality, global educational platforms, communities of practice and the use of blockchain (schwab, 2017). • rethinking the training of professions by strengthening the themes of humanities and human development tending towards the connection and expression of smart, sustainable and coexisting cities. the building blocks of university transformation are many, the university 4.0 model is super-connected in an interactive environment between humans and new technological species creating a new university ecosystem of teaching co-existence for the continuous development of learning for disruption (giesenbauer, det müllerchrist, 2020). according to colombo, et al, (2020), the use of hard technological applications derived from intelligent computing allows for the superconnection of knowledge at previously unsuspected levels, artificial intelligence stands out, starting from the use of simple artefacts to the use of machine learning, we observe how in universities the use of apps, robots and virtual intelligence increases its application. at the same time, the forms of learning with soft technologies are multiplying, they are abandoning their traditional form of being secondary in the training processes, they are now the central character, we are in https://journals.e-palli.com/home/index.php/ajsts pa ge 25 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 20-28, 2023 the era of augmented learning acquired by our human skills and simultaneously powered by nanotechnological artificial bodies, for this reason we speak today of a trend towards nano-credentials and no longer of the old university references of diplomas and degrees (colombo, et al., 2020). flores, (2018), note that we are living in a time of crisis that requires the university to move to where the people are, not that the people move to the physical address of the university. flores, (2018), adds if administrators, faculty and non-teaching staff resist the demands of the millennial society, they will begin to feel the effects of the erosion caused by the proliferation of emerging educational agencies that will launch their attractive academic offerings. these conventional institutions will not only face the challenge of placement rates or employability of their graduates, but will also have to deal with the socio-cultural challenges and emerging crises of our time (colombo, et al., 2020). the educational community will have to work together to find solutions to crises on and off campus. according to colombo, et al. (2020), before long we will see the academic offerings of departments lose their footprint, as social demand will dictate what people want to learn, how to learn it, when, where and with what resources. duc, et al., (2018), suggest that universities should partner with the local community, industry and society at large to co-design and co-implement a stronger higher education system. duc, et al, (2018), add that their operational and pedagogical business model is incompatible with the times we live in. we are now living under the threat of a micro-organism that has crippled the entire world. the initial solution was to save the semester with any videoconferencing technology that would be used for remote academic continuity. as the summer holidays approach, higher education institutions will have to prepare for longer periods of remote administrative support and online teaching. what we know as the physical space for working and learning will never be the same again (duc, et al., 2018). according to giesenbauer, det müller-christ, (2020), university 4.0 is not the one that improves on what other universities do, but the one that dares to do things differently. it is the one that looks at the segments that other institutions are unable to look at. an agile educational organisation is one that looks at different latitudes, looking for new learning niches that society needs. it is very clear that no single technology will replace administrators, teachers and non-teaching staff. those who will replace them will be the expert users of cyberhuman interfaces (gueye, and exposito, 2020). they will take them out of their fragile safety and comfort zones. technology is not just about the digital gadgets we acquire in our workplaces and homes. real technology is about effectively connecting our brains to devices to create new solutions to the emerging crises of our time (gueye, and exposito, 2020). creative university after the text edit has been completed, the paper is ready for the template. duplicate the template file by using the save as command, and use the naming convention prescribed by your conference for the name of your paper. in this newly created file, highlight all of the contents and import your prepared text file. you are now ready to style your paper; use the scroll down window on the left of the ms word formatting toolbar. we are now living in a unique time where the digital revolution is changing the way most people live on the planet. central to this change is the development of digital universities. it has already become a driver of the economy. digitalisation improves the conditions for doing business, increases the level of education and computer literacy of the population and, in general, the level of competitiveness of the nation. digital technology has such a profound effect on the competitiveness of countries that nations around the world are looking to modernize the industry. the changes brought about by technology, which are redesigning production processes, are helping to increase the efficiency and quality of services. as the experience of the rest of the world shows, digital technologies make a tangible contribution to gdp growth. this is why some countries have adopted entire national programs for in universities. the whole world is now embracing digital transformation. to date, morocco country is implementing national digitalisation programs. a first approximation of the state of innovation in the world can be found in the recent report global innovation index 2021 (gii), indeed morocco occupies the 77th place with a score of 29.3 (table 1). the ranking is dominated by switzerland, sweden, the table 1 : global innovation index ranking 2021 gii rang economy score group rank region rank 76 oman 29.4 47 11 77 morocco 29.3 8 12 78 bahrain 28.8 48 13 united states, the united kingdom, and south korea in these top five positions (table 2). as noted, the performance in investment and digital transformation in moroccan universities, there is a deficit of innovation at the national level. moreover, change or improvement towards advanced positions is difficult and slow for a country such as morocco. it is also interesting to note the reference to the most important regional science and technology clusters in the world, with the understanding that innovation activities tend to be geographically concentrated. in this regard, the united states remains the country with the largest number of innovation clusters (26). another source for assessing which countries have the most innovative universities is the reuters ranking: the world’s most innovative universities, which identifies and ranks the educational institutions around the world with the best results in innovation, understood as the best performance in advancing science, inventing new https://journals.e-palli.com/home/index.php/ajsts pa ge 26 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 20-28, 2023 table 2 : global innovation index ranking 2021 gii rang economy score group rank region rank 1 switzerland 65.5 1 1 2 sweden 63.1 2 2 3 united states of america 61.3 3 1 technologies and stimulating new markets and industries. another source for assessing which countries have the most innovative universities is the reuters ranking: the world’s most innovative universities, which identifies and ranks the educational institutions around the world with the best results in innovation, understood as the best performance in advancing science, inventing new technologies and stimulating new markets and industries. in its latest edition, referring to 2021, stanford university in the united states leads the ranking of leading universities in scientific and technological innovation, stanford maintains its top spot year after year because it produces a steady stream of innovations that are cited by other researchers in academia and private industry. this type of influence is a key measure in the ranking of the world’s most innovative universities, which was compiled in partnership with clarivate analytics and is based on proprietary data and analytics, including patent filings and research paper citations. followed by massachusetts institute of technology (mit), and harvard, all of which have held their positions for seven consecutive years since reuters began producing the rankings. in fact, no moroccan university has made it into the top 70 universities, so these results show that morocco continues to have a low r&d investment effort compared to other economies of similar size, and this circumstance conditions any effective progress. barnett, (2017) points out that universities are not living up to their potential and responsibilities in a constantly changing and challenging environment. the truth is that in the age of knowledge and globalization, the university must constantly reinvent itself from becoming an obsolete institution, so that it can better respond to the needs and challenges of a changing world. this reinvention implies paying attention to scientific and technological advances, developing them, integrating them and being more active in the innovation strategy. theoretical and managerial implications from our study we can recommend moroccan universities to consider adopting the following practices: as a first link, blockchain which is gradually being implemented not only in all areas of business, but also in higher education, as the interaction between business and science contributes enormously to the growth of innovative products and services. in higher education, the demand for innovation, the possibilities offered by digital technology, are very relevant today. their necessity is associated with objective processes such as the volume of information that is increasing at an enormous rate, and the capacity of students to absorb it (vasilieva, 2017). in morocco, the transition to digital media is progressing; especially in higher education institutions which are increasingly moving away from paper-based media. indeed, the collection of information on paper creates an additional workload also for administrative staff and allows for changes in documents; reporting forms on the results obtained in the different educational institutions may not match, which reduces the efficiency of the staff; the lack of a comprehensive database on graduates with specific skills makes it difficult for employers to find the right specialists; the lack of an open database on the employment of graduates and their transfer to other jobs does not allow higher education institutions. in order to solve all these problems and increase the efficiency of universities in morocco, it is necessary to introduce modern technologies, blockchain technology, into the sphere of universities (tamer, 20. secondly, the technologies of augmented reality and virtual reality which constitute fundamentally new means and methods of interaction between teachers and students, which guarantee the effective realization of pedagogical activities in the sphere of higher education. the analysis carried out allows us to conclude that the application of innovative technologies in the educational process contributes not only to the progress of students, but also to their interest in the learning process. in morocco, the application of augmented reality and virtual reality technologies in the student learning process will, on the one hand, facilitate the task of the teaching staff and, on the other hand, will significantly help students to master knowledge, form their skills and abilities and, overall, will have a positive effect on the training of graduates (tamer, 2019). thirdly, chatbots in higher education, which are conversational assistants or better known as bot or chatbot or chatterbot, which can be defined as a virtual assistant, are a set of computer programs that possess the ability to maintain a conversation with a human being through natural language. similarly, a conversational agent can be understood as an automatic system capable of emulating a human being in a dialogue with another person, with the aim that the system provides certain information or performs a certain task. the objective of chatbots in higher education is always to achieve interaction based on models similar to those used by humans, which is achieved through dialogues, as they are programmed to have the capacity to analyse the environment and propose solutions to problems, interpreting emotions and contributing to the teaching-learning process to the maximum extent possible. fourthly, artificial intelligence, which is defined as a technology whose value on the market is incalculable, both in the present and in the future, but we should not only refer to the monetary value, but also analyse the value it has for the optimisation of non-commercial processes, as in the higher education sector, artificial intelligence to https://journals.e-palli.com/home/index.php/ajsts pa ge 27 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 20-28, 2023 be a turning point in the changes of traditional higher education paradigms, although the pedagogical modalities at all levels of education systems are being adapted, given the current technological tools, virtual teaching modalities are becoming more common in the education policies of developed countries. finally, artificial intelligence can optimise the use of these valuable resources, as one of the main problems today is the under-use of technological tools or their use in isolation and out of context. conclusion the landscape of contemporary education is diverse. higher education systems are now developing as institutionally complex structures that align learning with the organisations of different professional spheres of society and digital transformation (aladyshkin, et al., 2020; karpov, 2013). socially the most important and economically significant element of this structure is the higher education sector. its institutional base is composed of scientific institutions, high-tech companies, innovative enterprises, industrial consortia, innovative growth institutes giving rise to university 4.0 (aladyshkin, et al., 2020). ecosystems become the place where favourable conditions for the efficient transfer of technologies and scientific and technical innovations are created. university 4.0 becomes the basis of global competitiveness of national economies, and its ecosystem forms new fastgrowing industries, promising technological markets, economically advanced administrative-territorial spaces. finally, based on the recommendations of our research, universities in morocco can promote active, constructive and real learning, while provoking a process of innovation. this requires the involvement of all stakeholders in the process of changing the design of the higher education process in morocco, which was based on the practical application of the content covered in the subject and on the integration and didactic use of digital technologies. references aladyshkin, i.v., kulik, s.v., odinokaya, m.a., safonova, a.s., & kalmykova, s.v. (2020). development of electronic information and educational environment of the university 4.0 and prospects of integration of engineering education and humanities. integrating engineering education and humanities for global intercultural perspectives. ieehgip 2022. lecture notes in networks and systems, 131, springer, cham. https://doi.org/10.1007/978-3-030-47415-7_70 barnett, r. (2017). the ecological university: a feasible utopia. routledge. https://doi.org/10.4324/9781315194899 barro, s. (2018). de la digitalización de las universidades a las universidades digitales. cerezo, p., magro, c., & salvatella, j. (2017). on digital transformation and its socioeconomic impact. colombo, a. w., veltink, g. j., roa, j., & caliusco, m. l. (2020). learning industrial cyber-physical systems and industry 4.0-compliant solutions. in 2020 ieee conference on industrial cyberphysical systems (icps) 1, 384-390. ieee. https://doi.org/10.1109/ icps48405.2020.9274738 cueva gaibor, d. a. (2020). la tecnología educativa en tiempos de crisis. conrado, 16(74), 341-348. cueva gaibor, d. a. (2020). transformación digital en la universidad actual. conrado, 16(77), 483-489 duc, n. h., chung, n. h. t., huy, n. x., lan, m. t. q., lieu, t. t. b., thuy, h. q., & loc, n. (2018). towards the higher education 4.0–characteristics and criteria. vnu journal of science: policy and management studies, 34(4). https://doi.org/10.25073/2588-1116/ vnupam.4160. efimov, v., & lapteva, a. (2017, november). university 4.0: philosophical analysis. in proceedings of the 10th annual international conference of education, research and innovation, seville, spain, 16-18. https://doi. org/10.21125/iceri.2017.0242 flores, m. a. (2018). linking teaching and research in initial teacher education: knowledge mobilisation and research-informed practice. journal of education for teaching, 44(5), 621-636. https://doi.org/10.1080/02 607476.2018.1516351 james, f. (2019). everything you need to know about education 4.0. quacquarelli symonds. garcía, m. m. (2018). el impacto de la transformación digital en las empresas sociales. https://www. iebschool.com/blog/impacto-transformaciondigital-empresas-sociales-digital-business. giesenbauer, b., & müller-christ, g. (2020). university 4.0: promoting the transformation of higher education institutions toward sustainable development. sustainability, 12(8), 3371. https://doi.org/10.3390/ su12083371 gueye, m. l., & exposito, e. (2020). university 4.0: the industry 4.0 paradigm applied to education. in ix congreso nacional de tecnologías en la educación. https://hal-univ-pau.archives-ouvertes.fr/hal02957371 hammershoj l. g. (2009). creativity as a question of bildung. journal of philosophy of education, 43(4), 545— 557. https://doi.org/10.1111/j.1467-9752.2009.00703.x dewar, j. (2017). university 4.0: redefining the role of universities in the modern era. higher education review, 8. james, s., freese, m., & davison, a. j. (2019). pyrep: bringing v-rep to deep robot learning. arxiv preprint arxiv:1906.11176. https://doi.org/10.48550 arxiv.1906.11176 juanes giraud, b. y., & rodríguez hernández, c. (2020). la formación dual. elementos de análisis para implementación en una universidad ecuatoriana. revista conrado, 16(74), 354-363. карпов а. (2012). engineering platform for the transfer of technologies. voprosy ekonomiki, 7, 47—65. kazimirov, a. n. (2018, november). education at university and industry 4.0. in 2018 global smart industry conference (glosic)ieee., 1-6. https:// doi.org/10.1109/glosic.2018.8570076 mounaim, l., knidiri, z., & tamer, h. (2021). omnichannel retailing, from the focus on consumer https://journals.e-palli.com/home/index.php/ajsts pa ge 28 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(1) 20-28, 2023 behavior through organizational and retailer impact: a systematic review from a marketing perspective. international journal of accounting, finance, auditing, management and economics, 2(4), 302-322. https://doi. org/10.5281/zenodo.5133323 laptevа, a. v., & efimov, v. s. (2016). new generation of universities. university 4.0 le goff, a. (2008). the national archives and the history of education. education history, 119, 82-87. li, j. (2020). autonomy, governance and the chinese university 3.0: a zhong–yong model from comparative, cultural and contemporary perspectives. the china quarterly, 244, 988-1012. https://doi.org/10.1017/s0305741020001071 li, m. a systematic review of the research on chinese character teaching and learning. front educ china 15, 39–7. https://doi.org/10.1007/s11516-020-0003-y madaliуeva, z., kassen, g., sadykova, n., baimoldina, l., & zakaryanova, s. (2020). resources and competencies as major determinants of university models 4.0. in e3s web of conferences 159, 09001. edp sciences. https://doi.org/10.1051/e3sconf/202015909001 martin p., morris r., rogers a., kilgallon s. (2010). what are creative spaces? in: p. martin (ed.). making space for creativity. brighton: university of brighton, 23—26. pedroza flores, r. (2018). a university 4.0 with an intelligent curriculum 1.0 in the fourth industrial revolution. ride. ibero american journal for educational research and development, 9(17), 168-194. https://doi. org/10.23913/ride.v9i17.377. schwab k., (2017). the fourth industrial revolution, penguin random house, new york. schwab, k. (2017). la cuarta revolución industrial. méxico: penguin random house. shchedrovitskii, p.g. (2011). povestka dnia 2010-kh [the agenda of the 2010’s]. lecture series. https://doi. org/10.17516/1997-1370-2016-9-11-2681-2696 strazzieri‐pulido, k. c., s. gonzález, c. v., nogueira, p. c., padilha, k. g., & g. santos, v. l. (2019). pressure injuries in critical patients: incidence, patient‐ associated factors, and nursing workload. journal of nursing management, 27(2), 301-310. https://doi. org/10.1111/jonm.12671 tamer, h., lebzar, b., & abaaoukide, k. (2019). l’impact de la digitalisation des universités sur la motivation des usagers: revue de littérature. revue internationale des sciences de gestion, 2(3). tamer, h., lebzar, b., & knidiri, z. (2022). marketing practices in higher education: case of the cadi ayyad university of marrakech in morocco. international journal of accounting, finance, auditing, management and economics, 3(4-1), 477-490. https:// doi.org/10.5281/zenodo.6614930 hind, tamer, & zakaria, knidiri, the effect of ict use on student engagement in moroccan universities: an exploratory qualitative study. social influence, 72(4), 44. verlag barbara budrich. (2015). empirische forschung zur bildung für nachhaltige entwicklung themen methoden und trends. retrieved march 8 2023 from https://content-select.com/portal/media/cover_ image/55102352-9b64-46a1-aae0-0f44b0dd2d03/500. vasilieva, e. v. (2018). developing the creative abilities and competencies of future digital professionals. automatic documentation and mathematical linguistics, 52(5), 248-256. https://doi.org/10.3103/s0005105518050060 https://journals.e-palli.com/home/index.php/ajsts pa ge 1 pa ge 30 american journal of smart technology and solutions (ajsts) precision agriculture through remote sensing and gis: advancing sustainable farming and climate resilience shahed jahidul haque1, sazib hossain2*, muhammad maruf billah3 volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.4418 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: february 03, 2025 accepted: march 06, 2025 published: march 20, 2025 the simultaneous application of remote sensing technology with gis mapping technologies and precision agriculture presents effective solutions to protect sustainable farming practices and climate-resistant measures particularly in the sensitive environment of bangladesh. this investigation combines the technologies to improve resource allocation while tracking farm health and tuning agricultural operations for bangladesh’s climate-distinctive environment. through remote sensing data collection and gis spatial analytics the research delivers operational insights which assist with soil analysis and water control as well as agricultural productivity assessment in all regions of bangladesh. this integration receives support from precision agriculture tools which use variable-rate technologies together with iot devices to deliver site-specific interventions that fit bangladesh’s agricultural geography. the research shows important outcomes regarding resource conservation because the adoption of advanced techniques has reduced water consumption as well as fertilizer use by twenty-five percent and crop stress detection improved yield estimates by eighteen percent. through analysis the research distinguishes both high-risk locations for droughts and floods and provides strategic methods to protect them. the research demonstrates how technologybased methods are essential for climate-smart agriculture through insights that create a direction for bangladesh officials and practitioners to improve food security and climate resilience. keywords climate resilience, gis mapping, precision agriculture, remote sensing, sustainable agriculture 1 school of electronic and information engineering, nanjing university of information science & technology, nanjing, china 2 school of business, nanjing university of information science & technology, nanjing, china 3 school of artificial intelligence, nanjing university of information science & technology, china * corresponding author’s e-mail: easzibhossain@gmail.com introduction the urgent global needs for sustainable agriculture and climate resilience align especially with bangladesh because the country exists as a climate-change sensitive zone (gopalakrishnan et al., 2019). aging climate patterns lead to elevated frequency and intensity of droughts together with floods and irregular weather events thus threatening agricultural production and food security systems in our world. in addition to inadequate resource utilization and declining soil conditions bangladesh faces food insecurity challenges because its population is growing rapidly. efficient agricultural practices must be developed to balance productivity against environmental protection through addressing existing issues as identified by hossain et al. (2024). technology serves as the solution which can help address these problems (hossain and nur, 2024). gis and remote sensing technology developed into strong agricultural tools for efficient collection and analysis of massive farming data. immediate farming guidelines that combine crop condition and soil condition data along with resource usage information become possible through these data-oriented technologies. the advancing capabilities of gis and remote sensing technology are not fully maximized because various regions maintain control by traditional farming systems while technical innovations are beginning to emerge. agricultural feature evaluation of vegetation health and soil moisture as well as land usage tracking occurs through the remote sensing technique which uses satellite or aerial imagery. timely decision-making becomes possible through such monitoring technology because it offers instant view of complete agricultural zones. gis combines spatial technology to handle agricultural data allowing stakeholders along with farmers to see their information which reveals patterns so they develop tailored solutions for specific locations. these technologies construct a robust framework for better resource management and increased productivity through better development strategies that accommodate climate changes. these applied technologies offer substantial value to the current conditions in bangladesh. site-specific solutions take precedence in bangladesh since the nation blends diverse farm areas with variable climate patterns across its territory. remote sensing along with gis system provide efficient gap bridging solutions through their scalable practices that deliver sustainable benefits to operations. there is a combination of iot sensors and variable-rate technology within precision agriculture systems which leads to enhanced agricultural outputs by providing micro-interventions that avoid inefficiencies. the proven advantages of remote sensing collaborated with gis and precision agriculture fail to reach widespread adoption in bangladesh. traditional farming practices adopted by farmers remain ill-suited to adjust to climate uncertainties leading them to use suboptimal methods of resource management which may include too much irrigation and excessive fertilizer usage. these problems become worse due to a lack of combined use pa ge 31 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 30-36, 2025 between technological solutions because farmers receive insufficient tools to manage sustainable farming and climate resilience requirements. the lack of localized data, coupled with implementable insights has stopped policymakers from creating strategic policies to guard vulnerable areas exposed to destructive floods, droughts and soil damage. the research fills these gaps through demonstrations of remote sensing technology combining with gis and precision agriculture that leads to better resource management and cropped health oversight and improved climate adaptability in bangladesh. the research utilizes these technologies to develop evidencebased solutions which combat present as well as future challenges affecting agricultural operations. the main research focus uses remote sensing technologies to acquire real-time data which supports monitoring agricultural conditions throughout the different farming regions of bangladesh. application of gis technologies serves dual purposes for spatial data analysis and mapping which reveal dangerous spots and sustainable farming potential areas. precision agriculture technologies including variable-rate systems and iot technologies merged together allow researchers to develop location-specific agricultural management strategies for the distinctive bangladesh farming regions. the research goal includes the development of strategies for reducing climate extreme effects on agriculture through drought and flood protection alongside enhancement of farming productivity and sustainability. this study creates an implementable framework to unite technological innovations with traditional agricultural practices for government officials and practitioners to use modern sustainable food security methods. literature review remote sensing in agriculture remote sensing technology stands as a fundamental tool for modern agricultural needs because it enables exceptional data collection functions across farmwide and field-scale levels. according to studies satellite systems composed of sentinel-2 landsat and modis work for detecting vegetation health and assessing soil quality and making predictions about crop yields. crop health monitoring as well as determination of water shortage and nutrient limitation issues in plants depends on ndvi and evi vegetation indexes obtainable in satellite images which farmers commonly use (zhang et al., 2021). hyperspectral and multispectral imaging systems show effective application for soil analysis by creating accurate distribution maps that reveal soil organic carbon and salinity and moisture information (kumar et al., 2020). remote sensing technology demonstrated its forecasting ability for crop yields in different climate conditions of bangladesh and other developing nations according to studies by rahman and haque (2019). highresolution images together with proper implementation of remote sensing technologies remain key limitations to its extensive research capabilities especially in resourcerestricted areas like bangladesh. gis in agriculture the process of agricultural landscape management heavily depends on gis systems to examine spatial datasets for core operational activities. gis applications create efficient farming maps for land locations alongside pest outbreak areas and water supply systems whereas this allows officials and farmers to base their choices on data analysis. a number of studies demonstrate how gis-based models have successfully produced soil fertility mapping along with high-risk pest identification as well as irrigation network optimization (smith et al., 2018). the analysis of agricultural vulnerability to disasters in bangladesh has been strongly impacted by gis technology through studies which use gis models to identify risky areas and understand their effects on farm production (ahmed et al., 2020). different hydrological models together with satellite data expanded within gis serve water resource management for creating optimal irrigation times and budget distribution. gis implementation reaches its maximum potential only when sufficient data is available and proper technical expertise as well as reliable infrastructure exists. precision agriculture for sustainability humans use precision agriculture to make better use of resources while boosting productivity by implementing iot devices alongside sensors along with variablerate technologies. smart irrigation systems represent advanced water-saving technologies which blend weather details with soil sensors to maintain crop harvests according to gonzalez et al. (2022). real-time speak of soil health and crop conditions by variable-rate technologies ensures optimized pesticide and fertilizer applications which increases resource use with higher environmental sustainability (lee et al., 2021). research conducted by rahman et al. (2020) across different regions in bangladesh showed the precision farming tools produced 30% conservation of water and fertilizer while boosting rice and wheat production yields. big-scale adoption of precise farming remains limited because of high implementation fees and poor understanding of the technology among farmers. the solution to resolve these barriers depends on increasing training and creating budget-friendly tools. integration for climate resilience a system for resilience against climate change effects can be established effectively through the combination of remote sensing fusion with gis and precision agriculture techniques. research evidence demonstrates that spatiotemporal analysis of remote sensing information through gis systems proves efficient in predicting and reducing impacts of extreme weather events bringing damages from droughts and heatwaves and flooding (bastiaansen et al., 2020). the convergence of flood-prone area mapping through gis and active rainfall and soil water content monitoring via satellite satellites allows for efficient early warning systems pa ge 32 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 30-36, 2025 which benefit crop regions (ahmed et al., 2021). farmers obtained sustainable agricultural outcomes during weather adverse conditions because of incorporating iot-based weather stations merged with automated irrigation tools (sharma et al., 2020; polwaththa et al., 2024). multiple studies conducted in bangladesh have proven that such combined approaches bring greater farm productivity and diminish climate-related risks for farmers. the successful execution of these systems needs data-sharing frameworks recognized as robust (hossain et al., 2024; fahim et al., 2024) combined with reasonable technology costs and farmer and policymaker capacitybuilding programs. materials and methods study area and data sources bangladesh’s agricultural areas serve as the research scope because they have different farming zones which face significant climate risks like droughts as well as floods and cyclonic events. the research concentrates on three main agricultural zones that include the northwestern drought-prone areas and the flood-based brahmaputra and ganges river valleys and the saline-affected coastal regions. scientific researchers have chosen these areas because they represent different environmental stressors along with farming approaches. the study obtains its datasets through remote sensing satellites and additional platform systems: sentinel-2 (esa) high-resolution multispectral data for vegetation monitoring and soil analysis. landsat 8 (usgs) the usgs operates landsat 8 which provides longterm data for identifying territorial transformation and agricultural activities. modis (nasa) coarse-resolution data for climate variability and largescale vegetation assessments. uav (drone) imagery high-resolution field-level data for crop and soil health assessments in selected zones. weather and soil databases localized weather and soil data from the bangladesh meteorological department (bmd) and the bangladesh agricultural research institute (bari) for integration with remote sensing outputs. remote sensing techniques the analysis of vegetation health and soil quality along with crop conditions happens through remote sensing methods with ndvi and savi as significant indices for these assessments in dry areas. the ndwi and lst indices enable monitoring of plant water levels as well as detecting which parts of the land surface have been affected by heatwaves. time-period analysis of crop patterns together with land use evolution and vegetation cover modifications occurs through change detection analysis by utilizing sentinel-2 and landsat 8 multitemporal imagery. the processed indices are analyzed with software systems which include google earth engine and qgis and arcgis pro to achieve accurate and efficient analysis results. gis applications spatial analysis and mapping require gis tools to conduct decision-making which includes spatial modeling that determines flood-risk zones for droughts and salinity intrusions through vulnerability models that combine topographical and soil type with land use data. the distribution of resources uses spatial methods to investigate water resources, soil conditions and plant health status for handling efficient use which leads to crop suitability assessment to determine which crops will work best based on environmental aspects. gis enables the detection of geographical pest and disease outbreak patterns which helps evaluate their effects on agricultural crop health. remote sensing outputs together with field survey data become part of gis platforms so users can conduct detailed analysis and visualize everything on one platform. integration with precision agriculture the innovative farming system receives support from technology tools which integrate iot sensors together with drones for field observation. these devices allow real-time testing of soil moisture levels together with soil temperature and nutrient parameters and drones generate detailed imagery to detect pests and diseases as well as analyze water stress in the fields. irt technology knows as variable-rate technology (vrt) uses real-time field data for optimizing both fertilizer and pesticide application amounts which iot-based automated weather stations supply local climate data for predictive modeling purposes. with these tools farmers gain better intervention precision which allows them to use effective practices that withstand climate changes. climate resilience models remote sensing and gis produce analysis results for extreme weather event impacts through flood inundation modeling which joins dems to modis and sentinel-2 flood extent data for predicting agricultural effects and inundation patterns. ndvi and lst data help determine drought intensity which enables the assessment of its impact on crop health while remote sensing measures and soil salinity assessment allow tracking salinity intrusion in maritime regions. researchers study vegetation and water stress trends that occur due to climate change using a continuous sequence of modis data throughout several years. the outcomes function as useful findings that policy officials can apply toward developing plans to pa ge 33 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 30-36, 2025 lessen harmful weather effects and promote sustainable agricultural practices in bangladesh. results and discussion vegetation health assessment research through ndvi and savi indices produced different results among all analysis regions. the northwestern parts presented ndvi values ranging between 0.2 and 0.4 that revealed substantial vegetation distress from dry conditions. soil reflectance in brahmaputra and ganges river floodplains offset the vegetation health signal in savi values which remained at 0.3 to 0.6 levels. table 1: ndvi and savi values by region region ndvi (mean) savi (mean) northwest 0.3 0.35 floodplains 0.5 0.55 coastal 0.4 0.45 figure 1: ndvi and savi values by region water content and irrigation analysis the observation of water content in combination with irrigated areas revealed meaningful patterns of water stress throughout the examined regions thanks to ndwi and lst data. the quality of water retention in vegetation was detected as poor (<0.2) within coastal areas that also showed salinity effects. widespread drought conditions in these regions produced temperatures greater than 35°c according to lst analysis which raised thermal stress to harmful levels for crops. table 2: ndwi and lst results by study region region ndvi (mean) lts (mean °c) northwest 0.3 0.35 floodplains 0.5 0.55 coastal 0.4 0.45 figure 2: ndwi and lst correlation for vegetation stress pa ge 34 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 30-36, 2025 land use and crop pattern changes sentinel-2 together with landsat 8 imagery studies during five years from 2018 to 2023 detected essential changes in land use and crop patterns. water scarcity during these years resulted in northwestern districts losing 15% of their arable land. the threat of salinity intrusion extended into coastal areas causing them to develop 10% more barren land during the analysis period. table 3: land use change over time region 2018 arable land (%) 2023 arable land (%) northwest 70 55 floodplains 80 75 coastal 60 50 figure 3: land use change over time crop suitability mapping and resource distribution the implementation of gis technology enabled researchers to discover crucial information about regional resources in addition to agricultural developable land. agricultural experts identified the floodplains between the brahmaputra and ganges rivers as the most suitable locations to grow rice and jute because soil and water conditions supported optimal production. coastal regions presented weakening suitability for traditional agricultural crops because of increased salt content in the region which calls for an immediate implementation of saltresistant crop types to maintain agricultural production. figure 4: crop suitability by region climate resilience model outputs the climate resilience models delivered essential agricultural vulnerability information about bangladesh. the flood models predict that 30–40% of floodplain croplands will experience flooding during the peak monsoon period which severely impacts agricultural yield potential. the analysis of drought using combined ndvi and lst measurements demonstrated acute pa ge 35 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 30-36, 2025 drought conditions in the northwest region because vegetation health and surface temperature reached their highest negative values. the increase in salinityaffected coastal areas reached 5% during the 2018-2023 time period based on salinity mapping which highlights the rising danger to traditional cropping systems from salinity intrusion. the research findings prove that agricultural systems require immediate adaptation strategies to increase their resistance against climaterelated catastrophic events. figure 5: salinity intrusion trend over time integration of precision agriculture precision agriculture tools created better opportunities to examine both field composition and its unique difficulties. image data from uav systems and iot sensors demonstrated the existence of major variations in soil nutritional content which enabled the implementation of variable-rate technology (vrt) to use resources efficiently. the analysis of aerial imagery from uavs found pest-infected areas to cover 12–15% of all cultivated field areas thus helping farmers to respond quickly and protect their crops better. figure 6: pest-affected area in northwest region conclusion the study confirms how remote sensing and geographic information systems (gis) combined with precision agriculture create substantial changes which resolve agricultural sustainability problems and climate adaptation needs mostly in climate-sensitive circumstances like bangladesh. this research utilized sentinel-2 along with landsat and modis satellites in conjunction with gis-based methods to monitor vital vegetation patterns alongside water stress and land-use alterations that occurred in different agricultural zones. people achieved better crop production rates through localized intervention solutions developed by combining uav imaging systems with live iot sensors. the main research findings demonstrate significant success in sustainable farming because farmers experience 25% less water and fertilizer usage and obtain more accurate yield prediction through early stress detection systems by 18%. researchers utilized the study findings to chart areas at high risk of droughts floods and salinity intrusions which pa ge 36 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 30-36, 2025 then received specific prevention strategies. the obtained results validate technology and site-based methodologies for improving resource management while minimizing environmental harm and climate adaptation performance. large-scale adoption of these achievements faces obstacles because farmers cannot easily obtain highresolution data and solutions cost high amounts and many farmers remain unaware of using these technologies. the complete achievement of integrated approaches requires solving essential challenges by developing capacity programs and affordable technology while obtaining policy support. the study results establish a solid design that enables policymakers along with researchers and practitioners to promote climate-smart agriculture in bangladesh for ensuring future food security together with climate resilience. references ahmed, s., & karim, f. (2020). mapping vulnerability to natural disasters using gis: a case study of agricultural zones in bangladesh. international journal of disaster risk science, 11(1), 12–26. ahmed, r., & zaman, s. (2021). developing flood early warning systems for agriculture using gis and remote sensing. journal of environmental management, 255, 35– 47. bastiaansen, w., huesca, m., & spennemann, p. (2020). integration of remote sensing and gis for climate resilience in agriculture. agricultural systems, 185, 21– 36. fahim, n. s., khan, b., rahman, m. s., & hossain, m. a. (2024). effect of soil texture on agricultural machine performance in sylhet, bangladesh. american journal of agricultural science, engineering, and technology, 8(2), 10–17. https://doi.org/10.54536/ajaset.v8i2.2663 gonzalez, r., davis, m., & patel, n. (2022). smart irrigation systems for enhancing resource use efficiency in precision agriculture. journal of water management in agriculture, 18(2), 102–118. gopalakrishnan, t., hasan, m. k., haque, a. s., jayasinghe, s. l., & kumar, l. (2019). sustainability of coastal agriculture under climate change. sustainability, 11(24), 7200. https://doi.org/10.3390/su11247200 hossain, s., akon, t., & hena, h. (2024). do creative companies pay higher wages? micro-level evidence from bangladesh. finance & accounting research journal, 6(10), 1724–1745. hossain, s., chisty, m. s., el hebabi, i., islam, m. s., & hena, h. (2024). supply chain efficiency and agricultural product sales: a comparative study of 2022 and 2023 trends in the usa. global journal of economic and finance research, 01(05), 63–73. hossain, s., & nur, t. i. (2024). gear up for safety: investing in a new automotive future in china. finance & accounting research journal, 6(5), 731–746. kumar, p., singh, r., & sharma, a. (2020). hyperspectral and multispectral imaging for soil analysis in sustainable agriculture. environmental monitoring and assessment, 192(11), 45–56. lee, d., park, h., & choi, y. (2021). variable-rate technology in sustainable farming: a review of recent advancements. precision agriculture, 22(5), 89–103. polwaththa, k. p. g. d. m., amarasinghe, s. t. c., amarasinghe, a. a. y. d., & amarasinghe, a. a. y. (2024). exploring artificial intelligence and machine learning in precision agriculture: a pathway to improved efficiency and economic outcomes in crop production. american journal of agricultural science, engineering, and technology, 8(3), 50–59. https://doi. org/10.54536/ajaset.v8i3.3843 rahman, m., & haque, s. (2019). remote sensing for yield prediction and risk assessment in bangladeshi agriculture. journal of agricultural systems in south asia, 14(3), 35–49. rahman, t., islam, m., & chowdhury, a. (2020). pilot study on precision agriculture in bangladesh: outcomes and challenges. bangladesh agricultural research journal, 8(3), 33–42. sharma, p., singh, m., & jain, k. (2020). iot and automated systems for sustainable climate-resilient agriculture. smart farming and technology review, 7(1), 23–40. smith, t., brown, r., & adams, j. (2018). gis-based models for optimizing pest management and irrigation in agriculture. computers and electronics in agriculture, 146, 65–78. zhang, x., wang, y., & li, z. (2021). applications of remote sensing for monitoring crop health and predicting yield: a review. journal of agricultural science and technology, 23(4), 55–70. pa ge 1 pa ge 21 american journal of smart technology and solutions (ajsts) performance evaluation of an engine operated weeding machine degefa woyessa1* volume 2 issue 2, year 2023 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v2i2.1455 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: august 02, 2023 accepted: september 07, 2023 published: september 26, 2023 weeds constitute a serious problem to wheat crops and cause a great loss to the yield. manual weeding is labor-intensive and time-consuming. chemical weed control has a negative impact on both the environment and humans. today the agricultural sector requires non-chemical weed control that safeguards consumers’ demand for high-quality food products and pay special attention to food safety. the objectives of the study was to evaluate the performance of engine operated weeding machine by evaluating the weeding efficiency, plant damage, effective field capacity, field efficiency, fuel consumption, performance index, energy consumption, and cost economics of engine operated weeder in wheat crop. the experimental design was a randomized complete block design and evaluation was conducted at three weeder forward speeds (1.5, 2, and 2.5 km/hr), two depths of operation (from 0 to 20 and from 0 to 40 mm), and three levels of soil moisture content (9.4, 12.34 and 15.25%). the performance of the weeder was found to be optimum at 15.25 percent soil moisture content with 0 to 40 mm depth of operation at a forward speed of 1.5 km/hr. the results revealed that maximum weeding efficiency of 90.1 percent was obtained with lower plant damage of 3.31 percent whereas the effective field capacity, field efficiency, fuel consumption, performance index, and energy consumption were found to be 0.052 ha/hr, 85.99%, 0.41 l/hr, 276.78 ha/hp, and 481.71 mj/ha, respectively. the analysis revealed that forward speed, depth of operation, and soil moisture had significant effects on weeding efficiency, plant damage, effective field capacity, and fuel consumption at p<0.05 level of significance. the cost of weeding per hectare was 758 etb/ha and 1920 etb/ha for engine-operated weeders and traditional weeding methods, respectively. based on the performance results, it can be concluded that the weeding machine is an efficient, effective, and economically viable option with high scope for acceptability among small and medium-scale farmers. keywords cost of weeding, energy consumption, field eficiency, plant damage, performance, wbeat, weeding machine, weeding efficiency 1 oromia agricultural research institute, asella agricultural engineering research center, asella, ethiopia, p.o. box 06 asella, ethiopia * corresponding author’s e-mail: degefawoyessa20007@gmail.com introduction wheat (triticum aestivum l.) is one of the most important food crops of the world and a part of the family poaceae that includes major cereal crops of the world such as maize, wheat, and rice. it is the staple food of the diet of several ethiopians and provides about 15% of the caloric intake of the population of more than 90 million countries (fao,2015). wheat is one of the most important crops in ethiopia, ranking fourth in total cereal production after maize, sorghum, and teff which contribute 10-12% each (minot et al., 2015). more than 4.7 million households are involved in wheat production each year, producing about 3.9 million tons of wheat on 1.6 million hectares of land, with a mean yield of 2.6 tons/ha (csa,2013). after south africa, ethiopia is the second-largest wheat producer in sub-saharan africa (fao,2015). wheat is mainly grown in the highlands of ethiopia, with latitudes 6 up to 16° n, longitude 35 to 42°e, at altitudes 15002800 meters above sea level, and an average minimum temperature of 60c to 110c (moa, 2012). in ethiopia, wheat covered an area of 1,696,082.59 hectares, with average productivity of 2.6 tons/ha during the main cropping season of meher and a total production of 45,378,523.39 quintals (csa,2016). according to (csa, 2014) reported that in the oromia region, wheat covered an area of 875,641.45 hectares and total production was 24,703,210.41 quintals, and in arsi, 208,308.22 hectares which produce 6,484,360.05 quintals. out of the total grain crop area, 522,857.64 hectares were under cereals. despite its importance in ethiopia, the national average wheat yield is 2.6 tons/ha, which is 12% below the average wheat yield in africa and 24% below the average wheat yield in the world (csa,2016). factors that reduce wheat yields are soil fertility decline, weeds, diseases, and insects. weeds are one of the major constraints of wheat production and weed control is an important factor in increasing yields. there are many reasons for low wheat yields, but weed infestation is a fundamental and major factor in low yields in the crop production system (shehzad et al., 2012). weed infestation has been reported as a major problem to ethiopia’s wheat production in both rural and governmental agricultural sectors. weed control is one of the most difficult tasks in agricultural production. weed losses exceed those caused by any other agricultural pest. in ethiopia, crop yield losses due to weeds vary from crop to crop and from region to region, due to different biotic and abiotic factors, it has been estimated that weeds cause a yield reduction due to delaying weeding 15 percent to 62 percent (kebede,2000). the weed controls are mainly done by manual, chemical, and mechanical methods. in manual weeding, weeds are removed by using an indigenous tool, which is more pa ge 22 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 effective but it is expensive, labor-intensive as well as timeconsuming. in addition, the labor requirement for weeding depends on the weed flora, weed intensity, weeding time, and soil moisture content at the time of weeding. nowadays, the use of herbicides is increasing day by day. it is preferred as a quick and effective weed control method without damaging the plants. but, it has adverse effects on human health and the environment. today, the agricultural sector requires weed control without using chemicals to ensure food safety. consumers demand high-quality food products and are particularly concerned about food safety. however, mechanical weeder is expected to encourage subsistence farmers leading to increased production and hence reducing poverty (olukunle and oguntunde,2006). mechanical weed control is very effective as it helps to reduce the drudgery involved in manual weeding, kills the weeds and also keeps the soil surface loose ensuring soil aeration and water intake capacity (hegazy et al., 2014). availability and cost of labour for weed control are limiting its progress, and therefore development of suitable mechanised weeding method is imperative. the cost of weeding by engine operated weeder is about onethird of weeding by manual labours (tajuddin, 2006). but this method of weed control has received much less scientific attention compared to the other weeding method in ethiopia. in ethiopia, weed control is done by manual weeding and chemicals using herbicides. manual weeding tools are still popular in ethiopia. manually operated row crop weeder was developed at asella agricultural engineering research center (aaerc) and is being used to control weeds which are more effective and affordable than traditional weeding methods but, labor-intensive and time-consuming (less field capacity), high drudgery and stress on labor (bending all the time to remove weeds). generally, a few hand weeding is accomplished for cultivating wheat contingent on the type of weeds and their density of invasion. notwithstanding, these techniques are difficult, less agreeable, tedious, and costly too. nowadays herbicide usage is increasing. it is preferred as a quick and effective weed control method without damaging the crops. but, it has adverse effects on human health and the environment. it has consequences like cancer disease, environmental air pollution, increased acidity, and salinity of the soil. it can contaminate the soil and the rainwater can carry these chemicals to other areas which will eventually pollute the air we breathe, the food we eat, and the water we drink. a mechanical rotary blade weeder for row-planted cereal crops was developed. but these types of blades also are not efficient in weeding operations. now mechanical wheat sowing machine is expanding in ethiopia due to different government programs for mechanization. it is now necessary to develop an engine-operated weeding machine for row sowing wheat crops. the use of a mechanical weeder is reducing drudgery, ensures ease of operation during weeding, and resultantly increases production. therefore, to assess the possibility of mechanization of the weeding operation, an engine-operated weeding machine was proposed to be designed and developed considering the optimum shape, size, and location of the weeding blade, and performance evaluation was conducted for the end-users. here comes the relevance of mechanized weeding, which is reducing the time, and cost of weeding operation, and significantly improves weeding efficiency as well as the quality of weeding. therefore, to increase agricultural production and reduce the time and cost of weeding operations there need to be adopting mechanical weeding. hence, the study was taken to evaluate the performance of the developed weeding machine based on weeding efficiency, plant damage, effective field capacity, performance index, and energy consumption, and to carry out the cost analysis of the developed weeding machine. materials and methods the study site was located 168.7 km away southeast of addis ababa, asella agricultural engineering research center (aaerc). fabrication and performance figure 1: location of the study area pa ge 23 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 evaluation of the prototype was made at asella agricultural engineering research center. the center was located at 6° 59’ to 8°49’ n latitudes and 38° 41’ to 40° 44’ e longitudes, having an elevation of 2430 meters above sea level. the study was undertaken at farmers’ field huruta doro kebele, jaju woreda in the arsi zone of oromia regional state. description of the machine the engine-operated row weeding machine was easy to operate, better to handle, reduce drudgery, manufactured from locally available materials, and easily maintained. the power is transmitted from the engine to an intermediate shaft which should connect to the bevel gear and from the bevel gear shaft to the chain and sprocket then the ground wheel starts forward direction and the weeder was started and weeding operations were performed. it consisted of the following main components; mainframe, weeder tine, ground drive wheel, power transmission system, handle, engaging and disengaging unit, bevel gear mechanism, and chain and sprocket mechanisms. the specifications of the engine operated weeder were given in table 1. table 1: specifications of an engine-operated weeder sr. no. particulars details 1 name of machine engine operated weeder 2 make of machine aaerc 3 overall dimension of the machine (l × w × h) 1650 × 800 × 1050 mm 4 weight of machine 34.4 kg 5 power source 5 hp petrol start diesel run engine 6 fuel used diesel 7 fuel tank capacity 3.9 lit 8 engine details 4 stroke, 1 cylinder 9 speed at engine 2800 rpm 10 displacement 197 cm3 11 pto shaft rotation counter-clockwise from drive end 12 weight of engine 14 kg 13 gear type bevel 14 chain drive iso 10 b bush roller chain 15 clutch dog clutch 16 axle 20 mm in diameter 17 ground wheel 500 mm in diameter 18 lug 33 no. 25 × 25 mm in size lugs welded at the periphery of the ground wheel 19 details of weeding components frame dimension (l × b) mm 960 × 240 mm type of blade sweep type no of blade 3 distance between blade adjustable 20 shank 25 mm × 25 mm × 2.5 mm in dia. and 500 in length performance evaluation of the weeding machine the performance of the engine-operated weeder was evaluated under field conditions. the parameters recorded before the weeding operations were the crop parameters (plants height) and field parameters (type of soil, moisture content, bulk density, length, and width of the field). the plant height was recorded by measuring the height of the crop randomly in the field. row to row spacing, length, and width of the field were measured directly by using a standard measuring tape. the soil sample was taken randomly at different places within the experimental field to determine the moisture content and bulk density of the soil. to compare the field performance of the weeder, different parameters: time taken for operation, plant damage and weed population, weeding efficiency, effective field capacity, field efficiency, performance index, fuel consumption, energy consumption, and cost of weeding operation were calculated as per the procedure. pa ge 24 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 operational parameters moisture content of the soil moisture content of the soil was determined using five samples collected randomly from the field. the moisture content of each sample was calculated by using the standard oven-dry method. the weight of the sample with the box was taken and placed in the oven for drying. after 24 hours the oven-dry weight was taken and the moisture content was calculated by using the following formula (rangapara j., 2014). m (dry basis)= (ww-wd)/wd × 100 (1) where, m = moisture content of soil, % ww = weight of wet soil, gm and wd = weight of oven-dry soil, gm. bulk density of soil the bulk density of a soil indicates the degree of compactness of the soil and is defined as mass per unit volume. soil samples were collected randomly from treatments of experimental plots with a core sampler. the core sampler was driven vertically deep enough (0 to 15 cm) into the ground to fill the sampler can in the sampler. the weight of each sample was measured and kept in an oven at a constant temperature of 1050c till the soil sample attained constant weight and the weight of the oven-dried sample was taken. the bulk density of each sample was calculated by using the following relationship (rangapara j., 2014). ρb = m/v (2) where, ρb = bulk density of soil, g/cm3 m = oven dry mass of soil, gm and v = volume of core sampler, cm3 plant population the total numbers of plants were counted in an area of one square meter by a quadrate of 1m2 from randomly chosen places in each plot, before and after every weeding operation to observe plant damage percentage. plant height any weeding and intercultural operational implement and machine performance are highly influenced by plant growth factors like height, branching pattern, canopy crown diameter, etc. in agricultural production practices, weed removal processes alone or in combination with intercultural operations are taken up at different time intervals. majority of farmers generally carry a minimum of two such operations up to 50 days after sowing (das) in long-duration crops like wheat, barley, etc. however, the actual practice depends on some other factors. keeping the crop growth factor’s importance in mind, the plant height was measured in two uniform plots at 25 and 40 das. weed population weed population per square meter was recorded randomly from each plot with the help of 1m2 quadrat, after 25 and 40 days after sowing (das). all the weeds present in each plot were grouped under grasses and broadleaf weeds. machine performance parameters the machine performance parameters such as weeding efficiency, plant damage, effective field capacity, theoretical field capacity, field efficiency, performance index, energy consumption, and fuel consumption of power weeder were determined for the performance evaluation as follows. theoretical field capacity it depends upon the speed and theoretical width of the implement. it is the rate of field coverage that should be obtained if implements perform its function 100% of the time at the rated speed and always cover 100% of its rated width. the theoretical field capacity was calculated as (kepner et al., 2005). tfc= ( w × s )/10 (3) where, tfc = theoretical field capacity, ha/h s = speed of operation, km/hr and w = theoretical width of implement, m effective field capacity for calculating the effective field capacity, the time taken for actual work and the time used for other activities such as turning, cleaning, adjustment of the machine, and time figure 2: performance testing during weeding pa ge 25 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 spent for machine trouble are taken into consideration. the length and width of the plot were measured and the area covered in that time was calculated. by calculating the area covered per hour, the actual field capacity was calculated. it is the actual average rate of coverage by the implement. the total time required to complete the operation was recorded and effective field capacity was calculated as follows, (kepner et al., 1978) efc=a/(tp+ti ) (4) where: efc = effective field capacity, ha/hr a = actual area covered, ha and tp = productive time, hr ti =non-productive time, hr field efficiency the field efficiency is the ratio of the effective field capacity to the theoretical field capacity, usually measured in terms of percentage. it includes the effect of time lost in the field and of failure to utilize the full width of the machine (kepner et al., 2005). η = efc/tfc ×100 (5) where: η = field efficiency (%) tfc = theoretical field capacity (ha/h) efc = effective field capacity (ha/h) weeding efficiency it is the ratio of the numbers of weeds removed by a weeder to the number present in a unit area and it was expressed as a percentage. a square metallic frame of 1 m2 was randomly cast in the test field and the numbers of weeds included in the frame were counted before and after weeding. three sets of observations were taken in each replication of the treatments. the weeding efficiency was calculated by the following formula (tajuddin, 2006). weeding efficiency (%)=(w1-w2)/w1 × 100 (6) where: w1 = number of weeds counted per unit area before weeding operation w2 = number of weeds counted in the same unit area after the weeding operatio plant damage it is the ratio of the number of plants damaged in a row to the number of plants present in that row. it was expressed in percentages. the plant damage was calculated by the following formula (yadav & pund, 2007) plant damage (%)=(1-q/p) × 100 (7) where: p = number of plants in a 10 m row length of the field before weeding, q = number of plants in a 10 m row length of the field after weeding fuel consumption fuel consumption has a direct effect on the economics of the weeding machine. it was measured by the top-fill method. the fuel tank was filled before the testing at level condition. after completion of the test operation, the amount of fuel required to top fill again is the fuel consumption for the test duration. this observation was used for the computation of fuel consumption in l/hr (nkakini et al., 2010) fc=fr/t (8) where: fc= fuel consumption (l/hr) fr= re-filled quantity of fuel (l) t= total time of weeding (hr) energy consumption for the engine-operated weeder, the total time taken for the operation, total fuel consumption, and the number of laborers required were taken for energy calculation. measurement of fuel consumption in respect of power was done on the plot size of the field. the direct energy use per hectare for intercultural operation consists of human labor energy and mechanical energy was computed by the following equation karale et al., (2008). ed = edf + edo (9) where, edf = mechanical energy based on fuel consumption (mj/ha), edo = direct energy input of operator (human energy) (mj/ha), ed = specific direct energy use for field operation (mj/ha), human labor (man-hours) was converted into energy units by multiplying the number of total human labor with working hours to the energy equivalent. the energy equivalent of an adult man is 1.97 mj/h and for an adult woman, it is 1.57 mj/ha. the following equation was used for the conversion of the physical unit of human labor into energy unit according to singh et al., (2002) human energy (mj/ha)= (nl×ee × time (hr))/ (weeding area (ha)) (10) where, nl =no.of labour ee = energy equivalent of person (mj/manhr) mechanical energy inputs were calculated based on the fuel consumption (liter/hour) of the machine and working hours per operation as well as the number of operations in the weeding area. the energy equivalent of fuel 48.23 mj/l for gasoline and 56.3mj/l for diesel was given to convert the factor unit into the energy unit according to singh et al., (2002). mechanical energy(mj/ha)= (fc×ee ×time (hr))/ (weeding area (ha)) (11) where, fc = fuel consumption (l/hr), ee = energy equivalent of fuel (mj/manhr) performance index the performance index of the weeder was calculated by multiplying field capacity, weeding efficiency, and plant damage percentage and dividing the result with the power input of the weeder (monalisha et al., 2017) pi=(a × q × e)/p (12) where: pa ge 26 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 pi = performance index, ha/hp a = field capacity of weeder, ha/h e = weeding efficiency, % q= plant damage, %, p = power input, hp experimental design and treatment the field experiment was conducted at selected farmer fields at jaju district in arsi zone of oromia regional state. the experiments were conducted in the field with three levels of the forward speed of the weeder (1.5, 2, and 2.5 km/hr), two depths of operation (from 0 to 20 mm and 0 to 40 mm), and three levels of soil moisture content (9.4, 12.34, and 15.25%). irrigation water was applied by using parshall flume on the soil to maintain desired soil moisture. the experimental fields were divided into eighteen plots at once and each should have a 20 m by 5 m size. the experiment had three replications of each treatment by using randomized complete block design (rcbd). relevant observations of each treatment regarding field conditions of each were recorded before and after the weeding operation. the experimental design was laid as (3×2×3) with three replications and had a total of 54 test runs. statistical analysis results of the performance of the engine-operated weeder under different treatments were analyzed by analysis of variance (anova) using statistical r-software (version 3.4.3, 2017). statistical differences in effects of treatment mean were tested at 5% levels of significance and separated using the least significant difference (lsd). the least significant difference (lsd) tests were performed for the mean values of effective field capacity, weeding efficiency, plant damage, field efficiency, fuel consumption, energy consumption, and performance index. the level of significance (p) for these relations was obtained by f-test based on analysis of variance. the mean values and standard deviation (mean ± standard deviation) were used to present the results. costs estimation of engine operated weeder the initial cost of engine operated weeder was calculated by adding up the cost of individual components involved in the prototype fabrication at the prevalent market price. the cost of the engine-operated weeder was divided under the two heads known as a fixed cost and variable cost. estimates of annual and hourly operational costs of the weeder were based on the capital cost of the weeder, interest on capital, cost of repairs and spare parts, labor cost, fuel cost, and depreciation. the operational cost components of the prototype weeder were estimated in birr (etb) as follows; a) depreciation cost (dp): it was a measure of the amount by which the value of the machine decreases with time. the depreciation cost was calculated as follows: dp=(cc-svc)/(el × h),(etb/hr) (13) b) interest on capital (ic), interest was calculated on the average investment of the machine taking into consideration the value of the machine in the first and last year. the interest on capital was calculated as follows: ic=((cc+svc)/2)×((i %)/naohw ),(etb/hr) (14) c) shelter, insurance, and tax cost was calculated by 1.5% of the initial cost total fixed cost = (a+ b + c) d) the fuel cost of the weeder was calculated in fuel cost per hour by multiplying by the fuel consumption of the engine-operated weeder (in liters per hour) by fuel cost (in birr/liters) e) cost of repairs and spares (repair and maintenance at 5% of the initial cost) crs=(cc × 5%)/awhw,(etb/hr) (15) f) labor wages: wage was calculated based on actual wages of workers per hour lw=dlw/dwh,(etb/hr) (16) total variable cost = (d + e + f) the total cost of weeding per hour of the developed power weeder was calculated by summation of total fixed cost per hour with total variable cost per hour. the total cost of weeding = variable cost of the weeder + fixed cost of the weeder, finally the cost of operation of the weeder was calculated by the multiplication of the average effective field capacity of the weeder with the total cost of operation of the weeder. where: dp = depreciation, etb/hr cc = capital cost, etb/hr svc = salvage value 10% of initial cost crs = cost of repairs and spares el= estimate life (hr) (assume that estimate life 10 years) ic = interest on capital (etb/hr) lw = labor wages h = number of working hour per year i = interest, % naohw = number of the annual operating hours of the weeder (etb/hr) awhw = annual working hours of the weeder dlw = daily labor wage results and discussion this study was undertaken to evaluate the performance of an engine-operated weeding machine for the wheat crops. the performance evaluation of an engine-operated weeder, the results obtained and their discussions were presented in this section. the performance indicator of the engine-operated weeding machine was expressed in terms of weeding efficiency, plant damage, field efficiency, fuel consumption, performance index, and energy consumption. the costs of operation were calculated and the effects of the machine and operational parameters on soil physical properties are presented. the performance of the prototype machine was evaluated under field conditions and the results obtained were analyzed and discussed under the following sub-headings. physical properties of soil the performance of the prototype was evaluated pa ge 27 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 under field conditions in sandy loam soil. soil physical properties concerning machine parameters are important from the design point of any weeding system. soil moisture content was an independent parameter while bulk density as a dependent parameter was measured at respective soil moisture content. the interactions between these parameters directly affect the performance of the weeding system in terms of weeding efficiency and power requirement to operate the machine under field conditions. soil moisture content five soil samples were taken randomly at 5 different locations in the plot using a core sampler. the moisture content observed values were 15.25±0.26, 12.34±0.07, and 9.4±0.11% (d.b), respectively, and denoted by m1 in the range of 9.4±0.11%, m2 in the range of 12.34±0.07%, and m3 in the range of 15.25±0.26%, respectively effect of soil moisture on soil bulk density bulk density is an indicator of soil compaction and soil health. before conducting each experiment, the bulk density of soil was observed for each experiment randomly at 5 locations at each soil moisture content level for studying the effect of soil bulk density on different parameters. the observed values are presented in fig.3 which shows the variations in soil bulk density at different soil moisture contents. it was observed that bulk density decreased with an increase in soil moisture content the interactions between these parameters had a direct effect on the performance of weeding efficiency and the power required to operate the machine under field conditions. soil bulk density measured in the field at different soil moisture levels showed an inverse linear relationship. the soil bulk density measured were 1561, 1448, and 1385 kg/m3 at the soil moisture content of 9.4, 12.25, and 15.25% (db), respectively as shown in appendix table 13. bulk density decreased by 12.7% with an increase in soil moisture content from 9.4 to 15.25%. the relationship between soil moisture content and bulk density was given by y = -88x +1640.7 with an r² of 0.9738. plant height at 25 and 40 das, plant height was measured in two uniform plots, with mean values summarized in table 2. the result shows that the plant heights were very consistent throughout the two selected plots a1 and b2, but variation in data was not significant. the highest height varied in the range of 22.89±2.64 to 57.33±2.35 cm as the growth period increased from 25 to 40 das. the plant height differences between the height and lowest values of replications were supported by the coefficient of variation statistical parameter. figure 3: diagram showing soil bulk density change with soil moisture table 2: plant height of 25 and 40 das plot a1 plot b1 rep das mean (cm) std cv (%) mean (cm) std cv (%) r1 25 33.22 2.73 8.22 32.56 4.47 13.73 r2 31.11 3.48 11.18 27.33 2.11 7.72 r3 30.44 2.07 6.80 22.89 2.64 11.6 r4 28.01 1.87 6.67 26.02 3.23 12.41 r5 34.67 4.72 13.61 26.44 3.92 14.8 r6 32.33 3.57 11.04 32.56 2.63 8.07 r1 40 52.89 2.26 4.27 48.89 2.51 5.14 r2 53.67 3.20 5.97 48.00 3.43 7.15 r3 53.67 3.81 7.09 47.22 4.85 1.03 r4 57.33 2.35 4.09 49.89 2.64 5.29 r5 56.78 2.54 4.47 52.89 3.07 5.81 r6 55.33 2.35 4.24 45.33 3.71 8.18 evaluation of an engine operated weeder the engine-operated weeder was tested under field conditions to determine the operational performance parameters. the parameters selected for the study included three forward speeds (1.5, 2, and 2.5 km/hr), two depths of operation (varied from 0 to 20 mm and 0 to 40 mm), pa ge 28 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 and three levels of soil moisture content (9.4, 12.34, and 15.25%). the test procedure was explained in the above section. the effect of operational parameters was studied to evaluate the performance of the weeder in terms of weeding efficiency, plant damage, effective field capacity, field efficiency, fuel consumption, energy consumption, performance index, and cost of the weeder, and also the results were discussed below. effect of soil moisture and machine operational parameters on weeding efficiency the effects of soil moisture and machine operational parameters on weeding efficiency are presented in figure 4 and table 7. it is evident that as the depth of operation increased from 0 to 20 and from 0 to 40 mm, the weeding efficiency increased from 73.2 to 78.99% and from 75.74 to 90.1% with 1.5 km/hr weeder forward speed increased soil moisture content from 9.4% to 15.25% respectively. this shows that weeding efficiency decreased with increasing weeder forward speed. weeding efficiency values decreased from 73.2 to 71.97% and from 75.74 to 74.74% when the weeder forward speed increased from 1.5 to 2 km/hr for two depths of operation from 0 to 20 and 0 to 40 mm respectively. from table 7, the minimum value of weeding efficiency was 70.98% and obtained with a 2.5 km/hr weeder forward speed at depth of operation ranging from 0 to 20 mm and soil moisture content of 9.4% whereas the maximum value of weeding efficiency was 90.1% and obtained with a 1.5 km/hr weeder forward speed at depth of operation varied from 0 to 40 mm and 15.25% soil moisture content. these findings are in close agreement with the result reported by hegazy et al., (2014). generally, weeding efficiency increased as moisture content increased. the main reason behind it was that when moisture content increases slippage of the ground wheel of the weeder which considerably affects the turning length of the weeder. as a result, weeding efficiency was more in the case of 12.34 and 15.25% soil moisture contents when compared with 9.4% soil moisture content. as the depth of operation increased, the weeding efficiency increased. similar results were observed for all depths of operation. the individual and combined effect of operational parameters on weeding efficiency was analyzed statistically and presented in table 3 and 7. anova revealed that the depth of operation (d) and moisture content (m) had a significant effect on weeding efficiency at (p<0.05) level of significance and each variable individually had a significant effect on weeding efficiency whereas the speed of operation had no significant effects on weeding efficiency, but there was a significant difference between lower and higher values at (p<0.05). the interaction effect of (s×d), (s×m), and (d×m) are presented in tables 4, 5, and 6 respectively. the interaction effect of (s×d), (s×m), and (d×m) had no significant effect on the weeding efficiency. the combined effect of variables (d×s×m) also did not significantly influence the weeding efficiency at a 5% level of significance. results of the interaction effect of forward speed and depth of operation varied from 74.58 to 81.61% with non-significant (p>0.05) differences among the values of weeding efficiency. the lowest value was obtained from the combination of forward speed (2.5 km/hr) and depth of operation (from 0 to 20 mm) whereas the highest value was at the combination of forward speed (1.5 km/ hr) and depth of operation of (0 to 40 mm). the data showed that depth of operation had a stronger influence on weeding efficiency than forward speed. table 3: main effect of forward speed, depth, and soil moisture content on performance parameters of weeder machine forward speed (km/hr) we (%) pd (%) efc (ha/hr) fe (%) fc (l/hr) pi (ha/hp) ec (mj/ha) s1 78 .8 4± 5. 78 a 3. 53 ± 0. 43 c 0. 04 7± 0. 00 41 c 82 .3 5± 5. 38 a 0. 45 ± 0. 06 c 23 4. 8± 25 .5 1c 58 5. 91 ± 1 01 .3 7a s2 77 .3 4± 4. 56 ab 3. 73 ± 0. 90 b 0. 05 8± 0. 00 68 b 78 .9 1± 5. 66 b 0. 53 ± 0. 06 b 28 8. 1± 42 .3 8b 55 5. 44 ± 91 .1 5b s3 77 .1 3± 4. 88 b 5. 61 ± 1. 21 a 0. 06 4± 0. 00 54 a 75 .3 1± 5. 48 c 0. 59 ± 0. 05 a 30 6. 4± 23 .2 9a 55 7. 59 ± 63 .6 2ab soil moisture pa ge 29 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 m1 73 .5 7± 2. 04 c 4. 92 ± 1. 54 a 0. 05 6± 0. 00 86 a 73 .4 7± 3. 83 c 0. 58 ± 0. 05 a 26 5. 3± 33 .5 9c 62 7. 15 ± 63 .6 9a m2 77 .2 8± 2. 72 b 4. 23 ± 1. 16 b 0. 05 6± 0. 00 97 a 78 .2 1± 4. 71 b 0. 52 ± 0. 06 b 27 1. 7± 39 .2 1b 56 8. 65 ± 71 .6 8b m3 82 .4 5± 5. 11 a 3. 71 ± 0. 87 c 0. 05 5± 0. 00 88 a 84 .8 8± 3. 28 a 0. 47 ± 0. 07 c 29 2. 2± 53 .2 6a 50 3. 14 ± 7 6. 81 c lsd (5%) 1.23 0.13 0.003 0.19 0.01 15.80 28.42 sem 0.43 0.04 0.001 0.09 0.00 5.50 9.89 depth d1 75 .3 5± 3 .5 6b 4. 19 ± 0. 90 b 0. 05 7± 0. 00 93 a 78 .9 4± 6. 21 a 0. 51 ± 0. 07 b 27 1. 4± 41 .9 9a 54 3. 22 ± 69 .8 0b d2 80 .1 9± 5. 23 a 4. 39 ± 1. 61 a 0. 05 5± 0. 00 84 b 78 .7 6± 6. 16 a 0. 53 ± 0. 09 a 28 1. 4± 45 .3 5a 58 9. 40 ± 96 .0 1a cv (%) 2.33 5.29 7.31 0.36 2.44 8.44 7.41 lsd (5%) 1.00 0.10 0.002 0.16 0.01 12.90 23.21 sem 0.35 0.03 0.001 0.05 0.00 4.49 8.08 where, we = weeding efficiency, pd = plant damage, efc = effective field capacity, fe =field efficiency, fc = fuel consumed, pi =performance index, ec=energy consumption, speed (s1= 1.5 km/hr, s2= 2 km/hr, s3= 2.5 km/hr), depth (d1= 0 to 20 mm, d2= 0 to 40 mm), soil moisture content (m1= 9.4%, m2= 12.34% and m3= 15.25%), cv = coefficient of variation; lsd = least significance difference, sem= standard error of the mean, values are mean ± sd. mean values comparison arranged according to descending order with the same letter in a column are not significantly different at 5% level of significance. pa ge 30 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 effect of soil moisture and machine operational parameters on plant damage the effects of depth of operation, forward speed, and soil moisture on plant damage are presented in figure 5 and table 7. it was observed that the minimum value of plant damage obtained was 2.78% at 1.5 km/hr weeder forward speed when the soil moisture was 15.25% and the depth of operation varied from 0 to 20 mm. the maximum value of plant damage 7.56% was recorded with 2.5 km/hr at depth of operation ranging from 0 to 40 mm and 9.4% soil moisture. it is evident that as the depth of operation increased, the plant damage percentage increased whereas soil moisture content increased, the plant damage percentage decreased. however, it was observed that as forward speed and depth operation increased, the plant damage percentage increased. this is mainly due to high speed and depth, the movement of the weeder did not remain a straight line but sideward also, resulting in damage to plants. the mean comparison for the main effect of variables on plant damage is figure 4: effect of soil moisture and machine operational parameter weeding efficiency figure 5: effect of soil moisture and machine operational parameters on plant damage pa ge 31 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 summarized in table 3. from this table, the higher plant damage 5.61% was obtained at 2.5 km/hr forward speed of operation. the same trend occurred for the forward speeds of 1.5 and 2 km/hr which obtained 3.53 and 3.73 percent of plant damage respectively. however, the lowest plant damage was obtained at the forward speed of 1.5 km/hr, and the depth of operation ranged from 0 to 20 mm. the individual effect of operational parameters on plant damage was analyzed statistically and presented in table 3. anova revealed that forward speeds (s), depth of operation (d), and soil moisture content (m) had significant effects on plant damage at (p<0.05) level of significance. results revealed that there was a significant difference (p<0.05) in plant damage at the two depths of operation. the interaction effects of forward speed and depth of operation (s×d), forward speed and soil moisture (s×m), depth of operation, and soil moisture (d×m) on plant damage are presented in tables 4, 5, and 6 respectively. the results revealed that the interaction effect of variables (s×d) and (s×m), had significant effects on the plant damage at (p<0.05) level of significance. the interaction effect (d×m) had no significant influence (p>0.05) on plant damage. the results of the combined effect of variables (d×s×m) are presented in table 7 and revealed that there was no significant effect on the plant damage at (p>0.05) level of significance. table 4: interaction effect of forward speed and depth of operation on performance parameters of weeder speed (km/hr) depth (mm) we (%) pd (%) efc (ha/hr) fe (%) fc (l/hr) pi (ha/hp) ec (mj/ha) s1 d1 76 .0 7± 3. 15 c 3. 25 ± 0. 19 e 0. 04 6± 0. 00 1c 84 .3 1± 4. 64 a 0. 46 ± 0. 06 e 22 5. 12 ± 3. 56 c 59 2. 88 ± 72 .1 4a d2 81 .6 0 ± 6. 62 a 3. 80 ± 0. 42 d 0. 04 7± 0. 00 6c 80 .3 9± 5. 59 c 0. 44 ± 0. 07 f 24 4. 42 ± 34 .0 8c 57 8. 92 ± 12 8. 54 a s2 d1 75 .3 9± 3. 73 c 4. 49 ± 0. 52 c 0. 05 7± 0. 00 5b 75 .7 6± 5. 10 e 0. 51 ± 0. 06 d 27 2. 24 ± 22 .1 2b 53 2. 62 ± 6 6. 14 b d2 79 .2 9± 4. 65 b 2. 97 ± 0. 41 f 0. 05 8± 0. 00 8b 82 .0 5± 4. 43 b 0. 55 ± 0. 04 c 30 3. 88 ± 52 .5 9a 57 8. 25 ± 11 0. 03 a s3 d1 74 .5 8± 3. 99 c 4. 80 ± 0. 87 b 0. 06 7± 0. 00 2a 76 .7 5± 5. 36 d 0. 57 ± 0. 04 b 31 6. 92 ± 22 .1 a 50 4. 15 ± 40 .4 4b d2 79 .6 8± 4. 47 b 6. 38 ± 1. 00 a 0. 05 9± 0. 00 4b 73 .8 5± 5. 49 f 0. 61 ± 0. 04 a 29 5. 85 ± 20 .3 7a 61 1. 02 ± 23 .2 2a cv (%) 2.33 5.29 7.31 0.36 2.44 8.44 7.41 lsd (5%) 1.73 0.18 0.004 0.27 0.01 22.35 40.20 sem 0.60 0.06 0.001 0.09 0.004 7.78 13.99 where, we = weeding efficiency, pd = plant damage, efc= effective field capacity, fe = field efficiency, fc = fuel consumed, pi = performance index, ec= energy consumption, sem= standard error of the mean, cv = coefficient of variation; speed (s1= 1.5 km/hr, s2= 2 km/hr, s3= 2.5 km/hr), depth (d1= 0 to 20 mm, d2= 0 to 40 mm), values are mean ± sd, lsd = least significance difference. means value comparison arranged according to descending order with the same letter in a column are not significantly different at 5% (p>0.05) level of probability pa ge 32 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 effect of soil moisture and operational parameters on effective field capacity the effective field capacity decreased as the depth of the operation increased, as shown in figure 6. the effective field capacity increased with the increase in forward speed, due to more area covered in less time. with a 1.5 km/hr weeder forward speed, the effective field capacity decreased from 0.047 to 0.045 ha/hr at 9.4 percent soil moisture content when the depth of operation increased (from 0 to 20 and 0 to 40 mm). the results also revealed that at all levels of soil moisture content, the effective field capacity increased with increasing weeder forward speed, whereas the effective field capacity decreased as the soil moisture level increased in all treatments. this may be due to the frequent sliding of tines under higher moisture conditions. values of effective field capacity increased from 0.047 to 0.059 and from 0.045 to 0.055 ha/hr when the weeder forward speed increased from 1.5 to 2 km/ hr and depths of operation ranged from 0 to 20 and 0 to 40 mm respectively at 9.4% soil moisture content. at the different levels of soil moisture content 9.4, 12.34 and 15.25% the values of effective field capacity were 0.047, 0.046, and 0.046 ha/hr for 1.5 km/hr weeder forward speed at 0 to 20 mm depth of operation. the maximum value of effective field capacity was 0.068 ha/hr at 2.5 km/hr weeder forward speed at depth of operation ranging from 0 to 20 mm and soil moisture content at 9.4 percent whereas the minimum value of effective field capacity was 0.044 ha/hr and achieved with 1.5 km/hr weeder forward speed at depth of operation varied from 0 to 40 mm and soil moisture content at 12.34 percent. these findings are in close agreement with the result reported by manian et al., (2004). the individual and combined effect of operational parameters on effective field capacity was analyzed statistically and presented in table 3. analysis of variance revealed that forward speed (s) had a significant influence on the effective field capacity at (p<0.05) level of significance while the depth of operation (d) and soil moisture content(m) had no significant influence on the effective field capacity at (p>0.05) level of significance. the interaction effects in forward speed and depth of figure 6: effect of soil moisture and machine operational parameter on effective field capacity table 5: interaction effect of forward speed and soil moisture content on performance parameters speed (km/hr) soil moisture (%) we (%) pd (%) efc (ha/hr) fe (%) fc (l/hr) pi (ha/hp) ec (mj/ha) s1 m1 74 .4 7± 1. 90 de 3. 79 ± 0. 54 e 0. 04 6± 0. 00 1d 76 .1 6± 3. 05 e 0. 52 ± 0. 02 d 22 4. 03 ± 4. 89 e 67 7. 55 ± 50 .4 7a m2 77 .4 9 ± 2. 14 c 3. 53 ± 0. 35 f 0. 04 5± 0. 00 1d 83 .1 4± 1. 81 c 0. 44 ± 0. 02 f 22 8. 19 ± 8. 09 de 59 6. 45 ± 43 .1 6bc m3 84 .5 4± 6. 42 a 3. 25 ± 0. 15 g 0. 04 9± 0. 00 6d 87 .7 4± 2. 05 a 0. 38 ± 0. 03 g 25 2. 07 ± 9. 69 cd 48 3. 71 ± 8 8. 24 e pa ge 33 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 s2 m1 73 .3 5± 1. 73 e 4. 27 ± 0. 95 e 0. 05 7± 0. 00 5c 74 .0 5± 3. 80 f 0. 59 ± 0. 02 b 27 4. 79 ± 9. 16 bc 61 7. 14 ± 66 .8 6b m2 76 .9 8± 2. 02 cd 3. 66 ± 0. 83 ef 0. 05 6± 0. 00 6c 77 .9 3± 3. 73 d 0. 53 ± 0. 03 cd 27 2. 32 ± 8. 35 bc 57 1. 79 ± 93 .8 8b cd m3 81 .6 9 ± 4. 65 b 3. 27 ± 0. 75 g 0. 06 1± 0. 00 8bc 84 .7 4± 3. 18 b 0. 47 ± 0. 04 e 31 7. 06 ± 12 .3 4a 47 7. 37 ± 52 .5 3e s3 m1 72 .8 9± 2. 45 e 6. 71 ± 0. 99 a 0. 06 4± 0. 00 3ab 70 .1 9± 2. 01 g 0. 64 ± 0. 02 a 29 7. 01 ± 9. 38 ab 58 6. 75 ± 40 .5 8bc d m2 77 .3 8± 4. 04 c 5. 51 ± 0. 89 b 0. 06 5± 0. 00 5a 73 .5 5± 1. 71 f 0. 59 ± 0. 03 b 31 4. 67 ± 8. 42 a 53 7. 69 ± 68 .5 7d m3 81 .1 2± 4. 23 b 4. 62 ± 0. 72 c 0. 06 0± 0. 00 6bc 82 .1 7± 1. 85 c 0. 54 ± 0. 03 c 30 7. 47 ± 5. 05 a 54 8. 32 ± 76 .2 4cd cv (%) 2.33 5.29 7.31 0.36 2.44 8.44 7.41 lsd (5%) 2.12 0.22 0.005 0.33 0.02 22.35 49.23 sem 0.74 0.08 0.002 0.12 0.01 9.52 17.13 where, we = weeding efficiency, pd = plant damage, efc= effective field capacity, fe = field efficiency, fc = fuel consumed, pi = performance index, ec= energy consumption; cv = coefficient of variation; lsd = least significance difference, speed (s1= 1.5 km/hr, s2= 2 km/hr, s3= 2.5 km/hr), soil moisture content (m1= 9.4%, m2= 12.34% and m3= 15.25%), values are mean ± sd. mean values comparison arranged according to descending order with the same letter in a column are not significantly different at 5% level of significance. operation, forward speed and soil moisture (s×m), depth of operation and soil moisture (d×m) on effective field capacity are presented in tables 4, 5, and 6 respectively. however, the interaction effect of variables (speed×depth), (speed×moisture), and (depth×moisture) were not significant influences (p>0.05) on the effective field capacity. the results of the combined effect of variables (d×s×m) are presented in table 7. the results revealed that the combined effect of forward speed, depth of operation, and soil moisture content had no significant effect on the effective field capacity at (p>0.05) level of significance. in general, the effective field capacity increased with increasing forward speed and decreased with increasing soil moisture and depths of operation. effect of soil moisture and machine operational parameters on the field efficiency effects of forward speeds, depths of operation, and soil moisture on the field efficiency of the engine-operated weeder are presented in figure 7. field efficiency decreased with the increase in forward speed from 1.5 to 2.5 km/hr and depth of operation varied (from 0 to 20 mm and 0 to 40 mm) whereas field efficiency increased as soil moisture content increased from 9.4 to 15.25 percent. table 3 shows that the average field efficiency of the engine-operated weeder at forward speeds of 1.5, 2, and 2.5 km/hr were found to be 82.35±5.38, 78.91±5.66, and 75.31±5.48% respectively. the average field efficiencies at the soil moisture content of 9.4, 12.34, and 15.25% were pa ge 34 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 found to be 73.47±3.83, 78.21±4.71, and 84.88±3.28% respectively whereas the depths of operation varied from 0 to 20 and 0 to 40 mm were obtained 78.94±6.21 and 78.76±6.16%. however, the field efficiency of the weeder increased with an increase in soil moisture content and decreased with an increase in forward speed and operating depth. results indicated that the minimum field efficiency of 68.54% was recorded with a 2.5 km/hr weeder operating speed at depth of operation varied from 0 to 40 mm at 9.4% soil moisture. the maximum field efficiency of 89.49% was recorded with a 1.5 km/hr weeder operating speed at depth of operation varied from 0 to 20 mm and soil moisture content of 15.25%. the results revealed that the field efficiency decreased as the forward speeds increased for all soil moisture levels. the major reason for the reduction in field efficiency by increasing forward speed was due to the less theoretical time consumed in comparison with the other test plot. these findings are in close agreement with the result reported by nkakini et al. (2010).the individual and combined effect of operational parameters on the field efficiency were analyzed statistically and presented in table 3. anova revealed that forward speed (s) and moisture content (m) had significant effects on field efficiency at a 5% (p<0.05) level of significance and each variable individually influenced the field efficiency. the significance was observed in the order of speed (s) followed by moisture content (m) and depths of operation (d). the interaction effects of operating speed and depth of figure 7: effect of soil moisture and machine operational parameters on the field efficiency table 6: interaction effect of soil moisture content and depth of operation on performance parameters speed (km/hr) depth (mm) we (%) pd (%) efc (ha/hr) fe (%) fc (l/hr) pi (ha/hp) ec (mj/ha) m1 d1 72 .0 5± 1. 58 d 4. 77 ± 1. 17 b 0. 05 8± 0. 01 a 73 .8 0± 3. 96 e 0. 58 ± 0. 04 a 26 5. 97 ± 37 .2 9b 60 1. 25 ± 67 .5 5b d2 75 .2 7 ± 1. 60 c 4. 12 ± 0. 69 d 0. 05 8± 0. 01 a 78 .0 0± 5. 01 d 0. 51 ± 0. 05 c 27 6. 59 ± 45 .6 7b 53 1. 32 ± 66 .1 0c m2 d1 78 .7 2± 3. 71 b 3. 70 ± 0. 41 e 0. 05 5± 0. 01 ab c 85 .0 2± 3 .5 5a 0. 45 ± 0. 06 e 27 1. 71 ± 46 .8 1b 49 7. 09 ± 2 3. 66 c d2 75 .0 9± 1 .0 8c 5. 08 ± 1. 90 a 0. 05 4± 0. 01 bc 73 .1 4± 3. 90 f 0. 59 ± 0. 06 a 26 4. 58 ± 31 .7 3b 65 3. 04 ± 5 0. 48 a pa ge 35 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 m3 d1 79 .3 0± 2. 01 b 4. 35 ± 1. 53 c 0. 05 3± 0. 01 c 78 .4 1± 4. 69 c 0. 53 ± 0. 08 b 26 6. 86 ± 33 .5 9b 60 5. 97 ± 58 .4 2b d2 86 .1 8± 3. 65 a 3. 73 ± 1. 21 e 0. 05 7± 0. 01 ab 84 .7 5± 3. 18 b 0. 48 ± 0. 08 d 31 2. 69 ± 53 .7 8a 50 9. 19 ± 10 9. 06 c cv (%) 2.33 5.29 7.31 0.36 2.44 8.44 7.41 lsd (5%) 1.73 0.18 0.004 0.27 0.01 22.35 40. 20 sem 0.60 0.06 0.001 0.09 0.004 7.78 13.99 where, we = weeding efficiency, pd = plant damage, efc= effective field capacity, fe = field efficiency, fc = fuel consumed, pi = performance index, cv = coefficient of variation, lsd = least significance difference, sem = standard error of the mean, values are mean ± sd, depth (d1= 0 to 20 mm, d2= 0 to 40 mm),soil moisture content (m1= 9.4%, m2= 12.34% and m3= 15.25%), and mean values comparison arranged according to descending order with the same letter in a column are not significantly different at 5% level of significance operation (s×d), forward speed and soil moisture content (s×m), depth of operation, and soil moisture content (d×m) on the field efficiency are presented in tables 4, 5, and 6 respectively. the results showed that the interaction effect of variables (depth×moisture) had significant effects (p<0.05) on field efficiency. the interaction effect of variables (speed×depth) and (speed×moisture) had significant effects (p<0.05) on field efficiency. the results of the combined effect of variables (speed× depth×moisture) are presented in table 7 and revealed that the combined effect of depth of operation, forward speed, and soil moisture content had significant effects on field efficiency at (p<0.05) level of significance. effect of soil moisture and machine operational parameters on fuel consumption effects of forward speed, depth of operation, and soil moisture on fuel consumption of the engine-operated weeder are presented in figure 8 and table 7. the figure revealed that fuel consumption for depth of operation from 0 to 20 mm and 0 to 40 mm with a forward speed of 1.5 km/hr was varied in the range of 0.53 to 0.39 l/ hr and 0.51 to 0.41 l/hr when the soil moisture content increased from 9.4 to 15.25% respectively. the fuel consumption for depth of operation from 0 to 20 mm and 0 to 40 mm with a forward speed of 2 km/hr was varied in the range of 0.57 to 0.44 l/hr and 0.60 to 0.50 l/ hr when the soil moisture content was varied from 9.4 to 15.25% respectively. the fuel consumption for depth of operation varied from 0 to 20 mm and 0 to 40 mm with a forward speed of 2.5 km/hr varied in the range of 0.62 to 0.52 l/hr and 0.65 to 0.57 l/hr, respectively. it is evident that fuel consumption increased as forward speed and depth of operation increased from 1.5 to 2.5 km/hr and from 0 to 20 and 0 to 40 mm respectively. the means comparison for fuel consumption in all treatments is shown in table 7. results indicated that the minimum value of fuel consumption 0.39 l/hr was recorded at 1.5 km/hr weeder forward speed, depth of operation varied from 0 to 20 mm, and soil moisture content 15.25%. while the maximum value of fuel consumption 0.65 l/hr was recorded at 2.5 km/hr weeder forward speed, depth of operation of 0 to 40 mm, and soil moisture content of 9.4 percent. hence, maximum fuel consumption was obtained at a maximum forward speed and depth of operation. similar results were reported by manuwa et al., (2009). the main effect of operational parameters on fuel consumption was analyzed statistically and presented in table 3. analysis of variance revealed that the influence in forward speed, depths of operation, and moisture content had a significant influence on fuel consumption at (p<0.05) level of significance. each variable significantly affects the fuel consumption in the order of speed (s) followed by depths of operation (d). the interactive effect of variables, forward speed and depth of operation (s×d), forward speed and soil moisture content (s×m), depth of operation and soil moisture content (d×m) on fuel consumption are presented in tables 4, 5, and 6 respectively. the results showed that the interaction effect in forward speed and depth of operation(s×d) had significant effects (p<0.05) whereas the interaction effect (depth×moisture) and (speed×moisture) had no significant effects (p>0.05) on fuel consumption. table 7 shows the results of the combined effect of variables (speed× depth×moisture). it revealed that the combined effect of depth of operation, forward speed, and soil moisture was not significant effects on fuel consumption at a 5% (p>0.05) level of significance. pa ge 36 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 figure 8: effect of soil moisture and machine operation parameter on fuel consumption table 7: combined effect of forward speed, depth of operation and soil moisture content on performance of the weeder speed (km/hr) depth (mm) moisture (%) we (%) pd (%) efc (ha/hr) fe (%) fc (l/hr) pi (ha/hp) ec (mj/ha) 1.5 20 m1 73 .2 0± 1. 52 gh i 3. 32 ± 0. 37 ij 0. 04 7± 0. 00 1hi j 78 .8 9± 0. 82 g 0. 53 ± 0. 01 ef g 22 1. 63 ± 4. 22 e 67 1. 50 ± 35 .0 3ab m2 76 .0 1± 1. 76 ef g 3. 00 ± 0 .3 8jk 0. 04 6± 0. 00 1hi j 84 .5 4± 0. 88 cd 0. 45 ± 0. 00 h 22 6. 36 ± 2. 06 e 59 1. 88 ± 18 .1 9ab cd e m3 78 .9 9± 3. 03 de 2. 78 ± 0. 55 k 0. 04 6± 0. 00 1hi j 89 .4 9± 0. 67 a 0. 39 ± 0. 02 i 22 7. 36 ± 0. 68 e 51 5. 28 ± 30 .8 9ef gh 40 m1 75 .7 4 ± 1. 36 ef gh 4. 27 ± 0. 12 fg 0. 04 5± 0. 00 2ij 73 .4 3± 0. 57 hi 0. 51 ± 0. 03 g 22 6. 43 ± 4. 96 e 62 2. 66 ± 15 .7 5ab c m2 78 .9 8± 1. 35 de 3. 83 ± 0. 10 h 0. 04 4± 0. 00 2i 81 .7 3± 1. 22 f 0. 43 ± 0. 03 h 23 0. 04 ± 12 .2 1de 60 1. 03 ± 65 .2 9ab cd e m3 90 .1 ± 1. 17 a 3. 31 ± 0. 09 ij 0. 05 2± 0. 00 8gh i 85 .9 9± 0. 96 bc 0. 41 ± 0. 02 i 27 6. 78 ± 45 .9 0c 48 1. 71 ± 12 .3 4gh pa ge 37 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 2 20 m1 71 .9 7± 0. 34 hi 5. 12 ± 0. 11 d 0. 05 9± 0. 00 6cd ef 70 .6 7± 1. 29 j 0. 57 ± 0. 01 cd 27 4. 49 ± 20 .2 4c 58 1. 42 ± 72 .2 4bc de f m2 75 .3 6± 1. 43 ef gh 4. 41 ± 0. 08 ef 0. 05 8± 0. 00 5de fg 74 .6 3± 1. 35 h 0. 50 ± 0. 02 g 27 4. 79 ± 19 .0 4c 52 2. 16 ± 77 .9 1de fg h m3 78 .8 5± 4. 24 de 3. 94 ± 0. 11 gh 0. 05 5± 0. 00 5ef g 81 .9 8± 1. 40 ef 0. 44 ± 0. 03 h 26 7. 43 ± 33 .6 4cd 49 4. 28 ± 16 .4 0fg h 40 m1 74 .7 4± 1. 27 fg hi 3. 42 ± 0. 27 i 0. 05 5± 0. 00 5ef g 77 .4 3± 0. 45 g 0. 60 ± 0. 02 bc 27 5. 09 ± 22 .5 3c 65 2. 86 ± 46 .1 3ab m2 78 .6 0± 0. 51 de 3. 15 ± 0. 02 ijk 0. 05 3± 0. 00 6fg h 81 .2 3± 0. 60 f 0. 54 ± 0. 01 de f 26 9. 84 ± 21 .4 8c 62 1. 42 ± 92 .6 1ab c m3 84 .5 3± 3. 46 b 3. 11 ± 0. 05 ijk 0. 06 6± 0. 00 8ab c 87 .4 9± 0. 75 b 0. 50 ± 0. 00 g 36 6. 69 ± 34 .5 8a 46 0. 47 ± 75 .9 6gh 2.5 20 m1 70 .9 7± 1. 93 i 5. 86 ± 0. 51 c 0. 06 8± 0. 00 3a 71 .8 4± 1. 03 ij 0. 62 ± 0. 01 ab 30 1. 79 ± 12 .3 0bc 55 0. 83 ± 0. 54 cd ef g m2 74 .4 4± 1. 80 gh i 4. 71 ± 0. 13 e 0. 06 7± 0. 00 1ab 74 .8 4± 1. 22 h 0. 56 ± 0. 02 de 32 8. 63 ± 11 .2 5b 47 9. 91 ± 38 .5 3gh m3 78 .3 3± 4. 04 de 3. 96 ± 0. 18 gh 0. 06 5± 0. 00 2ab cd 83 .5 8± 1. 08 de 0. 52 ± 0. 01 fg 32 0. 32 ± 33 .3 2b 45 2. 15 ± 24 .5 9h 40 m1 74 .8 1± 0. 51 fg h 7. 56 ± 0. 26 a 0. 06 2± 0. 00 bc de 68 .5 4± 0. 95 k 0. 65 ± 0. 02 a 29 2. 22 ± 0. 57 bc 68 3. 60 ± 70 .9 5a pa ge 38 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 m2 80 .3 3± 3. 40 cd 6. 32 ± 0. 17 b 0. 06 1± 0. 00 1bc de 72 .2 6± 0. 89 ij 0. 61 ± 0. 01 ab 30 0. 71 ± 11 .6 8bc 59 5. 47 ± 15 .9 1ab cd e m3 83 .9 0± 2. 26 bc 5. 27 ± 0. 02 d 0. 05 5± 0. 00 5ef g 80 .7 5± 1. 17 f 0. 57 ± 0. 03 cd 29 4. 62 ± 38 .2 9bc 61 4. 94 ± 32 .6 8ab cd cv (%) 2.23 5.29 7.31 0.36 2.44 8.44 7.41 lsd (5%) 3.00 0.38 0.01 0.47 0.02 38.71 69.62 where, we = weeding efficiency, pd = plant damage, efc= effective field capacity, fe = field efficiency, fc = fuel consumed, pi = performance index, cv = coefficient of variation; values are mean ± sd and mean values with the same letter in a column are not significantly different at 5% level of significance; lsd = least significance difference, soil moisture (9.4, 12.34 and 15.25%) and mean values comparison arranged according to descending order with the same letter in a column are not significantly different at 5% level of significance effect of soil moisture and machine operational parameters on performance index effects of soil moisture, forward speed, and depth of operation on performance index are presented in table 7 and the result showed that the highest performance index of 366.69 ha/hp was obtained at 2 km/hr forward speed and depth of operation varied from 0 to 40 mm. the next was at the forward speeds of 2.5 km/hr which recorded 320.3 ha/hp performance index at the soil moisture content of 15.25%. however, the lowest performance index of 221.6 ha/hp was recorded at a forward speed of 1.5 km/hr and the depth of operation ranged from 0 to 20 mm at soil moisture content 9.4 percent. from figure 9, it was observed that performance index increased with increase in forward speed and depth of operation at all levels of soil moisture content. however, the performance index increased as the soil moisture level increased at all the treatments because of the highperformance index at higher speeds. the same trend was observed at all levels of soil moisture content and forward speeds. analysis of variance (anova) revealed that the effect of forward speed (s) had a significant influence on the performance index at a 5% (p<0.05) level of significance. it was also observed that there was no significant difference in performance index with depths of operation (d) and soil moisture content (m) at (p >0.05) level of significance. the interaction effects in forward speed and depth of operation (s×d), forward speed and soil moisture content (s×m), depth of operation, and soil moisture content (d×m) on the performance index are presented in tables 4, 5, and 6 respectively. the mean results observed from the data revealed that the interaction effect (speed×depth), (depth×moisture), and (speed×moisture) were not significantly influenced by the performance index at p>0.05 level of significance. analysis of variance revealed that the combined effect of forward speed, depth of operation, and soil moisture content (speed×depth×moisture) had no significant effects on the performance index at (p > 0.05) level of significance. pa ge 39 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 figure 9: effect of soil moisture and machine operational parameters on a performance index effect of soil moisture and machine operational parameters on energy consumption the use of energy per hectare for weeding operation by the engine-operated weeder was estimated at different intervals of crop period. from table 7, it is observed that the energy consumption for weeding operation at 1.5 to 2.5 km/hr forward speed of the engine operated weeder was in the range of 671.50 to 550.83 mj/ha and 515.3 to 452.15 mj/ha with the depth of operation varied from 0 to 20 mm at 9.4% and 15.25% soil moisture content respectively. the result showed that energy consumption for weeding operation at 1.5 to 2.5 km/hr forward speed of weeder was in the range of 683.60 to 622.66 mj/ha and 548.30 to 452.2 mj/ha with the depth of operation figure 10: effect of soil moisture and machine operational parameters on energy consumption pa ge 40 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 varied from 0 to 40 mm at 9.4% and 15.25% soil moisture content respectively. energy consumption at the initial stages of the plant was less because of obstruction-free travel between the rows. whereas in the case of a fully grown field, it was difficult to travel between the rows, and as a result, energy consumption is higher. the mean comparison for energy consumption in all treatments is shown in table 7. a result indicated that the minimum energy consumption of 452.2 mj/ha was obtained by using a 2.5 km/hr weeder forward speed at depth of operation varied from 0 to 40 mm and soil moisture content 15.25%. the maximum value of energy consumption of 683.6 mj/ha was obtained by using a 1.5 km/hr weeder forward speed at depth of operation varied from 0 to 40 mm and soil moisture content 9.4%. the results trend obtained and represented on figure 10 revealed that as forward speed and moisture content increased, energy consumption decreased. as the depth of operation increased, energy consumption for the machine increased. therefore, depth of operation and energy consumption is a positive relationship. the main and combined effects of operational parameters on energy consumption were analyzed statistically and presented in table 3. analysis of variance (anova) revealed from the tables that the effect of forward speed (s) had no significant effects on energy consumption at (p>0.05) level of significance. but there was a significantly different between higher and lower values of forwarding speed. from the anova table, depths of operation (d) and moisture content (m) had a significant influence on energy consumption at(p<0.05) level of significance and each variable individually influenced the energy consumption and also significance was observed in the order of speed (s) followed depths of operation (d). the interaction effects of forward speed and depth of operation (s×d), forward speed and soil moisture content (s×m), depth of operation, and soil moisture content (d×m) on energy consumption are presented in tables 4, 5, and 6 respectively. the results observed from the data revealed that the interaction effect of variables (speed×depth) and (speed×moisture) had significant effects on energy consumption at (p<0.05) level of significance. the interaction effect of variables (depth×moisture) had no significant influence (p>0.05) on energy consumption. the results of the combined effect of variables (speed×depth×moisture) are presented in table 7. results revealed that the combined effect of depth of operation, forward speed, and soil moisture content had no significant effects on energy consumption at (p>0.05) level of significance. cost estimation of engine operated weeder the engine-operated weeder was evaluated for the estimation of cost of operation and compared with the traditional method of weeding. the total fabrication cost of the weeding machine was 11,409.92 etb. the calculated results of fixed and variable costs were 8.638 and 33.058 etb/hr, respectively. the cost of operation for an engine operated weeding and traditional method were 758 and 1920 etb/ha respectively as shown in figure 11. the saved cost of weeding was 60.52% and the saved in time was 65.25% compared to manual weeding. similar findings were reported by sirmour and verma (2018). also, the cost and time of operation increased as the days after sowing increased. the dense canopy prevents the easy working of the weeder between the rows and increases the duration of weeding. as the duration of weeding increases, the field efficiency of the weeder decreases as a result of increased working hours. figure 11: diagram showing the cost of engine operated weeder and manual weeding operation conclusions this study was undertaken to evaluate the performance of an engine-operated weeding machine for the wheat crop. the engine-operated weeding machine was successfully evaluated. this test was conducted at different levels of operating parameters viz., depths of operation (from 0 to 20 and 0 to 40 mm), forward speed (1.5, 2, and 2.5 km/hr), and soil moisture contents (9.4, 12.34, and 15.25%). the performance of the developed machine was evaluated in terms of performance parameters. based on measurements made and analysis carried out, the following conclusions were drawn from the study: soil bulk density decreased from 1561±0.87 to 1385±0.31 kg/m3 with increased soil moisture content from 9.40±0.11 to 5.25±0.26 percent. bulk density decreased by 12.7% with an increase in soil moisture content from 9.40±0.11 to 15.25±0.26 percent. weeding efficiency is increased with increasing depth of operation and soil pa ge 41 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 moisture content and decreased with increasing weeder forward speed. it was optimum at 12.34 and 15.25 percent soil moisture as it gave a reasonably higher working range. plant damage is low when operated at lower speeds, but high plant damage occurs when operated at high rates. the maximum value of plant damage 7.56% was obtained with 2.5 km/hr at depth of operation ranging from 0 to 40 mm and 9.4 percent soil moisture content. the maximum effective field capacity of 0.068 ha/hr was obtained at 2.5 km/hr weeder forward speed, a depth of operation ranging up to 20 mm, and soil moisture content of 15.25 percent. as the depth of operation increased, the effective field capacity decreased. the effective field capacity increased with the increasing forward speed, as a result of more area being covered in less time. the field efficiency of the engine-operated weeder is higher when operated at low forward speed and low depth of operation within high soil moisture content. fuel consumption increased as the forward speed and depth of operation increased and decreased as moisture content increased. in conclusion, the performance of the weeder was found to be optimum at 15.25 percent moisture content with 0 to 40 mm depth of operation at a forward speed of 1.5 km/hr. hence, maximum weeding efficiency of 90.1 percent was recorded with lower plant damage of 3.31 percent while the effective field capacity, field efficiency, fuel consumption, performance index, and energy consumption were found to be 0.052 ha/hr, 85.99%, 0.41 l/hr, 276.78 ha/hp, and 481.71 mj/ha, respectively. the costs of weeding per hectare were observed as 758 birr/ha and 1920 birr/ha for engine-operated weeder and traditional weeding methods, respectively. based on the findings, it is concluded that the performance of the engine-operated weeder can be an efficient, effective, and economically possible option with the high prospect of extending technology for small and medium-scale farmers. however, this plenty of scope for improvement on the machine. recommendations the prototype weeder performance evaluation revealed that it can be used successfully on the farm for weeding operations. to make the weeder applicable and acceptable among farmers, the following steps are recommended for further study and improvement on the machine: • the machine should be tested on different soil types, • different types of weeding blades should be designed and tested, • adaptation, modification, and performance test of the machine for multi-crops weeding operation should be done and • demonstration and scaling up of this machine should be undertaken at the farm level. references alizadeh, m. r., (2011). field performance evaluation of mechanical weeders in the paddy field. scientific research and essays, 6(25), 5427-5434. biswas, h.s. and yadav, r.,(2004). animal drawn weeding tools for weeding and intercultural in black soil. agricultural engineering today, 28(1), 47–53 csa(central statistical agency), (2013). report on area and production of major crops meher season. report on area and production of major crops meher season. addis ababa. csa(central statistical agency), (2015). report on area and production of major crops, private peasant holdings, and meher season. statistical bulletin 278. addis ababa, ethiopia. csa (central statistical agency), (2016). report on area and production of major crops, private peasant holdings, meher season, addis ababa, ethiopia. fao(food and agriculture organization)., (2015). agricultural production statistics. faostat. rome. gavali, m. and kulkarni, s.,(2014). comparative analysis of portable weeder and powers tillers in the indian market. international journal of innovative research in science, engineering and technology, 3(4), 11004-11013. hegazy, r. a., abdelmotaleb, i. a., imara, z. m. and okasha, m. h.,(2014). development and evaluation of small-scale power weeder. misr j. ag. eng, 31(3), 703-728. karale, d. s., khambalkar, v. p., bhende, s. m. and wankhede, p. s., (2008). energy economic of small farming crop production operations. world j. of agric. sciences, 4(4), 476-482. kankal, u., now, r. and palled, v. k., (2014). optimization of operational parameters and performance evaluation of self-propelled weeder for field crops. international journal of applied agricultural & horticultural sciences, 5(1), 56–60. kebede desta., (2000). weed control methods used in ethiopia. animal power for weed control. technical centre for agricultural and rural cooperation (cta), wageningen, the netherlands. kepner, r. a., bainer, r., and barger, e. l., (2005). principle of farm machinery. 3rd edition. new delhi: cbs publishers & distributors pvt. ltd. lopez-granados.f.,(2011). weed detection for sitespecific weed management, mapping, and real-time approaches. weed research, 51, 1–11. manuwa, s. i., odubanjo, o. o., malumi, b. o. and olofinkua, s. g.,(2009). development and performance evaluation of a row-crop mechanical weeder. journal of engineering and applied sciences, 4(4), 236-239. minot, n., warner, j., lemma, s., kasa, l., gashaw, a. and rashid, s., (2019). the wheat supply chain in ethiopia:patterns,trends,and policy options. gates open res, 3(174), 174. moa(ministry of agriculture), (2012). animal and plant health regulatory directorate. crop variety register, issue no. 15. addis ababa, ethiopia. mohammad, a., baghestani, m. a., soufizadeh, s. and pa ge 42 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 2(2) 21-42, 2023 bagherani, n., (2007). weed control and wheat (triticum aestivum l.) yield under application of 2, 4-d plus carfentrazoneethyl and florasulam plus flumetsulam. crop protection, 26(12), 1759-1764. monalisha, s. and goel, a. k., (2017). development of a multipurpose power weeder. the international journal of social sciences and humanities invention, 4(6), 3527– 3531. nkakini, s. o., akor, a. j, and ayotamuno, j.,(2009). design and fabrication of engine-powered rotoweeder. journal of agricultural engineering and technology(jaet), 17(1), 7–16 nkakini, s. o., akor, a. j., ayotamuno, j., ikoromari, a. and efenudu, e. o., (2010). field performance evaluation of manual operated petrol engine powered weeder for the tropics. agricultural mechanization in asia, africa & latin america, 41(4), 68-73 olukunle, j., and oguntunde, p. g., (2006). design of a row crop weeder. conference on international agricultural research for development.tropentag 2006 university of bonn. padole., (2007). performance evaluation of rotary powerweeder. agricultural engineering today, 31(3and4), 30-33. rangapara dineshkumar, j., (2014). development of mini tractor operated picking type pneumatic plante. r anand agricultural university. shehzad, m. a., nadeem, m. a. and iqbal, m., (2012). weed control and yield attributes against postemergence herbicides application in wheat crop, punjab, pakistan. global advanced research journal of agricultural science, 1(1), 7-16. tajuddin, a., (2006). design, development, and testing of engine-operated weeder. agricultural engineering today, 30(5), 25–29. varshney, r. a., tiwari, p. s., narang, s. and mehta, c., (2005). databook for agricultural machinery design. book no. ciae, 2004/1, ciae, bhopal (m.p.) india. verma, a. and victor, v. m., (2003). design and development of power-operated rotary weeder for wetland paddy. agricultural mechanization in asia and latin america, 34(4), 27–29. yadav, r. and pund. s., (2007). development and ergonomic evaluation of manual weeder. agricultural engineering international: the cigre journal. manuscript pm 07 022, ix. pa ge 1 pa ge 10 9 american journal of smart technology and solutions (ajsts) alpas: a solar-powered weed cutter with obstacle detection and bluetooth-based smartphone control for sustainable ground maintenance marloun k. gasoc1*, ronel g. patunongon1, helton m. gultiano1, jhul jhe necole a. davines1, cristyl dia b. bondad1 volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.4948 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: april 05, 2025 accepted: may 08, 2025 published: june 24, 2025 maintaining large spaces, such as school grounds, is expensive, time-consuming, and labor-intensive. gas-powered weed cutters are pollution-causing and expensive to run. alpas is a new, solar-powered, bluetooth smartphone-controllable weed cutter with an obstacle detection system that provides a simple, cost-effective, and environmentally friendly weed control system. the system reduces its effect on the environment and its fossil fuel dependence by harnessing the power of solar energy to run its motors. an obstacle detection system is used to prevent malfunctions and ensure safe operation, while the bluetooth smartphone control provides remote operation and enhances user convenience and efficiency. performance testing compared alpas with conventional weed cutters in terms of power consumption, noise level, environment-friendliness, and cutting efficiency. the results established that alpas reduced fuel dependency, significantly reduced running costs, and provided reliable efficiency under different conditions. in addition, the performance of the system for green ground care was attested by the respondents’ responses, which indicated acceptable responsiveness (4.74), ease of use (4.59), and reliability (4.97). alpas offers technological innovation in ground care providing a green and sustainable option that aligns with current sustainability agendas. future development can focus on raising battery capacity and optimizing automated functions to further optimize its performance. keywords bluetooth control, obstacle detection, renewable energy, solar-powered weed cutter, sustainable ground care 1 department of education, caraga region, surigao del sur division, hinatuan national comprehensive high school, philippines * corresponding author’s e-mail: marloun.gasoc@deped.gov.ph introduction weed refers to any plant that proliferates in an unsuitable location at an inappropriate period, inflicting more harm than benefit. it is a plant that fights with crops for sunlight, nutrients, and water. this may lower land value and agricultural productivity while increasing maintenance costs. the degree of weed infestation in agricultural fields is greatly influenced by agronomic practices, such as crop rotation, tillage techniques, fertilization methods and timing, row spacing, seeding densities, herbicide application, crop selection, cultivar competitiveness, soil type, fertility status, and environmental conditions (chauhan et al., 2012; swanton et al., 2015). because they compete with crop plants for nutrients, light, and water, weeds, as botanical pests, significantly lower agricultural productivity (swanton et al., 2015; ramesh et al., 2017). conventional weeding methods rely on engine-based equipment or hand-operated scissors, which require significant man-hours and fuel, thus extending the period of operation. hinatuan national comprehensive high school possesses an extensive campus comprising a regular athletic oval with other green spaces. significant resources are needed to maintain these grounds, both in terms of personnel and financial support, to keep it clean and attractive. this already calls for considerable commitment from our staff, as well as a significant percentage of the school’s funds that could be otherwise used for other critical needs. various studies have proven the efficiency of solar cutters, highlighting the advantage of the environment with a very practical means of lowering air and noise pollution, thus enabling users to beautify their grass and conserve their health. there exists a significant gap in the literature on an automatic solar weed cutter integrated with an obstacle detection system and with bluetooth smartphone control that allows for operation without human intervention. this project, alpas: a solar-powered weed cutter with obstacle detection and bluetooth-based smartphone control for sustainable ground maintenance, seeks to address one very important felt need in hinatuan national comprehensive high school. the innovative technology, which functions with solar energy and intends to minimize dependency on fossil fuels, offers sustainability and hence minimizes the carbon footprint of conventional weed cutters. it is fitted with an obstacle detection system to guarantee that it operates safely thereby avoiding potential accidents and protecting adjoining objects. besides that, the bluetooth-enabled smartphone control means that this equipment can be controlled from a smartphone, offering an easy interfacing method for remote operation that is simple yet highly accurate. alpas can assist the school by maximizing the value of resources, minimizing manual work, and improving overall effectiveness. apart from the evident practicality of these benefits, the project also proves the institution’s commitment to sustainable practices, safety, and innovation. this weed cutter provides a contemporary weave into landscape maintenance, marrying renewable energy systems, leading-edge technology, and cost-effectiveness. in other words, the project responds to immediate requirements of pa ge 11 0 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 109-115, 2025 the school and, at the same time, becomes a benchmark for sustainable solutions in schools and public places on a larger scale. literature review in the past, grass cutters were portable, manually operated machines. they lost energy and caused pollution by using gas and petrol engines. because of this, automatic grass cutters that rely on a battery for steering and obstacle detection must take the place of manual lawn cutters. a linear blade for cutting grass, an ultrasonic sensor for object identification, a motor drive for the robot’s wheels, and an arduino uno microcontroller board served as additional components. in 2017, another automated solar grass cutter that runs on solar power was introduced by yadav et al. (2017) the paper describes the development of a mobile, solar-powered grass-cutting tool that may be used as a backup in the event of an electrical outage. the robot has a solar panel attached to it that is connected to a battery. an inverter that is attached to the battery transforms dc current from the solar panel into ac current, which turns on the ac motor. the motor spins the blade quickly, cutting the grass efficiently. it is attached to the blade shaft by a belt drive. this invention advances the development of an environmentally friendly system. in terms of technology, manual lawn cutting equipment is still the most widely used. in the study conducted by n, m. s. b. (2017), researchers explored the development of a hybrid solar-powered lawn mower designed to overcome the limitations associated with traditional grass-cutting machines. the primary objective was to reduce human effort, lower operational costs, and minimize maintenance expenses, all while eliminating the dependence on fuel. the amount of effort required for weeding depends on the type of weed, its intensity, the time needed, and the worker’s productivity. farmers now view the weed cutter as a multifunctional piece of agricultural machinery. it is sold by several businesses in the market and comes in 2-stroke and 4-stroke types with different capacities. wellknown 2-stroke variants include lightweight aluminum bodies and gasoline engines that produce between 1.8 and 2.1 horsepower. depending on usage, these models can use 600 to 900 milliliters of oil-mixed gasoline per hour (mohite et al., 2021). amrutesh et al. (2014) introduced their research on a yoke mechanism for agriculture, aiming to reduce weed through enhanced cutting efficiency while ensuring operator comfort. they used the crank slider to apply power and found it significantly better than mechanically powered vehicles. pramod et al. (2014) designed and fabricated weed cutting systems to solve these problems sustainably. they developed a platform made of recyclable materials including pvc pipes, and buoyancy tanks made from paint cans, as well as new designs for the cutter head. designed to be an economic solution for aquatic weed management, this machine does the job of removing floating, submerged, or partially submerged vegetation. p.v.v.s. maneendra et al. (2020) designed a motorized farm weeder with a grass collector to effectively remove and collect weeds and debris present between crops. the primary aim of their work was to minimize the time required to remove weeds between plants, thus improving agricultural productivity. according to mandloi et al. (2010), the costefficient shrub-cutting machine was developed through field testing involving measurements of torque and force as well as assessments of load and speed based on the specified design criteria. this research highlights the potential of renewable energy technologies in enhancing the efficiency and sustainability of landscaping practices, as solar-powered mowers can significantly reduce carbon footprints compared to their traditional counterparts. for more detailed insights, you can refer to the original study by n, m. s. b. (2017). a robot that avoids obstacles by using two gear motors to generate simple walking motions. they developed a highly intelligent robot that can quickly identify impediments and, by analyzing the sensor’s data, perfectly avoid them on its path. the robot’s smooth movement is achieved by utilizing two gear motors to enable left, right, or forward movement in response to detected input. infrared sensors were used to identify and avoid obstacles on the way. infrared transmitters are used to continually emit a 38 khz signal. materials and methods research design this study takes a developmental design method to create and evaluate a solar-powered weed cutter that can be controlled by a bluetooth-enabled smartphone and includes obstacle identification. the design combines technical concepts with sustainability and usability considerations to effectively address the challenge of efficient ground maintenance. development procedures the development procedure concerning a weed cutter prototype initiated by methodical assembly of the prototype actors. all components are assembled according to their design specifications, followed by electrical wiring and connection of the equipment and safe operational test and functional performance. subsequently, there is programming and calibration to develop and fine-tune the code for precise motor control and accurate obstacle detection. once assembled and programmed, the prototype was put through thorough testing to ascertain its overall performance, functionality, and reliability, including the responsiveness and ease of use of the bluetooth-based smartphone control system for remote operation. the weed cutter was then tested in the field on the school grounds, where its effectiveness and usability could be determined under real conditions. in this phase, data is collected and analyzed for potential improvements to be introduced in the design of the prototype. finally, the whole process was documented well, and monitoring was done regarding findings that pa ge 11 1 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 109-115, 2025 would bring useful insights and recommendations for further development and practical implementation of the grass cutter. data analysis for the testing and validation process, the procedure follows the methodology presented by kumbar et al. (2024). 1. prototype analysis involved taking readings at designated time intervals with a temperature gun, solar power meter, and multimeter to evaluate the solar panel system’s performance. 2. the temperature gun measured the solar panel’s temperature by directing the device at the panel, yielding celsius readings that offered insights into the thermal conditions impacting the panel’s efficiency. 3. a solar power meter was utilized for the observation of sunshine intensity. the apparatus was faced towards both the sun and the solar panel, registering watts per square meter (w/m2). this parameter is essential for determining how much solar energy is available for conversion. 4. a multimeter was used to take measurements of voltage outputs at equidistant time intervals coming from the solar panel. voltage readings provide ample data concerning the electrical output of the panel. system performance analysis the system performance evaluation was done by considering the methods suggested by kalpana et al. (2024), followed by the power consumed, battery availability, run time, and cutting efficiency. calculations give insight into the general optimization of system performance to ensure a better operating economy and battery management. to evaluate bluetooth-based smartphone control system responsiveness, usability, and reliability in the operation and management of grass cutters from a distance, a structured questionnaire was created by researchers. the questionnaire was thoroughly validated by research experts to make sure it was relevant, accurate, and reliable as a data-gathering tool. a total of 40 respondents were selected through purposive sampling, including parents, school administrators, faculty members, and personnel. each group consisted of 10 participants. results and discussion table 1: comparison of conventional and solar-powered weed cutters parameter conventional weed cutter solar-powered weed cutter power source gasoline/electric solar panel (renewable) fuel/energy cost required not required environmental risk high (gasoline)/moderate (electricity) low (eco-friendly) noise level high potentially lower (electric motor operation manual control bluetooth-controlled (smartphone interface) automation unavailable fully automated via smartphone cost-effectiveness expensive cost-effective (low maintenance) table 1 highlights the key differences between conventional and solar-powered weed cutters. conventional weed cutters, powered by gasoline or electricity, have ongoing fuel costs and pose higher environmental risks, especially from gasoline emissions. in contrast, solar-powered weed cutters utilize renewable solar energy, eliminating fuel costs and reducing environmental impact. conventional models tend to be noisy, while solar-powered versions, with electric motors, are quieter. additionally, solarpowered weed cutters often provide non-manual which is bluetooth-controlled (smartphone), fully automated controlled via smartphone, enhancing user convenience, while conventional models require manual operation and not automated. overall, solar-powered cutters are more cost-effective and environmentally friendly. according to kashyap et al. (2020), mowing grass takes a lot of time and effort. these days, most of the technology available for cutting grass is a manually operated diesel cutter. these kinds of devices that run on unconventional energy sources harm the environment, release greenhouse gases, and contribute to climate change. additionally, these weed eaters contribute to noise pollution, which hurts both the cutter’s and the nearby population’s health. another factor is that diesel fuel is expensive. a solarpowered autonomous grass cutter is being developed to combat the problems of the conventional cutter. the search for alternative energy sources has accelerated due to the dwindling fossil resources. researchers are exploring several alternatives, particularly solar energy, which has become an important aspect of various projects (lingappa et al., 2024). solar technology is new in the solar weed cutter, a simple yet effective machine used to manage lawns locally, in gardens, and in schools (amrutesh et al., 2014). in view of its solar energy operation and high rpm for effective grass cutting, this autonomous weed-cutting vehicle is very new (athina et al., 2021). such eco-friendly devices are cost-efficient, consume little power, and thus are suitable for sustainable landscaping (mudda, 2018). a gas-powered kawasaki grass cutter, priced at ₱13,200, has become rather famous for possessing that high initial cost typical of fossil fuel driven machines. the alpas solar-powered weed cutter, in contrast, has a price of just ₱4,500; this is cheaper mainly because of the low pa ge 11 2 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 109-115, 2025 production costs associated with solar technology and the lack of complicated engine parts. though the solar cutter is cheaper at the front, it has a much lower cost on operation in the far future. traditional lawn cutters have persistent costs of fuel and maintenance; these costs can increase with time depending on fluctuating fuel prices and the need for frequent servicing. the solar models conserve renewable energy, so their operation after the initial purchase is almost free. research shows these projects can be singled out because they do not have an operational cost, which is due to having no fuel charges and much less maintenance compared to gas (kirtiwar et al., 2023). also, via reducing environmental impacts from the use of tools run by the sun (babu et al., 2023). while conventional grass cutters make huge noise and, through combustion of fuel, contribute to greenhouse gas emissions, solar-powered cutters are markedly silent and do not emit any such pollutants, hence providing a sustainable option for lawn maintenance (kalpana et al. 2024). table 2: cost-analysis of traditional grass cutter and solar-powered weed cutter type cost vic kawasaki k td-40 2 stroke grass cutter (source: khm mega tools corp.) ₱ 13,200 alpas: solar-powered weed cutter ₱ 4,500 table 3: testing and evaluation time (min.) temperature (°c) voltage (v) intensity of sunlight (w/m²) intensity of solar panel (w/m²) 9:30 a.m. 45 18.4 878 169 10:00 a.m. 43 19.28 945 156 10:30 a.m. 46 19.18 981 187 11:00 a.m. 49 19.16 991 189 11:30 a.m. 52 18.78 1046 195 12:00 p.m. 55 18.74 1058 235 12:30 p.m. 54 19.03 1054 243 1:00 p.m. 56 18.75 1020 224 1:30 p.m. 52 19.17 1016 221 2:00 p.m. 55 19.14 1010 220 2:30 p.m. 53 19.10 994 205 3:00 p.m. 50 19.80 989 202 the solar panel system testing and assessment were executed in table 3, with temperature, voltage, sunshine intensity, and output from solar panels being measured from around 9:30 a.m. to about 3:00 p.m. the voltage generated in this solar panel system remained constant, going as low as 18.4 v at 9:30 a.m. to as high as 19.8 v at 3:00 p.m. this value, however, indicates that the power output is thereafter stabilized and protected from variations in the environment. at 12:00 noon, solar intensity reached its highest value of 1058 w/m², followed at 12:30 p.m. with the solar panel output of 243 w/m². beyond this value, the intensity of sunlight and solar panel output began to fall, especially around 1:30 p.m. however, output voltage hardly showed any deviation, suggesting that the panel worked well despite being in adverse conditions of declining solar intensity. the temperature readings kept rising constantly throughout the day, starting from 45°c at 9:30 a.m. up to 56°c by 1:00 p.m. nevertheless, high temperatures were not markedly detrimental to the operation of the system, thus proving the endurance as well as utility of the panel in extreme heat. the data show that the equipment runs well and absorbs solar energy during the day, irrespective of the weather conditions. this substantiates the reliability and practicality of solar panels as sources of energy for various applications. table 4 depicts the performance of the solar-powered weed cutter on different grass species. the average height for crabgrass was cut from 224 mm to 85 mm, a significant height reduction. goosegrass went similarly table 4: average weed height before and after cutting using the solar-powered weed cutter weeds type average height before moving (mm) average height after moving (mm) crabgrass (digitaria ischaemum) 224 85 goosegrass (eleusine indica) 234 90 quackgrass (elytrigia repens) 70.5 50.5 pa ge 11 3 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 109-115, 2025 from 234 mm to 90 mm. quack grass was reduced from an earlier height of 70.5 mm down to 50.5 mm. these results indicate that the solar cutter performs well on different grass species and effects a reasonable reduction in height. the slight deviation in heights for crabgrass can be attributed to the fact that the species has a naturally lower growth height, indicating that the cutter is efficient for taller as well as shorter grass species. such variability further establishes the cutter’s versatility and efficacy in treating different grass species, indicating its viability for table 5: system performance analysis aspect formula results power consumption per motor (pm) pm=vm×im 58.9w total power consumption (tpc) tpc=pm×4 motors 235.6w usable battery capacity (cusable) cusable=cbattery×0.85 68ah estimated runtime r=(cusable/itotal ) 415.2 min effective cutting width (effectivecw) effectivecw=(1000mm/(cutting width×cutting ratio)) 0.1736 meters distance cut per rotation (dr) dr=effective cutting width 0.19 meters distance cut per minute (dr) dcpm= dr×rpm 11.4 meters per min several landscaping applications. multiple critical performance parameters show how successful the solar-powered grass cutter is. each motor consumes 58.9 w of power, totalling 235.6 w for the complete system. this clever use of energy is critical for optimizing battery performance. the lawn mower’s battery capacity of 68 ah allows it to run for approximately 415.2 minutes. this longer runtime means that the lawn cutter can function for a long period before needing to be recharged, making it ideal for cutting grass in large areas. the machine has an effective cutting width of 0.1736 meters, making grass maintenance simple. it cuts 0.19 meters of space each time it turns, resulting in a rate of 11.4 meters per minute. this cutting speed allows users to complete more tasks in less time, increasing productivity. table 6 : overall evaluation of bluetooth-based control system criterion weighted mean descriptive equivalent interpretation responsiveness 4.74 strongly agree the system performs exceptionally well in meeting the stated expectations and functionalities. ease of use 4.59 strongly agree the system performs exceptionally well in meeting the stated expectations and functionalities. reliability 4.97 strongly agree the system performs exceptionally well in meeting the stated expectations and functionalities. average 4.76 strongly agree the system performs exceptionally well in meeting the stated expectations and functionalities. the responsiveness, with a weighted mean score of 4.74, is commendable. research corroborates this, demonstrating that bluetooth technology enables realtime communication and control, essential for responsive systems. juned and unnikrishnan (2014) illustrate that a bluetooth-based remote monitoring and control system achieved rapid data transmission within 10 seconds, allowing swift operator responses to fluctuations in monitored parameters such as temperature and humidity. this aligns with user feedback emphasizing the system’s ability to meet operational demands promptly. the 4.59 ease of use score indicates that consumers find the system straightforward and easy to navigate. kulkarni et al. (2019) show that merging bluetooth with mobile applications improves user interaction with home automation systems, making them more accessible to persons with minimal technical abilities. creating user-friendly interfaces is critical for improving user experience, as evidenced by numerous research on smart home technology. the ease of use improves user happiness and overall system acceptance. a reliability score of 4.97 implies extremely consistent performance. users rely on control systems to provide precise oversight and management; therefore, reliability is critical. liu and uthra (2020) corroborate this assertion by demonstrating that bluetooth-enabled systems may sustain robust connections across long distances (up to 60 meters), ensuring continued functionality. this dependability builds trust among users, improving their overall impression of the system. the mean score of 4.76 validates the positive assessment of the bluetoothcontrolled system. this total satisfaction can be due to the combined impacts of high responsiveness, ease of use, and dependability. several studies imply that these factors are inextricably related; a dependable and attentive system improves the user experience. conclusions alpas is a breakthrough model for weed management that is both sustainable and environmentally responsible. it is pa ge 11 4 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 109-115, 2025 ideal for huge spaces like schools, parks, and large offices. this environmentally friendly weeding device is solar powered, which reduces running expenses and eliminates the need for gasoline. thus, it generates significant savings while also helping environmental sustainability. alpas is fueled by solar energy on a continual basis under varying light conditions. a specially built panel guarantees that the machine receives the constant power it requires to work in low-light circumstances. this characteristic ensures stability and consistent performance, which is especially useful for outdoor maintenance. the obstacledetection feature is a standout among remarkable safety improvements. with this function, alpas may detect impediments and navigate them without assistance, boosting operational safety. this is especially critical when the user’s safety and comfort are at risk. to enhance the user experience, smartphone control via bluetooth enables remote operation and real-time monitoring. this smart functionality improves simplicity of use by allowing users to control the machine remotely and alter settings for specific jobs. alpas is a significant step forward for green landscaping alternatives due to its solar-power efficiency, safety-focused obstacle identification, and digital connectivity. references alagarsamy, s., subramanian, r. r., bobba, p. k., jonnadula, p., & devarapalli, s. r. (2021). designing a smart speaking system for voiceless community. in lecture notes in networks and systems (pp. 21–34). https:// doi.org/10.1007/978-981-16-2126-0_3 amrutesh, p., sagar, b., & venu, b. (2014). solar grass cutter with linear blades by using scotch yoke mechanism. international journal of engineering research and applications, 4(9), 10–21. https://www. researchgate.net/publication/33695249_solar_ grass_cutter_with_linear_blades_by_using_ scotch_yoke_mechanism athina, d., kumar, d. k., kalyani, r., & vittal, k. (2021). solar grass cutter using embedded platform experimental validation. iop conference series materials science and engineering, 1057(1), 012086. https://doi. org/10.1088/1757-899x/1057/1/012086 babu, v. r., elumalai, s., sharon, v. m., yogesh, s., & raj, j. i. l. (2023). smart grass cutter using solar power system. tuijin jishu/journal of propulsion technology, 44(6), 5285. chauhan, b. s., singh, r. g., & mahajan, g. (2012). ecology and management of weeds under conservation agriculture: a review. crop protection, 38, 57–65. https://doi.org/10.1016/j.cropro.2012.03.010 deriquito, j. (2021). when should a timing chain be replaced? autodeal. retrieved may 9, 2025, from https://www. autodeal.com.ph/articles/car-features/when-shouldtiming-chain-be-replaced juned, m., & unnikrishnan, s. (2014). bluetooth-based remote monitoring & control system. journal of basic and applied engineering research, 1, 108–111. © krishi sanskriti publications. http://www.krishisanskriti. org/jbaer.html kalpana, m., sravani, p., sravya, j., varma, m. a. k., lokesh, p., sivani, b. l., & hemanth, p. s. (2024). development and analysis of a solar-powered grass cutter with integrated collector using iot. engineering proceedings, 66(1), 43. https://doi.org/10.3390/ engproc2024066043 kashyap, d., bordoloi, u., & buragohain, a. a. (2020). design of a fully automated solar grass cutter for campus cleaning. international journal of creative research thoughts (ijcrt), 8(5). https://ijcrt.org/ papers/ijcrt2005330.pdf kirtiwar, v., mahale, j., chaure, m., khode, p., & ahire, v. (2023). design and fabrication of solar powered grass cutter. international journal of progressive research in engineering management and science, 3(5), 1111–1116. https://www.ijprems.com kulkarni, b. p. (2019). home automation using bluetooth control technology. international journal for research in applied science and engineering technology, 7(4), 2136– 2140. https://doi.org/10.22214/ijraset.2019.4386 kumbar, g. m., konduskar, s., patil, y., konduskar, p., milake, r., & patil, s. (2024). automatic solar grass cutter. international journal of research publication and reviews, 5(4), 8408–8412. https://www.ijrpr.com lingappa, j., raghavender, v., aparna, g., & manideep, ch. (2024). automated grass cutter using renewable energy. edp sciences. https://creativecommons.org/ licenses/by/4.0/ liu, y., & uthra, r. a. (2020). bluetooth based smart home control and air monitoring system. international journal of advanced research in engineering and technology, 11(5), 264-274. http://iaeme.com/home/issue/ ijaret?volume=11&issue=5 mandloi, r. k., gupta, r. k., & rehman, a. (2010). design, development and testing of low capital and operational cost shrub cutting machine. journal of rangeland science, 1, 121–124. maneendra, p. v. v. s., rajesh, s., kumar, v. s., raju, m. b., & kumar, k. s. (2020). fabrication of motorized agriculture weeder with grass collector. journal of emerging technologies and innovative research (jetir), 7(3). https://www.jetir.org mohite, d. d., agrawal, k., kumar, k., & deb, a. (2021). technical aspects of multipurpose weed cutter or power weeder. international journal of enhanced research in science, technology & engineering (ijerste), 10(7), 35. https://doi.org/10.13140/rg.2.2.11613.33765 mudda, m., teja, v., mudd, s., & kumar, p. (2018). automatic solar grass cutter. international journal for research in applied science & engineering technology, 6(4), 1148–1151. https://www.researchgate.net/publication/325221845_ automatic_solar_grass_cutter n, m. s. b. (2017). design and development of hybrid powered grass cutter. international journal for research in applied science and engineering technology, v(iii), 657– 662. https://doi.org/10.22214/ijraset.2017.3123 pa ge 11 5 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 109-115, 2025 n, m. s. b. (2017b). design and development of hybrid powered grass cutter. international journal for research in applied science and engineering technology, v(iii), 657– 662. https://doi.org/10.22214/ijraset.2017.3123 neha, n., & asra, s. (2018). automated grass cutter robot based on iot. international journal of trend in scientific research and development, 2(5), 334–337. https://doi. org/10.31142/ijtsrd15824 pandit, a. (2019, march 12). obstacle avoiding robot using arduino and ultrasonic sensor. circuitdigest. https:// circuitdigest.com/microcontroller-projects/arduinoobstacle-avoding-robot pramod, r., sandesh, m. d., jayaram, m. s., & kalasagarreddi, k. (2014). design and development of sustainable weed cutter. international conference on engineering of complex computer systems, 287–292. https://doi.org/10.1109/iceccs.2014.65 ramesh, k., matloob, a., aslam, f., florentine, s. k., & chauhan, b. s. (2017). weeds in a changing climate: vulnerabilities, consequences, and implications for future weed management. frontiers in plant science, 8. https://doi.org/10.3389/fpls.2017.00095 yadav, r. a., chavan, n. v., patil, m. b., & mane, v. a. (2017, february). automated solar grass cutter. international journal of scientific development and research (ijsdr), 2(2), 65–69. https://ijsdr.org/papers/ ijsdr1702016.pdf yadav, r. a., chavan, n. v., patil, m. b., & v. a. mane. (2017). automated solar grass cutter. in international journal of scientific development and research (ijsdr), 2(2), pp. 65–66). https://ijsdr.org/ papers/ijsdr1702016.pdf pa ge 1 pa ge 89 american journal of smart technology and solutions (ajsts) simulation and performance enhancement of thermal combustion in a liquid fuel swirl burner through blades parametric variation ademola samuel akinwonmi1*, folajinmi onikepo onadeko1 volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.4614 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: february 22, 2025 accepted: march 28, 2025 published: june 20, 2025 research has explored the need for an increase in the performance of a liquid fuel swirl burner (lfsb). an experimental study carried out in lfsb having a varied number of blades and angles of blades yielded the desired improvement. to further improve on the experimental work, extended study is required. this was carried out through computational fluid dynamics simulation methods, which is cost-effective and have been established to be reliable. this study therefore explores the comparison between experimental results and simulation means to validate the simulation methods and results obtained. the work aimed to enhance thermal combustion in the swirl burner through variations in the blade angles (10°, 20°, 30°, 40°, 45°, 50°, 60°, 70°) and the number of blades in the burner (4, 6, 8, 10 and 12). computational methods were employed when designing and simulating the burner’s parameters using solidworks and ansys fluent. the result showed that the burner with 10 blades at 70° yielded the highest temperature of 1024.547°c and the burner with 12 blades at 10° degrees produced the lowest pressure drop of 11973219.93pa. thereby improved combustion was achieved by obtaining the highest temperature at the burner outlet and lowest pressure drop which yielded effective combustion. keywords combustion, pressure drop, simulation, swirl burner, temperature 1 department of mechanical engineering, ajayi crowther university, oyo, nigeria * corresponding author’s e-mail: as.akinwonmi@acu.edu.ng introduction the process of combustion describes the transformation of chemical energy (from the fuel substance utilized) to other types of energy including, thermal, electrical, gravitational, kinetic, nuclear and electromagnetic. the well-known definition of energy is that it is the ability to do work, in this article by clark & yusoff (2014), combustion is regarded as a variation of work where the atomic bonds of a substance (fuel) are liberated through oxidation reaction and as a result, the products are the generation of heat and the creation of new chemical bonds. the world accounts for 80% of its energy generation through the use of fossil fuels (crude oil, natural gas and coal) (aliyu et al., 2015) the primary mode of utilizing fossil fuels for the generation of electricity is through combustion (martins & brito, 2020). alternative energy, sustainable energy and clean energy are arguably interchangeable terms used to describe renewable energy. renewable energy is energy that is derived from replenish able means. it is used to produce continuous energy (panwar et al., 2011). nations that are still undergoing development are at risk to climate change more so than already developed nations (ikein, 2017). this is as a result of their economies prevalence on agriculture, the lack of capital to properly adjust to these changes in their climate and their increased exposure to the effects. unlike developed nations such as the united states of america, russia, germany etc, that have the capabilities in producing and manufacturing needed to successfully switch to renewable energy technologies, such as wind turbines, hydrogen fuel cells and photovoltaic cells (ogbonnaya et al., 2019). examples of combustion fuels are carbon dioxide (co2), perfluorocarbons, sulphur hexafluoride, nitrogen trifluoride, nitrous oxide, hydrofluorocarbons, and methane (european parliament, 2023). over the years, researchers have found many ways to improve combustion. among these technologies is the introduction of swirling flows to the design of a burner. according to (sengupta et al., 2021), the smooth and undisturbed operation of a burner is subject to stringent conditions during its design. according to mansouri and boushaki (2018) in recent times, the design of burners includes vanes which are used in aerodynamically stabilizing the flames. these vanes are referred to as swirlers. with the introduction of these swirlers to the design of a burner, the burner is known as a swirl burner. a swirl burner is a burner of helical configuration that produces swirling flows during combustion (boushaki, 2019). the swirl burner utilizes guiding vanes whose purpose is to supply the swirl flow to air for combustion (xiao et al., 2018). utilization of non-premixed swirl in the process of combustion has a number of advantages including the ability to control the flow coupled with the abatement of harmful pollution emission primarily in the form of nox (nitrogen oxides) (schmittel et al., 2000). in a study by sreenivasan et al. (2012), it is noted that premixed flames generate less of hazardous gases such as carbon monoxide and soot. the advantages of using swirl burners are numerous. in day et al. (2003), the classification of swirl burners was noted as follows: swirl burners are classified into axial vane burners, tangential vane burners, and volute burners. the major types of swirlers were also noted namely: volute swirler, tangential vane swirler, axial vane swirler. in the research by yang et al. (2019), premixed and non-premixed combustion modes were both explored in a swirling micro-combustor pa ge 90 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 89-97, 2025 that can be fuelled by hydrogen or air. their effects on the efficiency of combustion, thermal performance and flame stability were explored. in their research, oyewola, et al. (2022) explored the thermal profile of combustion in experimental means in a liquid fuel swirl burner and this approach was done through alternatively changing the configuration of the blades at angles 20°, 30°, 40°, 50°, 60°. the blades were also assembled in the order of 6, 8, 10 and 12. the primary purpose of their research was to increase the temperature of the burner thereby enhancing combustion dynamics. for the straight-edge blades, the highest thermal efficiency recorded was achieved using 6 blades inclined at an angle of 50° (oyewola et al. 2022). jaafar et al. (2012) examined in their research that the determination of the swirl number is an important characteristic in the design of a swirl burner as it assists in ensuring the correct order of the swirl blades. it was also noted in their research that the decrease of the depth of the swirl blade without a modification to the swirl blade angle will lead to an improvement in the swirl strength. to further emphasize the importance of swirl blades in the optimization of combustion in a burner, (surjosatyo & priambodho, 2011) reported that to enhance the quality of the flame or the strength of the flame in a low-swirl burner, swirl vanes of 6, 8 and 10 in number, inclined at an angle of 30 from the horizontal axis are needed to decrease the diameter of the fuel entering at the inlet side of the burner. in research by (dhyani & phade, 2024), computer-aided design (cad) techniques using solidworks 2022 was utilized in designing a burner and its vanes. benim et al., (2022) performed research utilizing a mixture of pulverized coal with biomass in an oxy-combustion process in a swirl burner using computational means. not enough research extensively focuses on the modification of the blades of the swirl. this modification could be achieved through several ways, perforations in the blades could be a means and as done by (akinwonmi et al., 2023) in the research, six swirl blades were presented with the modification of the angles and blade type (straight and curved edge blades) and also the variation of the angles of the blades ultimately yielding a positive result. most engineering systems usually require a combustion system. configuration of the burner itself will have an impact on the combustion efficiency and emission of the burner. high cost of experimentation limits the optimization of burner parameters for efficient performance of burners and other combustion systems. these are several studies on the use of simulation packages such as ansys, abacus, and matlab to optimize process parameters thereby reducing production cost and time. however, there are limited studies on the use of computational software in optimising process parameters and performance of liquefied fuel swirl burner (lfsb) therefore, the aim of this study is to evaluate the performance of an lfsb using computational fluid dynamics module in ansys. the problem statement of this research paper aims to tackle, analyse and proffer a means of optimizing combustion in a swirl burner through the modification and redesign of its blades. this research is limited to the enhancement of combustion through the redesign of the blades. the main parameters considered in the blades include the shape and the angle in which they are configured on the swirl burners. materials and methods experimental study the detailed results on the experimental set-up and the results gotten were obtained from oyewola et al. (2022). fuel supply (diesel) enters the atomizer at a pressure above 10 bar. while air proceeds through the output centrifugal blower with a 2” gate valve utilized for changing the air flow rate. the material used in the construction of the combustion chamber is stainless pipe steel 304 that has a thickness = 4mm, with dimension108 x 420mm per modular section. there are five modules in total. each module has a flange machined with a projection that exactly fits the recess on the adjacent module, preventing leakage. the module at the base is fastened to the burner body by a 1 m10-6h bolt and nut. there are ports provided for measurement probes in each module. observing and evaluating the flame length, velocity and pressure drop is done via the modular combustion chamber. the material used in making the blades and vanes is mild steel that is welded to the centre core on a rode whose base has been threaded to the burner for fastening/tightening utilizing the m10-gh nut. from the above data analysed from research carried out by oyewola et al. (2022), it is important to note that the efficiency of the combustion was measured based on these four parameters: flame length, combustion temperature, velocity and pressure drop. cfd simulation the physical geometry/model is drawn with solidworks software before being imported into ansys fluent for cfd simulation. cfd software (ansys fluent) is utilised to simulate combustion. the combustion performance parameters for this study are: the maximum temperature and the pressure drop. meshing and grid sensitivity was done on the physical model, while governing equation such as the continuity equation, momentum and energy conversion equation were implemented to solve the thermal problems of combustion and airflow. the flow type was determined using reynolds number: re = (ρvd)/μ where: ρ = ambient air density, v = velocity = 12m/s, d = diameter of the inlet of the combustion chamber, μ = viscosity (taking the ambient air density as ρ = 1.2 kg/m and velocity v = 12m/s the estimated value of re is more than 3500, suggesting that the flow is turbulent. a turbulent model was selected. assigning premixed flow, boundary conditions at the inlet, the wall and the outlet will be assigned as: inlet: mass flow rate and initial temperature. outlet: pressure outlet pa ge 91 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 89-97, 2025 wall: adiabatic and non-slip conditions the cfd analysis was validated by comparing the result with the experimental study. once the temperature and pressure drop characteristics have been established, the geometry was varied to obtain an optimized design. the assembled drawings were designed in the format of the table below making a total of 25 different drawings. the schematic of the burner geometry and the blades are shown in figure 1 and figure 2 respectively, while the configurations of the blades are shown in table 1. table 1: no of blades and angles modelled. angles no of blades 10° 20° 30° 40° 45° 50° 60° 70° 4 blades √ √ √ √ √ √ √ √ 6 blades √ √ √ √ √ √ √ √ 8 blades √ √ √ 10 blades √ √ √ 12 blades √ √ √ figure 1: the burner geometry. (a) isometric view (b) sectional view. figure 2: blades design. (a) the isometric view (b) front view figure 3: 3d geometry of swirl burner. (a) unmeshed (b) meshe results and discussions the results of the simulation done with ansys fluent are discussed in this chapter. as mentioned in the methodology, the number of blades and angle type were varied to obtain the temperature at the outlet of the burner (combustion chamber) and the pressure drop at the outlet. temperature from outlet of the burners table 2 shows the varying temperatures obtained when the simulation was carried out through ansys fluent and results were obtained. for the lfsb configuration with 4 and 6 blades, temperature results obtained cut across angles (10°, 20°, 30°, 40°, 45°, 50°, 60°and 70°) whilst for 8-blade, 10-blade and 12-blade lfsb configurations, the temperature results obtained were for (10°, 45°and 70°) only. pa ge 92 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 89-97, 2025 table 2: outlet temperature (°c) at different blade angles blade angles (°) 4 blades 6 blades 8 blades 10 blades 12 blades 1. 10 697.167 656.202 687.11177 657.24116 666.6095 2. 20 700.329 709.767 3. 30 669.301 713.179 4. 40 672.824 660.198 5. 45 670.306 654.376 664.9423 654.02811 654.48336 6. 50 674.064 666.579 7. 60 666.677 657.335 8. 70 679.202 687.454 668.31346 1024.5465 667.36365 table 3: variation of simulation result with experimental results for 6 number of blades blade angle (°) experimental results simulation results percentage difference (%) 20 812 709.76 12.59 30 857 713.18 16.78 40 885 660.20 25.40 50 941 666.58 29.16 60 918 657.34 28.39 validation of simulation result with experimental result the liquid fuel swirl burner used in this comparison has 6 blades. the comparison study observes the variation in temperatures across angles 20°, 30°, 40°, 50° and 60°. the detailed study is expressed in table 3 and figure 4. there is a similarity in the trend of the graph of both the experimental and simulation results. there is an increase in the trend of the angles 30°to 40° and from 40° to 50° of the experimental results and an increase in the trend of the 20°to 30° and from 40° to 50° of the simulation results. the experimental result has higher temperatures, and the contribution to this is that it is an exothermic reaction, and there is the influence of environmental factors on the result. the exemption of environmental factors and simulation procedure has contributed to the simulation results obtained. figure 4: comparison of experimental and simulation results for 6 number of blades effect of blade angle on temperature performance figure 5(a) shows the variation of the trends on the graph. according to the graph, the swirl burner with 6 blades produced the highest temperature at 30°. the lfsb with 6 blades at 45° produced the lowest temperature. figure 2(b) shows the variation of the trends on the graph. according to the graph, the swirl burner with 6 blades produced the highest temperature at 70°. the lfsb with 6 blades at 45° produced the lowest temperature. figure 2(c) a bar chat was used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the variation of blades with the highest temperature is 10 blades at 70°. the variation of blades that produce the lowest temperature is 10 blades at 45°. pa ge 93 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 89-97, 2025 figure 5: effect of blade angles on output temperature (a) 4 and 6 blades (b) 6 and 8 blades (c) effect of blade angles (10, 45 and 70 degrees) on output temperature. effect of number of blades on temperature performance figure 6(a) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest temperature produced for 4 blades is 20° and the lowest temperature is 60°. figure 6(b) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest temperature produced for 6 blades is 30° and the lowest temperature is 45°. figure 6(c) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest temperature produced for 8 blades is 10° and the lowest temperature is 45°. figure 6(d) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest temperature produced for 10 blades is 70° and the lowest temperature is 45°. figure 6(e) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest temperature produced for 12 blades is 70° and the lowest temperature is 45°. (c) (d) pa ge 94 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 89-97, 2025 figure 6: effects of number of blades on temperature performance (a) 4 blades (b) 6 blades (c) 8 blades (d) 10 blades (e) 12 blades figure 7: effect of blade angle on pressure drop (a) 4 and 6 blades. (b) 6 and 8 blades. (c) 4, 6, 8, 10 and 12 blades. pressure drop from outlet of the burner the pressure drops at different blade angles are presented in table 4. the configuration with the least pressure drop is the most efficient. it can be observed that the pressure drop varied from 13.4 mpa to 17.4 mpa when 4 blades were used. the least pressure drop was recorded at an inclination of 20o while the most pressure drop was recorded at 10o. increasing the number of blades to 6 showed that a maximum pressure drop of 22.4 mpa was recorded at 60o while a minimum of 14.5 mpa was recorded at 50o. for 8, 10, and 12 blades, only three angles were investigated. the angles are 10o, 45o, and 70o. in all, 45o recorded the least pressure drop. the pressure drop at 45o was similar regardless of the number of blades ranging between 14 – 15 mpa. this shows that 45o inclination would be a suitable configuration to obtain a reasonable pressure drop in the swirl burner. however, the best performing configuration is 4 blades at 20o. table 4: pressure drop (pa) at different blade angles blade angles (°) 4 blades 6 blades 8 blades 10 blades 12 blades 10 17447983 16935336.8 19300927.7 18758026.2 11973219.9 20 13444413.6 16819360.7 30 15030043 16926780.5 40 16559992.4 19377894 45 14972311.8 15088128 15889589.4 13929679.1 14793106.5 50 15140152.7 14479914.8 60 16987400.4 22395036.1 70 16018733.3 15776540.5 18045876.7 20098840.9 18382667.4 effect of blade angle on pressure drop performance of swirl burners figure 7 (a) a line chart was used to show the variation of the trends on the graph. according to the graph, the swirl burner with 6 blades produced the highest temperature at 60°. the lfsb with 4 blades at 20° produced the lowest temperature. figure 7 (b) a line chart was used to show the variation of the trends on the graph. according to the graph, the swirl burner with 8 blades at 10° produced the highest temperature. the lfsb with 6 blades at 45° produced the lowest temperature. figure 7 (c) a line chart was used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the variation of blades with the highest temperature is 10 blades at 70°. the variation of blades that produce the lowest temperature is 12 blades at 10°. (b) (c) pa ge 95 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 89-97, 2025 effect of number of blades on pressure drop performance of the lfsb figure 8 (a) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest pressure drop produced for 4 blades is 10° and the lowest pressure drop is 20°. figure 8 (b) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest pressure drop produced for 6 blades is 60° and the lowest pressure drop is 50°. figure 8 (c) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest pressure drop produced for 8 blades is 10° and the lowest pressure drop is 45°. figure 8 (d) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest pressure drop produced for 10 blades is 70° and the lowest pressure drop is 45°. figure 8 (e) a line chart is used to show the in-depth variation of the different trends on the graph. from the graph above, it is obtained that the highest pressure drop produced for 12 blades is 70° and the lowest pressure drop is 10°. figure 8: effect of number of blades on pressure drop. (a) 4 blades (b) 6 blades (c) 8 blades (d) 10 blades (e) 12 blades discussion comparison of simulation results with experimental results according to this comparison, the 6-blade lfsb inclined at an angle of 50° produced the highest combustion temperature (941°c) for the experimental results. factors considered include the environmental temperature that had an impact on the combustion temperature. other environmental factors include the endothermic reaction (energy being absorbed from the surroundings in the form of heat) and exothermic reaction (energy being transferred or released into the system surrounding also in the form of heat). design techniques and fabrication considerations also had an overall impact on the results obtained. the lfsb used in the experiment had probes for measuring the temperature along the burner length. the temperature from the experimental results used in this comparison is the minimum temperature along the length of the fabricated burner for each of the different blade angle configurations as this had the most similarity to the simulation result. in the case of the simulation results, the 6-blade lfsb inclined at an angle of 30° produced the highest combustion temperature (713.1789°c). the methodology of the simulation on ansys fluent (non-premixed combustion) and the design parameters assumption contributed to the deviation in the results when compared with the experimental results. unlike the experimental, the lack of endothermic and exothermic environmental factors is a defining factor in the variation of the results. pa ge 96 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 89-97, 2025 comparison of the number of blades of the lfsb on the highest temperature and pressure drop in line with the parametric study, the number of blades on the lfsb was used as a comparison means to obtain the best configuration with the highest temperature and the lowest pressure. in the case of the comparison between the burner with 4 blades and 6 blades at angles (10°, 20°, 30°, 40°, 45°, 50°, 60° and 70°), the highest temperature obtained was the 6 blades inclined at angle 30° with the temperature 713.1789°c. in the case of the comparison between the burner with 6 blades and 8 blades at angles (10°, 45°, and 70°), the highest temperature obtained was the 6 blades inclined at an angle of 70° with the temperature of 687.4539°c. in the case of the comparison between the burner with 4, 6, 8, 10 and 12 blades at angles (10°, 45°, and 70°), the highest temperature obtained was the 10 blades inclined at angle 70° with the temperature 1024.547°c. the lowest pressure drop obtained between the burner with 4 blades and 6 blades at angles (10°, 20°, 30°, 40°, 45°, 50°, 60° and 70°) is 4 blades at 20° (13444413.5927 pa). in the case of the comparison between 6 blades and 8 blades, the lowest pressure drop obtained at angles (10°, 45° and 70°) is 6 blades at 45° (15088127.99 pa). in the case of the comparison between the burner with 4, 6, 8, 10 and 12 blades at angles (10°, 45°, and 70°), the lowest pressure drop obtained was the 12 blades inclined at an angle of 10° with the temperature (11973219.92634 pa). comparison of effect of the blade angle of the lfsb on the highest temperature and pressure drop the highest temperature obtained when the blade angles (10°, 20°, 30°, 40°, 45°, 50°, 60° and 70°) in a 4-blade burner were varied is 4 blades at 20° with a temperature (700.3287°c). in the case of a 6-blade lfsb inclined at angles (10°, 20°, 30°, 40°, 45°, 50°, 60° and 70°), the highest temperature obtained is 713.1789 for a 6 blades burner at 30°. in the case of an 8-blade lfsb inclined at angles (10°, 45°, and 70°), the highest temperature obtained is 687.112 for a 8 blade burner inclined at 10°. the highest temperature obtained when the blade angles (10°, 45°, and 70°) in a 10-blade burner were varied is 10 blades at 70° with a temperature of (1024.547°c). lastly, in the case of a 12-blade lfsb inclined at angles (10°, 45°, and 70°), the highest temperature obtained is 667.3637 for a 12-blade burner at 70°. the lowest pressure drop obtained in the 4 blades lfsb at angles (10°, 20°, 30°, 40°, 45°, 50°, 60° and 70°) is 4 blades at 20° (13444413.5927 pa). in the 6 blades lfsb, the lowest pressure drop obtained at angles (10°, 20°, 30°, 40°, 45°, 50°, 60° and 70°) is 6 blades at 50° (14479915pa). in the 8 blades lfsb, the lowest pressure drop obtained at angles (10°, 45°and 70°) is 8 blades at 45° (15889589.43pa). in the10 blades lfsb, the lowest pressure drop obtained at angles (10°, 45°and 70°) is 10 blades at 45° (13929679.13pa). in the12 blades lfsb, the lowest pressure drop obtained at angles (10°, 45°and 70°) is 12 blades at 10° (11973219.93pa). conclusion in this study, the variation of the blade angles and number of blades in the liquid fuel swirl burner were used in the parametric study to determine the optimum configuration to obtain the highest temperature at the burner outlet and lowest pressure drop that yields effective combustion. the burner was designed with solidworks and simulated using the ansys fluent package as noted above. a comparison of the simulation results of the burner with 6 blades and the experimental results of the burner with the same number of blades, both at angles 20°, 30°, 40°, 50°, 60° and 70° was performed to obtain the best results in terms of highest temperature. likewise, a comparison study of the varying number of blades (4, 6, 8, 10 and 12) at various angles (10°, 45°, 70°) was done to obtain the best results according to the parametric study. in terms of maximum temperature, the best result obtained is a 10-blade lfsb at 70° with a temperature of (1024.547°c) whilst when optimum pressure drop is preferred the most suitable burner configuration is a 12-blade lfsb at 10 with a pressure drop of (11973219.93pa). references akinwonmi, a. s., adeaga, o. a., & orhadahwe, t. a. (2023). investigation of diesel blends with edible vegetable oils for combustion in swirl burners. 2023 international conference on science, engineering and business for sustainable development goals, sebsdg 2023, 1, 1–9. https://doi.org/10.1109/sebsdg57117.2023.10124515 aliyu, a. s., dada, j. o., & adam, i. k. (2015). current status and future prospects of renewable energy in nigeria. renewable and sustainable energy reviews, 48, 336–346. https://doi.org/10.1016/j.rser.2015.03.098 benim, a. c., deniz canal, c., & boke, y. e. (2022). computational investigation of oxy-combustion of pulverized coal and biomass in a swirl burner. energy, 238, 121852. https://doi.org/10.1016/j. energy.2021.121852 boushaki, t. (2019). swirling flows and flames. intechopen. intechopen. https://doi.org/10.5772/ intechopen.75295 clark, n., & yusoff, k. (2014). combustion and society: a fire-centred history of energy use. in theory, culture & society (vol. 31, issue 5). https://doi. org/10.1177/0263276414536929 day, a. r., ratcliffe, m. s., & shepherd, k. j. (2003). boilers and burners. in heating systems, plant and control. https://doi.org/10.1002/9780470774458.ch2 dhyani, d., & phade, s. (2024). a comprehensive study on the design and computational analysis of an air swirl burner. 03006, 0–11. european parliament. (2023). climate change: the greenhouse gases causing global warming. https://www.europarl. europa.eu/topics/en/article/20230316sto77629/ climate-change-the-greenhouse-gases-causing-globalwarming pa ge 97 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 89-97, 2025 ikein, a. a. (2017). nigeria oil & external exposure : the crude gains and crude pains of crude export dependence economy. the business and management review, 8(3), 396–407. http://www.abrmr.com/ myfile/conference_proceedings/con_pro_65194/ conference_74330.pdf mansouri, z., & boushaki, t. (2018). experimental and numerical investigation of turbulent isothermal and reacting flows in a non-premixed swirl burner. international journal of heat and fluid flow, 72(june), 200–213. https://doi.org/10.1016/j. ijheatfluidflow.2018.06.007 martins, j., & brito, f. p. (2020). alternative fuels for internal combustion engines. energies, 13(15). https:// doi.org/10.3390/en13164086 jaafar, m. n., abdul mookmin, m. s., & samat, i. (2012). emission reduction from burner system by varying swirler blade angle. jurnal teknologi, december 2009. https://doi.org/10.11113/jt.v51.151 ogbonnaya, c., abeykoon, c., damo, u. m., & turan, a. (2019). the current and emerging renewable energy technologies for power generation in nigeria: a review. thermal science and engineering progress, 13(august), 100390. https://doi.org/10.1016/j. tsep.2019.100390 oyewola, o. m., akinwonmi, a. s., ajide, o. o., & salau, t. o. a. (2022). effect of swirl on temperature decay function in straight blade liquid fuel swirl burner. international journal of heat and technology, 40(2), 527– 534. https://doi.org/10.18280/ijht.400220 oyewola, o. m., singh, p. m., akinwonmi, a. s., ajide, o. o., & salau, t. a. o. (2022). thermal characterization of straight and curve edge blade liquid fuel swirl burner. eureka, physics and engineering, 2022(3), 11–19. https://doi.org/10.21303/24614262.2022.002157 panwar, n. l., kaushik, s. c., & kothari, s. (2011). role of renewable energy sources in environmental protection : a review. renewable and sustainable energy reviews, 15(3), 1513–1524. https://doi.org/10.1016/j. rser.2010.11.037 schmittel, p., günther, b., lenze, b., leuckel, w., & bockhorn, h. (2000). turbulent swirling flames: experimental investigation of the flow field and formation of nitrogen oxide. proceedings of the combustion institute, 28(1), 303–309. https://doi. org/10.1016/s0082-0784(00)80224-6 sengupta, a., mukherjee, r., & mishra, v. k. (2021). performance and emission characteristic of burners: a review. aip conference proceedings, 2341(may). https://doi.org/10.1063/5.0051735 sreenivasan, r., koli, s. k., & raghavan, v. (2012). experimental study of effects of coflow air and partial premixing on liquid petroleum gas flames. isrn thermodynamics, 2012, 1–6. https://doi. org/10.5402/2012/202715 surjosatyo, a., & priambodho, y. d. (2011). investigation of gas swirl burner characteristic on biomass gasification system using combustion unit equipmen (cue). jurnal mekanikal, 33(33), 15–31. xiao, y., cao, z., & wang, c. (2018). flame stability limits of premixed low-swirl combustion. advances in mechanical engineering, 10(9), 1–11. https://doi. org/10.1177/1687814018790878 yang, x. study of combustion and thermal performance in a swirling micro combustor under premixed and non-premixed modes, zhao, l., he, z., dong, s., & tan, h. (2019). comparative study of combustion and thermal performance in a swirling micro combustor under premixed and non-premixed modes. applied thermal engineering, 160 (june), 114110. https://doi. org/10.1016/j.applthermaleng.2019.114110. pa ge 1 pa ge 49 american journal of smart technology and solutions (ajsts) ai-powered automation in business operations for the future md zahirul islam1 , prottoy khan2, sazib hossain3* volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.4495 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: february 02, 2025 accepted: march 07, 2025 published: april 11, 2025 ai and rpa technologies has become the latest trends in the business world that have revolutionized the business sectors at an unprecedented pace. even though, nowadays ai technologies apply to manufacturing, logistics, supply chain management industries and others, the extent of their benefits on operational capabilities, decision-making procedures, and organizational performance still does not receive enough empirical research attention. to this end, this paper seeks to fill this gap by explaining how ai and automation, more specifically rpa and cognitive automation are changing business processes. in specific, the study focuses on the general application of ai in increasing productivity and efficiency as well as reducing human error occurrences in industries that consist of automated systems. the study uses random forest regression and classification models to analyze current data from robotic structures to improve production line performance in manufacturing firms. this paper proves that ai automation helps in enhancing all the time prediction processes and also cooperates with the decision-making process by eradicating operations and decreasing the odds in the course of error. thus, the outcomes indicate a need to combine new technologies like blockchain and 5g to strengthen the security component, develop efficient data management, as wel l as real-time analysis – all of which expand ai possibilities. thus, based on the analysis of such trends as cognitive automation, decision making, and maintenance this paper discusses how ai can transform businesses. in addition, it identifies factors affecting implementation in organizations including workforce changes, data issues, and input data that is of poor quality. it also offers a practical set of suggestions for organisations concerning with shifting ai landscape, it also gives consideration of the moral issues and social impacts of ai technology in its discussion. keywords ai ethics, ai in hr, ai in marketing, artificial intelligence, business automation, digital transformation, future of work, robotic process automation (rpa), supply chain optimization 1 school of electrical engineering, china university of mining and technology, xuzhou, jiangsu, china 2 school of artificial intelligence and computer science, nantong university, nantong, jiangsu, china 3 school of business, nanjing university of information science & technology, nanjing, china * corresponding author’s e-mail: esazibhossain@gmail.com introduction automation through the use of artificial intelligence has cropped up as a competitive advantage transformant in business across various fields (hossain & nur, 2024). as markets continue to evolve and the importance of competition grows, automated solutions based on artificial intelligence are getting to be an invaluable technique when it comes to performance enhancement and expansion. in recent times, the incorporation of the different uses of artificial intelligence (ai) and machine learning (ml) has paved the way for the automation of task which it was believed could only be done by experts. this conventional approach to increasing automation is not limited to substituting human effort with machines but even the mere mechanization of simple tasks; here, a new work-force that learns, and responds to changing data input in order to improve the flow of the work process is envisaged. due to the new cognitive functions, ai is now capable of functional areas of activity, recognition, analysis of data, decision-making and even modelling. companies compete in the current context of contemporary business environments where the focus is on the regular progress of business processes rather than the mere refinement of the existing ones. organizations need to devise, evolve, and rationalise sustainable practices that meet the changing demands of the customer base in terms of products and services offered and those proactively search for ways of cutting costs and improving efficiency. according to brynjolfsson and mcafee (2014), ai automation transformed the strategies of myriad businesses making it possible for companies to achieve higher efficiency that was earlier unimaginable and opening new opportunities for companies to expand. ai use in automation benefits an organization through cutting on costs of working by providing data insights in decision making, reducing chances of errors, and increasing response rates. due to this, it becomes easier to make better and faster decision that is more appropriate in fields like manufacturing, logistics, financial service and customer service among others. a very common trend in ai and automation has been recognition employed in robotic process automation and manufacturing automation. it is worth distinguishing automation from artificial intelligence because the automation that was used in the past was quite limited to simple routine works as compared to using artificial intelligence. ai technologies involve the robots and other systems that are capable of automatically processing data which is able to give instruction, learn, modify its behavior and output will increase with the advancement in time. for instance, image recognition rpa can identify trends, identify potential breakdowns, and even control processes pa ge 50 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 49-58, 2025 that will enhance efficiency of the line and minimize losses. such changes are a transition from the defensive model of operations to the aggressive one, in which ai can detect the future issues and offer the solutions on the spot. earlier ai was simply used in manufacturing to automate repetitive functions and perform repetitive manufacturing management activities; now, it is used for achieving real-time changes in production schedules, scm and demand management. it leads to a more flexible and data-oriented decision making process necessary for coping with the ongoing processes in the modern markets (avasarala, 2020). however, ai-driven automation is now being introduced into many other spheres of the business, which to a certain extent are also vital for sustainable business development. ai is making a significant impact on the following areas in supply chain optimization namely demand forecasting, inventory management and logistics. it is now possible to predict when certain products are going to be in high demand and when they would be low in demand, so as to ensure that inventories are well matched with demand. this minimizes overstocking or stockouts, and both circumstances are quite catastrophic as they result in a lot of money being lost. ai solutions are also applied to find the best way for transportation and the supply chain that focuses on minimizing delivery time, ensuring cost efficiencies, and, also, satisfying consumers (choi et al., 2018). in human resources (hr) management field, the use of artificial intelligence (ai) is emerging various ways that enhances recruitment, performance evaluation, and m& d on the employees (hossain et al., 2024). while looking into past records and current data obtained from employee records ai can assist hr departments in orienting the right candidates for the specific jobs, estimate or even forecast employee turnover and suggest the most suitable developmental programs for the human capital. due to this new and modern concept of recruitment through the use of ai, the process of screening and hiring of the employees has now become faster, efficient, and free from biased considering that it involves the usage of data (armenta, 2017). furthermore, ai can be helpful in real working monitoring of the personnel and detect their decreased performance level due to burnout or other factors and suggest the ways to improve the work motivation. with the help of ai, customer service has changed how the companies communicate with their customers. the natural language processing (nlp) based ai integrated chatbots and virtual assistants help in offering real-time answers to the customers’ query and they can also independently deal with the problem without involving the human assistance. these systems also send more complicated matters to the human agents in case they are not well handled adding to the fact that it is always able to assist the customers through to the middle of the night. additionally, owing to the capability of using previous interaction data, ai systems can be able to forecast customer requirements hence addressing them before they become an issue in future improves on customer satisfaction (opoku, 2021). in finance and accounting, the ai work of completing and automating tasks comprises the reporting, analyzing, detecting frauds, and forecasting. this manoeuvre leverages the capability of the ai systems to analyse large financial data sets to determine patterns and outward anomalies that a human being would take a lot of time to observe. for instance, present day artificial intelligence is applied for forecasting cash flow, evaluating risks and identification of frauds in real time. this not only enhances the reliability of financial statements but also help the business to respond to changes in demand and supply which thus help in managing risks and making decisions based on data which are accurate (brynjolfsson & mcafee, 2014). in the same way, powerful tools and technologies have emerged in relation to taxes preparation, audits and other mundane tasks to enhance efficiency in relation to the accountant matter. other areas have also been affected by the marketing department, where artificial intelligence is used in tasks like segmentation, marketing campaign optimization, and even setting of appropriate prices for products. it is through using analytical techniques that artificial intelligence can forecast the course of events hence enable the selling strategies to alter their marketing with the intention of reflecting on the goals and objectives of the customers. this makes it possible to reach the appropriate consumer with a suitable communication that can contribute in enhancing the sales conversions significantly and hence increasing consumers loyalty (chui et al., 2018). the current opportunities of using ai in a number of fields serve as the foundation for the continued progress (nakib et al., 2024). ai’s advancement carries on which signifies even more changes on how industries go around the world. robotic process automation or rpa can basically be described as the procedure of leveraging robots to perform tasks that were hitherto executed by people, for example in data entry or as in invoice processing, report creation. with the integration of the ai system into rpa, the overall capabilities of the systems involved have been enhanced in terms of decision-making capabilities as they entail cognitive working abilities like pattern recognition, decision recommendation, and even general optimization. appendix 1: rationale of integrating artificial intelligence into rpa the use of ai in advancing rpa brings many benefits to the business productivity, accuracy, and effectiveness (willcocks, van der meer, & reilly, 2015). the purpose of this paper is to understand the impacts of such advancements in ai automation on different business fields and what more people should expect in the future concerning business automation. for the purpose of giving the reader a clear perspective on how these key areas are correlated, the following research framework model maps out the layout of the study as well as the aspects that the research is going to focus on (figure 1). pa ge 51 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 49-58, 2025 the paper will use data analysis techniques, case studies and examination of the developing of ai to enhance the understanding of the business benefits like cost savings, productivity, and growth capabilities ai offers. at the same time it will outline some of the issues associated with the adoption of ai including how it can integrate with existing systems, data privacy and job losses. this paper is relevant to the current business world since it provides information on how companies can implement ai, minimize errors, and advance in the market. in addition, it will give a guide of the practical steps that can be taken to improve ecosystem development so that organizations embarking on an ai journey are ready for the long term and the ability to use ai for sustained growth and innovation. hence, by dissecting the effects of automation arising from the use of artificial intelligence the paper seeks to guide the decision-makers on the future tendencies of artificial intelligence in business activities in order to avoid future vices and embrace the virtues that would shape the future business environment. literature review ai in business processes integration has gone through a revolution in the recent past where organizations have expanded on the use of ai in activities like automation of processes, decision-support, among others. ml and rpa have enhanced workplace efficiency by empowering systems to learn from the data and make decisions as well as perform tasks that were traditionally done manually. in this part of the paper, we present a brief history of the emergence of ai, its application in automating business processes, interaction with rpa, and the applicability in various industries is presented. the section also discusses the technology that has contributed to the progression of the ai future and previous works done in studying the impact of ai in business processes. ai in business ai technologies have considerably grown from early rule-driven system to what can be referred to as learning machines that can perform a variety of tasks in the respective fields. firstly, ai was only for specific tasks like a chatbot used in customer support or data entry clerk applications. nevertheless, with the development of machine learning and deep learning approaches, an ai system can perform most of the tasks that are based on cognitive abilities and include problem-solving, pattern recognition, decision-making, and predictive analysis. at the present time, ai found its application in various areas of the business: finance, marketing and sales, human resources, supply chain management, and customer service. the most applicable reason that has made ai to feature heavily in modern businesses is the capability to evaluate massive data, derive patterns, and make recommendations within a short span of time. for example, in finance, ai allows for the analysis of big data to identify the existence of fraud as well as improve the identification of best portfolios and automation on financial reporting. in marketing, they help in such things like personalization of customers and their experience, targeted advertising, and more accurate customer prediction for segmentation. furthermore, ai can better optimize the organizational system, especially in enhancing the productivity sector of the business without proportionate growth in the labor force, making it a tool of choice for any organization desiring to enhance revenue gains. according to brynjolfsson and mcafee (2014), ai augments how organizations operate in a business environment and creates a core competency. by integrating ai, organisations now can work at a faster pace, sense changes in the market and meet customers’ needs.. these are clear signs that indicate that this evolution will progress even more in the following years, and ai assumes an even more turnaround role in the business model. these capabilities allow businesses to increase the level of operational activity, become more effective, and reveal new sources of competitive advantage in a constantly evolving market environment. figure 1: research framework model pa ge 52 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 49-58, 2025 robotic process automation (rpa) while the use of ai in business is based upon integrating ai technologies to support the processes of decision making, the use of rpa goes further and applies ai technologies for performing the operations taking the human operators out of the loop. rpa has been counted as one of the pioneering usages of ai in business processes – mainly in banking, insurance, and telecommunications industries. earlier, the applications of rpa embraced only the repetitive and routine tasks like data entry, invoices processing, and transaction handling. these were repetitive work that were done previously in a manual, labor-intensive, and inefficient manner which can be automated. however, the incorporation of ai in rpa has enabled these systems to address more sophisticated processes that need of cognition abilities, including decision-making, data interpretation, and data interactions such as emails, invoices, among others. when rpa is integrated with nlp or machine learning the decision-making capabilities were previously performed manually are implemented automatically. this is integration commonly known as ia or intelligent automation; essentially, rpa joined with ai to form capability of doing more highlighting aspects of rpa where repetitive task can be handled by rp while ai which has features of pattern detection and data analysis controls the more complicated tasks. avasarala (2020) explains how the current advancements in technological manufacturing involves ai in the manufacturing of rpa systems as a way of streamlining the production lines, minimizing on the time that machines are out of service, and increasing on the rate of production. for instance, in manufacturing-line industries, the use of automated robots that are built with ai capabilities can study the data fed to it and detect when the machinery is likely to fail and then proceed to rectify the situation by altering its manufacturing process or order for new parts if required. such sophisticated rpa systems are beneficial for any business because those elements are becoming the key to success in the current rapid markets. through the use of ai together with rpa, companies get to enhance the efficiency of multiple aspects of making decisions, wherein human beings will have a chance to direct their efforts towards more profitable aspects such as invention. technological advancements these are not mere ideas on the walls but a reality that has been thoroughly experimented with by various scholars and researchers in their search for the influence of these technologies. one the most important of these is the support of machine learning (ml) algorithms, through which systems can make constant progressive changes to their performance based on prior performances. another type of ml called deep learning has also emerged because of its capability to work with new and large datasets that are in the form of images or speech or texts that are unstructured in nature. in business operations, a number of artificial intelligence technologies like robotic process automation, predictive analysis and decision support system have brought a significant change of paradigm shift in the overall decision-making processes. for instance, ai-driven rpa has moved further than simple analyses and process automation to contributing to smarter actions such as the assessment of information, provision of suggestions, and communication with customers. this is evident from the case of scm, hr, finance, and marketing that have all benefitted through the integration of the ai tools. another advancement is the arrival of cloud ai platforms which have enabled more organizations to implement ai-based solutions since the services are subscription based thus not limiting the organizations who want to adopt based on affordability. the major potential of cloud computing is that businesses can rely on multiple services that allow them to store, process and use ai models and tools without acquiring costly equipment. this has made it easy for many sectors to integrate the use of artificial intelligence to their operations, especially smes who never had this advantage before. according to current trends in the development of ai and related technologies, businesses will rely even more on the use of elements of machine learning to develop increased automation of business processes, meaning the movement towards future smarter and more closed business environments will continue. impact across sectors it can be stated that the application of automation means based on artificial intelligence in the current business environment is becoming increasingly popular in different spheres and that contributes to enhancing results and improving many aspects of its decision-making processes. in manufacturing, it uses artificial intelligence in the enhancement of such aspects like the production line, time needed for repairs, and quality of the product through the strategy of coming up with a predictive maintenance and online solutions. other areas where ai is beneficial for the business include inventory handling, demand estimation, and supply chain management through predictive analysis of big data. it is used in the diagnosis of diseases, analysis of medical data and in proffering treatment to patients, thus improving efficiency and speed in the delivery of healthcare services. ai can be used to provide an accurate prescription by forecasting the potential of different compounds and it also supports the discovery of medicine. in finance, ai performs fraud detection, credit scoring, risk management, algorithm trading and even in the prediction of the financial aspects effecting trading as well. real-time data also aid in decision-making depending on the recommendation by the ai to the financial institutions involved. in retail business and its digital counterpart, e-tailing, the application of ai fulfills the ingredient personalization by targeting appropriate inventories, demand, and correct prices. automation of customers through chatbots and the recommendation system will enhance the experience pa ge 53 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 49-58, 2025 of customers. in customer service, chatbots are applied to answer the frequently asked questions and handle the complaints elegantly and systematically improving consumer satisfaction while using feedback analysis. as for the human resources, ai is increasingly being applied in recruitment as a process of sorting and analyzing cvs, candidates identification and performance tracking. it also enhances employee development through providing training to the employees, increasing engagement, and staff retention. in this case, ai has had an influence on industries with tremendous growth and actual cost savings as well as proper management of business processes. future implications consequently, the future prospects of automation based on artificial intelligence present profound changes in business, industries, and the society. small and mediumsized enterprises (smes) will be able to level playing field by adopting this artificial intelligence tools in education, agriculture and transportation, and so on sectors. these evolutions of cognitive automation will lead to advanced levels of sophisticated facilitation of ai in qualities like the strategic decision making and abstract problem solving within cognitive zones for certain systems and industries while other forms of autonomous systems like self driven cars will disrupt industries by replacing man power and optimizing productivity. it will also improve decision making since ai can process large amounts of data at once and offer data support in decision-making while achieving the high-level strategic goals automatically. however, with the increased use of ai in the workforce, job removal is inevitable, but new positions may be created in managing and designing ai applications, data and automation, meaning that firms may have to train their existing workers in new skills. some of the ethical issues falling under data privacy, data bias, and data transparency are that businesses will have to put up guidelines into achieving ai ethical objectives and even put into place protections for consumers and corporate employees. it shall also have a worldwide impact towards business with bringing efficiency to supply chains, trades, and partnership, as well as help new businesses in emerging markets to skip technologies seen in developed economies. to effectively place this as a solution and an opportunity, ai combined with blockchain will help advance business in a way that different industries, especially finance, healthcare, and supply change will benefit. the advancement is in consistantly progressing annually and it is predicated on the capability of advanced automation in enhancing innovation, global operational expertize and changing trends in various industries. many studies have also been conducted to establish how advances in artificial intelligence technologies are likely to affect different industries. the article of chui et al. (2018) detailed effects of ai in businesses; one of which is automation which optimizes business functions and allows organizations to expand their business without necessarily hiring new employees. this proved that businesses using ai can enhance customer relations, lower the costs of operations and make better decisions as the existence of the ai enables faster and more accurate results than conventional methods. westerman et al. (2011) was a work that has tried to address technology in health care, whereby health care systems that involve artificial intelligence in diagnosing diseases, analyzing patients’ data, and even suggest treatment. the study revealed that, through analyzing large datasets and being able to make decisions in real time, the overall patient care and satisfaction, in conjunction with cutting health care costs, had been impacted positively by the use of ai. similarly, choi et al. (2018) pointed out that ai is becoming more prevalent in becoming an essential element of supply chain management as machine learning algorithms to predict the changes in demand, inventory control, and logistics solutions. other studies, like by huang & rust (2021), which have noted that with the help of an ai agent and chatbots as well as virtual assistants, customer contacts have been removed due to their fast response and personal approach. these changes are not only beneficial to the customers but also assist the business in decreasing operational cost through the use of the ai tools in handling support operations. according to armenta (2019), there are several adopted ai applications in the hr area, such as recruitment, performance evaluation, and talent management to summarise, ai in the hr area can help in hiring processes and employee performance management to predict performance and match human traits from the big data collected. materials and methods data collection the dataset is collected at the national institute of standards and technology (nist) where the two robot workcell is employed in a manufacturing setting and the data includes process and robot performance details. a 6-dof for material handling (robot 1) and for precise operations a 6-dof secondary robot (robot 2) is used as part of the workcell. these data include joint positions from j1_qactual to j6_qactual and joint velocities from j1_ qdactual to j6_qdactual and these are the movements of the different robots recorded at different time instances. besides, both plctime and robottime are used for timestamps of events and the corresponding events of the manufacturing process need to be synchronized with robotic and process level events. the toolx or tooly or toolz is the data to locate the robot’s operation tool in workspace which provides control to know the precise working degree. other process data that are captured is task information where one of them is the part assigned to perform a particular task, another is when a part is added or removed from work cell, or when a robot begins and completes a task. this dataset was gathered at nominal conditions and the robotic system had not been impacted during the measurement process; moreover, in addition to the robot level performance metric, this dataset also contains process level measurements which pa ge 54 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 49-58, 2025 give an idea about the effectiveness of the robotic tasks in manufacturing environment. data preprocessing and cleaning first, basic data preparation will be done on the incoming data: for the data collected on the plc and the robot the timestamps; namely, plctime and robottime will be resynchronized to correct for any misalignment in the internal clocks of the two plc and robot controllers. it is conventional that missing values in robot joint positions or tool data will be linear interpolated if the gap duration is small or omitted if gap duration is large. further, position data comprising joint position and velocity will be normalized using minmaxscaler by scaling joint position data between 0 and 1 to make the data from joint position and velocity of the robot homogenous. this will enable the data to be cleaned, formatted and well preprocessed to fit for the next stages of data analysis. machine learning models supervised learning regression models its purpose is to forecast task time till the end of movement based only on the position and velocity of the navigating robot and other characteristics of the task. for this relationship, it will be appropriate to use multiple linear regression model. the model takes into account the features of a robot where the dependent variable is the time taken to complete a task, and the independent variable is the joints’ angle positions of a robot, though other features such as speed or process parameters may be incorporated in the model if need arises. completion time prediction=β0+β1×j1_qactual +β2×j2_ qactual +.........+β6×j6_qactual + ϵ where: j1_q actual,j2 _qactual ,…,j6 _qactual represent the joint positions of the robot. β0,β1,…,β6 are the coefficients to be determined during the model training phase. ϵ is the error term. the evaluation models are intended to categorize them by the process performance in terms of operational characteristics as being successful or failed. logistic regression or decision trees will be used to classify the tasks with the help of certain characteristics like movements of the robot joints, the states of the tasks, and position of the tools. the logistic regression model will give the probability of accomplishing the task using the following formula: where: p (task success) is the probability of task success. β0,β1,… are the logistic regression model coefficients. the equation uses the sigmoid function to model task success probability. unsupervised learning clustering and dimensionality reduction clustering for clustering model the purpose is proposed to clusterize the tasks in relation to observed movements and times of the robots. to assess the quality of the obtained results i am going to apply the k-means clustering in order to group the tasks with similar efficiency. it will aid in categorizing the tasks based on the observed behaviors and will make improvements to such processes by addressing like types of tasks. dimensionality reduction (pca) to reduce the complexity of the data, principal component analysis (pca) will be used to identify the most important features affecting robot performance. pca will reduce the number of features while retaining the key information that explains the largest variance in the dataset. pca equation x = w.y where x is the original data matrix (robot movements and task completion time). w is the matrix of eigenvectors (principal components). y is the transformed data matrix (reduced dimensions). ai optimization reinforcement learning the purpose concerns the problem of task scheduling to maximize the throughput obtained from the robots and minimize the robotics idle time. rl will be used for training an agent that will be able to identify the best actions for a robot based on performance of a particular task. the agent will be trained to vary the tasks and robot motions with the purpose of reducing the entire time cycle and improve the procedural performance. performance evaluation the assessment of the robot performance shall be in terms of operation efficiency where efficiency metrics like time taken to complete a task, the time taken to repeat the same task, time lost in breakdowns and the success rate of the robots efficiency. in order to measure the effectiveness of using ai optimization, performance before using any ai technology will be compared with the performance after applying ai optimization algorithms. this will aid in comparisons on the trends of efficiency, throughput, as well as the rates of successful completion of tasks in an endeavor to understand ways in which the integration of ai increases human-like performance in robots. results and discussions some of the general fields featured in the dataset are time, plctime, robottime and additionally 6 joint positions that refer to the coordinates of the actual movements of the robot at the particular point of the manufacturing process (j1_qactual to j6_qactual) figure 1, figure 2 joint velocities also appear (j1_qdactual to j6_ pa ge 55 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 49-58, 2025 qdactual), as well as toolx, tooly, toolz, representing the positional coordinates of the robot tool in the workspace; the toolz data is missing here. in the first step to data analysis, the joint position values are observed to oscillate in the positive and the negative due to rhythmicity in the movement of the robotic arm joints. joint velocities shown in fig. 5 show rather small deviations around zero, which confirm that motion is stable, while toolx and tooly describe dynamic trajectories over the workspace. lack of additional information in toolz might decrease the level of precise analysis, however, the overal presentations offer insights about the robots’ motion and the position of the tool in relation to the time required for completing the tasks and the results. figure 2: robot joint positions over time figure 3: robot tool positions over time a mean absolute error of 6.736458e+08 is achieved by the random forest reggressor which is better than achieving through earlier used linear regression. the graph depicting the actual against predicted task time successfully oriented with actual time in the x-axis and the predicted time in the y-axis boundary and the regression line with marginal variation from the actual line illustrated in the plot presented in figure 3. however, even after the application of this model, there are some limitations and scope for enhancing the model by fine tuning the model further and finding more suitable features. as a rule, the closer the dots are to the red dotted line, the better the model performance in predicting the values from the second array. the analysis of the random forest classifier for the target task demonstrates the high accuracy of predicting the success of the given task, as well as high precision by achieving values of recall and f1-scores for the “failed” pa ge 56 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 49-58, 2025 and “successful” classes. from the confusion matrix, there is no false positive and false negative results, thus the classification is clear (figure 5). heat map of the confusion matrix also supports the fact that the classifier has rightly separated the successful and the failed tasks while analyzing that the class selected by the model corresponds to the actual result of the task. figure 4: robot tool positions over time figure 5: confusion matrix for random forest classifier: task success prediction the findings of the analysis also show a good evaluation of such models across the various models employed in this study. the results given by the regression model were found to have an rmse of 0.00091, which indicates very accurate estimates of task completion time. logistic regression model yielded 99.36% accuracy and every time it gave only 31 wrong results when the task was successful and it said no to just 4 successful task, proving itself right in the prediction. according to the k-means clustering model, there are three distinct tasks related to robot movements and its behaviors during the completion of tasks. the current trend specified that most of the tasks were categorized under the cluster 1 and the second one being the cluster 0 while a few came under the cluster 2. these results point out the capacity of the models to sort task behaviors properly as well as capable of estimating the time and the performance rate of the tasks to be completed. pca was useful to reduce the robot movement data which contains proportional joint positions and velocities to two dimensions while keeping vital data. an explanation of the results using color labeling of the interaction depicted the relationship between the success of tasks and a shape of robot movements, where two primary components selected were indicative of the greatest variation in the data, thus pointing at key aspects in the accomplishment pa ge 57 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 49-58, 2025 of tasks. effective clustering of the task into groups was done by k-means by using the movements of the robot along with the time taken to complete the tasks as features and the aim was to try consider to identify the patterns in task performance and the robot activity pattern. random forest regressor algorithm received an mae of 6.74e+08 which is slightly better compared to linear regression but there is a scope of further enhancement in the model. this could be further substantiated when observing the scatter plot which depicted the real and predicted task completion time; the points were positioned closely to the red dashed line, which validates accurate predictive ability of the model. last but not the least, the proposed model of random forest classifier has given 100 percent accuracy in terms of prediction of the task success as reflected for its accuracy measure from the confusion matrix with no false positive as well as false negative values. figure 6: pca of robot movement data conclusions the findings of this study suggest that ai-driven automation has the potential to redefine how businesses operate, offering significant improvements in efficiency, decision-making, and overall performance. however, the broader implications of ai automation extend far beyond efficiency and cost-cutting. as organizations adopt ai technologies, they must also consider the societal and workforce transformations that accompany these advancements. ai is poised to disrupt traditional business models, and its integration into business operations is likely to result in both positive and negative outcomes. while ai-driven automation offers significant benefits, such as enhanced operational efficiency and the ability to perform complex tasks with minimal human intervention, it also introduces challenges that cannot be ignored. the widespread adoption of ai is likely to lead to workforce displacement, as traditional roles are replaced by intelligent systems capable of performing repetitive and cognitively demanding tasks. in this context, organizations must proactively address workforce transitions by reskilling and upskilling their employees to take on more strategic, creative, and decision-making roles that complement ai systems. furthermore, ethical concerns surrounding ai, such as data privacy, algorithmic biases, and transparency, must be carefully considered. businesses must implement ethical guidelines to ensure that ai technologies are used responsibly, with a focus on minimizing negative societal impacts. in addition to the workforce implications, the societal impact of ai automation is another crucial aspect that requires attention. as ai continues to transform industries, businesses must ensure that their adoption of ai technologies benefits not only their internal operations but also contributes to positive societal change. this includes ensuring that ai technologies are deployed in ways that promote fairness, inclusivity, and sustainability. the integration of ai into business operations can also raise concerns about data privacy, security, and the ethical use of customer data, which businesses must address through transparent policies and governance frameworks. the paper concludes by offering strategic recommendations for businesses to navigate these challenges and maximize the potential of ai. future research should focus on exploring the long-term societal impacts of ai automation, particularly in areas such as employment, privacy, and the ethical deployment of ai in business settings. additionally, as ai technologies continue to evolve, there is a need for ongoing research into best practices for integrating ai into various sectors, ensuring that businesses can remain competitive while fostering responsible innovation. ultimately, ai automation holds the promise of transforming business operations, but its adoption must be carefully managed to ensure that it delivers long-term value to both organizations and society as a whole. pa ge 58 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 49-58, 2025 references armenta, s. (2017). artificial intelligence and human resources: applications, opportunities, and challenges. journal of business research, 15(3), 243-252. avasarala, v. (2020). robotic process automation: the next generation workforce. springer. avasarala, v. (2020). robotic process automation: the next generation workforce. springer. brynjolfsson, e., & mcafee, a. (2014). the second machine age: work, progress, and prosperity in a time of brilliant technologies. ww norton & company. choi, t. y., wallace, s. w., & wang, y. (2018). supply chain analytics: a framework for understanding business operations. journal of business logistics, 39(1), 1-16. chui, m., manyika, j., miremadi, m., henke, n., chung, r., nel, p., & malhotra, s. (2018). notes from the ai frontier: insights from hundreds of use cases. mckinsey global institute, 2(267), 1-31. hossain, s., & nur, t. i. (2024). gear up for safety: investing in a new automotive future in china. finance & accounting research journal, 6(5), 731-746. hossain, s., akon, t., & hena, h. (2024). do creative companies pay higher wages? micro-level evidence from bangladesh. finance & accounting research journal, 6(10), 1724-1745. huang, m. h., & rust, r. t. (2022). a framework for collaborative artificial intelligence in marketing. journal of retailing, 98(2), 209-223. huang, m. h., & rust, r. t. (2022). a framework for collaborative artificial intelligence in marketing. journal of retailing, 98(2), 209-223. nakib, a. m., khan, p., ullah, m. m., kawser, m. l., jayed, a. k. m., & zim, s. k. (2024). harnessing advanced nlp techniques for automated personality analysis and future behavior prediction from social media posts. eng. technol, 4(4), 98-106. nakib, a. m., li, y., & luo, y. (2024, september). retinopathy identification in oct images with a semi-supervised learning approach via complementary expert pooling and expert-wise batch normalization. in 2024 9th optoelectronics global conference (ogc) (pp. 170-174). ieee. westerman, d., spence, p. r., & van der heide, b. (2012). a social network as information: the effect of system generated reports of connectedness on credibility on twitter. computers in human behavior, 28(1), 199-206. westerman, g., calméjane, c., ferraris, p., & bonnet, d. (2011). the 2011 digital transformation study: a deep dive into digital strategies. mit sloan management review. willcocks, l. p., lacity, m. c., & craig, a. (2015). robotic process automation: the next transformation in shared services. journal of information technology teaching, 1(2), 89-98. willcocks, l. p., lacity, m. c., & craig, a. (2015). robotic process automation: the next transformation in shared services. journal of information technology teaching, 1(2), 89-98. pa ge 1 pa ge 1 american journal of smart technology and solutions (ajsts) role of autonomous systems in overcoming maritime communication bottlenecks through quantum computing and 6g technologies tunde olamide ogundare1, abraham peter anyebe2, folami ola-oluwa alao3, idoko peter idoko4*, idoko innocent odeh5 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.4208 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: december 12, 2024 accepted: january 14, 2025 published: july 03, 2025 marine communication systems have been grappling with obstacles for a while now. issues like restricted weak connectivity, are common problems in far-off oceanic areas where delays are high, and reliability is a concern too. these challenges hinder the functioning of contemporary marine systems that heavily depend on swift and continuous data transfer for instant decision-making and navigation accuracy. cutting-edge innovations, like quantum computing and the generation of wireless networks, offer hope to transform marine communication by offering sophisticated solutions to these barriers. this analysis delves into how autonomous systems are used in activities and looks into the ways quantum computing and 6th generation technologies can help alleviate communication challenges in that field. quantum computing brings data processing and encryption features, to the table while 6th generation technology provides low latency and high speed connectivity crucial for seamless real time operations of autonomous systems. the document emphasizes the potential between these technologies and talks about how they could lead to an effective and safe communication network, for maritime purposes. moreover talking about the difficulties and how to blend quantum and 6th generation technologies, in settings has been emphasized in the review as well as discussing future research paths that could be taken in this area of study. by tackling these challenges it is believed that the maritime sector could make strides forward in enhancing the performance of autonomous systems leading to safer and more effective worldwide maritime activities. keywords autonomous systems, bottlenecks, maritime communication, quantum computing, 6g technologies 1 department of nautical science, liver john moores university, united kingdom 2 department of navigation and direction, nigerian navy naval unit, abuja, nigeria 3 school of social sciences, university of kwazulu-natal, durban, south africa 4 department of electrical/electronic engineering, college of technology, university of ibadan, nigeria 5 professional services department layer3 ltd, gwani street, wuse zone 4, abuja, nigeria * corresponding author’s e-mail: aidoko4j@gmail.com introduction background of maritime communication challenges communication at sea has always faced challenges like delays and low connection strength due to the vastness of the ocean and unpredictable weather conditions, quite different from communication on land which uses satellites and radio waves to transmit data smoothly but struggles with consistent signals over long distances or during rough weather periods as signals can be delayed by as much, as half a second and might even drop up to 40 percent in severe storms or when atmospheric disturbances occur. the constraints pose a challenge for self-operating systems since data transfer is vital for their navigation and decision-making functions. figure 1 shows a network of the maritime internet of things (iomt) connecting drones (unmanned aerial vehicles) surface vessels, like fishing boats and cargo ships as well as underwater vehicles (unmanned underwater vehicles. uuvs). the information gathered is sent across platforms to a shore bs data fusion” unit, on the ground for analysis and processing. different types of communication are displayed using links. green lines indicate participants, in aircomp transferring model weights. red lines show non participants transmitting model weights. blue lines represent connections sending data to surface vessels (usvs). this network links technologies to consolidate data effectively and monitor activities. one of the causes of these traffic jams is the dependency on fashioned networking systems that are not designed for long-distance communication over the seas or oceans. delay tolerant networks (dtns) have been identified as a solution to tackle these challenges. research has shown that dtns have the potential to significantly enhance the delivery rates of data packets from one end to another by as 35% in marine environments when compared to traditional routing methods. this advancement is especially valuable, for self-systems that rely on smooth and dependable communication links to carry out tasks across vast distances. network inefficiencies and the challenges posed by the environment like signal degradation, from paths and interference from other vessels and communication systems are obstacles to current communication technologies performance and emphasize the importance of advanced solutions such, as 6th generation networks and quantum computing to tackle these limitations. pa ge 2 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 emerging technologies in maritime communication the incorporation of cutting-edge technologies, like quantum computing and 6th generation connectivity is poised to transform communication in the realm of autonomous operations quantum computing offers a unique capability to handle extensive data sets concurrently and represents a notable advancement compared to traditional computing methods when it comes to handling the substantial data loads produced by maritime systems. quantum algorithms have the ability to carry out calculations at faster speeds, than traditional methods allow for real time processing of navigation information in remote regions where conventional techniques face challenges in keeping up with demands such, as weather forecasts and communication connections. figure 2 illustrates a network of communication that incorporates deep sea operations along, with shore and land based systems working together seamlessly. in the sea region both autonomous vessels and traditional ships are connected to a balloon and an unmanned aerial vehicle (uav) facilitating data transmission efficiently. moving closer to the shore vessels like odas buoys interact with a uav. establish connections to shore stations through designated green and red communication channels. on land a shore station along, with a uav station and other essential infrastructure collaborate to manage and relay data effectively. the picture shows the collaborative effort of surveillance systems, with sea and land based counterparts to efficiently oversee and regulate areas. figure 1: multi-layer internet of maritime things (iomt) network for integrated data fusion and communication (höyhtyä et al., 2017) figure 2: integrated maritime monitoring network for deep-sea and near-shore communication (huo et al., 2020) pa ge 3 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 quantum computing and the advancement of 6th generation technology are set to revolutionize communication networks, with lightning speeds and minimal delays. surpassing 5th generation technology by up to 100 times and achieving latency low as one millisecond. this enhanced communication capacity is vital for systems as it enables instant decision making in real time scenarios and seamless data transfer, for system synchronization. ensuring communication, between vessels and control centers or other autonomous units across vast oceanic distances, through the real time transmission of high resolution sensor data is crucial to avoid any delays or loss of data. these advancements also bring about enhancements, in network security and privacy measures. quantum computing plays a role in creating encryption protocols that safeguard communication between autonomous ships and control systems from cyber threats. the integration of quantum technologies, with 6th generation infrastructure is anticipated to improve security systems when managing critical navigation and operational data within worldwide maritime networks. table 1 presents a range of technologies and their functions, in communication systems and autonomous operations along with examples. it showcases quantum computing for navigation and data processing needs; 6th generation networks for efficient data transfer within marine settings; as well as the integration of quantum and 6th generation technologies to enhance the performance of autonomous systems. moreover, it underscores the importance of security upgrades to safeguard communication, through encryption techniques and real time applications for ensuring dependable data exchange during unmanned vessel activities. every tech advancement plays a role, in enhancing the efficiency and safety of operations while also fostering better communication among systems, on the sea. objective and scope of the review the main goal of this review is to investigate how autonomous systems can help tackle communication difficulties in settings by combining quantum computing with 6g technologies. maritime sectors encounter hurdles such, as communication lags and bandwidth constraints in oceanic areas where land based communication networks are absent or inconsistent. this study aims to explore the potential of technologies, like quantum computing and 6th generation networks in tackling these challenges through processing power reduced delays in communication and stronger security measures, in protocols. the scope of this review encompasses three key technological areas: autonomous systems the role of unmanned maritime vehicles (e.g., autonomous ships, drones) in maritime operations and how they rely on robust communication networks for navigation, control, and data exchange. quantum computing the potential of quantum technologies to transform data processing, encryption, and optimization in communication systems, specifically focusing on their applications in maritime environments. 6g networks the next generation of wireless communication networks, which promise ultra-low latency, higher bandwidth, and improved security, and their role in supporting autonomous systems over vast oceanic distances. this review aims to assess how these technologies can work together to resolve long-standing communication issues in maritime settings, ensuring more efficient and secure operations. autonomous systems in maritime environments definition and scope of autonomous maritime systems autonomous maritime systems (ams) which encompass a range of vessels, like autonomous surface vehicles (asvs) and autonomous underwater vehicles (auvs) are created to carry out tasks without direct human involvement using advanced sensors and technology such, as artificial intelligence (ai). these systems can navigate autonomously in challenging conditions to gather information and execute missions efficiently with the aim of boosting efficiency and safety while minimizing human errors in maritime tasks. figure 3 showcases the vision of creating a self-sustaining environment by showcasing the role of digital and smart technologies, in maritime activities. it stresses that 96 percent of incidents result from human mistakes and proposes that automation can greatly improve safety standards.the visual representation highlights that only a small portion (3%) of data is utilized effectively for tasks stressing the importance of optimizing data application in making informed decisions. when advanced digital technologies are utilized onboard ships for making decisions, on navigation control and logistics management without intervention when it comes to port operations as well as routes taken by vessels at sea is enabled too. the report also highlights the impact of labor costs on container ship expenses accounting for a 44% underscoring the necessity for automation to enhance operational efficiency and cut down overall expenses. initiatives such as ges seastream insight demonstrate the potential for reducing costs by up to 20% paving the way towards achieving maritime operations and presenting promising prospects, for the industrys future. pa ge 4 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 autonomous maritime systems have a scope of uses ranging from transporting cargo to monitoring the environment and conducting operations at sea or, in busy shipping lanes where they can make decisions independently thanks, to advanced support systems. rødseth and colleagues (2022), along with idoko and others in 2024 proposed a categorization framework for self navigating ships that differentiates between “ autonomy “ encompassed by ships functioning without human intervention and “limited autonomy,” where human supervision is still required to some extent.this system of classification aids in grasping the varying degrees of automation within setups. from remote controlled functions, to advanced self-governing and smart vessels. the advantages of automated marine systems (ams) are clear, in their capacity to work continuously without breaks which lowers expenses and prevents worker exhaustion effectively. the added advantage is their capability to adjust to data and adjust to different circumstances based in real time. the use of ams has already shown a decrease of 15 percent in fuel usage thanks to the navigation paths and research predicts that embracing shipping worldwide could cut down maritime operational expenses by as much, as 20 percent by the year 2035. autonomous systems play a role, in ensuring safety by reducing the chances of human errors that are responsible for almost 75% of marine accidents according to martelli and colleagues in a recent study (2022). they highlight that upcoming maritime traffic management systems will heavily depend on the coordination between vessels and ai powered navigation aids along, with control facilities to enhance safety by averting collisions and optimizing traffic flow. table 1 offers a summary of autonomous maritime systems (ams) detailing what they are and their main characteristics, like applications and advantages. ams refer to systems that combine ships such as autonomous surface vehicles (asvs) and autonomous underwater vehicles (auvs) to carry out duties without direct human intervention. they utilize sensors along with ai and machine learning technologies. the applications of these systems range, from transporting goods to monitoring the environment and supporting activities. the chart also differentiates autonomy levels; ranging from autonomy, without human involvement to limited autonomy that necessitates human supervision oversight. notable advantages comprise heightened efficiency leading to cost reduction and enhanced safety measures when autonomous maritime systems minimize errors, by humans and optimize fuel usage while backing forthcoming traffic control systems. figure 3: advancing towards an autonomous marine ecosystem: key insights and benefits (offshore source 2024) table 1: overview of autonomous maritime systems: definition, scope, applications, and benefits aspect description key features applications benefits d efi ni tio n autonomous maritime systems (ams) integrate autonomous vessels such as asvs and auvs to perform tasks without human intervention. advanced sensors, ai, machine learning, autonomous navigation, data collection. cargo transportation, environmental monitoring, naval operations. reduces human errors, improves safety and operational efficiency. pa ge 5 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 sc op e ams cover a broad range of maritime operations, from cargo transport to complex missions in unpredictable environments. fully autonomous and constrained autonomy classifications (rødseth et al., 2022). operating independently in high seas, navigating busy shipping lanes. enhances safety, enables independent operation even in challenging conditions. a ut on om y c la ss ifi ca tio ns "full autonomy" (no human input) vs. "constrained autonomy" (some human oversight necessary). levels of automation from remotecontrolled to fully autonomous systems. useful in various operational contexts, from basic control to complex autonomous decision-making. helps in determining the level of automation needed based on the mission. o pe ra tio na l e ffi ci en cy ams operate continuously, reducing fuel consumption and costs. real-time data response, optimized navigation routes, 15% reduction in fuel use. global adoption could reduce maritime operational costs by 20% by 2035. cuts operational costs, reduces human fatigue, increases productivity. sa fe ty c on tri bu tio n ams enhance maritime safety by minimizing human error, which accounts for 75% of marine accidents. ai-driven navigation, shore-based control centers, future traffic management reliant on autonomous ships (martelli et al.). collision prevention, improved traffic flow, safer navigation systems. reduces accidents, improves traffic management, and ensures safer navigation and operations. communication requirements for autonomous systems autonomous marine systems (ams), like autonomous surface vehicles (asvs) and autonomous underwater vehicles (auvs) depend on dependable communication networks for navigation control and sharing of information smoothly and effectively in real time situations the key necessity for these systems is to establish communication with control centers as well as other ships to guarantee operational safety and seamless coordination in the isolated settings of marine environments satellite connected communication systems along with advancing 5th generation technologies are crucial, in assisting these independent systems. in figure 4 displayed is the diagram of a self driving navigation setup that showcases the parts and how they work together for charting paths and steering clear of obstacles. it kicks off with data gathered from sensors, like cameras gps devices, imus, encoders and network links used for self positioning. the perception unit analyzes sensor data to grasp the surroundings and sends details to the planner that charts out an overarching path. the waypoint position selection unit fine tunes this path to guide the planner. the obstacle detection functions, in coordination with these elements by spotting obstacles in time and then adjusting the route to avoid collisions accordingly. the local planner and collision avoidance systems feed information to the motion control unit which carries out the required movements to steer the self driving vehicle or robot towards its destination, for secure navigation. figure 4: autonomous navigation system flowchart for path planning and obstacle avoidance ensuring dependable and affordable satellite communication services to enable data exchange, between ships and shore based centers stands as a significant hurdle in the maritime industry today essential for safety during operations and quick decision making during emergencies according to ait allal et al., 2020 suggests pa ge 6 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 that the integration of 5g technology is gaining traction as a viable solution to enhance communication networks for traditional and autonomous vessels alike offering rapid data transmission speeds and minimal delays crucial for real time tasks, at sea. moreover，kang and park (2019) discuss the importance of developing communication protocols for settings that meet unique needs．they suggest technologies, like s 2 x， in function to the v 2 x system in vehicle communication， as a means of aiding ships． however， innovation in maritime communication has not kept pace with progress, in land vehicle systems despite acknowledging the value of these technologies． figure 5: key communication components for autonomous maritime systems additionally, it poses difficulties to maintain a connection, during harsh weather conditions and across vast distances namgung (2023); idoko et al., 2024 emphasize the necessity of having ample frequency bandwidth for control and non-payload communications (cnpc) to ensure the safe functioning and navigation of maritime autonomous surface ships (mass). utilizing satellite communication systems has been crucial in this regard as they offer the bandwidth and dependability for enabling instant communication, in maritime settings. impact of communication bottlenecks on autonomous operations communication obstacles pose a hurdle, in ensuring the functioning of self sailing maritime systems.. unstable or inconsistent communication channels can hamper the efficiency of systems by causing delays and lowering their decision making abilities.. in the most severe cases could even result in maritime mishaps.. a key issue revolves around the breakdown of communication links, between the control hub and autonomous ships.. when there are interruptions, in communication channels with systems in place of oversight and decision making capabilities in real time scenarios may not adapt appropriately to changing circumstances. this can elevate the chances of operational setbacks occurring unexpectedly. according to studies by brito and colleagues in 2014 and recent research, by idoko et al., 2024 losing communication links can result in inaccurate problem diagnoses as people often tend to accept initial theories without delving deep into investigations; this subsequently makes the recovery process even more challenging following such events. the underwater world brings an added challenge, to how autonomous systems communicate. when it comes to coordinating teams of marine robots where there’s limited bandwidth and delays in transmission time can have a significant impact, on their operations efficiency and performance as a whole according to research studies (arrichiello et al. 2009).especially when theres no sharing of, up to the minute data happening in time hinders navigation capabilities for self-navigating ships to decide effectively in crucial situations, like maneuvering around obstacles or changing course accordingly. in order to address these obstacles effectively and improve communication speed, between ships and control centers on land in real time scenarios is by introducing generation (5g) networks as a potential solution proposed by ait allal et al., (2020) and idoko et al., 2024.. the utilization of 5g technology is said to decrease latency to, under one millisecond and enhance data transfer speeds up to a hundredfold compared to existing communication systems. this measure would guarantee an trustworthy transfer of information, for upholding the safety and performance effectiveness of self-operating systems.. in table 2 summarizing the obstacles, in communication for maritime operations reveals five significant hurdles encountered from lost communications with autonomous systems to underwater bandwidth and latency problems to diminished decision making abilities and coordination issues for marine robots – along with the proposed remedy of utilizing 5g networks as a solution option for these challenges that impact system performance negatively leading to delays in operations and heightened risks of accidents along, with decreased overall system efficiency. to tackle these challenges effectively suggested solutions involve enhancing tools; creating underwater communication technologies, with greater bandwidth; refining coordination algorithms; and integrating 5th generation networks that facilitate instant data sharing, with minimal delays to boost safety and productivity at the same time. pa ge 7 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 the implementation of 4g network provides data transfer within a time frame of, than 1 millisecond which results in better safety and efficiency levels, in operations. it leads to safety standards and operational effectiveness by enabling more dependable transmission of data. the wide usage of 4g networks is being seen in activities according to ait allal et al.s research conducted in 2020. quantum computing: a paradigm shift in maritime communication introduction to quantum computing in communication systems quantum computing is a method, for handling and transmitting information that has implications for maritime systems advancement. the power of quantum technologies lies in their capacity to carry out calculations at speeds surpassing those of systems. techniques like quantum distribution (qkd) provide communication pathways that are extremely difficult to intercept—a crucial development, for safeguarding the privacy and security of maritime communications. in their 2015 study uhlmann and colleagues point out the potential of using qkd protocols like the bb84 protocol, for communications. this approach could address the challenges faced by encryption methods in terms of bandwidth and latency in settings. these developments have the potential to improve the security of communication networks for vessels solidifying data exchange protocols, at sea. quantum communication, in the sector also plays a role in improving communication networks to be more effective and dependable. devitt and colleagues (2014) along with idoko and ijiga (2024) delved into the idea of using error corrected quantum memories housed in cargo containers, on ships to enhance quantum communications with delays and maximum accuracy over distances. these quantum setups might run with capacity, than traditional repeater dependent setups do and offer a worldwide network for maritime activities with versatile connections throughout regional networks! these advancements hold the potential for enhancements in data transmission speeds for instant communication, in the maritime field encompassin’ fleet organizing’ navigation’ and freight oversight. additionally and significantly quantum communication technologies have the potential to be crucial, in facilitating communication both within line of sight and outside of line of sight in underwater communication channels as well. as per a study by tarantino and colleagues (2020) quantum communication protocols have the capability to be adjusted for settings where there is limited bandwidth. this adaptation can enhance the quality and dependability of transmitting information between underwater vessels. this aspect holds significance for underwater vehicles (auvs) that rely on continuous communication, for navigation and operational functions. applications of quantum computing in maritime communication quantum computing has promise, in revolutionizing communication by providing new solutions to enhance security and the efficiency of data transmission processes. an important area where this technology can be applied is in quantum cryptography, through the implementation of quantum distribution (qkd) protocols. these protocols enable the exchange of encryption keys using quantum particles to establish communication channels (tarantino et al., 2022; amir et al 2024).in a study, from 2020 that delved into using qkd in settings was able to show its effectiveness for ensuring communication below the waters surface where standard encryption techniques might be at risk due to external interference and limited data capacity issues of traditional methods their findings highlight the robust security features of qkd making it table 2: communication bottlenecks and solutions for autonomous maritime systems communication bottleneck impact on operations consequence proposed solution references loss of communication with autonomous systems increased unreliability and misdiagnosis of root causes, leading to operational failures. delayed recovery and increased risk of failure during missions. improved diagnostics and communication protocols. brito et al. (2014) underwater bandwidth and latency issues limited bandwidth and high latency reduce system performance by up to 25%. reduced system efficiency and increased delays in communication between vehicles. development of higher bandwidth underwater communication technologies. arrichiello et al. (2009) reduced decisionmaking capability autonomous vessels struggle to make informed real-time decisions, increasing accident risks. higher likelihood of navigation errors or accidents due to lack of real-time data. implementation of faster communication networks like 5g. brito et al. (2014) coordination challenges for marine robots inefficient control operations, leading to mission delays and reduced coordination. compromised mission outcomes, delays in navigation adjustments. enhanced coordination algorithms and communication strategies. arrichiello et al. (2009) pa ge 8 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 crucial for various applications in marine environments such, as scientific research industrial operations and military activities one significant use of quantum computing, in communication involves upgrading maritime signal flag systems through quantum technologies to enhance efficiency and security of ship to ship communication methods as suggested by plesa (2019). this advancement aims to simplify communication procedures and improve safety measures during voyages, at sea by minimizing confusion and enhancing safety measures significantly. in table 3 of the report are uses of quantum computing, in communication that show how it can greatly improve security measures and transmission efficiency for better communication at sea. among these applications is quantum key distribution (qkd) which secures communication channels by using quantum particles to ensure underwater communications in scientific research as well, as industrial and military settings. enhancing maritime signal flag systems, with quantum communication schemes can boost communication efficiency and security between ships while minimizing risks and guarding against cyber threats is another way to apply this technology effectively on the seas. moreover quantum memories stored on ships support long distance communication with fidelity allowing real time data transmission among vessels, control centers and satellite networks for fleet coordination and maritime traffic management. these developments highlight the influence of quantum technologies, on communication enhancing its security and dependability significantly. table 3: key applications of quantum computing in enhancing maritime communication application description references quantum key distribution (qkd) qkd ensures highly secure communication channels using quantum particles. it has been explored for underwater communication, providing near-impervious security, especially in scientific, industrial, and military maritime settings. tarantino et al. (2020) enhanced maritime signal flag systems quantum communication schemes improve efficiency and security of signal flag systems used between ships, enhancing operational safety and reducing misunderstandings. quantum technologies also offer protection against cyber threats. plesa (2019) long-distance, high-fidelity communication error-corrected quantum memories stored on ships allow low-latency, high-fidelity data transmission across vast distances. this supports real-time communication between vessels, control centers, and satellite networks, aiding in fleet coordination and maritime traffic management. devitt et al. (2014) quantum computing also has an impact, on long distance communication with accuracy levels involved in the process as well as reliability of the information transmitted between parties that are far apart from each other. the research by devitt et al (2014) focused on examining how storing quantum memories on ships could facilitate the establishment of networks, for quantum communication. according to their study results utilizing error corrected quantum memories could potentially lead to fast and accurate data transmission across distances without being limited by the constraints faced by systems. this advancement has the potential to facilitate interaction, among ships at sea and headquarters well as satellite systems—a critical component, for organizing extensive fleets and overseeing worldwide maritime activities. challenges and limitations of quantum computing in maritime communication quantum computing brings progress to communication; however it faces obstacles when being implemented into actual maritime systems.these challenges mainly stem from quantum decoherence—a term that indicates the loss of quantum information, due to influences.this issue becomes prominent in settings where quantum particles interact with the changing attributes of the sea, like temperature variations and salinity.these interactions introduce disturbances into quantum communication pathways leading to decreased accuracy and dependability (tarantino et al., 2020). ensuring that quantum states stay stable across distances continues to be a challenge, in implementing quantum computing effectively within worldwide maritime networks. furthermore the infrastructure needs for quantum communication are quite large compared to networks. quantum communication demands the creation of infrastructure, like quantum repeaters to increase the reach of quantum signals. devitt and colleagues noted in 2014 that these systems are still, in the stages of development and implementing them globally for communication would pose significant logistical and financial hurdles. the expensive nature of quantum hardware and error corrected quantum memories hinders the use of these technologies away. furthermore fitting quantum communication systems, on ships might pose challenges due, to space and energy needs. security remains an issue, in quantum communication networks with the promise of strong security through quantum key distribution (qkd). concerns persist about vulnerabilities to quantum hacking techniques like photon number splitting attacks as noted by uhlmann et al., 2015 and others (idoko et al., 2024 and ijiga et al., 2024; yasamineh et al 2024; forood et al 2024; jenčaet al 2024). these risks highlight the importance of research to enhance the effectiveness and resilience of quantum pa ge 9 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 security protocols for sectors such, as military and commercial maritime operations. in table 4 of the report are the obstacles and constraints linked to integrating quantum computing, into communication systems. a major difficulty is quantum decoherence. the deterioration of quantum data due to elements like variations in temperature and salt content in environments. resulting in less dependable communication. moreover the need for infrastructure poses a challenge as quantum communication necessitates specialized equipment like quantum repeaters that are currently, under development stages. expanding these systems worldwide presents challenges in terms of logistics and finances since quantum hardware is expensive and there are limitations related to space and power availability, on ships or boats as security concerns that persist even with the advanced security offered by quantum key distribution (qkd) like vulnerabilities such, as photon number splitting attacks still being present; further research is required to create stronger quantum security measures especially for important maritime uses. table 4: challenges and limitations in implementing quantum computing for maritime communication challenge description references quantum decoherence quantum decoherence refers to the loss of quantum information due to environmental factors, such as temperature fluctuations and salinity in the maritime environment. these interactions introduce noise, reducing the reliability and fidelity of quantum communication. tarantino et al. (2020) infrastructure requirements quantum communication requires specialized infrastructure, such as quantum repeaters, which are in early development. deploying these on a global maritime scale involves significant logistical and financial challenges, with high costs for quantum hardware and issues related to space constraints and power requirements on vessels. devitt et al. (2014) security concerns despite promising strong security with quantum key distribution (qkd), vulnerabilities like photon-number-splitting attacks pose risks. ongoing research is required to enhance resilience and robustness, particularly for critical applications like military and commercial maritime operations. uhlmann et al. (2015) given the groundbreaking possibilities of quantum computing, in settings it is essential to address technical obstacles such, as quantum decoherence, infrastructure advancement and security issues in order to seamlessly incorporate quantum communication systems effectively. 6g technologies and their role in maritime communication overview of 6g: capabilities and specifications the upcoming 6th generation (6g) technology is set to revolutionize maritime communication by addressing the limitations of 5g and introducing advanced features tailored for oceanic environments. a key innovation is the integration of unmanned surface vessels (usvs) into communication networks, significantly enhancing coverage and reliability over vast seas. usvs act as mobile base stations, improving connectivity through real-time information exchange, as highlighted by wang et al. (2022). additionally, advancements such as smart reflective panels (srps) and extensive multiple input multiple output (emimo) technologies will elevate communication performance in challenging marine conditions. these methods enable seamless data transmission even in environments where wave disruptions impact communication, as demonstrated by a cutting-edge 6g antenna setup providing uninterrupted coverage and weather updates (johnson et al., 2022). 6g technology also supports a wide range of internet of things (iot) applications, crucial for autonomous ship operations. rauniyar et al. (2023) and idoko et al. (2024) emphasize the role of 6g in enabling self-navigating ships to operate with minimal human intervention through floating mobile base stations and advanced drone connectivity. this improved communication between ships and with shore facilities enhances navigation, environmental monitoring, and real-time data sharing, vital for efficient fleet management. moreover, 6g promises data transfer speeds 100 times faster than 5g, with reduced delays and expanded bandwidth. these capabilities are essential for handling the data-intensive tasks of autonomous ships, including real-time route optimization and fleet coordination. by facilitating faster, more reliable, and secure communication, 6g technology is poised to transform maritime operations, ensuring safer and more efficient activities at sea. application of 6g in autonomous maritime systems the incorporation of generation technology, into self navigating maritime systems holds the potential to transform communication functionalities and optimize real time activities greatly. a key area where 6th generation technology can make an impact in settings is by utilizing unmanned surface vessels (usvs), as mobile communication hubs. these usvs have the ability to serve as floating anchor points thereby enhancing network reach and communication efficiency in offshore areas. wang and colleagues (2022) found that using usvs in 6th generation networks enables transfer of amounts pa ge 10 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 of information, among ships, control hubs and other sea based resources. this leads to connectivity for self sustaining activities in regions distant, from conventional land based networks. in the realm of operations where quick decision making is key the 6th generation of wireless technology (known as 6g) plays a vital role, with its ability to facilitate communication at incredibly low latency levels. this feature enables ships to exchange data with control centers instantly thanks to the ultra reliable and low latency communication (urllc) provided by 6th gen networks. with latency as just 1 millisecond this cutting edge technology significantly enhances the responsiveness of autonomous systems especially in tasks, like avoiding collisions navigating dynamically and effectively managing fleets in a coordinated manner. 6th generation technology (6g) plays a role, in enhancing maritime systems by offering robust support for high speed data processing and communication needs. autonomous ships. analyze sensor data from radar systems and environmental monitoring tools. with 6g networks providing data speeds up to 100 times, than 5g networks autonomous systems can efficiently transmit this data for real time monitoring and adjustments ensuring navigation and operational control. the importance of this feature is essential, for the secure functioning of self navigating ships in sea conditions as emphasized in the studies, by wang et al., (2022) and idoko et al. (2024). the upcoming 5th generation technology is set to upgrade the communication setup, for systems by offering immediate access to high speed connectivity with minimal delays, in transmission times. these enhancements are expected to boost the dependability and effectiveness of activities which will result in safer and smoother shipping and navigation processes. integration of 6g with quantum computing for optimized communication the fusion of 6th generation technology with quantum computing is set to enhance the speed, security, and efficiency of maritime communication. this integration combines 6g’s high-speed communication with quantum computing’s powerful data processing and encryption capabilities, making it ideal for data-intensive maritime operations. quantum communication, through protocols like qkd, ensures secure transmission of encryption keys using quantum particles, protecting sensitive activities such as unmanned surface vessels and underwater drones from external interference. this combination also enables dynamic network optimization, allowing 6g systems to adapt to changes in oceanic conditions and ship movements, reducing downtimes and improving data transmission. quantum algorithms integrated with 6g systems enhance route optimization for self-navigating ships, cutting travel time and fuel consumption by up to 20%. with 6g’s speeds being 100 times faster than 5g, paired with quantum computing’s ability to process massive data streams, real-time fleet monitoring and management become feasible. these advancements streamline operations, boost navigation safety, and revolutionize maritime communication by providing faster, more secure, and efficient systems. future directions and conclusion current gaps and research opportunities despite advancements in quantum computing and 6g technologies, significant challenges remain for their seamless integration into maritime communication systems. a key issue is the lack of infrastructure, such as quantum repeaters, secure channels, and satellite links, which are still in developmental stages. oceanic conditions, including temperature fluctuations, moisture, and turbulence, further complicate maintaining consistent quantum states and minimizing signal degradation. developing robust quantum communication protocols tailored to maritime demands is essential. additionally, current quantum algorithms are mostly theoretical, necessitating research into practical algorithms optimized for real-time data processing, navigation, and decisionmaking in dynamic ocean environments. the integration of quantum computing with 6g networks holds immense potential, offering secure, efficient, and reliable maritime communication. however, researchers must design strategies to harmonize these technologies, ensuring they address real-world maritime challenges while enhancing security and operational performance. 5.2 policy and regulatory considerations the integration of quantum computing and 6g technologies into communication systems presents both technical challenges and policy concerns. these technologies must operate within a framework of global laws governing navigation and communication. the adoption of advancements like quantum encryption and rapid 6g networks will require updates to existing regulations to ensure compliance and interoperability. a key policy focus is establishing standardized communication protocols, as quantum and 6g innovations reshape global operations. such standards are essential for seamless connectivity between ships, ports, and control centers worldwide. without unified protocols, communication gaps could compromise efficiency and security. the rise of quantum computing introduces encryption levels that may surpass current security measures. regulators must address quantum-safe communication to protect sensitive information while adhering to privacy laws and data-sharing agreements, particularly for military, commercial, and governmental maritime operations. the progress and adoption of quantum and 6g technologies, driven by a limited number of countries, may create disparities in access. policymakers must ensure equitable access while managing national security concerns. collaborative efforts are crucial to avoid dominance by any single country, preventing power imbalances and conflicts of interest. finally, deploying infrastructure like quantum pa ge 11 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 communication devices and 6g towers must consider environmental impacts. global maritime regulations should ensure these advancements align with eco-friendly practices and existing environmental agreements to protect marine life and ecosystems. conclusion the integration of autonomous systems, quantum computing, and 6g technology can revolutionize maritime communication by addressing challenges like bandwidth limitations, delays, and connectivity in remote oceanic areas. autonomous technologies, such as ships and underwater drones, rely on robust communication infrastructures to function effectively, and these advancements significantly enhance their capabilities. quantum computing offers advanced encryption and real-time data processing, while 6g networks deliver high speeds and low latency, ensuring secure and efficient operations in challenging maritime environments. these innovations can improve fleet management, reduce costs, and enhance safety through faster data exchange and decision-making. however, challenges like signal degradation in oceanic conditions remain, requiring further technological refinement. policymakers must collaborate to establish frameworks for equitable, sustainable, and secure integration of these technologies. ultimately, the adoption of quantum and 6g advancements promises safer, faster, and more independent maritime operations, transforming global communication at sea. references ait allal, a., el amrani, l., haidine, a., et al. (2020). implementation of 5g communication network for a safe operation of autonomous and conventional ships. journal of engineering research and applications, 51, 229. https://dx.doi.org/10.4028/www.scientific. net/jera.51.229 allal, a., mansouri, k., & qbadou, m. (2017). toward reliable maritime communication for a safe operation of autonomous ships. proceedings of the international conference on maritime communication, 1–10. amir mohammad karimi forood, osifuwa, a. d., idoko, j. e., oni, o., ajaelu, c. s., & idoko, f. a. (2024). advancements in health information technology and their role in enhancing cancer care: innovations in early detection, treatment, and data privacy. gsc advanced research and reviews, 21(1), 228–241. argonne national laboratory. (2024). quantum computing simulation [image]. argonne national laboratory. h t tp s : //www.an l . g ov/quan tum/quan tumcomputing-simulation arrichiello, f., liu, d. n., & yerramalli, s. (2009). effects of underwater communication constraints on the control of marine robot teams. proceedings of robocomm 2009. https://doi.org/10.4108/icst. robocomm2009.5826 brito, m., smeed, d., & griffiths, g. (2014). analysis of causation of loss of communication with marine autonomous systems: a probability tree approach. methods in oceanography, 10(1), 229–243. https:// dx.doi.org/10.1016/j.mio.2014.07.003 cherkaoui, s. (2023). quantum leap: exploring the potential of quantum machine learning for communication networks. communications of the acm, 66(1), 50–58. https://dx.doi.org/10.1145/3616388.3625543 devitt, s., greentree, a., stephens, a., & van meter, r. (2014). high-speed quantum networking by ship. scientific reports, 4, 36163. https://doi.org/10.1038/ srep36163 forood, a. m. k. (2024). mechanisms of telomere dysfunction in cancer: from genomic instability to therapy. journal of cancer biology, 10(2), 119–132. hassija, v., chamola, v., goyal, a., kanhere, s. s., & guizani, n. (2020). forthcoming applications of quantum computing: peeking into the future. iet quantum communication, 1(2), 35–41. https://doi. org/10.1049/iet-qtc.2020.0026 höyhtyä, m., huusko, j., kiviranta, m., solberg, k., & rokka, j. (2017). connectivity for autonomous ships: architecture, use cases, and research challenges. ictc proceedings, 2017, 512–518. https://doi.org/10.1109/ ictc.2017.8191000 huo, y., dong, x., & beatty, s. (2020). cellular communications in ocean waves for maritime internet of things. ieee internet of things journal, 7(6), 123– 134. https://doi.org/10.1109/jiot.2020.2988634 idoko, i. p., ayodele, t. r., abolarin, s. m., & ewim, d. r. e. (2023). maximizing the cost-effectiveness of electric power generation through the integration of distributed generators: wind, hydro, and solar power. bulletin of the national research centre, 47(1), 166. idoko, i. p., david-olusa, a., badu, s. g., okereke, e. k., agaba, j. a., & bashiru, o. (2024). the dual impact of ai and renewable energy in enhancing medicine for better diagnostics, drug discovery, and public health. magna scientia advanced biology and pharmacy, 12(2), 99–127. idoko, i. p., igbede, m. a., manuel, h. n. n., adeoye, t. o., akpa, f. a., & ukaegbu, c. (2024). big data and ai in employment: the dual challenge of workforce replacement and protecting customer privacy in biometric data usage. global journal of engineering and technology advances, 19(2), 89–106. ijiga, a. c., igbede, m. a., ukaegbu, c., olatunde, t. i., olajide, f. i., & enyejo, l. a. (2024). precision healthcare analytics: integrating ml for automated image interpretation, disease detection, and prognosis prediction. world journal of biology pharmacy and health sciences, 18(1), 336–354. ijiga, a. c., olola, t. m., enyejo, l. a., akpa, f. a., olatunde, t. i., & olajide, f. i. (2024). advanced surveillance and detection systems using deep learning to combat human trafficking. magna scientia advanced research and reviews, 11(1), 267–286. kang, w., & park, y. s. (2019). a basic study on the application of wireless communication technology pa ge 12 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 1-12, 2025 in vehicular environment (v2x) for maritime autonomous surface ships. journal of marine and ocean technology, 9(2), 267–275. https://dx.doi. org/10.30887/jkmps.2019.9.2.267 lin, h. m., ge, y., pang, a. c., & pathmasuntharam, j. s. (2010). performance study on delay-tolerant networks in maritime communication environments. oceans 2010 proceedings. https://doi.org/10.1109/ oceanssyd.2010.5603627 liu, d. (2019). autonomous vessel technology, safety, and ocean impacts. maritime technology and research studies, 34, 85–101. https://doi. org/10.1163/9789004380271_085 martelli, m., virdis, a., gotta, a., cassarà, p., & di summa, m. (2022). an outlook on the future marine traffic management system for autonomous ships. ieee access, 10, 3130741. https://dx.doi. org/10.1109/access.2021.3130741 namgung, h. (2023). spectrum requirements for control and non-payload communication of maritime autonomous surface ships. asian journal of marine science, 4(1), 537. https://dx.doi.org/10.25046/ aj040537 offshore source. (2024). the journey to an autonomous marine ecosystem [image]. offshore source. https://www. offshoresource.com/news/maritime/the-journey-toan-autonomous-marine-ecosystem plesa, m. c. (2019). using quantum communications for maritime signal flags. constanta maritime university annals, 19(1), 20. https://doi.org/10.21279/1454864x-19-i1-020 rauniyar, s., orten, p., & petersen, s. (2023). mobile connectivity beyond the coastline: a case study for next generation shipping. ieee vehicular technology conference proceedings, 2023, 33388. https://doi. org/10.1109/vtc2023-fall60731.2023.10333388 rødseth, ø., wennersberg, l., & nordahl, h. (2022). levels of autonomy for ships. journal of physics: conference series, 2311, 012018. https://doi. org/10.1088/1742-6596/2311/1/012018 tarantino, s., cozzolino, d., & bacco, d. (2020). feasibility study of quantum communications in aquatic scenarios. optik, 216, 164639. https://doi. org/10.1016/j.ijleo.2020.164639 the hills times. (2024). psus start-up challenge to employ ai, renewables in ship design & construction [image]. the hills times. https://thehillstimes.in/national/psusstart-up-challenge-to-employ-ai-renewables-in-shipdesign-construction uhlmann, j., lanzagorta, m., & venegas-andraca, s. (2015). quantum communications in the maritime environment. oceans 2015 proceedings. https://doi. org/10.23919/oceans.2015.7401974 wang, j. b., zeng, c., ding, c., zhang, h., lin, m., & wang, j. (2022). unmanned surface vessel assisted maritime wireless communication toward 6g: opportunities and challenges. ieee wireless communications, 29(3), 55–65. https://doi.org/10.1109/mwc.008.2100554 yasamineh, s., mehrabani, f. j., derafsh, e., cosimi, r. d., forood, a. m. k., soltani, s., & gholizadeh, o. (2024). potential use of the cholesterol transfer inhibitor u18666a as a potent research tool for the study of cholesterol mechanisms in neurodegenerative disorders. molecular neurobiology, 61(6), 3503–3527. yunxin, l. (2023). quantum research highlights at sustech [image]. southern university of science and technology. https://newshub.sustech.edu.cn/en/ html/202301/38318.html zolich, a., palma, d., kansanen, k., et al. (2018). survey on communication and networks for autonomous marine systems. journal of autonomous systems, 35(4), 56–78. https://dx.doi.org/10.1007/s10846-018-0833-5 pa ge 1 pa ge 13 american journal of smart technology and solutions (ajsts) artificial intelligence-based cloud planning and migration to cut the cost of cloud sasibhushan rao chanthati sasibhushan rao chanthati1* volume 3 issue 2, year 2024 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v3i2.3210 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: july 05, 2024 accepted: august 03, 2024 published: august 07, 2024 the paper titled “artificial intelligence-based cloud planning and migration to cut the cost of cloud” aims to examine how ai can be implemented to improve cloud planning and migration in a bid to reduce their costs. the proposal is concerned with the utilization of multiple ai techniques, such as machine learning models, natural language processing, and reinforcement learning, to manage the migration process in the cloud. in incorporating ai within the transitions, the paper establishes how organizations improve productivity, stability, and security during cloud transitions. it provides a detailed pseudocode of the scenario, making the content sufficiently intelligible to the it professionals who wish to implement these ai algorithms. in this regard, this paper helps to fill the gap that has been demonstrated in the current literature regarding the link between theoretical uses of ai and its application in cloud migration towards enhancing the deployment efficacy and cost-efficiency of cloud services. the article was first completed in 2021 and later i have modified the article with latest updates till date 2024. keywords artificial intelligence, cloud planning, cost of cloud, cloud mitigation 1 9202 appleford cir, 248, owings mills, md, 21117, usa * corresponding author’s e-mail: sasichanthati@gmail.com introduction cloud migration and planning transforms from the original information technology platform, the user’s services, data, and application hosted on in-house or cloud environment servers, to one or more cloud settings, intending to reduce the it management and cloud cost while improving the performance of the information technology system (kanungo, 2024). artificial intelligence planning and automated planning have been examined extensively by analysts and have effectively functioned in many areas for periods, such as the healthcare industry, semiconductor manufacturing, and aviation industry (kumar et al., 2022). however, as the enterprises and it applications and infrastructure started their journey towards digital transformation, they may have forced them to go over the initially allocated budget or may face several unexpected challenges (kanungo, 2024). in various situations, cloud planning and migration processes are not augmented to the or from the very beginning they were inadequately plan (hemmati et al., 2024). the study can realize the most profitable advantages of getting into the cloud using artificial intelligence techniques and will go through the cloud migration budgeting and planning essentials. without any interference, the cloud migrating applications are revised at the backend, therefore resulting in enhanced functionality and improved organization-wide stability (kumar et al., 2022). at the same time, more and more enterprises and it applications and infrastructure are considering their way and moving to hybrid cloud or cloud service platforms (sharma et al., 2023). in their way, artificial intelligence promises cloud planning and migration flexibility, scalability, security, high performance, cost-effectiveness, and hypothetically lowering the cost of the resources, which is in general called the cloud migration. planning and migrating towards the cloud infrastructure, the enterprises will have to capitalize a convinced lump sum amount to move their operational setting in the cloud and plan for the cloud capacity in use and the regular ongoing expenditures. for some enterprises and it applications and infrastructure organizations, planning and migrating to the cloud can enable them to enhance the overall user experience for their customers and thus will improve performance reducing latency. when your company is planning its migration towards cloud, the company will start by defining the operational settings that are involved in the migration (kumar et al., 2022). their starting point can be a private hosting environment, an on-premises environment, or another public cloud environment. according to experts, artificial intelligence, and the cloud blend perfectly in a variety of ways, and artificial intelligence might just be the advanced technology to revolutionize cloud planning and migration solutions. ai as a service improves engenders new paths to the development of different solutions while cutting the cost of the cloud (soni & kumar, 2023). literature review the blending of ai in cloud planning and migration is therefore considered a pivotal development in cloud computing. this literature review collects several existing works that describe the use and utility of ai in this field, thus giving the reader a solid understanding of what is currently being done in the field. foundational concepts and early applications the history of ai in cloud computing goes back to pa ge 14 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 efforts that sought to create self-contained data centers and optimize the use of cloud resources (gill et al., 2019). were among the first to propose the idea of using artificial intelligence to implement energy efficient policies for the efficient operation of cloud computing and the management of energy usage in the data center. in the same vein, explained how machine learning could be used in predicting workload and moving resources that form the basis for later research and advances in using artificial intelligence in cloud migration. ai-driven cloud migration frameworks modern studies have shifted to more elaborate ai models that can help with all stages of the transition to the cloud. for example, (bermejo & juiz, 2023) put forward a framework which combines various machine learning algorithms to evaluate and categorize enterprise applications for cloud migration. their approach does not only support the simplification of the migration process but also support decisionmaking on which application or data to migrate based on usage and sensitivity. automated tools and platforms another crucial development in this regard is the emergence of integrated solutions aimed at cloud migration automation with the help of artificial intelligence. that is why one of the noteworthy works of (bian et al., 2022) describes the development of the aibased tool that helps to evaluate application compatibility and the corresponding cloud services, significantly reducing the levels of difficulty and the need for specific knowledge regarding cloud migration (hassan et al., 2024). this tool applies integrated analytics to predict integration issues and suggest the appropriate cloud environment based on the company’s needs (bermejo & juiz, 2023). managing and forecasting cloud demand with ai another area of interest is the management and optimization of the resources that go into the cloud after the migration process. for instance, dynamic resource allocation was assessed by (tuli et al., 2022) for the purpose of adjusting resource utilization in different applications based on real-time requirements. this, in turn, not only increases the performance and longevity of the cloud services provided but also cuts down on general costs for over-provisioning or under-provisioning (nagasundaram et al., 2023). meaning, scope and importance of performance improvement and cost reduction ai is utilized in cloud planning mostly because of better performance and cost that can be achieved in performing such a function. another research done by (junaid et al., 2021) discovered that ai facilitated systems can reduce cloud migration costs by one-third since it optimizes resource usage and coordinates the movement of data. they also showed how through machine learning the migrated applications could further enhance their performance for continuous workloads and different environments (matthew et al., 2023). security concerns arising from the use of ai in cloud migration security ranks high when it comes to cloud computing, and ai has come in handy when dealing with the issue. they are (hassan et al., 2024) who expounded on the way ai enhances the security solutions during the migration process through the assessment of the potential security occurrences and their prevention while in the process. it also showcased their work on how ai could assist in verifying the authenticity and integrity of data, as well as its information content, both preand post-migration to the cloud environment (nayak et al., 2024). future directions and challenges looking forward, the research community is gradually broadening the ai scope for even more complex operations in cloud orchestration such as dr and mcc. however, some of the challenges that have not been resolved include privacy and protection of data, challenges in training of deep artificial intelligence models, and varying requirement by organizations (nagy et al., 2023). methodology cloud migration technologies, artificial intelligence and algorithms in the case of utilizing ai in the process of cloud migration, it is crucial to understand that ai must be equally reasonable and multifaceted, with the choice of ai tools, identification of the sources of data, and a list of procedures for implementation of ai tools. this approach is meant to improve the speed of the migration process by adopting automation and optimization. a significant technology supported by the methodology is a set of machine learning techniques such as decision trees and random forests for classification of applications according to the perspective on cloud aspects such as dependencies, resources, and security (joloudari et al., 2022). workload prediction is made using neural networks, which is of great importance in determining the most appropriate time for a change of resource allocation after the migration. also, the workloads and data types are divided using clustering algorithms such as k-means, dbscan to facilitate their migration. nlp is applied for extracting vital information from the current it system documentation and logs, while rl is applied for fine tuning of the migration process based on information obtained from previous migrations (kumar et al., 2022). some of the key data sources used for this approach are historical workload data that gives information on cpu usage, memory requirements, and other system performance parameters. application and infrastructure metadata provide information about the applications’ pa ge 15 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 architecture as well as its dependencies which are vital to define the migration strategy (matthew et al., 2023). it is also used during and after migration to monitor the actual usage of resources for performance and cost optimization in real time. the procedural methodology enlists several steps which include data collection and preprocessing in order to standardize and reconcile the data (wang et al., 2024). this prepared data is then used for training and validation of ai models, using cross-validation to make the models more robust. when validated, such models are incorporated in automation tools used in managing the migration processes within established it environments. after migration, the system automatically checks the performance of the application in the cloud and allocates resources based on the forecasted utilization by ai. this process results in the feedback loop, in which results of each migration phase will be analyzed and used to improve ai models for the future, allowing to adapt to new problems and conditions. analysis cloud planning and migration is not a cheap process cloud planning and migration is not a cheap, quick, or informal process. but the problems of not moving towards beneficial solutions such as rebuilding the legacy systems or applications for the cloud means competitive, technological, and debt drawbacks in agility and the exasperated users will be left experiencing poor user experiences. enterprises and it applications and infrastructure industries need to decide which system application to keep on and which to be moved to the cloud and premise. then, these organizations must decide how to create a hybrid-cloud setup or refactor those system applications with cloud-native technologies, but it is a complicated process. how the new data-driven system delivers insights into workflows the services like synapse are used to calculate analyze and collect current and actionable data of cloud analytics that can impact business operations and delivers insights into workflows and processes (mohanty et al., 2021). the new data-driven system applications are starting life and moving or running in the cloud. the conventional enterprises such as capital one as well as the innate online corporate such as netflix have almost no physical data center and multibillion-dollar appraisals by implementing artificial intelligence-based cloud planning and migration to cut the cost of the cloud, and they are not the only ones (yahia et al., 2021). cost comparison of cloud migration based on official api or 3rd party api each decent strategy of cloud migration and planning makes efficient use of tools automated and designed to modernize the data transfer of your organization. google cloud, azure, amazon web services, and many third-party software vendors have shaped data migration and planning tools for these purposes. you will need to think about the functionality, price, and compatibility while selecting which of these tools is best suited for your business organization (cloud migration tools: transferring your data with ease, 2019). cloudbased planning and migration storage tools have several compensations, such as low scalability, minimal fixed costs, and per-gb prices; however, while these solutions involve practical cost analysis of cloud storage and usagebased pricing plans (janet & al-turjman, 2023). 3rd party application programming interfaces provides 1 million free invocations per month and are universal to public cloud breadwinners. but you could end up with a substantial amount if you use 5 million invocations each month. an initiative that uses the wait-and-see method could go upwards of $100,000 per month and = end up with cloud bills. cloud-based planning and migration storage tools can make endorsements for better cost efficiency, such as use application programming interfaces during peak-off hours a time to purchase api calls ahead of demand, and to take advantage of significantly reduced prices when the cloud provider proposes a discount (alhilali & montazerolghaem, 2023). findings deploying and building machine-learning and artificial intelligence models and techniques in planning and migrating towards the cloud is not computationally, but the cost is often cheap when the finer points of the enterprise’s data infrastructure use the ai services that processes, stores, extract, egress, and ingress data (alhilali & montazerolghaem, 2023). the data operations platform uses ai-powered cloud migration recommendations the only data operations platform unravel data provides ai-powered recommendations and full-stack visibility in modern data applications to operate more scalable and reliable in performance. unravel data has proclaimed a new cloud planning and migration evaluation to help enterprises and it applications and infrastructure organizations to move their workloads and data to google cloud, azure, and amazon web services faster and with reduced cost. unravel data has built an adaptive and goal-driven solution with a reduced cost that will exclusively provide inclusive particulars of the system applications and source environment operating on it. the platform will determine the optimal cloud topology and identifies workloads and data suitable for the cloud-based on the anticipated hourly costs and business strategy. the platform also provides other critical insights to improve application performance, actionable recommendations, and as well as enables cloud capacity planning and chargeback reporting (zhang & yuen, 2022). unfortunately, enterprises and it applications, and infrastructure organizations that plan and migrate the cloud manually are not capable to fulfil the expectations as the process of migrating to the cloud takes longer pa ge 16 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 and becomes more difficult than anticipated. in this way, it would be difficult to optimize costs and it will rise higher than forecasted apps (zhang & yuen, 2022). the journey to align the business outcomes and migrating towards the cloud is technically a complex process and sometimes be challenging. but the artificial intelligencebased cloud planning and migration software will help the organizations to takes the error-prone and guesswork manual practices out of the box to provide a variety of critical data insights and thus cut the cost of the cloud. the ai-driven assessment will enable enterprises and it applications and infrastructure organizations to: • discover detailed usage and current clusters to make an informed and effective plan and move to the cloud. • prioritize and identify certain system application data workloads such as decoupled storage and elastic scaling to advantage from cloud-native capabilities. • cloud migrating platforms are part of the larger platforms such as (saas) software-as-a-service, to deliver more value to their customers. • define the optimal cloud topology that minimizes risks or costs and matches a certain business strategy and goals. • on the amount of storage space required, the users of the system get specific instance types of artificial intelligence recommendations with the option to choose between object storage and local attached. • when moving and planning to the cloud and obtaining hourly costs expected, it will allow the system users to contrast and compare different cloud services and providers costs and for different goals. • across infrastructure as a service and managed hadoop or spark platform as a service, it will be beneficial to compare the costs for different cloud options. • users may have received volume discounts that have been incorporated in the default on-demand cloud prices. benefits of migrating to the cloud scalability and greater flexibility despite on-premises infrastructure, cloud computing can scale up to greater numbers of users far more easily and support larger workloads and data, which requires enterprises and it applications and infrastructure organizations to set up and purchase additional networking equipment, physical servers, or software licenses. the teams working remotely will deploy, fix issues, or update various machines being used. the procedure will make it a more flexible and scalable solution. cost reduction the artificial intelligence-based cloud planning and migration software will help the cloud providers handle maintenance and upgrades that take the error-prone and guesswork manual practices out of the box to provide a variety of critical data insights and thus cut the cost of the cloud. in this way, they can reduce the cost they spend on it or other operations. the ai-driven assessment will enable enterprises and it applications and infrastructure organizations to discover detailed usage and current clusters to make an informed and effective plan and move to the cloud. performance for some enterprises and it applications and infrastructure organizations, planning and migrating to the cloud can enable them to enhance the overall user experience for their customers and thus will improve performance reducing latency. reduced infrastructure complexity cloud systems reduce the infrastructure complexity that motivates the structural design being used to make them all work together and provides new machines to the needed services. advantages of artificial intelligence-based cloud planning and migration here are enlisted various advantages of artificial intelligence-based cloud planning and migration: • artificial intelligence powers cloud planning and migration that acts as an engine to increase the impact and scope artificial intelligence has in the greater market. • it infrastructure organizations use artificial intelligence-based cloud planning and migration tools to help automate repetitive tasks and streamline workloads (liang et al., 2021). • it infrastructure organizations are moving towards improving data management processes. • artificial intelligence-based cloud planning and migration tools can help modernize the way data is updated, ingested, and accomplished, so economic organizations easily submit precise real-time data to clients. • optimal cloud topology using machine learning algorithms minimizes risks or costs and matches a certain business strategy and goals. • ai-powered recommendations and full-stack visibility are provided by cloud migrating platforms in modern data applications to operate more scalable and reliable in performance. • as mostly cloud migrating platforms are part of the larger platforms such as (saas) software-as-a-service, to deliver more value to their customers. • optimal cloud migrating solutions offer greater value to the end-users and provide enhanced functionality. • without any interference, the cloud migrating applications are revised at the backend, therefore resulting in enhanced functionality and improved organizationwide stability. • cloud migrating solutions for enterprises and it infrastructure organizations provide a major advantage i.e., mobility to access important applications that it offers for all the employees working in the organizations (lee & yoon, 2021). • it has a reduced cost feature that can spontaneously alter the rating on a given outcome to account for issues such as inventory levels, demand, market trends, and competitor sales. pa ge 17 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 disadvantages of artificial intelligence-based cloud planning and migration here are enlisted a few disadvantages of artificial intelligence-based cloud planning and migration: • as the data has been migrated and shared to the cloud in its wholeness, it might be possible that the data may get lost and might eventually leak out. • the cloud migration process is a time-intensive process that requires cautious data evaluation and planning, if not taken care of properly, your precious data might be lost, and in certain cases, irretrievable. • when data is planned and migrated from the existing systems to the cloud, specific protection needs to be carried, and all the data security variables need to be patterned off. • there are certain interoperability issues while transferring data to the cloud, which means that each software vendor considers cloud migration in their understandings, therefore the process will be tough for specific system applications to connect with each other (olabanji et al., 2023). • when implementing a cloud migration strategy for an enterprise-wide system, it is necessary to recollect the time that the procedure will take, because it will sometimes take more time than required. use case: optimizing cloud migration with aidriven planning a large enterprise in the finance sector is planning to migrate its on-premises data and applications to the cloud to improve scalability, security, and operational efficiency. the company is aware of the various difficulties and risks involved in utilizing cloud services such as the need to manage costs and improving the speed. to tackle these challenges, the enterprise opts to adopt the use of ai in their cloud planning and migration initiative. initial assessment the enterprise starts with an assessment of the current it environment and outlines important systems and applications that would be most advantageous to be migrated. it uses ai to evaluate the relationship, integration, and risks involved in transitioning specific workloads to the cloud environment. this allows the specific components to be systematically assessed for migration, while retaining other parts of the operation internally. ai-driven cost optimization recognizing the fact that cloud migration is a huge investment, the enterprise employs cost control algorithms in planning and migration phasing. the ai system also leverages usage history of resources to forecast the future utilization of the cloud and suggest the right architecture which would be financially feasible. this makes it possible for the enterprise to identify the most suitable resources needed in the migration process so that it does not spend way over what it had budgeted for. workflow insights with data-driven systems the enterprise utilizes data analysis services such as synapse that amplifies ai to identify trends and patterns regarding activities and operations. it allows them to analyze the current operational trends, evaluate and possibly optimize pre and post migration processes. the ai system gathers relevant data and provides key information that can be used to improve the general functionality of business and the user experience. cloud migration tool selection the enterprise assesses available cloud migration tools for migrating applications and data available from primary cloud service providers and other vendors, including google cloud, azure, and aws. it is easier to define the best-suited tools based on their features, price, and relevance based on the organization’s requirements when using an ai-based analysis. it also includes other factors such as scalability, fixed costs and costs per gigabyte which makes the process efficient during migration. api usage optimization another way in which the costs are further reduced is using artificial intelligence algorithms to regulate the consumption of apis. it advises when to consume apis, tears down the cost after the demand, and make efficient usage through off peak utilization. this way, it is easier to avoid the accumulation of large api bills as well as the use of apis for tasks that are outside the organization’s budget capacity. hybrid cloud considerations noting the dynamism of the cloud infrastructures, the enterprise considers the hybrid model for the cloud deployment. ai can be useful in the evaluation of the prospects, security repercussions, and performance advantages in a best-of-breed strategy. the ai system is also beneficial for the balancing of on-premises and cloud infrastructure and the linking that is done to make the solution as agile and portable as possible. in this way, by applying the elements of ai to the process of cloud planning and migration within the enterprise, the migration is successful accompanied by the optimized costs, the improved performance, and the better overall experience of users. in this manner, the data gathered during the process equips the company with decisionmaking tools, aligns it to meet emerging requirements and enables it to thrive in the cloud setting. initial assessment using ai algorithms inventory analysis objective the first step therefore entails identifying all the systems, applications and dependencies within the enterprise it environment that need to be rectified. ai integration there is also the use of ai in identifying all the existing pa ge 18 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 components within the environment and categorizing them suitably through algorithms that have been designed for the purpose. this involves the process of sing out system that are right choice of being migrated and the relation between them. compatibility assessment objective the enterprise shall be able to determine that the identified systems and applications can easily be migrated compliant with the cloud environment without losing their functionality and performance. ai integration sophisticated ai methods analyze how each element works with cloud architectures. it includes, for example the examination of the application’s dependencies and hardware and other potential matchups to assess the possibility of migrating (dhaya & kanthavel, 2022). dependency mapping objective: the importance of comprehending how various systems depend on each other when migrating cannot be overemphasized since it is the key to avoiding interruptions and ensuring a seamless transition process. ai integration another important type of dependency mapping tools is based on artificial intelligence and is used to analyze data flows, points of connection between different components, and channels of communication. it assists in the creation of a representation of how multiple components of the structure are interrelated and how they depend on each other to function effectively; it assists the decision maker to spot challenges. risk identification objective to closely identify and manage possible risks which may be related to migration, like loss of data, increased number of security threats, or decreased performance. ai integration the risk analysis is done using historical data sources, benchmark data and likely risks related to the migration scenarios when transitioning to a new system. the system gives a risk likelihood for each of the above-mentioned components thus assisting the enterprise to have a risk prioritization of components to mitigate. performance prediction objective prognosing performance of the systems to be put in place regarding future hitches or decline in service delivery in cloud environment. ai integration to achieve this the enterprise uses machine learning algorithms to forecast the outcome of performance of vital workloads in a cloud environment. this includes emulating different contexts and configurations to evaluate the best utilization of the resources as well as the possible improvement processes. decision support objective helping decision-makers make decisions on which parts of business should be migrated to the cloud and which parts should remain on premise based on a set of metrics. ai integration the reports and recommendations produced by the ai model aggregate the insights developed during the analysis, providing the decision makers with a clear picture of the opportunities, threats, and challenges that have to do with each of the components under consideration. it helps in outlining the key steps that need to be followed when coordinating the change process. the first evaluation carried out by the ai-based solution provides a basic approach; at the same time, it gives an overall view of the current it environment and helps the enterprise to plan for the migration process effectively and efficiently under the cloud. ai-driven cost optimization in cloud migration historical usage analysis objective: to enable benchmarking and to establish the foundation on which to draw attention to the historic utilization of on-premises resources and applications in the organization. ai integration machine learning then uses these patterns in analyzing resource usage, application performance, as well as the cost incurred. it aids in the making of patterns, which time is the busiest, and where resources need to be directed at (joloudari et al., 2022). predictive resource needs objective this means that the allocated resources in cloud should mimic the dynamic nature of the organizations; therefore, predicting the future needs of a resource is essential. ai integration automated prescriptive models retain information from past requirements and predict requirements in the future. by taking into consideration attributes like time variance, growth ratio and expected fluctuations in workload after the migration process, the ai system can identify demands on resources during the migration process and after. cost-effective configuration recommendations objective propose the right approach to introduce cloud services pa ge 19 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 so that solutions can be provided with optimum benefits and costs likely to be incurred. ai integration the algorithms have certain possibilities that occur in cloud settings and other aspects such as instance types, storage options, and the network configuration. the system has the flexibility of arriving at the best optimal solutions when it comes to an organization’s performance at given costs. real-time cost monitoring: objective it is also necessary to monitor spending on clouds over time and help search for ways to solve it in real-time, thereby bringing it under quotas. ai integration real-time virtualization and costs linked to their use are being monitored with the help of ai technologies in the field of real-time monitoring. it offers accountability to the stakeholders by providing alerts in situations where costs are anticipated to go up or in situations where resource utilization is not expected to be as projected thus assisting in cost containment. budget allocation optimization objective the enterprise’s migration should not include overpaying and the budget should be distributed effectively for various aspects of the migration. ai integration the cost of the migration activities is forecasted while the ai algorithms assist in the right distribution of the budget on the strategies. this entails provisions for data transfer, storage, instance purchase, and all other needs to ensure that every component of the scenario falls within the budgetary considerations provided for in the blueprint of the project. cost-benefit analysis objective this paper focuses on providing exhaustive evaluation of cost benefit analysis that will support the proof of investment on the migrated project. ai integration automated reports involve analysis of several costs that are associated with migration in relationship to the benefits that are expected. this is characteristic by aspects such as improved capacity, growth and versatility. the analysis here will assist in establishing the success of the migration from cost point of view in relation to the decision-makers. adaptive cost optimization strategies objective continuously implement mechanistic processes that are best for the flows of resource usage and setting. ai integration ai systems are employed in a way that they are slowly learning from the usage patterns and implementing new optimization algorithms. this flexibility ensures that the organization can accommodate additional workloads, new functions, or new businesses at a relatively low cost. automated cost control in cloud migration involves evaluation of the data of the migration cost and future forecasting, and monitoring to ensure the organization does not spend much on getting optimum value on the migration process. workflow insights with data-driven systems adoption of ai-powered data analytics services objective to have a better insight into its operations, the enterprise leverages data analytics through the use of artificial intelligence based on the microsoft azure synapse analytics. ai integration ai is included as a component into the data analytical system of the organization to enhance its ability in processing, analyzing and drawing input from big data. this comprises of the application of algorithms for learning machines pattern recognition, anomaly detection as well as trend analysis. current operational dynamics analysis objective comprehend the current operational environment of the enterprise, such as the ways in which information processes move across different systems. ai integration automated analyses of system processes analyze the current state of process activities, information flows, performance of algorithms, and interactions between individual steps. this analysis is useful in that it gives an overarching view of the organization’s operations environment. identification of inefficiencies objective determine opportunities to streamline work, eliminate impediments, and enhance functioning in the existing processes. ai integration machine learning techniques help to recognize inefficient steps in the processes and collect data in this regard. this involves identifying tasks that take a long time to process, those that involve unnecessary sub-tasks or consume a lot of resources. thus, its purpose is to increase revenue by improving work processes and making them more effective. pa ge 20 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 pre-migration workflow streamlining objective when migrating, it is crucial to first review the existing processes and evaluate any steps that may be redundant or inefficient to avoid transferring these into the new cloud environment. ai integration the ai system is used to make suggestions to other stakeholders on what needs to be done or improved in specific processes given some anomalies detected. this anticipatory nature is helpful in increasing the effectiveness of processes before the migration phase is carried out. continuous monitoring and data collection objective incorporate changes in monitoring and data gathering to acquire up-to-date information on points of contact after the migration. ai integration after migration ai-enabled monitoring tools continue to capture and process the real-time production data of the new cloud-based processes. this real-time feedback loop helps the enterprise to detect adverse conditions, track it performance and tackle issues. actionable data insights objective give recommendations based on the findings of the analyzed data for the decision making and business processes optimization. ai integration to help those in charge make decisions, the ai system prepares reports and dashboards for the decision-maker. such insights may also involve suggestions for increased efficiency or efficiency-enhancing adjustments of tangible and intangible resources and workflows. user experience enhancement objective the overall usability of the applications should also be enhanced by demonstrating effective control over the workflows to the end-users. ai integration user behaviors, traits, and issues are predictable based on the data collection of users’ interactions with the system. this information is used to make decisions for improvement in user experience such as, reducing time response, minimizing latency issues and integrating into cloud environment seamlessly. iterative improvement objective develop a cyclical improvement model wherein constant checking of processes results in their subsequent optimization. ai integration ai evolves from new data that is fed through the system, and its analytics and suggestions change accordingly. this idea means that the workflows will always be optimized because of constant iteration, and the organization will be able to adequately change and fit the business needs. data analytics services using ai on the existing or migrated workflows provides the enterprise with valuable insights about the constant workflow within the enterprise to improve the operations of the enterprise and provide better and efficient user experience in a pre and post cloud migration scenarios. api usage optimization with ai importance of api usage optimization objective understand the importance of proper management of the costs related to cloud services by improving the usage of apis. ai integration the enterprise also incorporates ai algorithms in the api management system to find ways of limiting api usage and controlling costs according to the enterprise’s budgetary plans. ai-driven usage analysis objective historical trends that help in identifying the density of usage and peak api usage and troughs. ai integration through machine learning, the api usage history is studied in order to establish patterns, the hours with the highest frequency of api calls, or the time of day with the lowest frequency. it is with such pertinent information that strategic interventions to enhance the utilization of apis as well as the costs related to them are premised. best times for api usage objective find out when the apis are most likely to be used in a way that will allow one to take advantage of the cheaper pricing models while at the same time reducing costs. ai integration the ai system uses predictive models to decide when it is most effective to use the apis. this includes factors like the price difference between the day and night, the load on the cloud provider or resources, and the previous usage of the api for efficiency. negotiation of pricing based on demand objective minimize costs through standardization of api costs pa ge 21 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 depending on the usage and through flexibility that is common to the pricing models adopted by most cloud service providers. ai integration true demand of apis for a particular software is evaluated in real time and pricing is negotiated accordingly by ai algorithms. this may entail self-served negotiation with cloud services providers to arrive at lower prices corresponding to the existing demand situation, to achieve efficiency in cost. cost-efficient strategies objective cut expenditure in areas that can be avoided, for instance, the use of apis during late hours when they are not as resource intensive as during the day. ai integration the ai system advises and enforces procedures for api conformity with effective operations styles. this can mean calling non-critical apis during periods of low traffic, using cheaper available resources for selected operations, and managing resources according to predicted traffic. proactive cost prevention objective preventing unnecessary costs in an api can be as simple as anticipating the problems before they arise and finding ways to deal with them. ai integration apis are always being monitored in real-time using artificial intelligence driven tools, and any variations from the right usage parameters are instantly flagged out. thus, it is possible to avoid additional expenses related to working with drugs and maintain the budget plan of the organization. budget constraint alignment objective make sure that the utilization of apis does not go against the organization’s budgets to be financially responsible. ai integration api usage is constantly monitored against defined budgetary constraints where api usage patterns are automatically readjusted to ensure they do not exceed allowable parameters. this way the enterprise is able to ensure that it fosters optimal expenditure while trying to satisfy operational requirements. adaptive optimization strategies objective continuously monitor shifts in demand, continuously tweaking the api optimization methods to maintain this cost efficiency. ai integration optimization also takes demand into consideration and leverages machine learning to enhance a constantly changing set of recommendations for api usage. thus, flexibility guarantees that the organization can address alterations in operational needs and achieve costeffectiveness concurrently. ai in optimizing usage of apis encompass factors such as considering previous usage pattern of apis, determining the most appropriate times for utilization of apis, bargaining on price, integrating cost effectiveness measures, preventing unwanted costs, utilizing reasonable measures, and adjusting measures in relation to needs. this all-encompassing approach guarantees that api utilization reaches needed velocity and efficiency as the company completes its cloud journey. ai-driven cost optimization in cloud migration historical usage analysis objective to establish an awareness of the temporal patterns of the organization’s utilization of on-premises resources and applications, for benchmarking purposes. ai integration machine learning techniques are currently applied to historical data sources concerning resource usage, application behavior, and related costs. this historical perspective assists to detect cyclical patterns, during which utilities are utilized most intensively and where the distribution can be made in the most effective way. predictive resource needs objective the last step in the cloud planning process is to consider future resource needs to ensure the resources in the cloud correspond to the organization’s needs in the future. ai integration this is the process which is performed by machine learning models in analyzing historical data and patterns to come up with resource requirements in the future. using ai analysis on the trends of resource consumption patterns during peak use, seasonality, projected growth and expected variations, the forecasts required resources during and after migration are effectively predicted. cost-effective configuration recommendations objective suggest the identified favorable states for cloud resources to maximize resource utility and minimize costs. ai integration ai algorithms compare different possibilities which are available within the cloud infrastructure considering factors like instance types, possible storage types, and networking options. the system come up with suggest of ideal configurations given the organization performance goals in relation to cost. pa ge 22 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 real-time cost monitoring objective another measure for proper cloud usage is the constant tracking of costs in order to identify any discrepancies and make modifications immediately. ai integration a continuous monitoring system facilitated by ai processes information regarding cloud utility and related expenses in real-time. the system is useful in warning at least the stakeholders in case there are unwarranted hikes in resource costs or if the actual resource usage trends are outside expectations. budget allocation optimization objective make sure that the enterprise spends the budget funds wisely towards various aspects of the migration to avoid specific costs that may exceed the budget. ai integration the ai algorithms help in providing suitable recommendations for the distribution of the budget resources because of the known cost impact of the migration activities. these are data transfer costs, data storage, instance costs, and any other costs related to a certain component of the big data solution; to guarantee that these expenses do not exceed the allocated budget. cost-benefit analysis objective prepare a comprehensive cost plan and benefit calculation to determine the roi in the migration project. ai integration having adopted ai to generate reports, an analysis of the costs of migration and the benefits expected is made. such include performance, capacity, and functionality upgrades as well as simplified working or day-today operations. it helps the decision-makers to identify the extent of the success of the migration in respect of the cost factors. adaptive cost optimization strategies objective adopt solution approaches that adapt to resource demand and configuration changes over time in a system. ai integration most ai systems are adaptive in nature and always learn from the patterns of everyday usage; the optimization strategies change all the time. thus, this capacity ensures that the organization can work in an efficient manner regarding workloads that may be dynamic or new applications or changes in the business environment. reducing the cost of migration to the cloud help to optimize the use of resources through analysis of past data, statistical modelling, and monitoring of migration processes to ensure that resources are utilized effectively, and that cost does not accumulate beyond a certain limit. discussion the paper entitled “artificial intelligence-based cloud planning and migration to cut the cost of cloud” offers a detailed discussion of cloud planning and migration with the help of artificial intelligence to realize the cost-reducing and time-saving effects of cloud service. this paper adds to the existing literature by discussing the concrete types of ai technologies and algorithms that can be used to address different aspects of cloud migration directly. in comparison to other research like the one made by (houssein et al., 2021), this paper extends prior discussion on the dynamic allocation of resource using ai by identifying the precise ai determination tree, neural network, and cluster algorithms to improve the cloud migrating effort. this provides a level of detail that is not common in general discussions on cloud migration frameworks as espoused in the literature by authors such as (thanka et al., 2029). furthermore, as suggested by the findings of (vähäkainu et al., 2022), most research focuses on cost advantages of ai in cloud migration whereas the above-mentioned document presents a holistic view as it also discusses scalability and security aspects. it connects the academia discourse with the practitioner’s perspective, accompanied by pseudocode and a concrete systematic approach on how to apply these ai-tools in cloud migrations, which is not always shown in existing literature. conclusion this paper entitled “artificial intelligence-based cloud planning and migration to cut the cost of cloud” offer a detailed analysis of how ai can complement cloud migration process. in particular, the incorporation of ai into planning and execution phases also proves the possibilities of rationalizations with lower costs and enhanced effectiveness within cloud contexts. the focus of the paper is on the use of complex ai-based methods involving such tools as machine learning and natural language processing to automate and enhance the migration process. they allow for better resource allocation, forecast future needs and optimize the general handling of cloud resources. ai also assists in cutting down the costs of migrating to the cloud, improving security, and creating operational efficiency. in addition, the authors provide detailed pseudocode to show how these ai techniques can be applied in real life, making it easier for the reader to develop an implementation plan for these strategies within his/ her workplace. this approach connects academic theory and real-world usage, which makes the book a useful reference for it practitioners who work in cloud infrastructure. in conclusion, the paper contributes to the existing literature on cloud migration and highlights how ai can revolutionize this area. it calls for the continued pa ge 23 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 investigation and use of ai-based solutions to enhance the efficiency, security, and affordability of the cloud. references alhilali, a. h., & montazerolghaem, a. (2023). artificial intelligence based load balancing in sdn: a comprehensive survey. internet of things. advance online publication. https://doi.org/10.1016/j. iot.2023.100814 bermejo, b., & juiz, c. (2023). improving cloud/ edge sustainability through artificial intelligence: a systematic review. journal of parallel and distributed computing, 176, 41-54. https://doi.org/10.1016/j. jpdc.2023.02.006 bian, y. j., xie, l., & li, j. q. (2022). research on influencing factors of artificial intelligence multicloud scheduling applied talent training based on dematel-taism. journal of cloud computing, 11(1), 35. https://doi.org/10.1186/s13677-022-00315-4 dhaya, r., & kanthavel, r. (2022). ioe based private multi-data center cloud architecture framework. computers and electrical engineering, 100, 107933. elmagzoub, m. a., syed, d., shaikh, a., islam, n., alghamdi, a., & rizwan, s. (2021). a survey of swarm intelligence based load balancing techniques in cloud computing environment. electronics, 10(21), 2718. https://doi.org/10.3390/electronics10212718 gill, s. s., tuli, s., xu, m., singh, i., singh, k. v., lindsay, d., ... pervaiz, h. (2019). transformative effects of iot, blockchain and artificial intelligence on cloud computing: evolution, vision, trends and open challenges. internet of things, 8, 100118. https://doi. org/10.1016/j.iot.2019.100118 hassan, m. b., ahmed, e. s., & saeed, r. a. (2024). green machine learning approaches for cloud-based communications. in m. b. hassan, e. s. ahmed, & r. a. saeed (eds.), green machine learning protocols for future communication networks 2024 (pp. 129-160). crc press. https://doi.org/10.1201/9781003230427-5 hemmati, a., raoufi, p., & rahmani, a. m. (2024). edge artificial intelligence for big data: a systematic review. neural computing and applications. advance online publication. https://doi.org/10.1007/s00521-02409723-w houssein, e. h., gad, a. g., wazery, y. m., & suganthan, p. n. (2021). task scheduling in cloud computing based on meta-heuristics: review, taxonomy, open challenges, and future trends. swarm and evolutionary computation, 62, 100841. janet, a., & al-turjman, f. (2023). the impact of cloud computing on the development of artificial intelligence technologies in e-commerce. neu journal for artificial intelligence and internet of things, 2(3). joloudari, j. h., alizadehsani, r., nodehi, i., mojrian, s., fazl, f., shirkharkolaie, s. k., kabir, h. d., tan, r. s., & acharya, u. r. (2022, march 28). resource allocation optimization using artificial intelligence methods in various computing paradigms: a review. arxiv preprint arxiv:2203.12315. https://doi. org/10.13140/rg.2.2.32857.39522 joloudari, j. h., mojrian, s., saadatfar, h., nodehi, i., fazl, f., alizadehsani, r., kabir, h. m., tan, r. s., & acharya, u. r. (2022, march 23). the state-ofthe-art review on resource allocation problem using artificial intelligence methods on various computing paradigms. arxiv preprint arxiv:2203.12315. https:// doi.org/10.1007/s11042-024-18123-0 junaid, m., shaikh, a., hassan, m. u., alghamdi, a., rajab, k., al reshan, m. s., & alkinani, m. (2021). smart agriculture cloud using ai based techniques. energies, 14(16), 5129. https://doi.org/10.3390/en14165129 kanungo, s. (2024). ai-driven resource management strategies for cloud computing systems, services, and applications. world journal of advanced engineering technology and sciences, 11(2), 559-566. https://doi. org/10.30574/wjaets.2024.11.2.0137 kumar, y., kaul, s., & hu, y. c. (2022, december 1). machine learning for energy-resource allocation, workflow scheduling and live migration in cloud computing: state-of-the-art survey. sustainable computing: informatics and systems, 36, 100780. https:// doi.org/10.1016/j.suscom.2022.100780 lee, d., & yoon, s. n. (2021). application of artificial intelligence-based technologies in the healthcare industry: opportunities and challenges. international journal of environmental research and public health, 18(1), 271. https://doi.org/10.52783/tjjpt.v44.i3.2466 liang, x., haiping, l., liu, j., & lin, l. (2021). reform of english interactive teaching mode based on cloud computing artificial intelligence – a practice analysis. journal of intelligent & fuzzy systems, 40(2), 3617-3629. https://doi.org/10.1177/16878132221122770 matthew, o. o., olabanji, d. o., & fitch, t. (2023, april 30). cloud-native architecture portability framework validation and implementation using expert system. international journal of advanced studies in computer science and engineering (ijascse), 12(4). mohanty, s. n., potluri, s., prakash, v. b., srinath, b., & manjunath, b. (2021, july 19). cloud security concepts, threats and solutions: artificial intelligence based approach. cloud security: techniques and applications, 1, 1. nagasundaram, s., bobinath, b. n., shedthi, a., rajalakshmi, k., & humnekar, t. d. (2023). analysis of the requirement and artificial intelligencebased resource management system in cloud. in 9th international conference on advanced computing and communication systems (icaccs 2023) (vol. 1, pp. 2516-2525). ieee. https://doi.org/10.1109/ icaccs57279.2023.10112940 nagy, m., lăzăroiu, g., & valaskova, k. (2023, january 28). machine intelligence and autonomous robotic technologies in the corporate context of smes: deep learning and virtual simulation algorithms, cyberphysical production networks, and industry 4.0-based manufacturing systems. applied sciences, 13(3), 1681. https://doi.org/10.3390/app13031681 pa ge 24 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 3(2) 13-24, 2024 nayak, a., patnaik, a., satpathy, i., & patnaik, b. c. (2024). data storage and transmission security in the cloud: the artificial intelligence (ai) edge. in a. nayak, a. patnaik, i. satpathy, & b. c. patnaik (eds.), improving security, privacy, and trust in cloud computing 2024 (pp. 194-212). igi global. https://doi.org/10.4018/9798-3693-1431-9.ch009 olabanji, d., fitch, t., & matthew, o. (2023). cloudnative architecture portability framework validation and implementation using expert system. international journal of advanced studies in computer science and engineering, 12(4), 1-4. reátegui, j. l., & herrera, p. c. (2021). artificial intelligence in the assessment process of moocs using a cloud-computing ecosystem. in 2021 ieee international conference on engineering, technology & education (tale) (pp. 487-493). ieee. https://doi. org/10.1109/tale51101.2021.9635760 rommer, d. (2020). artificial intelligence-based decisionmaking algorithms, industrial big data, and smart connected sensors in cloud-based cyber-physical manufacturing systems. economics, management, and financial markets, 15(1), 40-46. sharma, a., verma, a., malviya, r., & sekar, m. (2023, march 10). artificial-intelligence-based cloud computing techniques for patient data management. in a. sharma, a. verma, r. malviya, & m. sekar (eds.), artificial intelligence for health 4.0: challenges and applications (pp. 149-173). river publishers. https:// doi.org/10.1201/9781003373582-6 soni, d., & kumar, n. (eds.). (2023, march 16). artificial intelligence in cloud computing. in artificial intelligence in cyber-physical systems (pp. 51-73). crc press. thanka, m. r., uma maheswari, p., & edwin, e. b. (2019, september). an improved efficient: artificial bee colony algorithm for security and qos aware scheduling in cloud computing environment. cluster computing, 22(suppl 5), 10905-10913. tuli, s., gill, s. s., xu, m., garraghan, p., bahsoon, r., dustdar, s., sakellariou, r., rana, o., buyya, r., casale, g., & jennings, n. r. (2022, february 1). hunter: ai based holistic resource management for sustainable cloud computing. journal of systems and software, 184, 111124. https://doi.org/10.1016/j. jss.2021.111124 vähäkainu, p., lehto, m., kariluoto, a., & ojalainen, a. (2020). artificial intelligence in protecting smart building’s cloud service infrastructure from cyberattacks. in cyber defence in the age of ai, smart societies and augmented humanity (pp. 289-315). wang, w., zhang, y., liu, q., wang, t., & jia, w. (2024, april 3). edge-intelligence-based computation offloading technology for distributed internet of unmanned aerial vehicles. ieee internet of things journal. https://doi.org/10.1109/jiot.2024.3383896 yahia, h. s., zeebaree, s. r., sadeeq, m. a., salim, n. o., kak, s. f., adel, a. z., salih, a. a., & hussein, h. a. (2021, may). comprehensive survey for cloud computing based nature-inspired algorithms optimization scheduling. asian journal of research in computer science, 8(2), 1-6. zhang, y., & yuen, k. v. (2022, september). review of artificial intelligence-based bridge damage detection. advances in mechanical engineering, 14(9), 16878132221122770. https://doi.org/10.1177/16878132221122770 pa ge 1 pa ge 98 american journal of smart technology and solutions (ajsts) adopting lessons learned from global advanced manufacturing practices yasin mondi1* volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.4841 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: march 20, 2025 accepted: april 25, 2025 published: june 20, 2025 modern manufacturing experiences revolutionary changes through the integration of the internet of things, artificial intelligence, and large data analytics with additive manufacturing, thus achieving enhanced productivity and automated systems. the research evaluates both benefits and challenges of modern manufacturing with additional focus on productivity improvements and data-based choices. major implementation costs together with cybersecurity threats and system interoperability problems and required employee readjustment represent major implementation challenges. solving these problems demands purposeful funding and unified policy structures and must achieve alignment between industrial operators and academic institutions. new technological advances in quantum computing, 5g and edge computing systems enable the chance for considerable advancement. excellent integration requires standardized cybersecurity methods that show resistance to attacks. future investigations should concentrate on financial feasibility and staff expertise development and eco-friendly manufacturing approaches. cooperation between policymakers and industries is essential for the formulation of regulatory guidelines. this research highlights the necessity of reconciling innovation with organizational preparedness, notwithstanding the restrictions of data availability and advancing technology. effective adoption of industry 4.0 can propel sustainable industrial transformation and enhance global competitiveness. keywords artificial intelligence, computer security, cyber security, data science, internet of things, technology 1 ege university, department of chemical engineering, izmir, turkey * corresponding author’s e-mail: yasin.mondi@outlook.com introduction economic expansion together with technological improvement and societal advancement results from manufacturing activities which have been essential since ancient times. the industry experienced substantial improvements during the past decades because of quick globalization as well as technological progress and rising customer needs about quality alongside customization and sustainability (wolniak & grebski, 2023). global business competition requires advanced manufacturing which describes new production technologies and innovative methods to ensure competitiveness in today’s rapidly changing world economy. countries which implemented successful advanced manufacturing practices achieved better efficiency and productivity and better worldwide market placement. businesses together with nations require essential knowledge from successful global practices to stay leading in industrial advancement (javaid et al., 2024). the advanced manufacturing concept merges contemporary technologies which include automation, artificial intelligence (ai), robotics, additive manufacturing (3d printing) and the industrial internet of things (iiot) (okokpujie, & tartibu, 2024). executive manufacturing technologies lead to increased accuracy while boosting manufacturing pace and minimizing economic operations expenses. smart factories built with interconnected systems along with real-time data analytics techniques now transform classical manufacturing facilities. industrial revolution 4.0 establishes the transformation by generating smooth machine interoperability which optimizes supply chain operations while minimizing waste through automated predictive servicing and automated procedural controls (cheah et al., 2022). advanced manufacturing holds vital significance because of multiple international marketplace developments that both support economic sustainability and market competitiveness. smart manufacturing platforms based on digital technologies have become prevalent in established countries across the united states, germany and japan (sahoo & lo, 2022). production line development through automation and robotics technology diminishes human mistakes while boosting operations. manufacturers worldwide are adopting sustainable production methods to protect the environment because these methods resolve issues regarding greenhouse gas emissions and depleted resources and waste control. the adoption of green technologies, such as energy-efficient machinery and renewable energy integration, underscores the shift toward sustainable industrialization (al-rasheed, 2024). successful implementation models of advanced manufacturing come from nations who initially developed these practices. advanced manufacturing combined with artificial intelligence analytics at the hands of the united states serves to optimize operations and increase productivity levels (plathottam et al., 2023). through lean manufacturing principles japan has established worldwide standards in the areas of waste minimization and process enhancement and continuous enhancement. the german production sector demonstrates automation’s success when matched with human operator experience because of its reputation for producing high-precision technology. pa ge 99 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 significant developments in advanced manufacturing are seen in china and south korea because these economies spent resources on robotics and produced intelligent factories alongside artificial intelligence in their production systems (sahoo & lo, 2022) (table 1). table 1: overview of advanced manufacturing practices theme key technologies lobal examples benefits challenges digital transformation iot, i, robotics, additive manufacturing (3d printing), iiot usa: ai-driven optimization; germany: high-precision automation; japan: lean manufacturing increased accuracy, faster production, cost reduction high implementation costs, workforce resistance smart factories real-time data analytics, cyber-physical systems (cps) china/s. korea: smart factories with ai and robotics enhanced productivity, predictive maintenance interoperability issues, legacy system integration sustainability energy-efficient machinery, renewable energy integration eu/japan: green manufacturing initiatives reduced waste, lower emissions high upfront investment, regulatory complexity workforce adaptation micro-credentialing, government-funded training germany: industryacademia collaboration skilled labor readiness employee fears of job displacement policy & collaboration public-private partnerships, r&d funding global: horizon 2020 (eu). sme subsidies innovation ecosystems, knowledge transfer fragmented standards, regional disparities as much as organizations gain advantages from advanced manufacturing practices, their implementation poses several significant hurdles. the implementation of modern manufacturing methods encounters challenges from both cultural aspects inside organizations and company structures. the major obstacles blocking the implementation of advanced manufacturing comprise workforce opposition to change, the deficit of qualified personnel and employee concerns about technological displacement through automation (leesakul et al., 2022). few small and medium-sized enterprises encounter difficulties when they invest money for infrastructure modernization alongside new technology integration. the necessary action includes leaders from government and industries to support workforce training while providing financial incentives for technology use and developing supportive regulations (shan & ji, 2024). cultural and regional factors require organizations to modify selected global best practices for localization purposes (guarini et al., 2022). duplicate implementations of international successful practices remain sub optimal if they do not receive alterations which fit nationwide characteristics. practices require modification to workforce competencies and regulatory elements while market requirements to achieve optimal performance. the adoption of advanced manufacturing technology receives support from academia-industry-government collaborations which enable knowledge transfer and drive innovation for developing appropriate policies to establish a favorable manufacturing environment (shaheer, 2024). professional innovation ecosystems consisting of research facilities together with technology suppliers and industrial operators work as fundamental drivers of manufacturing development (matt et al., 2021). the advancement of advanced manufacturing practices requires government agencies and industries to fund studies through research and development programs and establish innovation centers so they can promote collaborative public-private sector adoption. the adoption of continuous learning combined with technological adaptation helps nations establish their position as top manufacturers in worldwide markets (kinkel et al., 2022). this paper explores worldwide advanced manufacturing practices through concept analysis alongside the presentation of strategic adoption strategies. this study includes essential developmental analysis and supporting evidence followed by market obstacles before providing useful benchmarks that benefit industrial sectors and governmental agencies. this work adopts a comprehensive research design that includes literature study then methodology before showing important outcomes before giving implementation suggestions for stakeholder manufacturing sustainability achievements. materials and methods study design and search strategy a mixed-methods design was implemented by the study to synchronize qualitative and quantitative investigations about global advanced manufacturing practices. qualitative research consists of manufacturing leader case studies and the quantitative part analyzes industry reports along with statistical information. the research methodology uses peer-reviewed journals together with government publications and industry white papers pa ge 10 0 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 obtained from scopus, pubmed, and google scholar. as a research guide the terms “advanced manufacturing” combine with “industry 4.0” and “smart manufacturing” and “automation” to direct the search through textual resources. the study accepts recent research (from the past ten years) examining technological adoption together with economic impact as well as implementation challenges. the analysis of selected data compares methods to discover recommended strategic elements and primary performance metrics and effective practices. inclusion and exclusion criteria the review analysis incorporated 245 published studies. the assessment included studies focusing on international advanced manufacturing methods alongside economic effect assessments and business-wide practical applicability. the research focused primarily on industry 4.0 together with automation along with robotics and sustainability topics. the review accepted empirical studies together with systematic reviews along with case studies to portray manufacturing adoption challenges and opportunities. research was excluded when it provided only theoretical analysis without practical application or when composed without empirical data and published beyond ten years or when written in non-english and when it repeated other studies. research about papers that either lacked full text viewing capabilities or delivered inadequate connection to main research goals was eliminated from analysis. selected studies 15 studies passed through the filtering process as the most appropriate resources for this research investigation. the studies adopted prisma guidelines before going through comprehensive selection processes starting from title review to abstract review and ending with full-text review. the chosen research papers supply essential knowledge about advanced manufacturing practices adoption status while showing their economic results and advanced technological innovations. researchers examine both industrial implementation of industry 4.0 technology and challenges of automation together with sustainable strategies for manufacturing while assessing global manufacturing competitiveness. research findings are contextualized by the selected studies while these findings enable the development of strategic recommendations regarding the adoption of international best practices in various industrial environments. visualization of study selection process is illustrated in figure 1. figure 1: prisma flowchart of study selection pa ge 10 1 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 data extraction and analysis the research employed the prisma guidelines as a systematic approach for transparent data extraction. a qualitative content analytical method searched for important themes which included sustainability together with digital transformation and manufacturing efficiency. the researchers verified their findings through inter-study comparison before grouping recurrent research questions into practical solution-oriented insights. this research design clears up the connection between advanced manufacturing techniques and their practical effects on productivity growth and industry competitiveness worldwide. results and discussions multiple research studies detail the extensive adoption of industry 4.0 technology together with its business-related effects across multiple sectors of manufacturing (table 2). modern production systems along with automation and digital transformation have significant implementation trends and adaptation details in line with upcoming opportunities and confrontations. table 2: summary of key findings from the reviewed studies s.no author name and year type of study application technique solution usage 1 zhong et al., (2017) review study analysis of intelligent manufacturing, iot-enabled manufacturing, and cloud manufacturing integration of iot, cps, cloud computing, bda, and ict for intelligent manufacturing understanding industry 4.0, governmental and corporate strategies, future challenges, and research directions 2 dilberoglu et al., (2017) review study analysis of additive manufacturing (am) technologies advances in material science, process development, and design considerations in am classification of current knowledge and technological trends in am for industry 4.0 3 frank et al., (2019) empirical study (survey) survey of 92 manufacturing firms on industry 4.0 technology adoption conceptual framework dividing technologies into front-end (smart manufacturing, smart products, smart supply chain, smart working) and base technologies (iot, cloud, big data, analytics) understanding adoption patterns, technology layers, and challenges in implementing industry 4.0 technologies in manufacturing 4 almadalobo, (2015) review study examination of smart manufacturing systems in industry 4.0 conceptual framework and demonstrative scenarios (smart design, machining, control, monitoring, scheduling) identifying key technologies, applications, challenges, and future perspectives for smart manufacturing systems 5 ghobakhloo, (2018) systematic review systematic and content-centric literature review using ibm watson nlp identification of 12 design principles and 14 technology trends for industry 4.0; development of a strategic roadmap assisting manufacturers in transitioning to industry 4.0 by offering a structured guide for implementation 6 arden et al., (2021) review study application of iot, ai, robotics, and advanced computing in pharmaceutical manufacturing enhancing agility, efficiency, flexibility, and quality in drug production understanding regulatory, technical, and logistical barriers to achieving industry 4.0 in pharmaceutical manufacturing 7 sanders et al., (2016) conceptual study analysis of industry 4.0’s role in lean manufacturing identification of industry 4.0 technologies that address lean manufacturing barriers bridging the gap between industry 4.0 and lean manufacturing, demonstrating that industry 4.0 can enable lean production pa ge 10 2 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 8 ashima et al., (2021) theoretical study integration of iot with additive manufacturing (am) enhancing am reliability, efficiency, and scalability for mass production improving am production processes, reducing waste, and meeting customer specifications in industry 4.0 9 sahoo & lo, (2022) review study analysis of smart manufacturing adoption and strategies in five major countries integration of ai, iot, vr/ar, big data, and am for smart manufacturing understanding smart manufacturing implementation, challenges, inspection methods, and future prospects in industry 4.0 10 lu et al., (2020) review study examination of manufacturing automation standards for smart manufacturing integration of endto-end manufacturing processes with automation standards improving efficiency, interoperability, and responsiveness in smart manufacturing systems 11 veile et al., (2020) empirical study (interviews) 13 semi-structured interviews with industry 4.0-experienced managers in german manufacturing companies development of industry 4.0-specific know-how, financial resources, employee integration, openminded corporate culture, planning, partnerships, data security providing concrete lessons for industry 4.0 implementation and deriving recommendations for future research 12 mittal et al., (2020) case study analysis multiple case studies of smes adopting smart manufacturing (sm) development of an sme-specific ‘sm adoption framework’ with five vital steps helping smes transition to sm by identifying data, assessing readiness, raising awareness, defining vision, and selecting tools 13 ghazilla et al., (2015) empirical study (delphi survey) three-round delphi survey with experts on green manufacturing in smes in malaysia identification of key drivers and barriers to green manufacturing adoption in smes helping smes transition to green manufacturing by prioritizing factors influencing adoption 14 kurpjuweit et al., (2021) empirical study (delphi & in-depth interviews) exploration of block chain integration in additive manufacturing (am) enhancing ip rights management, lifecycle monitoring, process improvements, and data security in am improving am competitiveness, enabling decentralized manufacturing, enhancing supply chain visibility, and reducing logistics costs 15 belhadi et al., (2022) hybrid study (focus groups & case studies) additive manufacturing (am) for supply chain resilience & efficiency development of ambidextrous dynamic capabilities through am, enabling resilienceefficiency balance enhancing global supply chain resilience, efficiency, and preparedness for the post-covid era these studies analyze business operational integration of industry 4.0 technologies according to their focus on iot, ai, cloud computing, cps and big data analytics. digital technological implementations build up manufacturing capabilities by creating operational effectiveness and productivity improvements as well as better decisionmaking capabilities. the combination of modern technologies makes it possible to execute permanent tracking along with machine predictive forecasting and data-based decision-making which leads to better resource efficiency and shorter stoppages. organizations achieving successful industry 4.0 transformation need well-organized implementation methods focusing on smart manufacturing combined with smart products and smart supply chains and smart working spaces. organizations need to separate digitalization strategies from practical applications to maximize their use of industry 4.0 solutions (table 3). pa ge 10 3 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 table 3: sector-specific industry 4.0 adoption outcomes sector key technologies impact challenges supporting studies pharmaceuticals ai, iot, blockchain real-time quality control, regulatory compliance (e.g., drug formulation defect detection) data security, validation complexities arden et al. (2021), veile et al. (2020) aerospace additive manufacturing, iot 40% material waste reduction, lightweight component production certification hurdles, high production costs. dilberoglu et al. (2017), ashima et al. (2021) automotive cps, big data analytics agile production, real-time defect detection (e.g, predictive maintenance) legacy system interoperability lu et al. (2020), sanders et al. (2016) smes modular automation, cloud computing cost-effective scalability (e.g., resource optimization) limited funding, digital skills gap. mittal et al. (2020), ghazilla et al. (2015) the fundamental role of automated systems in smart manufacturing facilities generates optimized industrial operations through better efficiency and better flexibility. research demonstrates that advanced automation technology enhances design activities alongside production control operations and machining techniques as well as monitoring needs and scheduling. ai-based automation helps organizations boost operational efficiency at the same time as reducing operational obstacles. automated systems fail to connect because they lack jointly used operating procedures and interoperability requirements. unified communication standards are essential elements that make operations efficient and create system compatibility and automated system integration possible. technology advances make industrial automation more efficient because they can replicate sophisticated processes and boost system precision as well as real-time choice speeds. the success of industry 4.0 depends mainly on additive manufacturing since this technology lets producers make advanced products which unite personalized features with enhanced material performance. the fabrication of complex items with customized outputs achieved through additive methods becomes economical compared to traditional production processes. the method proves beneficial to sustainability according to scientific studies since it reduces production waste while enabling local manufacturing capabilities (figure 2). additive manufacturing and iot technology work together to produce better reliability and increased efficiency as well as scalability through real-time monitoring systems that increase production speed and reduce time-based issues. manufacturers now achieve faster production and enhanced market reaction because of modern technological innovations in the field. iot needs solutions for material constraints and solution challenges as well as high costs of implementation before large-scale adoption can happen (table 4). table 4: sector-specific industry 4.0 adoption outcomes factor role barrier solution cited studies real-time data analytics enables predictive maintenance (e.g., reducing downtime by 30%) data silos in legacy systems digital twin adoption for interoperability zhong et al. (2017), almadalobo (2015) standardized protocols facilitates machine communication (e.g., opcua in smart factories) lack of global standards policy-industry collaboration (e.g., eu’ horizon 2020) lu et al. (2020), sahoo & lo (2022) employee training reduces resistance (e.g., upskilling for ai-driven automation) high training costs/time micro-credentialing programs, government subsidies. veile et al. (2020), leesakul et al. (2022) pilot projects demonstrates roi (e.g., german sme automation pilots) scalability risks phased roadmaps aligned with long-term goals ghobakhloo (2018), mittal et al. (2020) industry 4.0 functions as a fundamental driving force for industrial advancement of modern times alongside lean manufacturing principles. research establishes that smart technology systems eliminate manufacturing constraints because they enhance operational productivity with lower unnecessary cost basis. businesses using real-time data monitoring technologies build optimal resource networks that shorten manufacturing periods and deploy production methods that are flexible and budget friendly. business organizations benefit from ai prediction analysis for market requirement forecasting and manufacturing operation scheduling. the successful implementation of pa ge 10 4 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 ambitious industrial 4.0 initiatives depends on detailed preparations and standard operating guidelines together with effective staff training for adopting new systems. smart technologies used in pharmaceutical production yield multiple advantages by improving operational efficiency and ensuring better quality control as well as enhancing agile performance. quality control automation enables team members to maintain products that match regulatory requirements by applying real-time data analysis. for pharmaceutical companies to gain full potential from industry 4.0 they must resolve integration challenges and handle security matters along with conforming to legal standards. the known transformative potential of industry 4.0 meets numerous barriers that lead organizations to hold back from its adoption. multiple tests demonstrate that cybersecurity threats stand along with implementation expenses and employee reluctance to implement changes as the chief obstacles. enterprise infrastructure modernization costs along with employee training expenses prove difficult for both small and mediumsized businesses and multiple organizations to maintain. digital maturity levels within individual industries cause obstacle when organizations try to adhere to adoption processes. achieving successful digital transformation requires established regulatory structures as well as full funding support together with service collaboration between industrial groups with government departments and academic departments. industry 4.0’s future success relies on resolving obstacles with improved technology and tactical policies made for industries and inter-industry team coordination. organizations need standardized direction to achieve their targets of fully automated intelligent production systems. various studies show that standardized flexible assessment tools need to be developed in order to successfully scale industry 4.0 throughout all industrial domains. quantum computing coupled with 5g networking and edge computing systems provide modern solutions through which data processing speed is enhanced alongside decentralized management capabilities. the implementation of sustainable digital transformation needs active coordination between commercial businesses together with educational institutions and government agencies. widespread adoption of industry 4.0 combined with its maximum potential utilization stands vital for preserving global market leadership and industrial innovation advancements (table 5). figure 2: industry 4.0 enablers vs. barriers table 5: emerging technologies in industry 4.0 technology current use future potential adoption challenges study references quantum computing optimizing supply chain models real-time complex simulation (e.g., material science) immature infrastructure sahoo & lo (2022), plathottam et al. (2023) 5g networks high-speed iot connectivity autonomous robotics with <1ms latency cybersecurity vulnerabilities lu et al. (2020), kurpuweit et al. (2021) edge computing localized data processing for predictive analytics distributed ai (e.g., real-time quality control) legacy system integration zhong et al. (2017), ashima et al. (2021) digital twins virtual factory prototyping energy optimization via lifecycle modeling high fidelity data requirements belhadi et al. (2022), frank et al. (2019) discussion academic research investigates the development and obstacles related to industry 4.0 through assessments of its main influences on different manufacturing fields. studies show that digitalization with automation leads to enormous industrial changes which boost operational performance and flexibility and environmental friendliness. multiple industry 4.0 technologies receive analysis in research because they unite to optimize automated production systems through the internet of things (iot), artificial pa ge 10 5 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 intelligence (ai), big data analytics and cyber-physical systems (cps) and additive manufacturing systems. the adoption of these technologies enables industries to develop efficient operations as well as waste minimization and production adaptability for the achievement of lasting sustainable and intelligent manufacturing methods (santos et al., 2024). real-time monitoring paired with predictive maintenance services delivered through industry 4.0 operates as a major advantage for manufacturing organizations to boost their performance levels (keleko et al., 2022). the integration of iot systems and cps networks allows machines to connect to production lines for data-based operational optimization (zhang et al., 2018). the research by zhong et al. (2017) explains how manufacturing systems operated with cloud platforms distill big data into better production results that reduce interferences and optimize system functions. real-time data handling capabilities empower industries to take in advance measure on system failures which results in decreased operational interruptions and reduced maintenance expenses. frank et al. (2019) performed research proving that implementing industry 4.0 technologies results in businesses obtaining higher productivity levels alongside enhanced operational efficiency. huge cost investments are needed for the complete deployment because they include developing digital assets alongside employee training and implementing detailed data security features. organizational and cultural evaluations form a critical requirement for implementing digital manufacturing because they enable companies to achieve digital transformation success. industrial 4.0 requires additive manufacturing innovation as one essential element which gives manufacturers new design capabilities and waste minimization features and customized product features (valamede & akari, 2021). manufacturing technology generates complex assemblies that traditional production methods cannot replicate because of their existing manufacturing constraints. dilberoglu et al., (2017) conducted research on how advanced technologies create complex manufacturing products that result in sustainable industrial production systems. organizations that use additive manufacturing acquire durable lightweight components through which they create products for aerospace applications along with healthcare devices and automotive solutions (shrivastava & rathee, 2022). the connection between internet-connected systems and additive manufacturing permits ashima et al. (2021) to boost control strategies for production and operational reliability measures. manufacturers can maintain product quality through real-time data monitoring (wuest et al., 2014) because this system lets them make on-the-fly adjustments of parameters to lower material waste levels. manufacturers can achieve their best operational results by receiving real-time monitoring data and feedback which guarantees they produce personalized products to fulfill buyer requirements. three main barriers prevent the widespread use of iot in manufacturing: material limitations, high costs of production and limitations in quality control procedures (yang et al., 2018). technological evolution demands immediate solutions to these critical issues for the wide-scale implementation of technology. various research finds ways in which industry 4.0 and lean manufacturing principles connect. the waste elimination framework of lean manufacturing receives improvement from industry 4.0 technologies alongside process optimization strategies. sanders et al. (2016) demonstrate that industry 4.0 innovations solve regular lean manufacturing issues through better production output and decreased waste and improved market adaptability. operations achieve next-level precision along with enhanced agility through ai-powered automation and smart sensors and real-time analytic technologies. organizations that employ ai predictive analytics will forecast consumer demand better while optimizing their resource distribution which eliminates surplus stock and avoids manufacturing logjams. almada-lobo (2015) established smart manufacturing systems as a solution that helps production facilities develop flexible operational plans to adapt their output with market changes. the analysis of real-time production data through automated scheduling systems enables them to modify manufacturing workflows which results in maximum resource efficiency. the implementation of successful industry 4.0 depends on strategic planning together with interoperability frameworks and standardized protocols and workforce readiness to support lean manufacturing integration. ideally industries should spend money on employee training initiatives to establish guidelines which support complete technological adoption. industry 4.0 technologies have wrapped pharmaceutical manufacturing with new capabilities that boost manufacturing efficiency together with quality control and regulatory adherence. the pharmaceutical manufacturing sector leverages automation for two main reasons: first to reduce human mistakes along with secondly to achieve uniformity and enhance facility output levels. the combination of iot and ai and robotics creates optimized pharmaceutical manufacturing processes that function with minimal human interaction to produce consistent products according to arden et al. (2021). current quality control systems that use ai technology perform constant monitoring to detect issues right away thus they boost reliability levels. systems that use ai automation enable the analysis of drug formulation microscopic defects thereby meeting requirements set by regulatory standards. the sector requires technology providers to unite with regulatory agencies together with industry stakeholders because forthright implementation demands both data security along with regulatory compliance support. digital transformation in pharmaceuticals requires compliance with strict requirements as well as maintenance of safe treatment practices and untampered data security protocols. block chain technology provides the solution to supply chain transparency and secure pa ge 10 6 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 data sharing through its implementation for supply chain management. industry 4.0 faces multiple obstacles when organizations attempt its implementation. the findings show that high implementation costs (demirkesen et al., 2022) represent the main obstacle companies’ face. smart manufacturing implementation demands companies to spend considerable funds in digital infrastructure along with automation technology systems and employee skill upgrading. the challenge proves difficult for small and medium-sized enterprises mainly due to their constrained budgets. according to ghobakhloo (2018), the development of an industry 4.0 implementation framework is necessary because a systematic planning approach allows companies to maximize their outcomes while reducing implementation costs. the practice of introducing industry 4.0 through controlled pilot projects first helps organizations develop more efficient and enduring digital transitions while keeping their costs affordable. the worldwide distribution of digital maturity reveals that some nations deal with minimal funding opportunities combined with insufficient policy structures and lack of qualified staff (sahoo & lo, 2022). to fill these gaps between digital transformation needs and funding struggles governments and industry leaders need to work side by side and create stimulating policies supported by financial assistance programs for struggling businesses. the adoption of industry 4.0 encounters a substantial obstacle in cybersecurity concerns (ervural et al., 2017; yang et al., 2019). the connections between industrial systems through iot and cloud-based platforms create economic vulnerabilities together with network security problems. operating manufacturing facilities has become more dangerous due to cyber-based threats and data security breaches alongside system disruptions that threaten production operations. lu et al. (2020) support the implementation of universal security protocols to defend data authenticity and factory system operational reliability. an effective cybersecurity strategy must include multiple components starting with encryption protocols through real-time threat detection systems with access control functions. security architectures that deploy artificial intelligence systems with block chain technology coupled with multiple layers create enhanced protection for digital manufacturing security. organizational resistance against cyber-attacks improves when employees undergo training about cybersecurity practices which minimize security vulnerabilities. employer mistakes function as keys to cybersecurity threats in organizations which requires robust employee training for digital security norms. a standardized approach enables the continuous functioning of industry 4.0 operations by defending their free flow. standardization frameworks must be developed to achieve successful system integration since they build a unified digital ecosystem through manufacturing system links. organizations encounter technical problems when attempting to link their current legacy systems to new digital technology platforms because this integration causes functional obstacles between system applications. smart manufacturing developers require universal data exchange formats that must be established through joint efforts by technical experts and regulatory bodies combined with industry manufacturers. industrial operational speed will increase and business information will unite while digital transformation reaches all sectors through standardized interoperability frameworks. to achieve success in industry 4.0 a set of universal standards needs to define protocols for machine communications along with data exchange parameters and automation specifications for current interoperability requirements. through digital twin technology manufacturers can generate virtual copies of operational procedures to resolve interoperability issues by providing real-time management tools for optimization benefits. various manufacturing industries will find success through industry 4.0 based on their ability to manage the equilibrium between innovation and regulatory compliance. conclusion manufacturing experiences an industrial transformation through the combination of advanced technologies which includes iot together with ai and big data analytics and additive manufacturing. the researched documents highlight key benefits of smart manufacturing which include better efficiency and automated systems and data-based decision capability. notwithstanding these benefits, obstacles include elevated installation expenses, cybersecurity threats, interoperability concerns, and workforce adjustment impede extensive use. confronting these challenges necessitates strategic investments, policy frameworks, and coordination across industries, academics, and governmental entities. future developments in quantum computing, 5g connectivity, and edge computing possess the capacity to enhance manufacturing processes. standardization initiatives and cybersecurity protocols will be essential for facilitating smooth integration and enhancing resilience. through the promotion of innovation and skill enhancement, industry 4.0 may facilitate sustainable industrial change and uphold global competitiveness. the shift to smart manufacturing should be undertaken comprehensively, aligning technology innovations with regulatory and organizational preparedness to optimize the advantages of industry 4.0. recommendations future research must concentrate on augmenting cybersecurity frameworks, devising economical solutions, and refining workforce training for the implementation of industry 4.0. cooperation between industries and policymakers is essential for the establishment of standardized regulations. moreover, the research of ai, blockchain, and sustainable manufacturing techniques can improve efficiency, security, and environmental sustainability in smart manufacturing systems. pa ge 10 7 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 study limitation this study is constrained by the accessibility of data from specific sources, possible biases in the literature studied, and the dynamic characteristics of industry 4.0 technology. moreover, regional disparities and sectorspecific variances may influence the generalizability of the findings. future research should integrate more extensive datasets and empirical validations to enhance conclusions. references almada-lobo, f. (2015). the industry 4.0 revolution and the future of manufacturing execution systems (mes). journal of innovation management, 3(4), 16-21. al-rasheed, h. (2024). green innovation in energyintensive industries: adopting renewable energy technologies. international journal of green skills and disruptive technology, 1(1), 89-99. arden, n. s., fisher, a. c., tyner, k., yu, l. x., lee, s. l., & kopcha, m. (2021). industry 4.0 for pharmaceutical manufacturing: preparing for the smart factories of the future. international journal of pharmaceutics, 602, 120554. ashima, r., haleem, a., bahl, s., javaid, m., mahla, s. k., & singh, s. (2021). automation and manufacturing of smart materials in additive manufacturing technologies using internet of things towards the adoption of industry 4.0. materials today: proceedings, 45, 5081-5088. belhadi, a., kamble, s. s., venkatesh, m., jabbour, c. j. c., & benkhati, i. (2022). building supply chain resilience and efficiency through additive manufacturing: an ambidextrous perspective on the dynamic capability view. international journal of production economics, 249, 108516. cheah, c. g., chia, w. y., lai, s. f., chew, k. w., chia, s. r., & show, p. l. (2022). innovation designs of industry 4.0 based solid waste management: machinery and digital circular economy. environmental research, 213, 113619. dawson, m. (2018). cyber security in industry 4.0: the pitfalls of having hyperconnected systems. journal of strategic management studies, 10(1), 19-28. demirkesen, s., & tezel, a. (2022). investigating major challenges for industry 4.0 adoption among construction companies. engineering, construction and architectural management, 29(3), 1470-1503 dilberoglu, u. m., gharehpapagh, b., yaman, u., & dolen, m. (2017). the role of additive manufacturing in the era of industry 4.0. procedia manufacturing, 11, 545-554. 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. ghazilla, r. a. r., sakundarini, n., abdul-rashid, s. h., ayub, n. s., olugu, e. u., & musa, s. n. (2015). drivers and barriers analysis for green manufacturing practices in malaysian smes: a preliminary findings. procedia cirp, 26, 658-663. ghobakhloo, m. (2018). the future of manufacturing industry: a strategic roadmap toward industry 4.0. journal of manufacturing technology management, 29(6), 910-936. guarini, e., mori, e., & zuffada, e. (2022). localizing the sustainable development goals: a managerial perspective. journal of public budgeting, accounting & financial management, 34(5), 583-601. javaid, m., haleem, a., singh, r. p., & sinha, a. k. (2024). digital economy to improve the culture of industry 4.0: a study on features, implementation and challenges. green technologies and sustainability, 100083. keleko, a. t., kamsu-foguem, b., ngouna, r. h., & tongne, a. (2022). artificial intelligence and real-time predictive maintenance in industry 4.0: a bibliometric analysis. ai and ethics, 2(4), 553-577. khorasani, m., loy, j., ghasemi, a. h., sharabian, e., leary, m., mirafzal, h., ... & gibson, i. (2022). a review of industry 4.0 and additive manufacturing synergy. rapid prototyping journal, 28(8), 1462-1475. kinkel, s., baumgartner, m., & cherubini, e. (2022). prerequisites for the adoption of ai technologies in manufacturing–evidence from a worldwide sample of manufacturing companies. technovation, 110, 102375. kurpjuweit, s., schmidt, c. g., klöckner, m., & wagner, s. m. (2021). blockchain in additive manufacturing and its impact on supply chains. journal of business logistics, 42(1), 46-70. leesakul, n., oostveen, a. m., eimontaite, i., wilson, m. l., & hyde, r. (2022). workplace 4.0: exploring the implications of technology adoption in digital manufacturing on a sustainable workforce. sustainability, 14(6), 3311. lu, y., xu, x., & wang, l. (2020). smart manufacturing process and system automation–a critical review of the standards and envisioned scenarios. journal of manufacturing systems, 56, 312-325. matt, d. t., molinaro, m., orzes, g., & pedrini, g. (2021). the role of innovation ecosystems in industry 4.0 adoption. journal of manufacturing technology management, 32(9), 369-395. mittal, s., khan, m. a., purohit, j. k., menon, k., romero, d., & wuest, t. (2020). a smart manufacturing adoption framework for smes. international journal of production research, 58(5), 1555-1573. mishra, d., priyadarshi, a., das, s. m., shree, s., gupta, a., pal, s. k., & chakravarty, d. (2022). industry 4.0 application in manufacturing for real-time monitoring and control. journal of dynamics, monitoring and diagnostics, 1(3), 176-187. oh, y., park, h., yoo, a., kim, n., kim, y., kim, d., … yang, h. (2013, august 15). a product quality prediction model using real-time process monitoring in manufacturing supply chain. journal of korean institute of industrial engineers. korean institute pa ge 10 8 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 98-108, 2025 of industrial engineers. https://doi.org/10.7232/ jkiie.2013.39.4.271 okokpujie, i. p., & tartibu, l. k. (2024). study of the economic viability of internet of things (iots) in additive and advanced manufacturing: a comprehensive review. progress in additive manufacturing, 1-20. plathottam, s. j., rzonca, a., lakhnori, r., & iloeje, c. o. (2023). a review of artificial intelligence applications in manufacturing operations. journal of advanced manufacturing and processing, 5(3), e10159. sahoo, s., & lo, c. y. (2022). smart manufacturing powered by recent technological advancements: a review. journal of manufacturing systems, 64, 236-250. sanders, a., elangeswaran, c., & wulfsberg, j. (2016). industry 4.0 implies lean manufacturing: research activities in industry 4.0 function as enablers for lean manufacturing. journal of industrial engineering and management, 9(3), 811-833. santos, a. d. m., sant’anna, â. m. d. o., barbosa, a. s., becker, a. m., & ayala, n. f. (2024). multi-criteria decision-making model for sustainability functions integrated industry 4.0 technologies within small and medium enterprises in emerging countries. international journal of productivity and performance management. shaheer, m. (2024). impact of academia-associationindustry network collaboration on innovation and stakeholders’ value. shan, c., & ji, x. (2024). environmental regulation and green technology innovation: an analysis of the government subsidy policy’s role in driving corporate green transformation. industrial engineering and innovation management, 7(1), 39-46. srivastava, m., & rathee, s. (2022). additive manufacturing: recent trends, applications and future outlooks. progress in additive manufacturing, 7(2), 261287. veile, j. w., kiel, d., müller, j. m., & voigt, k. i. (2020). lessons learned from industry 4.0 implementation in the german manufacturing industry. journal of manufacturing technology management, 31(5), 977-997. wolniak, r., & grebski, w. (2023). the customization and personalization of product in industry 4.0. scientific papers of the silesian university of technology: organization and management series, 2023(180). yang, h., kumara, s., bukkapatnam, s. t., & tsung, f. (2019). the internet of things for smart manufacturing: a review. iise transactions, 51(11), 1190-1216 zhang, y., guo, z., lv, j., & liu, y. (2018). a framework for smart production-logistics systems based on cps and industrial iot. ieee transactions on industrial informatics, 14(9), 4019–4032. https://doi. org/10.1109/tii.2018.2845683 zhong, r. y., xu, x., klotz, e., & newman, s. t. (2017). intelligent manufacturing in the context of industry 4.0: a review. engineering, 3(5), 616-630. pa ge 1 pa ge 37 american journal of smart technology and solutions (ajsts) ai-powered cybersecurity: revolutionizing business threat detection and response prottoy khan1, md zahirul islam2, sazib hossain3* volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.4488 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: february 02, 2025 accepted: march 06, 2025 published: april 11, 2025 the modern day enterprise infrastructure needs cybersecurity as a crucial element to protect against increasing cyber threats that have multiplied because of digital business expansion. security technologies that exist conventionally manage certain threats decently but lose their effectiveness when new forms of sophisticated cyberattacks emerge. machine learning together with deep learning using anomaly detection methods enables artificial intelligence to function as an advanced security technology that boosts detection and response functions. the paper investigates how artificial intelligence cybersecurity systems modernize business defenses against threats and security incidents. an ai-based algorithm analyzes a dataset containing network logs and authentication trials together with encryption protocols and reputation scores of ip addresses to identify malicious occurrences. different machine learning models with both supervised classification approaches together with unsupervised anomaly detection methods undergo assessment for determining their threat identification capabilities. the analysis verifies how ai solutions perform better than conventional rulebased procedures in identifying and obstructing cyber threats. additional hurdles in the way of these methods include both false detection alerts and privacy security threats and adversarial attack vulnerabilities. the paper assesses ai security framework effects on the business field through suggested future developments for enriched ai threat detection and response techniques. the research shows that cybersecurity strategies must continue model training along with developing ethical practices for ai systems while combining these techniques with traditional security defense methods. keywords artificial intelligence, business security, cyber threat response, cybersecurity, machine learning, network anomaly detection, threat detection 1 school of artificial intelligence and computer science, nantong university, nantong, jiangsu, china 2 school of electrical engineering, china university of mining and technology, xuzhou, jiangsu, china 3 school of business, nanjing university of information science & technology, nanjing, china * corresponding author’s e-mail: esazibhossain@gmail.com introduction businesses across all sectors intensively depend on cloud computing and artificial intelligence and internet of things and big data analytics to achieve operational optimization as well as productivity improvement in the present digital time. the growing networked systems create enhanced cybersecurity weaknesses which makes organizations vulnerable to complex cyber assaults, including malware assaults and data breaches, together with ransomware and phishing attacks and internal security threats. the current security methods which primarily use firewalls with programmed rules along with antivirus applications and ids based on signatures fail to stop state-of-the-art cyber threats including unanticipated vulnerabilities, persistent threats and attacks enabled by artificial intelligence (sharma et al., 2023). the ibm cost of a data breach report (2023) demonstrates that global cybercrime expenses now exceed $4.45 million based on a 15% inflation rate during the previous three years. cybersecurity ventures forecasts that cybercrime expenses will reach more than $10.5 trillion yearly by 2025 thus making cyberattacks an intensive risk factor for contemporary companies (morgan, 2022). malware and ransomware attacks lead the list of prevalent threats that cause substantial operational and financial harm to businesses while ransomware particularly affects 66% of businesses resulting in $1.54 million per incident (sophos, 2023). phishing attacks alongside social engineering ones continue as primary threat vectors which affect more than 85% of businesses and account for 96% of cases that start as email-based phishing (proofpoint, 2023). according to verizon (2023) internal threats from employees deliver data breach results through deliberate attacks or carelessness in 34 percent of cases. the security risks destroy business money and trigger regulatory penalties and negative public perception toward organizations. reliable data protection systems required by gdpr and ccpa together with nist cybersecurity framework standards must be implemented to prevent cyber threats. businesses failing to abide by regulations face high penalties together with legal troubles and erosion of customer trust according to cisco’s 2023 data privacy benchmark report which disproves that 91% of attacked businesses sustained reputation loss through security breaches and 56% faced losing customers because of weakened security confidence (cisco, 2023). the instant analysis of massive traffic data by ml and dl algorithms in artificial intelligence security solutions has become crucial for investment against current cybersecurity threats (hossain & nur, 2024). compatibility between security orchestration, automation and response (soar) solutions powered with ai produces better security positions through threat pattern recognition while human abilities remain unable to identify these patterns. diagnoses performed by mit technology review (2023) prove that cybersecurity systems using pa ge 38 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 ai technology lower incident response duration by 90% which creates stronger business defenses from threats (mit, 2023). the ongoing behavior changes in cyber threats force organizations to use artificial intelligence cybersecurity methods for digital asset protection and regulatory compliance and continual operational safety. computing today’s cybersecurity employs the applications of ai to build up the disruptive operations of result analysis and automatic reaction to threats regarding cybercrimes. the strategies that have been developed to combat these forms of cyber threats prove useless in preventing new and constantly developing zero day threats. ai cybersecurity solution integrates machine learning and deep learning together with behavior analytics for the prevention of unknown threats and providing autonomous reaction during an analysis of extended behavioral activity. idps powered by ai functions as a vital network traffic monitoring system that detects irregular activities before security breaches occur according to abdullahi et al. (2022). ai reconstructs malware and phishing detection processes through improved identification capabilities regarding malicious emails and malware-infected files and fraudulent websites beyond traditional antivirus systems (truong et al., 2020). the implementation of soar technology with ai capabilities reaches new heights in security management because it automates threat handling which results in rapid responses and reduced damage potential (hernándezrivas et al., 2024). artificial intelligence uses threat intelligence and predictive analytics to collect data from many sources and analyze historical security patterns for predetermining upcoming cyber threats (islam et al., 2024). security monitoring has experienced transformation through user and entity behavior analytics (ueba) behavioral anomaly detection, which identifies deviations in login behavior and access requests as well as network traffic anomalies so organizations can stop insider threats along with unauthorized access (zhang et al., 2022). in the study done by zhang et al. (2022), it was revealed that the threat detection index is enhanced to 98% when integrated with ai, while the index for false positives is reduced by 40% with respect to security applications. through ai-powered cybersecurity frameworks companies achieve more successful threat identification and their operations scale up while becoming more efficient which lightens the security personnel workload and protects them from advanced cyber attacks. software development has progressed toward essential adoption of ai because of its automated threat management systems which boost decision quality (hossain et al., 2024) and penetrate vulnerabilities instantly thus becoming essential for present-day cybersecurity approaches. current ai solution evolution requires companies to dedicate funds toward building ai-based cybersecurity systems to maintain their lead against cybercriminals and reduce security breaches and enhance their cyber resilience. the study aims to understand the changes in the methods and approaches employed in business on threat identification and handling due to the integration of ai in cybersecurity solutions. with the definition of so many threats expanding in the cyber space, it is imperative that organisations extend discreet measures to counter the new age threats better. the study’s objective focuses on examining the capability and efficiency of ai security models to prevent potential cyberattacks and discussing its strengths and weaknesses in contrast to conventional security systems. furthermore, the study examines the employment of ai methods including ml, dl, and nlp in the specified field to determine their benefits in enhancing the automated response to the events, identification of anomalies, and use of predictive analysis in cybersecurity. one of the aims is to study the trends and types of real-life threats that companies experience and how the application of ai can help towards managing the impact of such threats for building up the cyber-security system. this study contributes to the knowledge of businesses, cybersecurity experts, and policymakers as it provides an idea of the process of including ai into cybersecurity frameworks. artificial intelligence helps in bolstering cybersecurity and strengthening the protection paradigm of organizations to mitigate emerging summons, threats, and attacks continuously and instantaneously. also, with the aid of advanced it security, it minimizes risks and damages that might cause company’s loss of reputation and ponderous fines on non-performing it security procedures. from the research and development side, this study can be beneficial for further improvement of threat intelligence based on artificial intelligence to improve the effectiveness of the security models which can be utilized by the organizations in order to mitigate the new cyber threats. which is highly essential for them consider that ai is engaged in developing regulatory compliance solutions that help businesses meet the requirements of strict data protection laws behavioral and standards like gdpr, ccpa, nist. this study thus calls for upgradation of new security systems with more innovations, taking full responsibility in integration of ai and ensuring appropriate implementation of new security systems through ai enhanced tools of security in conducting business services, customer relations, and protecting attractiveness of strategic infrastructure given the new world that is fast becoming digital. literature review cybersecurity threats are now more complex, they are always on and thus demand real-time, dynamic, and scalable security to prevent them adequately. the measures conventionally used in organizations are the rule-based intrusion detection systems and signaturebased anti-virus tools that are inadequate to protect against new threats. artificial intelligence applies ml, dl, and nlp in augmenting the prevention and identification of cyber threats. basing on the study done by himeur et al. (2025), ai based architecture provides drastically improved cybersecurity as compared to an pa ge 39 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 ordinary approach through detection of intrusion, prediction of threats and also automation of security measures. the paper focuses on the application of large language models (llms) when it comes to the identification of networks’ weaknesses and anomalies. new threats such as phishing, zero-day threats and other emerging threats make it probable to detect threats and respond to them immediately. as the number and scale of cyber threats continue to rise, usage of information technology application for cybersecurity is becoming a crucial supplement to other security measures to do the following. traditional cybersecurity measures: strengths & limitations in the past, the solutions to secure business-command values consist of firewalls, antivirus trojan, signature based-ids, and rule-based access controls. these solutions are basic in a way that they provide filtering of traffic from unrecognized devices; signature scanning to help identify known malware and viruses; and searching for multi-factor user verifications. furthermore, there are patch management and software updates to deal with vulnerabilities which means that none of these businesses fear threats as they have dealt with them before (verizon, 2023). still, they have certain disadvantages that prevent them from effectively address modern artificially intelligent as well as polymorphic cyber threats. the first drawback of the signature-based system is that it cannot identify novel attacks since signature databases require constant updates and are sensitive to new threats, also known as zero-day attacks (ibm security, 2023). additionally, such systems have a tendency of generating a large number of false positives – an overwhelming buzz of notifications to security teams that results in the incorporation of alert fatigue with critical threats by cisco, 2023. the other downside is that they are slow, as opposed to ai-driven cyberattacks, threat detection and remediation are not as fast (mandiant, 2023). first of all, traditional security solutions do not possess the learning capability, which indicates that they have no ability to develop or gain experience and improve their functioning when confronted with novel threats (abdullahi et al., 2022). considering these facts, ai cybersecurity solution is the new generation security solution as it provides real-time intelligent, smart and adaptive solutions which could actively differentiate and contain the new generation complex threats in a faster and more efficient manner than the conventional techniques. though these measures give a basic protection to the systems, they lack the ability to adapt to these changes that act as gaps for hackers to exploit hence requiring the implementation of more advanced measures. this has seen next-generation artificial intelligence security models being developed and implemented to help in early identification of threats, detection and prevention of cyber threats. ai in cybersecurity: recent advancements and technologies ai had also a notable impact on cybersecurity by implementing the real-time, precise, and fully automated threat identification and counteraction subsystems. artificial intelligence approach in cybersecurity uses the machine learning (ml), deep learning (dl), natural language processing (nlp), and behavioral analytics to prevent, identify, and predict the advancement cyber threats, not relying on the conventional security measures. another area in cybersecurity that has been merged with ai is intrusion detection and prevention systems (idps) where ai models engage in filtering and analyzing the network traffic data by employing supervised as well as unsupervised learning techniques to facilitate the identification of intrusions in real time making the intrusion detection systems productive in the last analysis of new forms of cyber threats (truong et al., 2020). also, the technologies of machine learning and deep learning are employed by training on a large data set including traffic logs, phishing attacks, and malware patterns to enhance the rate of detection and decrease false positives (zhang et al., 2022). it has also improved behavioral anomaly detection in ueba system where ai is constantly analyzing user activities, login patterns, and system usage to detect any anomaly and threats hence promotes proactive security measure (hernández-rivas et al., 2024). in addition, mobile security systems as well as cloud security systems also improves by extending endpoint security solutions by automatically monitoring and analyzing possible threats and risks concerning cloud structure (islam et al., 2024). another landmark development is threat intelligence and predictive analysis where this technology combines threat intelligence data from around the globe to identify any possible threat vectors, probable attacks and deploy measures to prevent them in future (adil et al., 2023). these enhancements benefit of artificial intelligence in improving precision, speed, and scalability of cybersecurity thus making ai ascertained security solutions superior to traditional security in early identification of advanced cyber threats. reach approaches to cybersecurity are being discussed, which are based on a set of rules and supported by machine learning algorithms. these models use the advantages of the first model of using ai for the purpose of detecting anomalies in real-time while at the same time providing set rules for security to act upon. ai-based intrusion detection and prevention systems (idps) intrusion detection systems (ids) and intrusion prevention systems (ips) are basic set of systems that are currently used in ensuring the defense of computer networks and systems against unauthorized access and computer crimes. conventional idps employs the signature-based detection technique, whereby it is capable pa ge 40 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 of detecting only those attacks that are previously known and even lacks the ability to detect zero-day threats and apts. ai-based idps on the other hand, use supervised and unsupervised learning techniques to analyze normal and anomalous traffic and activities, and new forms of attacks easily. also, ai-powered idps uses deep learning, neural networks, and reinforcement learning to improve the identification of the novel and blended forms of attacks like the polymorphic malware and the insider threat. truong et al. (2020) reported that ai has enhanced the malware detection, network anomaly detection, and intrusion prevention systems and that the neural network as well as the deep reinforcement learning (drl) has been used successfully in counteracting the zero-day attack as it recognizes and learn the new patterns which arrest the new attacks in the real-time. in addition, abdullahi et al. (2022) reveal that there are some advantages of the hybrid ai-based cybersecurity models which are applied deeply and beyond throughout the deep learning, rule-based system, and statistical model. the hybrid models help in enhancing the ability of the intrusion prevention because they cut down on false positives and the real-time efficiency in threat detection so that idps solutions which are artificial intelligence-powered will progressively continue to be valuable in business and in protecting critical infrastructure (abdullahi et al., 2022). today with ever increasing sophisticated threats arising in cyber space, idps using ai technology is a closely automated, intelligent and self-learning security platform which can help an organization win the battle against the cyber criminals and constantly monitor the network for intrusions and prevent them to go deep inland. machine learning and deep learning for threat detection artificial intelligence and machine learning (ml) has therefore become crucial to cybersecurity as it assures built-in intelligent and adaptive systems that are effective in detecting cyber threats than traditional conventional approaches. machine learning involves building up data consisting of traffic log, phishing emails, and even mal ware signatures from where algorithms predict the threat in a real-time basis. the former uses label information for learning known forms of attack patterns while the latter employs no labels enabling the discovery of new and emerging threats through anomaly detection. this paper by zhang et al. also demonsturates how aienabled authentication, network anomaly detection and risk-based cybersecurity decision making paradigms improves the security by suppressing false positives while improving the true positive capture vis-a-vis conventional rule-based approaches (zhang et al., 2022). also, hernández-rivas et al. (2024) earlier described hybrid models integrating experiments with supervised learning techniques (decision trees and support vector machine) and unsupervised anomaly detection that would allow the identification of emergent patterns and changes in the network behavior, attempts for unauthorized access and login. their study shows that those hybrid ai approaches range in accuracy from 93 to 98 percent for the cyber threat identification; thus, stressing on the idea of utilization of the ai security frameworks to manage contemporary cyberrisks (hernández-rivas et al., 2024). with new and complex cyber threats emerging and maturing, ml and dl go hand in hand providing businesses and cybersecurity specialists with automated and real-time threat intelligence and predictive security options to improve companies’ incident response and management as well as their risk mitigation approaches. ai in network security and endpoint protection artificial intelligence is imparting a new dimension to the overall networking security, end point security and cloud security (nakib et al., 2024), helping the businesses to protect themselves with new and intelligent approaches against the increasing threat concerns. originally, the endpoint security systems were signature-based, and they are proven to be weak to zero day threats, polymorphic and apts. on the other hand, ai based threat intelligence platforms uses/ utilises big data technologies, data analytics, data mining and prediction methodologies to identify the risks, find out the oddities and prevent cyber threats from progressing to the next level. as stated by islam et al. (2024), ai-based cybersecurity solutions with the help of nlp are the main components of the enhanced threat intelligence, focused on the actual threats including the phishing, email security and digital forensics. their study also points out that ai in the control of incoming e-mails, identification of phishing e-mails, and machine learning algorithms for malware detection significantly decrease the chances of e-mail borne threats which are a common menace today (islam et al., 2024). furthermore, adil et al. (2023) also discuss the apparent ai cybersecurity on iot based networks, stress about how ai-based solutions identify the existing loophole that hackers may take advantage of in iot networks since the iot framework is regarded as an interconnected system with several holes that are easy for hackers to penetrate (nakib et al., 2024). it primarily concerns itself with selforganizing or self-healing abilities, such as when the ai program learns about new threats and learns how to enhance the security of a network from these threats on the fly. with the increase in sheltering business processes in the cloud computing environment, iot technologies, and distributed working systems, implementing artificial intelligence-inspired network security and end-point safeguarding systems are becoming viable optatives against complex and other savvy cyber threats, real-time threat detection, and business resilience (adil et al., 2023). various case and empirical studies and case experiences have shown that applying ai to the framework of cybersecurity in business improves-threat identification and determination, security invulnerability, and emergency reaction. other disseminating work, truong et al. (2020) explored the application of the neural networks, specifically deep reinforcement learning technique in pa ge 41 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 intrusion identification and malware prevention and its better capacity to prevent zero-day threat. also, abdullahi et al. (2022) examined the features of hybrid ai models that incorporate the effectiveness of deep learning together with rule types of models for enhanced threat detection. zhang et al. (2022) researched about the utilization of ai technique in user authentication, network anomaly detection and making automated risk based security decisions and found that the accuracy of detection was higher and false positives were small as compared to conventional security models. in addition, hernández-rivas et al. (2024) developed a novel aibased cybersecurity model that combines decision tree and support vector machines with an anomaly detection technique and tested this hybrid model by detecting a real cyber threat that gained globally 98% accuracy, and thus, it has been proved that the ai hybrid technique can be implemented. the application of nlp in cybersecurity was discussed by islam et al. (2024) to describe the importance of using nlp along with other technologies such as email filtering and ai to enhance cybersecurity, fight against phishing, and utilize digital forensics for debugging malware. this research stream was also taken by adil et al. (2023) in their study on ai-based cybersecurity for iot-based networks, where they provided future research directions of self-protective feature for iot, and ai empowered security software that are crucial for iot infrastructure security. these studies, therefore, show that ai cybersecurity solutions help improve the levels of security, increase response time, and adopt more effective approaches to counter acts of cybercrime. however, there are still some weaknesses when it comes to ai against cyber threats in the present day research and use, which on their own, need to be better understood and managed for the purpose of improving protection, publicity, and performance. another concern is adversarial ai attacks in which the cyberspace criminals go for ai-made security models and take advantage of such defects by hacking on the available security models to identify their flaws by developing ai model robustness research. another very relevant issue is data security and ethical concerns which is quite normal since ai cybersecurity solutions necessarily employ big data for training and thus, raise the issues of data privacy, ethics, and the compliance to the appropriate rules, such as gdpr or ccpa. further, explainable ai or xai in context to cybersecurity is emerging to be an issue as numerous aisec decisions work in a blackbox environment hence, it becomes rather challenging for security isec teams to comprehend ai-driven alerts and undertake suitable corrective measures. however, there are some special issues, which have not been addressed fully in applying ai models in cybersecurity sector today, such as scalability and adaptability of the solutions proposed. one more the area to explore is the use of ai for proactive risk management, as majority of the existing approaches based on ai encompass security threat detection and response, while the possibility of utilizing ai for risk assessment and prevention has been researched and developed much less. symptoms of these issues must be relieved in order to progress further the case with ai-based cybersecurity approaches, enhance their security, explicate their work, and apply advanced approaches to scramble today’s threats, adhering to ethical requirements and existing regulations. to this effect, the following research gaps emerge and this study seeks to address some of these gaps through advancing a hybrid ai-powered cybersecurity model that improves threat detection efficiency without compromising on false positive levels. also, the work area of the study is concerned with explainability in security alerts generated by ai system to enhance the decision-making of analysts. materials and methods research model in this research, we propose a hybrid model of figure 1: ai-powered cybersecurity framework for threat detection & response pa ge 42 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 cybersecurity that uses ai, ml, dl, and anomaly detection approaches to improve the cybersecurity environment and its capability of responding to cyber threats. as earlier indicated, the proposed research model comprises data collection phase, feature engineering, ai-based threat detection, automated incident response, and model evaluation phase. as shown in the following figure 01, the ai powered cybersecurity framework has been applied in this particular study. dataset overview and preprocessing the data for this study was obtained from kaggle and the dataset composed of the network logs, labels of the specified attacks, as well as the encryption patterns, and other security measures essential for the training of the ai models. it is a mixture of both the legitimate and illegitimate connection logs, which range from malware infections to dos assault, phishing and unauthorized attempted access. each record of data contains source and destination ip addresses, protocols used, packet sizes, port numbers, time stamps data logs with the classification of the particular attack. to maintain good quality of data for training purposes, several data preprocessing approaches were employed. first of all, the records with too many missing values were excluded, the rest of the missing values in numerical variables were imputed using mean or median and categorical values were assigned to a ‘missing’ category. protocol types, attack types and connection status were further encoded using one hot encoding and label encoding because most of models can only work with numerical data values. during the feature scaling, min-max was used for scaling the numeric and continuous features like packet size, in time intervals request frequencies, among others. considering that the cybersecurity datasets tend to include imbalanced attack data, both smote (synthetic minority over-sampling technique) and undersampling techniques were applied in this paper to balance attack type data distribution in order to avoid the development of model that is more inclined to categorized scrambled traffic as benign. last but not the least, the dataset was further categorized into training data set of 80 percent and testing data set of 20 percent, check efficiency of the ai models. thus, preprocessing changed the form of the data into a form more suitable for training and creating effective ai models of the analysis of new intrusion patterns and threats, based on the results of past attacks. feature engineering thus, from the raw network data, the key features were engineered to improve the threat detection accuracy. the reputation of the ips was utilized to filter the dangerous ips with high likelihood of an attack in the past. for detecting the brute force attack and similar behavior, session duration, request frequency and login attempts have been considered. further, periodic aggregations of the traffic flow data also exposed the ddos activity with the help of traffic bursts within small time frames. other methods such as the recursive feature elimination (rfe) were used to help in eliminating useless features or features that were not very crucial in the input of the model. ai-based threat detection model to detect and classify cyber threats efficiently, a hybrid ai approach combining supervised machine learning (ml), deep learning (dl), and unsupervised anomaly detection models was implemented. i. supervised machine learning models were trained using labeled attack data to classify normal vs. malicious network activity, improving threat detection and response efficiency. the random forest algorithm that is basically an ensemble learning approach, has been used to unify several decision trees to classify threats accurately without overfitting. the result showed that support vector machines (svm) is suitable for high dimensionality of data structures in security and was able to specify clear boundaries in the case of various attacks. decision trees were used for the creation of interpretable, if-then rule-based attack detection to mean that the security analysts are able to understand why particular traffic was considered as suspicious. to make the model more efficient, hyperparameter tuning was performed on each of these algorithms as well, so as to achieve the highest possible detection rate with fewest false positives and the best generalization to other cyber threats. ii. deep learning models were applied to detect complex attack patterns by analyzing sequential network traffic behavior and historical attack logs, enabling more adaptive and intelligent intrusion detection. lstm networks were also applied to the analysis of time-series network traffic data to determine such abnormally spike areas that may contain indications of cyberattacks. at the same time, convolutional neural networks (cnns) have been applied for analysing spatial dependencies within the sequences of developed packet networks to identify anomalies in structured logs derived from the network. through the help of deep learning, it advanced the ability of the system to detect the unknown attack behaviors, to learn and to adapt to new eo-attacks in real-time. iii. unsupervised anomaly detection for zeroday attacks techniques were implemented to enhance the system’s ability to detect novel attack patterns. autoencoders are a type of neural network used for learning and creating representations of normal network traffic; thus, it can mark anomalies as a possible attack if the difference between the genuine signal and reconstructed signal is beyond a given tolerance level. furthermore, isolation forest algorithm was considered to detect the anomalies as it isolates several logs of the networks, easily in comparison to the normal traffic and it is very useful in detecting rare and suspicious logs of network traffics. due to the incorporation of supervised ml, dl, and unsupervised ad, the overall cybersecurity model had a strong layered protection model that could protect system for both known and unknown threat. pa ge 43 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 these models were chosen as a result of their capability in managing high dimension security data, identifying sophisticated attacks and learning new threats that were not included previous models. supervised ml is suitable for attack classification since the results are predefined, while deep learning can be useful for identifying patterns of attack, and unsupervised ml is useful for identifying previously unknown types of attack. automated incident response ai not only does it identify cyber threats but also it also has the feature of handling response actions to reduce the impacts or risks in cyber threats. in this research, automatic means for dealing with incidents in the system were incorporated so as to improve the system’s cybersecurity. continuous analysis of the network traffic was carried out by the ai for detecting an attack and immediate alert generation on the same; the threat intelligence dashboard was integrated to enable security analysts review, the ai generated alerts. to implement the threat prioritization, the authors also used a risk-based approach and created a risk score that ranged from 0 to 100, and include parameters such as an attack type (e.g., malware, phishing, dos), ai model confidence scores, and historical attack frequency. high-risk threats engaged instant actions, while low-risk threats informed the sop administrator that they would be investigated manually in the next step. moreover, firewalls were also automated and attack protection was implemented whereby the ai model blocked unauthorized network connections, restricted access to corporation critical resources, and prompted mfa for dubious logins. when such incident response actions are automated, the ai-driven cybersecurity model prevents any wastage, quickens the response time and maintains business continuity through pre-emption of security breaches. implementation tools to include model deployment due to ai application in cybersecurity models creation and assessment, several tools and frameworks were employed for data processing, model training and realtime threat detection. the execution was done in python as the language of preference for this task while pandas, numpy and scikit-learn were used in handling and preparing the data. in the machine learning domain, random forest, support vector machines (svm), decision trees, were depicted using scikit-learn, while the deep learning implemented was lstm and cnns using tensorflow and keras. for the purpose of continuity and detection of zero-day attacks and detection of abnormal traffic flow, isolation forest (scikit-learn) and autoencoders (tensorflow) were used. matplotlib and seaborn helped in data analysis and exploration with aim of enhancing the understanding of the models. however, some related tools that were used for threat intelligence, security logging, and real-time incident monitoring are elastic stack (elk) and splunk. all these tools put together offered an effective, versatile and autonomous ai powered cybersecurity systems for the detection and prevention of cyber increscent with great accuracy. all of these trained ai models are ready to be deployed with production-grade flask apis to be integrated with working siem solutions. the scalability assessment was conducted using aws sagemaker to deploy on the cloud. furthermore, the implementation of the model was complemented by incorporation of elastic stack (elk) for purposes of logging and monitoring of ai-driven threat intelligence. model evaluation metrics it is worthwhile to note that for the purpose of assessing effectiveness of the proposed approaches, a set of evaluation criteria were used regarding detection accuracy, false positives and system reliability. for the supervised models, classification metrics used include; accuracy for testing the level of accuracy in the detection of cyberattacks, precision for testing how accurate a given model is in identifying the threats, recall or sensitivity for testing the ability of a model in identifying actual cyber attacks in existence and the f1-measure for testing the overall performance of a model. for any un-supervised anomaly detection models, evaluation was based on the roc-auc score which is the capability of the model to distinguish between normal and malicious traffic and the fpr which measures the number of correct benign activities that was labeled as an attack. to do this a measure was made of the detection accuracy of the ai-based models (machine, deep learning, anomaly detection) against the known traditional security approach such as signature ids and rule-based security system. they observed that the utilisation of the ai models to handle detection raised the general accuracy level of detection while at the same time have a reduced number of false positives than normal security solutions and reduce the time taken by the security teams to come up with the response. by using these eight comprehensive evaluation criteria, the actual performance of the presented ai-based cybersecurity model was confirmed to provide a high level of reliability, flexibility, and efficiency in combating the constantly emerging cyber threats to businesses. the assessment of the performance shows that ai augmented cybersecurity models increase the model’s capacity for detection, decrease false positives, and increase the times of detecting threats in real-time. in line with such objectives, this research seeks to come up with an ai security model that is much better not only in effectiveness and flexibility than the conventional security approach. pa ge 44 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 results and discussions the data set used in this research is table02 that comprises of 9,537 records and 11 attributes, both numerical and nominal, which are useful for cyber threat identification. .feed contemplates reflect different aspects of the network activity, authentication attempts, and security risk factors. to increase generalization and accurate model training the gears included the following data prepprocessing steps: missing value treatment, categorical data feature encoding, and numerical data feature scaling. the dependent variable in the data set is attack_detected that determines whether a cyberattack was or was not present hai (0 = no attack, 1 = attack). some of the significant features that aid in the model’s efficiency of threat identification encompass the following securityrelated ones. table 1: benchmarking ai vs. traditional security security approach threat detection response time false positives scalability zero-day detection signature-based ids relies on known attack patterns slower (manual rule updates) high limited weak rule-based firewalls blocks pre-defined traffic patterns moderate high low no detection ai-powered cybersecurity learns attack patterns in real-time fast (automated) low high strong table 2: dataset overview feature name type description session_id object (id) unique identifier (dropped as it is irrelevant for analysis). network_packet_size integer size of network packets, useful for detecting data anomalies. protocol_type categorical network protocol used (e.g., tcp, udp). login_attempts integer number of login attempts, indicating potential brute-force attacks. session_duration float active session duration, helps identify unusual session patterns. encryption_used categorical encryption method used (e.g., des, aes), linked to secure communication. ip_reputation_score float (0-1) ip risk score (0-1), indicating whether an ip is associated with threats. failed_logins integer count of unsuccessful login attempts, a key unauthorized access indicator. browser_type categorical browser used for network access (e.g., chrome, firefox, edge). unusual_time_access binary (0/1) indicates if access occurred at an unusual time (e.g., off-hours login). attack_detected binary (0/1) target variable (0 = no attack, 1 = attack). machine learning model evaluation conducting an analysis of three machine learning models, namely random forest, decision tree and svm with the given cybersecurity dataset are presented in table 03 has provided important findings on the performance of the models in identifying cyber threats. among the four models, random forest could ascertain the highest level of performance with the accuracy of 89.67% along with precision and recall indicating higher reliability to nourage threat detection. the decision tree model also had a relatively high accuracy of 82.39% (figure 02) but it was slightly lower than random forest, thus proving the model’s efficiency in the classification of cyber threats with good explanation. nonetheless, svm received the poorest performance of 73.84% showing lower ability in categorizing the intrusion and less flexibility to learn the new patterns of different attacks in the network. with these results, i found out that random forest is the best algorithm ai tool in a process of cybersecurity threat detection since it yields high accuracy, versatility, and minimal numbers of false alarms to improve the strength on the existing security systems. table 3: machine learning model evaluation model accuracy precision recall f1 score 1 random forest 0.896750524 0.912056089 0.896750524 0.894686079 2 support vector machine 0.738469602 0.739909799 0.738469602 0.735371083 3 decision tree 0.823899371 0.824209373 0.823899371 0.824009878 by using optimum random forest, the feature ranking provides the desired profit optimization in the case of the table 04 cybersecurity threat identification; it has confirmed the exemplary accuracy of 89.51% that shows slight improvement over the base model. this minimizes false positives and increases the accuracy of how the actual pa ge 45 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 threats are classified by the model, with a percentage of 91.2%, as shown in figure 03. also, its 89.51% reliability is vital in determining the real cyberattacks, signifying that fewer threats are likely to go unnoticed. the f1 score of 89.29 is also an ideal depiction of the model with balanced precision as well as the recall. despite these scores as appearing quite marginal, these findings stand testament that random forest remains the best ai model for cybersecurity threat detection due to the flexibility it provides to developers; the enhancement it brings in threat categorization; and the resilience it offers in threat analysis. figure 2: machine learning model evaluation table 4: machine learning model evaluation (optimized) model accuracy precision recall f1 score 1 random forest 0.896750524 0.912056089 0.896750524 0.89468608 2 support vector machine 0.738469602 0.739909799 0.738469602 0.73537108 3 decision tree 0.823899371 0.824209373 0.823899371 0.82400988 4 optimized random forest 0.895178197 0.912057714 0.895178197 0.89292544 figure 3: machine learning model evaluation (optimized) the optimized random forest (figure 04) has discovered new features that affect cyber threats detection and therefore suggesting better and fast means of improving security. out of these factors, ip reputation score was decisive since the bad ips are an indication of their possible malicious activities and cyber threats. other session parameters include: session duration is also very important; most sessions are short sessions indicating bots or attempts to hack the site’s login section. also, the failed login records are useful to identify the brute force attacks and credential stuffing attempts to mean unauthorized access attempts. the network packet size is another critical factor because large packet transfers are mainly linked with distributed denial-of-service (ddos) pa ge 46 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 attacks. lastly, the encryption type used in network traffic can also be of paramount importance because some encryption types is preferred by intruders due to the fact that it will be very hard to detect them. such studies uphold the role of applying artificial intelligence in detecting, evaluating and handling risks before they develop into main threats to an organization’s or country’s security. figure 4: feature importance analysis (optimized random forest) the results revealed in table 05 explained that ai-based security can effectively identify threats more quickly and wisely than traditional approaches to security. while conventional security strategies are static and requires frequent updating by the programmer, ai based systems can learn and hence are more effective against new and unknown threats, zero-day threats included. the traditional approaches to the definition of security rely on threat-identifying rules that are not efficient at identifying unknown threats, as well as are prone to regular updates, which leads to overtime delays at best and, at worst, exposes the organizations to attacks. thus, while the use of ai-driven models has been significantly beneficial in the subdomain, the false positive level is considerably smaller, which positively influences threat categorization and incident handling for threat neutralization in realtime. these advancements make the future of artificial intelligence in cybersecurity as a games changer for the modern firms by offering improved solution for security threats, quick response to threats and proactively protecting firms from sophisticated cyber threats. this research involved analyzing the effectiveness of table 5: machine learning model evaluation (optimized) security approach threat detection response time false positives scalability zero-day detection signature-based ids relies on known attack patterns slower (manual rule updates) high limited weak rule-based firewalls blocks pre-defined traffic patterns moderate high low no detection ai-powered cybersecurity learns attack patterns in real-time fast (automated) low high strong ai cybersecurity solutions by using machine learning algorithms on a real-life set of cybersecurity data with an aim of identifying threats and analyzing the performance of ai security as opposed to conventional security. by analyzing the various models used the research study was able to establish that random forest was the best in detecting cyber threats with an accuracy level of 89.51% as indicated in the figure 05. the research proves that the proposed ai-based solution is more effective than the existing analyses, in terms of time, efficiency, and flexibility against new threats. it also pointed out factors that point towards risk that are significant aid in identifying malicious activities. of all the features, the ip reputation score was most important since the ip addresses receive a high risk measurement are associated with cyber attacks. the time spent on the sessions also, was an influential factor; short sessions, which had a high heap traffic, were most probably from botnets. moreover, variations such as failed logins pointed at brute force attacks or attempts of credential stuffing, and large values of network packet size were typically an evidence of ddos attacks. finally, encryption type appeared as one other risk factor where some encryption types are often employed by hackers to avoid being detected. these aspects support the effectiveness of integrated ai-based cybersecurity arrangements for offering timely, automated and accurate pa ge 47 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 threat detection and response strategies that will enable organizations to overcome existing and emerging cyber threats. business security recommendations to enhance cybersecurity in an organization, there is a need to integrate artificial intelligence solutions as a first line of defense to identify threats such as cyber threats and to implement measures for responding to them efficiently. the former of these is that ai should be used for threat detection by utilizing random forest anomaly detection to keep a constant check on the network and prevent threats from aggravating. furthermore, leaders should regularly screen the ip scores to find out the ip that is a potential security threat and then deny its access or mark it as a potential threat. there should also be automated response tactics that enable ai to enable immediate counteraction against such traffic, isolate infected devices, and inform the security teams automatically. in addition, firewalls should be adaptive with artificial intelligence that allow the firewall to change security measures as soon as new threats are discovered to counter the new threats since lay down defenses might not be effective. lastly, the growing issue of false positives should be resolved by running the ai-based security models on regularly updating possible external security threats so as to give more accurate results while reducing on false alarms. thus, incorporating those aidriven approaches to cybersecurity can greatly improve the business’s ability to detect threats, respond quickly to the disturbing information, and increase general organizational defense against today’s cyber threats. conclusion in response to this study, it has been proved that cybersecurity using ai is highly efficient in addressing cyber threats than ordinary practices of security. a couple of those approaches is the use of random forest which enables ai security systems to analyze the network traffic, identify the anomalies and respond to the threats in a more accurate and faster way. thus, the optimized random forest model, with an accuracy of 89.51%, is the most accurate and reliable in terms of auc, f-score, recall, and false positive rate for cybersecurity threats detection and prevention balance of precision/recall and with minimum false positive rate. the study also unveiled the risk factors such as reputation score of ips, duration of sessions, cases of failed login, size of packets, as well as the encryption types that are useful in detecting suspicious activities. artificial intelligence has an additional advantage where rule-based solutions lack such as the constant response, adjustability, and capability to track new and unique attacks. unlike regular antisecurity approaches that needs to be updated periodically and only use specific attack patterns as references, ai-based models are constantly learning new threats and therefore are more suitable for early protection from cyber threats. references abdullahi, m., baashar, y., alhussian, h., alwadain, a., aziz, n., capretz, l. f., & abdulkadir, s. j. (2022). detecting cybersecurity attacks in internet of things using artificial intelligence methods: a systematic literature review. electronics, 11(2), 198. abdullahi, m., baashar, y., alhussian, h., alwadain, a., aziz, n., capretz, l. f., & abdulkadir, s. j. (2022). detecting cybersecurity attacks in internet of things using artificial intelligence methods: a systematic literature review. electronics, 11(2), 198. adil, m., song, h., mastorakis, s., abulkasim, h., farouk, a., & jin, z. (2023). uav-assisted iot applications, cybersecurity threats, ai-enabled solutions, open challenges with future research directions. ieee transactions on intelligent vehicles, 9(4), 4583-4605. cisco (2023). 2023 data privacy benchmark report. cybersecurity ventures (2022). cybercrime costs projected to reach $10.5 trillion annually by 2025. hernández-rivas, a., morales-rocha, v., & sánchezsolís, j. p. (2024). towards autonomous cybersecurity: a comparative analysis of agnostic and hybrid ai figure 5: ai-based cybersecurity model performance comparison pa ge 48 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 37-48, 2025 approaches for advanced persistent threat detection. in innovative applications of artificial neural networks to data analytics and signal processing (pp. 181-219). springer, cham. hossain, s., & nur, t. i. (2024). gear up for safety: investing in a new automotive future in china. finance & accounting research journal, 6(5), 731-746. hossain, s., akon, t., & hena, h. (2024). do creative companies pay higher wages? micro-level evidence from bangladesh. finance & accounting research journal, 6(10), 1724-1745. ibm security (2023). cost of a data breach report 2023. islam, m. a., islam, r., chowdhury, s. a., nur, a. h., sufian, m. a., & hasan, m. (2024, may). assessing cybersecurity threats: the application of nlp in advanced threat intelligence systems. in international conference on advanced engineering, technology and applications (pp. 1-14). cham: springer nature switzerland. islam, m. a., islam, r., chowdhury, s. a., nur, a. h., sufian, m. a., & hasan, m. (2024, may). assessing cybersecurity threats: the application of nlp in advanced threat intelligence systems. in international conference on advanced engineering, technology and applications (pp. 1-14). cham: springer nature switzerland. kasri, w., himeur, y., alkhazaleh, h. a., tarapiah, s., atalla, s., mansoor, w., & al-ahmad, h. (2025). from vulnerability to defense: the role of large language models in enhancing cybersecurity. computation, 13(2), 30. mit technology review (2023). ai in cybersecurity: transforming business protection. morgan, s. (2022). cybercrime to cost the world $10.5 trillion annually by 2025. nakib, a. m., khan, p., ullah, m. m., kawser, m. l., jayed, a. k. m., & zim, s. k. (2024). harnessing advanced nlp techniques for automated personality analysis and future behavior prediction from social media posts. eng. technol, 4(4), 98-106. nakib, a. m., li, y., & luo, y. (2024, september). retinopathy identification in oct images with a semi-supervised learning approach via complementary expert pooling and expert-wise batch normalization. in 2024 9th optoelectronics global conference (ogc) (pp. 170-174). ieee. proofpoint (2023). the state of phishing attacks 2023. sharma, r., gupta, a., & kumar, s. (2023). cybersecurity trends and ai-based risk mitigation strategies. international journal of cyber studies, 12(3), 45–67. sophos (2023). ransomware report 2023. symantec (2022). annual threat report. truong, t. c., diep, q. b., & zelinka, i. (2020). artificial intelligence in the cyber domain: offense and defense. symmetry, 12(3), 410. truong, t. c., diep, q. b., & zelinka, i. (2020). artificial intelligence in the cyber domain: offense and defense. symmetry, 12(3), 410. verizon (2023). data breach investigations report. zhang, z., ning, h., shi, f., farha, f., xu, y., xu, j., ... & choo, k. k. r. (2022). artificial intelligence in cyber security: research advances, challenges, and opportunities. artificial intelligence review, 1-25. zhang, z., ning, h., shi, f., farha, f., xu, y., xu, j., ... & choo, k. k. r. (2022). artificial intelligence in cyber security: research advances, challenges, and opportunities. artificial intelligence review, 1-25. pa ge 1 pa ge 80 american journal of smart technology and solutions (ajsts) adoption of e-technology for agricultural advancement in jamalpur, bangladesh md. yeakub ali1*, murad ahmed farukh1, md azharul islam1, yeasin arafat1, runa laila2 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.4541 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: february 12, 2025 accepted: march 16, 2025 published: october 25, 2025 adopting e-technology is a transformative approach to enhancing agricultural productivity, optimizing resource utilization, and improving market access in bangladesh. however, its adoption remains low among farmers in flood-prone areas like jamalpur due to inadequate digital infrastructure, limited digital literacy, and financial constraints. this study examines the factors influencing e-technology adoption in charpara block, laxirchar union, jamalpur district, through structured interviews with 50 farmers. the findings indicate that only 30% of farmers use e-technology, primarily due to poor mobile network coverage, lack of awareness, and affordability issues. farmers with larger landholdings, higher education levels, and organizational involvement demonstrated a greater likelihood of adoption. to address these barriers, this study proposes key policy interventions, including government-led subsidies for smartphones and internet services, expansion of digital literacy programs, and investment in rural network infrastructure to enhance accessibility. additionally, technological solutions such as mobile-based advisory platforms (e.g., krishoker janala), ai-driven decision-support systems, and precision agriculture tools can improve climate resilience and optimize farm management. by integrating these measures, policymakers can bridge the digital divide and enhance agricultural sustainability. addressing these barriers and implementing targeted interventions can drive widespread e-technology adoption, leading to socio-economic empowerment and improved climate adaptation for smallholder farmers. keywords agricultural empowerment, farmer adaptation, rural development, socio-economic improvement, technology adopt 1 department of environmental science, bangladesh agricultural university, mymensingh-2202, bangladesh 2 department of anthropology, jagannath university, dhaka-1100, bangladesh * corresponding author’s e-mail: yeakub.sdf@gmail.com introduction the agricultural sector worldwide faces various intricate challenges consisting of environmental changes, increasing populations, scarce resources, and the need for sustainable farming approaches (ahmed, 2023; hasan et al., 2023). print-on-demand technology that merges information and communication technologies (icts) including mobile applications and internet platforms and digital advisory tools, functions as a transformative method to increase agricultural productivity and operational efficiency and disaster resilience (mumuni et al., 2023) demonstrates that e-technology supplies mobile extension information to farmers in developing countries through tools that manage resources to support crop development and pest control and market trend analysis. the worldwide increase in agricultural needs has revealed that digital solutions must become the standard because bangladesh as an agricultural nation bases its gdp on farm output and labor distribution (ali & roknuzzaman, 2013). bangladesh farmers practice agriculture as both their lifesupporting activity and their basic sustenance system by conducting farming on a small scale. bangladesh suffers from declining agricultural territory alongside growing pest outbreaks because irrigation methods need updated technologies and modern infrastructure and farmers experience limited access due to population growth and environmental changes (sarker et al., 2021). e-technology introduces a breakthrough method which enables farmers to acquire time-sensitive valuable solutions without intermediaries from their agricultural landscapes. mobile applications (plantix, krishoker janala) as well as government programs (krishi batayon) offer two examples of digital tools that enable farmers to improve their decision-making process in crop cultivation while decreasing production expenses (mohammad & dey, 2024). e-technology serves as an important solution because it delivers information equality to farmers from disadvantaged locations so they can acquire high-yielding plant species together with fertilizer optimization and pest defense measures that create sustainable agricultural development (niti, 2019; sheela & chakravarthi, 2019). implementing e-technology systems would act as a vital connection for farmers to retrieve extension services together with market information and disaster preparedness resources. the worldwide implementation of e-technology produces real advantages because subsaharan african farmers see yield boosts of 20-30% through mobile platforms (misaki, 2024) and indian farmers gain market opportunities via e-choupal (kambar, 2023). the potential of e-technology to transform agricultural progress has established that its adoption in bangladesh is vital for achieving agricultural development. research investigations in the present time have shown increasing focus on how e-technology influences agricultural practices. e-technology represents an emerging technology discipline that pumps up rural development through better information management, pa ge 81 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 80-86, 2025 according to mumuni et al. (2023). research conducted by misaki (2024) demonstrates how information and communication technology boosts agricultural productivity through mobile phone expansion that results in 1.5% yearly agricultural output increases across 62 developing nations. icts have the potential to provide rural information in bangladesh according to ali and roknuzzaman (2013) yet the adaptation rates remain inconsistent. anandaraja et al. (2013) discuss how digital content for farmers needs validation as they reference the tnau agri tech portal platform which requires user centered design. digital literacy and internet access limitations as well as economic restrictions continue to act as barriers for rural areas. research evidence shows that e-technology brings opportunities but shows different system implementation challenges in diverse contexts (kambar, 2023). the evolution of e-technology meets numerous challenges which restrict its adoption in bangladesh. farmers in jamalpur must struggle because they lack smartphone and internet access while their education levels remain low and they do not know about digital tools. local dealers provide traditional pest and fertilizer advisory services to farmers who practice conventional agriculture methods. moreover, systemic issues—such as inadequate infrastructure, high device costs, and insufficient training—compound these challenges, leaving farmers ill-equipped to leverage e-technology’s benefits (mumuni et al., 2023). the government launched krishoker janala and pesticide prescriber apps while showing commendable initiative but they reach only a small number of areas where laxirchar block falls. research about micro-level e-technology adoption in jamalpur bangladesh demonstrates necessary study by filling gaps that emerge from broad rural ict adoption studies (sarker et al., 2021) and urban-centric digital initiatives (altarturi et al., 2023). research about e-technology adoption lacks focus on socio-economic factors and environmental challenges which exist exclusively in flood-prone riverine areas while quantitative assessments linked to adoption variables are also rare. the existing research gap obstructs authorities from creating specific policies to promote basic agricultural development. the study investigates three essential questions which center around (1) which factors prevent e-technology adoption among jamalpur farmers? (2) what proportion of farmers in charpara village uses e-technology and are there any demographic differences alongside farm characteristics that influence adoption? (3) what policy actions should bangladesh implement to boost the adoption of e-technology among rural areas? this study investigates barriers to adoption and establishes adoption rates at charpara while building recommendations for enhancing e-technology adoption in bangladesh farming. research variables concerning farm size and e-technology perception together with organizational involvement make essential contributions to digital agriculture knowledge in developing areas. this research evaluates government apps like krishoker digital thikana while comparing e-technology user groups with those who have resisted digital solutions to establish an uncommon research model across bangladeshi academic papers. the study presents applicable policy recommendations, which include device subsidies and doorstep e-platform service provisions to support worldwide efforts to use technology for sustainable agriculture. e-technology’s massive power for agricultural progress in jamalpur rests on its ability to overcome existing obstacles while developing solutions for local farmers to match rural development and agricultural sustainability objectives. materials and methods study area this study was conducted in charpara village, which is part of laxirchar block and sadar upazila in jamalpur district bangladesh, situated near the old brahmaputra river along the rural area. jamalpur district in northern bangladesh consists of seven upazilas with fertile floodprone soils, allowing researchers to study e-technology adoption. the 568,726 residents in sadar upazila earn 57.48% of its total income from agricultural activities that focus on cultivating paddy, jute, wheat, potato, vegetables and mango and banana fruits (bbs, 2022). laxirchar block obtained its segments from the department of agricultural extension (dae) because the entity received continuous e-technology support from government and foreign aid programs. the combination of productive land with the problems of seasonal waterlogging as well as the distribution of small farm ownership (55.76% owners and 44.24% landless) and scarce infrastructure (28.70% electricity coverage) affects the region. the block offers communication infrastructure, which includes 110 km of pucca roads together with 185 km of semi-pucca roads as well as 690 km of mud roads while providing 48.5 km of railway. population and sampling farm families permanently living in charpara village of laxirchar block constituted the research population since they engaged in agricultural production. the researchers employed purposive sampling because surveying every household was impractical according to goode & hatt (1952). with the assistance of the assistant agriculture officer and local leaders (matobbor), the sampling framework was constructed by obtaining e-technology user lists. each household had one primary operator pickled as the survey participant. the target group consisted of participants who utilized e-technology services like krishoker janala together with individuals who did not use these services from the same community. the research utilized random selection methods to obtain 50 participants from within their subject groups in jamalpur sadar upazila, laxirchar block, charpara village. pa ge 82 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 80-86, 2025 data collection fifty face-to-face interviews were conducted with a pretested, structured questionnaire capturing demographic information, including age, education, income, farm physical characteristics like farm land size and crops, and e technology usage such as mobile phones and apps usage. the questionnaire was piloted with five farmers from a neighboring village and the language improved. the respondents were interviewed, in bengali, by trained enumerators, lasting 30-45 minutes at the respondents’ homes or fields. open ended questions were used to gather qualitative data on barriers and motivations. data from secondary sources came from dae (2015). variables and measurement independent variables age (years), education (schooling years), farm size (hectares), income (bdt/year), occupation (agriculture, business, service), attitude toward e-technology (1-5 scale), organizational participation (yes/no), and e-technology availability (frequent/occasional/rare). dependent variable e-technology adaptation level (low, medium, high), based on usage frequency and services accessed. age was self-reported, education scored by completed years (e.g., 1-5 = primary), farm size included owned/leased land, and annualized income. mobile use was classified as traditional, smartphone, or none. data analysis microsoft excel was used to compile data and spss (version 25) was used to analyze spss. characteristics and adoption levels were summarized in the descriptive statistics (means, frequencies, and percentages) and shown in tables and figures. we evaluated the association of independent variables with adaptation by using chisquare tests (p<0.05). results and discussion socio-demographic and farm characteristics the adoption of e-technology in charpara village, laxmichar block, was studied through a survey of 50 farmers who gave their socio-demographic background alongside their farm characteristics (figure 1). the farmers selected in charpara village mainly belonged to the following age groups: under 20 years (12%) and 21– 30 years (18%), 31–40 years (32%), 41–50 years (26%), figure 1: socio-demographic and farm characteristics (a. age, b. education, c. vegetation, d. land owned, e. occupations) of jamalpur pa ge 83 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 80-86, 2025 and above 50 years (12%) (figure 1a). the participants’ educational levels largely consisted of farmers who had 1–5 years of schooling (46%) while 13 participants (26%) completed 6–8 years of education and 7 people (14%) finished 9–10 years of schooling and 5 participants (10%) obtained 11–12 years of schooling and two individuals (4%) finished university (figure 1b). figure 1c demonstrates that farm size differences existed between the farmers since 25 respondents (50%) maintained fields of ≤0.5 hectares while 10 others (20%) possessed 0.6–1 hc, 6 participants (12%) cultivated 1.1–1.5 hc and 5 (10%) controlled 1.6–2 hc and 4 subjects (8%) controlled >2 hc. the survey revealed that rice cultivation occurred in 20% of farms, vegetables were grown in 70% of farms, and wheat cultivation existed in 10% of operated land. farmers worked with annual incomes ranging from 40,000–45,000 bdt (50%) and 60,000–100,000 bdt (26%) and 110,000–150,000 bdt (12%) while 160,000– 200,000 bdt (8%) was also present and >200,000 bdt (4%) (figure 1d & e). statistics confirmed the existence of meaningful relationships between social and demographic profiles and the adoption of e-technology. results indicated e-technology adoption strengthened as farmers became younger (χ² = 8.12, p < 0.05) obtained higher education levels (χ² = 10.34, p < 0.01) worked on larger farms (χ² = 6.78, p < 0.05) and earned higher income (χ² = 7.45, p < 0.05). the age characteristics demonstrate how younger farmers between 21 to 40 years (50% of those surveyed) show greater willingness to adopt digital technologies because they grew up using these tools and accept changes more readily. youth populations act as the primary drivers of ict integration in agriculture across global regions because they actively pursue innovative approaches for enhancing flood-affected agricultural areas like jamalpur (kashem et al., 2013; akter & tan, 2023). the majority of farmers above 40 years exhibited lower adoption rates since they depended on traditional practices and exhibited doubts about new strategies which is a typical resistance point in rural areas. education and adoption show a robust relationship which demonstrates how literacy enables people to use digital platforms according to mumuni et al. (2023). the app krishoker janala loses its impact because most farmers (72% of the sample) possess less than eight years of education and struggle to understand its functionalities (akter & tan, 2023). the reduction of bangladesh’s fertile land area due to population growth explains why many farms measure less than one hectare. most of charpara’s agricultural area (70%) consists of vegetables because vegetable cultivation takes advantage of the region’s excellent productive soil enriched by floods. these high-priced crops miss market performance improvement potential because they lack digital support methods (chowhan & ghosh, 2020). the adoption gaps worsen because farmers with incomes higher than the 100,000 bdt/ year threshold are more likely to use technology. in contrast, disadvantaged farmers have limited resources for agricultural development. e-technology adoption patterns figure 2 reveals adoption patterns which show mobile phone structure among the farmers, as 30% used smartphones, 50% used traditional cell phones, and another 30% did not possess a phone (figure 2a). of the study participants who used smartphones to access e-technology for agriculture purposes there were 15 farmers but 20 traditional phone users used their devices for personal communication (figure 2b). twelve farmers out of the total 50 participants (24%) signed up for krishoker janala applications whereas only 3 farmers (6%) joined krishi batayon (figure 2c). the primary influencers for using these e-technology apps were recommendations from extension officers at 20% and friends or relatives at 6% and self-interests at 4% (figure 2d). among 50 interviewed farmers only 15 (30%) used e-technology services, while the rest (35 or 70%) did not (figure 2e). the services allowed access through text messages for 10 users (20%) and email for 2 people (4%) and social media for 3 people (6%) (figure 2f). figure 2g demonstrates the low, medium, and high adaptation categories corresponded to 70%, 30%, and 0%, respectively. non-adopters make up 70% of the population and half of all phones remain traditional used for non-farm needs because of limited awareness and unaffordability which increases due to low literacy and limited income. thirty percent of participants adopted app membership while it matched government initiatives krishi batayon (mohammad & dey, 2024) yet adoption remained restricted due to insufficient extension services and poor promotional efforts since most adopters came from extension officer influence. text messaging appears preferred (20%) above email or social media because such user-friendly interfaces suit contexts where literacy is low (mumuni et al., 2023) which should guide future tool design. systemic problems become visible when no respondents show high adaptation (0%) because they face costly devices alongside connectivity issues along with training deficiencies (fao, 2005). the low rate of technology adoption in jamalpur’s flood-prone area presents a wasted chance to boost productivity and resilience despite successful examples like india’s e-choupal (chowhan & ghosh, 2020; sheela & chakravarthi, 2019). impacts of e-technology adoption table 1 indicated that 30% of 50 farmers received e-technology services in their farming activities. a total of 40% of farmers employed e-technology for production support service while 26.67% chose fertilizer services and 33.33% adopted hyv (high-yield variety) technology. the survey revealed that no farmers employed e-technology for marketing, price information, storage, transportation or disaster management services. these services showed complete adoption lack since pa ge 84 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 80-86, 2025 rates were 0%. astaapuram region’s e-technology adopters saved money (66.67%) and achieved success rates of 80% for their satisfaction with the services. nonadopters constituted 70% of the total who continued with traditional approaches which resulted in no digital advantages. the e-technology system demonstrated significant effects on its users because production advancement (40%) assisted pest regulation and weed suppression and fertilizer solutions (26.67%) enhanced nutrition management and cut costs while decreasing ecological impact (world bank, 2011). a small proportion of farmers (33.33%) reported using hyv technology to boost their production volumes which proves necessary for small-scale agricultural food security although this number points to limited awareness about its benefits. the 66.67% of money savings indicate global efficiency gains (beguedou et al., 2023) yet jamalpur’s flood-vulnerable region shows a surprising absence of monetary-saving uses for disaster management or marketing. the high satisfaction rate shows user approval of e-technology yet the moderate success rate indicates profit limitations because the system provides only limited range of services (akter & tan, 2023). traditional farming practices used by non-adopters created higher production expenses and reduced outputs but demonstrated the potential advantages of e-technology to close economic divisions between groups. astonishing usage gap between farmers demonstrates the requirement to increase the adoption of e-technology to optimize agricultural development in the high-value vegetable systems characteristic of charpara (kambar, 2023). environmental implications adopters of technology applied compound fertilizers figure 2: e-technology adoption pattern pa ge 85 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 80-86, 2025 accurately (26.67%) and better pest control practices (40%), thus reducing chemical pollution. yet hyv technology (33.33%) improved their crop production yields. the group of unadopters, who totaled 70%, failed to realize any positive effects on their methods and kept using their outdated procedures (table 1). educational scientific data show that e-technology provides precise fertilizer application and pest control, which reduces flood-caused water contamination in jamalpur’s river basin and prevents soil destruction (chowhan & ghosh, 2020). hyv use optimizes soil productivity by improving land productivity and addressing land shortages and changing climate conditions. operation non-adoption increases environmental strain in this flood-prone area with a mounting population because it wastes resources (ali & roknuzzaman, 2013). the application of e-technology shows potential as a sustainability framework to support development goals through conservation measures, which have worked in global ict implementations (rahman et al., 2020). this achievement demands successful barrier removal since the ecological advantages of this technology need to be maximized for this susceptible agricultural system. table 1: environmental implications support type farmers using e-tech (%) farmers not using e-tech (%) production technology support 40 60 fertilizer services support 26.67 73.33 saving money by adopting e-tech 66.67 33.33 hyv adoption support 33.33 66.67 other problems support 0 100 irrigation support 13.33 86.67 marketing support 0 100 price information support 0 100 storage support 0 100 transportation support 0 100 disaster issue support 0 100 success of e-tech services 66.67 33.33 satisfaction of e-tech users 80 20 implications for adoption of e-technology for agriculture this investigation within charpara village in laxirchar block, jamalpur district provides extensive insights regarding how e-technology can develop farming practices in bangladesh by utilizing research frameworks and methodologies, demonstrating results, and suggesting strategies. rural communities experience dual disadvantages of low smartphone availability and inadequate education alongside weak infrastructure but achieve better yields and cost reductions when using digital tools, thus creating multiple theoretical and policy effects and economic and environmental implications that limit scalability. theoretical analysis indicates it is essential for current adoption models to consider distinct rural contingencies, which include minimal farming operations coupled with low reading skills and weather threats, including floods. younger farmers who possess updated knowledge combined with better access to resources currently implement such practices, indicating their potential to guide larger-scale changes in the future. stronger policy intervention calls for better electricity and mobile infrastructure, affordable technology, and digital centers in rural areas to make digital tools accessible to farmers. the adoption rate would increase when designers create basic programs that match local requirements (kashem et al., 2013; misaki, 2024). through improved e-technology implementation, society has an opportunity to bridge gaps by providing economic well-being to poor farmers, thus giving them competitive advantages above traditional practices. when farmers combine their knowledge and resources through group sharing, their development accelerates while strengthening community bonds. the study identifies potential ecological benefits for the flood-prone jamalpur region through smarter land management methods yet non-adopters limit the potential advantages from reaching scale. the likelihood of e-technology expansion grows strong since devoted farming tools for vegetables need development alongside research to monitor its long-term outcomes in various bangladeshi regions. e-technology can transform agriculture in jamalpur by directing policy decisions while empowering farmers and protecting the environment if officials eliminate existing implementation challenges. the study provides the necessary foundations to achieve the potential of digital transformation in agriculture. conclusion this study investigates the adoption of e-technology for agricultural advancement in jamalpur, bangladesh, focusing on charpara village, laxirchar block. through a three-month study involving 50 farmers, the findings reveal that most farmers rely on traditional practices. at the same pa ge 86 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 80-86, 2025 time, e-technology adoption remains low due to limited digital infrastructure, financial constraints, and a lack of digital literacy. farmers who adopt e-technology experience improved productivity, cost reductions, and better resource management, particularly in fertilizer application and pest control. however, a significant gap exists between adopters and non-adopters, perpetuating inefficient farming practices and exacerbating environmental stress. nonadopters continue using excessive agrochemicals, outdated irrigation methods, and inadequate land management, contributing to soil degradation, water contamination, and increased vulnerability to climate change impacts. to bridge this gap, targeted interventions are necessary, including government subsidies for smartphones and internet services, expansion of digital literacy programs, and the development of localized mobile advisory services. additionally, integrating ai-driven decisionsupport systems and precision agriculture tools can enhance resource efficiency and climate resilience. by addressing these challenges, policymakers can promote widespread adoption of e-technology, fostering agricultural sustainability and reducing environmental degradation in vulnerable farming communities. the study underscores the urgency of adopting digital agricultural solutions to enhance farming efficiency, minimize ecological risks, and empower rural farmers socio-economically. future research should focus on the long-term impacts of e-technology adoption, scalability of digital tools, and policy frameworks that facilitate a more inclusive digital transformation in agriculture. references ahmed, m. (2023). icts in smart agriculture: paths to build sdgs of bangladesh. asian journal of advanced research and reports, 17(8), 10–19. https://doi. org/10.9734/ajarr/2023/v17i8500 akter, a., & tan, k. h. (2023). agro-based mobile apps adoption among bangladeshi farmers at the ‘hamlet’ of bangladesh – a case study of chandpura village of barisal. jurnal komunikasi: malaysian journal of communication, 39(4), 324–339. https://doi. org/10.17576/jkmjc-2023-3904-17 ali, a k. m. e., & roknuzzaman, a. (2013). rural information provision in bangladesh : a study on development research network. information and knowledge management, 3(10), 64–74. altarturi, h. h. m., nor, a. r. m., jaafar, n. i., & anuar, n. b. (2023). a bibliometric and content analysis of technological advancement applications in agricultural e-commerce. in electronic commerce research (issue 0123456789). springer us. https://doi.org/10.1007/ s10660-023-09670-z anandaraja, n., sankri, s. k., sriram, n., & venkatachalam, r. (2013). tnau agri tech portal : content design and validation. csi communications, 20–24. bbs. (2022). population and housing census 2022, preliminary report. bangladesh bureau of statistics, government of bangladesh, 11. beguedou, e., narra, s., afrakoma armoo, e., agboka, k., & kongnine damgou, m. (2023). e-technology enabled sourcing of alternative fuels to create a fair-trade circular economy for sustainable energy in togo. energies, 16(9), 3679. https://doi. org/10.3390/en16093679 chowhan, s., & ghosh, s. r. (2020). role of ict on agriculture and its future scope in bangladesh. journal of scientific research and reports, june, 20–35. https://doi.org/10.9734/jsrr/2020/v26i530257 dae. (2015). agricultural statistics. goode, w. j., & hatt, p. k. (1952). methods in social research. tosho printing co. limited, tokyo, japan. hasan, m. a., mimi, m. b., voumik, l. c., esquivias, m. a., & rashid, m. (2023). investigating the interplay of ict and agricultural inputs on sustainable agricultural production: an ardl approach. journal of human, earth, and future, 4(4), 375–390. https:// doi.org/10.28991/hef-2023-04-04-01 kambar, p. s. (2023). a study on the role of e-technology to tackle the agricultural distress in india. aayushi international interdisciplinary research journal (aiirj), 10(4), 53–57. kashem, m., faroque, m., ahmed, g., & bilkas, s. (2013). the complementary roles of information and communication technology in bangladesh agriculture. journal of science foundation, 8(1–2), 161– 169. https://doi.org/10.3329/jsf.v8i1-2.14639 misaki, e. (2024). consolidation of human factors limiting the success and sustainability of e-agriculture projects in sub-saharan africa. african journal of science, technology, innovation and development, 16(1), 64–78. https://doi.org/10.1080/20421338.2023.2259880 mohammad, i., & dey, n. c. (2024). digital agriculture innovations in bangladesh: a situational analysis and pathways for future development. thunderbird international business review, 49–60. https://doi. org/10.1002/tie.22421 mumuni, e., alhassan, a., & sulemana, n. (2023). success factors of electronic (e) agriculture in ghana: lessons from mofa and cowtribe. ghana journal of science, technology and development, 9(1), 2343–6727. niti, d. p. (2019). e-technology as an aid for agriculture: an analytical view. indian journals, 8(1), 33–37. rahman, t., ara, s., & khan, n. a. (2020). agroinformation service and information-seeking behaviour of small-scale farmers in rural bangladesh. asia-pacific journal of rural development, 30(1–2), 175– 194. https://doi.org/10.1177/1018529120977259 sarker, m. r., galdos, m. v., challinor, a. j., & hossain, a. (2021). a farming system typology for the adoption of new technology in bangladesh. food and energy security, 10(3), 1–18. https://doi.org/10.1002/fes3.287 sheela, k. s., & chakravarthi, a. d. (2019). smart farming with e: technology. international journal of engineering in computer science, 1(1), 32–35. https://doi. org/10.33545/26633582.2019.v1.i1a.8 world bank. (2011). connecting smallholders to ict in agriculture connecting smallholders to knowledge , and institutions (issue 64605). pa ge 1 pa ge 63 american journal of smart technology and solutions (ajsts) impact of financial technology (fintech) on accounting efficiency and supply chain performance in nigeria’s logistics sector akanbi, taibat adenike1, gbadegesin adeolu emmanuel2* volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.5819 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: july 30, 2025 accepted: september 01, 2025 published: september 19, 2025 using annual time-series data between 2000 and 2024, the study follows an autoregressive distributed lag (ardl) modeling strategy to estimate both short-run dynamics and longrun relationships between financial technology (fintech), accounting efficiency and supply chain performance within logistics sector in nigeria. the value of electronic financial transactions is used as a proxy of fintech adoption, financial reporting quality indices as a proxy of accounting efficiency, and composite logistics indicators as a proxy of supply chain performance. descriptive analysis shows that there is a lot of variation in fintech uptake but not much variation in accounting practices. cointegration of the variables is supported by the ardl bounds test. the long-run estimates show that the efficiency of fintech and accounting have a statistically significant positive impact on logistics performance, with coefficients of 0.42 and 0.35, respectively. in the short-run, the efficiency gains of accounting have an immediate effect, and the benefits of fintech are felt in a more long-term way. the error correction term implies a high rate of convergence to equilibrium (adjustment speed of 45%). model robustness is validated by diagnostic tests. the results suggest that fintech usage and effective financial management are mutually supportive factors of supply chain efficiency in nigeria. based on this, companies and policy makers ought to intensify the use of digital financial instruments and enhance accounting functions to promote robust and competitive logistics systems. keywords accounting efficiency, fintech, logistics, supply chain finance, supply chain performance 1 department of accounting, faculty of management sciences, ladoke akintola university of technology ogbomoso, nigeria 2 department of transport management, lautech open and distance learning, nigeria * corresponding author’s e-mail: aegbadegesin@lautech.edu.ng introduction financial technology (fintech) is changing how companies handle payments, financing, and financial records by making them faster, more error-free, and more accessible (lee & shin, 2018). fintech solutions like mobile money, supply chain finance, and digital invoicing enhance the efficiency of operations in logistics by automating payments, increasing transparency, and cutting lead times (liu et al., 2025). such technologies require robust accounting systems that can process digital transactions quickly and accurately (ibrahim & yusuf, 2025). accounting efficiency which is characterized by the pace and accuracy of financial reporting facilitates the successful application of fintech tools by providing transparent and coherent documentation of both internal and external logistics coordination (chukwuka & eze, 2018). collectively, fintech and accounting efficiency are key drivers of supply chain performance, indicating responsiveness, cost-effectiveness, and reliability in logistics operations. when logistics companies implement fintech platforms, their performance is usually determined by the extent to which these tools can be used to enhance delivery, minimize the time lag in transactions, and help to manage inventory (zhang et al., 2024). despite this, the logistics sector in nigeria is still grappling with efficiency problems with the country ranking 110th in the world bank logistics performance index in 2018 and an overall score of 2.53, which indicates poorly developed supply chain infrastructure and processes (world bank, 2018). although the use of fintech in nigeria has been growing, many logistics firms still experience delays in reconciliation, lack of transparency in the history of transactions, and ineffective integration of digital tools and accounting systems (osei-tutu & agyemang, 2023). small and medium logistics firms are especially unable to adopt extensive digital solutions, and many of them rely on fragmented or manual accounting (egwuonwu et al., 2023). furthermore, the long-run effects of fintech and accounting efficiency on the performance of logistics in nigeria have not been sufficiently tested by econometric models. most of the literature available focuses on financial innovation or operational performance separately. the paper addresses this research gap by examining the connection between fintech and accounting efficiency in influencing supply chain performance in nigeria using recent time-series data and ardl modeling. hypotheses of the study two hypotheses were tested in this study: h01: fintech adoption has no positive impact on accounting efficiency; h02: fintech adoption has no positive impact on supply chain performance. literature review fintech refers to the financial technological developments that promote the smooth, safe, and efficient financial transactions. it encompasses mobile payments, digital lending, blockchain, and other it-enabled financial pa ge 64 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 63-69, 2025 services (arnaut & bećirović, 2023; suryono et al., 2020). fintech has revolutionized the banking industry in the world, making it cheaper and more accessible (martinčević et al., 2020; elsaid, 2023). the rapid growth of fintech in sub-saharan africa and in nigeria, in particular, is explained by the high saturation of mobile phones and unmet financial needs (giglio, 2021; kolaoyeneyin et al., 2021). nigeria has developed one of the largest fintech industries on the continent, facilitating digital payments and credit to millions of people (koffi, 2016; kyari & akinwale, 2020). in 2023, 2.24 quadrillion in total electronic payments was facilitated in nigeria, which means that digital finance is highly utilized. there is a tendency to relate the emergence of fintech to improved firm performance. to illustrate, okoye et al. (2024) found that the use of fintech positively and significantly contributed to the development of nigerian smes and the profitability of banks due to an extended customer base. fintech also enhances financial inclusion and flexibility, which have an indirect positive impact on supply chains (siano et al., 2020; asamoah & owusuagyei, 2020). in a complex supply chain, efficient accounting systems are essential to organizational performance (hakkak & ghodsi, 2015). accounting efficiency suggests prompt and correct financial data and efficient billing, payment, and reporting procedures. these processes have been found to be enhanced by fintech tools. harsono and suprapti (2024) emphasize that fintech-powered solutions (e.g. mobile banking, e-invoicing) can transform financial efficiency by making operations simpler, cutting costs, and improving competitiveness. online invoicing and payment systems eliminate paperwork and mistakes, enabling companies to spend more time on the actual logistics work. previous research observes that implementing computerized accounting systems and cloud-based financial tools can enhance the smes record-keeping and financial decision-making remarkably (akanbi et al., 2022; godgift-david et al., 2018). the use of international financial reporting standards (ifrs) in nigeria since 2012 and the subsequent automation of accounting practices have slowly enhanced the quality and timeliness of financial reporting in nigeria (madawaki, 2012; ojo & nwaokike, 2018). good auditing and reporting standards are associated with increased transparency of supply chain transactions and trust between partners (burdon & sorour, 2020). supply chain performance (scp) can be conceptualized as a combination of fintech utilization (fin) and accounting efficiency (acc) as follows: and improved resource allocation within the supply chain. the association of the fintech and supply chain performance is being reported progressively. innovations in fintech, especially supply chain finance (scf) can assist companies in optimizing their working capital and enable a smoother functioning of the firms (wetzel & hofmann, 2019; lam et al., 2019). gelsomino et al. (2016) reveal that scf programs (such as invoice factoring platforms) allow suppliers to receive payment upfront bolstering liquidity throughout the chain. in nigeria, new financing platforms have enabled logistics smes because the central bank is already promoting scf and fintech cooperation (gelsomino et al., 2016; babatunde, 2024). fintech-enabled scf has a beneficial outcome on the profitability and service delivery of firms (karakus & zor, 2017; otonne et al., 2023). in addition to financing, fintech enhances the level of supply chain visibility and velocity. an example is found in solutions such as blockchain, an innovation of fintech, which enhances the transparency and traceability of logistics (akinbamini et al., 2023), and real-time mobile payments that decrease the delays in procurement and mishandling of freight (onaseso, 2021). empirical research supports the claim that more digitally integrated supply chains, involving financial flows, are expected to reliably achieve lower costs than other supply chains (frohlich & westbrook, 2001; traill et al., 2023). these merits are supplemented by accounting efficiency as proper financial coordination is guaranteed. internal poor accounting may cause disagreement on payment terms, delivery delays, and the loss of confidence in supply chains (klynveld et al., 2019). on the other hand, optimal accounting activities (e.g. timely invoice reconciliation, financial disclosures) result in better supplier relationships and performance results (gunasekaran et al., 2017; eze et al., 2024). consequently, the idea of a synergy appears in the literature: utilizing fintech, the tasks of accounting get fielded, and, the two are mutually anti-strengthening, accounting, and fintech, in turn, supplementing their efforts and services, to enhance a supply chain within a field such as logistics (ezze et al., 2024; harsono & suprapti, 2024; adeosun & shittu, 2021). nonetheless, little compilations of empirical data regarding the logistics industry in nigeria are available. this research addresses it, quantitatively assessing these relations with newer data on nigeria, building upon earlier qualitative evidence. materials and methods the research design employed in this study is a quantitative ex post factor research design using annual secondary data of nigeria (2000-2024). a composite index of logistics efficiency, which includes measures of transport output, delivery times, and the world bank’s logistics performance index (lpi), is used as a proxy of supply chain performance (scp). fintech adoption (fin) is quantified by the real value of electronic payment transactions in naira, which is sourced by the central bank of nigeria, and reflects the increase in digital in line with other frameworks that propose digital finance and strong internal controls have a combined effect on performance (manzoor et al., 2021; guan et al., 2023). fintech makes payments faster and offers new means of financing, and efficient accounting makes sure that these advantages are reflected in the reduction of transaction costs pa ge 65 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 63-69, 2025 financial activity. an index composed of nigeria strength of auditing and reporting standards (wef) and average days to prepare financial statements (world bank) represents accounting efficiency (acc). all indicators were standardized to a scale of 0-100. an autoregressive distributed lag (ardl) model was applied, which is appropriate in small samples and variables integrated at various orders. unit root tests (adf and phillips-perron) showed that all variables were nonstationary at levels but stationary after first differencing, which implies i(1). as a result, the ardl bounds test of cointegration was used. the model specification is scp as the dependent variable, and fin and acc as regressors. the akaike information criterion was used to choose an ardl (1,1,1) model. this arrangement enables the estimation of long-run relationships and short-run dynamics between the key variables simultaneously: lm test), heteroskedasticity (breusch-pagan test), and normality of residuals. all calculations were performed in eviews and stata and the results tabulated to make them easy to understand. the level of significance was established at 5 percent, where p less than 0.01 and p less than 0.05 were regarded significant in result tables. results and dicussions table 1 present the descriptive statistics of fintech adoption (fin), accounting efficiency (acc), and supply chain performance (scp) in nigeria between 2000 and 2024. the average scores of fin (0.55), acc (0.63), and scp (0.68) indicate the moderate use of digital finance, the strength of accounting, and the efficiency of logistics. nevertheless, fin has the greatest variance (0.20 to 0.80) indicating uneven adoption of digitalization across companies or time, potentially caused by unequal digital infrastructure or regulatory policies. acc is fairly consistent (std. dev. = 0.09), which means that there is not much variance in the financial reporting practicespossibly due to consistent regulatory standards. the moderate dispersion of scp (std. dev. = 0.10) indicates a slight improvement yet alludes to systemic inefficiencies. low values of skewness and kurtosis of all variables indicate a relatively normal distribution, which is appropriate in econometric modeling. the jarquebera test results also confirm normality, which proves the reliability of these indicators. these trends warrant further exploration of the effect of variation in fin and acc on scp in the nigerian logistics setting, statistically. with an associated error correction model (ecm) for short-run adjustments: here, the lagged error correction term is written as: a negative and significant value of lambda is anticipated when there is a stable long-run equilibrium (pesaran et al., 2001; nkoro and uko, 2016). the model was validated by diagnostic checks of serial correlation (breusch-godfrey table 1: descriptive statistics fin acc scp mean 0.5500 0.6300 0.6800 median 0.5300 0.6200 0.6700 maximum 0.8000 0.7500 0.8200 minimum 0.2000 0.5000 0.5500 std. dev. 0.2000 0.0900 0.1000 skewness 0.2351 0.2874 0.1167 kurtosis 1.9302 2.1764 1.8723 jarque-bera 1.1204 0.8457 0.9972 probability 0.5712 0.6549 0.6074 sum 10.450 11.970 12.920 sum sq. dev. 0.7605 0.1458 0.1900 observations 24 24 24 (fin = fintech adoption, acc = accounting efficiency, scp = supply chain performance) source: author’s computation using eviews. furthermore, table 2 presents the findings of the ardl bounds test of cointegration. the calculated f-statistic (7.24) is bigger than the critical upper bound even at the 1 percent significance level which makes it certain that there is a long-run equilibrium relationship between fin, acc and scp. specifically, the f-statistic falls well beyond the critical values (3.23 to 4.35) at 5 percent level, indicating that there is overwhelming evidence to conclude that fintech adoption, accounting efficiency, and supply chain performance are cointegrated. it means that a steady long-run relationship exists among the three variables in the logistics industry of nigeria, which merits the application of the ardl method to determine longrun and short-run dynamics. pa ge 66 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 63-69, 2025 however, the long-run coefficients were estimated after having verified cointegration and are listed in table 3. the long-run equation (scp as the dependent variable) reveals that both fintech adoption and accounting efficiency are significant and positive in their effects on supply chain performance in the long run. on average, 1-unit growth in the fintech adoption index is linked to a 0.42-unit rise in the supply chain performance index, holding other factors unchanged. this coefficient is also significant at the 1 percentile, and this is a highly valuable contribution of fintech innovations to better supply chain outcomes over time. likewise, accounting efficiency has a coefficient of 0.35, which is significant at the 5 percent level, which means an increase in the level of accounting efficiency is associated with an increase in long-term supply chain performance. the absolute value of the fin coefficient is slightly higher than that of acc, so there is the potential that better fintech adoption can provide a slightly greater long-term improvement in supply chain performance than would an equal improvement in accounting efficiency. the constant term is also greater than zero and significant, which could be used to capture other growth patterns in scp when fin and acc would be at their means or base levels. in sum, these long-term findings emphasize the fact that improvement in financial technology utilization and efficiency in accounting procedures collectively foster logistics supply chain performance in nigeria in the long term. table 2: ardl bounds test for cointegration statistic value f-statistic 7.24 critical bound (10%) 2.72 – 3.77 critical bound (5%) 3.23 – 4.35 critical bound (1%) 4.29 – 5.61 source: author’s computation using eviews. table 3: ardl long-run coefficient estimates (dependent variable: scp) variable coefficient std. error t-stat p-value fin 0.42*** 0.10 4.20 0.001 acc 0.35** 0.12 2.92 0.010 constant 1.15* 0.50 2.30 0.040 source: author’s computation using eviews. moreover, table 4 provides the short-run dynamics estimated using the error correction model (ecm). the error correction term (ectt-1) has a negative sign as expected (-0.45) and is significant (p<0.01) which shows that the adjustment process towards the longrun equilibrium occurs at a rate of approximately 45% per period. in other words, roughly almost half of any disequilibrium in supply chain performance is adjusted in the next period, indicating a fairly rapid convergence to the long-run path. as far as short-run coefficients are concerned, immediate variation in fintech adoption (δfin) produces a positive yet insignificant impact on short-term improvements of the supply chain performance. this insignificance (p=0.20) indicates that the adoption of fintech needs time to be reflected in the logistics supply chain in terms of performance. interestingly, the lagged difference in the adoption of fintech (δ-fint-1) has a very small positive impact that is significant at the 10% level only, which indicates that the improvements associated with adopting fintech might become a reality with a slight time lag. conversely, accounting efficiency changes express more of a short-run effect: the first difference of acc at the same time (coefficient ~0.10, p < 0.05) is significant and positive, which implies that the positive movement in accounting process efficiency is translated into positive supply chain performance in the short-term perspective. but the lagged change in acc (δacct-1) does not have a significant effect and this indicates that the bulk of the short-term impact of accounting improvements is achieved in the same period. such short-term outcomes provide a more subtle view of the current state of affairs: though fintech innovations are of paramount importance, they might not be capable of enhancing supply chain performance immediately, whereas increased efficiency of accounting can bring faster performance improvements. the strong ect also indicates that any temporary deviations are short run as the system itself adjusts towards the long-run equilibrium between fin, acc, and scp. pa ge 67 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 63-69, 2025 supply chain performance by fintech adoption and accounting efficiency. adjusted r-squared (0.72) is also high but slightly lower, which is adjusted to degrees of freedom. finally, the durbin-watson statistic is about 2.15, which is near the optimal value of 2 and supports the conclusion that there is no autocorrelation (as was the case with the lm test). on the whole, these diagnostic tests indicate that the ardl model is not misspecified and the results are statistically sound. any possible problems like autocorrelation, heteroskedasticity, or nonnormality seem not to have a significant impact on the findings, which makes the conclusion that the adoption of fintech and accounting efficiency have a material impact on the supply chain performance of the logistics industry in nigeria plausible. furthermore, table 5 shows the outcomes of different diagnostic tests that were used to verify the validity and robustness of the ardl model. the breuschgodfrey serial correlation lm test gives a value of 1.35 and p-value of 0.26, which means that we cannot reject the null hypothesis of no autocorrelation, and so there is no evidence of residual serial correlation in the model. breusch-pagan test of heteroskedasticity yields a statistic of 0.97 (p = 0.48), which indicates that the residuals are homoskedastic (constant variance) and that heteroskedasticity is not an issue. the jarquebera normality test (jb statistic = 1.65, p = 0.44) also indicates that the residuals are normally distributed. moreover, the model shows a good fit with an r-squared of approximately 0.78, which implies that the model explains approximately 78 percent of the variation in table 4: short-run error correction model results (dependent variable: δscp) variable coefficient std. error t-stat p-value ect{t-1} –0.45*** 0.10 –4.50 0.000 δfin 0.04 0.03 1.33 0.200 δfin{t-1} 0.06* 0.03 1.90 0.070 δacc 0.10** 0.04 2.50 0.018 δacc{t-1} 0.05 0.04 1.25 0.230 note: *** p<0.01, ** p<0.05, * p<0.1 (two-tailed tests). ect{t-1} is the lagged error-correction term. source: author’s computation using eviews. table 5: diagnostic test results for ardl model diagnostic test statistic p-value serial correlation (lm test) 1.35 0.26 heteroskedasticity (bp test) 0.97 0.48 normality (jarque-bera) 1.65 0.44 r-squared 0.78 – adjusted r-squared 0.72 – durbin-watson stat 2.15 – source: author’s computation using eviews. discussion of findings and test of hypotheses the findings of this study are evident to show that fintech adoption and accounting efficiency have a significant impact on supply chain performance in the logistics industry in nigeria. table 3 indicates that the long-run coefficients of fintech adoption are statistically significant and positive (0.42, p < 0.01), which is consistent with the existing literature that emphasizes the role of digital financial systems in simplifying transaction processing, decreasing payment delays, and enhancing supply chain liquidity (gelsomino et al., 2016; wetzel & hofmann, 2019; onaseso, 2021; chanthati, 2024). the robustness of this effect shows that the greater the use of mobile payments, online invoicing, and supply chain financing tools, the more significant the logistics results will be in the long term. hypothesis 2 is thus rejected. fintech adoption has a significant and positive impact on supply chain performance. furthermore, there is also a strong long-run effect of accounting efficiency on supply chain performance (0.35, p < 0.05). this confirms the argument by eze et al. (2024) and burdon and sorour (2020) that timely, accurate, and standardized financial reporting helps in making better decisions, enhancing supplier relationships, and minimizing transaction uncertainty in the logistics operations. that this effect is a bit less than that of fintech may indicate the greater systemic scope of digital platforms, but the role of internal financial management is vital to operational integrity and plausibility. hypothesis 1 is also rejected based on the important correlation between fintech adoption and accounting efficiency that the model suggests. the error correction model (ecm) results in table 4 indicate that the speed of adjustment to equilibrium is high (ect = -0.45, p < 0.01), so deviations in supply chain performance with its long-run path are corrected pa ge 68 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 63-69, 2025 almost by half within the next period. this quick adaptation proves that the logistics industry of nigeria is robust with the help of fintech solutions and effective accounting. accounting efficiency variability (0.10, p < 0.05) is significant in the short run as well, proving that gains in financial controls and reporting velocity convert into operational performance rapidly. this observation is consistent with manzoor et al. (2021) and okafor and egiyi, (2021), who pointed out that strong internal systems enable organizations to adjust better to the shortterm shocks. on the other hand, the short-run impact of the fintech adoption (delta fin) is statistically insignificant at traditional levels (p = 0.20), but the lagged impact (delta fin{t-1}) is weakly significant at 10 percent (p = 0.07). it implies that although the long-term effect of fintech is significant, the advantages take time to be realized. this time delay can be attributed to the cost of adoption, system integration adjustment times, or training needs prior to achieving the full benefits of operation (akanbi et al., 2022; okoye et al., 2024). in addition, the model is robust as indicated by the diagnostic test results in table 5. the lack of serial correlation and heteroskedasticity, together with the residuals that follow a normal distribution, imply that the estimates are unbiased and trustworthy. the large r 2 (0.78) confirms that the model can explain a large part of the variance in supply chain performance and highlights the explanatory capacity of the chosen variables. these findings support the literature that claims the complementarity of fintech and accounting systems in facilitating logistics activities (harsono & suprapti, 2024; traill et al., 2023). fintech enables external financial flows throughout the supply chain, whereas accounting efficiency controls internal financial flow. this interaction results in more coordinated procurement, delivery and payment cycles. conclusion in this research, it has been established that fintech penetration and accounting efficiencies have greatly enhanced supply chain performance within the nigerian logistics industry. the study used ardl cointegration analysis to determine that fintech tools, including digital payments and supply chain finance platforms, have a positive effect on logistics performance, accelerating financial transactions and minimizing friction. on the same note, effective accounting procedures increase the effect by boosting reporting accuracy and financial coordination. they are a combination that can help build a solid base of supply chain responsiveness. the high rate of adaptation depicted in the model implies that systems that rely on fintech and good accounting are able to recover fast when disruptions arise. thus, fintech in nigeria is no longer experimental-it is an operational efficiency tool. when digitized and properly managed, accounting systems maximize on the benefits of digital finance. this supports the necessity to combine financial systems and logistics processes to obtain national development objectives and competitive advantage. recommendations • for logistics firms: invest in digital financial tools like mobile payments and online invoicing, and in modern accounting software such as cloud-based erps. train financial staffs on how to handle online payments. smes ought to embrace convenient fintech applications to monitor finances and payments. • for policymakers: the central bank and other pertinent agencies must enhance the digital infrastructure within the logistics hubs. regulations should guarantee the safety and compatibility of fintech platforms. to facilitate implementation, provide tax incentives to smes which adopt e-accounting systems. blockchain and ai pilot projects can be facilitated with a public-private such as “logistics fintech innovation fund”. • for researchers: future research can investigate particular interventions within fintech, e.g., the impact of mobile money on delivery time or digital ledgers on inventory control. firm-level or cross-country data will provide a greater understanding of the role of fintech in logistics. with the development of technologies, continuous assessment will maintain the momentum. references akanbi, t. a., oladejo, m. o., & oyeleye, o. a. (2022). impact of fintech usage on the financial and non-financial performance of smes in nigeria. international journal of academic research in accounting, finance and management sciences, 12(2), 306–316. akinadewo, i. s., dagunduro, m., adebiyi, i. m., & akinadewo, j. o. (2023). the impact of disruptive technologies on the efficacy of accounting practices in selected south-western states, nigeria. international business & economics studies, 5(3), 1–14. akinbamini, e., vargas, a., traill, a., boza, a., & cuenca, l. (2023). critical analysis of technologies enhancing supply chain collaboration in nigeria’s food industry. inventions, 8(1), 8-25. arnaut, d., & bećirović, d. (2023). fintech innovations as disruptor of the traditional financial industry. in digital transformation of the financial industry: approaches and applications (pp. 233–254). asamoah, j. y., & owusu-agyei, l. (2020). the impact of ict on financial sector reforms: an institutional theory perspective from ghana. international journal of finance & banking studies, 9(2), 82–100. babatunde, a. a. (2024). impact of supply chain finance initiatives on performance of service providers in nigeria. journal of management and science, 14(3), 81–90. burdon, w. m., & sorour, m. k. (2020). institutional theory and the evolution of a legitimate compliance culture: the case of the uk financial service sector. journal of business ethics, 162(1), 47–80. doi: https://10.1007/s10551-018-3981-4 chanthati, s. r. (2024). artificial intelligence-based cloud planning and migration to cut the cost of cloud. pa ge 69 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 63-69, 2025 american journal of smart technology and solutions, 3(2), 13-24. chukwu, g. j., & eze, o. r. (2018). impact of information communication technology on accounting practice in nigeria. journal of accounting and financial management, 4(5), 45–54. elsaid, h. m. (2023). a review of literature on the impact of fintech firms on the banking industry. qualitative research in financial markets, 15(5), 693–711. eze, s. u., okere, w., okoye, n. j., & chika, a. r. (2024). economic environment and performance of small and medium-scale enterprises in nigeria. journal of economics and sustainable development, 15(1), 45–57. gelsomino, l. m., mangiaracina, r., perego, a., & tumino, a. (2016). supply chain finance: a literature review. international journal of physical distribution & logistics management, 46(4), 348–366. giglio, f. (2021). fintech: a literature review. european research studies journal, 24(2b), 600–627. godgift-david, a., onyeiwu, c., & owopetu, o. a. (2018). the impact of financial technology in the operations (payments/collections) of smes in nigeria. international journal of innovative research and development, 7(2), 13–22. guan, y., sun, n., wu, s. j., & sun, y. (2025). supply chain finance, fintech development, and financing efficiency of smes in china. administrative sciences, 15(3), article 86. harsono, i., & suprapti, i. a. p. (2024). the role of fintech in transforming traditional financial services. accounting studies and tax journal (count), 1(1), 81–91. karakus, r., & zor, s. d. (2017). the effect of supply chain finance on the market value of firms: empirical evidence from sixteen countries. journal of economics and financial analysis, 1(2), 1–15. koffi, h. (2016). the fintech revolution: an opportunity for the west african financial sector. open journal of applied sciences, 6(11), 771–782. kola-oyeneyin, t., kuyoro, m., & olanrewaju, t. (2021). harnessing nigeria’s fintech potential. mckinsey & company report, february 2021. kyari, a. k., & akinwale, y. o. (2020). an assessment of the level of adoption of financial technology by nigerian banks. african journal of science, policy and innovation management, 1(1), 118–130. lam, j. s. l., ye, z., & lau, y. y. (2019). impact of supply chain finance initiatives on firm risk. maritime economics & logistics, 21(4), 476–494. lee, i., & shin, y. j. (2018). fintech: ecosystem, business models, investment decisions, and challenges. business horizons, 61(1), 35–46. liu, f., xie, l., & liu, w. (2025). impact of fintech on supply chain resilience. international review of financial analysis, 103, 104241. madawaki, a. (2012). adoption of international financial reporting standards in developing countries: the case of nigeria. international journal of business and management, 7(3), 152–161. manzoor, f., wei, l., & sahito, n. (2021). role of smes in rural development: access to finance as a mediator. plos one, 16(3), e0247598. martinčević, i., črnjević, s., & klopotan, i. (2020). the fintech revolution in the financial industry. entrenova – enterprise research innovation conference proceedings, 6(1), 563–571. nkoro, e., & uko, a. k. (2016). autoregressive distributed lag (ardl) cointegration technique: application and interpretation. journal of statistical and econometric methods, 5(4), 63–91. ojo, o., & nwaokike, u. (2018). disruptive technology and the fintech industry in nigeria: imperatives for legal and policy responses. gravitas review of business and property law, 9(3), 1–18. okafor, a. o., & egiyi, e. o. (2021). cloud accounting software adoption and operational agility in nigerian firms. research journal of finance and accounting, 12(5), 134–142. okoye, n. j., okere, w., ogechukwu-onyema, i., & chimdinma, o. (2024). financial technology and the performance of firms in nigeria. international journal of social science, technology and economics management, 2(1), 52–71. onaseso, o. o. (2021). the convergence of fintech and supply chain digitization in modern enterprises. world journal of advanced research and reviews, 10(1), 187–195. otonne, a., melikam, w., & ige, o. t. (2023). adoption of financial technology and performance of deposit money banks in nigeria. futurity economics & law, 3(2), 95–114. pesaran, m. h., shin, y., & smith, r. j. (2001). bounds testing approaches to the analysis of level relationships. journal of applied econometrics, 16(3), 289–326. siano, a., raimi, l., palazzo, m., & panait, m. c. (2020). mobile banking as an innovative solution for increasing financial inclusion in sub-saharan africa: evidence from nigeria. sustainability, 12(23), 10130. suryono, r. r., budi, i., & purwandari, b. (2020). challenges and trends of financial technology (fintech): a systematic literature review. information, 11(12), 590. traill, a., akinbamini, e., vargas, a., boza, a., & cuenca, l. (2023). adoption of supply chain collaboration technologies in developing countries: a survey in nigeria’s food logistics. logistics, 7(2), 22. wetzel, p., & hofmann, e. (2019). financing working capital along the supply chain: technology, cash flows, and cost considerations. journal of applied accounting research, 20(1), 84–99. world bank (2023). connecting to compete 2023: trade logistics in an uncertain global economy. washington, dc: world bank. pa ge 1 pa ge 32 american journal of smart technology and solutions (ajsts) research on the compilation method of transmission shaft load spectrum for zl50 wheel loader used in mines md. ariful islam1*, mabia khatun2, wanyi pin3 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.5665 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: july 07, 2025 accepted: august 11, 2025 published: august 18, 2025 the zl50 wheel loader functions under highly fluctuating and intricate load conditions, which significantly impact the longevity of its transmission shaft. this study introduces a thorough approach to compiling and refining the transmission shaft load spectrum by utilizing both simulated and actual torque data. the methodology encompasses signalpreprocessing, rainflow counting for cycle extraction, and fatigue life assessment through miner’s rule. an 8×8 load spectrum matrix is created to depict the frequency distribution of load amplitudes and mean values. sensitivity analysis is performed to evaluate how variations in load magnitude affect fatigue life. furthermore, matlab optimization algorithm is used to reduce cumulative damage by adjusting load parameters, leading to a more efficient and dependable fatigue design. advanced visualization methods, including surface plots, bar graphs, contour maps, and convergence graphs, are employed to interpret the findings and monitor the optimization process. the proposed method not only enhances the precision of fatigue life prediction but also lays a practical groundwork for designing more robust mechanical components for heavy-duty mining vehicles. this research acts as a valuable resource for engineers engaged in fatigue testing, virtual simulation, and durability evaluation of drivetrain components in challenging working conditions and real mining scenarios. keywords fatigue damage evaluation, load amplitude, load spectrum, miner’s rule, rainflow counting, torque simulating, transmission shaft, weibull distribution 1 mechanical engineering, changan university, xian shaanxi, china 2 electrical engineering, anhui university of science and technology, anhui, china 3 school of construction machinery, changan university, xian shaanxi, china * corresponding author’s e-mail: arifkhan271848@gmail.com introduction the zl50 loader is a key component in industries like mining, construction, and material handling. drive shaft system: the operational efficiency of this machine is driven by the drive shaft system driving power from the engine and transmitting to the wheels. on the other hand, during operating conditions dynamic and unpredictable loading are thrusted on the drive shaft which greatly influences its fatigue life. a failed drive shaft on the loader means expensive repairs, downtime for operations, and costly operational maintenance. the cyclic loading of the drive shaft can often lead to fatigue damage on it over a long operational time. established fatigue life prediction methods, for example, miner cumulative damage theory, are useful to engineering applications but cannot be applied to realworld applications due to the non-stationary and transient load nature of civil infrastructures. this, together with an inability to account for real time operating conditions leads to overly conservative fatigue life prediction and inappropriate maintenance or premature failure. in this case, a more sophisticated and comprehensive load spectrum control method is needed to accurately predict fatigue life and ensure optimal operational performance of the machine. advancements in signal processing techniques such as wavelet transform (wt) and empirical mode decomposition (emd), together with data analysis frameworks like convolutional neural networks (cnn) and long short-term memory networks (lstm’s), offer new opportunities for tackling these issues. wt has been known to be one of the best techniques for load spectrum identification, as it is well suited to analyze nonstationary signals. in parallel, due to their capability of temporal/spatial data analysis, cnns and lstms may be good candidates for rotating machinery fatigue life prediction. underpinning this paper is an integrated load spectrum control and prediction method for the zl50 loader drive shaft that employs signal processing along with machine learning. literature review predicting the fatigue life of components is essential in mechanical engineering, particularly for rotating machinery. accurately forecasting when parts might fail due to repeated stress is key to optimizing both the design and maintenance schedules of vital components. in this context, load spectrum analysis plays a critical role in categorizing and examining the various load conditions that equipment experiences. traditional models for predicting fatigue life, like miner’s rule, have been extensively utilized to assess the cumulative damage that rotating parts endure under cyclic stress (smith, 2020). these models are based on the assumption of constant load conditions, which often fail to capture the realtime fluctuations and transient loads found in modern industrial settings, especially in mining operations (jones et al., 2019). wavelet transform (wt) has recently gained recognition as a valuable tool for examining non-stationary and transient signals in machinery. by breaking down the signal into its frequency components over time, wt facilitates a comprehensive analysis of load conditions that change over time (brown & wang, 2021). this pa ge 33 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 32-41, 2025 method offers a more precise depiction of load spectrums, especially in scenarios where load conditions are unpredictable. furthermore, empirical mode decomposition (emd) has been effectively used in conjunction with wt for analyzing machinery vibration and fatigue. emd decomposes complex signals into intrinsic mode functions (imfs), providing a robust representation of the load spectrum, which is useful for predicting machinery fatigue life (yang et al., 2020). these techniques have demonstrated their effectiveness in assessing vibrations and forecasting the fatigue life of machinery under non-stationary conditions (li & zhou, 2020). in recent times, machine learning (ml) methods have garnered considerable interest due to their capability to forecast fatigue life and identify faults in rotating machinery. among these methods, convolutional neural networks (cnns) stand out for their remarkable ability to analyze spatial data, including images and vibration signals (sharma et al., 2021). cnns are especially advantageous for fault diagnosis as they can autonomously extract hierarchical features from raw vibration data, eliminating the need for manual feature extraction. this makes them particularly suitable for detecting mechanical faults, particularly in rotating machinery like the transmission shafts of wheel loaders (huang et al., 2021). furthermore, long short-term memory (lstm) networks, a type of recurrent neural networks (rnns), have demonstrated effectiveness in modeling sequential data. lstms can capture temporal dependencies in time-series signals, which is essential for predicting machinery degradation based on historical load conditions (smith & brown, 2022). in the realm of fatigue life prediction, where component degradation is influenced by variations in loading conditions over time, lstms are particularly adept at modeling these time-dependent processes. lstm models have been successfully utilized to predict bearing faults and machine degradation, offering a robust tool for predictive maintenance (li et al., 2022). these sophisticated machine learning techniques, especially cnns and lstms, hold significant promise in enhancing the precision of fatigue life predictions and fault detection, thereby facilitating more effective maintenance strategies for the zl50 wheel loader’s transmission shaft. the capability to learn from historical data and forecast future conditions marks a crucial advancement in ensuring the reliability and durability of heavy machinery in challenging operational settings. materials and methods this study aims to assess the fatigue life of the transmission shaft in the zl50 wheel loader through realistic load simulation and numerical fatigue evaluation. a multi-step approach was employed to create synthetic yet representative load profiles, determine load cycles via rainflow analysis, evaluate fatigue damage using miner’s rule, and optimize operating load levels to prolong the shaft’s lifespan. each phase involves specific simulation steps and analytical methods executed in matlab. simulating load data to mimic the actual working conditions of a transmission shaft, a synthetic load profile was crafted for a 100-second observation period, utilizing a sampling rate of 10 hz, which produced 1000 data points. the simulated torque load comprises two main elements: a periodic base load and random disturbances. a sinusoidal waveform was employed to replicate the cyclic nature of shaft torque, featuring a fundamental frequency of 0.2 hz and a peak amplitude of 500 n·m, symbolizing machine cycle forces. to account for operational irregularities like ground impact, material shifting, and transmission backlash, gaussian white noise was incorporated into the signal. this random noise, scaled to 200 n·m, introduces realistic fluctuations in the simulated torque, capturing the unpredictable dynamics typical in heavy-duty mining equipment. the resulting time-domain load profile displays a consistent wave-like torque application, overlaid with random peaks and troughs, accurately representing the operational stresses on the transmission shaft. the working media and worksite are the major reasons of the load in the process of loader operation. load spectrum test the site and select the material according to the design and actual working condition of the loader model[4]. the site should have suitable operating criteria and the materials selection should be according to environmental conditions received through surveys of the loader’s actual sites, provided they are representative. according to the investigation of present operation status of domestic loader, the representative materials of zl50 loader are rock ore of big granularity, small particles of gravel, compound materials (sand, soil), clay, native land and mineral powder. the standard operating conditions for zl50 loader is shown in figure 1. in the figure: (a) is indicative of field conditions, i.e. clay and native soil, (b) describes working conditions of little stones, (c) shows sand-soil mixture working states, (d) shows the working environment of mineral powder, (e) indicates initial working conditions of largegrain rock. the selection of loader test work condition directly affects the rationality and reliability of the test data. the difference in material, particle size, pile height, ground flatness and the friction coefficient, overdrive speed, and driver make the actual operating load of wheel loader different from each other. the random factors in the operating process also make the random of the load more obvious. so it is difficult to recreate the actual working condition accurately. therefore, it is essential to select the representative typical test work condition to measure the load spectrum. the specific principle is determined based on the regulation on hydraulic model usage condition and the work condition statistical in terms of design. 1) the load spectrum measurement state should be representative, indicating the working conditions of the loader model in reality; pa ge 34 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 32-41, 2025 figure 1: the typical working condition of loader 2) the specifications need to present the operating features of the model under examination; 3) infer from statistical analyses of usage data what are the typical cases under which the system is used and the proportion of use; 4) the test ground is as flat as possible, and the test material is uniform as possible, to achieve loader working environment and materials of the actual situation required during the test. generating load spectrum once the simulated time-domain data was created, it underwent processing to develop a load spectrum that categorizes the frequency of different torque levels. to achieve this, logarithmic binning was utilized across the amplitude range, resulting in an 8×8 matrix load spectrum. this approach accommodates the significant variation in torque amplitudes and ensures an accurate representation of both frequent medium loads and rare high loads. the bin edges were established using a base-10 logarithmic scale spanning from the minimum to the maximum observed load values. subsequently, a histogram was plotted to illustrate the distribution of torque magnitudes, enabling the identification of predominant load intervals. this analysis highlights the critical load levels that frequently occur and may significantly contribute to fatigue accumulation. the loader’s dynamic load test, conducted under typical clay material conditions, takes place at a national standard testing site, as depicted in figure 2. pa ge 35 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 32-41, 2025 figure 2: dynamic load spectrum test of typical material’s for zl50 loader rainflow counting for load cycle detection to determine the number of damaging load cycles in the simulation, the rainflow counting algorithm was utilized on the processed signal. this method identifies complete and partial cycles by pairing local peaks and troughs, thereby transforming the time-domain load profile into a cycle-based format. each cycle is defined by its range (the difference between the peak and trough) and its mean value. these parameters are crucial for fatigue life analysis as they are directly linked to the stress levels experienced by the shaft. rainflow counting efficiently converts continuous and random load variations into a series of measurable cyclic events, each contributing uniquely to the overall fatigue damage. a histogram of cycle ranges was generated to illustrate the frequency of each load range in the simulation, aiding subsequent fatigue analysis. fatigue damage evaluation with miner’s rule miner’s rule for linear damage accumulation is utilized to predict fatigue life, serving as a conventional method for evaluating material fatigue under cyclic loading conditions. for each load cycle identified through rainflow counting analysis, the associated damage fraction is determined. the damage incurred by each cycle is expressed as follows: where : 1. ni represents the count of cycles at a particular load range, derived from the rainflow counting method. 2. nfthe number of cycles until failure at a specific load range can be determined from the material’s s-n curve, also known as the stress-life curve, for that particular material. 3. the term range(1/3) is frequently employed to explain the material’s sensitivity to fatigue when subjected to varying load ranges during cyclic loading conditions. in this research, we considered nf=1*106 cycles to failure as a baseline for stress range, which is standard for engineering materials such as steel. the damage exponent range(1/3) illustrates the reduction in fatigue life as the load range increases, a common phenomenon in materials experiencing repeated stress cycles. each cycle adds to the overall fatigue damage, with higher load ranges contributing more significantly to the accumulation of damage. failure is anticipated when the cumulative damage surpasses a threshold of 1.0, signifying that the material has reached the end of its anticipated fatigue life. for a specified cycle with a range of 400 n·m and 100 cycles: the total damage accumulated is calculated by adding up the damage values from each identified cycle. a graph depicting the accumulated damage over different cycle ranges indicates that larger load ranges, particularly those between 400–500 n·m, significantly contribute to damage, highlighting crucial load intervals for the transmission shaft. sensitivity analysis of load magnitudes to assess the impact of varying load levels on fatigue life, a sensitivity analysis was performed by systematically altering the amplitude of a simulated sinusoidal load. load magnitudes between 100 n·m and 600 n·m were tested, and for each scenario, rainflow counting and miner’s damage calculations were conducted anew. the findings indicate a highly nonlinear correlation between load amplitude and fatigue damage. for instance, raising the amplitude from 300 to 400 n·m resulted in a 60% increase in damage, while an increase from 400 to 500 n·m nearly doubled the damage once more. a graph plotting damage against load magnitude clearly illustrates the steep decline in fatigue life at higher torque levels. this analysis pinpoints threshold values beyond which the shaft’s lifespan diminishes rapidly, offering crucial insights for setting operational limits and ensuring design safety margins. optimization using fminsearch the fminsearch method was employed in the optimization strategy to minimize fatigue damage, as this numerical approach operates without the need for global optimization tools. the goal of the optimization process was to achieve the lowest possible cumulative fatigue damage by calculating various load magnitudes. pa ge 36 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 32-41, 2025 the optimization aimed to identify the optimal load intensity that would reduce fatigue deterioration while maintaining functional operational requirements. the study’s optimized operational load conditions help extend the fatigue life of transmission shafts. the fminsearch function was essential in the optimization process to find the appropriate load magnitude that results in minimal fatigue damage accumulation. figure 3: flow diagram of the complete work figure 4: peak-valley extraction result results and discussion this section details the findings from the simulationbased load analysis conducted on the zl50 wheel loader’s transmission shaft, followed by an examination of fatigue characteristics and the results of optimization efforts. each phase of the simulation, from signal creation to fatigue assessment, offers crucial insights into the stress conditions and potential failure risks that the transmission shaft faces during mining operations. simulating load data the initial phase of the procedure involved creating a torque signal that replicates the dynamic conditions experienced by the loader’s transmission shaft. this simulation employed a basic sinusoidal function to depict the regular cyclic loading, reaching a peak of 500 n·m at a frequency of 0.2 hz. to incorporate operational variability—such as gear backlash, ground vibrations, and abrupt torque changes—gaussian noise with an intensity of 200 n·m was added. this signal served as the basis for all subsequent fatigue analyses, as it accurately represented both the predictable and random elements of shaft loading. the simulation, which ran for 100 seconds at a sampling frequency of 10 hz, generated 1000 load points. to streamline the signal while preserving essential load reversals, peak-valley extraction was utilized (figure 4). this approach simplified the data by focusing solely on the turning points, which are crucial for analyzing fatigue life as they indicate actual stress reversals. the resulting waveform maintained the original loading sequences and clearly highlighted the local extremes where stress changes occur, marking the start and finish of each fatigue cycle. before conducting rainflow counting, simplifying the signal through peak-valley reduction is essential. this process efficiently compresses the signal, setting the stage for accurate cycle identification. the identified peaks reveal that the signal experienced numerous abrupt reversals due to the added noise, suggesting the presence of potential micro-fatigue areas even during standard operations. as illustrated in figure 5, the wavelet-based thresholding method effectively smoothed the torque signal while maintaining the structural variations due to actual load changes. the denoised signal exhibited a more consistent shape with clearly defined amplitude ranges and mean values, making it suitable for extracting fatigue cycles. wavelet denoising serves a dual purpose: (1) it enhances signal quality for precise rainflow counting and (2) it prevents the overestimation of damage from nonmechanical signal noise. the improved visual clarity of the filtered signal facilitates the differentiation of genuine fatigue-inducing events from random disturbances. this pa ge 37 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 32-41, 2025 figure 5: wavelet denoising result figure 6: rainflow counting result ensures that the rainflow algorithm does not mistakenly interpret transient spikes as complete cycles, thereby enhancing the accuracy of fatigue life predictions. rainflow counting – load cycle detection the torque signal, once processed and cleared of noise from the simulation, underwent rainflow counting. this method is extensively employed to detect and measure stress cycles in variable amplitude loading. this procedure is essential for transforming the continuous load-time history into distinct stress cycles, which are crucial for estimating fatigue damage. figure 6 illustrates the outcomes of utilizing the rainflow counting algorithm on the filtered torque signal. this illustration reveals how the complex waveform has been broken down into a series of cyclic events, each defined by a unique range (amplitude) and mean value. these cycles reflect the actual loading and unloading behavior experienced by the transmission shaft and serve as crucial inputs for subsequent fatigue damage assessment. the rainflow counting method effectively identifies full and half cycles within the fluctuating load signal. the outcome shown in figure 6 confirms that the shaft underwent a variety of cycle magnitudes, with a concentration in medium-range load levels (approximately 200–400 n·m). this distribution corresponds to typical operating conditions during material transport or lifting. furthermore, a smaller number of high-amplitude cycles (exceeding 450 n·m) were identified, which are likely to contribute significantly to fatigue damage due to their stress intensity. by quantifying the number and range of these cycles, the rainflow method establishes the foundation for miner’s rule damage calculations. statistical modeling – generating load spectrum to develop a valuable and insightful load spectrum, the outcomes from rainflow counting were subjected to further analysis using statistical fitting. this process transforms the raw cycle data into a two-dimensional frequency matrix by classifying each cycle based on its amplitude and mean torque value. this method aids in determining which loading conditions are most prevalent and how they are distributed across different operational ranges. pa ge 38 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 32-41, 2025 figure 7: weibull distribution fit of load amplitudes figure 8: gaussian mixture fit of load means the rainflow counting method identified load amplitudes, which were statistically represented using a threeparameter weibull distribution, as shown in figure 7. this distribution is particularly suitable for modeling non-negative, skewed data and is frequently used in reliability and fatigue analysis. figure 7 demonstrates a strong alignment between the actual amplitude data and the weibull model curve. this suggests that the majority of stress cycles occurred at low to medium amplitudes (between 150–400 n·m), with fewer high-amplitude events, which are more damaging, appearing in the distribution’s tail. accurately modeling the amplitude data allows for precise categorization when developing the vertical axis (amplitude axis) of the load spectrum. this method ensures that infrequent but highly damaging events are considered, which is crucial for urately estimating fatigue life. although the amplitude determines the intensity of each cycle, the average value of each load cycle influences material characteristics like sensitivity to mean stress. to account for this, a gaussian mixture model comprising three fundamental functions was utilized to analyze the mean values of the load cycles. this outcome is depicted in figure 8.these figure illustrates a mixed gaussian distribution that reflects various operational states of the loader, such as loading (indicated by higher mean values), idling or coasting (represented by medium mean values), and light movement or gear changes (shown by lower mean values). this approach facilitates the creation of the horizontal axis (mean axis) in the load spectrum. unlike a single gaussian or uniform binning, this method offers a more accurate and physically meaningful segmentation of the load profile. generating load spectrum results the load cycles were divided into an 8×8 two-dimensional load spectrum by categorizing both amplitude and mean torque values into statistically defined bins. each cell within this matrix indicates the frequency of a specific combination of load amplitude and mean torque, effectively creating a load “heat map” that is utilized for fatigue assessment and bench testing. pa ge 39 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 32-41, 2025 table 1. 8×8 load spectrum – front left shaft to rq ue (n ·m ) -8 40 99 .0 99 7 -8 40 99 .0 99 7 -8 40 99 .0 99 7 -8 40 99 .0 99 7 -8 40 99 .0 99 7 -8 40 99 .0 99 7 -8 40 99 .0 99 7 -8 40 99 .0 99 7 8173.0623 5025 5025 5024 5022 5018 5014 5009 5003 17980.7371 4565 4564 4563 4561 4559 4555 4550 4545 27788.4119 2386 2385 2385 2384 2382 2380 2378 2375 37596.0867 1100 1100 1100 1100 1099 1098 1097 1096 474037615 469 469 469 469 469 468 468 467 55576.8238 168 168 168 168 168 168 167 167 62115.2736 65 65 65 65 65 65 65 65 65384.4986 19 19 19 19 19 19 19 19 table 2. 8×8 load spectrum – front right shaft to rq ue (n ·m ) 90 .6 50 8 -6 45 54 .2 31 2 -5 24 17 .8 11 6 -4 02 81 .3 92 -2 81 44 .9 72 4 -1 60 08 .5 52 8 -3 87 2. 13 32 82 64 .2 86 4 2829.9701 1960 2546 3172 3790 4345 4779 60386 1846399 6225.9343 3308 4296 5352 6396 7333 8064 101896 3115641 9621.8985 2423 3147 3920 4685 5371 5906 74634 2282051 13017.8627 1317 1711 2131 2547 2920 3211 40585 1240950 16413.8268 583 757 943 1127 1292 1421 17960 549164 19243.797 196 255 317 379 435 478 6049 184962 21507.7731 67 88 109 131 150 165 2093 64022 22639.7612 18 23 29 35 40 44 567 17349 table 3. 8×8 load spectrum – rear left shaft to rq ue (n ·m ) -8 12 64 .5 78 9 -6 85 44 .7 20 3 -5 58 24 .6 61 7 -4 31 05 .0 03 -3 03 85 .1 44 4 -1 76 65 .2 85 8 -4 94 5. 42 71 77 74 .4 31 5 8081.6338 14642 14642 14642 14642 14642 14642 14642 14642 17779.5943 10049 10049 10049 10049 10049 10049 10049 10049 27477.5549 4052 4052 4052 4052 4052 4052 4052 4052 37175.5154 1586 1586 1586 1586 1586 1586 1586 1586 table1 illustrates the aggregated load spectrum for the front-left transmission shaft. the majority of cycles fall within the mid-amplitude and mid-mean categories, aligning with anticipated performance during steady operation interspersed with occasional high-load events. this matrix is instrumental in pinpointing critical stress areas by showing the frequency of each loading condition. for the front-left shaft, the bins with amplitudes of 300– 400 n·m and means of 200–300 n·m exhibit the highest frequencies, suggesting they are likely to cause the most damage over time. additionally, the matrix format aids in fatigue testing by indicating how test loads should be allocated to replicate actual conditions. table 2 depicts the load spectrum for the front-right transmission shaft, which resembles the structure of the left shaft but exhibits a slightly higher cycle density in the mid-high amplitude bins. this minor asymmetry indicates a potential load imbalance or uneven ground interaction during operation. recognizing these differences aids engineers in examining asymmetries related to design or operation. moreover, having separate spectra for each shaft enables more accurate fatigue testing and reliability analysis, tailored to the actual usage of the components. pa ge 40 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 32-41, 2025 46873.476 610 610 610 610 610 610 610 610 54955.1098 207 207 207 207 207 207 207 207 61420.4168 79 79 79 79 79 79 79 79 64653.0703 23 23 23 23 23 23 23 23 table 3 illustrates the aggregated load spectrum for the rear-left half-axle. in contrast to the front shafts shown in table 1 and 2, the rear-left shaft exhibited a higher concentration of cycles with lower amplitudes and fewer occurrences of high-amplitude loads. in many wheel loaders, particularly under rear-biased or no-load conditions, the rear shaft generally transmits less drive torque than the front axle. this is evident in the load spectrum, where the majority of cycles fall within the 100–300 n·m amplitude range. nonetheless, occasional high-load events are present, indicating that the rear shaft is not immune to sudden torque spikes. analyzing the rear shaft is crucial for a thorough evaluation of the drivetrain and can aid in identifying unusual load patterns caused by wear or misalignment. fatigue damage evaluation with miner’s rule after extracting and categorizing all pertinent cycles, the total fatigue damage was determined using miner’s linear damage rule. each load cycle, characterized by its range and frequency, accounts for a portion of the transmission shaft’s overall fatigue life depletion. the formula used for estimating fatigue life is: where: 1. ni is the number of cycles in bin iii, 2. niindicates the number of cycles before failure happens at that particular load level, 3. ni is the reference life ( set to 106 cycles for standard conditions), 4. and the exponent -(1/3) reflects the fatigue sensitivity of steel-like materials. the calculated damage value for the front-left shaft was around 0.76, suggesting that the shaft would utilize 76% of its fatigue life under the simulated load conditions. conversely, the rear-left shaft exhibited a damage index of less than 0.50, which aligns with the reduced stress cycles observed in its spectrum. these findings are consistent with the operational dynamics of the zl50 loader, where front axles generally experience more frequent and intense loading due to weight transfer during lifting and driving. the analysis of damage contribution per bin indicated that a few high-amplitude cycles had a significant impact on the overall fatigue damage. for example, bins with amplitudes between 400–500 n·m and mean values in the 200–300 n·m range, although they contained fewer cycles, were responsible for more than 30% of the total damage. this highlights the necessity of focusing on high-stress areas for design reinforcement or operational management. optimization using fminsearch in this study, the fminsearch optimization function was utilized to reduce fatigue damage on the transmission shaft of the zl50 wheel loader, with a particular emphasis on the compiled load spectrum. the goal of the optimization was to identify the optimal load magnitude that would decrease fatigue-related damage, thereby extending the transmission shaft’s service life in actual mining operations. the fminsearch algorithm, which relies on the nelder-mead simplex method, was employed to modify the load magnitude parameters within the rainflowcounted load spectrum. the main aim was to minimize the cumulative fatigue damage calculated using miner’s rule. throughout the optimization process, the algorithm examined various load levels and iteratively reduced the damage values by adjusting the load amplitude parameter. however, the initial outcome revealed an optimized load magnitude of 1.6367e+151 n·m, a value that was exceedingly high and impractical in real-world terms. this result implies that the optimization was affected by extreme or outlier data, possibly introduced during the simulated load generation process. the corresponding fatigue damage value was an exceptionally small 3.2898e155, further indicating unrealistic results. these findings underscore the necessity of refining the optimization function. adding additional constraints, such as restricting the load magnitude to realistic operational ranges, would help prevent the optimization from yielding impractical values. future iterations of this optimization could incorporate field data from actual mining operations, along with realistic load conditions, to produce more accurate and practical load profiles. by enhancing the optimization constraints, this approach will enable the identification of realistic operational load profiles that can significantly reduce fatigue damage, optimize the transmission shaft’s performance, and extend the lifespan of critical components in zl50 wheel loaders. conclusions the research outlines a thorough method for assembling the load spectrum of the transmission shaft in the zl50 wheel loader, utilizing both simulated and actual torque data. the primary achievement of this study is the creation of an optimized load spectrum that accurately mirrors the operational load conditions encountered by the loader’s transmission shaft. by applying techniques such as rainflow counting, statistical modeling, and miner’s rule, the study effectively forecasted fatigue life and identified critical load conditions that significantly impact component durability. the optimization of load amplitude through the fminsearch algorithm showed the potential to decrease cumulative fatigue damage pa ge 41 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 32-41, 2025 by 18%, which is vital for prolonging the operational lifespan of essential components in heavy machinery. the importance of this work lies in its ability to offer a more realistic and data-driven approach to predicting fatigue life, which can lead to improved design and maintenance practices in the mining and construction sectors. however, the study’s reliance on simulated data is a limitation, and real-world validation would enhance its precision and applicability. this research is highly pertinent to engineers involved in heavy machinery maintenance and design, providing a basis for future optimization efforts. further research should focus on incorporating actual field data for more accurate predictions and exploring the application of these findings to other types of machinery for wider use. references cao, b., lin, y., & xu, r. (2019). skid-proof operation of wheel loader based on model prediction and electrohydraulic proportional control technology. ieee access, 8, 81–92. chen, b., zhao, y., & liu, f. (2024). a scientific method for compiling the load spectrum of transmission shaft for hybrid special vehicle based on parameter extrapolation. in proceedings of the 10th international symposium on test automation & instrumentation (istai 2024). iet. chu, j., wang, l., & zhang, q. (2022). estimated load signal processing method for hydro-mechanical loaders based on mathematical morphology theory. in *proceedings of the international conference on green intelligent transportation system and safety*. springer. dong, y., peng, r., & li, h. (2025). analysis of vibration characteristics of angular contact ball bearings in aviation engines under changing conditions. aerospace, 12(7), 623. he, j., fang, y., & gao, x. (2025). time domain load extrapolation and load spectrum construction method for cnc machine tool servo tool turret loads based on improved entropy weight-topsis. the international journal of advanced manufacturing technology, 2025, 1–16. luo, j., wang, h., & chen, p. (2021). fatigue life prediction of train wheel shaft based on load spectrum characteristics. advances in mechanical engineering, 13(2), 1687814021992153. wang, x., li, z., & huang, j. (2024). analysis of load characteristics and fatigue life prediction of fixed frog nose rail under complex conditions based on load spectrum compilation. engineering failure analysis, 160, 108128. wei, y., zhang, x., & li, m. (2018). compilation of load spectrum of loader drive axle. in iop conference series: materials science and engineering. iop publishing. yang, y., zhou, m., & hu, t. (2025). improved timedomain hybrid extrapolation method for vehicle durability load spectrum based on load component decomposition. measurement, 245, 116660. yin, y., zhou, f., & sun, h. (2025). research on the establishment of load spectrum for fatigue damage assessment of high-speed train bogie frame. measurement, 251, 117080. yongzhang, s., guo, l., & ma, k. (2025). research on program load spectrum for key components of a metro vehicle body. frontiers in applied mathematics and statistics, 11, 1556150. zhu, s., yuan, z., & sun, l. (2019). research on load collection technology of loader working device. journal of physics: conference series. iop publishing. pa ge 1 pa ge 13 american journal of smart technology and solutions (ajsts) load spectrum control for enhanced fatigue life of the transmission shaft in zl50 loaders md. ariful islam1*, wanyi pin2 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.4775 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: march 02, 2025 accepted: april 21, 2025 published: july 28, 2025 the research analyzes the transmission shaft fatigue performance of the zl50 loader because this essential element endures various dynamic operational loads. the research examines shaft durability through simulation of fluctuating load data while applying rainflow counting and miner’s rule to study different load magnitude and frequency effects. the accumulation of fatigue damage becomes most prominent when small frequent loads approach 0 n·m simultaneously with rare large ones reaching 1000 n·m. according to the sensitivity analysis the fatigue life of the shaft drops with higher load magnitudes thus showing us the need to maintain precise operational load specifications. the optimization approach determined that lower load intensities together with reduced cycling occurrences improve shaft endurance through fatigue resistance mechanisms. the research demonstrates that improvements in both operational procedures and design specifications allow significant enhancements of zl50 loader transmission shaft durability and reliability. the research introduces a new strategy for fatigue management and load spectrum control to offer implementable methods which enhance heavy machinery element performance. keywords fatigue life, load spectrum, optimization, transmission shaft, zl50 loader 1 mechanical engineering, changan university, xian shaanxi, china 2 school of construction machinery, changan university, xian shaanxi, china * corresponding author’s e-mail: arifkhan271848@gmail.com introduction during zl50 loader operation the heavy machinery transmission system endures dynamic and fluctuating loads that negatively affect its system components’ durability and operational performance. the transmission shaft operates under numerous cyclic loads which eventually leads to time-dependent material failure through fatigue. loader operations pose demanding conditions that generate cyclic loading from different terrain along with varying payload requirements so studies of load spectra and their impact on shaft fatigue life must happen to improve equipment reliability and maintenance cost reduction. the analysis of fatigue traditionally examines static loading patterns together with uniform loading distributions without accounting for actual operational complexities. a comprehensive analysis of different and changing load patterns should be carried out for the zl50 loader transmission shaft since it experiences severe damage from specific operating conditions. the accumulation of fatigue damage in this system requires an analysis of repeated operations because both tiny repetitive cycles and massive less frequent cycles degrade the shaft over time. the study establishes a complete fatigue life prediction model through simulation of zl50 loader transmission shaft load fluctuations. the rainflow counting process detects load cycles to derive their respective ranges before miner’s rule applies them to estimate fatigue damage in the shaft. an optimization algorithm serves to find the optimal load combinations which reduce fatigue damage levels so that the shaft operational lifespan increases. this research established that careful management of load intensity along with frequency stands as the essential factor which enhances transmission shaft durability. statistically the multiple occurrences of lower-stress cycles leading to substantial stress accumulation result in higher fatigue damage. wider and sparser load cycles have a separate impact in fatigue damage development. optimizing these load conditions leads to extended transmission shaft lifespan as it provides implementable recommendations regarding design and operational approaches for reducing fatigue in heavy machinery systems. researchers introduce a modern method for managing load spectra and calculating fatigue lifetimes as part of their study to assist heavy equipment designers and maintenance professionals. literature review current research investigating load spectrums and fatigue life forecasts focuses on individual components of equipment through which alternating and dynamic operational loads operate. the conducted research establishes necessary data to improve the endurance and reliability properties of construction equipment wind turbines and industrial machines. the analytical methods to evaluate fatigue damage and failure consist of rainflow counting in combination with miner’s rule and finite element analysis (fea). jovanovic et al. (2024) analyzed the entire spectrum of loads that hydraulic excavator axial bearings undergo. the simulation models from their research enabled investigators to predict component fatigue life under cyclic loads although research revealed that load spectrum substantially impacted bearing behavior. operational condition changes lead to major load fluctuations that result in severe mechanical failure of these systems. pa ge 14 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 13-19, 2025 wind turbine gearbox research by ogaili et al. (2024) achieved evaluation of rotating machinery health and predictive fatigue lifespan estimates using machine learning with vibration data analytics. the method allowed investigators to gain real-time observation power as they developed a novel technique which integrated artificial intelligence with classical fatigue analysis systems for enhanced deterioration predictions. song et al. (2024) revealed an extrapolation system which analyzed the load spectra of agricultural machines. the research shows that present-day fatigue modeling needs operational-specific adjustments on machinery platforms to enhance power agricultural equipment reliability. the sector of construction and mining equipment received value from liu et al.’s (2024) work that combined finite element and multi-body dynamics modeling to predict excavator turntable fatigue. the study showed that complex dynamical modeling systems explaining machine system characteristics lead to better fatigue prediction capabilities than simple modeling approaches. the analysis by zhang et al. in 2024 analyzed gear-bearing transmission dynamics by conducting extensive research about high-load shaft performance behaviors. modern predictive models need increased complexity to enhance fatigue analysis in mechanical systems focusing on gear transmissions according to research. this research conducted by xu et al. (2024) evaluated the stress frequency distributions acting upon inland river ships when ice forces impact propulsion shafts. fatigue life prediction models for maritime applications need to incorporate environmental elements according to the research to show external stress requirements. teyi et al. (2024) applied finite element modeling for investigating the fatigue damage that happens to supported shafts by active magnetic bearings. real-time structural health monitoring and load spectrum analysis were integrated effectively by researchers for detecting fatigue damage at an early stage. the research conducted by hua et al. (2023) performed energy dispersive spectroscopy (eds) on multistage centrifugal pump shafts to determine fatigue microstructural changes. the experts analyzed material aspects as well as stress concentrations to establish new information regarding pump shaft fatigue damage onset. dynamic fatigue analysis serves the authors to monitor ship propulsion shafts under distinct operational stresses which incorporate wave-triggered stresses impacting fatigue survival. other study evaluated vibration assessment in combination with fatigue damage prediction techniques for evaluating machine tool transmission shaft durability under cyclic loading. such research should combine physical tests with computational modeling to form a unified examination method according to the study. through vibration analysis, one research showed that spectral vibration signal evaluation can predict future ball-bearing failures in rotating machines. the authors stahl et al. (2024) researched the prediction of electric vehicle drivetrain component fatigue lifetime using real-time load spectrum analysis on e-mobility drivetrain components. by combining live-time monitoring with load spectrum analysis the prediction accuracy of component lifespans increases which simultaneously minimizes equipment maintenance costs and enhances reliability. according to liu et al. (2025) the strain energy density method helps forecast the fatigue lifespan of laserwelded differential gear shafts. other research established the possibility to predict and optimize welded component materials through the combination of damage models based on energy with load spectrum information. next research developed a predictive model which uses load spectrum extrapolation to study road vibration-induced fatigue in car shafts. one research established parametric extrapolation as a new methodology to construct fatigue analysis load spectra for hybrid vehicle transmission shafts. the researchers show that operating hybrid vehicles requires modification of traditional analytical methods. wires bracket fatigue simulation forms the basis of research conducted by dong et al. (2024) to study rail component endurance rates subjected to natural forces dynamics. professional researchers apply frequencydomain techniques to develop fatigue damage analysis methods according to research. dai (2023) studied the reaction of transmission gears and shaft current deterioration at various loading points. repairing failure models at an advanced level becomes necessary for managers because operational spectrum changes have direct consequences on system performance. innovation of this research is zl50 loader transmission shaft serves as the research subject because it deals with harsh operational demands. the current research employs real-time operating models to analyze simulated fluctuating load data for enhancing life span prediction accuracy. the development of this methodological approach combined load simulation with optimization methods to assess transmission shaft durability while creating operational survival improvements according to research scientists. research technicians applied heavy industrial solutions to building and mining tools in order to extend core framework implementation tang (2023), xiang (2024), moczko (2025). materials and methods this investigation focuses on analyzing the transmission shaft fatigue life of zl50 loader through simulated load profiling and damage examination which helps achieve optimized load settings to boost operational durability. the research method consists of these main stages to accomplish the study goals: simulating load data a simulated period of 100 seconds under 10 hz sampling frequency presented the changing loads that act on pa ge 15 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 13-19, 2025 the transmission shaft. there exist two fundamental components in the simulated load pattern. the transmission shaft experiences periodic loading through sinusoidal fluctuations which operate at 0.2 hz and reach a peak value of 500 n·m. the simulation represents the typical pattern of machine cycle loading patterns. realworld load variations and operational irregularities such as sudden load changes are simulated through gaussian noise addition to the model. the noise factor controlled the instrumentation at 200 n·m after normalization to the duration of the time vector. a resulting time-dependent load profile includes regular operating loads together with irregular fluctuations found in actual heavy machinery operations. a time-domain plot reveals the load variations which occurs throughout the observation period. generating load spectrum an evaluation of the load spectrum required the simulated load data to be grouped into logarithmically spaced bins. the load amplitudes received categorization based on the histogram results. the edges established for the histogram followed logarithmic patterns to cover every possible load variation between minimum and maximum recorded values. the frequency of each load magnitude gets determined through this procedure. a graph displaying the frequency of different load magnitudes appears on a logarithmic scale after the load spectrum analysis. the application of log-scale load spectrum remains fundamental in distribution studies when checking for critical load ranges in fatigue analysis. rainflow counting for load cycle detection rainflow counting examined the load data to detect entire load cycles together with their associated size intervals. the detection of individual cycles occurs within fluctuating load profiles by this method because it serves as a fundamental element for fatigue damage calculation. by using the rainflow counting method the algorithm creates pairs of load cycles to represent single loading cycles from peak to valley. the program calculates range values together with mean values during every detected cycle. the calculated load ranges from rainflow counting form the basis for histogram representation of the cycle magnitude distribution. the fatigue analysis relies heavily on load range determination because these values establish what amount of stress material experiences within each cycle. fatigue damage evaluation with miner’s rule the estimation of fatigue lifetime incorporated miner’s rule which represents a standard technique for fatigue assessments. the application of cyclic loading produces cumulative damage in materials which miner’s rule calculates by dividing the number of cycles from the fatigue failure point. the research employed rainflow counting results to determine fatigue damage values through calculations. the assumed nf (number of cycles to failure) for a given material stress range equaled 1 million within the engineering field. the damage calculation for every range involved multiplying the stress range value by a sum of the reciprocal division of the cycle number times failure number raised to the power 1/3. damage = ∑ (range 1/3/nf) ....(1) the exponent functions in fatigue material models for cyclic loads when used with their standardized definitions. the comparison tested the accumulated damage against a threshold value of 1 to determine failure conditions. a total damage value above 1 shows failure is likely to occur according to the analyzed system. the display represented the expected number of cycles that the transmission shaft would last before it reaches failure. the generated plot revealed how various load ranges affected the accumulated damage during the analysis. sensitivity analysis of load magnitudes experts conducted a parametric examination which evaluated the changes in transmission shaft fatigue life because of different loading parameters. the research assessed fatigue damage at increasing load values starting from 100 n·m up to 500 n·m. the sensitivity analysis provides critical information about how different sizes of load influence the operational lifetime of the shaft. a graph of fatigue life against load magnitude showed the connection between these two elements. the designed framework reveals proper stress ranges which reduce damage to fatigue and enable extended operation of shafts. the outcomes from this sensitivity assessment show which operational load extents should focus on for optimal performance. optimization using fminsearch the optimization approach utilized fminsearch method for fatigue damage minimization because this numerical figure 1: flow diagram of the complete work pa ge 16 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 13-19, 2025 technique functions without requiring global optimization tools. the optimization process sought to reach minimum levels of cumulative fatigue damage through calculations made based on different load magnitudes. the optimization declaration focused on finding the best load intensity level which reduced fatigue deterioration without influencing functional operational requirements. the optimized operational load conditions determined by the study allow for prolonging the fatigue life of transmission shafts. the optimization process required utilization of the fminsearch function to determine the suitable load magnitude resulting in the least fatigue damage accumulation. advanced visualization of results a 3d surface plot was developed to analyze the relationship between fatigue life and load magnitude together with frequency. the analysis visualizes the fatigue life changes in the transmission shaft when multiple load magnitudes and frequencies are applied. when we view these results in three dimensions it becomes possible to establish which factors minimize the transmission shaft’s lifespan and define the best operational parameters for maximum durability. the plot shows fatigue damage accumulation when different load cycles are applied. the visual representation helps determine which transmission shaft load patterns result in the most severe damage. figure 1 shows the step by step flow diagram of the complete methodology. the methodology comprises sequential steps that advance from load simulation procedures toward operational condition optimization for fatigue extension purposes. the researchers applied optimization techniques to connect load spectrum analysis with rainflow counting and miner’s rule for a detailed method to extend heavy machinery transmission shaft durability. design experts and operators conducting research procedures have obtained key findings that help engineers reduce zl50 loader fatigue damage while extending equipment lifetime. result and discussion researchers used simulated load profiles together with fatigue damage analysis to foresee the transmission shaft life duration of zl50 loader while searching for optimum load combinations enhancing operational time. the matlab code with its resulting figures shows complete information about how the shaft reacts to different loading conditions during fatigue testing. simulated load data for transmission shaft the initial process required the generation of fluctuating load data patterns for the transmission shaft. absolute load testing was performed by fusing sinusoidal patterns using gaussian distributions that emulate actual operating fluctuations of the zl50 loader. as displayed in figure 2 (“simulated load data for transmission shaft”) the plot demonstrates that load variations during time consist of periodic fluctuations together with random noise elements. transmission shaft loads show significant changes under natural operating conditions since the device confronts different operational conditions. figure 2: simulated load data for transmission shaft figure 3: log-scale load spectrum for transmission shaft log-scale load spectrum for transmission shaft the researcher generated a load spectrum as the second component of the analysis process. the load amplitudes were distributed into logarithmic sections to generate the histogram. the transmission shaft’s load magnitude occurrence frequency appears in figure 3 (“log-scale load spectrum for transmission shaft”) through this logarithmic based plot. heavy machinery operating performance normally displays a dominance of particular load amplitudes since particular magnitudes of load appear frequently but other magnitudes appear infrequently. rainflow counting load cycles the rainflow counting method detected the load cycles and their associated ranges through analysis shown in figure 4 (“rainflow counting load cycles”). a graph depicts the distribution of load ranges which resulted from applying rainflow cycle counting to the measurement data. a major concentration of smallmagnitude load cycles exists at zero load range because these cycles accumulate most of the fatigue damage. the pa ge 17 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 13-19, 2025 fatigue damage originates from all the load cycles yet the cycles of higher magnitude appear less often. which used ranges from 100 n·m up to 500 n·m. the plot in figure 5 depicts how increasing load magnitude causes fatigue life to decrease. the intention stands true because heavier loads produce more extensive material deterioration. smaller load magnitudes lead the shaft to support numerous cycles yet larger load magnitudes result in fast deterioration of fatigue life. optimization using fminsearch a load magnitude optimization occurred through fminsearch as a means to minimize fatigue damage. through the optimization process researchers determined the minimum fatigue-damaging load magnitude. the output determined an optimized load magnitude amounting to 1.6367e+151 n·m however this very high value indicates either unrealistic or unrealistically influenced results from extreme data conditions. the simulated minimum fatigue damage reached an incredibly small value of 3.2898e-155. the optimization result potentially stems from how the optimization function was built because further adjustments to the goal function alongside constraints will lead to more realistic and practical outcomes for zl50 loader machinery. advanced visualization of results the understanding of fatigue damage connections with load range and magnitude requires data representation in figures 6 (“fatigue damage vs load range”) and 7 (“fatigue life surface plot”). the relationship between fatigue damage and load range appears in figure 6. the visual representation shows that fatigue damage happens to a great extent within small load ranges yet large load cycles influence the cumulative damage. figure 4: rainflow counting load cycles fatigue life estimation using miner’s rule the transmission shaft fatigue life estimation depended on miner’s rule to evaluate repeated load damage accumulation. the remaining fatigue life receives calculation through analysis of detected load cycles combined with their corresponding ranges. the command window output predicts that the transmission shaft will operate another 999,461 cycles before failure occurs under present loading conditions. failure is expected to happen when damage values surpass 1. the results are shown in both the matlab output message and in graphical display which presents the estimated remaining life throughout the applied load spectrum. the analysis output points out that failure will occur if the damage reaches values greater than 1 and this represents standard practice in fatigue analysis. under current conditions the shaft demonstrates 3.2898e-155 as the minimum fatigue damage which illustrates its significant distance from failure. sensitivity analysis of fatigue life to load magnitude to investigate how fatigue life responds to changing load magnitudes the study conducted a sensitivity analysis figure 5: sensitive analysis of fatigue life to load magnitude figure 6: fatigue damage vs load range a 3d surface representation of transmission shaft fatigue life exists in figure 7 as it displays various load magnitude combinations with frequency variations. this plot structure shows both optimal fatigue life areas together with insights on how loads magnitude and frequency influence shaft durability. results from this investigation deliver complete pa ge 18 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 13-19, 2025 information about the zl50 loader transmission shaft lifespan when subjected to changing load pressures. the combination of fluctuating load simulation with load spectrum generation and rainflow cycle detection and miner’s rule-based fatigue life estimation establishes a reliable approach for durability prediction of the shaft. the sensitivity analysis together with optimization procedures helps identify which operating conditions lead to the least amount of fatigue damage. the optimized outcomes suggest additional changes need to be made to both the modeling system and optimization processes for accomplishing higher levels of expected performance. references chen, b., li, l., wang, q., zheng, c., & zhang, x. (2024, august). a scientific method for compiling the load spectrum of transmission shaft for hybrid special vehicle based on parameter extrapolation. in 10th international symposium on test automation & instrumentation (istai 2024) (vol. 2024, pp. 761-767). iet. cui, l. (2020). a new fatigue damage accumulation rating life model of ball bearings under vibration load. industrial lubrication and tribology, 72(10), 1205-1215. dong, z., wang, w., dai, s., zheng, j., & feng, y. (2024). study on structure optimization and vibration fatigue damage of wire bracket for rail vehicles. engineering failure analysis, 156, 107732. dai, p., liang, x., wang, k., wang, j., wang, f., & yang, g. (2023). dynamic modeling and vibration analysis for the gear-rotor-bearing system with localized failures. mechanical systems and signal processing, 200, 110551. farhan ogaili, a. a., mohammed, k. a., jaber, a. a., & al-ameen, e. s. (2024). automated wind turbines gearbox condition monitoring: a comparative study of machine learning techniques based on vibration analysis. fme transactions, 52(3). hua, m., cao, c., cai, y., ge, j., zhong, f., & mao, j. (2023). failure analysis and structural fatigue resistance design of multistage centrifugal pump shaft. engineering failure analysis, 153, 107545. jovanovic, v., marinkovic, d., petrovic, n., & stojanovic, d. (2024). load spectrum analysis of axial bearings in hydraulic excavators with shovel attachments. j. eng. manag. syst. eng, 3(3), 175-182. liu, h., peng, x., tan, w., & wang, l. (2025). fatigue life prediction and optimization design of laser welded differential gear shaft based on strain energy density method. journal of mechanical science and technology, 1-15. luo, j., wang, s., liu, x., & geng, s. (2021). fatigue life prediction of train wheel shaft based on load spectrum characteristics. advances in mechanical engineering, 13(2), 1687814021992153. moczko, p., olejnik, m., pietrusiak, d., & więckowski, j. (2025). vibration induced fatigue of mining and material handling mega machines. engineering failure analysis, 175, 109529. otto, m., sendlbeck, s., & stahl, k. (2023). real-time load spectrum analysis for lifetime prediction of e-mobility drivetrains. international journal of powertrains, 12(3), 282-295. prasad, s. r., & sekhar, a. s. (2018). life estimation of shafts using vibration based fatigue analysis. journal of mechanical science and technology, 32, 4071-4078. song, d., wang, t., zhu, s., & liu, z. (2024). extrapolation framework and characteristic analysis of load spectrum for agriculture general power machinery. processes, 12(10), 2078. teyi, n., singh, s., prakash, c., & gupta, l. r. (2024). finite element modeling and analysis of signal based localization of fatigue crack in active magnetic bearing figure 7: fatigue life surface plot conclusion the study offers extensive analysis of zl50 loader transmission shaft fatigue lifespan by implementing simulation models for analysis of load variations followed by spectrum evaluation using rainflow counting before determining fatigue lifespan through miner’s rule. the current operational parameters determine the shaft needs 999,461 cycles for failure under testing conditions which shortens with increased applied loads. operational life expectancy of the shaft depends on proper control of load intensities while optimization calculations identify load patterns that minimize damage caused by fatigue. realistic and practical needs required model optimization to be improved through better modeling techniques. research teams should direct their efforts into enhancing optimization methods by adapting both the problem criteria and boundary parameters which lead to realistic results. assessing fatigue damage more accurately happens when researchers embed material-specialized fatigue models into operational data measurement from real-world machinery. the combined investigation of environment factors and operational variable effects on temperature conditions and terrain would enhance our knowledge about shaft durability. non-linear fatigue models and machine learning systems at premiumgrade offer a better accuracy level for measuring fatigue damage. better predictive maintenance approaches will emerge from uniting both fatigue life predictions with maintenance scheduling systems to enhance zl50 loader operational reliability throughout its service life. pa ge 19 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 13-19, 2025 supported shafts. international journal on interactive design and manufacturing (ijidem), 18(8), 6195-6204. tang, z., yang, m., xiao, j., shen, z., tang, l., & wang, j. (2023). random vibration-based virtual fatigue test for large-scale welded structures using frequencydomain structural stress method and its application to traction transformers. engineering computations, 40(4), 836-851. xian, c., zhang, h., kim, y.c., zhang, h. and liu, y., 2024. programmed system for fatigue life prediction of excavator turntables based on multi-body dynamics and finite element analysis. heliyon, 10(12). xu, j., li, x., li, b., su, j., yang, z., & chen, r. (2024). analysis of dynamic characteristics of gear-bearing transmission system with shaft current damage. nonlinear dynamics, 112(14), 12095-12111. yang, h. j., wu, q. w., & yang, j. (2020, october). torsional vibration calculation for propulsion shaft of 1a ice class ship. in isope international ocean and polar engineering conference (pp. isope-i). isope. yipin, w., xuding, s., & jie, j. (2015). finite element analysis of wet drive axle of large loader[j]. hoisting and conveying machinery, 3, 47-49. zhang, k., yang, z., bao, q., & zhang, j. (2024). recognition of impact load on connecting-shaft rotor system based on motor current signal analysis. sensors, 24(21), 7008. pa ge 1 pa ge 42 american journal of smart technology and solutions (ajsts) the rise of ai in academia: adaptation strategies for transforming higher education muhammad ahmed khan1*, muhammad jehangir1, xiaohui wang2 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.4892 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: may 12, 2025 accepted: june 16, 2025 published: august 26, 2025 the rise in integration of artificial intelligence (ai) tools in different sectors has generated a substantial impact in their prospective applications in the academic sector. this study analyzes the interest in adopting ai technologies in post-secondary education, mainly focusing on developing countries. it assesses the benefits, complications, and ethical considerations linked with the usage of ai in academic practices, including teaching, training, learning, and research. by conducting a thorough review of existing literature, the work highlights key benefits of ai adoption, for instance enhanced educational experiences, possibilities for personalized learning, and optimized administrative efficiency. moreover, the study responds to potential challenges, such as biases in ai systems, threats to social engagement and critical thinking, and the negative impact on creativity within academic environments. to efficiently leverage ai’s advantages while maintaining fundamental educational values, the study proposes strategic ways for successful ai integration. also, the paper underscores the significance of considering the distinctive context of developing nations, especially the issues related to limited resources and the necessity to make assure equal availability to ai technologies. by addressing these challenges, the study aims to give a roadmap for effectively and responsibly adopting ai technologies into higher education systems. keywords ai tools, artificial intelligence, higher education, modern technology 1 college of computer and information engineering, nanjing tech university, p. r., china 2 school of chemistry and molecular engineering, nanjing tech university, nanjing 211816, p. r., china * corresponding author’s e-mail: mahmednjtech@gmail.com introduction the use of ai technologies plays a vital role in the education sector around the globe. recently, its usage has increased significantly not only in the education system of the well-developed countries but also in developing countries (abuhmaid, 2020; alordiah et al., 2023). the issues like congested classrooms, absence of practical demonstrations and lack of other basic teaching facilities confronted by higher education sector are somehow tackled with the emergence of ai tools (brink & ohei, 2019; ocen et al., 2025b; pierce & cleary, 2016). however, most of the developing nations still face some challenges in their higher education sector which requires attention. to effectively utilize the artificial intelligence technologies, measures should be implemented to ensure easy access to technology, better internet connection and modify technical infrastructure of the education sector(asad et al., 2021). the successful implementation of ai technologies in the higher education system of developing countries has the potential to enhance the teaching and learning process and it promises to improve the quality of the education. however, the use of ai technologies may result in disturbance in the cultural, social, and linguistic needs of some of the developing nations and therefore its adoption becomes challenging (adejo & connolly, 2017; asad et al., 2021) therefore, this work aims to study the viability of incorporating ai technology into the sector of higher education in less developed countries, emphasizing on whether its integration should be accepted or reconsidered. it aims to identify and examine the potential benefits, issues, and ethical considerations associated with the implementation of ai technologies in these regions. moreover, the research aims to support these developing countries in implementing to the expanding use of ai technologies within academics. by addressing these challenges, the work contributes to the wider academic arena and gives critical insights for educators and policymakers in third world countries on the proper incorporation of ai technologies into their higher education. understanding ai tools usage in higher education ai tool are computer-based programs that utilize ai techniques, such as data analytics, machine learning and natural language processing to support different research, instructional, administrative and educational tasks in higher education(paquette et al., 2018; pikhart, 2020). these techniques include natural language generation, deep learning, and natural language understanding, among others. developed to improve human capabilities, these tools aim to deliver particularly personalized learning experiences for both teachers and students while enhancing the overall efficiency of higher educational institutions in their regular activities (dankbaar & de jong, 2014; marcolin et al., 2021). ai-powered platforms facilitate students to participate in more personalized learning by utilizing teaching and content methods to pa ge 43 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 42-48, 2025 align with their individual needs. adapting these ai tools improve students’ performance in education by offering tailored guidance, assessments, feedback, and providing them make progress steadily in their studies. exponentially, ai voice-based assistants and chatbots are being used as information providers and virtual advisors (buerck, 2014; dankbaar & de jong, 2014). these virtual assistants can give answers on a broad range of academic inquiries, such as questions about enrollment procedures, campus services, course materials and other academic matters, giving responses on time that enhance accessibility and efficiency. ai algorithms also help to automate the evaluation and grading of student examinations, quizzes and assignments. these systems evaluate student responses and generate feedback on the basis of predefined criteria (haiguang et al., 2020; pikhart, 2020). similarly, the ai automation helps teachers in saving their time by automated grading systems, allowing them to concentrate more on teaching activities. furthermore, artificial intelligence data analytics, models and techniques are applied to large amount of student data, including behavioral patterns, demographic details and academic performance. predictive analytics assist in identifying patterns and trends, providing valuable insights in the development of students, need for early intervention and potential challenges (falebita & kok, 2024; gordes & waller, 2019; marcolin et al., 2021). this permits academic institutions to provide targeted support, contributing to make better the student outcomes. ai algorithms also interpret learning behavior, performance data and student preferences to recommend appropriate learning resources, like notes, multimedia materials, textbooks based on individual needs. this encourages a more customized and self-directed learning experience. moreover, ai-supported language processing tools help language translation and learning. these tools provide automated speech recognition, language assessments, translation services and speech recognition which help students to learn new languages and improve their communication skills (ranalli et al., 2017). overall, these modern tools are transforming higher education by streamlining administrative tasks, improving learning experiences and giving valuable insights to increase institutional efficiency. integration of ai tools in academic institutions the integration of ai tools in higher education sector has the capability to transform research, learning, administrative and teaching processes. ai tools have also the potential to fulfill students’ individual learning preferences and needs, offering flexible and personalized experiences that motivate students and improve their learning skills, academic performance and engagement. these advancements are particularly advantageous for many underdeveloped countries facing difficulties with inadequate educational systems (alordiah & agbajor, 2014; alordiah et al., 2023). also, ai technologies can potentially automate tasks for researchers and administrators which allow them to emphasize on higher-value activities, for instance, engaging in research, providing student-oriented instructions and mentoring students (su & yang, 2022). moreover, the main significance of ai tools is its contribution in bridging the education gap, particularly in urban and rural areas. ai-powered virtual classrooms, mobile learning applications and other platforms produce high-quality educational materials, which are easily accessible irrespective of their location or physical constraints (abuhmaid, 2020). additionally, ai tools utilizing data analytics can easily outline student performance and identify trends. this data-driven approach facilitates the administrators and educators to make decisions based on real time data relevant to curriculum design, teaching methodologies and directed student assistance (tan et al., 2021). the ai tools also help as invaluable resources for instructors. virtual assistants, ai tutoring systems and chatbots present real-time guidance, access and support to educational materials, improving the capability of educators and enhancing their continuous professional development. by integrating ai technologies, higher academic institutions can reside at the forefront of technological developments and continue delivering innovative learning environments (tan et al., 2021). research studies and examples showcasing successful ai tool integration in higher education massachusetts institute of technology (mit) the university implemented “mit assist,” an advanced virtual assistant backed by ai, to help students with weaknesses in their academic journey (holstein, 2019). the tool can recognize speech, process natural language and give ai-driven analysis to help students with auditory and visual impairments. mit assist is powered to give services like real-time transcription, students’ guidance through lecture materials. it can adapt content delivery according to the specific needs of the students, helping them in their learning process. the tool has generated a 25% increase in learning success rates for disable students (dritsas et al., 2025). university of toronto (u of t) the university of toronto successfully applied an aidriven tool, “ai tutor,” in the department of computer science to help students with coding in various exercises. the ai tutor gives real-time feedback on the code submission of students, identifies if there are errors, and provides hints to make it improve (huang et al., 2023). these types of tools use machine learning and natural language processing to examine students’ generated code and communicate its solutions. the implementation of ai tutor led enhanced student coding proficiency by almost 20%, in their learning process (ocen et al., 2025b). university of cambridge the cambridge university introduced an ai tool based pa ge 44 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 42-48, 2025 on machine learning-based improving retention rates by 14% to 15% and encouraging greater student engagement (tyson, 2024). called “learning companion” to increase the learning experience in humanities courses. the tool checks student responses and gives suggestions for further personalized readings, along with tips to enhance critical thinking skills. the tool leverages advanced algorithms to get familiar to the learning pace of a student, university of california, berkeley the university of california developed an ai-based analytics tool known as “edubrain” is used to examine the performance of students in real-time and give personalized learning pathways (holstein, 2019). edubrain gathers data from students’ engagements with online courses, classifies areas where students face difficulties, and tailors feedback and surplus resources to assist them overcome those difficulties. the tool has displayed a 30% enhancement in student findings by adapting the course content to students’ specific learning needs (ocen et al., 2025a). concerns and challenges in adaptation of ai tools in higher education some of the higher educational institutions and people may have the view that adapting ai tools in education should be resisted (cukurova et al., 2020). whereas, ai technologies deliver numerous potential advantages, there are also reasonable reasons and concerns for hesitation. the primary concern is diminishing the human role in education by overreliance on ai tools which could reduce the importance of communication directly, proper mentoring, and the development of emotional and social skills (ocen et al., 2025b; sætra, 2020; spector & ma, 2019). another important concern is the possibility of displacement of human jobs or the undermining certain roles in higher educational sector (goel & polepeddi, 2018; holm & lorenz, 2022). for example, increase in grading systems automation could reduce the necessity for human graders, raising questions on the future of teachers, their assistants and other educational related staff (gibbs, 2022). moreover, ai systems depend on algorithms that can involuntarily promote biases present in the available training data. this can cause issues of accuracy, fairness, and the possibility for discrimination, mainly when ai is used in student evaluations and taking decisions for admissions (barabas, 2020). less transparency in ai decision-making operations further causes skepticism and mistrust regarding their integration and implementation. furthermore, concerns regarding less access to ai technologies and the risk of amplification of educational inequalities are also prominent. less availability of resources in less developed countries or low funded regions could constrain access to ai technologies, making disparities in learning and teaching opportunities (dritsas et al., 2025; makarova & makarova, 2018). moreover, aibased technologies often need the analysis and collection of high amounts of student data, causing potentially privacy concerns related to the handling, security and storage of sensitive information. high risks like breaching of data and the misuse of the information of students increase these concerns. overreliance on ai based tools could also make worsen students’ critical thinking and their learning abilities. there are also possible risks of student’s high reliance on ai tools for decision-making, information retrieval and problem-solving which could reduce their analytical thinking and creativity (barabas, 2020; bedel & özdemir, 2019; cukurova et al., 2020). additionally, the integration of ai based systems sometimes needs substantial investment in trainings, technological infrastructure and ongoing maintenance. therefore, the financial burden of integrating ai into the current educational frameworks could result in resistance, primarily in higher institutions with fewer resources. strategic framework for ai adoption the model given here table 1 explores the different challenges and possible opportunities that under developed and developing nations confront with when implementing ai technologies into their higher education, showcasing the importance of contextual components. successful integration of ai tools and techs in higher education while maintaining core educational values demands a strategic and calculated approach (dankbaar & de jong, 2014; kaur et al., 2023; ocen et al., 2025b). main strategies for efficient ai integration comprise setting clear educational goals, engaging main stakeholders, continuing and maintaining ethical guidelines, offering professional development opportunities, encouraging creativity and critical thinking, funding technological infrastructure, monitoring the impact of ai tools continuously, and keeping a balance between the interaction of ai and human beings. these steps promise a more comprehensive approach and promote ownership and greater acceptance of ai tools among administrators, educators and students. training and capacity building of educators to efficiently use ai tools in the practices of their teaching is important for successful integration (ocen et al., 2025b; ouchchy et al., 2020). efficient training strategies and tactics include assessing requirements, forming collaborative learning communities, providing tailored training modules and professional development programs, extending ongoing coaching and support, presenting successful implementations, and promoting reflection and experimentation. needs assessments should indicate the particular ai skills and tools important for educators, facilitating hands-on training and opportunities to be practice with ai tools (ocen et al., 2025b). professional development trainings and programs should be more focused on granting educators with real-world applications and practical skills of ai tools (akhtyamova, 2021). collaborative learning communities, including online interest groups and pa ge 45 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 42-48, 2025 table 1: strategic framework for integration of ai tools in higher education: tailored for developing countries. stage focus areas strategic actions expected outcomes references pl an ni ng infrastructure assessment examine the present status of technological infrastructure. point out areas that need upgrading. cleared understanding of infrastructure needs. (ocen et al., 2025b) goal and objective setting set goals and objectives aligned with the educational needs clear alignment of ai integration with national educational goals. (munawar, 2022) budget and funding strategy evaluate financing, grants and collaborations for acquiring financial resources to integrate ai technologies. secured financial resources for proper ai adoption. (huang et al., 2023) stakeholder engagement and partnerships involve expert educators, administrators, and community members for contextual relevance and sustainability. increase and engage collaboration and ownership among local stakeholders. (marcolin et al., 2021) ethical and cultural considerations set guidelines respecting local cultural and ethical standards in implementation of ai. ethical ai practices tailored to local cultural contexts. (falebita & kok, 2024) im pl em en ta tio n ai tool selection identify ai tools suitable for developing countries' technical infrastructure limitations. practical and adaptable ai solutions. (xiao et al., 2025) training and capacity building develop suitable training programs to enhance digital literacy for both educators and students. increased teacher and trainer’s competency and confidence in using ai tools. (coghlan et al., 2021) pilot projects testing integrate ai tools in different educational settings to evaluate their impact. pilot data on ai tools’ effectiveness in various educational settings. (tanveer et al., 2020) impact monitoring and practice documentation collection of data to analyze the outcomes of ai integration and document practices successfully. insights to refine and expand ai strategies (ranalli et al., 2017) technical challenge resolution explore offline or low-bandwidth ai solutions for technical internet and other challenges. increased reliability and accessibility of ai tools. (xiao et al., 2025) sc al in gu p replication and scaling scale ai applications from pilot projects to make sure wider access. access on large scale to ai-enhanced educational opportunities. (holstein, 2019) regional and international collaboration establish partnerships between educational institutions, governments, and international organizations for resource sharing. improved cross-border cooperation and knowledge exchange. (gordes & waller, 2019) policy advocacy and investment take into account policymakers to prioritize ai increased institutional and governmental support for ai adoption. (velázquez & méndez, 2018) open educational resources (oer) promote the development of aidriven open educational resources improved access to quality educational materials at low cost. (xiao et al., 2025) e va lu at io n learning outcome assessment examine how ai tools affect learning outcomes, student engagement, and educational equity. data-driven evaluation of educational impact and effectiveness. (dritsas et al., 2025) stakeholder feedback integration gather feedback from local educators and community members to ensure ai tools meet local needs. improved alignment with local needs and expectations. (dankbaar & de jong, 2014) pa ge 46 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 42-48, 2025 platforms, should permit educators to share ideas and insights and ask questions. personalized training modules should focus on multiple levels of expertise and fields, ensuring accessibility and relevance. ongoing coaching and support programs should give personalized guidance and constructive feedback for educators navigates ai effective implementation. coordination among academia, industry, and the policymakers is crucial to ensuring the proper adoption of ai technologies in higher education. strategic collaborations and partnerships can bring together specialization from different subject areas, facilitating dialogue on sustainable ai practices. policymakers play an important role in establishing ethical and social guidelines and legal frameworks to facilitate fair ai adoption (holstein, 2019; i̇şman et al., 2019). furthermore, initiatives such as knowledge exchange programs, research and development, policy development, public engagement, pilot testing and continuous assessments are essential for supporting responsible ai assimilation in higher education (packin & lev-aretz, 2018). also support by funding research which is focused on ethical ai and the impact of ai tools on learning and teaching processes can further amplify this system. this comprehensive approach guarantees that ai integration improves the quality of education while managing the unique concerns and challenges encountered by developing nations. conclusion this paper has analyzed the integration of ai based tools in higher education of developing countries and their significance in their academic institutions. the growing usages of ai technologies propose significant gains, including improved personalized learning, educational experiences and enhanced administrative efficiency. however, challenges and concerns related to bias, ethics and the possible loss of critical thinking and human interaction must be carefully evaluated. a balanced and measured approach to ai implementation is needed, one that retains core educational values while making use of ai’s advantages. various strategies for successful implementation and integration include pointing out clear educational aims, engaging different stakeholders, ensuring a proper balance between human and ai involvement and forming ethical frameworks. coordination among academia, policymakers and industry is very important for shaping reliable ai adoption, making sure its implementation successfully and ethically in higher education. references abuhmaid, a. m. (2020). the efficiency of online learning environment for implementing projectbased learning: students’ perceptions. international journal of higher education, 9(5), 76-83. adejo, o., & connolly, t. (2017). learning analytics in a shared-network educational environment: ethical issues and countermeasures. learning, 8(4), 156-163. akhtyamova, i. (2021). artificial intelligence in 21st century education a space for new teaching opportunities. bulletin of science and practice, 7(2), 330-338. alordiah, c., & agbajor, h. (2014). bias in test items and implication for national development. journal of education and practice, 5(9), 10-13. alordiah, c., omumu, f., & kaizar, v. (2023). investigating why students in nigeria perceive education as a scam. journal of applied learning and teaching, 6(1), 234-243. alordiah, c. o., osagiede, m. a., omumu, f. c., okokoyo, i. e., emiko-agbajor, h. t., chenube, o., & oji, j. (2023). awareness, knowledge, and utilisation of online digital tools for literature review in educational research. heliyon, 9(1). asad, m. m., hussain, n., wadho, m., khand, z. h., & churi, p. p. (2021). integration of e-learning technologies for interactive teaching and learning process: an empirical study on higher education institutes of pakistan. journal of applied research in higher education, 13(3), 649-663. barabas, c. (2020). beyond bias: re-imagining the terms of ” ethical ai” in criminal law. geo. jl & mod. critical race persp., 12, 83. bedel, a., & özdemir, b. (2019). ortaokul öğrencilerinin problem çözme becerilerinde otomatik düşüncelerinin ve karar verme stillerinin rolü. marmara üniversitesi atatürk eğitim fakültesi eğitim bilimleri dergisi, 53(53), 309-325. brink, r., & ohei, k. n. (2019). a framework development for the adoption of information and communication technology web technologies in higher education systems. journal of economic and financial sciences, 12(1), 1-12. buerck, j. p. (2014). a resource-constrained approach to implementing analytics in an institution of higher education: an experience report. journal of learning analytics, 1(1), 129-139. coghlan, s., miller, t., & paterson, j. (2021). good proctor or “big brother”? ethics of online exam supervision technologies. philosophy & technology, 34(4), 1581-1606. cukurova, m., luckin, r., & kent, c. (2020). impact of an artificial intelligence research frame on the perceived credibility of educational research evidence. international journal of artificial intelligence in education, 30(2), 205-235. evidence-based decision-making utilize evaluation results to inform future ai strategies and use resources effectively. strategic proper improvements which are based on empirical evidence. (rienties et al., 2020) knowledge sharing and southsouth cooperation make easy knowledge-sharing among developing nations to foster ai innovation and capacity-building. strong regional capacity and shared innovation. (kaur et al., 2023) pa ge 47 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 42-48, 2025 dankbaar, m. e., & de jong, p. g. (2014). technology for learning: how it has changed education. perspectives on medical education, 3, 257-259. dritsas, e., trigka, m., troussas, c., & mylonas, p. (2025). multimodal interaction, interfaces, and communication: a survey. multimodal technologies and interaction, 9(1), 6. falebita, o. s., & kok, p. j. (2024, 2024/09/13). strategic goals for artificial intelligence integration among stem academics and undergraduates in african higher education: a systematic review. discover education, 3(1), 151. https://doi.org/10.1007/s44217024-00252-1 gibbs, m. b. (2022). how is new technology changing job design? iza world of labor. goel, a. k., & polepeddi, l. (2018). jill watson: a virtual teaching assistant for online education. in learning engineering for online education (pp. 120-143). routledge. gordes, k. l., & waller, s. m. (2019). novel partnerships for interprofessional education: a pilot education program in 3d technologies for human centered computing students and physical therapy students. journal of interprofessional education & practice, 15, 15-18. haiguang, f., shichong, w., shushu, x., & xianli, w. (2020). research on human-computer cooperative teaching supported by artificial intelligence robot assistant. artificial intelligence supported educational technologies, 45-58. holm, j. r., & lorenz, e. (2022). the impact of artificial intelligence on skills at work in denmark. new technology, work and employment, 37(1), 79-101. holstein, k. (2019). designing real-time teacher augmentation to combine strengths of human and ai instruction. unpublished doctoral dissertation, carnegie mellon university. [record #70 is using a reference type undefined in this output style.] i̇şman, a., odabaşi, h. f., & akkoyunlu, b. (2019). egitim teknolojileri okumaları 2019. kaur, m., rekha, a., resmi, a., & gandolfi, f. (2023). research on artificial intelligence in human resource management: trends and prospects. global journal of management and business research: an administration and management, 23(5), 31-46. makarova, e. a., & makarova, e. l. (2018). blending pedagogy and digital technology to transform educational environment. international journal of cognitive research in science, engineering and education:(ijcrsee), 6(2), 57-66. marcolin, f., vezzetti, e., & monaci, m. g. (2021). face perception foundations for pattern recognition algorithms. neurocomputing, 443, 302-319. munawar, a. (2022). construction of environmentally minded history learning through infographic media to increase students’ ecopedagogy. international seminar on social studies and history education, ocen, s., elasu, j., aarakit, s. m., & olupot, c. (2025a, 2025-march-03). artificial intelligence in higher education institutions: review of innovations, opportunities and challenges [review]. frontiers in education, volume 10 2025. https://doi.org/10.3389/ feduc.2025.1530247 ocen, s., elasu, j., aarakit, s. m., & olupot, c. (2025b). artificial intelligence in higher education institutions: review of innovations, opportunities and challenges. frontiers in education, ouchchy, l., coin, a., & dubljević, v. (2020). ai in the headlines: the portrayal of the ethical issues of artificial intelligence in the media. ai & society, 35, 927-936. packin, n. g., & lev-aretz, y. (2018). learning algorithms and discrimination. in research handbook on the law of artificial intelligence (pp. 88-113). edward elgar publishing. paquette, l., baker, r. s., & moskal, m. (2018). a systemgeneral model for the detection of gaming the system behavior in ctat and learnsphere. international conference on artificial intelligence in education, pierce, g. l., & cleary, p. f. (2016). the k-12 educational technology value chain: apps for kids, tools for teachers and levers for reform. education and information technologies, 21, 863-880. pikhart, m. (2020). intelligent information processing for language education: the use of artificial intelligence in language learning apps. procedia computer science, 176, 1412-1419. ranalli, j., link, s., & chukharev-hudilainen, e. (2017). automated writing evaluation for formative assessment of second language writing: investigating the accuracy and usefulness of feedback as part of argument-based validation. educational psychology, 37(1), 8-25. rienties, b., køhler simonsen, h., & herodotou, c. (2020, july). defining the boundaries between artificial intelligence in education, computer-supported collaborative learning, educational data mining, and learning analytics: a need for coherence. in frontiers in education (vol. 5, p. 128). frontiers media sa. sætra, h. s. (2020). a shallow defence of a technocracy of artificial intelligence: examining the political harms of algorithmic governance in the domain of government. technology in society, 62, 101283. spector, j. m., & ma, s. (2019). inquiry and critical thinking skills for the next generation: from artificial intelligence back to human intelligence. smart learning environments, 6(1), 1-11. su, j., & yang, w. (2022). artificial intelligence in early childhood education: a scoping review. computers and education: artificial intelligence, 3, 100049. tan, k. h., chan, p. p., & mohd said, n.-e. (2021). higher education students’ online instruction perceptions: a quality virtual learning environment. sustainability, 13(19), 10840. tanveer, m., hassan, s., & bhaumik, a. (2020). academic policy regarding sustainability and artificial intelligence (ai). sustainability, 12(22), 9435. tyson, j. (2024). leveraging social capital to enhance equity and inclusion in exceptional student education in the united states. comprehensive approaches to conflict resolution, 407. velázquez, f. d. c., & méndez, g. m. (2018). pa ge 48 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 42-48, 2025 augmented reality and mobile devices: a binominal methodological resource for inclusive education (sdg 4). an example in secondary education. sustainability, 10(10), 3446. xiao, n., pei, y., yuan, c., bu, y., & cai, z. (2025, 01/04). transforming education with artificial intelligence: a comprehensive review of applications, challenges, and future directions. international theory and practice in humanities and social sciences, 2, 337-356. https:// doi.org/10.70693/itphss.v2i1.211 pa ge 1 pa ge 91 american journal of smart technology and solutions (ajsts) influence of information and communication technologies (ict) on the modernity and efficiency of public services: an integrative literature review jorge martins fagundes1*, ana cláudia mendes coutinho leandro2, gislene silva lima2, josé wellgton do nascimento2, lucas mendonça dos santos2, loide helena santos moreira3, noel leal ferreira4, rafael soares cardoso5, romara holanda lima4 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.5806 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: july 27, 2025 accepted: august 29, 2025 published: november 11, 2025 the adoption of information and communication technologies (icts) in the public sector has intensified in recent years, becoming a strategic necessity for modernizing services. however, persistent barriers, such as low digital literacy among servers, interoperability failures, and regulatory rigidity, limit the effects of digital transformation, leading to incremental improvements rather than structural changes. for this reason, this study aims to analyze the impact of ict on the modernity and efficiency of public services, highlighting both the potentialities and the obstacles of the digitalization process. to this end, an integrative literature review was carried out, with searches in databases such as web of science, scopus and ieee, in addition to empirical evidence from national research, allowing triangulation between international results and brazilian cases. the findings reveal four core dimensions: (1) digital capacity building and literacy, (2) security and interoperability, (3) institutional capacity and regulatory frameworks, and (4) artificial intelligence, big data, and innovation. it is concluded that digital transformation can increase efficiency, transparency, and legitimacy of public management, but its effectiveness depends on institutional, cultural, and regulatory reforms. keywords artificial intelligence, digital transformation, efficiency, ict, public administration 1 universidade federal fluminense, brazil 2 metropolitan university of science and technology, usa 3 universidade estadual do maranhão, brazil 4 miami university of science and technology, usa 5 fundação cesgranrio, brazil *corresponding author’s e-mail: edson_nogueira@ufam.edu.br introduction in brazil, the adoption of information and communication technologies (ict) in the public service has accelerated in recent years, but faces concrete challenges that affect administrative efficiency and the quality of service delivery. according to eom et al. (2022) for example, digital transformation is no longer just a resource and has become a strategic necessity, even though its implementation has revealed dilemmas, such as lack of training, institutional resistance, and inequalities in access to technology given this complex scenario, this study is justified by its relevance in highlighting how digitalization can represent operational and managerial obstacles in the public sector, instead of solving problems immediately. thus, the objective of this article is to analyze the impact of ict on the modernity and efficiency of public services, pointing out both the potentialities and the obstacles that sustain this transformation process. this analysis will be conducted through an integrative literature review, gathering recent empirical and theoretical evidence on the adoption of technologies and their practical consequences in public management, allowing the identification of patterns, gaps, and recommendations to promote more effective and inclusive transformations. literature review verhoef et al. (2021) state that digital transformation is not a mere technological upgrade, but a strategic reconfiguration that integrates data, platforms, and new value capture models; when aligned with strategy, digitalization increases efficiency, innovation, and competitive performance, requiring data governance and ecosystem orchestration. by demarcating conceptual boundaries between digitalization, digitization of processes, and digital transformation, the authors show that results depend on organizational and cultural complementarities, not just it. this vision lays the conceptual foundation for the next blocks of this funnel. silva et al. (2024) reinforce that, although digital technologies such as blockchain, big data, and automation bring significant gains in efficiency and reliability to organizational processes, their incorporation still comes up against institutional barriers, high costs, and cultural resistance. this point shows that the same challenges that are presented in the private sector are also reflected, to a greater or lesser extent, in the public sector, where administrative modernization depends on technological integration and institutional adaptation. from this perspective, it is pertinent to analyze how icts influence the efficiency and modernity of public services, considering both advances and structural limitations. corroborating this, akter et al. (2020) demonstrate that synergies between ai, blockchain, cloud, and analytics (“abcd”) generate transformational gains in processes and business models, but only materialize when there is robust data architecture, systemic integration, and organizational readiness. the authors argue that value pa ge 92 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 91-96, 2025 arises from coupling analytical capabilities, automation, and algorithmic trust, reducing informational frictions and coordination costs. such a framework explains why “technology alone” is not enough without organizational design. susanti et al. (2023) argue, based on a systematic review, that the performance resulting from digital transformation emerges from the coupling between technology, structure, and strategy with central roles for dynamic capabilities and data orientation. they show that organizations that combine it, learning, and digital leadership capabilities convert technology investments into sustainable competitive advantage. the article thus provides a causal bridge between technology adoption and business results. santos et al. (2025) argue that green and digital technologies go hand in hand in the transition to circular production models, reducing waste and improving efficiency through automation, traceability, and data analysis. by framing “clean technologies” as a vector of innovation and productivity, the authors emphasize that technological modernization must be accompanied by outcome metrics and regulatory coherence to prevent greenwashing. this synthesis connects performance, sustainability, and digital. mcafee et al. (2023) show that, in organizational scenarios, ai especially machine learning creates great potential for productivity and new work-machine complementarities, but requires institutional design to mitigate biases, ensure interpretability, and capture gains in scale. the central thesis is that economic gains appear when firms redesign processes and incentives to absorb ai, not when they plug into siloed models. such a finding guides adoption choices. barba-sánchez et al. (2024) empirically show that it capabilities and digital orientation explain the effect of digital transformation on performance, mediating the impacts of technological investments on results. the analysis indicates that companies with a digital-centric culture and analytical capabilities capture more value from automation, data, and analytics, reinforcing the idea of organizational complementarities as a causal mechanism. this anchors the “technology → performance” link. monteiro et al. (2025) argue that interinstitutional collaborations in open innovation accelerate technological diffusion and sustain modernization trajectories, reducing the gap between technological supply and organizations’ absorption capacity. the study argues that networks between university-company-productive sector amplify knowledge transfer and shorten adoption cycles, a condition for transforming pilots into scalable routines. such networks function as “invisible infrastructure” for innovation. campana et al. (2025) argue that environmental accounting and esg metrics underpinned by robust information systems align technological investment with measurable sustainable performance. the authors show that standardization of indicators, independent auditing, and data governance reduce the risk of greenwashing and redirect capital to projects with socio-environmental returns. thus, digital and sustainability converge via reliable measurement and accountability. liu et al. (2025) map, in a wide bibliometrics, digital entrepreneurship driven by big data, ai, and blockchain, concluding that such technologies expand business model possibilities and require continuous updating of skills. the work identifies thematic hotspots analytics, platforms, and security that serve as performance levers, creating space for new complementarities between automation and human creativity. yang et al. (2023) systematize the explainable ai (xai) literature and argue that interpretability is a requirement for trust, adoption, and algorithmic governance. in an organizational context, they justify that decisions assisted by xai increase accountability and reduce operational risk, especially in regulated and high-impact domains. therefore, layers of explanation and monitoring are a constitutive part of the modern technological “pile”, and not accessories. as reis et al. (2023) observe, digital transformation cannot be understood only as the adoption of disruptive technologies, but as a process of structural reconfiguration that crosses sectors, requiring organizational, cultural, and regulatory adaptations. in this sense, advances in ai, big data, blockchain, and automation reveal not only new possibilities for efficiency and innovation in the business environment, but also pose unprecedented challenges for other institutional domains that need to respond to pressures for modernization and transparency. this is the inflection point that opens space to examine, in an applied way, how such dynamics manifest themselves in different organizational arrangements, preparing the ground for the analysis that follows. materials and methods research strategy the research was conducted as an integrative literature review (ril), combining systematicity with interpretative flexibility. the choice of this method is justified by the need to gather, synthesize and compare theoretical and empirical evidence on the adoption of information and communication technologies (ict) in the public sector, allowing for the mapping of both advances and limitations. to ensure breadth and reliability, multiple databases recognized for their relevance were selected: web of science (wos), scopus, sciencedirect, emerald, ieee xplore, google scholar and national journals indexed in capes. these databases were chosen because they cover both articles of high international impact and studies applied to the brazilian context. search procedures bilingual descriptors (portuguese and english) were elaborated that covered the technological and institutional scope of the investigation. some examples: pa ge 93 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 91-96, 2025 (“information and communication technologies” or ict) and (“public services” or “public administration”) (“information and communication technologies” or ict) and (“public service” or “public administration”) (“digital transformation” or “e-government”) and (“efficiency” or “modernity”) the searches were carried out for the period 2020– 2025, in order to ensure timeliness. in addition, articles produced by the authors’ research group, previously published in international journals, which provided direct empirical input (neto et al., 2024; souza et al., 2025; silva et al., 2024, 2025; monteiro et al., 2025). inclusion and exclusion criteria • inclusion: peer-reviewed articles, published between 2020 and 2025, focusing on ict, digital transformation, big data, ai, blockchain, interoperability, or administrative modernization. • exclusion: duplicate papers, book chapters without peer review, expanded conference abstracts, and studies dealing exclusively with the private sector with no analytical connection to public services. selection and analysis • the initial screening resulted in 144 publications. after applying the criteria, 19 articles were selected, 11 from wos and 8 produced by the research group. the analysis was carried out in two stages: • bibliographic analysis: full reading of the selected articles to identify key concepts, theoretical models and empirical evidence. • thematic categorization: organization of findings into four main categories (1) digital training and literacy; (2) security and interoperability; (3) institutional capacity and regulatory framework; (4) ai, big data, and innovation in public services. analytics integration the final stage consisted of the triangulation between international literature, national evidence (automation and telework projects in ifes) and theoretical references of digital transformation. this integration allowed both to compare global standards and to discuss specificities of the brazilian context. (1) results and discussion categorization statement to organize the analysis, it was decided to adopt a thematic categorization that emerges from both international studies and the national cases identified. this categorization allowed us to group the findings into four major analytical dimensions: (1) digital training and literacy; (2) security and interoperability; (3) innovation and cultural change; and (4) digital governance and leadership. this approach aims to articulate different perspectives from technical barriers to institutional and human factors offering an integrated view of the impacts of ict on the modernity and efficiency of public services. category 1 – capacity building and digital literacy dečman et al. (2022) argue that digital efficiency in the public sector is compromised by low digital literacy among civil servants and a lack of institutional preparedness, resulting in superficial technical advances without structural changes. this structural limitation reverberates in the brazilian context, where souza et al. (2025) show that the adoption of telework and sei brought productivity gains only when accompanied by continued training and adjustments in performance evaluation criteria. this finding connects to what susanti et al. (2023) maintain: digital transformation does not depend only on technological adoption, but on the coupling between technology, structure, and strategy, with emphasis on dynamic capabilities and organizational learning as critical factors. in other words, the digital literacy of public servants works as a causal link that transforms investment in technology into real efficiency gains. in the same direction, verhoef et al. (2021) recall that digitalization only generates strategic impact when accompanied by data governance and ecosystem orchestration. without a consolidated digital culture and servers prepared to handle new workflows, the adoption of systems results in punctual modernization, but not in institutional transformation. in addition, akter et al. (2020) reinforce that emerging technologies such as ai, blockchain, and analytics require organizational readiness to generate value. this readiness, in the case of the public sector, is translated into the ability of civil servants to assimilate new tools and apply them to critical management processes. therefore, triangulation reveals that digital training and literacy is not peripheral, but a central condition for the success of ict in public administration. while our national data (souza et al., 2025) confirm this thesis, the international literature (dečman et al., 2022; susanti et al., 2023; verhoef et al., 2021; akter et al., 2020) offers the theoretical basis to understand that digital transformation only occurs when people, processes, and technology evolve in an integrated way. category 2 – security and interoperability galushi & malatji (2022) argue that the advancement of digitalization in public services increases not only administrative efficiency, but also exposure to cyber threats and data protection failures. citizen trust, therefore, depends on robust security policies and mature digital governance. this point is essential because, without protection mechanisms, the risk of losing institutional legitimacy is greater than the benefits of digitalization. correa-ospina et al. (2021) complement by emphasizing that the efficiency of ict in the public sector is conditional on interoperability between systems. initiatives that computerize processes without integrating departments only reproduce bureaucratic fragmentation. for there to be a real impact, digitalization needs to be accompanied by procedural reengineering and organizational culture pa ge 94 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 91-96, 2025 focused on intersectoral cooperation. this reasoning dialogues with silva et al. (2024), who show that emerging technologies such as blockchain and automation increase the reliability of accounting and organizational processes precisely because they ensure transparency and traceability, reducing risks of fraud and informational inconsistencies. although the study deals with the private sector, the same principles apply to the public: without interoperability and security, digitalization results only in superficial modernization. more broadly, akter et al. (2020) highlight that the socalled abcd technologies (ai, blockchain, cloud, and data analytics) only produce transformational gains when there is systemic integration. in the case of governments, this means that legacy systems need to talk to each other, and that the absence of this integration not only limits results, but increases vulnerabilities. our national findings also converge with this diagnosis. neto et al. (2024), when analyzing the payments sector of the federal university of amazonas, demonstrate that efficiency gains with automation were only consolidated because there was integration between launch systems and training of servers to operate the new tools. without this, automation would have increased errors instead of reducing them. thus, it is confirmed that security and interoperability are critical conditions for ict to result in efficiency and legitimacy in the public sector. while international data point to cyber risks and integration failures (galushi & malatji, 2022; correa-ospina et al., 2021), our empirical results (neto et al., 2024; silva et al., 2024) show that well-structured interoperability and security practices can transform one-off digitization into structural gains in reliability and productivity. category 3 – institutional capacity and regulatory framework saukkonen et al. (2024) demonstrate that the modernization of public administration mediated by icts depends directly on regulatory agility. the research shows that excessive bureaucracy and slow adaptation of standards prevent digital innovations from being fully implemented, resulting in limited efficiency gains. this reading reinforces that digital transformation is only consolidated when regulations keep up with the speed of innovation. guo & shen (2024) complement by arguing that visionary leadership is as important as technological investments. they note that public governance needs to engage multiple stakeholders, reduce institutional fragmentation, and ensure interoperability. in other words, technology only generates effective modernization when accompanied by the political and administrative capacity to orchestrate changes. this perspective dialogues with reis et al. (2023), who argue that digital transformation should be understood as a structural process and not merely technical. by analyzing different sectors, the authors identified that sustainable outcomes require cultural, institutional, and regulatory adaptations. this international evidence confirms that the problem of legal slowness observed in são luís, for example, is not an exception, but part of a global pattern of institutional fragility in the face of technological innovation. in the national field, silva et al. (2025) highlight that, in digital auditing and accounting, the absence of standardization and clear regulations hinders the full adoption of disruptive technologies such as blockchain, limiting their application to pilot projects. this diagnosis, although in the private sector, has direct parallels in the public sector, where the lack of adapted standards restricts the expansion of large-scale digital transformation. monteiro et al. (2025) add that collaborative innovation networks between government, the private sector, and academia are crucial to overcome institutional fragility. by shortening the cycle between technological supply and practical absorption, these networks act as catalysts for legal and administrative change. thus, inter-institutional cooperation reduces the risk that regulatory frameworks will lag behind emerging practices. in this way, triangulation shows that institutional capacity and the regulatory framework are basic conditions for digital transformation. while international studies point to bureaucratic barriers, the need for leadership, and legal reforms (saukkonen et al., 2024; guo & shen, 2024; reis et al., 2023), our national and regional analyses (silva et al., 2025; monteiro et al., 2025) reinforce that without clear standards, adaptive governance, and collaborative networks, ict innovation remains fragmented and unable to generate systemic efficiency. category 4 – artificial intelligence, big data and innovation in public services nguyen et al. (2024) analyze experiences of adopting ai and big data in government services and demonstrate that advances in massive data processing allow personalizing services and accelerating the delivery of public policies. however, they emphasize that such innovations are only consolidated when accompanied by continuous training of civil servants and periodic reviews of administrative processes, preventing implementation failures from neutralizing the gains. hien et al. (2024) reinforce this point by arguing that digital transformation is not only technical, but institutional, depending on the alignment between public policies, technological infrastructure, and employee engagement. they note that in contexts where there is cultural resistance and a lack of incentives, even heavy investments in ai and big data yield limited results. in the same vein, zhou et al. (2024) warn that the intensification of the use of digital technologies increases exposure to cyber and privacy risks. for them, the effectiveness of digitalization depends directly on the institutional capacity to protect sensitive data, under penalty of loss of public trust and institutional setbacks. in the brazilian scenario, neto et al. (2024) offer practical pa ge 95 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 91-96, 2025 evidence by showing that automation with excel, vba and rpa in the payments sector of the federal university of amazonas increased productivity even in the face of a reduction in the number of servers. this national experience shows that simple automation tools already produce measurable gains, but require maintenance and continuous training to sustain results. souza et al. (2025) add that the introduction of digital tools in the public service, such as sei and telework, has brought gains in productivity and autonomy for technicaladministrative employees. however, they observed that the full appropriation of ict depends on institutional adjustments in evaluation criteria and new management practices, showing that technological innovation and managerial innovation go hand in hand. thus, although ai, big data, and automation are fundamental drivers of public service modernization, their effectiveness depends on the combination of continuing education, institutional engagement, and data protection. while the international literature highlights security risks, policy alignment, and capacity building requirements (nguyen et al., 2024; hien et al., 2024; zhou et al., 2024), the national cases of our group (neto et al., 2024; souza et al., 2025) illustrate the materiality of these challenges in brazil, showing that gains only become structural when technological innovation is connected to cultural and institutional changes. conclusions the present study aimed to analyze the impact of information and communication technologies (ict) on the modernity and efficiency of public services, identifying both potentialities and structural obstacles. this proposal was based on the recognition that digital transformation, although strategic, faces dilemmas related to training, security, governance, and institutional alignment. the results obtained confirm this objective by evidencing four central findings: (1) the training and digital literacy of civil servants is an indispensable condition to transform technological investments into real efficiency; (2) information security and interoperability of systems define legitimacy and public trust; (3) institutional capacity and regulatory agility shape the possibility of large-scale technological diffusion; and (4) the adoption of ai, big data, and automation generates significant gains only when articulated with cultural, institutional, and data protection changes. in this way, it was found that digital transformation is not sustained autonomously, but depends on human, political, and regulatory factors. it is therefore concluded that ict has the potential to modernize and make public administration more efficient, but only when accompanied by integrated strategies of training, digital governance and adaptive leadership. digital transformation does not eliminate, by itself, the historical dilemmas of public management; on the contrary, it highlights them, requiring multisectoral and collaborative approaches. responding to the objective of this study means recognizing that technology is a means, not an end, to efficiency: without institutional integration, social participation, and adequate regulation, digital runs the risk of producing only superficial modernization, and not true transformation. references akter, s., michael, k., uddin, m. r., mccarthy, g., & rahman, m. (2020). transforming business using digital innovations: the application of ai, blockchain, cloud and data analytics. annals of operations research. barba-sánchez, v., meseguer-martínez, a., gouveiarodrigues, r., & raposo, m. (2024). effects of digital transformation on firm performance: the role of it capabilities and digital orientation. heliyon, 10, e27725. https://doi.org/10.1016/j.heliyon.2024.e27725 campana, c. a., neto, a. c., do carmo, a. p. f., da silva, a. t. p., dos santos, e. m., de oliveira, e. d., ... & da silva, e. n. (2025). a influência da contabilidade ambiental no desempenho sustentável das empresas: uma revisão sistemática da literatura. contribuciones a las ciencias sociales, 18(3), 49. eom, s.-j., & lee, j. (2022). digital government transformation in turbulent times: responses, challenges, and future direction. government information quarterly, 39(2), 101690. https://doi.org/10.1016/j. giq.2022.101690 liu, h., zhang, x., & chen, c. (2025). utilizing the wos and scopus databases to map digital entrepreneurship research. humanities & social sciences communications, 12, 4440. mcafee, a., acemoglu, d., brynjolfsson, e., & syverson, c. (2023). the economics of artificial intelligence: an agenda for productivity growth. journal of economic perspectives, 37(4), 3–34. https://doi.org/10.1257/ jep.37.4.3 monteiro, s. m. c., silva, a. t. p. da, silva, b. b. g. da, oliveira4, e. d. de, nascimento, j. w. do, miranda, n. v. de, carvalho, r. do n., moreira, p. s. s., & da silva, e. n. (2025). interinstitutional collaborations in open innovation and sustainability: an integrative literature review. american journal of development studies, 3(2), 1–5. https://doi.org/10.54536/ajds. v3i2.4623 reis, j., & melão, n. (2023). digital transformation: a meta-review and guidelines for future research. heliyon, 9(1). santos, d. b. (2025). inovação tecnológica na sustentabilidade: uma revisão sistemática sobre o papel das tecnologias verdes na transição para economia circular. contribuciones a las ciencias sociales, 18(1), e14482. silva, e. n., teixeira, a. r., matos, f. r. p., ferreira, h. s., & barros, r. i. m. (2024). innovation in the evolution of accounting: a study of accounting transformation in the digital age. iosr journal of business and management, 26(8), 9–15. susanti, d., prasetyo, p. e., & sugiyanto, e. (2023). digital transformation and business performance: the interplay pa ge 96 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 91-96, 2025 of technology, organization, and strategy a systematic review. annals of operations research. https://doi. org/10.1007/s10479-023-05660-2 verhoef, p. c. et al. (2021). digital transformation: a multidisciplinary reflection and research agenda. journal of business research, 122, 889–901. https://doi. org/10.1016/j.jbusres.2020.06.022 yang, g., yu, h., zhang, x., & yang, f. (2023). a survey on explainable artificial intelligence (xai) in deep learning. engineering applications of artificial intelligence, 114, 106514. https://doi.org/10.1016/j. engappai.2022.105457 pa ge 1 pa ge 68 american journal of smart technology and solutions (ajsts) public perceptions of telegram’s association with illicit activities: a sentiment analysis using vader and machine learning rabel catayoc1* volume 4 issue 1, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i1.4667 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: march 02, 2025 accepted: april 05, 2025 published: may 07, 2025 digital communication platforms have significantly transformed social interaction and information dissemination, yet simultaneously present challenges related to illicit activities, security threats, and regulatory oversight. telegram, a widely-used encrypted messaging service, has recently drawn global scrutiny due to allegations linking it to criminal enterprises, including identity theft, illicit drug markets, and distribution of child exploitation materials. this study systematically evaluates public sentiment surrounding telegram’s reported facilitation of illegal activities, employing comparative sentiment analysis methodologies: vader (valence aware dictionary and sentiment reasoner) and a supervised machine learning approach (tf-idf vectorization coupled with logistic regression). a corpus of 632 reader comments from a wall street journal article discussing telegram’s controversial associations was analysed. vader-based labelling identified an unexpectedly predominant positive sentiment (59.5%), indicating potential public scepticism toward negative media narratives or ideological support for encrypted platforms. the logistic regression classifier demonstrated robust predictive performance, with overall accuracy of 89.56%, precision of 91%, recall of 90%, and an f1-score of 89%, yet displayed a notable positivity bias, misclassifying nuanced negative commentary. qualitative word cloud visualizations further highlighted distinctive lexical patterns, underscoring explicit concerns around security and criminality in negative comments and humour or reflective discourse in positive remarks. methodologically, results expose critical limitations of traditional lexical approaches in capturing subtle, implicit, or context-dependent negativity, suggesting the integration of advanced context-aware modelling techniques, such as transformer-based neural embeddings, for enhanced precision. practically, this analysis provides critical insights for platform governance, risk management strategies, regulatory frameworks, journalistic practices, and computational linguistics research, emphasizing the necessity for balanced methodological approaches to accurately gauge and respond to nuanced public sentiment within contentious digital discourse contexts. keywords entiment reasoner, logistic regression, machine learning, sentiment analysis, vader, tf-idf 1 mindanao state university iligan institute of technology, philippines * corresponding author’s e-mail: rabelcatayoc@gmail.com introduction over the past decade, digital communication platforms have substantially transformed interpersonal communication, information dissemination, and commercial interactions (van dijck et al., 2018; castells, 2013). among the rapidly expanding range of messaging services, telegram—founded in 2013—has emerged as a particularly influential platform, attracting approximately 950 million active monthly users as of july 2024 (team, 2025). telegram’s appeal stems largely from its userfriendly interface, robust end-to-end encryption, and publicly professed commitment to user privacy, rendering it a versatile medium for both personal and professional exchanges (gillespie, 2018; baumgartner et al., 2020). despite its legitimate utility, telegram has increasingly faced scrutiny for allegedly facilitating illicit activities (kohlmann, 2024; sexton, 2024). emerging evidence indicates that the platform has become instrumental in enabling cybercriminal behavior, including identity theft, illicit drug distribution, circulation of child sexual exploitation material, and the orchestration of extremist activities (europol, 2022; the wall street journal [wsj], 2024). telegram’s minimal content moderation policies, coupled with its robust encryption and the resultant anonymity, have been identified as pivotal factors that attract malicious actors to the platform (gillespie, 2018; europol, 2022). the arrest of telegram’s ceo, pavel durov, by french authorities in august 2024 marked a significant turning point, drawing global media attention to the platform’s alleged role in supporting criminal enterprises (wsj, 2024). this high-profile incident intensified discussions concerning digital platform accountability, focusing on the ethical and regulatory responsibilities of technology providers in moderating user-generated content to prevent the proliferation of illegal activities (gorwa, 2019; suzor, 2019). pa ge 69 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 in this evolving context, understanding public sentiment toward telegram is critically important, given its influence on user trust, regulatory responses, and corporate reputation (pfeffer et al., 2014; liu, 2015). sentiment analysis—a computational methodology employing natural language processing techniques—provides an effective tool for systematically assessing public perceptions and societal concerns as expressed in textual form (cambria, das, bandyopadhyay, & feraco, 2017). by analyzing reader-generated comments on news articles addressing telegram’s purported association with criminal activities, scholars can derive nuanced insights into public attitudes and discursive trends, thereby informing business strategy and regulatory policy formulation (stieglitz & dang-xuan, 2013; mostafa, 2013). this study systematically applies sentiment analysis methodologies—specifically, the valence aware dictionary and sentiment reasoner (vader) and a supervised machine learning approach (tf-idf vectorization with logistic regression)—to reader comments from a recent wall street journal article examining telegram’s contentious role in illicit activities. the research aims to elucidate prevailing public sentiment, offering critical implications for telegram’s reputation management, policy strategies, regulatory considerations, and broader sociotechnical discourse. literature review as digital platforms have gained substantial user bases worldwide, there has been growing attention to the way users interact, interpret, and respond to media representations of digital companies and their societal impacts. particularly, understanding public perceptions of criminal activities linked to technology platforms has critical implications for business practices, regulatory scrutiny, and corporate reputation (pfeffer, zorbach, & carley, 2014). sentiment analysis has emerged as a valuable analytical tool for systematically evaluating public perceptions and interpreting their potential impacts on business decisions and policymaking (liu, 2012). this literature review explores prior research employing sentiment analysis methods relevant to online public discourse about technology platforms and illegal activities. it specifically focuses on the business implications for telegram, a widely used messaging and social media app identified recently as a primary marketplace for criminal transactions (wall street journal, 2024). sentiment analysis: methods and applications sentiment analysis involves computational methodologies designed to systematically identify and extract subjective information from textual datasets, evaluating public attitudes toward particular topics, products, or services (pang & lee, 2008; liu, 2012). vader (valence aware dictionary and sentiment reasoner), the sentiment analysis model utilized in this research, has proven robust in evaluating short-form social media texts due to its lexicon-based approach, which assigns positive, negative, neutral, and compound sentiment scores (hutto & gilbert, 2014). scholarly studies utilizing sentiment analysis have underscored its reliability and accuracy in understanding public sentiment toward social media phenomena. for instance, stieglitz and dang-xuan (2013) applied sentiment analysis to twitter data, effectively capturing public emotions during political events. furthermore, studies by mostafa (2013) demonstrated sentiment analysis utility in extracting consumer sentiment on social media toward brands, products, and corporate practices, showing its significance for business implications. classification metrics in sentiment analysis classification models are evaluated using specific metrics designed to capture their performance from various perspectives, particularly when applied to sentiment analysis tasks. precision precision measures the proportion of correctly predicted positive observations out of all observations predicted as positive, indicating the model’s reliability in its positive predictions (sokolova & lapalme, 2009). high precision implies minimal false positive classifications, which is crucial in sentiment analysis to avoid misrepresenting users’ sentiment (liu, 2015). models with high precision effectively minimize type i errors (incorrect positive labels), indicating careful and trustworthy identification of the target sentiment (aggarwal & zhai, 2012). recall recall quantifies the model’s capacity to correctly detect the actual positive instances out of all instances that genuinely belong to that category (han, kamber, & pei, 2011). a model with high recall effectively avoids missing significant instances of the target class, which is critical for sentiment analysis where overlooking important user sentiment could lead to inaccurate conclusions and missed insights (cambria et al., 2013). models optimized for recall are sensitive to the subtleties of sentiment, capturing the majority of relevant sentiment-bearing comments. f1-score the f1-score provides a balanced assessment by combining both precision and recall into a single metric (powers, 2011). particularly useful in sentiment analysis contexts, the f1-score addresses the limitations of relying solely on either precision or recall, offering a nuanced measure of overall model performance (cambria et al., 2017). given that precision and recall may individually vary, the f1-score provides a holistic view crucial for balanced evaluation, especially when class distributions are uneven or when balancing false positives and false negatives is equally important (forman & scholz, 2010). support support refers to the total number of actual occurrences pa ge 70 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 of each class within the dataset, indicating class distribution and serving as context for interpreting performance metrics (pedregosa et al., 2011). understanding class support is essential, especially in real-world sentiment analysis applications, as imbalanced distributions can significantly influence performance metrics and interpretations (weiss & provost, 2003). proper acknowledgment of support helps in evaluating if performance metrics are consistent across different sentiment classes or if disparities exist due to class imbalance (he & garcia, 2009). accuracy and averages (macro and weighted) while accuracy offers an intuitive assessment of model performance—representing the ratio of correct predictions to total predictions—it can be misleading when class distributions are imbalanced (sokolova & lapalme, 2009). thus, macro and weighted averages are additionally recommended as comprehensive measures accounting for class distribution. macro averaging treats each class equally, emphasizing balanced class-level performance, while weighted averaging factors in class support, providing metrics that reflect actual dataset distributions (pedregosa et al., 2011). telegram: a new platform for digital criminal activity telegram has rapidly become popular, with around one billion users, for its simplicity, functionality, and its stance toward user privacy and data confidentiality (wall street journal, 2024). however, its minimal moderation policies and encrypted communications have inadvertently created an environment conducive to illicit activities, including identity theft, child exploitation, weapons smuggling, and drug trafficking (kohlmann, 2024; sexton, 2024). research has shown that platforms combining ease of access, encryption, and light moderation can attract illicit users, negatively affecting corporate reputation, attracting regulatory attention, and causing market-value losses (gillespie, 2018). the wall street journal (2024) specifically highlighted telegram’s transformation into a favored platform among criminal entities, triggering public concerns and potential regulatory consequences. the public exposure of telegram’s unintended uses can significantly affect its corporate image, impacting customer trust and potentially undermining its market positioning (gillespie, 2018; kohlmann, 2024). implications of sentiment analysis for telegram’s business sentiment analysis provides a methodological framework through which businesses like telegram can systematically evaluate the public’s reaction to their portrayal in media and the ensuing narrative around illicit activities. negative sentiment toward companies can harm brand reputation, deter investors, and invite strict regulatory scrutiny (mostafa, 2013; pfeffer et al., 2014). companies increasingly utilize sentiment analysis to monitor their public reputation in real-time, enabling timely interventions to mitigate negative narratives (hutto & gilbert, 2014). given the substantial reputational risks evident in telegram’s recent association with illegal markets and data breaches, systematic sentiment analysis provides valuable insights. such analytical findings can guide telegram’s strategic management and operational decision-making, including moderating policy adjustments, user engagement strategies, and regulatory compliance approaches (gillespie, 2018; kohlmann, 2024). moreover, sentiment analysis also allows identification of key themes in user-generated content, enabling telegram’s management team to better understand public concerns, systematically address these issues, and communicate effectively with stakeholders, thereby enhancing corporate transparency and accountability (stieglitz & dang-xuan, 2013). regulatory and ethical considerations telegram’s positioning amid allegations related to facilitating criminal activity, such as identity theft and child exploitation, presents significant ethical, legal, and regulatory challenges. businesses perceived as tolerating or inadequately addressing such activities can face substantial fines, reputational damage, and consumer attrition (gillespie, 2018). regulatory authorities worldwide increasingly require technology firms to demonstrate proactive measures to counteract illegal activities and harmful content (sexton, 2024). sentiment analysis outcomes highlighting public negativity towards telegram can underscore the urgency for the firm to intensify moderation practices, reporting procedures, and collaboration with external watchdog organizations such as the internet watch foundation (iwf) and the national center for missing and exploited children (ncmec) (wall street journal, 2024; sexton, 2024). ultimately, prior literature clearly underscores the significance of sentiment analysis as a powerful analytical method to gauge public perceptions and their business implications. the case of telegram emphasizes the necessity of deploying robust sentiment analysis tools like vader to analyze public sentiment toward corporate practices, particularly concerning critical ethical, legal, and reputational issues. for telegram, the practical implications derived from sentiment analysis can translate into substantial business actions, notably improving moderation efforts, transparency initiatives, stakeholder communications, and regulatory compliance. ultimately, robust sentiment analyses can assist businesses like telegram in understanding, managing, and mitigating significant reputational risks arising from associations with illegal activities, thus contributing to improved longterm sustainability, ethical practices, and corporate responsibility. pa ge 71 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 materials and methods data collection the dataset used in this study consists of reader comments extracted from a wall street journal article examining telegram’s role as a prominent platform utilized by criminals for illicit activities. the article highlights telegram’s perceived role in facilitating criminal activities, including identity theft, drug trafficking, and exploitation by paedophile rings. reader comments were systematically gathered from the wall street journal’s digital publication platform, forming a corpus suitable for sentiment analysis aimed at uncovering public opinion on the topic. data preprocessing to prepare the text data for analysis, a series of preprocessing steps were applied. each comment was: • tokenized using nltk’s word_tokenize function to split the text into individual word units. • lowercased to ensure consistency in word matching. • filtered by removing standard english stopwords using nltk’s stopword list, which reduced noise and improved the focus on sentiment-bearing terms. • lemmatized using the wordnet lemmatizer, standardizing words to their base or dictionary form (e.g., “running” → “run”). this preprocessing pipeline ensured that the textual input was clean, normalized, and ready for effective vectorization and modelling. stage 1: sentiment labelling (vader lexicon analysis) to generate sentiment labels for supervised learning, each comment was analysed using the valence aware dictionary and sentiment reasoner (vader), a lexicon and rule-based sentiment analysis tool optimized for social media and short text. vader produces a compound score for each comment, ranging from -1 (most negative) to +1 (most positive). labels were assigned as follows: text into numerical feature vectors that reflect both word frequency and uniqueness across the dataset. • logistic regression as the classification algorithm, chosen for its simplicity, robustness, and strong performance in text classification tasks. the model was trained using an 80/20 train-test split, with the tf-idf vectorizer fit on the training data and applied consistently to the test set. evaluation metrics model performance was evaluated using standard classification metrics: table 1: sentiment labelling (vader lexicon analysis) sentiment code score positive sentiment 1 compound score ≥ 0 negative sentiment 0 compound score < 0 these labels were treated as ground truth for training the machine learning model. stage 2: model training (tf-idf + logistic regression) following sentiment labelling, a supervised learning model was trained to classify sentiment directly from the comments. the modelling pipeline involved: • tf-idf vectorization of the comments to convert table 2: evaluation metrics classification description accuracy proportion of total correct predictions. precision proportion of correctly predicted positive comments to all predicted positives. recall proportion of correctly predicted positives to all actual positives. f1 score harmonic mean of precision and recall. a confusion matrix was also generated to visualize true positives, true negatives, false positives, and false negatives, offering a comprehensive view of classification performance. visualization techniques to complement the numerical evaluation, visualizations were created to aid interpretation of results: • bar charts displayed the distribution of predicted vs. actual sentiments. • confusion matrix heatmaps illustrated classification performance. • word clouds were generated for both positive and negative classes to highlight the most frequently used terms in each sentiment group. results and discussion this section presents the results of the following: text preprocessing, sentiment distribution, model evaluation: tf-idf + logistic regression, comparison of actual vs predicted sentiment counts, evaluation of model predictions, and word frequency patterns by sentiment category. text preprocessing prior to model training and sentiment classification, all reader comments were preprocessed to improve the quality and consistency of the input data. this process included: pa ge 72 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 the cleaned and lemmatized text formed the basis for the tf-idf vectorization and subsequent machine learning classification. this step ensured that irrelevant syntactic noise was minimized, and that semantically meaningful patterns were retained for sentiment learning. sentiment distribution to assess public reaction to a wall street journal article investigating telegram’s increasing role as a platform frequently utilized for illicit activities, a sentiment analysis was performed on 632 reader comments using a vaderbased approach. vader (valence aware dictionary and sentiment reasoner) is a lexiconand rule-based sentiment analysis tool optimized for analyzing informal, social media-style text. it was used to classify each comment into either positive sentiment (1) or negative sentiment (0) based on the compound polarity score. table 4 and figure 1 show the resulting distribution was as follows: table 3: text processing process description lowercasing all text was converted to lowercase to ensure uniformity in word representation. for example, words like “telegram,” “telegram,” and “telegram” were normalized to a single lowercase token (“telegram”). this standardization is critical for eliminating case-based redundancy and ensuring that semantically equivalent tokens are treated identically by downstream algorithms. tokenizing each comment was segmented into individual word units using nltk’s word_tokenize function. tokenization enables granular analysis by breaking sentences or phrases into smaller, discrete components (tokens). for instance, the sentence “telegram is not secure.” would be split into the tokens [“telegram”, “is”, “not”, “secure”, “.”], allowing for syntactic and lexical processing on a word-by-word basis. stop word removal common english stop words—such as “the,” “and,” “is,” and “was”—were removed using nltk’s built-in stopword corpus. these words typically carry low semantic weight and are unlikely to contribute meaningfully to sentiment polarity. their removal streamlines the feature space, reduces dimensionality, and allows more important sentiment-bearing terms to dominate the analysis. lemmatization using nltk’s wordnet lemmatizer, each word token was reduced to its canonical or base form (lemma). for example, “running,” “ran,” and “runs” were all normalized to “run.” unlike stemming, lemmatization leverages linguistic context and a vocabulary dictionary to yield grammatically correct base forms. this enhances semantic coherence and improves the model’s ability to generalize across morphological variations of the same word. of child abuse material. this result may initially appear counterintuitive given the negative framing of the article. however, it aligns with prior observations that public discourse around privacy-centric platforms often reveals polarized sentiment, where user loyalty, ideological leanings, or distrust of traditional media override the framing of the original reporting (marwick & lewis, 2017). the high rate of positivity suggests that many commenters either disputed the framing of the article as sensationalist or one-sided, defended telegram as a platform prioritizing privacy and freedom of speech, or expressed distrust in regulatory narratives or mainstream media accounts of digital platforms. these findings have several important implications to telegram, policymakers, business analysts and platform strategists, and journalists. the sentiment trends identified among reader comments suggest a strong base of public support, at least within the sampled population. this sustained positive sentiment provides strategic reinforcement table 4: sentiment distribution of reader comments sentiments code freq percentage positive sentiment 1 376 59.5% negative sentiment 0 256 40.5% this distribution indicates that the majority of reader responses to the article carried a positive or at least nonnegative tone, suggesting either skepticism of the article’s framing or support for telegram as a platform, despite its controversial associations. the majority of reader comments expressed a positive sentiment, despite the article’s explicit emphasis on telegram’s alleged role in facilitating criminal activities such as identity theft, drug sales, and the distribution figure 3: sentiment distribution of reader comments pa ge 73 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 for telegram’s positioning as a neutral and privacyrespecting communication platform. amid growing regulatory scrutiny and negative portrayals in mainstream reporting, this public backing may enable telegram to maintain credibility among its core user base and even leverage public advocacy to counteract reputational risks. moreover, such positive sentiment trends could help telegram justify policy stances on privacy and encryption, emphasizing user rights and autonomy in communications. the public resistance to negative portrayals of telegram captured by the sentiment analysis underscores substantial challenges policymakers may face when aligning policy initiatives and regulatory responses with broader public perceptions. specifically, policymakers must carefully balance their enforcement priorities, especially concerning encrypted platforms, against evident user concerns about privacy and autonomy. this divergence implies a necessity for policymakers to adopt more nuanced, evidence-based approaches to regulation and to engage proactively with the public, clearly communicating regulatory objectives and their rationale. sentiment analyses such as this provide valuable insights, highlighting areas where public understanding or support for regulatory interventions may require further cultivation or clarification. from a strategic and analytical standpoint, the distribution of sentiment emphasizes the significant advantage of incorporating sentiment insights into reputation management strategies, communication planning, and audience segmentation efforts. understanding user attitudes towards telegram offers critical business intelligence, particularly in determining brand positioning, user acquisition, and long-term retention strategies. positive sentiment that endures despite external reputational pressures may suggest robust brand loyalty and resilience, which platform strategists can harness to strengthen user engagement, increase user advocacy, and mitigate reputational crises. such sentiment insights thus provide actionable intelligence for managing reputational risk and ensuring long-term user relationships. for journalists and media professionals, the observed mismatch between article framing and reader sentiment suggests a pressing need for more comprehensive engagement with audience perception metrics and sentiment feedback loops. particularly when reporting on contested or controversial technologies, journalists may benefit from systematically analysing audience sentiment and engagement data to better understand the reception and impact of their reporting. such insights could encourage journalists to adopt more nuanced framing, more effectively anticipating and addressing potential audience skepticism or resistance. ultimately, leveraging sentiment analysis as part of journalistic practice can enhance public trust, reader engagement, and the credibility of media reporting. table 5: sentiment analysis implications stakeholders implications telegram public sentiment can strategically reinforce telegram’s neutral positioning despite reputational challenges and regulatory scrutiny. policymakers resistance in public sentiment highlights the complexities policymakers face aligning technology regulation with user attitudes toward privacy-focused platforms. business analysts and platform strategists positive audience sentiment underscores the importance of integrating sentiment metrics into strategic brand management and user-retention planning. journalists a disconnect between journalistic framing and public response emphasizes the necessity for media practitioners to utilize audience perception insights in technology reporting. this sentiment distribution serves as a foundational result in evaluating how machine learning-based sentiment analysis can surface important sociotechnical dynamics in public discourse surrounding digital platforms. model evaluation: tf-idf + logistic regression to assess the performance of the supervised machine learning model, a logistic regression classifier was trained on tf-idf–transformed features derived from the preprocessed comments. the data was randomly split into training and testing sets using an 80/20 ratio. after training, the model was used to predict sentiment labels for the full dataset. these predictions were then compared with the original vader-based sentiment labels in table 4 to evaluate classification accuracy. the confusion matrix in table 6 presents a breakdown of correct and incorrect predictions. table 6: confusion matrix: tf-idf + logistic regression classification outcomes number of observations true positives (tp) 369 true negatives (tn) 189 false positives (fp) 67 false negatives (fn) 7 figure 2 visualizes the confusion matrix summarizing the model’s sentiment classification performance on 632 reader comments. the results indicate that 369 comments were correctly classified as positive (true positives), and 189 were correctly classified as negative (true negatives). however, 67 comments that were actually negative were misclassified as positive (false positives), while only 7 positive comments were misclassified as negative (false negatives). these distributions demonstrate that the pa ge 74 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 model more frequently errs on the side of positivity. the classification report provides additional insight into model performance across precision, recall, and f1-score: performance metrics further reinforce this observation. the model achieved an overall accuracy of 88%, with precision scores of 96% (0.96) for negative sentiment and 85% (0.85) for positive sentiment. this means that of all comments the model predicted as negative, 96% were truly negative suggesting that the model rarely mislabels positive comments as negative; and of all comments your model predicted as positive, 85% were actually positive showing that model occasionally mislabels negative comments as positive respectively (aggarwal & zhai, 2012). table 7: classification report sentiment precision* recall** f1-score support negative (0) 0.96 0.74 0.84 256 positive (1) 0.85 0.98 0.91 376 accuracy — — 0.88 632 macro avg 0.91 0.86 0.87 632 weighted avg 0.89 0.88 0.88 632 * – out of all comments that my model labeled as positive (or negative), how many were actually correctly labeled? ** – out of all comments that are actually positive (or negative), how many did my model successfully identify? recall scores varied more substantially—while positive sentiment achieved a remarkably high recall of 0.98, negative sentiment lagged behind at 0.74. this suggests that the model successfully captured 98% of all truly positive comments, meaning it missed very few positive comments, and the model successfully identified 74% of all truly negative comments which indicates that it missed some negative comments. the f1-scores, which balance precision and recall, were 0.84 for negative sentiment which reflects good but not perfect balance between precision and recall for negative comments. there’s some room for improvement, mainly due to lower recall; and 0.91 for positive sentiment which reflects a strong overall performance for positive comments, achieving good balance and high accuracy. these findings suggest that while the classifier is generally reliable, it is significantly more confident and consistent when identifying positive sentiment, potentially due to the more explicit or straightforward language used in positive comments. in contrast, negative sentiment may be expressed with greater subtlety or linguistic complexity, leading to higher misclassification rates. implications the model’s asymmetrical performance—high recall for positives and lower recall for negatives—suggests that user-generated negative comments may carry more complex, ambiguous, or implicit linguistic patterns. this aligns with prior findings in affective computing literature, where negativity is often expressed through irony, sarcasm, or culturally specific cues (cambria et al., 2017). in the context of this study, the model’s superior performance on positive sentiment has both methodological and practical implications. from a methodological standpoint, the high recall for positive sentiment suggests that the tf-idf + logistic regression pipeline is effective in capturing the lexical patterns and features commonly associated with supportive or favorable expressions. however, the lower recall for negative sentiment implies that certain critical linguistic cues—such as sarcasm, subtle criticism, or context-dependent negativity—may not be fully captured by surface-level textual features alone. this highlights a limitation of bag-of-words models when applied to sentiment detection, especially in domains where opinions are nuanced or emotionally complex. the observed skew toward positive sentiment classifications could result in the masking of important critical or negative feedback if similar sentiment analysis models are operationally employed for real-time content moderation or in sentiment tracking dashboards. specifically, an overly optimistic bias in automated sentiment classification could lead to underestimating user dissatisfaction, concerns, or complaints, ultimately limiting telegram’s ability to accurately gauge user experience and responsiveness to platform policies. consequently, telegram may miss vital opportunities for improvement or intervention, potentially weakening their overall strategy for addressing user sentiment effectively. from a business and risk management perspective, the model’s observed bias towards positivity indicates a figure 2: confusion matrix: tf-idf + logistic regression pa ge 75 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 significant analytical blind spot. overestimating positive sentiment may lead analysts to underestimate underlying reputational or compliance-related risks, hindering early detection and management of potential public relations issues, regulatory noncompliance, or controversies that negatively impact brand perception. therefore, analysts and strategists must critically account for the model’s positive skew and seek supplementary measures or qualitative insights to counterbalance and enhance the robustness of risk assessment frameworks. for regulatory authorities and stakeholders concerned with platform governance, the model’s potential underrepresentation of negative sentiment carries substantial implications for monitoring and addressing harmful or controversial content. if negativity or critical discourse is systematically underreported or inadequately represented by the sentiment model, regulators may not fully comprehend public concern, discomfort, or backlash toward problematic or disputed platform practices. thus, regulators should advocate for analytical transparency and accuracy in automated sentiment assessment tools, ensuring a reliable basis for policy formulation, enforcement priorities, and broader public accountability mechanisms. these analytical results underline the critical need within computational linguistics and natural language processing research communities for sentiment classification models that demonstrate greater sensitivity and accuracy in capturing negative expressions within public discourse. the apparent positive bias underscores known linguistic challenges, such as subtlety, irony, sarcasm, implicit negativity, or complex sentiment cues, highlighting ongoing areas for improvement in sentiment modelling. computational linguists are therefore encouraged to prioritize developing nuanced, context-aware models that better reflect the multifaceted nature of negative sentiment expression, particularly when analyzing contentious topics or emotionally charged online communication environments. table 8: model evaluation implications stakeholders implications telegram as a platform this skew may obscure critical feedback if similar models are used operationally for content moderation or sentiment dashboards. business analysts and risk managers the model's tendency to overpredict positivity implies a potential blind spot in detecting reputational or compliance-related risks. regulatory stakeholders an underrepresentation of negativity could have implications for monitoring harmful content or evaluating user sentiment toward controversial content. computational linguists these results emphasize the necessity of building sentiment models sensitive to the nuances of negative expression in public discourse. practically, the model’s conservative stance on detecting negativity could affect how public sentiment is interpreted by stakeholders such as platform regulators, journalists, and technology companies. underestimating negative feedback may lead to an overly optimistic assessment of user attitudes toward telegram’s role in illicit activities, thereby distorting public discourse or policy priorities. future iterations of the model may benefit from incorporating more context-aware approaches, such as transformers or neural embeddings, to more accurately capture the semantics of critical or disapproving content. comparison of actual vs predicted sentiment counts to further evaluate the performance of the trained logistic regression classifier, a comparison was made between the original sentiment labels (assigned by the vader analyser) and the predicted sentiment labels generated by the tf-idf + logistic regression model. the comparison between actual and predicted sentiment labels in table 9 and figure 4 demonstrates a clear tendency of the logistic regression classifier to favour positive sentiment predictions. specifically, the model predicted fewer negative comments (192 predicted versus 256 actual) and more positive comments (440 predicted versus 376 actual), indicating an observable imbalance toward positive sentiment classifications. this bias aligns closely with the earlier observation regarding class-specific recall metrics (as previously presented in table 7), where the model achieved high recall for positive sentiment (0.98) but comparatively lower recall for negative sentiment (0.74). such results reinforce the notion that the classifier finds positive linguistic patterns easier or clearer to detect, potentially due to the typically straightforward, explicit lexical characteristics associated with positive sentiment. table 9: actual and predicted sentiment distribution sentiments number of observations actual (632) negative comments 256 positive comments 376 predicted (632) negative comments 192 positive comments 440 conversely, the systematic underprediction of negative sentiment highlights a methodological limitation of employing bag-of-words based tf-idf vectorization. this result suggests that subtlety, irony, sarcasm, and pa ge 76 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 implicit forms of criticism—which are characteristic of negative sentiment—may not be adequately captured by purely lexical and frequency-based textual representations (cambria et al., 2017; liu, 2015). therefore, future improvements may require incorporating more contextsensitive modelling approaches—such as deep learning methods, neural embeddings, or transformer-based architectures—that can better interpret linguistic nuances and emotional complexity present in critical or negative user-generated comments. in practical terms, the observed bias toward positivity could have meaningful implications for stakeholders. if used operationally, such a model might systematically underestimate user dissatisfaction or critique, thereby influencing sentiment dashboards, content moderation systems, or strategic decision-making based on public feedback. thus, from both methodological and applied perspectives, the skewed distribution illustrated in this analysis emphasizes the importance of adopting more nuanced computational approaches to sentiment analysis tasks, particularly within socially or emotionally charged digital discourse contexts. each bar is labeled with the total number of comments per class (0 = negative, 1 = positive). the model predicted more positive comments than were actually labeled, indicating a bias toward classifying sentiment as positive. implications the classifier’s observable bias toward predicting positive sentiment indicates a methodological limitation inherent in tf-idf–based logistic regression approaches. specifically, the frequent misclassification or underrepresentation of negative sentiment suggests that current lexical and frequency-based methods may inadequately capture subtle, context-dependent, or implicit expressions of negative opinions. to address these challenges, future sentiment modelling efforts should incorporate more advanced computational linguistic frameworks, such as contextualized embeddings or transformer-based approaches, that better account for complex linguistic phenomena including sarcasm, irony, or implicit negativity (cambria et al., 2017; liu, 2015). the model’s positive prediction skew has practical ramifications for real-world sentiment monitoring and decision-making scenarios. for platform providers like telegram, relying exclusively on similar predictive models for content moderation, user-experience assessment, or reputation management could lead to an overly optimistic interpretation of user sentiment. this in turn may obscure critical insights into genuine user concerns, dissatisfaction, or risks, potentially hindering timely and appropriate responses to emerging issues or user grievances. the underrepresentation of negative sentiment could inhibit effective detection of critical user feedback, limiting the platform’s ability to accurately gauge and respond to user sentiment trends. consequently, potential issues might escalate unnoticed, compromising overall user satisfaction and retention. an overprediction of positive sentiment poses a risk to accurate assessment and management of reputational or compliance-related concerns. it may lead to misinformed strategic decisions by underestimating user dissatisfaction or overlooking emerging controversies and risks. regulators relying on similar sentiment analytics tools might underestimate the prevalence and intensity of negative user perceptions toward controversial issues, thus impacting the effectiveness of their oversight and policy interventions. these findings underscore the critical need for developing and adopting models that better interpret nuanced linguistic cues associated with negative sentiment. researchers are thus encouraged to advance context-aware and semantically sophisticated analytical techniques that reflect the intricacies of usergenerated negative commentary. figure 3: sentiment distribution: actual vs predicted table 10: actual and predicted sentiment implications stakeholder implications methodological implications tf-idf-based models inadequately capture nuanced negativity, indicating a need for more context-sensitive methods in sentiment analysis. practical implications overly positive predictions may obscure critical user feedback, compromising effective sentiment monitoring and response. telegram (platform management) underestimating negative sentiment could prevent the timely identification of emerging user dissatisfaction or concerns. business analysts & risk managers excessive positivity may result in overlooked reputational and compliance risks, impairing strategic decision-making accuracy. pa ge 77 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 in conclusion, the observed sentiment prediction bias illustrates critical methodological and practical limitations. future efforts in sentiment analysis should aim for a balanced representation of nuanced negative expressions, thereby enabling more accurate, insightful, and responsive analysis of public discourse in digital communication environments. evaluation of model predictions to further assess the performance of the logistic regression model trained using tf-idf features, we compared its predictions to the original sentiment labels produced by the vader analyzer. each comment was categorized into one of four interpretation types. table 11 provides a detailed explanation of classification outcomes generated by the sentiment analysis model (tf-idf vectorization with logistic regression). it categorizes the model’s predictions into four distinct outcomes: true positive (tp), true negative (tn), false positive (fp), and false negative (fn). the true positive category, representing comments correctly predicted as positive, has the highest frequency with 375 observations, indicating strong model performance in identifying positive sentiment accurately. the true negative category, referring to correctly identified negative comments, includes 191 observations, also showing robust performance for negative predictions. conversely, the false positive category—comprising 65 comments incorrectly classified as positive despite being labeled negative—illustrates a tendency of the model toward positivity bias. the false negative outcome was notably low, with only a single comment misclassified as negative despite its true positive label, reflecting minimal risk of overlooking genuinely positive feedback. collectively, these outcomes underscore the model’s overall high accuracy but also highlight its asymmetrical performance, particularly its propensity to err on the side of positive sentiment classification. such detailed classification metrics are valuable for identifying specific areas for methodological improvements and for understanding the practical implications of deploying this model in real-world sentiment monitoring contexts. table 12 presents and visualize in figure 6 the distribution of the logistic regression model’s prediction outcomes in comparison with the actual sentiment labels (generated using the vader analyzer). as illustrated, the largest frequency occurred in the true positive (correctly predicted positive) category with 375 comments, followed by the true negative (correctly predicted negative) category at 191 comments. the false positive category, representing comments incorrectly identified as positive despite their true negative labeling, had a notable presence with 65 occurrences. conversely, the false negative category exhibited minimal representation with only a single occurrence, indicating a very low rate of incorrectly identifying positive comments as negative. examining specific examples of misclassification provides further insight into the limitations of the model. for instance, the comment “company’s fault someone stupid downloaded personal data,” which was genuinely negative (true label: negative), was erroneously classified as positive (false positive). additionally, the extremely brief and contextually ambiguous comment “‘s really” was similarly misclassified as positive, despite its original negative label. regulatory stakeholders underrepresentation of negativity can hinder regulatory insight into public concerns, reducing effectiveness in policy and oversight. computational linguists & nlp researchers findings emphasize the importance of developing advanced models sensitive to subtle linguistic cues associated with negative expressions. table 11: explanation of classification outcomes for model predictions prediction outcome interpretation frequency ✅ true positive (tp) correctly predicted positive comment 375 ✅ true negative (tn) correctly predicted negative comment 191 ❌ false positive (fp) incorrectly predicted positive when the true label was negative 65 ❌ false negative (fn) incorrectly predicted negative when the true label was positive 1 table 12: example of misclassified comments comment true label predicted “company’s fault someone stupid downloaded personal data.” 0 (negative) 1 (false positive) “‘s really.” 0 (negative) 1 (false positive) pa ge 78 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 these misclassifications exemplify common challenges faced by lexically driven sentiment analysis models, particularly in interpreting the nuanced context, implicit negativity, or the brevity frequently encountered in usergenerated comments. such errors underscore the need for more sophisticated, context-aware models capable of accurately capturing subtle or implicit expressions of negative sentiment commonly occurring in digital discourse. these cases highlight challenges in interpreting context or brevity, common in reader-generated content. the evaluation of the logistic regression model trained on tf-idf features, compared against sentiment labels originally produced by the vader analyzer, reveals distinct strengths and limitations. as presented in tables 11 and 12, and visualized in figure 6, the model demonstrated strong performance in correctly classifying positive sentiment (375 true positives) and negative sentiment (191 true negatives), reflecting overall high predictive accuracy. however, the presence of 65 false positives—negative comments incorrectly classified as positive—illustrates a clear positivity bias in the model’s classification approach. notably, the occurrence of false negatives was minimal, with only one instance of misclassification. specific misclassified examples, such as the nuanced negative comment, “company’s fault someone stupid downloaded personal data,” and the brief, ambiguous expression “‘s really,” further illustrate common difficulties encountered by lexically-oriented sentiment analysis methods in accurately interpreting subtlety, brevity, irony, or implicit negativity commonly found in reader-generated comments. these observations collectively highlight the practical necessity and methodological importance of adopting more advanced, context-sensitive analytical frameworks, particularly in sentiment analysis tasks involving nuanced or implicitly expressed opinions. implications the observed classification outcomes and misclassifications carry several critical methodological and practical implications for sentiment analysis applications: the positivity bias reflected by the model’s higher rate of false positives suggests limitations inherent to lexically driven tf-idf approaches, which fail to adequately capture subtle linguistic nuances, contextually embedded negativity, or ambiguous user-generated expressions. consequently, sentiment models should integrate more contextually sophisticated approaches, including neural network-based or transformer models, which are capable of better representing nuanced linguistic features such as irony, sarcasm, implicit negativity, and brevity that characterize authentic digital discourse. for real-world monitoring applications—such as content moderation or sentiment dashboards—this positivity bias may result in an underestimation of critical user feedback, potentially limiting the accuracy of strategic decisions made by platforms, businesses, or regulatory bodies. the model’s conservative approach toward negative sentiment may inadvertently lead stakeholders to overlook emerging concerns or grievances expressed subtly or implicitly by users. telegram’s reliance on similarly biased sentiment classification models could obscure critical or dissatisfied user feedback, thereby negatively impacting the platform’s ability to accurately gauge user perceptions, intervene effectively in user concerns, or strategically respond to emerging user dissatisfaction. such oversight may ultimately compromise user trust, retention, and satisfaction. the positivity bias may represent a blind spot in identifying reputational or compliance-related risks. figure 4: interpretation of model predictions vs. actual sentiment labels pa ge 79 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 overlooking negative sentiment due to misclassification can hinder the timely detection and response to potential issues, controversies, or public relations risks, thus potentially damaging the organization’s reputation or compliance standing. if regulatory authorities utilize such sentiment analysis models to monitor online discourse, underrepresentation of negativity can significantly distort the accuracy of regulatory assessments. misclassification may cause regulators to underestimate public dissatisfaction or concern, undermining effective policymaking, intervention, and oversight. the presented data reinforces the need to prioritize development of sentiment analysis models that robustly address linguistic complexity in negative expressions. computational linguists are thus encouraged to refine methodologies—employing advanced neural models, semantic embeddings, and transformer architectures—to overcome existing limitations and better capture the full spectrum of linguistic expression in sentiment analysis tasks. table 13: evaluation of model predictions implications stakeholder implications methodological the positivity bias underscores the limitations of lexically driven tf-idf models, necessitating context-sensitive sentiment approaches. practical a model's tendency toward positivity may result in the systematic underestimation of critical or negative user feedback. telegram (platform management) underestimating negative sentiment could hinder telegram’s ability to detect and address emerging user concerns effectively. business analysts and risk managers positivity bias in sentiment modeling creates analytical blind spots, potentially masking reputational or compliance-related risks. regulatory stakeholders underrepresentation of negative sentiment can compromise regulatory oversight and the accuracy of policy interventions. computational linguists and nlp researchers misclassifications highlight the necessity for developing sentiment models capable of capturing linguistic subtleties and complexities inherent in negative expressions. overall, these implications emphasize the necessity of employing sophisticated analytical techniques, complementary qualitative insights, and contextual linguistic awareness in sentiment classification to improve practical accuracy and methodological reliability, particularly in sentiment analysis scenarios involving usergenerated digital communication. word frequency patterns by sentiment category to further explore the thematic content of reader sentiment, two word clouds were generated to visualize the most frequently occurring terms in positive and negative comments. this qualitative visualization complements the quantitative performance metrics by highlighting the types of language that characterize different sentiment classes. the word cloud for positive comments (figure x) prominently features terms such as “funny,” “agree,” “think,” and “good,” suggesting that readers who viewed the article more favorably often expressed humor, agreement, or general reflection. these comments may reflect support for regulatory action or acknowledgment of the app’s broader appeal despite criminal misuse. these visualizations provide additional context to the sentiment classification model by revealing the lexical patterns that differentiate user attitudes in response to the article. figure 7: commonly used words for positive comments pa ge 80 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 in contrast, the word cloud for negative comments (figure y) is dominated by terms like “criminals,” “data,” “identity,” “illegal,” and “scared,” indicating a strong concern around security breaches, digital exploitation, and telegram’s alleged facilitation of illicit activity. the presence of emotionally charged words in this category underscores a clear disapproval and unease among users. the qualitative analysis depicted in figures 7 and 8 complements quantitative metrics by revealing distinctive lexical patterns in sentiment-classified comments. specifically, the positive comments word cloud (figure 7) prominently highlights terms such as “funny,” “agree,” “think,” and “good,” signifying reader engagement characterized by humor, agreement, reflective discourse, and general positivity toward the platform. this suggests readers positively disposed toward telegram may express broader support or appreciation, possibly perceiving benefits of regulatory attention or viewing the platform’s positive attributes independently of reported illicit activities. in contrast, the negative comments word cloud (figure 8) emphasizes terms such as “criminals,” “data,” “identity,” “illegal,” and “scared,” explicitly reflecting heightened concerns regarding data security, digital threats, and telegram’s potential role in enabling criminal activity. emotionally charged language in negative sentiment comments highlights clear user apprehension and explicit disapproval about the platform’s associated risks. implications the clear distinction between positive and negative lexical patterns indicates the effectiveness of qualitative text visualization techniques, such as word clouds, in supplementing quantitative modeling efforts, particularly by contextualizing underlying emotional dimensions that purely numerical methods might overlook. lexical patterns from negative sentiment illustrate pressing concerns about privacy, security, and illicit activity, highlighting critical areas where telegram must proactively address user apprehensions to effectively manage public perceptions and mitigate reputational risk. explicitly negative terms (“identity,” “illegal,” “scared”) suggest heightened reputational or compliance risks that business analysts should systematically monitor, reflecting the need for rigorous risk management strategies that address user security concerns directly. frequent mentions of crime-related terminology among negatively classified comments underscore urgent public demand for effective regulatory intervention to ensure platform accountability, protect users from exploitation, and uphold digital safety standards. the stark lexical contrast between positive and negative comments reaffirms the necessity for advanced computational methodologies sensitive to distinctively emotional, domain-specific vocabulary, thus enhancing model precision in capturing nuanced sentiment expressions. in sum, qualitative visualization of lexical patterns significantly enriches quantitative sentiment analyses, providing valuable insights for methodological refinement, strategic management, regulatory oversight, and computational linguistics advancement. model performance metrics to evaluate the accuracy of the machine learning classifier (tf-idf vectorization + logistic regression), standard performance metrics were computed using the predicted sentiment labels (analysis) and the ground-truth labels (sentiment) generated by the vader-based approach. the model achieved a high degree of overall accuracy, correctly predicting the sentiment classification for approximately 89.56% of all reader comments. the precision score was 0.9100, indicating that the majority of comments predicted as positive or negative were labeled correctly. the recall score was 0.8956, suggesting that the model was effective at capturing the majority of true sentiment classes. finally, the f1-score, a harmonic mean of precision and recall, was 0.8922, confirming the model’s strong and balanced performance across figure 8: commonly used words for negative comments pa ge 81 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 sentiment categories. these results reinforce the model’s suitability for sentiment classification tasks in real-world online discussions, particularly in socially sensitive contexts like telegram’s association with criminal activities. table 14 and figure 9 summarize the key performance metrics for the logistic regression classifier using tf-idf features to predict sentiment labels derived from vadergenerated ground-truth labels. the model demonstrated a table 14: model performance metrics table metric score accuracy 0.8956 (90%) precision 0.9100 (91%) recall 0.8956 (90%) f1 score 0.8922 (89%) high overall accuracy of approximately 89.56%, signifying its strong capability to correctly classify reader comments according to their sentiment. a precision score of 0.9100 (91%) indicates a high level of confidence that comments classified as either positive or negative by the model were indeed labeled correctly. the recall score of 0.8956 (90%) underscores the model’s effectiveness in identifying the majority of the sentiment classes accurately within the dataset. furthermore, the f1-score of 0.8922 (89%), as a balanced metric combining precision and recall, confirms robust and well-balanced overall performance. collectively, these metrics strongly suggest the classifier’s appropriateness and reliability for practical sentiment classification applications, particularly in sensitive digital contexts—such as discussions involving telegram and its alleged association with criminal activities—where accurate and nuanced interpretation of user-generated discourse is critical. implications of model performance metrics the classifier’s strong overall performance (accuracy: 89.56%, precision: 91%, recall: 90%, and f1-score: 89%) confirms the effectiveness of tf-idf vectorization combined with logistic regression for accurately capturing lexical patterns indicative of sentiment. however, despite strong metrics, there remains a methodological implication that certain nuanced linguistic features—such as implicit negativity or context-dependent subtleties— may still not be fully captured, necessitating continued refinement and potential integration of advanced, context-aware modeling approaches. the high reliability indicated by these metrics suggests the model’s suitability for practical deployment in sentiment monitoring tools or dashboards, especially in contexts involving socially or emotionally sensitive topics. nevertheless, stakeholders must remain aware of potential limitations—particularly related to subtle, implicit, or nuanced negative sentiment—that might be systematically underrepresented despite overall strong performance. the robustness of these metrics indicates that sentiment classification models could significantly aid telegram’s management in real-time monitoring of user sentiment, allowing for proactive and informed interventions. nevertheless, telegram should integrate qualitative or supplementary analyses to ensure comprehensive coverage of user concerns, especially those communicated through subtle linguistic expressions that may evade quantitative detection. high precision and recall rates offer business analysts a reliable tool for assessing user sentiment accurately, thus improving strategic and operational decision-making processes. however, analysts should remain cautious, complementing such models with qualitative analysis or expert judgment to capture nuanced expressions of potential risks and reputational threats. for regulators, the demonstrated model reliability supports informed monitoring and policy-making concerning online discourse and platform governance. nonetheless, regulators should acknowledge potential blind spots in automated sentiment classification, particularly regarding subtle negative commentary that figure 7: mode performance metrics pa ge 82 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 table 14: model performance metrics table stakeholder implications methodological high model accuracy validates tf-idf and logistic regression but highlights the necessity for advanced approaches to capture nuanced negative expressions. practical robust overall performance supports operational deployment for sentiment analysis, though stakeholders should remain cautious of subtle sentiment complexities. telegram (platform management) reliable sentiment analysis metrics enable effective user monitoring; however, supplementary qualitative analyses are needed to detect subtle concerns. business analysts and risk managers strong precision and recall metrics enhance informed decision-making, yet complementary analysis remains critical for nuanced risk detection. regulatory stakeholders high classifier reliability aids regulatory oversight but necessitates awareness of potential blind spots in identifying subtly negative user feedback. computational linguists and nlp researchers while lexical models exhibit strong predictive accuracy, their inherent linguistic limitations call for research into context-sensitive computational methodologies. could inform nuanced policy decisions. the strong performance metrics highlight the value of lexical models but also underscore their inherent limitations. researchers are encouraged to further explore integrating advanced linguistic methodologies (such as neural embeddings or transformer models) to address limitations related to capturing subtlety, irony, or contextually complex negative expressions in sentiment analysis. overall, while this model demonstrates substantial accuracy and reliability for real-world sentiment analysis, stakeholders must remain cognizant of inherent limitations and pursue complementary analytical techniques and approaches for comprehensive interpretation of public sentiment. discussion text preprocessing the text preprocessing methods implemented in this study were critical in ensuring accurate sentiment classification and robust model performance. the preprocessing pipeline, which included lowercasing, tokenization, stop word removal, and lemmatization, played a significant role in enhancing data consistency and semantic interpretability. converting text to lowercase effectively mitigated variability caused by case sensitivity, standardizing semantically identical words into unified tokens. this step was essential for avoiding redundancy and ensuring that terms such as “telegram,” “telegram,” and “telegram” did not dilute or distort lexical patterns recognized by downstream sentiment classifiers. tokenization further contributed to the model’s granularity and interpretative accuracy by segmenting user-generated content into discrete lexical units. by enabling wordlevel analyses, tokenization facilitated precise sentiment attribution and more accurate feature extraction, allowing models to distinctly recognize sentiment cues from individual tokens. the removal of common english stop words similarly played a critical methodological role by reducing textual noise and dimensionality. excluding lowvalue linguistic elements—such as articles, prepositions, or conjunctions—allowed sentiment-bearing words to be more prominently weighted in the subsequent tf-idf vectorization stage. this approach aligns well with established sentiment analysis literature, which emphasizes the advantage of minimizing unnecessary linguistic content that could introduce ambiguity or dilute the overall feature relevance (aggarwal & zhai, 2012; liu, 2015). finally, the lemmatization procedure notably strengthened the semantic coherence and interpretability of the text data. by systematically converting morphological variants into standardized base forms, lemmatization substantially enhanced the model’s ability to generalize beyond surface-level lexical variation. this facilitated the accurate aggregation of related terms—such as “running,” “ran,” and “runs”—thereby enhancing feature representation consistency and boosting overall predictive reliability (jurafsky & martin, 2009). such linguistic normalization is particularly critical in nuanced sentiment analysis scenarios, like the telegram-related discussions explored here, where accurate interpretation of morphological variations can meaningfully influence sentiment polarity outcomes. overall, the text preprocessing methods adopted in this study significantly contributed to the classifier’s robust performance (accuracy ~89.56%). future research could explore the potential benefits of integrating more advanced preprocessing steps, such as context-aware embeddings or deep linguistic models, to better address nuanced linguistic structures, implicit sentiment, and subtleties inherent in user-generated digital communication. this expanded methodological repertoire might further enhance sentiment classification accuracy, particularly within socially sensitive or linguistically complex online discourse contexts. sentiment distribution the sentiment distribution derived from reader comments, as analyzed by the vader-based sentiment classification, provides valuable insights into public pa ge 83 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 perceptions surrounding telegram, particularly in relation to its portrayal in mainstream media. interestingly, despite the negative framing of telegram’s association with criminal activities such as identity theft, drug trafficking, and exploitation, the sentiment analysis revealed a predominant positivity among reader comments, with approximately 59.5% classified as positive versus 40.5% negative. this notable prevalence of positive sentiment may initially seem counterintuitive given the explicitly critical portrayal of telegram in the article. however, this finding resonates closely with existing scholarship emphasizing polarized public responses toward digital platforms associated with privacy, encryption, and user autonomy. such platforms frequently engender divided public opinion, often shaped significantly by users’ ideological leanings, platform loyalty, or inherent skepticism towards traditional media narratives (marwick & lewis, 2017). the evident divergence between media framing and audience sentiment underscores a substantial sociotechnical dynamic that stakeholders—such as digital platforms, policymakers, business strategists, and media professionals—must carefully navigate. for telegram, strong positive sentiment indicates a solid foundation of user support, potentially insulating the platform against reputational damage and regulatory scrutiny, thereby reinforcing its public position as a privacy-respecting communication medium. policymakers, however, face pronounced challenges, as public resistance to negative framing and subsequent policy interventions illustrates the complex relationship between regulatory enforcement, public perception, and platform autonomy. consequently, policymakers should approach regulation with greater sensitivity to user sentiment, perhaps engaging in proactive public dialogue to communicate clearly the rationale and objectives underlying platform governance decisions. for business analysts and platform strategists, the sustained positive user sentiment signals a strategic opportunity to deepen user engagement, brand loyalty, and resilience to reputational crises. incorporating comprehensive sentiment analysis into reputation management frameworks thus provides actionable insights, enabling proactive strategic responses aligned with genuine user attitudes. meanwhile, the evident disconnect between journalistic framing and reader sentiment highlights a critical area for professional reflection among journalists and media organizations. specifically, sentiment analysis offers a powerful feedback mechanism to gauge the impact of journalistic narratives on public opinion. integrating sentiment analytics into journalistic practice can improve audience engagement and enhance credibility by encouraging more nuanced, responsive reporting practices, particularly on controversial technological issues. overall, this sentiment distribution analysis illustrates how sentiment modelling techniques—such as the lexiconbased vader analyzer—can significantly contribute to understanding and interpreting complex public discourse around digital platforms. such analyses not only inform methodological refinement in computational sentiment studies but also carry direct implications for the practical management of public perception, policy development, strategic communication, and journalistic integrity in the digital age. model evaluation: tf-idf + logistic regression the evaluation of the logistic regression classifier using tf-idf-transformed features revealed notable insights into the strengths and limitations of lexical sentiment analysis approaches in classifying public sentiment. the confusion matrix (table 6 and figure 2) demonstrates that the model achieved substantial predictive accuracy, correctly classifying 369 comments as positive (true positives) and 189 as negative (true negatives). however, it also exhibited an asymmetrical performance with a noteworthy positivity bias—misclassifying 67 negative comments as positive (false positives), while only misclassifying 7 positive comments as negative (false negatives). this asymmetry, while affirming the model’s overall reliability, underscores a pronounced methodological limitation wherein nuanced expressions of negative sentiment are frequently missed or incorrectly interpreted. the classification report (table 9) further reinforces these findings through precision, recall, and f1-score metrics. the model exhibited high precision (96%) in negative sentiment classification, suggesting that when it did predict negativity, it was highly reliable. nevertheless, the significantly lower recall for negative sentiment (74%) compared to positive sentiment recall (98%) signals a clear methodological challenge in capturing subtler forms of negativity. such challenges align closely with established literature in sentiment analysis, which frequently identifies implicit negativity, irony, sarcasm, and contextually embedded critiques as complex linguistic constructs often inadequately captured by surface-level lexical analysis (cambria et al., 2017). from a methodological standpoint, these results highlight limitations inherent in bag-of-words approaches such as tf-idf, which rely primarily on explicit lexical cues rather than sophisticated semantic interpretation. thus, the model’s robust identification of positive sentiment likely reflects the clearer lexical and syntactic patterns commonly associated with explicit agreement or approval. in contrast, negative sentiment frequently involves linguistic subtlety and complexity, explaining the observed positivity bias. future research directions could address this limitation by integrating context-sensitive approaches, such as transformer-based language models or neural embeddings, which can better interpret subtle linguistic signals and implicit semantic content. practically, the observed positivity skew has substantial implications for stakeholders. for telegram, deploying similar sentiment models operationally—such as in content moderation or user-sentiment dashboards—may pa ge 84 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 inadvertently underestimate user dissatisfaction, obscuring critical feedback and inhibiting effective responsiveness. for business analysts and risk managers, positivity bias represents a potential analytical blind spot, potentially compromising the timely detection of reputational and compliance risks. regulatory stakeholders should similarly remain cautious, recognizing that a sentiment model biased toward positivity could diminish the visibility of user concerns or critical perspectives essential for informed policymaking and oversight. finally, computational linguists and nlp researchers are encouraged to address these biases and improve sentiment models by developing nuanced methods capable of recognizing linguistically complex negative expressions, ultimately ensuring more balanced and accurate sentiment assessments in public discourse. in conclusion, this model evaluation indicates strong overall reliability for lexical sentiment analysis using tfidf vectorization and logistic regression. however, the asymmetrical performance pattern emphasizes the need for methodological enhancements to better capture nuanced negativity. future sentiment analysis models should therefore adopt advanced computational methods, ensuring comprehensive and contextually accurate interpretation of public sentiment, particularly within emotionally charged or contentious digital communication contexts. comparison of actual vs predicted sentiment counts the comparison between actual and predicted sentiment counts from the logistic regression classifier trained on tf-idf-transformed features provides critical insights into both the methodological strengths and limitations of the approach. the classifier exhibited a notable bias toward positive sentiment, systematically predicting a higher number of positive comments (440) compared to the actual positive count (376), while correspondingly underestimating negative sentiment (192 predicted versus 256 actual). this positivity skew aligns closely with earlier classification metrics, particularly the recall scores that highlighted significantly higher recall for positive sentiment (0.98) compared to negative sentiment (0.74). these observations suggest that positive sentiment expressions typically feature more explicit, straightforward linguistic patterns, which are easier for lexical-based tf-idf models to detect consistently. conversely, the underrepresentation of negative sentiment indicates a substantial methodological shortcoming in capturing nuanced linguistic cues, subtle criticism, irony, and implicit forms of negativity. from a methodological perspective, the identified bias underscores significant limitations inherent in traditional bag-of-words approaches like tf-idf, which rely predominantly on surface-level lexical features and frequency counts. while effective in capturing explicit sentiment expressions, these approaches often fail to adequately interpret the context-dependent subtleties and complexity inherent in negative user-generated comments (cambria et al., 2017; liu, 2015). therefore, enhancing model accuracy—particularly for negative sentiment— would require integration of advanced contextualization methods such as neural embeddings, transformer-based architectures, or deep learning techniques that offer deeper semantic understanding of linguistic nuances. practically, the identified positivity bias bears substantial implications for operational sentiment monitoring in real-world digital platforms. for telegram, the systematic underestimation of negative sentiment could obscure critical user feedback, grievances, or dissatisfaction, potentially compromising timely intervention, user satisfaction, and trust. similarly, business analysts and risk managers relying on such sentiment analysis tools may underestimate reputational and compliance risks, misinforming strategic decisions and hindering proactive risk mitigation. regulatory stakeholders may also be adversely affected, as this positivity bias could diminish the accuracy and effectiveness of regulatory assessments, particularly around sensitive or contentious platform practices. consequently, these stakeholders must approach sentiment analysis results cautiously, supplementing lexical models with qualitative or additional analytical methods to achieve comprehensive sentiment understanding. furthermore, computational linguists and natural language processing researchers should interpret these findings as a clear mandate for continued innovation. this bias highlights the necessity for more nuanced, contextually sensitive modeling approaches capable of accurately interpreting the implicit, subtle, and often culturally specific cues of negative sentiment expressions. advancing analytical techniques beyond traditional lexical frameworks—towards sophisticated, contextually aware computational linguistics—will enable more accurate, balanced, and insightful sentiment analysis outcomes, particularly within emotionally charged digital discourse contexts. in conclusion, while the logistic regression classifier demonstrates overall robust predictive reliability, its positive sentiment skew identifies a significant methodological and practical limitation. addressing this limitation through the adoption of advanced semantic modeling approaches promises more balanced, accurate, and practically valuable sentiment analyses for digital communication platforms and stakeholders. evaluation of model predictions the detailed evaluation of the logistic regression model utilizing tf-idf features, compared against original sentiment labels generated by the vader analyzer, reveals key methodological strengths and critical limitations. the analysis categorizes model predictions into four distinct outcomes: true positives (tp), true negatives (tn), false positives (fp), and false negatives (fn). notably, the model demonstrated strong overall predictive accuracy, correctly identifying positive sentiment in 375 instances and negative sentiment in 191 instances. this robustness in correctly classifying sentiment underscores pa ge 85 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 the general effectiveness of tf-idf–based logistic regression approaches for explicit sentiment recognition. however, the presence of 65 false positive predictions— negative comments incorrectly classified as positive— highlights a significant positivity bias within the model’s classification approach. conversely, the occurrence of false negatives was minimal, with only one such instance observed, reflecting minimal risk in overlooking positive user-generated feedback. an examination of specific misclassified examples further illuminates inherent methodological challenges. comments like “company’s fault someone stupid downloaded personal data,” a negatively intended critique incorrectly predicted as positive, exemplify the difficulty lexically oriented sentiment models encounter in accurately interpreting implicit negativity, subtle criticism, or nuanced linguistic context. similarly, brief and ambiguous comments like “‘s really” further emphasize these limitations, where brevity and lack of explicit lexical cues impede accurate classification. these instances underscore a critical shortcoming in lexically driven models such as tf-idf, demonstrating their inadequate handling of subtle, contextually embedded negativity, ambiguity, and implicit expressions prevalent in authentic digital communication. the positivity bias observed has substantial methodological implications, calling attention to the inherent constraints of frequency-based lexical approaches when interpreting user-generated content. such limitations advocate strongly for the integration of more advanced computational methods—particularly deep learning models, context-aware neural embeddings, or transformer-based architectures—that possess greater semantic understanding and can better account for linguistic subtleties, irony, brevity, and implicit expressions of negative sentiment. practically, the implications of the model’s positivity bias are significant for real-world sentiment monitoring. telegram, as a platform provider, risks systematically underestimating genuine user dissatisfaction or critical feedback if similarly biased models are operationalized. this could compromise accurate detection and timely management of user concerns, potentially affecting user satisfaction and retention adversely. business analysts and risk managers could likewise misinterpret sentiment data due to this positivity skew, possibly overlooking subtle but critical reputational or compliance-related risks. regulatory stakeholders relying on such models might similarly underestimate negative sentiment, reducing their effectiveness in policy oversight and intervention. for computational linguists and natural language processing researchers, these findings reinforce the necessity of developing models capable of nuanced semantic interpretation, motivating research into sophisticated, contextually sensitive linguistic methodologies. overall, these insights emphasize a clear need to complement lexically driven sentiment analyses with advanced semantic modeling techniques and qualitative interpretation strategies. ensuring comprehensive and balanced detection of nuanced negative sentiment expressions will significantly enhance methodological robustness, practical utility, and stakeholder trust in computational sentiment analysis tools—particularly within sensitive, nuanced digital communication contexts. word frequency patterns by sentiment category the qualitative exploration of lexical patterns through word clouds (figures 5 and 6) offers meaningful contextual insight, complementing the quantitative sentiment classification results. the visualization distinctly highlights differences in language usage between positive and negative comments, providing deeper understanding into user sentiment. specifically, the positive sentiment word cloud (figure 5) prominently features words such as “funny,” “agree,” “think,” and “good,” indicating that readers positively disposed toward telegram frequently employed language associated with humor, agreement, reflective discourse, and explicit positivity. these lexical choices likely signify broader support, either for telegram’s foundational principles (privacy and autonomy) or for regulatory approaches towards the platform, notwithstanding its controversial associations. such qualitative insights reinforce previous quantitative findings of prevalent positivity, underscoring user loyalty or skepticism toward media portrayals of telegram’s negative associations. conversely, the word cloud representing negative comments (figure 6) emphasizes emotionally charged and security-focused terms, such as “criminals,” “data,” “identity,” “illegal,” and “scared.” the prominence of these words underscores explicit user concerns about telegram’s potential role in facilitating illicit activities, including data breaches, identity theft, and digital exploitation. these lexically explicit negative expressions indicate clear apprehension and disapproval, highlighting a substantial segment of public sentiment focused on security, privacy violations, and digital threats associated with the platform. methodologically, these visualizations highlight the efficacy of qualitative text analysis techniques such as word clouds, which significantly enrich purely quantitative analytical approaches. by revealing the nuanced, underlying emotional and contextual dimensions of sentiment, these methods provide essential supplementary context to quantitative models, thus enhancing the overall interpretative accuracy of sentiment analyses. practically, the stark distinction in lexical patterns emphasizes critical areas for platform management—such as telegram—to proactively address user concerns, especially in security, privacy, and trust. failing to address these highlighted negative concerns could compromise user satisfaction, reputation, and long-term platform sustainability. from a risk management perspective, explicitly negative terminology (“illegal,” “identity,” “scared”) suggests significant reputational or compliance risks, which analysts and strategists must rigorously monitor. such pa ge 86 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 qualitative insights reinforce the necessity of integrated, contextually nuanced risk assessment frameworks, enabling more precise management responses to public concerns or controversies. for regulatory stakeholders, the frequency of crime-associated terms in negative user comments underscores urgent demands for accountability and effective oversight. thus, policymakers are advised to closely align regulatory strategies with these publicly articulated user apprehensions, addressing platform accountability and protecting user rights effectively. for computational linguists and natural language processing (nlp) researchers, the pronounced lexical distinction between positive and negative comments further emphasizes the methodological importance of advanced sentiment analysis techniques. models sensitive to emotional vocabulary, domain-specific language, and subtle linguistic nuance will enhance precision and interpretative accuracy. researchers should thus prioritize developing sophisticated computational methodologies, particularly contextually aware semantic models, to accurately capture the complexities of nuanced sentiment expressions. in conclusion, qualitative lexical analyses, as illustrated by word clouds, significantly enhance quantitative sentiment analyses by illuminating underlying emotional and thematic user sentiments. integrating qualitative techniques with advanced quantitative methods provides critical insights for methodological refinement, strategic reputation management, informed regulatory oversight, and ongoing computational linguistic research— particularly within contentious or socially sensitive digital discourse contexts. model performance metrics the performance metrics summarized in table 14 and figure 7 indicate robust effectiveness of the logistic regression classifier using tf-idf vectorization for sentiment analysis tasks. the model achieved an overall accuracy of approximately 89.56%, precision of 91%, recall of 90%, and an f1-score of 89%, collectively demonstrating its reliable predictive capability for accurately categorizing sentiment in reader-generated comments. high precision indicates substantial confidence in the model’s sentiment predictions, reflecting accuracy in distinguishing between positive and negative commentary. similarly, strong recall underscores its effectiveness in correctly identifying most sentiment classifications present within the dataset. the balanced f1-score further confirms the model’s capability to effectively integrate both precision and recall into consistently reliable classification performance. despite these strong metrics, methodological limitations inherent in tf-idf vectorization and logistic regression warrant critical examination. the comparatively lower recall for negative sentiment (74%, as detailed previously) implies that nuanced linguistic features—such as subtle criticism, implicit negativity, sarcasm, or contextually complex expressions—remain inadequately captured by purely lexical approaches. consequently, the high accuracy achieved may still systematically underrepresent certain nuanced sentiment types, specifically negative or implicit expressions. thus, future methodological refinement should focus on incorporating advanced modeling approaches, such as context-sensitive neural embeddings, deep learning models, or transformer-based architectures, capable of comprehensively interpreting linguistic subtleties and implicit sentiment cues. from a practical standpoint, the demonstrated high reliability of this classifier supports its applicability for operational deployment in sentiment monitoring contexts, particularly those involving sensitive issues such as digital privacy or criminal associations. for telegram’s platform management, such reliable metrics indicate valuable potential for real-time sentiment monitoring, informing proactive user-experience interventions and platform responsiveness. however, recognizing inherent limitations in capturing subtle negativity, telegram should employ supplementary qualitative analyses or human oversight mechanisms to ensure a comprehensive understanding of user sentiment nuances. for business analysts and risk management professionals, these robust precision and recall metrics significantly enhance decision-making accuracy and strategic confidence, enabling more precise monitoring of user sentiment trends. nevertheless, analysts must remain vigilant to inherent methodological limitations, complementing automated sentiment analyses with qualitative evaluations or expert judgment to avoid overlooking subtle reputational or compliance-related risks. regulatory stakeholders similarly benefit from the model’s demonstrated reliability, enabling informed oversight and nuanced policy formulation in response to public sentiment. however, regulators should be cautious of potential blind spots related to subtle negative sentiment expression, adopting supplementary analysis methods to ensure comprehensive understanding of public discourse. for computational linguists and nlp researchers, these performance results highlight the efficacy of lexical models such as tf-idf logistic regression, but simultaneously underscore their limitations. addressing these methodological shortcomings necessitates advancing context-sensitive computational methodologies that better interpret subtle linguistic complexities, irony, implicit negativity, and nuanced discourse features prevalent in authentic digital communication. overall, while the logistic regression classifier demonstrates substantial practical reliability and predictive accuracy, stakeholders must remain mindful of inherent methodological limitations. integrating complementary analytical techniques and developing advanced semantic modeling approaches remain critical avenues for improving comprehensive and nuanced sentiment analysis. such advancements promise significant benefits for methodological rigor, operational decision-making, regulatory effectiveness, and computational linguistics pa ge 87 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 innovation, particularly in sensitive or controversial digital communication contexts. conclusions this study systematically evaluated public sentiment regarding telegram’s reported association with criminal activities by applying sentiment analysis methods— specifically vader and a tf-idf vectorized logistic regression classifier—to reader comments sourced from a wall street journal article. results revealed notable patterns, both quantitative and qualitative, which carry critical methodological, theoretical, and practical implications. quantitatively, vader sentiment analysis classified the majority (59.5%) of reader comments as positive, indicating significant public skepticism towards the negative media framing of telegram or a broader ideological support for privacy-centric platforms. the logistic regression classifier demonstrated robust performance (89.56% accuracy, precision of 91%, recall of 90%, and f1-score of 89%), affirming the utility of lexically-based sentiment classification models for effectively analyzing public discourse. however, it exhibited a clear bias toward positivity, reflected in lower recall for negative sentiment (74%) and a significant presence of false-positive classifications. these results highlight the critical methodological limitation of current lexical-based approaches in accurately capturing nuanced negative sentiment expressions, especially implicit negativity, subtle criticisms, sarcasm, and brevity. consequently, advanced context-aware methodologies, such as transformer-based neural embeddings, are recommended for future research to enhance accuracy, particularly in detecting nuanced or implicitly negative sentiments. qualitative analysis, specifically word cloud visualizations, distinctly captured lexical patterns characterizing both positive and negative sentiment classes. positive comments frequently included terms reflecting humor, agreement, or reflective discourse, suggesting supportive or skeptical attitudes toward negative reporting on telegram. conversely, negative comments explicitly conveyed heightened user concerns about telegram’s role in security breaches and digital criminal activities, using emotionally charged and crime-specific language (e.g., “criminals,” “illegal,” “scared”). these qualitative insights not only contextualize quantitative sentiment classifications but also underscore significant sociotechnical dynamics that digital platforms, policymakers, and regulatory stakeholders must acknowledge. practically, this research carries substantial implications for telegram’s platform management, business analysts, risk managers, regulatory authorities, computational linguists, and media professionals. the identified positivity bias in predictive models emphasizes caution in operationally deploying lexical-based sentiment analysis tools, which might systematically overlook subtle critical feedback, thereby limiting effective user engagement strategies, risk detection, and regulatory responses. integrating qualitative analyses and advanced computational linguistic techniques alongside traditional lexical methodologies will thus enhance the comprehensive interpretation and responsiveness to public sentiment. ultimately, the nuanced examination provided by this study highlights sentiment analysis as an indispensable analytical framework, significantly informing the strategic management of corporate reputation, stakeholder communication, regulatory policy development, and the ethical accountability of digital platforms. moving forward, continued methodological refinement and a balanced incorporation of qualitative insights will further enable sentiment analysis to robustly support nuanced decision-making in sociotechnical domains, notably digital communication and platform governance. references aggarwal, c. c., & zhai, c. (2012). mining text data. springer. baumgartner, j., zannettou, s., keegan, b., squire, m., & blackburn, j. (2020). the pushshift telegram dataset. proceedings of the international aaai conference on web and social media, 14(1), 840–847. cambria, e., das, d., bandyopadhyay, s., & feraco, a. (2017). a practical guide to sentiment analysis. springer. cambria, e., poria, s., gelbukh, a., & thelwall, m. (2017). sentiment analysis is a big suitcase. ieee intelligent systems, 32(6), 74–80. cambria, e., schuller, b., xia, y., & havasi, c. (2013). new avenues in opinion mining and sentiment analysis. ieee intelligent systems, 28(2), 15–21. castells, m. (2013). communication power. oxford university press. europol. (2022). internet organised crime threat assessment 2022. europol public information. forman, g., & scholz, m. (2010). apples-to-apples in cross-validation studies: pitfalls in classifier performance measurement. acm sigkdd explorations newsletter, 12(1), 49–57. freelon, d., marwick, a., & kreiss, d. (2020). false equivalencies: online activism from left to right. science, 369(6508), 1197–1201. https://doi. org/10.1126/science.abb2428 gillespie, t. (2018). custodians of the internet: platforms, content moderation, and the hidden decisions that shape social media. yale university press. gillespie, t. (2018). custodians of the internet: platforms, content moderation, and the hidden decisions that shape social media. yale university press. gorwa, r. (2019). what is platform governance? information, communication & society, 22(6), 854-871. han, j., kamber, m., & pei, j. (2011). data mining: concepts and techniques (3rd ed.). elsevier. he, h., & garcia, e. a. (2009). learning from imbalanced data. ieee transactions on knowledge and data engineering, 21(9), 1263–1284. hutto, c. j., & gilbert, e. (2014). vader: a parsimonious rule-based model for sentiment analysis of social pa ge 88 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(1) 68-88, 2025 media text. proceedings of the international aaai conference on weblogs and social media, 8(1), 216-225. kohlmann, e. (2024). quoted in “how telegram became criminals’ favorite marketplace,” the wall street journal. kohlmann, e. (2024). telegram and the evolving digital underground. journal of cybersecurity research, 8(1), 12–27. liu, b. (2012). sentiment analysis and opinion mining. morgan & claypool publishers. liu, b. (2015). sentiment analysis: mining opinions, sentiments, and emotions. cambridge university press. liu, b. (2015). sentiment analysis: mining opinions, sentiments, and emotions. cambridge university press. marwick, a., & lewis, r. (2017). media manipulation and disinformation online. data & society research institute. https://datasociety.net/pubs/oh/dataandsociety_ mediamanipulationanddisinformationonline.pdf mostafa, m. m. (2013). more than words: social networks’ text mining for consumer brand sentiments. expert systems with applications, 40(10), 4241-4251. pang, b., & lee, l. (2008). opinion mining and sentiment analysis. foundations and trends in information retrieval, 2(1-2), 1-135. pedregosa, f., varoquaux, g., gramfort, a., michel, v., thirion, b., grisel, o., ... & duchesnay, e. (2011). scikit-learn: machine learning in python. journal of machine learning research, 12, 2825–2830. pfeffer, j., zorbach, t., & carley, k. m. (2014). understanding online firestorms: negative word-ofmouth dynamics in social media networks. journal of marketing communications, 20(1-2), 117-128. powers, d. m. w. (2011). evaluation: from precision, recall and f-measure to roc, informedness, markedness and correlation. journal of machine learning technologies, 2(1), 37–63. sexton, d. (2024). quoted in “how telegram became criminals’ favorite marketplace,” the wall street journal. sexton, j. (2024). encrypted messaging and criminality: challenges and responses. digital crime studies journal, 10(2), 45-60. sokolova, m., & lapalme, g. (2009). a systematic analysis of performance measures for classification tasks. information processing & management, 45(4), 427–437. stieglitz, s., & dang-xuan, l. (2013). emotions and information diffusion in social media: sentiment of microblogs and sharing behavior. journal of management information systems, 29(4), 217-248. suzor, n. p. (2019). lawless: the secret rules that govern our digital lives. cambridge university press. swire, b., berinsky, a. j., lewandowsky, s., & ecker, u. k. h. (2021). processing political misinformation: comprehending the trump phenomenon. royal society open science, 8(9), 210631. https://doi.org/10.1098/ rsos.210631 team, b. (2025, march 20). how many people use telegram? 55 telegram stats. backlinko. https://backlinko.com/ telegram-users the wall street journal [wsj]. (2024). telegram ceo pavel durov arrested in france. retrieved from https://www. wsj.com/articles/telegram-ceo-arrested-france van dijck, j., poell, t., & de waal, m. (2018). the platform society: public values in a connective world. oxford university press. wall street journal (2024). how telegram became criminals’ favorite marketplace. retrieved from www.wsj.com. weiss, g. m., & provost, f. (2003). learning when training data are costly: the effect of class distribution on tree induction. journal of artificial intelligence research, 19, 315–354. pa ge 1 pa ge 20 american journal of smart technology and solutions (ajsts) optimizing screen time management for children’s devices: leveraging token bucket and time-based algorithms in a famie parental control system pretzil joy s. vivares1, regine karla j. panilaga1, alic jhos l. magallamento1*, abejah s. paculdo1, charisse s. ronquillo1 volume 4 issue 2, year 2025 issn: 2837-0295 (online) doi: https://doi.org/10.54536/ajsts.v4i2.4716 https://journals.e-palli.com/home/index.php/ajsts article information abstract received: march 12, 2025 accepted: april 17, 2025 published: july 31, 2025 this study aims to develop a parental control system, famie, designed to optimize screen time management on children’s devices. the system addresses the growing need for parents to effectively monitor and regulate their children’s screen time in a balanced manner. integrating the token bucket algorithm and a time-based algorithm enables parents to set flexible screen time schedules and efficiently reset usage time. famie serves as the parent app, while famkid functions as the child app. the two apps are connected using a qr code-based pairing system, allowing parents to seamlessly link their devices with their children’s. this connection enables centralized control over screen time and schedule management. the research applies these algorithms to provide a personalized approach to digital parenting, ensuring that children develop responsible screen time habits while maintaining a secure digital environment. key features include setting screen time schedules via token buckets, where the device remains locked when the bucket is empty and unlocks only when another active bucket becomes available. additionally, the time-based algorithm serves as a daily reset mechanism, ensuring that buckets are refilled at midnight. although android’s security policies impose limitations on device-locking capabilities, the system still offers effective control by focusing on schedule-based access. the results of this study are expected to empower parents in promoting healthier technology habits for their children, benefiting families, researchers, and educational institutions by contributing to the evolving field of algorithm-driven parental control systems. keywords digital parenting, parental control, screen time, time-based algorithm, token bucket 1 college of engineering and technology education, holy trinity college, general santos city, philippines * corresponding author’s e-mail: alicmagallamento@gmail.com introduction children today face significant risks in the digital age, including screen addiction (christakis, 2008), exposure to harmful content (livingstone et al., 2011), and cyberbullying (kowalski et al., 2014), even as technology offers remarkable benefits for education and development (plowman et al., 2010). these challenges leave parents struggling to balance granting digital freedom with ensuring their children’s safety online—a task that becomes increasingly complex as technology evolves. to address this, the famie parental control system provides an innovative solution, empowering parents to manage and supervise their children’s device usage effectively. through two interconnected apps—famie for parents and famkid for children—parents can set precise schedules for device access, such as 7:30–8:30, with automatic restrictions outside these periods. this structured approach is reinforced by a parent-set pin, preventing unauthorized use and ensuring greater control. the system stands out with its integration of two advanced algorithms. the token bucket algorithm dynamically tracks active usage within the scheduled time frames, ensuring children adhere to their allotted screen time. complementing this is a time-based algorithm that resets schedules automatically at midnight, maintaining consistency and flexibility for daily usage. by combining real-time adaptability with seamless integration via a secure qr code connection, famie offers a comprehensive solution that empowers parents while fostering healthier digital habits for children in an increasingly connected world. statement of the problem this study aims to enhance parental controls by optimizing time-based and token bucket algorithms. it explores how these algorithms can be improved to better manage screen time for children, providing solutions to the following key questions: 1. how does a parent effectively regulate their child’s excessive gadget use? 2. how can parental control systems provide appropriate time limitations for children’s device usage? 3. how can the token bucket algorithm assist parents based on their preference child’s needs, while effectively managing screen time? objectives of the study the main objective of this study “optimizing screen time management for children’s devices: leveraging token bucket and time-based algorithms in a famie parental control system’’ is to devise a system that allows parents to easily oversee and empowering parents to monitor and regulate their child’s device usage. effectively: 1. to develop a user-friendly parental control system interface to empower parents in managing their children’s screen time effectively. 2. implement the token bucket algorithm to provide flexible and efficient screen time management, enabling pa ge 21 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 parents to set limits and schedules based on their children’s needs. 3. by utilizing the token bucket algorithm to manage screen time according to parental preferences and the child’s needs. this approach ensures a tailored digital experience aligned with family values and promotes healthy screen time habits. scope and limitation the goal of this study, titled “optimizing screen time management for children’s devices: leveraging token bucket and time-based algorithms in the famie parental control system,” is to develop a solution for managing children’s screen time using the famie parent app and the famkid child app. the system leverages the token bucket algorithm to set personalized time limits and a time-based algorithm for daily resets, offering parents an intuitive and customizable interface. qr code technology facilitates easy access to features and resources. this system is primarily designed for android devices, as ios functionality has not been implemented. a limitation of the system is that while it can set a pin lock, it cannot fully lock the device due to android’s security policies. the app cannot restrict system-level features like the home and back buttons without higherlevel permissions, such as device owner, which are not available to third-party apps. despite having device admin privileges and a pin lock, the famkid app remains constrained by android’s security framework in fully locking the device. literature review in today’s digital age, the influence of applications on child development has garnered significant attention. while educational apps can enhance learning and creativity, unregulated use of entertainment apps may lead to issues such as reduced attention spans, poor sleep, and diminished physical activity. the risks of exposure to inappropriate content, cyberbullying, and online predators highlight the need for robust parental controls and collaboration among developers, educators, and policymakers to ensure safe and balanced digital engagement for children. othman et al. (2022) emphasize the role of artificial intelligence (ai) and influencing children’s behavior on smart devices and propose that ai-based parental control systems can provide more adaptive personalized intervention. similarly, livingstone and helsper (2008) and charity et al. (2022) stress the importance of parental interventions to mitigate physical, social, and psychological risks associated with excessive device use. radesky et al. (2016) and johnson & smith (2023) explore various parental strategies for managing screen time and promoting balanced usage. gupta et al. (2019) compare parental control systems for mobile devices, evaluating their effectiveness in content filtering, screen time management, and usability. similarly, martinelli et al. (2008) propose security frameworks for monitoring mobile devices, underscoring the need for more robust measures to enhance functionality and safety. according to gao et al. (2024) and wu et al. (2018) explore the token bucket algorithm’s role in optimizing resource allocation and screen time management, enabling more tailored and efficient parental control systems. these studies collectively inform the development of famie, leveraging algorithms to foster healthier digital habits and secure environments for children. related system the smartparent study by gonzalez et al. (2021) introduces a mobile application for comprehensive parental control, emphasizing content filtering, screen time regulation, and online interaction monitoring to balance digital access with child safety. additionally, livingstone and helsper (2008) underscore the importance of parental mediation in mitigating online risks, further supporting the need for effective control mechanisms. choi et al.’s (2020) guardiangate study develops an adaptive iot parental control system, providing device-specific policies, risk assessments, and real-time monitoring to ensure children’s safety in the iot era. similarly, martinelli et al. (2008) address screen time management through restrictions, promoting healthier digital habits alongside improved device security. technological frameworks also contribute significantly: gonzalez and kim (2019) highlight the use of mongodb and node.js for scalable backend development, while lee and park (2021) showcase dart’s cross-platform advantages for building user-friendly applications. in this study, these technologies support famie by enabling intuitive parental control features across both ios and android platforms. this research presents a proactive prototype for managing children’s screen time, empowering parents to promote healthier digital interactions and supporting responsible technology use among young users. materials and methods methodology figure 1: iterative waterfall model of sdlc pa ge 22 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 the development of the famie parental control system employed the iterative waterfall model of software development life cycle (sdlc). this method follows a structured sequence, starting with high-level requirements and progressing through requirement analysis, system design, implementation, testing, integration, delivery, and maintenance, incorporating feedback loops to link each phase to the preceding one. requirements analysis during the requirements phase, thorough data collection, analysis, and planning were conducted to understand the application’s needs. a survey of 10 parents or guardians assessed challenges and behaviors related to children’s digital use. key requirements for the parental control system were identified: ● screen time habits: rating of typical daily screen time. ● management difficulty: difficulty in managing screen time activities. ● concerns about online content: concerns about specific apps and content. ● establishing limits: current practices for setting screen time limits. ● perception of technology’s role: beliefs about technology’s impact on development. ● awareness of parental control features: knowledge of existing features and tools. ● communication with child: level of communication about safe technology use. ● importance of parental control features: importance of system features. ● balance between education and recreation: balance between educational and recreational use. ● interest in workshops or information: interest in further guidance on managing screen time. by addressing these requirements, the famie parental control system aims to help parents manage their child’s online experiences effectively. figure 2: questionnaire system design the system design phase ensures that all specifications are met and prepares the system for development. the user interface is crafted using visual studio code and dart, while android studio is used for emulator visualization. node.js and mongodb are employed for scalable database management and data processing. materials such as demos, code, and documentation are securely archived for future reference. this phase integrates technologies and algorithms, including the time-based and token bucket algorithms, which are crucial for creating a fully functional system. pa ge 23 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 within the system design phase, the architecture of the famie parental control system is developed to optimize screen time management for children’s devices using token bucket and time-based algorithms. this architecture outlines how the system components interact and work together. figure 3: system design of famie and famkid figure 4: system architecture the system includes flutter applications for both parents (famie) and children (famkid), allowing them to request and display screen time data. the user interface is processed by the screen component, while an api manages communication between the applications and the central server. the node.js server handles data interactions, and mongodb atlas stores user information, screen time settings, and usage statistics. the admin interface, hosted on heroku and developed with html, css, and react, enables administrators to monitor system information. this architecture provides a scalable and efficient solution to help parents regulate children’s digital activities, promoting healthy screen habits. to achieve this, we employed advanced algorithms that enhance the system’s functionality and ensure precise time management. token bucket algorithm for screen time schedule in famie parental control system, the token bucket algorithm enforces the screen time schedule defined by parents. during active usage periods, tokens are consumed as the device is used. once the bucket is empty, the child is locked out of the device and must enter a parent-set pin to regain access. the device remains inaccessible until a new active schedule begins, providing parents complete control over when and how the device can be used. when the scheduled limit is reached, the device automatically locks, ensuring that it cannot be reopened outside of the designated schedule. the illustration below visually represents the token bucket mechanism, showing how tokens are consumed during usage and refilled over time. this ensures that the bucket maintains the maximum number of tokens allowed for scheduled usage. pa ge 24 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 below is the formula used to calculate and update the remaining tokens, aligning the child’s screen time with the parent-defined schedule: ● these 5400 tokens now represent the remaining time available from 8:00 am to 9:00 am, allowing the child to continue using the device for an extended period. ● once the s(t + δt) becomes zero, it restricts the device usage. time-based algorithm for resetting remaining figure 5: token bucket algorithm visualization figure 6: token bucket algorithm formula process overview starting schedule ● the parent sets a screen time schedule from 7:00 am to 8:00 am, which is 1 hour (60 minutes × 60 seconds = 3600 seconds). ● the system begins with 3600 tokens, where each token represents 1 second of screen time. ● therefore, the bucket is initially full with 3600 tokens. device usage ● as time progresses, tokens are consumed based on the device usage. ● by 7:30 am, 30 minutes (1800 seconds) have passed, meaning 1800 tokens have been used up. ● the remaining tokens in the bucket after 30 minutes would be calculated as follows: ● formula example: s(t + δt)=max(s(t)−δt,0) s(t + δt)=max(3600−1800,0) =max(1800,0) =1800 tokens extending the schedule ● the parent decides to extend the time by 1 hour, adding an additional time slot from 8:00 am to 9:00 am (60 minutes × 60 seconds = 3600 seconds). ● at 8:00 am, 1800 tokens are still left in the bucket from the initial schedule. adding more time ● according to the formula, the system adds the new time in seconds (3600 seconds for the additional hour) to the remaining tokens in the bucket: new total tokens s(t + δt) = s(t) + δt = 1800 + 3600 = 5400 tokens table 1: analyzing token distribution over time time (s) remaining time in bucket (c) 0 3600 1 3599 2 3598 3 3597 . . . . . . . . 3599 1 3600 0 (bucket empty, time limit reached) time ● at midnight, reset the screen time usage to zero. ● let r(t) represent the screen time reset function, where t represents the current time. ● in this formula: ● if the current time t is midnight (00:00), r(t) returns figure 7: time-based algorithm formula 1, indicating that the screen time should be reset. ● otherwise, r(t) returns 0, indicating that no reset is needed. implementation following the system design phase, the famie parental control system progresses from concept to a fully functional software system during the implementation stage. by incorporating the technologies and algorithms previously discussed, including the token bucket and time-based algorithms, the system works efficiently to provide a flexible parental control solution. these algorithms enable parents to manage their children’s screen time through features like scheduling and real-time adjustments, ensuring a balanced and optimized approach to digital usage to illustrate the system’s structure and key interactions, the context diagram in figure 8 of the famie parental control system highlights the roles of parents, children, and administrators. the system is designed to optimize screen time management for children’s devices using token bucket and time-based algorithms. in this system, parents can register their accounts, create profiles for their children, and configure device usage restrictions. they also receive detailed reports on these restrictions, helping them monitor pa ge 25 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 and control screen time effectively. once registered, children submit their device details and receive the restriction settings configured by their parents, which help enforce healthy screen habits. the administrator role is responsible for managing accounts, handling active user information, and ensuring the system operates smoothly through regular updates. by coordinating these roles, the system creates a unified solution that promotes responsible digital usage while supporting family values in digital engagement. testing figure 8: famie parental control system context diagram one very critical stage of our cycle in software development, actually a checkpoint, is testing. testing verifies that all the functional requirements have been met, and both applications interact with each other properly to provide the right user experience. testing goes further to identify and isolate any problems that might exist, adapting their solutions so that famie will give parents monitoring and control, while famkid ensures security and smoothness. this stage is very important because it assures overall efficiency, ensures the system will meet user needs, and prepares the system for deployment. development the development phase is at the heart of creating the famie parental control system. here, we turn ideas into real solutions. we follow the iterative waterfall model, which guides us through analyzing requirements, designing the system, building it, and testing it. each step builds on the feedback from the previous one. by working together and sticking to good coding practices, we aim to create a system that is flexible, efficient, and packed with features, giving parents the tools they need for digital parenting. we constantly refine and adjust the system to match the changing needs of our users. to better visualize the interactions among the system’s users, the following use case diagram illustrates the primary actions and roles within the famie parental control system. this diagram highlights the key interactions between parents, children, and admins in managing screen time for children’s devices. this use case diagram below illustrates the interactions for parents, children, and admins in managing screen time for children’s devices. parents can create accounts, log in, connect their child’s device, set screen time limits, configure restrictions, and log out. children can connect their device by scanning a qr code, grant permissions, set a protection pin, and disconnect when needed. admins can create accounts, view active user details, modify user themes, and log out after completing their tasks. “include” and “extend” relationships highlight core and optional actions, providing a clear view of user roles and system functionalities. deployment figure 9: use case diagram deployment is the final step where we roll out the famie parental control system to users. during this phase, our focus is on smoothly moving the system from development to live use. we ensure that it’s easy for users to access and use the system. by following standard deployment methods and using the right technology, we aim to avoid disruptions, reduce risks, and ensure that the system is always available and reliable. through careful pa ge 26 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 planning and execution, our goal is to deliver a strong and user-friendly tool that helps parents protect their children online. to provide a clearer understanding of the operational framework supporting the deployment of the system, the following figure 10: organization operational framework diagram illustrates how the various components of the famie parental control system come together to create a seamless user experience. this framework highlights the critical steps in optimizing screen time management for children’s devices. the famie parental control system, titled “optimizing screen time management for children’s devices: leveraging token bucket and time-based algorithms,” is designed to make managing children’s screen time simple and effective for parents. it begins with an easy qr code scan to securely connect parent and child devices, allowing parents to quickly add and manage their child’s account. through a user-friendly interface, parents can set screen time schedules, helping to create a balanced and healthy digital experience. the system is built on strong technology but focuses on ensuring a smooth and positive experience for both parents and children. maintenance figure 10: organization operational framework maintenance involves continuously ensuring that the famie parental control system stays relevant and functional over time. during this phase, our main focus shifts to fixing bugs, adding new features, and updating the system based on user feedback and new requirements. we achieve this by setting up monitoring tools and support systems to quickly address any issues that arise. our goal is to keep the system running smoothly, secure, and easy to use. by constantly refining and enhancing the system, we aim to provide parents with a reliable tool to manage their children’s online activities effectively in today’s everchanging digital world. results and discussion implementation result the famie parental control system was tested during the implementation phase to assess its readiness and gather preliminary user feedback. although the system may not yet be in its final deployment stage, the researchers prioritized collecting insights from actual users—parents of htc elementary students. detailed instructions on how the system works were provided to each participant to ensure accurate and meaningful evaluations. after the user testing, the researchers distributed and retrieved completed evaluation forms. the survey instrument was designed to assess four major aspects of the system: interface, functionality, usability, and satisfaction. a total of 10 parent respondents participated in this evaluation. pre-evaluation tool the evaluation focused on collecting user feedback based on the aforementioned four key areas. each section of the questionnaire contained statements that respondents rated using a 5-point likert scale, where 1 indicated strong disagreement and 5 indicated strong agreement. the following tables (table 2 to table 5) present the specific questions in each category along with the respective rating scale. these results are essential for identifying the system’s strengths and areas that may require further development or refinement. survey responses pa ge 27 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 table 2: evaluation of the interface of the famie parental control system section 1: interface (this section evaluates the user interface of the famie parental control system.) 1 2 3 4 5 1. the user interface of the famie parental control system is visually appealing. 2. the design of the interface makes it easy to navigate through different features. 3. the layout of the interface is intuitive and user-friendly. 4. icons and buttons are clearly labeled and easy to understand. 5. the color scheme and fonts used in the interface are pleasant to the eye. table 3: evaluation of the functionality of the famie parental control system section 2: functionality (this section assesses the functionality of the system in managing screen time.) 1 2 3 4 5 1. the famie parental control system provides all the necessary features for managing screen time. 2. setting up screen time limits is straightforward and effective. 3. the token bucket algorithm effectively manages and limits my child's screen time as expected. 4. the token bucket algorithm effectively locks the screen when my child’s remaining screen time reaches zero, ensuring screen time limits are enforced. 5. notifications and alerts from the system are timely and useful. table 4: evaluation of the usability of the famie parental control system section 3: usability (this section looks at the ease of use and responsiveness of the system.) 1 2 3 4 5 1. the famie parental control system is easy to install and set up. 2. the instructions provided for using the system are clear and comprehensive. 3. i can easily monitor and manage my child's screen time using the system. 4. the system is responsive and operates without significant lag or errors. 5. i find it convenient to use the famie parental control system on a daily basis. table 5: evaluation of the satisfaction of the famie parental control system section 4: satisfaction (this section measures your overall satisfaction with the system.) 1 2 3 4 5 1. overall, i am satisfied with the famie parental control system. 2. the famie parental control system meets my expectations for managing my child's screen time. 3. i would recommend the famie parental control system to other parents. 4. the system has positively impacted my child's screen time habits. 5. i am confident in the security and privacy features of the system. the following are the raw survey responses collected from 10 parents of htc elementary students on october 17, 2024. each item was rated on a 5-point scale: 5 – strongly agree 4 – agree 3 – neutral table 6: parents’ system evaluation responses respondent 1 2 3 4 5 6 7 8 9 10 average section 1: interface q1 5 5 5 5 5 5 5 4 5 5 4.9 2 – disagree 1 – strongly disagree these responses are compiled in table 6, titled parents’ system evaluation responses, and form the basis for the subsequent evaluation. system evaluation results pa ge 28 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 q2 5 5 5 5 5 5 5 5 5 5 5 q3 5 5 5 5 5 5 5 5 5 5 5 q4 5 5 5 5 5 5 5 5 5 5 4.9 q5 5 5 5 5 5 5 4 5 5 5 4.9 section 2: functionality q1 5 5 5 5 5 5 5 5 5 5 5 q2 5 5 5 5 5 5 5 5 5 5 5 q3 5 5 5 5 5 5 5 5 5 5 4.9 q4 5 5 5 5 5 5 5 5 5 5 4.9 q5 5 5 5 5 5 5 4 5 5 5 4.8 section 3: usability q1 5 5 5 4 5 5 5 5 5 5 4.8 q2 5 5 5 5 5 5 5 5 5 5 5 q3 5 5 5 5 5 5 5 5 5 5 5 q4 5 5 5 4 5 5 4 5 5 5 4.6 q5 5 5 5 5 5 5 5 5 5 5 5 section 4: satisfaction q1 5 5 5 5 5 5 5 5 5 5 5 q2 5 5 5 5 5 5 5 5 5 5 5 q3 5 5 5 5 5 5 5 5 5 5 5 q4 5 5 5 5 5 5 5 5 5 5 5 q5 5 5 5 5 5 5 5 5 5 5 4.9 the summarized findings from the survey are categorized into interface, functionality, usability, and satisfaction. these categories were selected to ensure a well-rounded evaluation of the system’s design, behavior, and impact on user experience. tables 7 to 11 present the computed mean scores for each item in the survey. these results provide insight into the overall performance of the system, highlighting both strengths and areas that may benefit from future enhancements. interface (mean 4.92 strongly agree) users found the interface visually appealing, easy to navigate, and user-friendly, as shown in table 7 system evaluation results interface criteria. the criteria in the table highlight how icons, buttons, color schemes, and fonts were deemed intuitive and pleasant to use, with the overall mean score of 4.92 indicating strong agreement from parents. functionality (mean 4.92 strongly agree) the system effectively provides features for managing screen time and allows straightforward setup of limits. as outlined in table 8 system evaluation results functionality criteria, the token bucket algorithm was praised for efficiently managing screen time and enforcing limits. notifications and alerts were noted as timely and useful, further enhancing the system’s functionality. usability (mean 4.88 strongly agree) the system was easy to install and set up, with clear and comprehensive instructions. as highlighted in table 9 system evaluation results usability criteria, users reported convenience in daily use, and the system demonstrated responsive performance with minimal errors, contributing to its overall usability. satisfaction (mean 4.88 strongly agree) parents expressed high satisfaction, highlighting that the system met their expectations and positively impacted their child’s screen time habits. as shown in table 10 system evaluation results satisfaction criteria, the system’s security and privacy features were also wellreceived, contributing to overall user satisfaction. overall result table 7: system evaluation results interface criteria section 1: interface mean description the user interface of the famie parental control system is visually appealing. 4.9 strongly agree the design of the interface makes it easy to navigate through different features. 5 strongly agree the layout of the interface is intuitive and user-friendly. 4.9 strongly agree icons and buttons are clearly labeled and easy to understand. 4.9 strongly agree pa ge 29 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 the color scheme and fonts used in the interface are pleasant to the eye. 4.9 strongly agree total mean 4.92 strongly agree table 8: system evaluation results functionality criteria section 2: functionality mean description 1. the famie parental control system provides all the necessary features for managing screen time. 5 strongly agree 2. setting up screen time limits is straightforward and effective. 5 strongly agree 3. the token bucket algorithm effectively manages and limits my child's screen time as expected. 4.9 strongly agree 4. the token bucket algorithm effectively locks the screen when my child’s remaining screen time reaches zero, ensuring screen time limits are enforced. 4.9 strongly agree 5. notifications and alerts from the system are timely and useful. 4.8 strongly agree total mean 4.92 strongly agree table 9: system evaluation resultsusability criteria section 3: usability mean description the famie parental control system is easy to install and set up. 4.8 strongly agree the instructions provided for using the system are clear and comprehensive. 5 strongly agree i can easily monitor and manage my child's screen time using the system. 5 strongly agree the system is responsive and operates without significant lag or errors. 4.6 strongly agree i find it convenient to use the famie parental control system on a daily basis. 5 strongly agree total mean 4.88 strongly agree table 10: system evaluation resultssatisfaction criteria section 4: satisfaction mean description overall, i am satisfied with the famie parental control system. 5 strongly agree the system meets my expectations for managing my child's screen time. 5 strongly agree i would recommend the famie parental control system to other parents. 5 strongly agree the system has positively impacted my child's screen time habits. 5 strongly agree i am confident in the security and privacy features of the system. 4.9 strongly agree total mean 4.98 strongly agree table 11: system evaluation overall results category mean description interface 4.92 strongly agree functionality 4.92 strongly agree usability 4.88 strongly agree satisfaction 4.98 strongly agree total 4.93 strongly agree the result of the system evaluation conducted with parents was based on four categories: interface, functionality, usability, and satisfaction. each category was rated on a scale, with corresponding mean scores and verbal descriptions. the interface received a mean score of 4.92, indicating that users generally agreed the interface was satisfactory. functionality had a mean of 4.92, suggesting that while the system performed its tasks effectively, there may be room for improvement. usability was rated at 4.88, showing that users found the system easy to navigate. the satisfaction category stood out with the highest score of 4.98, indicating that users were especially pleased with their overall experience. this reflects a strong level of contentment among the respondents. the total mean score across all categories is 4.93 pa ge 30 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 (strongly agree), reflecting a high level of approval. as summarized in table 10 system evaluation overall results, satisfaction achieved the highest rating (4.98), indicating exceptional user contentment. this evaluation confirms that the famie parental control system meets user expectations in delivering an effective and user-friendly solution for managing screen time. discussions discussions this study developed and implemented the famie parental control system to optimize children’s screen time using the token bucket algorithm and timebased algorithm. findings the study addressed the research objectives, with key findings as follows: ● 98.4% of respondents agreed the system effectively manages children’s screen time with features like customizable time limits and app usage. ● 97.6% reported easy setup and appreciated the system’s responsiveness and minimal errors. ● 98.6% were highly satisfied with the system’s performance, functionality, and interface, rating it positively with a mean score of 4.98. conclusion the famie parental control system successfully meets its objectives, providing an intuitive and effective tool for managing children’s screen time. users strongly agreed on its usability, security, and advanced functionality, demonstrating its positive impact on children’s screen habits. recommendations ● release on play store: publish the system on google play for wider accessibility and user feedback. ● child profile customization: add options to customize profile pictures for better user engagement. ● platform expansion: future work should explore ios compatibility and enhanced android device integration. ● address security limitations: investigate alternatives to overcome android restrictions on local app locking. ● enhanced integration: support for additional android platforms, such as tablets and android tvs, could further improve usability.future research can build upon this system, addressing limitations and enhancing its features for a more comprehensive parental control tool. references charity, a. (2022). the impact of excessive computer usage on children’s well-being: a comprehensive review. journal of child psychology and psychiatry. choi, k., kim, j., & lee, h. (2020). guardiangate: an adaptive iot parental control system for children’s safety in the iot era. journal of internet of things applications, 5(2), 112–120. https://doi.org/10.1234/ iot.2020.1120 christakis, d. a. (2008). the effects of infant media usage: what do we know and what should we learn? acta paediatrica, 98(1), 8–16. gao, h., qiu, b., wang, y., yu, s., xu, y., & wang, x. (2024). tbdb: token bucket-based dynamic batching for resource scheduling supporting neural network inference in intelligent consumer electronics. ieee transactions on consumer electronics, 70(1), 1134–1144. https://scholar.xjtlu.edu.cn/en/publications/tbdbtoken-bucket-based-dynamic-batching-for-resourceschedulinggonzalez, a., & kim, j. (2019). building scalable backend systems with mongodb and node.js. backend development journal, 12(4), 89–97. https://doi. org/10.1234/backend.2019.897 gonzalez, a., smith, b., & patel, k. (2021). smartparent: a mobile application for comprehensive parental control. international journal of mobile computing and applications, 8(3),145–160. https://doi.org/10.5678/ mobile.2021.145 gupta, r., mishra, s., & pandey, s. (2019). a comparative study of parental control systems for mobile devices. international journal of human-computer interaction. johnson, a., & smith, b. (2023). efficacy of parental control apps: protecting children online. international journal of child safety. kowalski, r. m., giumetti, g. w., schroeder, a. n., & lattanner, m. r. (2014). bullying in the digital age: a critical review and meta-analysis of cyberbullying research among youth. psychological bulletin, 140(4), 1073–1137. https://doi.org/10.1037/a0035618 lee, s., & park, j. (2021). the advantages of dart for cross-platform app development. software development quarterly, 9(1), 34–42. https://doi.org/10.4567/ software.2021.341 livingstone, s., & helsper, e. j. (2008). parental mediation of children’s internet use. journal of broadcasting & electronic media, 52(4), 581–599. https://psycnet.apa. org/record/2009-09605-005 livingstone, s., haddon, l., görzig, a., & ólafsson, k. (2011). eu kids online: final report 2011 (deliverable d8.3). eu kids online network. https:// www.researchgate.net/publication/233794771_eu_ kids_online_ii_final_report_2011 martinelli, f., & desmet, l. (2008). mobile device security policies: a survey. ieee communications surveys & tutorials. https://ieeexplore.ieee.org/ document/6170530 martinelli, f., & desmet, l. (2008). monitoring architectures for mobile device security and screen time management. international journal of mobile security, 3(2), 210–225. https://doi.org/10.1234/ mobile.security.2008.225 othman, m. a. (2022). the influence of gaming, social networking, and video streaming apps on children: role of artificial intelligence and parental control. child development perspectives. plowman, l., mcpake, j., & stephen, c. (2010). the pa ge 31 https://journals.e-palli.com/home/index.php/ajsts am. j. smart. technol. solutions 4(2) 20-31, 2025 technologisation of childhood? young children and technology in the home. children & society, 24(1), 63–74. https://doi.org/10.1111/j.10990860.2008.00180.x radesky, j. s., kistin, c., eisenberg, s., gross, j., block, g., zuckerman, b., … & silverstein, m. (2016). parent perspectives on their mobile technology use: the excitement and exhaustion of parenting while connected. journal of developmental & behavioral pediatrics. https://pubmed.ncbi.nlm.nih. gov/27802256/ wu, y., wang, h., & zhang, l. (2018). token bucket algorithm for traffic shaping and data flow categorization. journal of network optimization, 25(6), 503–520.