Enhancing Labor Productivity of Garment Firms by Mixed Arena Simulation and Taguchi Design of Experiments (Case Company: MAA Garment and Textile Factory)

Abstract

This study presented a mixed Arena simulation and Taguchi experimental design approach to improve productivity in garment factory, with a focus on the round neck T-shirt production line at MAA Garment and Textile Factory. This production line was observed to influence the overall performance of the company highly for its low operator output, longer idle times and lower average resource utilizations. Realistic modeling of the production environment was ensured by fitting appropriate statistical distributions (e.g., normal, triangular, exponential and etc.) to each operation’s processing time, based on data collected through time studies. These distributions were selected using Arena Input Analyzer, which applies goodness-of-fit tests with the Kolmogorov Smirnov test method to determine the best fit for the observed data. Informal interviews with operators and line supervisors were conducted as well to gain an understanding of the variability and operational arrangements. The development of a verified simulation model and the development of dependable Taguchi experiments were made feasible with these inputs. Discrete-event simulation was used to model the existing production system using Arena software, enabling detailed analysis of flow dynamics, bottlenecks, and resource utilization. Two key operational factors; batch size and operator skill level were selected for experimental analysis using the Taguchi method with an L9 orthogonal array. Performance was evaluated based on three response parameters: throughput, average waiting time, and resource utilization. Signal-to-Noise (S/N) ratio analysis and main effects plots were employed to identify optimal factor settings that deliver stable and improved system performance. The results revealed that a batch size of 20 and high operator skill level yielded the most consistent gains in throughput and utilization while reducing delays. The optimized scenario, when re-simulated, showed notable improvements; 835-860, 779-743 and 68.3-70.8 (i.e., +3% throughput, -4.6% Average Waiting Time and +2.5% utilization) respectively, compared to the existing system model. This study demonstrated the effectiveness of integrating Arena Rockwell simulation and Taguchi design of experiments (DOE) in identifying data-driven productivity strategies, offering a practical framework for performance enhancement in garment production lines. However, points such as estimated input distributions due to lack of secondary data sets, limited factor analysis because of the limitations of Taguchi method were seen as the basic limitations of this study.