Optimal Production Planning Models for Garment Industries in Bangladesh under Stochastic Atmosphere

Abstract

A key pillar of Bangladesh’s economy, the ready-made garment (RMG) industry makes a substantial contribution to the country’s GDP, foreign exchange earnings, and job creation. This study presents a thorough modeling framework for production planning optimization in Bangladesh’s RMG sector under both deterministic and stochastic circumstances. In order to maximize profit and optimize production planning of the RMG industry, this study develops first a deterministic linear programming (LP) model. The model finds the optimal product mix and output levels to effectively meet demand while lowering production costs and increasing profitability when it is applied to a real-world factory setting in Gazipur, Dhaka. Uncertainties in global demand trends, variable manufacturing costs, volatile raw material prices, and dynamic international trade rules etc. are the most vital reasons of the continuously increased challenges to the industry. This research seeks to develop strong decision-making frameworks to support strategic and tactical decision-making under such uncertainty. Taking into consideration economic fluctuations, the deterministic LP is expanded in this stage into stochastic programming models (SLP) in which all significant factors are represented as random variables, including cost coefficients, demand levels, labor availability, and processing durations. Another stochastic model also formed in this research by combining all the scenarios, as it is not actually predictable which situation would be come. This approach provides a strong planning for RMG under a variety of circumstances by capturing holistic uncertainty. A two-stage stochastic linear programming model (TSLP), in which only a chosen subset of parameters remains stochastic, is developed next in this study to reduce uncertainty. Following the revelation of parameter realizations, decisions taken in stage one are modified in stage two. This model improves tractability and increases adaptability. Expected Value of Perfect Information (EVPI) and Value of the Stochastic Solution (VSS), two important parameters in stochastic programming that measure the advantages of uncertainty modeling and perfect foresight, are used to assess the performance of deterministic, general stochastic, and two-stage stochastic models. xiii In next, two separate two-stage stochastic (TSLP) models are formulated in this study, where the demand uncertainty represents various scenarios, including seasonal variations. In this case, the fluctuation of one product’s demand is presented. Key uncertainties in demand, export prices, labor costs, raw material costs, and operational costs related to that specific product are included in scenarios that are created using historical data and probabilistic distributions. These models are then expanded into a multistage stochastic programming (MSLP) model, which simulates a decision process by allowing decision variables to change over several stages as demand unfolds and gradually adapting decisions at each stage to capture dynamic changes and integrating learning over time. By applying these models, a RMG factory can be able to decide the best production planning along with the profit and cost optimization. A comparison between the two-stage and multistage is presented next. LINDO and AMPL (A Mathematical Programming Language) are used to solve all the deterministic and stochastic models, while Excel Solver is applied for preparing all the graphical presentations of this study. In terms of predicted profit, robustness to uncertainty, and overalsl operational efficiency, the stochastic models perform noticeably better than the deterministic model, according to computational results validated using real-world data gathered from RMG companies in Bangladesh. By offering an integrated optimization framework that enables RMG stakeholders to use scenario-based planning, efficient, uncertainty-aware design tools from deterministic LP to comprehensive multistage stochastic programming. Optimization tools to increase profitability, the study adds both theoretically and practically significant, as global markets continue to vary. By presenting a thorough modeling framework designed to address the unique difficulties faced by Bangladesh’s RMG industry in order to facilitate adaptable and proactive decision-making. In the geopolitically unstable environment, where flexibility and risk-aware planning are critical also in the post-COVID atmosphere, these findings are especially pertinent. And provide resiliency in a world economy that is becoming more and more uncertain.