Sustainability-Driven Decision Making: A Multi-Objective Optimisation Approach to Optimise Closed-Loop Supply Chains of Electrical Drives

Abstract

Because of growing concerns about global hike in e-waste and resource scarcity, this thesis explores how a manufacturing company can make better decisions for end-of-life (EoL) electrical drives. Electrical drives are widely used in industrial processes. They often become obsolete due to rapid technological change, creating a critical need for sustainable and cost-effective solution.

The research is based on a case study of electrical drive manufacturer who are exploring different alternatives in the principles of the circular economy and focuses on a key practical question: should a used electrical drive be repaired or disposed of when the objective is to minimize the cost and greenhouse gas (GHG) emission? A data-driven, multi-objective optimization model was developed using a Mixed Integer Linear Programming (MILP) and the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to answer the question.

The model aims to minimize both total costs and GHG emissions while considering realistic constraints such as budget limits, quality of returned products, and age. As part of the case study, a batch of 50 EoL products were analyzed. The result indicated a trade-off between environmental and economic objectives - lower emissions with higher costs, and vice versa. The model was further validated by performance indicators and parameter tuning. The performance indicators confirmed the model’s performance while the robustness of the model was tested through parameter tuning. Three different configuratios were selected (large, small and medium population) for parameter tuning. The configuration with large population provided the best result with cost being € 12031,60 and GHG emission being 15658,50 KgCO2-e. The study also conducted sensitivity analysis to evaluate the influence of key variables. Product quality and transport distance were found to be the most significant factors. Improving the quality of returned products significantly reduced both costs and emissions, while longer transport distances had a measurable negative effect. Interestingly, changing the budget had little impact on decisions, confirming that technical feasibility outweighed financial flexibility in such constrained environments.

This thesis contributes a practical framework for manufacturers like the case company which are seeking to align circular economy goals with operational decision-making. The model serves as a decision-support tool that helps identify feasible and sustainable EoL decisions by offering a view of cost and environmental trade-off. The findings are also relevant for other manufacturers, policymakers, and sustainability-driven enterprises aiming to reduce e-waste and promote circular business practices.