Multi-objective meta-heuristics to optimize end-of-life laptop remanufacturing decisions under quality grading of returned parts.

Research on multi-objective discrete optimization of Waste Electrical and Electronic Equipment (WEEE) remanufacturing remains under-studied in the literature. Remanufacturing laptops to extend their useful life is viewed as the best End-Of-Life (EOL) alternative considering environmental and social factors. This paper develops a model to decide the best EOL option, namely reuse, conditional repair, and disposal of quality-graded laptop parts, with economic and environmental objectives. The first objective is to maximize the profit of remanufactured laptops over a multi-period planning horizon. The second objective is to minimize the emissions associated with remanufacturing. A Multi-Objective Discrete Particle Swarm Optimization (MODPSO) algorithm and Non-dominated Sorting Genetic Algorithm II (NSGA-II) are embedded as a decision support tool in Microsoft Excel with a user interface to yield Pareto optimal solutions, and the results are compared. The Taguchi approach is applied to find the optimum value of the control parameters of the proposed algorithms. The approach is tested with inputs from an authorized remanufacturer in Bangalore, India. The performance of the algorithms is further investigated using randomly generated test problems. The MODPSO algorithm provided better solutions for all problem instances based on the convergence and diversity metrics. The inclusion of conditional repair options and parts with low and medium-quality grades in remanufacturing leads to higher profit, albeit with more emissions. A variation in the quality grade assigned to the conditional repair option for the parts needed for higher profit margin laptops is observed. A sensitivity analysis is conducted to observe the impact of supply, demand, repair cost, shortage cost, and emissions on the two extreme Pareto solutions. The decision-making tool offers a continuum of trade-offs to help a remanufacturer choose the EOL options, depending on economic and environmental performance preferences. [ABSTRACT FROM AUTHOR]