A hybrid genetic algorithm-based decision support system for enhancing transportation efficiency in reverse logistics

Abstract

Two application case studies were also conducted so as to evaluate the proposed models and algorithms. The first case is motivated by the distribution and recycling of water carboys, while the second case investigates the collection of waste from the commercial and industrial sectors. The computational experiments performed in the two case studies show that the proposed models and algorithms allow fleet managers to determine cost-effective transportation plans. Particularly, the system enables a diverse control of various economic and environmental costs and a flexible approach so as to provide relevant information to enable fleet managers to consider the compromises or trade-offs among different cost indicators, such as transportation cost, manpower, service level and even the value of returned products. The models and methods enable logistics decision makers to determine proper and optimal logistics strategies. Notably, they can also be generalized to any other type of Reverse Logistics activity in practice. The contribution of this study is twofold. For industry and in general practice, a decision support system is proposed to evaluate the possible economic and environmental significance of real-world transportation problems and to take action at different levels to carry out Reverse Logistics. For academic development, this research is distinguished and featured by proposing a new variant of vehicle routing model that is characterized by optional backhauls and multiple objectives is proposed. The model aims to minimize transportation cost, balance the driver workloads, maintain high service levels, and to maximize the number of recycled products. In addition, a hybrid Genetic Algorithm characterized by a greedy look-ahead heuristics and a Pareto Ranking Scheme is proposed to seek Pareto Optimality for multi-objective optimization.