50 / 2024-04-17 11:47:08

Jointly study storage assignment, order batching, and pod selection in robotic mobile fulfillment systems considering picker fatigue

warehousing, robotic systems, picker fatigue

Human-machine coordination is prevalent in e-commerce order fulfillment systems, particularly in the Robotic Mobile Fulfillment System (RMFS) widely used by major retail e-commerce platforms such as Amazon, JingDong, and Cainiao. From the perspective of human-machine coordination, we aim to jointly address storage assignment, order batching, and pod selection to minimize both pickers' total energy expenditure and robots' total transport distance, abbreviated the integrated problem as the JIOPP. We formalize this integrated problem as a mixed-integer programming model. Additionally, we introduce an improved multiple objectives algorithm called NSGAII-ILS to solve real-world instances of the problem. Extensive numerical experiments demonstrate the competitiveness of our NSGAII-ILS compared to state-of-the-art algorithms like IKnEA and KnEA, as it can identify the Pareto solution set closer to the true Pareto front. The values of hypervolume (HV) and Coverage (C-metric) support our conclusions. Furthermore, we assess the impact of batch sizes, number of pod layers, and various pod selection policies on the JIOPP using real-world e-commerce order instances. Finally, we provide meaningful management implications for warehouse managers, including adopting a reasonable batch size to save more than 35% of pickers' energy expenditure and over 70% of robots' transportation distance. Proper pod layering and pod selection policies are vital factors for balancing objectives and enhancing order picking efficiency.