Correlated SKU assignment in warehouses using the joint demand probability distribution: a metaheuristic algorithm approach

Abstract

In warehouse management, picking orders from storage locations quickly and in the shortest time has become even more important with the development of e-commerce. Thus, efficiently assigning affined products to storage locations within the warehouses is crucial in reducing operational costs and preserving product quality. In this study, a Mixed-Integer Linear Programming model (MILP) is developed to minimize in-warehouse picking distances. Based on demand data, inter-product relationships are analyzed, and correlation coefficients are estimated for product pairs with a high tendency to be ordered together. These correlation values are then integrated into the objective function to optimize storage location decisions. To obtain faster and near-optimal solutions from the MILP model on large-scale data sets, a genetic algorithm (GA)-based approach has been developed. A set of computational experiments conducted on medium and large-scale instances compares the performance of the proposed GA approach with the Random-Based Correlated Skus Assignment Model (RBC-SAM). The GA approach under different scenarios shows an improvement of up to 22%.