Modelling and optimising a multi-depot vehicle routing problem for freight distribution in a retail logistics network

Abstract

An efficient freight distribution network is critical for enhancing competitiveness by lowering transportation costs and increasing profitability. This study adopts a case-based modelling approach to tackle a real-world Multi-Depot Vehicle Routing Problem (MDVRP) faced by a UK-based retailer aiming to expand its operations in northern UK. Due to high fixed costs and a limited branch network, the retailer seeks to improve operational efficiency by reducing transportation costs without establishing additional facilities. A novel mixed-integer programming model is developed to optimise the existing distribution network by incorporating realistic operational constraints. The model addresses key complexities such as driver costs, inter-depot routing, transportation hubs, multiple depots, dynamic demand, a heterogeneous fleet, cross-docking, multiple product types, vehicle capacity and travel time restrictions. Using an exact solution method, the model yields optimal results demonstrating significant reductions in transportation costs while maintaining service constraints. The findings provide valuable research insights and practical recommendations for optimising freight distribution networks under realistic and resource-constrained conditions.