A dynamic closed-loop vehicle routing problem with uncertainty and incompatible goods

This study investigates a dynamic closed-loop vehicle routing problem (VRP) with uncertain pickup and deterministic delivery of incompatible goods, which is an extension of the VRP with simultaneous pickup and delivery (VRPSPD) in closed-loop logistics, where the incompatibility between goods of pickup and delivery is considered. The problem involves minimizing transportation cost, incompatibility and number of customers visited twice. A solution method based on variable neighborhood search (VNS) is developed for solving the VRPSPD. The pickup uncertainty is handled in two stages: first, a priori routes are generated by solving a VRPSPD whose pickup demands are estimated; second, the a priori routes are simulated by dynamically satisfying the pickup demand under incompatibility, and the second-round routes are generated to meet the unmet demands. The effects of considering the incompatibility are examined by experiments. A case of centralized tableware disinfection and logistics services in China's catering industry is used for demonstration. Disinfected tableware for delivery and used tableware for pickup are incompatible because of potential cross-contamination. The experimental results quantitatively provide insights for managers who must solve the dynamic closed-loop VRP with uncertain pickup and incompatible goods. The proposed method also proves competitive for the VRPSPD.