Fleet Management for Pickup and Delivery Problems with Multiple Locations and Preferences

Abstract

To provide more routing flexibility and improve service in delivery processes, we extend the Pickup and Delivery Problem with multiple time-location combinations for service. Furthermore, we introduce preference possibilities for each option, and aim for finding solutions that balance minimizing total travel costs and customer or operator dissatisfaction. We compare an Adaptive Large Neighborhood Search metaheuristic with solving the problem exactly. Simulation experiments indicate that a multiple-location scenario is highly beneficial compared to the corresponding single-location scenario and that the metaheuristic always finds the optimum if this could be computed by the exact solver.