Truck Routing for Parcel Delivery
Solving a Multi-depot Pickup and Delivery Problem with Occasional Drivers using ALNS
More Info
expand_more
Abstract
This research presents a mathematical model for routing that incorporates multiple depots, occasional drivers, and multiple depot visits, a problem faced by many companies in reality. An Adaptive Large Neighborhood Search (ALNS) algorithm was employed, using Random and Worst removal operators, Basic Greedy and Regret-2 insertion strategies, a roulette wheel for operator selection, and simulated annealing for acceptance criteria. Computational experiments validated the effectiveness of the ALNS algorithm and model.Sensitivity analysis revealed that adding depots (ODs) can reduce routing costs by up to 15.21%, with various OD capacities offering additional savings (7.86% for 60% capacity and 9.97% for 70% capacity). A case study with the Dutch e-commerce company Ochama, scaled to 150 requests, confirmed practical applicabil- ity, achieving cost reductions of 11.37% for small vehicles and 6.85% for medium-sized vehicles. The results highlight that integrating ODs into routing strategies can significantly lower costs, with optimal outcomes dependent on market research to balance savings, service quality, and OD capacity.