A Variable Neighborhood Search Algorithm for the Truck-Drone Routing Problem

The rise in e-commerce sales has increased the demand for logistics companies to handle the problem of last-mile delivery, which is complex and incurs large operational expenses. The use of drones in the development of modern distribution networks opens new avenues for improving existing delivery systems. The integration of drones with traditional delivery trucks that form the truck-drone systems are being proposed by logistics firms for use in delivery procedures. In this paper, we consider a single truck and drone delivery system and address the associated routing problem. As a result, a variable neighborhood search algorithm is developed and tested on small and medium instances. The algorithm shows its potential ability to handle instances in a short time. Finally, the introduced truck-drone delivery system with multiple visits per drone dispatch generates cost-savings in comparison with the case of a single visit per delivery trip.