Exact and heuristic approaches for the Truck–Drone Team Logistics Problem

Abstract

Truck-and-drone routing problems are a topic of increasing interest due to advancements in drone technology. Initially, drones were thought to serve only one customer per flight; however, recent improvements in drone technology have led to the definition of multi-visit truck-and-drone routing problems, where a drone can serve multiple customers in each flight. These routing problems, which require a careful synchronization between the truck and the drone, are challenging to solve, even with a small number of customers. In this work, we address a multi-visit, single-truck, single-drone routing problem also known in the literature as the Truck–Drone Team Logistics Problem (TDTLP). In particular, we study two versions of the TDTLP arising from the possibility for the drone to land or hover at the recollection points. Our study proposes a novel mixed-integer linear programming formulation for the TDTLP, solved by a Branch-and-Cut (B&C) algorithm. Then, we introduce a matheuristic approach that incorporates our B&C algorithm to solve medium- and large-size instances. Computational results on benchmark instances demonstrate the robustness and efficacy of the proposed methods.