Benchmark Instances and Optimal Solutions for the Traveling Salesman Problem with Drone

TL;DR

This paper introduces benchmark instances and optimal solutions for the truck-and-drone routing problem, addressing a gap in standardized testing and validation tools for this emerging logistics optimization field.

Contribution

It defines nine basic problem settings for truck/drone routing and provides optimal solutions for existing instances, creating publicly available benchmarks.

Findings

Optimal solutions for various problem settings

Organized benchmark instances for validation

Enhanced tools for future research validation

Abstract

The use of drones in logistics is gaining more and more interest, and drones are becoming a more viable and common way of distributing parcels in an urban environment. As a consequence, there is a flourishing production of articles in the field of operational optimization of the combined use of trucks and drones for fulfilling customers requests. The aim is minimizing the total time required to service all the customers, since this has obvious economical impacts. However in the literature there is not yet a widely recognized basic model, and there are not well assessed sets of instances and optimal solutions that can be considered as a benchmark to prove the effectiveness of new solution methods. The aim of this paper is to fill this gap. On one side we will clearly describe some of the most common components of the truck/drone routing problems and we will define nine basic problem settings, by combining these components. On the other side we will consider some of the instances used by many researchers and we will provide optimal solutions for all the problem settings previously identified. Instances and detailed solutions are then organized into benchmarks made publicly available as validation tools for future research methods.