-
Routing One Million Customers in a Handful of Minutes
Authors:
Luca Accorsi,
Daniele Vigo
Abstract:
In this paper, we propose a new dataset of Capacitated Vehicle Routing Problem instances which are up to two orders of magnitude larger than those in the currently used benchmarks. Despite these sizes might not have an immediate application to real-world logistic scenarios, we believe they could foster fresh new research efforts on the design of effective and efficient algorithmic components for r…
▽ More
In this paper, we propose a new dataset of Capacitated Vehicle Routing Problem instances which are up to two orders of magnitude larger than those in the currently used benchmarks. Despite these sizes might not have an immediate application to real-world logistic scenarios, we believe they could foster fresh new research efforts on the design of effective and efficient algorithmic components for routing problems. We provide computational results for such instances by running FILO2, an adaptation of the FILO algorithm proposed in Accorsi and Vigo (2021), designed to handle extremely large-scale CVRP instances. Solutions for such instances are obtained by using an a standard personal computer in a considerably short computing time, thus showing the effectiveness of the acceleration and pruning techniques already proposed in FILO. Finally, results of FILO2 on well-known literature instances show that the newly introduced changes improve the overall scalability of the approach with respect to the previous FILO design.
△ Less
Submitted 25 June, 2023;
originally announced June 2023.
-
Very large-scale neighborhood search for drone routing with energy replenishment
Authors:
Catherine Lorenz,
Nicola Mimmo,
Alena Otto,
Daniele Vigo
Abstract:
The Drone Routing Problem with Energy replenishment (DRP-E) belongs to a general class of routing problems with intermediate stops and synchronization constraints. In DRP-E, the drone has to visit a set of nodes and routinely requires battery swaps from a (potentially) mobile replenishment station. Contrary to widespread restrictions in the drone routing literature, several destinations may be vis…
▽ More
The Drone Routing Problem with Energy replenishment (DRP-E) belongs to a general class of routing problems with intermediate stops and synchronization constraints. In DRP-E, the drone has to visit a set of nodes and routinely requires battery swaps from a (potentially) mobile replenishment station. Contrary to widespread restrictions in the drone routing literature, several destinations may be visited in between two consecutive battery swaps. In this paper, we propose a nontrivial very large-scale neighbourhood for DRP-E, which synergetically leverages two large-sized polynomially solvable DRP-E SubProblems (SP1 and SP2). The number of feasible solutions in the resulting neighborhood is a multiple of those in SP1 and SP2, and, thus, exponential in the input size of the problem, whereas the computational time to search it remains polynomial. The proposed polynomial two-stage dynamic programming algorithm VLSN to search this neighborhood can be flexibly adjusted to the desired trade-off between accuracy and computational time. For instance, the search procedure can be converted into an exact algorithm of competitive runtime for DRP-E. In computational tests, the developed solution methods outperform current state-of-the art heuristics for DRP-E by a significant margin. A case study based on a search for missing persons demonstrates that VLSN easily accommodates additional practice relevant features and outperforms the state-of-the-art solution in disaster relief by 20%.
△ Less
Submitted 28 July, 2022;
originally announced July 2022.
-
Guidelines for the Computational Testing of Machine Learning approaches to Vehicle Routing Problems
Authors:
Luca Accorsi,
Andrea Lodi,
Daniele Vigo
Abstract:
Despite the extensive research efforts and the remarkable results obtained on Vehicle Routing Problems (VRP) by using algorithms proposed by the Machine Learning community that are partially or entirely based on data-driven analysis, most of these approaches are still seldom employed by the Operations Research (OR) community. Among the possible causes, we believe, the different approach to the com…
▽ More
Despite the extensive research efforts and the remarkable results obtained on Vehicle Routing Problems (VRP) by using algorithms proposed by the Machine Learning community that are partially or entirely based on data-driven analysis, most of these approaches are still seldom employed by the Operations Research (OR) community. Among the possible causes, we believe, the different approach to the computational evaluation of the proposed methods may play a major role. With the current work, we want to highlight a number of challenges (and possible ways to handle them) arising during the computational studies of heuristic approaches to VRPs that, if appropriately addressed, may produce a computational study having the characteristics of those presented in OR papers, thus hopefully promoting the collaboration between the two communities.
△ Less
Submitted 28 September, 2021;
originally announced September 2021.
