Title: Distance metrics matter: analysing optimisation algorithms for the last mile problem

Authors: Jay R. Brown; Maxim A. Bushuev; Alfred L. Guiffrida

Addresses: Sellinger School of Business and Management, Loyola University Maryland, Baltimore, MD 21210, USA ' Department of Information Science and Systems, Earl Graves School of Business and Management, Morgan State University, Baltimore, MD 21251, USA ' Department of Management and Information Systems, Kent State University, Kent, OH 44242, USA

Abstract: This research compares and contrasts the performance of 11 optimisation algorithms and two new combined optimisation algorithms for solving the last mile delivery problem. The last mile delivery problem is modelled as a circular region with a central depot and customers randomly distributed throughout the region. The performance of the optimisation algorithms is studied with respect to expected travel distance and computation time needed to achieve a solution for problems with varying numbers of customers and where travel distance is measured with Euclidean and Manhattan distance metrics. The paper represents the first analytical comparison of different algorithms for optimising last mile delivery when the number of customers varies stochastically under both Manhattan and Euclidian distance metrics. Findings show that the number of customer deliveries and the metric used to measure travel distance impacts a decision maker's choice of the best algorithm and that employing multiple algorithms is recommended.

Keywords: last mile problem; LMP; supply chain optimisation; algorithms; heuristics; Manhattan distance.

DOI: 10.1504/IJLSM.2021.113233

International Journal of Logistics Systems and Management, 2021 Vol.38 No.2, pp.151 - 174

Received: 24 Apr 2018
Accepted: 19 May 2019
Published online: 25 Feb 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article