EURO 2025 Leeds
Abstract Submission

1257. Service network design with uncertainty: A simulation-optimization approach

Invited abstract in session MC-59: Freight Transportation and Logistics, stream Transportation.

Monday, 12:30-14:00
Room: Liberty 1.14

Authors (first author is the speaker)

1. Javier Durán-Micco
Department of Maritime and Transport Technology, Delft University of Technology
2. Bilge Atasoy
Maritime and Transport Technology, Delft University of Technology

Abstract

In this study, we address the Service Network Design (SND) problem from the perspective of a logistics service provider operating within a multimodal network. The goal is to solve this tactical problem while incorporating information and feedback from the operational level. To achieve this, in addition to the typical capacity planning decisions in SND, we also account for uncertain travel times, routing decisions for a fleet of trucks, and dynamic reaction and re-planning to face disruptions. As such, we propose an approach that explores the interactions between tactical and operational decisions. Specifically, we present a simulation-optimization approach capable of handling all these factors. We employ a simulated annealing algorithm to optimize capacity planning, while an agent-based simulation model is used to estimate costs by capturing uncertainties and dynamic decision making. Since simulations are time-consuming, we introduce a surrogate function, trained on simulation data, to estimate costs efficiently. An adaptive learning technique further refines predictions during optimization, also using the simulation model. The method is tested using benchmark instances from the literature. Preliminary results indicate that our adaptive methods achieve strong performance while significantly reducing computational time. Future work will focus on refining models to better reflect real-world conditions and incorporating dynamic interactions between logistics actors.

Keywords

Status: accepted

Back to the list of papers