Dr. Davod Hosseini, PhD
Service Time Window Design in
Routing Optimization under Uncertainty
Assistant Professor, Saint Mary's University
Date: March 22, 2023
Time: 1:00-2:00 PM
Venue: "I" Building, Room 220, Sexton Campus
Online MS Team: Click here to join the meeting
In this paper, we present a new routing optimization approach where a service provider aims to design service time windows to visit a set of customers within a given time budget in a network with stochastic and possibly correlated travel times. To construct the service time window for each customer, we introduce two criteria to address the length of the time window and the risk of violating it. To tackle the latter, we define two performance metrics to quantify the early and late arrivals that do not need prior knowledge or assumptions about the route and time windows but account for both the frequency and magnitude of the time window violations. The service provider can allocate different penalties indicating risk preferences/tolerance to each of these criteria, resulting in various routes and time windows with different levels of service guarantee. We provide two modeling frameworks based on stochastic and distributionally robust optimization to handle uncertainty in travel times. In each setting, we derive closed-form solutions for the optimal time windows that are functions of the service provider's risk preference on both criteria. Moreover, these closed-form solutions provide opportunities to reformulate the proposed models and make them more tractable to commercial solvers. We develop two decomposition-based algorithms to find the exact optimal routing and time window assignment solutions. We show the efficacy of our proposed models on a rich collection of instances derived from some well-known datasets in the literature. While a small portion of the time windows designed by the stochastic model was violated on the out-of-sample test instances, the distributionally robust model generates more reliable routes and time windows whose violation rates never exceeded the risk tolerance of the service provider. Our computational experiments also demonstrate the efficiency of our proposed algorithms in improving the performance of a state-of-the-art commercial solver.
Dr. Davod Hosseini joined the Sobey School of Business, Saint Mary’s University as an assistant professor of Management Science in July 2021. He holds a PhD in Management Science from the DeGroote School of Business, McMaster University (Dec 2018). Prior to that, he did an MSc and BSc in Industrial Engineering in Iran. His research interests are in the application of machine learning, optimization under uncertainty, and risk analysis to the solution of transportation planning and logistics management problems.
Industrial Engineering seminar series contact person:
Dr. Floris Goerlandt
Ms. Susan Russell McGrath