EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

3525. Evolutionary Approaches to Solve the Dynamic and Stochastic Inventory-Routing Problem

Invited abstract in session WA-50: Retail Distribution I, stream Retail Operations.

Wednesday, 8:30-10:00
Room: M2 (building: 101)

Authors (first author is the speaker)

1. Francisco Maia
Industrial Engineering and Management, FEUP, INESC TEC
2. Gonçalo Figueira
INESC-TEC, Faculty of Engineering of Porto University
3. Fábio Moreira
INESC TEC, INEGI, Faculty of Engineering, University of Porto

Abstract

Sequential-decision making is an active area of research in Operations Research (OR), where real-time decision-making, machine learning and data-driven approaches are stepping forward.
The Dynamic and Stochastic Inventory-Routing Problem (DSIRP) is one of the most fundamental problems companies seek to optimize, given its meaningfulness at strategic and operational levels.
The challenge at hand involves the integration of inventory management and vehicle routing problems while at the same time effectively handling the dynamic and stochastic nature of customer demands unveiled over time.
A promising research path lies in combining Reinforcement Learning and OR to achieve novel inventory and routing policies with significant cost savings. Accordingly, Policy Function Approximation (PFA) emerges as an encouraging strategy.
This research intends to give new insights into solving the single-item, single-vehicle DSIRP with a one-to-many endpoint structure, where decisions must be released periodically. Therefore, we propose novel delivery policies based on PFA, exploring two evolutionary algorithms, i.e. Genetic Programming and Genetic Algorithms.
The proposed methodology enabled us to derive novel rules that present competitive results compared to several literature benchmarks, optimizing the balance between holding and stockout costs, reducing waste and decreasing travel distances.

Keywords

Status: accepted

Back to the list of papers