EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

1983. Dynamic Programming-based formulations and algorithms applied to inventory management: A systematic review

Contributed abstract in session MA-33: Optimal control in supply chain management, stream Optimal Control Theory and Applications.

Monday, 8:30-10:00
Room: 42 (building: 303A)

Authors (first author is the speaker)

1. Edson Souza
Operations Management, IESEG School of Management
2. Stefan Creemers
UCLouvain
3. Tanja Mlinar
IESEG School of Management, Univ. Lille, CNRS, UMR 9221 - LEM - Lille Economie Management
4. Maud Van den Broeke
Operations and Supply Chain Management, Ieseg School of Management

Abstract

Dynamic programming (DP) has been consistently employed in inventory management, especially in multi-period formulations, contributing to our understanding of the field. Despite the challenges posed by the curse of dimensionality, DP-based formulations have evolved over time to alleviate its effects, especially as inventory problems become more complex. This review aims to analyze the use and evolution of DP-based methods in inventory management, encompassing 287 articles. To provide a comprehensive overview of DP-based formulations for inventory problems, we propose a typology classification, centered inventory- and DP-based criteria, by categorizing inventory problems based on the types of decisions explored and the main characteristics of the problem investigated, and by identifying DP approaches employed. We emphasize the practical implications of DP approaches in industry, specify limitations in existing studies, and suggest future research directions. In particular, we stress the relevance of considering uncertainties beyond demand-related factors, integrating sustainability aspects into DP approaches such as incorporation of government subsidy levels and regulations, and exploring inventory problems in reverse logistics and remanufacturing. Furthermore, we highlight the need for further development of applied studies utilizing Approximate and Hybrid DP in combination with data-driven methodologies to address challenges in real-life inventory scenarios.

Keywords

Status: accepted


Back to the list of papers