EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

3209. A heuristic for an inventory control policy in a closed-loop distribution multi-echelon supply-chain with returns

Invited abstract in session WA-49: Stochastic inventory systems, stream Lot Sizing, Lot Scheduling and Production Planning.

Wednesday, 8:30-10:00
Room: M1 (building: 101)

Authors (first author is the speaker)

1. Rodrigue FOKOUOP
Computational and Data Sciences Lab, Operational Research team, Air Liquide R&D

Abstract

This study tackles an inventory control challenge in a complex multi-echelon distribution system, drawing insights from Air Liquide's closed-loop supply chain with product returns. The inherent complexity arises from the distribution structure and the handling of returned cylinders. The goal is to frame this as a multi-echelon periodic review inventory control problem and identify the optimal control parameter.

Conventional stochastic programming approaches prove impractical due to the curse of dimensionality resulting from unmet customer demand and potential lost sales. In response, we introduce a heuristic approach, validated through simulation-optimization. This heuristic involves breaking down the multi-echelon distribution system into serial systems, optimizing target stock levels for each echelon, and aligning cumulative backorder quantities by aggregating shared echelon stock levels. The heuristic achieves a 0.74% average reduction in total inventory cost, demonstrating comparable results to simulation-optimization, with the notable advantage of rapid computation in seconds for practical real-world applications.

Applying the heuristic to real-world data enhances inventory control in the multi-echelon distribution system, leading to a 12.63% reduction in stock costs for regularly demanded products and a 21.89% reduction for sporadic demand. These findings highlight the benefits of a global inventory control approach in a multi-echelon distribution system.

Keywords

Status: accepted

Back to the list of papers