A heuristic approach for multi-echelon inventory optimisation in a closed-loop supply chain

Abstract

This study deals with a closed-loop supply chain where inventory levels are controlled by an order-up-to inventory policy. The system under consideration is the cylinder-packaged gas supply chain of Air Liquide company, where empty cylinders used by customers are returned to be filled again by company plants. First, we examine the goodness of fit for demand distributions based on company real data. This enables us to better characterise demands pertaining to different classes of products. Then, we formulate the multi-echelon serial inventory model to be optimised and propose a heuristic to compute the target inventory levels that helps in achieving the desired customer service level while minimising the total inventory cost. The proposed heuristic is easy to implement in the field and gives results close those obtained using a simulation-optimisation approach that is more time-consuming. Finally, we perform a numerical analysis based on company real data and compare several methods that can be used to compute the target inventory levels by varying mainly two assumptions: parameters regarding demand distributions and metrics used to assess customer service levels.

Keywords:

• Closed-loop supply chain
• multi-echelon inventory control
• target stock calculation
• sustainable supply chain
• simulation-optimization approach

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data supporting this study's findings are available from the Air Liquide company. Restrictions apply to the availability of these data, which were used under license for this study. Data are available from the authors with the permission of Air Liquide company.

Additional information

Notes on contributors

Rodrigue Fokouop

Rodrigue Fokouop is an OperationsResearch Scientist at Air Liquide's Computational & Data Science Lab. He holds a diploma in industrial engineering (2014). He is currently doing his PhD at CentraleSupelec. He works on developing models and solutions adapted to industrial cases, the use of which will allow improving inventory management practices and optimisation.

Evren Sahin

Evren Sahin is a Full Professor of Supply Chain and Operations Management at CentraleSupelec. She is the Academic Director of the Supply Chain & Operations Management Program at CentraleSupelec. She received her doctoral degree from Centrale Paris on the impacts of RFID on supply chains, a joint project with the Auto ID Center of MIT. Her primary research interests are supply chain management, flow and inventory control under uncertainty and healthcare management. She also serves as expert for the French Research Ministry and is member of several research committees.

Zied Jemai

Zied Jemai is a full professor at Ecole Nationale d'Ingénieurs de Tunis where he is the director of the laboratory research OASIS. He is also a research associate at CentraleSupelec and responsible for the research activities of the chair Supply Chain. He holds a PhD degree in Industrial Engineering from CentraleSupelec and a Diploma degree in Industrial Engineering from ENIT. His research interests are in the fields of supply chain management, inventory control and stochastic models, with a special emphasis on modelling and optimisation.

Yves Dallery

Yves Dallery In Memoriam is a Professor of Supply Chain at CentraleSupelec. He is the Director of the Supply Chain Chair where his team works with large size companies of various sectors on innovation in supply chains. In the past, he has held positions at MIT, Boston University and HEC. Yves is also at partner at Diagma, a consulting company specialising in supply chain. Yves's main areas of expertise are in supply chain strategy, supply chain planning and flow management, agility in supply chains.