1503. Inventory Management and Assortment Planning in an Apparel Closed-Loop Supply Chain
Invited abstract in session MC-42: Sustainable apparel supply chains, stream Circular & Sustainable Supply Chains.
Monday, 12:30-14:00Room: Newlyn GR.02
Authors (first author is the speaker)
| 1. | Heletje van Staden
|
| Michael Smurfit Graduate Business School, University College Dublin | |
| 2. | Hannah Yee
|
| Technology and Operations Management department, Rotterdam School of Management |
Abstract
We consider an apparel rental platform with a fixed membership pool and non-stationary demand with partial observability. The platform aims to optimize garment purchasing decisions to best meet customer demand while minimizing operational costs. We model the problem as an ambiguous partially observable Markov decision process, which allows us to account for both uncertainty regarding the underlying demand distributions governing the demand and ambiguity regarding the platform members’ demand preferences. We analyze the structure of the optimal policy and use reinforcement learning to approximate the optimal inventory control and assortment decisions. For the instances considered, we find that the algorithm improves on the static order up to policy by 6%, on average. By adjusting the inventory position to changing demand, our policy is able to limit dissatisfaction among members at the expense of holding slightly more inventory. Our work contributes to the closed-loop supply chain literature and can be used to support rental platforms with inventory and assortment procurement decisions given ambiguous and uncertain trends in customer preferences.
Keywords
- Supply Chain Management
- Inventory
- Stochastic Models
Status: accepted
Back to the list of papers