Inventory management in a closed-loop supply chain with advance demand information

We study inventory control for rental operations in a closed-loop supply chain. In such a system (e.g., Netflix), customers create online queues in a service provider's website to indicate the items that they would like to rent. Leveraging this advance demand information, we propose effective forecast models and formulate a multi-item inventory control problem. We prove that the (L,U) policy is optimal in our multi-item setting. We also consider an aggregate service level constraint across all items and propose heuristics.