A heuristic procedure for a dynamic inventory routing problem
Date:
August 22-26, 2008. Co-authored with Gutiérrez J, Abdul-Jalbar B and Sicilia J. Book of Abstracts, page 100.
Abstract
We propose a Logistic Network where a central warehouse (vendor) distributes a single commodity to multiple geographically dispersed retailers (buyers) through a finite planning horizon consisting of T periods. Each retailer faces a deterministic and time-varying demand, which should be fulfilled at the beginning of each period and influences the replenishment policy at the warehouse within a Vendor Managed Inventory (VMI) framework. Furthermore, the maximum inventory level at each retailer is limited due to local storage constraints and the batches are delivered by a single capacitated vehicle. We also assume that the total cost function involves the following components: a unit holding cost per period and location, a transportation cost between each pair of distinct locations and a fixed shipment cost per period. When the fixed shipment cost is activated in one period, a unique route should be stated to distribute the corresponding delivery quantities to a subset of retailers. The model so stated can be seen as an extension of the classical Inventory Routing Problem to the case with time-varying demand, and hence it can be referred to as the Dynamic Inventory Routing Problem (DIRP). Accordingly, the goal is to determine a compromising inventory/distribution policy which guarantees the demands at each retailer are met on time without exceeding both local storage and vehicle capacities.