Multi-Echelon inventory optimization

Abstract

Traditionally, inventory optimization identifies optimal inventory at the individual levels of the supply chain without considering the dependencies between different levels. So, inventory at an echelon is determined by taking the demand and its variation from the immediate downstream echelon. In this case, each echelon is responsible for its fill rate, resulting in high safety stock being maintained at each level. The phenomenon of bullwhip effect explains the increased fluctuations in demand as one move from the retailer level to the manufacturer level due to traditional method of inventory optimization in silos. For example, if a supply chain contains the suppliers, manufacturer, warehouses and retailer, the inventory would be managed at each of the levels separately i.e. inventory at warehouse would consider the demand from immediate downstream echelon (retailer) and not the point-of-sales data. This approach, though optimizes the inventory at each echelon, doesn’t optimize the inventory for the complete supply chain. So, there is potential of reducing the inventory costs by centrally managing the inventory at the different levels which should be accompanied by information sharing across multiple echelons.