Determining Warehouse Storage Location Assignments Using Clustering Analysis

The methodology used to assign products to a storage location in a warehouse can have a significant impact on the amount of time required to retrieve all of the items needed to fill an order. This paper describes a methodology that uses a clustering approach to determine storage assignments, where the metric of the strength of the relationship between two stock-keeping units (SKUs) is the number of times that the SKUs appear in the same order. Clustering is performed to maximize the frequency with which SKUs in the same cluster are ordered together. In testing, the clustering assignments were compared to a demand-based assignment strategy and showed a reduction of 20-30% in the number of aisles visited to retrieve orders.