A DESIGN METHODOLOGY FOR CONVERTING A REGULAR WAREHOUSE INTO A ZONE-PICKING WAREHOUSE

In this paper, we study the problem of converting a regular warehouse into a zone-picking warehouse. A designed methodology that divides this problem into two sub-problems is proposed. The first sub-problem is to determine the items in each zone. A clustering procedure consisting of a similarity coefficient measure and a clustering algorithm is proposed to solve this problem. Six similarity coefficients and two clustering algorithms are proposed and studied. The second sub-problem is to determine the storage locations of the items in each zone. For this problem, two storage-location assignment rules are proposed and studied. Since it is assumed that order batching is adopted in the warehouse, two order-batching methods which are different seed-order selection rules are considered in this study. Furthermore, two route-planning methods are investigated. Totally, there are 96 combinations of these methods. Each combination is tested by thirty randomly generated test problems. The testing results are collected and analyzed to understand not only each method's total order-picking travel distance performance, but also their mutual effects. Finally, we also solve the storage-location assignment problem for each test problem without dividing the warehouse into different zones, so that we can see how much improvement can be achieved by converting a regular warehouse into a zone-picking warehouse.