Order Picking In A-3-D Warehouse Using Clustering Approach : A Case Study Of Gemadept

Abstract

In warehouse operations, it is believed that order picking is a time-consuming, labor intensive and expensive activity resulting in possible loss of profit for the company. As a result, efficiency in picking process has always been a subject of continuous enhancement in the eye of thriving warehouse management. In this case study, a sector of an actual 3-D warehouse is investigated, this means that the levels of rack have been accounted for as the height parameter. The layout is of the typical one with single-block and sizeable aisles width allowing a forklift to be able to make a U-turn thereby making them bidirectional. The picking process is treated as a capacitated vehicle routing problem. The purpose is to minimize the overall travel distance. Since it takes a considerable amount of time for Mixed Integer Programming to handle big instances making it impractical, we turn to the clustering approach based on known heuristics and clustering technique through machine learning. The algorithm making it possible to reach near optimality in no time, discovering realistic solutions that might possibly be put into actual warehouses to minimize order picking time and consequently, total warehouse expenses.