Order consolidation in warehouses: The loop sorter scheduling problem

To meet today’s ambitious order throughput targets, many distribution centers, especially those operated by online retailers, apply batching and zoning in their picker-to-parts warehouses. These order retrieval policies improve the pick density per tour by unifying multiple customer orders to larger pick lists and allow a parallelization of the picking process among multiple zones, respectively. The price for this is an additional consolidation stage, where picked products must be sorted according to customer orders, typically with the help of a sortation conveyor. In this context, we treat the loop sorter scheduling problem, which is defined as follows. Once a wave of orders, picked concurrently in multiple zones, has been inducted onto a closed-loop sorter, we have to assign the products that refer to the same stock keeping units (SKUs) to orders and orders to packing lanes, where they are prepared for shipping. Furthermore, we have to decide on the sequence in which the orders are channeled into their packing lanes. Our aim is to minimize the makespan until all orders of the current wave are readily sorted. We formulate the loop sorter scheduling problem, investigate computational complexity, and derive suitable solution algorithms. One important finding of our computational study is that simple priority rules, which are frequently applied in real-world warehouses and previous research, waste significant optimization potential.