Use of dynamic simulation to analyze storage and retrieval strategies

Abstract

In the second half of 1998, shipment volumes at one of Intel's warehouses had increased beyond the storage and retrieval capabilities of the facility. An engineering improvement team began studying changes to the Warehouse Management System (WMS) that would increase throughput. From observation it was unclear what WMS code changes would actually improve throughput, and nearly impossible to predict the amount of improvement that would be realized in the facility. To solve these issues, the algorithms for storing product, releasing orders, and routing vehicles were first analyzed in a dynamic simulation model. Strategies that showed a significant increase in throughput were recommended for coding into the WMS software. Using a simulation model not only allowed the strategies to be prioritized, but also predicted the performance of each strategy. The equipment and physical layout of the facility were comprehended in the simulation model. The storage area consisted of twelve aisles, each 112 bins long and 16 bins high. Product was stored in boxes, which were retrieved and stored by operators driving Stockpicker vehicles. Since both the storage and retrieval of material were entirely controlled by the WMS, it was imperative that a logical routing decision for each Stockpicker vehicle be made. The initial storage and retrieval strategies were first coded in the simulation model to ensure that the model outputs were valid. Then, numerous storage and retrieval strategies were coded and analyzed to determine which ones would increase throughput. The final simulation results showed that throughput could be increased by 110% per day by simply improving the WMS storage and retrieval strategies. No additional vehicles or headcount were required which resulted in a significant annual cost savings.