Automated Guided Vehicles for Material Flow in Fulfillment Centers

Abstract

The eCommerce industry utilizes fulfillment centers for product inventory, order packaging and distribution. The fulfillment process at Amazon has been highly automated within their Amazon Robotics (AR) Sortable facilities, mainly within the inventory and order picking processes. Although there has been significant progress in introducing technology into the fulfillment processes, there are still several opportunities for further integration. This work proposes the integration of automated guided vehicles (AGVs) in Amazon's fulfillment centers (FCs) to improve process efficiency, labor utilization, and improve employee safety. Through utilizing the Six-Sigma DMAIC method, the process path of Transport Support associate was selected as a focus because of the manual labor involved in their role often moving empty material throughout the facility. The improved process path is proposed with integration of AGVs and modeled using a process-based discrete-event simulation framework. The specific hardware and software requirements for an AGV to fit the proposed process path results in a recommendation for a small packet AGV which utilizes LiDAR scanners and vision-based navigation technology. The simulation results indicate that the integration of AGVs in Inbound Stow process can increase individual throughput by 4-6% per shift per associate and reduces total idle time. The results demonstrate the potential for AGVs to improve the productivity of FCs while contributing to reducing potential work-related injuries. The work concludes that AGVs can improve FC operations in the short and long term, with the potential for significant labor cost savings.