Evaluation of Collaboration Policy in a Warehousing System With Heterogeneous Robots

Abstract

ABSTRACT This study investigates a novel collaborative warehousing system that employs transportation robots and retrieving robots for order fulfillment. The system features a central platform with peripheral racks and workstations located at the ends. Retrieving robots transfer items from racks to transportation robots, which then deliver them to workstations for processing orders. We develop a fork–join queuing network with two semi‐open queuing networks as its branches to capture the collaboration between the two types of robots and the synchronization of robots and orders, thereby enabling an accurate representation of the overall system performance. An approximation algorithm is designed to solve the queuing network model, and its accuracy is validated through simulation. We analyze several key collaboration policies, including dwell point policies, the optimal transportation‐to‐retrieving robot ratio, and the location of battery charging zones. The study yields several managerial insights: (1) The effectiveness of dwell point policies depends on warehouse size: The point of service competition policy outperforms in small layouts, whereas the point of designated position policy excels in large systems. (2) The optimal location of the battery charging zone varies with warehouse dimensions and order arrival rates. (3) Sequentially operating homogeneous robotic systems show advantages in high‐demand scenarios and with small fleets, whereas parallel operating heterogeneous robotic systems perform better under moderate demand with large robot populations. Beyond a critical fleet‐size threshold, heterogeneous robotic systems consistently dominate across all demand levels. (4) Furthermore, based on our model, we determine the optimal ratio of transportation robots to retrieving robots under different demand scenarios and operational policies.