Cube-Based Automated Storage and Retrieval Systems: A Multi-Agent Path Finding Approach

Yi Zheng, Yimin Tang, T. K. Satish Kumar, Sven Koenig

Keywords: Warehouse Robotics, Cube-Based Automated Storage Retrieval Systems, Multi-Agent Path Finding

Abstract: Automated Storage and Retrieval Systems (AS/RS) require the coordination of multiple agents in a common warehouse environment. They often rely on Multi-Agent Path Finding (MAPF) technologies for the path-coordination of agents that transport inventory pods. In this paper, we study the more recent cube-based AS/RS and the relevance of MAPF technologies to them. In these systems, storage bins are stacked on top of each other in a tight 3D grid environment without aisles. Robots operate only at the top level to retrieve and organize the storage bins. While cube-based AS/RS maximize the storage density, they require the robots to "dig'' out the storage bins whenever necessary. Despite their advantages and recent adoption, limited research has been reported on the relevance of MAPF technologies to them. In this paper, we adopt novel MAPF perspectives on the cube-based AS/RS and propose two approaches for planning in such domains. In the first approach, the robots are treated as the agents: Plans are generated for them using 'move' and 'dig' actions. In the second approach, there are two phases. In the first phase, the storage bins are treated as the agents: Plans are generated for them to move autonomously. In the second phase, the robots are treated as the agents: Plans are generated for them to chaperone the movements of the storage bins. We demonstrate the value of MAPF technologies in cube-based AS/RS and experimentally show that the second approach marginally outperforms the first.

Email Sharing: We authorize the sharing of all author emails with Program Chairs.

Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.

Submission Number: 33