Robot Collaboration Techniques in Automated Warehouses

Abstract: As e-commerce grows, the need for efficient warehouses increase. One way of increasing warehouse efficiency is using item picking robots that collect all the items for the orders. This study expands on a study by Pinkam etal, which compares two different heuristics for the order picking strategies of these robots. Either the robots collaborate, or they work independently on separate orders. Our study repeats this comparison, with the addition of a new collaborative strategy. Further, we introduce an exhaustive search that finds the optimal movement for all the robots, and compare the three heuristics to this baseline. The comparison is performed by running simulations with a randomly generated queue of orders for the warehouse. Efficiency is measured by the time to completion of all orders. The new heuristic is found to have similar efficiency to the collaborative algorithm, while the non collaborative algorithm is significantly worse.