Unloading sequence planning for autonomous robotic container-unloading system using A-star search algorithm

Abstract

In autonomous unloading systems, one of the major challenges is planning the unloading sequence in cluttered logistics container environments. Proper unloading sequence planning is crucial to avoid damage to the packages caused by collision and collapse. At the same time, the planned sequence has to minimize the effort during the unloading process, such as energy consumption, time taken, and distance moved, to enhance efficiency. This paper presents a sequence planning method based on the A-star search algorithm applied to unloading randomly stacked unidentical boxes from logistics containers. To represent the general overlap relationships of boxes in a cluttered environment, we utilize an adjacency matrix structure and involve it in the state of the unloading sequence planning. Moreover, by adopting the A-star search algorithm, the method can minimize the cost of system movement for an efficient unloading sequence. As a result, our method can determine the unloading sequence while considering the relationships between packages. Simultaneously, it achieves better performance compared to the existing sequence planning method in terms of moving distance, both in structured and unstructured cases. Finally, we construct a robotic unloading system and present experimental results in the simulated container environment.