Design and Testing of an Integrated Robot for Harvesting, Stipe-Cutting, and Grading of Agaricus bisporus.

To address the limitations of current Agaricus bisporus harvesting robots, including low picking efficiency, susceptibility to mechanical damage, poor operational stability, and discontinuous harvesting processes, an integrated robotic system capable of mushroom detection and localization, picking sequence planning, stipe-cutting, and grading was developed. The robot adopts a left-right symmetrical configuration and operates along the side of the mushroom cultivation racks. It mainly consists of a mobile platform, a lifting mechanism, dual picking units, a receiving unit, a stipe-cutting device, a collection system, and an electronic control system. A picking sequence planning method based on YOLOv8n-USD and KD-tree nearest neighbor search was employed to determine the optimal harvesting order for densely clustered and adhered mushrooms. The dual robotic arms executed the picking operations, after which the mushrooms were transferred to the receiving unit for stipe-cutting and subsequently classified according to the detection results, thereby completing the integrated process of detection, picking, stipe-cutting, and grading. Experimental results show that the average detection accuracy reaches 96.64%, the picking success rate is 95.39%, and the harvesting damage rate is 1.63%, while the average interaction failure rate and grading error rate are 0.35% and 1.28%, respectively. These results indicate that the proposed robot can achieve accurate detection and efficient integrated operation for densely clustered Agaricus bisporus. The findings provide a reference for flexible and efficient harvesting of Agaricus bisporus with synchronized stipe-cutting and grading operations.