主催: 一般社団法人 日本機械学会
会議名: ロボティクス・メカトロニクス 講演会2021
開催日: 2021/06/06 - 2021/06/08
Picking systems for a wide variety of products are required in logistics fields. We have developed a picking robot with a hybrid hand, which consists of a suction and a pinching mechanism. The hybrid hand is an effective means to increase kinds of items that can be grasped. However, there are two problems. First problem is the grasping fails when the appropriate hand cannot be selected for each item. Second problem is when the item in the container cannot be selected properly. Therefore, in this paper, we propose a grasping strategy that selects an appropriate grasping method for each item and determines the priority of grasping for multiple items in the container to be picked. We show that the proposed grasping strategy is effective by testing whether the number of successful grasping patterns increases for dozens of test patterns that define the items to be grasped and their placement conditions.