Realizing an Assist-As-Needed Robotic Dressing Support System through Analysis of Human Movements and Residual Abilities

We propose a framework for a new control method for robotic dressing assistance systems that utilize the residual capabilities of the person being dressed. The proposed control method focuses on the swinging motion of the arm during dressing, and by passively controlling the arm in response to the human’s movement, the robot does not inhibit the human’s movement but only provides the minimum necessary support for dressing. The proposed control method is realized by assuming that the arm-swinging movements of the human during controlled dressing have some regularity and that modeling is possible. By acquiring and analyzing the skeletal information of the person being dressed during the dressing assistance by the robot, the possibility of the proposed method is discussed by proving the assumption, and the future prospects for the realization of a controller that makes use of residual abilities are also discussed.