Article ID: 17-00364
Human motion is very flexible for performing various tasks but has low speed and low precision; therefore, support and assistance of human motion by robots is desirable in some situations. In this study, in order to achieve such functions, we developed a new portable module for accurately controlling the position of a human hand and constructed a high-speed, high-accuracy positioning control system using image tracking via a high-speed vision system. In order to evaluate its performance, we executed tracing tests of constant position, circular or linear trajectory. Finally, we performed the task of catching a falling small ball, as a significantly difficult positioning task. Although this task was nearly impossible to perform with the human hand alone, the success rate was dramatically improved by using the proposed method.