Abstract
In this study, we improved a human coexistence feature for robotic vacuum cleaner Roomba through use of movement discrimination function by environmental camera. Specifically, the environmental camera recognizes movement of a human (No Human, Walking, and Staying) by using Cubic Higher-order Local Auto-Correlation. Then Roomba gets the movement information, and moves in accordance with it. In the case of No Human, Roomba continues cleaning. If the movement information is Walking, Roomba stops. When Roomba gets the information of Staying, Roomba avoids the human while cleaning by using potential field. We implemented this feature by using OpenRTM-aist. In order to confirm effectiveness of proposed function, we conduct experiment.
