Semi-Autonomous Robotic Avatar System for Patients with Motor Disabilities

This paper presents a method of semiautonomous navigation in a robotic avatar system for patients with physical disabilities. Gaze tracking system is used as an effective handsfree teleoperation interface. However, in the manual control, the operator needs to continuously control the robot by using eye movement. We therefore propose a method of environmental modeling and path planning to realize the adaptive semi-autonomous navigation in unknown environment. Moreover, we developed an intelligent controller dealing with allocation of the tasks, such as target tracking and collision avoidance. Furthermore, we show an experimental example to discuss the applicability of the system