Determination of Local Goal for a Mobile Robot with Sporadic Human Commands of Tele-operation

Abstract

In this study, we propose a probabilistic approach to overcome time delay for a mobile robot. The robot receives sporadic user input when using brain-computer interface, and related studies have been considered about various strategies to control the robot as intended. However, the most of experiments heavily depended on the driving environment because of the low information transfer rate. To solve this problem, we propose a probabilistic local goal determination method. Our approach accumulates user input sequence, and specifies it as a feasible path. Simulations in the room and on the track are provided. As a result, we show that the proposed method is robust to long time delay, and achieves the mission with fewer inputs.