Online turnover-free control for a mobile agent with a terrain prediction sensor

This paper proposes a turnover-free control method for a teleoperated mobile agent (or vehicle) moving through uneven terrain. The teleoperated agent is primarily driven by an operator at a remote site and is able to react autonomously when a possible turnover is predicted. In order to predict the turnover, a low-cost terrain prediction sensor has been developed using a camera vision with a structured laser light. Since it is difficult for an operator to predict the reactive motion of the agent, a force reflection technique with a force feedback joystick is employed to intuitively recognize the inconsistency between the intended motion and the reactive motion of the agent. Finally, to verify the feasibility and effectiveness of the proposed method, experiments with the ROBHAZ-DT (actual mobile agent) have been carried out. In the experiments, the operator could recognize the reactive motion of the agent for turnover prevention through force reflection while the agent was moving on slopped terrain.