Design and locomotion analysis of an arm-wheel-track multimodal mobile robot

The increasingly complex application environment has raised higher demands on the performance of ground mobile robots in terms of environmental adaptability, autonomous avoidance, and self-rescue. In addition to multi-sensor fusion and control strategies, novel locomotion systems are crucial research directions. Here we propose a novel hybrid locomotion ground mobile robot, called arm-wheel-track robot (AWTR). It combines two locomotion systems, wheeled and tracked locomotion. Two multiple-degree-of-freedom arms are mounted on the front and rear of its chassis. The arms can assist the robot in transforming locomotion modes, surmounting obstacles, fall recovery, etc. Two ultrasonic sensors and a tilt sensor are mounted on it to perceive the environment and self-posture. One of the ultrasonic sensors mounted on the forearm can achieve a more comprehensive perception of the environment ahead with the extra workspace provided by the forearm. We establish the relationship of terrains with sensor data and forearm posture and develop different locomotion strategies for different terrains, so that the robot can classify different terrain and accomplish the corresponding locomotion strategies autonomously. We have built a prototype and conducted experiments on different terrains. The results verified the robot’s movement performance, the effectiveness of the terrain perception method and the locomotion strategies for different terrains.