Mobile robot navigation using recursive motion control

Abstract

As a mobile robot travels towards its goal, sensor input is used to modify the path of the robot to prevent collision with obstacles. In conventional programming, when an obstacle is detected by the sensors, control loops are escaped according to the requirements of the executing navigation task. Thus navigational routines have to be written in a loop form with the necessary exit conditions built in. This paper presents an alternative approach, recursive motion control. Recovery paths around obstacles are planned and executed by recursively invoking the motion control software. The approach allows multiple obstacles to be handled in a consistent manner. On completion of the obstacle avoiding sub-path(s) the level of recursion is dropped to allow resumption of the original motion.