Avoiding Moving Persons by Using Simple Trajectory Prediction and Spatio Temporal Planning

Abstract. If a mobile robot operates within its environment, it should take other persons into account while moving around. This work presents an approach, which predicts the movements of persons in a very simple way, and uses the predicted trajectories to plan a motion path for the robot. The presented motion prediction and planning process is much faster than real time. A potential field is applied to predict the person’s movement trajectory, and a modified Fast Marching planner is used for the planning process. The aim of this work is, to create an early avoid-ing behavior of the robot, when the robot passes a person, to signal a ”busy”’-behavior towards the person.