A probabilistic approach to path planning with object boundary uncertainties

Abstract

In the presence of geometric uncertainties, collision avoidance of the robots in close proximity to or in contact with the obstacles in the environment is discussed from the probabilistic point of view. A hierarchical representation for geometric uncertainties is proposed; spatial uncertainties are described by occupation probability. An algorithm which generates robot trajectories by calculating collision probability between the robot and the obstacles is presented to show the applicability of the proposed framework to path planning.<>