Path planning based on inflated medial axis and probabilistic roadmap for duct environment

Abstract

This paper proposes an inflated medial axis based probabilistic roadmap (PRM) for the duct environment. The study intends to explore a sampling strategy for PRM toward clearance-driven path planning. Sampled nodes are constrained only inside the inflated medial axis area. The medial axis is generated from skeletonizing the map and inflate the skeletal diagram. To evaluate effectiveness of the proposed method, experiments were carried out on a reference map generated by simultaneous localization and mapping (SLAM) from an actual mockup duct environment. Experiments conducted showed the proposed method able to produce a path trajectory with clearance-driven behavior. The method is very useful in narrow spaces such as ducting, which required a safe distance when navigating autonomously in the environment.