Collision-Free Edge-Following Path Planner based on Adaptive Sampling Continuous Collision Detection

Outdoor robots often require exceptional edge-following(EF) mobility to perform tasks like sweeping and delivering goods along borders. However, moving along the edge poses challenges for fast and accurate collision detection and path planning, as the robot needs to approach obstacles while ensuring collision-free. To address these challenges, we propose an Edge-Following Path Planner (EF-Planner) integrating a novel Error-Limited Adaptive Sampling Continuous Collision Detection (ELAS-CCD) method, enhancing mobile robot's capability of following edges and simultaneously avoiding obstacles. The ELAS-CCD establishes a relationship between allowable error and sampling frequency(SF) to achieve the required detection accuracy through inserting discrete footprints. EF-Planner inte-grates ELAS-CCD to formulate multi-objective function, taking into account safety, smoothness, and efficient EF. Experimental results demonstrate that our approach achieves a reduced deviation distance when moving along the edge. Moreover, this method is applicable to other EF robots as well.