Path Planning for Planetary Surface Exploration Using Incremental A-r-Star Pathfinder

This paper addresses rover path planning for wide-area exploration on remote planetary surfaces using the Incremental A-r-Star pathfinder algorithm. Given an initial terrain map derived from satellite imagery or images captured by a lander during descent as well as onboard 3D range sensors, a rover plans a safe and suboptimal path from location to location and through boulder fields while avoiding mobility hazards. Initial paths are based on the knowledge base obtained from the initial map but as the rover navigates and discovers new obstacles that were not initially represented in the map, it re-plans a new path from its current location. This results in reduced computation as compared to re-planning an entirely new path and hence increases rover navigation rate and reduces power consumption.