Path planning algorithm for a South Pole lunar rover mission

Abstract

On the Moon, a rover needs to navigate around physical obstacles related to the topography such as boulders or steep slopes, to reach targets of scientific interest. In addition, it needs to avoid shadowed areas (which vary during the day), which prevent the rover from having access to the Sun for energy or to the Earth for communication. The combination of changing illumination and communication, as well as the rugged terrain, make the path planning quite complex. Moreover, the distance between the Moon and Earth causes a latency for commands to reach the rover. This makes it difficult to plan the rover’s actions, especially with many areas to explore. To tackle this problem, we propose a novel mission planner tool with two components: we first use a two-step Genetic Algorithm to compute a tentative order of exploration from a list of points of interest, allowing us to associate a time stamp for each explored waypoint. We then compute feasible trajectories between the ordered waypoints while ensuring that the rover is avoiding all static obstacles, staying in contact with Earth, and being powered by solar illumination. The simulation results show that this tool works well for different lunar sites and substantially reduces the workload for manual mission planning.