A Thematic Approach to Robotic Path Planning on the Moon

Abstract

Path planning in remote spaces presents a notable challenge as it requires significant pre-mission time and effort due to the multitude of inputs that must be considered. This challenge is especially difficult on the Moon due to highly complex terrain and sparsity of detailed remote sensing information. Importantly, paths must fulfill engineering, scientific, and operational constraints in order to ensure the successful outcome of the mission. We take a thematic STV (Science - Traverse - Survive) approach to this problem which allows for the optimization of scientific, traversability, and survivability outcomes. STV consists of three overarching themes, the first being the scientific merit or reward (‘Science’), the second being the complexity and difficulty of the terrain (‘Traverse’), and the third being inputs that damage the system and lead to catastrophic failure (‘Survive’). Rather than contend with multiple variables individually, the STV approach works in a way that requires tracking only three primary themes to derive an optimized path that meets the scientific goals of the mission while also remaining within engineering and operational constraints. To implement this approach, lunar remote sensing data sources were surveyed and aggregated from available data repositories, with additional data being derived from these existing datasets using geographic information systems (GIS) processes. Terrain based elements, such as slope, aspect, curvature, and ruggedness were derived from a digital elevation model (DEM). Other datasets, such as albedo and gravitational field strength, were obtained and indicate potential targets of scientific value. Each of these datasets are assessed for their impact on each of the STV themes, weighted, and combined via various distribution functions. GIS software and processes were also employed in the data management strategy. Each data layer is standardized and segmented into a tile-based scheme and subsequently stacked together to promote data accessibility and cooperation. Our method allows for the fine-tuning of weights and thresholds for each dataset, as well as each theme of STV, to produce cost surfaces and visualizations in either a mission planning or real-time context. This approach is demonstrated across a selected portion of the Endurance mission concept path near the South Pole, and includes visualization of the STV themes using lunar remote sensing data and GIS methods.