Extravehicular activity (EVA) is a highly hazardous phase of human spaceflight, yet one of the most scientifically rewarding. Ensuring the safety, efficiency, and autonomy of astronauts is essential for the success of Artemis mission EVAs. Prior work has developed automated EVA planning tools, but none incorporate monitoring of physiological data. The Navigational Ground Assessment & Trajectory Optimization Router (NaviGATOR) is designed to process surface environment data, physiological data, consumables levels, and preselected scientific regions of interest (ROI) to determine the EVA path that minimizes astronaut energy expenditure and maximizes scientific return. First, a grid-based representation of the lunar surface is constructed from slope data. Metabolic rate (MR) is predicted for each grid cell using slope data, locomotion speed, and mass of the astronaut-suit system. Grid cells with slopes that exceed allowable ranges as defined in Exploration EVA System's Concept of Operations are constrained as unusable. All other grid cells are assigned a normalized cost based on MR predictions. Once the coordinates of start points, end points, and ROIs are provided, the tool uses mixed integer linear programming to determine the optimal ordering of ROI visits. Optimal Rapidly-exploring Random Trees planpaths between ROIs to minimize energy expenditure while satisfying constraints. The tool can plan paths both before an EVA and adjust paths during EVA based on astronaut vital signs and suit consumable levels, thus facilitating better-informed mission planning and improving crew autonomy. The tool's performance was validated through comparison to Apollo EVAs, which showed that the tool orders ROIs identically to visitation order by Apollo astronauts, plans paths between ROIs similar to those the Apollo astronauts traversed, and predicts MRs comparable to Apollo mission-recorded data. Overall, this tool may be able to improve safety, efficiency, and autonomy of astronauts on surface EVAs. NaviGATOR is maintained on a public GitHub repository.
Richardson et al. (Sun,) studied this question.