Enhanced Autonomous Navigation on the Perseverance Mars Rover

This paper presents the \textit{Enhanced Autonomous Navigation}, or \textit{ENav}, the autonomous driving algorithm of NASA's Mars rover \textit{Perseverance}. % Since its landing, ENav has been successfully driving the rover on the red planet. A unique challenge for the autonomous driving of \textit{Perseverance} is to meet strict safety and performance requirements in a highly uncertain environment only with a single-core CPU with extremely limited computing resources. ENav overcame this challenge with a novel two-stage path selection approach that balances the path optimality and computational efficiency, combined with a unique collision checking algorithm that conservatively approximates computationally expensive kinematic settling \cite{ace}. In addition, ENav provides robustness against slip by expanding the bounding boxes for wheels used by the collision check. These new features, together with FPGA-accelerated vision processing, enabled \textit{Perseverance} to autonomously drive on substantially more rock-dense terrains and increased by an order of magnitude the average distance the rover can drive each day, compared to its predecessors, the \textit{Curiosity}, \textit{Spirit}, and \textit{Opportunity} rovers. \textit{Perseverance} has set several new records for autonomous driving on Mars, breaking those previously held by the \textit{Opportunity} rover. E.g., as of the 1312th Martian day since landing, or 28 October 2024 on the Earth calendar, 90\% of all 32.1 km of driving has used ENav to evaluate the terrain. This paper presents the algorithm, implementation, and testing of ENav, as well as its performance on Mars.