Abstract DGP2026-71

ExoMars Drive-Path Optimization Algorithms for the Rosalind Franklin Rover

Lucas Fernandez (1), Elena A. Favaro (2,3), Sam Fayolle (2), Alexander Barrett (3), Matt R. Balme (3), Peter Fawdon (3), Jack Wright (4), Luc Joudrier (2)

(1) Institute of Geological Sciences, Freie Universität Berlin, Berlin, Germany; (2) European Space Agency, European Space Research and Technology Centre, Noordwijk, Netherlands; (3) School of Physical Sciences, The Open University, Milton Keynes, UK; (4) Institute for Space, University of Leicester, Space Park Leicester, UK.

The ExoMars Rosalind Franklin rover will explore Oxia Planum to investigate ancient habitable environments and search for the chemical building blocks of life [1]. The mission’s success depends not only on the rover’s scientific payload, but also on the ability to safely and efficiently traverse the Martian terrain while preserving time and energy margins. Oxia Planum has been extensively characterized using orbital datasets including high-resolution digital elevation models (DEMs; e.g. [2]), high fidelity geologic [3] and mineralogic mapping [e.g. 4], and landscape classifications from machine-learning outputs of the Novelty and Anomaly Hunter - HiRISE (NOAH-H) [5,6]. However, transforming these geographic information system (GIS) and remote sensing products into datasets specifically tailored to rapidly characterize diverse rover traverse options remains a practical challenge for both strategic planning and mission operations at the surface.

During this pre-launch phase of the ExoMars rover mission, we tested whether rover traverses between two arbitrary points at Oxia Planum can be generated automatically in a GIS using Python-based algorithms. Our work was guided by three operational questions: (I) what is the safest path from point A to point B, (II) how quickly can we traverse that distance, and (III) what is the expected power demand along the route?

The algorithms combine topography from HiRISE DEMs, geomorphologic terrain information from NOAH-H [5,6] at Oxia Planum and Jezero Crater, traverse-time and power proxies, and spatially resolved solar radiation to enable energy-aware planning for solar-powered operations. Adjustable layer weights allow rapid re-optimization for different mission priorities (e.g., safety-first, timeline-driven, energy-aware, or balanced multi-objective routing) without the need to rebuild the data stack, which will be critical once in situ constraints and observations become available.

We validated the operational relevance of our work by comparing our algorithm-generated routes in Jezero Crater against actual traverse segments of the Perseverance rover. Our algorithm, which considered slope and terrain, generated a traverse that was nearly identical to the actual Perseverance traverse path. This result supports the framework’s ability to accurately model relevant traverse paths.

At Oxia Planum, we generated over 30 route configurations spanning single- and multi-objective optimizations. Time-focused solutions can markedly reduce traverse duration (sometimes at the expense of safety constraints), whereas constrained multi-objective routes may maintain safety thresholds while improving operational efficiency and lowering modeled energy demand. We continue to iterate on and streamline our process, introducing new operational considerations such as “Waiting Zones”, which identify safe, high-irradiance locations where the rover could be preferentially parked to take advantage of incoming solar radiation. Overall, our work bridges orbital site characterization and rover operations by delivering mission-ready traverse products and a transferable methodology for future rover missions on planetary surfaces.

[1] Vago et al. (2017) Astrobiology 17(6–7). [2] Volat et al. (2022) PSS 222. [3] Fawdon et al. (2024) Journal of Maps 20(1). [4] Bowen et al. (2022) PSS 214. [5] Barrett et al. (2022) Icarus 371. [6] Barrett et al. (2023) Journal of Maps 19(1).