DETERMINING EFFICIENT ROVER SCIENCE PROTOCOLS FOR ROBOTIC SAMPLE SELECTION: A GEOHEURISTIC OPERATIONAL STRATEGIES TEST IN GREATER CANYONLANDS, UTAH, US

In a recent field test at a potential Mars 2020 landing site analog, the GeoHeuristic Operational Strategies (GHOST) team tested two methods for data acquisition and science decision-making: (1) a “linear” scenario, where all sites are examined as they are encountered; and (2) a “walkabout-first” approach in which a region of interest is traversed multiple times, the first time to gain context and downselect targets for more detailed study in future loops. Our field site was a 1 km² area in Greater Canyonlands, Utah, USA. Local units contain micro and macroscopic biosignatures. Lacustrine deposits, travertine deposits, and exhumed curvilinear fluvial channels are common, similar to features documented on Mars.

GHOST adopts a “roverless roving” approach, in which we use a generalized suite of commercial, off-the-shelf instruments that provides visual, compositional and geochemical data similar to flight-ready instruments (digital SLR camera with macro lens, handheld spectrometer, field XRD). Humans provide mobility and run the instruments. This method is ideal for testing science decision-making protocols, as science requires as input only the data gathered by those instruments. Based on average resource availability for prior Mars missions, we assumed ~1 hour of active remote data acquisition (imaging, whole-rock multispectral data) and one choice of either a drive (50-100 m) or observations using the instruments mimicking those that contact the surface.

Both methods facilitated geologic characterization and interpretation. However, the walkabout-first method yielded a ~25% savings in sols, taking 56 sols to execute compared to 69 sols required by the linear approach. Additionally, since contextual information was acquired earlier in the process for the walkabout-first approach, team members had more time to discuss results and thus were more confident in their conclusions. By contrast, the team executing the linear approach was under pressure to decide immediately whether or not to sample, leading to less optimal samples being acquired in lieu of additional information. Our results indicate that (1) geologic context, provided as early as possible, will save mission time and resources; and (2) science must be given adequate time to produce robust results.