Adaptive Sampling with PIXL on the Mars Perseverance Rover

Planetary rovers can use onboard data analysis to adapt their measurementplan on the fly, improving the science value of data collected between commandsfrom Earth. This paper describes the implementation of an adaptive samplingalgorithm used by PIXL, the X-ray fluorescence spectrometer of the Mars 2020Perseverance rover. PIXL is deployed using the rover arm to measure X-rayspectra of rocks with a scan density of several thousand points over an area oftypically 5 x 7 mm. The adaptive sampling algorithm is programmed to recognizepoints of interest and to increase the signal-to-noise ratio at those locationsby performing longer integrations. Two approaches are used to formulate thesampling rules based on past quantification data: 1) Expressions that isolateparticular regions within a ternary compositional diagram, and 2) Machinelearning rules that threshold for a high weight percent of particularcompounds. The design of the rulesets are outlined and the performance of thealgorithm is quantified using measurements from the surface of Mars. To ourknowledge, PIXL's adaptive sampling represents the first autonomousdecision-making based on real-time compositional analysis by a spacecraft onthe surface of another planet.