Artificial intelligence helps scientists identify minerals present in rocks studied by the Perseverance rover.
Some scientists dream of exploring planets with “smart” spacecraft that know exactly what data to look for, where to find it, and how to analyze it. While that dream will take time to realize, the progress made by NASA’s Perseverance Mars rover is a promising step in that direction.
For nearly three years, the rover mission has been testing a form of artificial intelligence capable of searching for minerals in the rocks of the Red Planet. This is the first time that AI has been used on Mars to make autonomous decisions based on real-time analysis of the composition of the rocks.
The software supports PIXL (Planetary Instrument for X-ray Lithochemistry), a spectrometer developed by NASA’s Jet Propulsion Laboratory in Southern California. By mapping the chemical composition of minerals on a rock’s surface, PIXL allows scientists to determine whether the rock formed under conditions that could have supported microbial life in Mars’ ancient past.
Called “adaptive sampling,” the software autonomously positions the instrument near a rock target and then examines PIXL’s scans of the target to find minerals worth examining further. All of this happens in real time, without the rover communicating with mission controllers on Earth.
“We use PIXL’s AI to focus on key science,” said the instrument’s principal investigator, Abigail Allwood of JPL. “Without it, you would see an interesting clue in the data and then have to reanalyze the rock to study it in more detail. This allows PIXL to reach a conclusion without humans looking at the data.”
Data from Perseverance’s instruments, including PIXL, help scientists determine when to drill a rock core and seal it in a titanium metal tube so it, along with other high-priority samples, can be returned to Earth for further study as part of NASA’s Mars Sample Return campaign.
Adaptive sampling isn’t the only application of AI on Mars. About 2,300 miles (3,700 kilometers) away from Perseverance is NASA’s Curiosity, which has developed a form of AI that allows the rover to autonomously blast rocks with a laser based on their shape and color. Studying the gas that burns after each laser strike reveals a rock’s chemical composition. Perseverance has that same capability, plus a more advanced form of AI that lets it navigate without a specific direction from Earth. Both rovers still rely on dozens of engineers and scientists to plan the hundreds of individual commands each day, but this digital intelligence allows both missions to accomplish more in less time.
“The idea behind PIXL’s adaptive sampling is to help scientists find the needle in a haystack of data, freeing up time and energy to focus on other things,” said Peter Lawson, who led the implementation of adaptive sampling before retiring from JPL. “Ultimately, it helps us gather the best science data faster.”
AI helps PIXL in two ways. First, it positions the instrument exactly as it needs to be once it’s near a rock target. Located at the end of Perseverance’s robotic arm, the spectrometer sits on six tiny robotic legs, called hexapods. PIXL’s camera repeatedly checks the distance between the instrument and a rock target to help with positioning.
Temperature variations on Mars are large enough that Perseverance’s arm expands and contracts microscopically, which can disrupt PIXL’s aim. The hexapod automatically adjusts the instrument to get exceptionally close to the rock without making contact with it.
“We have to make adjustments at the micrometer level to get the accuracy we need,” Allwood said. “The device gets close enough to the rock to make the hair on the back of an engineer’s neck stand up.”
Once PIXL is in position, another AI system gets a chance to shine. PIXL scans a postage-stamp-sized area of a rock, firing an X-ray beam thousands of times to create a grid of microscopic dots. Each dot reveals information about the chemical composition of the minerals present.
Minerals are key to answering key questions about Mars. Depending on the rock, scientists might be looking for carbonates, which hold clues about how water may have formed the rock, or phosphates, which could have provided nutrients for microbes, if any existed in the Martian past.
Scientists have no way of knowing in advance which of the hundreds of X-rays will reveal a particular mineral, but when the instrument finds certain minerals, it can automatically pause to collect more data, an action called “long dwell.” As the system improves through machine learning, the list of minerals that PIXL can focus on with a long dwell grows.
“PIXL is kind of a Swiss Army knife in that it can be configured based on what scientists are looking for at any given moment,” says David Thompson of JPL, who helped develop the software. “Mars is a great place to test AI because we have regular communications every day, which gives us the ability to make adjustments along the way.”
As future missions travel deeper into the solar system, they will be out of contact for longer than current Mars missions. That is why there is a strong interest in developing greater autonomy for missions that roam and conduct scientific research for the benefit of humanity.
One of the primary goals of the Perseverance mission to Mars is astrobiology, including searching for signs of ancient microbial life. The rover will characterize the planet’s past geology and climate, pave the way for human exploration of the Red Planet, and be the first mission to collect and cache Martian rocks and regolith (broken rocks and dust).
Subsequent NASA missions, in cooperation with ESA (European Space Agency), would send spacecraft to Mars to collect these sealed samples from the surface and bring them back to Earth for further analysis.
The Mars 2020 Perseverance mission is part of NASA’s Moon-to-Mars exploration approach, which includes the Artemis missions to the Moon that will help prepare for human exploration of the Red Planet.
JPL, managed for NASA by Caltech in Pasadena, California, built and operates the Perseverance rover.
To learn more about perseverance:
mars.nasa.gov/mars2020/
Media Contacts
Andre Bon
Jet Propulsion Laboratory, Pasadena, California.
818-393-2433
andrew.c.good@jpl.nasa.gov
Karen Fox / Alana Johnson
NASA Headquarters, Washington
202-358-1600 / 202-358-1501
karen.c.fox@nasa.gov / alana.r.johnson@nasa.gov
2024-099