Artificial intelligence is helping scientists identify minerals 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. That dream will take some time to realize, but progress on NASA’s Mars rover Perseverance represents a promising step in that direction.
For nearly three years, the rover mission has been testing artificial intelligence to hunt for minerals in rocks on the Red Planet, but this will be the first time that AI will be used on Mars to make autonomous decisions based on real-time analysis of rock composition.
The software supports pixel PIXL (Planetary X-ray Rock Chemistry Instrument) is a spectrometer developed by NASA’s Jet Propulsion Laboratory in Southern California that maps the chemical composition of minerals on rock surfaces, allowing scientists to determine whether those rocks formed in Mars’ ancient past under conditions that could support microbial life.
Called “adaptive sampling,” the software automatically positions the rover near rock targets and examines PIXL’s target scans to find minerals worth exploring further — all in real time, without the rover needing to communicate with mission controllers on Earth.
“We use PIXL’s AI to zero in on the science that matters,” said JPL’s Abigail Allwood, the instrument’s principal investigator. “Without AI, we might find a hint of something interesting in the data, but we’d have to rescan the rock to find more. This allows PIXL to reach conclusions without humans examining the data.”
Data from Perseverance’s instruments, including PIXL, will help scientists decide when to drill a rock core and seal it in a titanium metal tube. High Priority SamplesNASA Mars Sample Return Campaign.
Adaptive sampling is not the only application of AI on Mars. 2,300 miles (3,700 kilometers) Following Perseverance came NASA’s Curiosity, a rover that pioneered AI to enable the rover to act autonomously. Burning rocks with a laser It will classify rocks based on their shape and color, and the gases that burn off after the laser will reveal the rocks’ chemical composition. In addition to this, Perseverance is equipped with more advanced AI, Navigate Without specific instructions from Earth, both probes rely on dozens of engineers and scientists to plan hundreds of commands each day, but this digital smarts will enable both missions to accomplish more in less time.
“The idea behind PIXL’s adaptive sampling is to help scientists find the needle in the haystack, saving them 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 collect the best science information, faster.”
AI helps PIXL in two ways. First, it helps position the instrument properly once it’s near the rock target. The spectrometer at the end of Perseverance’s robotic arm sits on six small robotic legs called a hexapod. PIXL’s cameras repeatedly check the distance between the instrument and the rock target to help with alignment.
Mars’s large temperature variations and the microscopic expansion and contraction of Perseverance’s arm could throw off PIXL’s aim, so the hexapod automatically adjusts the instrument to get very close to the rock without touching it.
“To get the precision we need, we need to get down to the micrometre,” Allwood says, “getting close enough to the rock to make the hairs on the back of an engineer’s neck stand up.”
Once PIXL is in place, it opens the door for another AI system to step in. PIXL scans a postage-stamp-sized area of rock, firing its X-ray beam thousands of times to create a grid of microscopic dots, each of which reveals information about the chemical composition of the minerals present.
Minerals are crucial for answering important questions about Mars: In some rocks, scientists might be looking for carbonates, which could provide clues about how water formed the rocks, or phosphates, which could have nourished microbial life if it existed in Mars’ past.
Scientists have no way of knowing in advance which of the hundreds of X-ray shots will find a particular mineral, but once the instrument finds a particular mineral, it can automatically stop and collect more data, a process called “long dwell.” As the system improves through machine learning, the list of minerals PIXL can focus on with long dwells continues to grow.
“PIXL is kind of a Swiss Army knife in that it can be configured depending on what scientists are looking for at any particular time,” said JPL’s David Thompson, who helped develop the software. “Mars is the perfect place to test the AI, because there’s regular, daily communication and the opportunity to make tweaks along the way.”
As future missions go deeper into the solar system, they will be out of contact for longer periods than current missions at Mars, so there is strong interest in making the missions more autonomy. Wandering and Conducting Science For the benefit of humanity.
The main goal of the Perseverance Mars exploration mission is AstrobiologyThis includes searching for signs of ancient microbial life. The rover will characterize Mars’ geology and past climate, pave the way for human exploration of Mars, and will be the first mission to collect and store Martian rocks and regolith (broken rock and dust).
A subsequent NASA mission, in partnership with the European Space Agency (ESA), will send a spacecraft to Mars to collect sealed samples from the surface and bring them back to Earth for detailed analysis.
The Mars 2020 Perseverance mission is part of NASA’s approach to explore Mars from the Moon. Artemis A mission to the Moon that will help prepare for human exploration of Mars.
JPL is managed for NASA by the California Institute of Technology in Pasadena, California, and builds and operates the Perseverance rover.
For more information about Perseverance, please visit:
News Media Contact
Andrew Good
Jet Propulsion Laboratory, Pasadena, California
818-393-2433
Email: 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
