Mars rover uses A.I. to decide what to zap with a laser

Robotic rover makes some of its own decisions on what to study on Mars

mars curiosity rover

This is a self-portrait of the Mars rover Curiosity, which is back to work after a software glitch forced the robot into safe mode on July 2. A.I. software has been added to the rover, allowing it to make decisions on its own on objects to study further.

Credit: NASA

NASA’s Mars rover Curiosity now has the ability to decide what targets it wants to capture with a camera or hit with its laser all on its own. No humans needed.

The space agency announced this week that using artificial intelligence (A.I.) software, developed by NASA’s Jet Propulsion Laboratory, the robotic rover is “frequently” choosing multiple targets per week all on its own.

Most targets are still chosen by scientists curious about particular rocks or areas of soil they’ve seen in images that Curiosity has sent back to Earth, but the fact that the rover can make some of its own decisions about what to focus on adds a new, and important, capability to the mission.

"This autonomy is particularly useful at times when getting the science team in the loop is difficult or impossible -- in the middle of a long drive, perhaps, or when the schedules of Earth, Mars and spacecraft activities lead to delays in sharing information between the planets," said NASA robotics engineer Tara Estlin, in a statement.

Curiosity, which has been working on Mars since it landed in August 2012, has been searching for evidence that the Red Planet was ever capable of sustaining life – even in microbial form.

The rover, which carries drills, cameras and scientific instruments, has had a successful run, finding evidence that ancient Mars had rivers and lakes of fresh water, along with various chemicals needed for life as we know it.

Curiosity is back to work after sustaining a software glitch earlier this month that temporarily placed the machine in safe mode.

With Curiosity back at work, NASA engineers are hopeful that by giving the rover a new level of autonomy it will be even more productive because it won’t have to wait for instructions from Earth and can take steps on its own.

After having the A.I. software uploaded to Curiosity, the rover can choose what rocks and areas are scientifically interesting and then use its laser and telescopic camera to investigate them.

Both the telescopic camera and the laser are part of Curiosity’s ChemCam instrument, which has fired more than 350,000 laser shots at about 10,000 points, according to NASA. Curiosity also has used ChemCam to investigate more than 1,400 targets, enabling scientists to identify the targets’ chemical compositions.

The AEGIS software, or Autonomous Exploration for Gathering Increased Science, had been used before in Curiosity’s predecessor, the Mars rover Opportunity, which has been working on Mars since 2004.

However with Opportunity, the A.I. software was used less often and was used mainly to analyze images taken with a wide-angle camera to help identify rocks that should be more thoroughly imaged.

NASA noted that with Curiosity, the AEGIS software is used to analyze images normally taken at the end of each drive, using the rover’s navigation camera.

The software uses criteria, such as size, shape and brightness, uploaded by scientists to decide if it should use its laser and telescopic camera to investigate a rock or an area.

The A.I. system also directs the ChemCam instruments in their work. For instance, it pinpoints the laser at fine-scale targets.

"Due to their small size and other pointing challenges, hitting these targets accurately with the laser has often required the rover to stay in place while ground operators fine tune pointing parameters," Estlin said. "AEGIS enables these targets to be hit on the first try by automatically identifying them and calculating a pointing that will center a ChemCam measurement on the target."

This story, "Mars rover uses A.I. to decide what to zap with a laser" was originally published by Computerworld.

To comment on this article and other CIO content, visit us on Facebook, LinkedIn or Twitter.
Download the CIO Nov/Dec 2016 Digital Magazine
Notice to our Readers
We're now using social media to take your comments and feedback. Learn more about this here.