autoevolution
 

Fuzzy Logic MIT Software Automatically Picks Best Mars Landing Sites for Rovers

NASA has been analyzing the pros and cons of various landing sites for the Mars 2020 rover mission for years and is now down to three choices: Columbia Hills, Jezero, and NE Syrtis. Next month, the agency will be hosting the fourth and final landing site workshop as a means to further narrow down the list.
Three possible landing sites for Mars 2020 left 1 photo
Photo: NASA via Space.com
Choosing these sites was not an easy task, as analysts had to take into account a multitude of factors including if the location would serve the rover’s scientific purpose and how geographically friendly it is.

Of course, software was involved in the selection process, but merely as a tool for providing data, not decision-making.

MIT researchers have developed a software that can determine the best landing sites for rovers faster and better than humans. While doing so, it takes into account geology, terrain, scientific priorities, and engineering constraints.

The use of this software is not meant to replace humans making the final decisions, but the tool might help speed up significantly a process that normally takes years to complete.

The accuracy of the software is proven, says principal research scientist Victor Pankratius, by the fact that it managed to identify on its own several of the landing sites previously considered by NASA for the Mars 2020 rover. It also identified locations that were not brought up at NASA’s workshops as often.

Pankratius says the software uses the so-called fuzzy logic, a mathematical logic scheme that groups data in probability-based fashion.

In simpler words, the software calculates the probability for the rover to navigate the terrain around the landing site, for instance. That’s different than the usual approach that deals with can or can’t scenarios.

“One day, if we have fully autonomous rovers, they can factor in all these things to know where they can go, and be able to adapt to unforeseen situations,” Pankratius said according to MIT News.

“You want autonomy, otherwise it can take a long time to communicate back and forth when you have to make critical decisions quickly.”

There’s no information whether NASA will use the software in the near future, but the agency did fund the research together with the National Science Foundation.
If you liked the article, please follow us:  Google News icon Google News Youtube Instagram X (Twitter)
About the author: Daniel Patrascu
Daniel Patrascu profile photo

Daniel loves writing (or so he claims), and he uses this skill to offer readers a "behind the scenes" look at the automotive industry. He also enjoys talking about space exploration and robots, because in his view the only way forward for humanity is away from this planet, in metal bodies.
Full profile

 

Would you like AUTOEVOLUTION to send you notifications?

You will only receive our top stories