NASA is looking for assistants to analyze images of Mars sent by its robots

In the last years Both NASA and other international space agencies have redoubled efforts to boost human presence on Mars. With new tentacles swirling around it, a Rovran running through it, and the development of futuristic missiles that will transport humans to this desolate world, interest in the Red Planet seems limitless.

even in, Artificial intelligence, or AI, is already at the service of this futuristic featIt has enormous potential to change the way NASA spacecraft study the universe.

But since all machine learning algorithms require training by humans, A new NASA project recently brings anyone in the United States or the world closer to being a part of the research that is being conducted daily, for example, by Curiosity and Perseverance rovers on the surface of Mars today.

Specifically, NASA is asking citizen scientists to name “features of scientific interest in images taken by NASA’s Mars Perseverance spacecraft.“. The project is called AI4Mars Hosted on the Zooniverse platform, it continues work that began last year with the Curiosity spacecraft We ended up creating an object classification algorithm.

“Perseverance pictures will improve [el algoritmo] By expanding the types of tags that can be applied to the features of the surface of MarsOfficials from NASA’s Jet Propulsion Laboratory (JPL) in Southern California, which operates the Curiosity and Perseverance missions, said. “AI4Mars now offers labels to identify more fine details, allowing people to choose options such as floating rocks (“islands” of rocks) or nodules (BB-sized balls, often formed by water, of minerals glued together),” the officials added.

The instrument that is produced by Curiosity Images is called SPOC ( Classification of soil properties and objects ). It was based on work in which people have tagged nearly half a million images, outlining landmarks such as sand and rocks. The tool gets these functions right 98% of the time, Rover drivers are already using SPOC to plan paths for the Red Planet, officials at the Jet Propulsion Laboratory said.

Perseverance has 23 cameras and sends dozens to hundreds of photos to Earth every day.. Mission team members like it Reduce the time between receiving images and uploading instructions to persistence from your computers. This could take hours, as engineers and geologists search the images for specific features of interest to them, as well as terrain that might be dangerous to the rover.

“No scientist can quickly look at all the images sent by the probe every dayexplained Vivian Sun, a JPL scientist who helps coordinate daily perseverance operations and is an advisor to the AI4Mars project. “It would save us time if there was an algorithm that could say, ‘I think I saw veins or nodules here,’ and then the science team could look at those areas in more detail,” he added.

Perseverance sends dozens to hundreds of images back to Earth every day for scientists and engineers to examine specific geological features. But time is short: After those images travel millions of miles from Mars to Earth, the team members have a few hours to develop the next set of instructions, based on what they see in those images, to send to persevere.

All of this classification work will feed into a database that will then be analyzed to be able to classify the terrain and thus help the robots better understand what they are capturing with their cameras. And moving on the red surface. Especially during this stage of development, SPOC requires a lot of validation by scientists to ensure its labeling is accurate. But even when you get betterAnd The algorithm is not intended to replace the more complex analyzes that human scientists perform.

“The key to any successful algorithm is a good data set. Hiro Ono, the AI ​​researcher at JPL who led the development of AI4Mars, explained that the more individual data available, the more the algorithm learns.”Machine learning is very different from regular software. This is not like making something from scratch. Think of it as starting with a new mind. The effort here is to get a good set of data to teach this brain and analyze the data so that it is better learned,” Ono concluded.

In the future, marking such geological features could also aid in the ongoing search for life on Mars, which includes in addition to the rover, an orbiter and a spacecraft that enables the return of Mars samples that is expected to be a reality in the next decade..

Read on: