It turns out that artificial intelligence may give us a better chance to find life on Mars.
A group of researchers led by the team of NASA Institute of Astrobiology (NAI) of SETI Institute recently released a study detailing how to use machine learning to more effectively determine where life may be hidden on the red planet, if it exists.
Scientists decided to use an area called Salar de Pajonales to run their project-this area is located at the intersection of Atacama desert and plateau, which is famous for being uninhabitable, but not uninhabitable-because it is considered as a good substitute for the surface of Mars. The team combined still images, samples and drone lenses to draw a map of life distribution in sparsely populated areas.
Then, the researchers brought the data back to the laboratory for training artificial intelligence. The goal is to teach computers to identify which areas are most likely to have life—such as small clusters of certain geological features or scarce water sources—and to be able to extrapolate this data to other locations. Theoretically, once AI knows the life-oriented details in Salar de Pajonales, it should be able to use these patterns to find the areas that are most likely to bear life in any terrain.
And the effect is surprisingly good. Once AI learns what it needs to recognize, researchers can combine machine learning with collected statistical ecological data to find signs of life 85% of the time. Without the help of artificial intelligence, the team was able to discover biometrics in less than 10% of the time.
You can imagine how helpful it is to find life on the surface of Mars. At present, we send rovers to the land of our next-door neighbors and look for life when they drive along their own paths. Although we try to choose a landing site that has a good chance of containing life (if it exists), it is different from the best and most promising area that the rover can microscopically observe around the planet it is driving.
The research team made it clear that although the speed at which artificial intelligence-assisted programs find life is incredible, it is equally exciting to find a process that can combine long-distance and short-distance data into useful maps to find microbial life. "Although the high biometric detection rate is the core result of this study, it is equally important that it successfully integrates data sets with different resolutions from orbit to ground, and finally links regional orbit data with microbial habitats," Nathalie A said. Cabrol, an astrobiologist and the lead researcher of NAI group of SETI Institute. "Our team showed a way to transition from the scale and resolution needed to characterize livability to the scale and resolution that can help us find life."
I hope we can integrate this technology into rovers, helicopters and drones going to Mars in the near future. Maybe this will be the ticket to finally find life on the red planet.