If University of Chicago geoscientist Patrick McGuire starts calling the shots, NASA won't have to decide between sending a person or a robot into space. They could just send a "cyborg astrobiologists" — or advanced artificial intelligence riding piggyback on an astronaut.
McGuire's team has been working on what's known as a Hopfield neural network, specifically programmed to identify new forms of life. Right now, the system can differentiate between different types of lichen based on color, and tell what it hasn't seen before. Textures are the next obstacle to overcome, and, in the future, McGuire would like to see the network being able to scan for new life on a microscopic and landscape-wide scope.
We probably won't see this kind of technology being integrated anytime soon, though a pair of researchers did spend two weeks in spacesuits at Utah's Mars Desert Research Station, where they fed McGuire's system using handheld microscopes and cellphone cameras. This would be useful in the long run if astronauts ever explore Mars on foot, though it will also help orbital craft seek out new life on other planets — something that McGuire is no stranger to as he worked on the CRISM instruments aboard the Mars Reconnaissance Orbiter. More information about the team's endeavors can be seen in an article published in Cornell University's arXiv.