Sending an exploratory rover to another planet is riddled with challenges, among them, what to investigate and what to ignore. Take the Spirit and Opportunity rovers on Mars for example. Ground control scientists decide where the rovers should go and what they should investigate but commands sent via radio signals from Earth can take up to 20 minutes to reach Mars, depending on where the rover is relative to Earth.
Patrick McGuire from the Freie Universität, Berlin, has a way to speed up extra-planetary exploration. His Cyborg Astrobiologist program teaches robots what’s worth looking and what to leave behind. Future rovers and probes could use it to check out otherwise ordinary-looking rocks for signs of life.
The program will be based on a huge database of images of geological features on Earth, the reasoning behind it being that rocks and such out there are similar to those down here. A rover would snap images of its surroundings and then compare those images to the ones in the database. If the computer program found something unusual or something that looked like a living organism on Earth — such as lichen — the rover would investigate more closely.
McGuire and his colleagues have tested the system in landscapes similar to Martian ones, such as around coalbeds and gypsum cliffs as well on sandstone, limestone and mudstone.
Some of those rocks were partly covered with lichen. Lichens are particularly important when seeking out alien life; it’s one of the few living things that could conceivably survive in a Martian environment. Anything that lives on marks is likely to look and act like lichens or algae mats.
Thus far, matching images with similar features in images from the database seems to have worked pretty well. McGuire said in a press release that the computer program agreed with human geologists that it was looking at a lichen nine out of ten times.
McGuire presented his results at the European Planetary Science Congress in London.
Credit: P.C. McGuire, L. Wendt, B. Foing, C. Gross /Freie Universität Berlin, /CSIC-INTA/U. Malta/ESTEC/U. Chicago