Autonomous drones learn to find 'hidden' meteor impact sites

It is easy to find great meteorites (or their craters) once they have arrived on Earth, but the little ones often neglect: scientists recover less than 2 percent of them. Soon, however, it could be a matter of sending a robot to do the work. The universe today reports that researchers have developed a system that has autonomous drones use the learning of the machine to find smaller meteorites at the impact sites that are “hidden” (even if the observers traced the fall) or simply inaccessible.

Technology uses a mix of convolved neuronal networks to recognize meteorites based on training images, both online images and from shooting on the team’s collection. This helps the AI that distinguishes between spatial rocks and ordinary stones, even with a variety of ways and terrain conditions.

The results are not impeccable. While a test drone did the planted meteorites correctly, there were also some false positives. It could be a time before robotic aircraft are sufficiently reliable to provide accurate results on your own.

However, the implications for spatial science are significant if the technology demonstrates precise, however. It would help scientists to observe and potentially recover meteorites that are too small to find or too remote. That, in turn, could help identify meteorite sources and identify rock compositions. In short, drones could fill the gaps in the understanding of the humanity of the cosmic debris that terminate at our door.