Autonomous drones learn to find ‘hidden’ meteorite impact sites

It’s easy to find large meteorites (or their craters) once they’ve reached Earth, but the smaller ones often go neglected — scientists recover fewer than 2 percent of them. Soon, however, it might just be a question of sending a robot to do the job. Universe Todayreports that researchers have developed a system that has autonomous drones use machine learning to find the smaller meteorites in impact sites that are either ‘hidden’ (even if observers traced the fall) or simply inaccessible.

The technology uses a mix of convolutional neural networks to recognize meteorites based on training images, both from online images as well as staged shots from the team’s collection. This helps the AI distinguish between space rocks and ordinary stones, even with a variety of shapes and terrain conditions.

The results aren’t flawless. While a test drone did correctly spot planted meteorites, there were also some false positives. It could be a while before robotic aircraft are trustworthy enough to provide accurate results all on their own.

The implications for space science are significant if the technology proves accurate, though. It would help scientists spot and potentially recover meteorites that are either too small to find or too remote. That, in turn, could help pinpoint meteorite sources and identify the rocks’ compositions. Simply put, drones could fill gaps in humanity’s understanding of the cosmic debris that lands at our doorstep.