A team of researchers from Columbia University and Safe Pro Group is developing a drone-based machine-learning technology to make land mine removal faster and safer. The drones fly over potentially mined areas, collecting images that are then processed by an algorithm trained on the visual characteristics of various types of land mines. The algorithm creates a detailed map that can recognize and locate explosives more quickly and accurately than a human reviewing the images. The drones have a reported detection rate of about 90 percent and are meant to complement traditional demining methods. The team is testing ways to detect deeply buried explosives and those covered by vegetation using thermal imaging and ground-penetrating radar. The new technology could significantly speed up the demining process, which currently could take over 750 years in some areas.