Using AI to map disaster damage for faster rescue and recovery

CLARKE, developed by researchers at Texas A&M University, can assess up to 2,000 homes in seven minutes. After a hurricane, drones can capture detailed images of damaged homes and roads. But getting those images to emergency responders is harder: The files are huge, and connectivity in the wake of a storm is often spotty. CLARKE, developed at Texas A&M University with funding from the National Science Foundation’s AI Institute for Societal Decision Making, is designed to speed up the process. After a drone lands, the CLARKE system can analyze images without an internet connection, and teams can then send compact reports detailing building damage and road conditions, rather than huge image files. The work is part of a broader effort to use AI in emergency response efforts. After Myanmar experienced a major earthquake in March 2025, Microsoft’s AI for Good Lab analyzed satellite imagery to help aid groups identify damaged buildings. By reducing delays in transferring and interpreting images, researchers estimate CLARKE could help emergency managers make decisions roughly a day earlier. This, in turn, allows them to focus on the hardest-hit areas first. In a serious crisis, that could