While questions remain about the effectiveness of AI in many fields, it does seem like it’s beginning to have a notable impact in some aspects of disaster recovery. In Turkey, in support of disaster recovery and ground rescue missions, the US Department of Defense has deployed their xView2 tool, which uses machine learning with satellite imagery to help identify and categorize infrastructure damage, as reported in Technology Review. Originally developed by the Pentagon’s Defense Innovation unit with Carnegie Mellon’s Software Engineering Institute, the tool had previously been deployed in response to wildfire and flooding disasters.
In Turkey, the tool has been used to support multiple search and rescue ground teams, as well as in support of identification of unreported infrastructure damage, which would previously relied on eyewitness reporting. Working by comparing adjacent pixels in satellite imagery, it’s estimated to be up to 90 per cent accurate in its evaluations, and can process large amounts of data rapidly.
Ritwik Gupta, the principal AI scientist at the Defense Innovation Unit, acknowledges there have been some challenges in getting ground teams to start using the tool, saying to Technology Review, “First responders are very traditional. When you start telling them about this fancy AI model, which isn’t even on the ground and it’s looking at pixels from like 120 miles in space, they’re not gonna trust it whatsoever.”
Source:
