LSU AI Wildfire Tool Wins XPRIZE Recognition
DeepFire, an AI wildfire detection system led by a Louisiana State University professor, received special recognition and $100,000 from the XPRIZE Wildfire competition. The system combines satellite data, smoke sensors, and monitoring cameras to forecast fires up to 35 days ahead, with 90 percent accuracy for forecasts four to seven days out. In tests in New South Wales, Australia, it detected wildfires before the state fire service in most cases.
DeepFire is led by Supratik Mukhopadhyay, a professor at Louisiana State University. The system earned special recognition and a $100,000 award from XPRIZE Wildfire. It combines satellite imagery, smoke-sensing equipment, and camera networks to anticipate fires as much as 35 days out, with 90% accuracy for forecasts four to seven days ahead.
During trials in New South Wales, Australia, DeepFire identified wildfires before the state’s firefighting agency in most cases. The tool arrives amid rising fire danger: Louisiana saw 600 fires in August 2023 alone, close to its prior-decade yearly average, while NASA found global extreme wildfire risk doubled from 2003 over the next 20 years.
If systems like DeepFire prove reliable, residents in fire-prone areas could gain earlier warnings, giving them more time to evacuate or protect property. Emergency responders and land managers may use forecasts to position resources and reduce false alarms. Insurers and local governments might also weigh such tools when assessing risk and planning. However, benefits depend on access, accuracy, and integration with existing agencies.