RADRFIRE Combines Artificial Intelligence with Satellite Imagery for Wildfire Detection

The wildfire situation in the western U.S. is obviously extremely problematic, but thanks to something called RADRFIRE, improvements are on the horizon.  RADRFIRE, or Rapid Analytics for Disaster Response for Wildfires, is a new Dept. of Energy (DOE) initiative that utilizes artificial intelligence (AI) combined with cloud computing to analyze images and provide earlier warning of impending disaster.

The Nuts and Bolts of RADRFIRE

RADRFIRE is useful because it not only predicts the direction and velocity of fire activity, but it can also pinpoint the localized areas that are particularly hard hit, in real time.  This helps with recovery efforts as well as identifying the safest evacuation routes.  The system is even useful for post-event activities, like identifying areas at-risk for landslide activity.

The nice thing about RADRFIRE is that it provides a fully-automated surveillance solution.  The system retrieves images and analytics for all fires in the U.S. in order to facilitate the most informed decisions possible.  Generally speaking, the images originate from satellites like the ISS, or via nocturnal aircraft or drone surveillance, which are interpreted by human analysts that also factor-in weather, vegetation, fuel and forecast data.

The system is already widely used across the U.S., but a global expansion is likely on the horizon (mainly because the satellites capturing the imagery are global satellites).  The overarching goal is to not only assess damage, but to predict it as well.  And because the data resides “in the cloud,” it can be retrieved by stakeholders in any location.

Obviously, I like this a lot!  Aside from the emergency preparedness perspective, seeing images of the California wildfires, etc., leaves a sick feeling in my gut.  If RADRFIRE can alleviate even 10% of this issue, it’s a worthwhile excursion in my book.  It’s certainly not the end-all-e-all, but it’s certainly a step in the right direction.