About 4.5 million U.S. homes are considered at high or extreme risk of wildfire. While wildfires mainly destroy unpopulated land and remain relatively “manageable” (fewer than two percent classified as “significant” or conflagrations), any wildfires of any size near populated areas wreak havoc on those who live there. Factors such as weather and geography make it difficult to predict fire movement and size. Limited reporting amplifies the impact of unpredictable fires.
First responders need accurate and timely information to help limit the destruction wildfires cause to homes and businesses. Until recently, technology and infrastructure limited people’s ability to act as early beacons or alarms for an encroaching fire. While the recent advent of smartphones improves people’s ability to call or report information, networks often overload during emergencies. Furthermore, no clear aggregator exists for the reported information making it hard for responders to synthesize information.
Read about how we helped a technology company use Machine Vision to automate the detection of wildfires within a crowdsourced warning system designed to help save lives and property. The solution we partnered to build can identify wildfires with 98% accuracy using cameras on mobile devices.