Bracing for a severe fire season, Western states and utilities are expanding AI-enabled camera networks to spot smoke sooner and speed initial attack. Arizona Public Service plans 71 cameras by summer, while Xcel Energy has installed 126 across the region; California’s ALERTCalifornia network counts roughly 1,240 AI-assisted cameras. Vendors such as Pano AI, now active in 17 U.S. states, say their systems detected 725 fires last year and can flag incidents an average 45 minutes before the first 911 call, as happened with Arizona’s Diamond Fire. The technology carries a price tag—about $50,000 per camera annually including monitoring—and still relies on humans to verify alerts and make tactical decisions, particularly when extreme winds or dense populations limit its utility. Researchers are also developing AI tools to forecast fire spread and smoke impacts to inform evacuations and closures. With hotter, drier conditions amplifying risks, AI is shifting from pilot projects to a standard part of the wildfire toolbox, even as questions of cost, accuracy and operational integration persist.
Related articles:
NASA FIRMS: Near Real-Time Fire Mapping and Data
AirNow: Wildfire Smoke and Air Quality Guidance
U.S. Forest Service: Managing Wildland Fire
Pano AI: Wildfire Detection and Situational Awareness Technology
