A network of AI-enabled cameras scanning the horizons of Arizona’s Coconino National Forest on a March afternoon detected a wisp of smoke. Within minutes, a human analyst confirmed it wasn’t dust or a cloud, and Arizona’s forest service and the state’s largest electric utility were on alert. Firefighters scrambled to the remote area and stopped what became the Diamond Fire at just 7 acres. The early warning didn’t come from a lookout tower or a 911 call. It came from a camera running artificial intelligence software that had been trained to spot the first sign of a wildfire.
Across the fire-prone West, states and utilities are racing to install such cameras ahead of what forecasters warn will be another severe fire season. From Arizona to Colorado to California, the technology is proving faster than the public’s first call for help — in some cases, firefighters have already extinguished a blaze before anyone thought to dial 911.
“Earlier detection means we can launch aircraft and personnel to it and keep those fires as small as we can,” said John Truett, fire management officer for the Arizona Department of Forestry and Fire Management. The department has deployed seven AI cameras of its own, while Arizona Public Service, the state’s largest electric utility, has nearly 40 active AI smoke-detection cameras and plans to have 71 by summer’s end.
In Colorado, Xcel Energy has installed 126 AI cameras and aims to have them in seven of the eight states it serves by year’s end. ALERTCalifornia, a network run by the University of California, San Diego, operates some 1,240 AI-enabled cameras across the state. “The AI that’s being run on the cameras is actually beating 911 calls,” said Neal Driscoll, a UC San Diego professor of geology and geophysics who founded the network.
The technology is particularly valuable in high-risk areas that are sparsely populated, rural or remote, where a blaze might not be quickly spotted by human eyes. Brent Pascua, a battalion chief for the California Department of Forestry and Fire Protection, or Cal Fire, described instances where a response began before a 911 call was made. “In a few cases, we’ve actually started a response, went there, put the fire out, and never received a 911 call,” he said.
Arvind Satyam, co-founder and chief commercial officer of Pano AI, whose technology combines high-definition camera feeds, satellite data and AI monitoring, said the company’s system detected 725 wildfires in the U.S. last year. Customers include forestry operations, government agencies and utilities, including Arizona Public Service. “In many of these situations, we hear from stakeholders that the visual intelligence, the time, really, really gives them a head start and some of these could have taken off into hundreds if not thousands of acres,” Satyam said.
Cindy Kobold, an Arizona Public Service meteorologist, said the technology notifies them about 45 minutes faster on average than the first 911 call. Satyam said the technology was developed in response to a lack of hardened solutions to combat worsening wildfires, which climate change is making burn hotter, faster and more frequently.
The systems do not come cheap. Pano AI charges around $50,000 annually per camera, a figure that includes fire risk analysis and round-the-clock monitoring. False alarms remain a challenge, with every suspected smoke plume requiring a human analyst to confirm it isn’t a cloud, dust or a glint of light. Patrick Roberts, a senior researcher with the nonprofit RAND who recently completed a project on accelerating innovation in wildfire management, said that even when the AI correctly identifies a fire, “do you send help right away? Do you monitor? Should you worry about it? Where do you send help? Do you think about evacuation? All this still requires people and decision support systems.”
In highly populated areas, people tend to spot fires quickly, and the technology is less useful when hurricane-force winds intensify and rapidly shift the flames, as happened in Los Angeles last year. But even then, Pascua said, the AI-fed information helps firefighters on the ground. “As the fire moves and shifts around, that’s where the human factor comes in and decides which tactics are best in fighting the fire,” he said. “AI can only do so much. It just provides that real time information where we can make better decisions on the fire ground.”
The technology is expanding beyond detection. Roberts said AI can also identify the best places to thin vegetation and burn cool fires, and sensors can monitor air quality for smoke with sensitivity far beyond a home carbon monoxide detector. At George Mason University, professor Chaowei “Phil” Yang is working with researchers in California and NASA’s Jet Propulsion Laboratory to build a system that forecasts where a fire will burn and which communities will be most affected by smoke, aiming to give agencies real-time maps for evacuations, school closures and air quality warnings. Yang said the goal is to have the system operational in three years.
“AI in wildfires, it’s no longer just speculative. It’s really being used,” Roberts said. “The future is AI everywhere, and the lines will blur between AI wildfire detection and just wildfire detection as the lines will blur in other areas of our life.”