As California’s wildfires alter landscapes, threaten homes and endanger communities, a new ally has emerged for firefighting agencies: artificial intelligence (AI).
Utilizing AI technology in both prevention and response has become a cornerstone of California’s firefighting efforts, according to Phillip SeLegue, the California Department of Forestry and Fire Protection’s chief of fire intelligence.
“On a daily basis, staff across the state in all of our emergency command centers are monitoring cameras through the ALERTCalifornia network and receiving detections of AI for a possible wildland fire,” SeLegue told ABC News.
Firefighters work to keep a spot fire from growing as the Park Fire burns along State Route 172 in the Mill Creek community of Tehama County, Calif., Aug. 7, 2024.
Noah Berger/AP
Known as a “camera quilt,” the network has more than 1,080 high-definition, pan-tilt-zoom cameras deployed across the state.
The camera network, based out of the University of California San Diego, provides a 24-hour backcountry monitoring system with near-infrared night vision to detect and surveil active wildfires.
The camera system has been a success for Cal Fire, SeLegue said, noting how the automated system is faster than placing a 911 call.
In 2023, TIME recognized California’s use of AI to detect wildfires as the “best invention” of the year.
“California is fighting fires smarter, combining cutting-edge technology with a world-class firefighting force — all to better protect our communities,” Gov. Gavin Newsom in a press release at the time.
Fire detection traditionally consisted of people being staffed at fire lookout locations around the state, which SeLegue said is sometimes “very challenging” due to remote locations, long shifts and potentially dangerous weather conditions.
“It’s reduced the personnel hours when it comes to the deployment of folks sitting up on a hillside or a ridge, knowing that there’s inclement weather or knowing that there’s high fire potential,” SeLegue said.
Inmate firefighters talk while monitoring a firing operation to slow the spread of the Park Fire near Mill Creek, California, August 7, 2024.
Fred Greaves/Reuters
When a wildfire is recorded, emergency command personnel then run fire spread predictions and fire spread simulations through another AI system to see the potential impacts of that fire, according to SeLegue.
Cal Fire has been utilizing Technosylva, a wildfire science and technology company, since 2019.
The AI system runs predictive analytics on historical fire patterns, weather patterns and vegetation types to provide context and a potential roadmap for how the initial blaze could spread.
“We’ve revolutionized fire modeling,” Bryan Spear, CEO of Technosylva, told ABC News.
“When you’re using the power of AI machine learning, you’re able to increase situational awareness and speed up decision-making to a matter of seconds versus minutes or hours,” Spear said.
The Technosylva system also integrates data from satellites and weather stations to model fire behavior and predict potential wildfire hotspots.
This allows agencies to allocate resources more effectively, implement preventive measures in high-risk areas and prepare for potential outbreaks with more precision, according to Spear.
“Ultimately, it saves lives and saves structures,” he added.
Firefighters monitor the Park Fire near Mill Creek, California, August 6, 2024.
Fred Greaves/Reuters
California firefighters have battled more than 5,500 wildfires this year, with a total of 830,000 acres burned, according to Cal Fire. The state’s wildfires have destroyed just over 1,200 structures and led to one civilian death, officials said.
The reality of wildfires in the state has changed over the last 10 years, forcing Cal Fire to adapt its response, according to SeLegue.
“We have already seen early, increased activity and the dryness of the fuels,” SeLegue said of the state’s landscape. “We continue to see the fuels more receptive to the burn environment, meaning they’re more receptive to roadside starts or any type of new ignitions — whether it’s lightning or someone even just mowing their grass.”
A study from the National Integrated Drought Information System released in 2023 found that human-caused climate change contributed to a 172% increase in burned areas from 1971 to 2021.
The 2024 Park Fire, which is now California’s fourth-largest wildfire ever, forced fire agencies to utilize a tremendous force of resources — including AI, according to Spear.
“They’re using it to predict spread and track resources so they can be coordinated and efficient,” Spear said.
