Firefighters respond to destroyed homes as a helicopter drops water as the Palisades Fire grows in Pacific Palisades, California on January 7, 2025.
AFP via Getty Images
Artificial intelligence, it turns out, is no match for Mother Nature.
Following California’s devastating wildfires in 2022, Governor Gavin Newsom announced a slew of new initiatives aimed at mitigating future disasters. “We’re using cutting-edge technology in our firefighting efforts, exploring how innovations like artificial intelligence can help us identify threats faster and use resources smarter,” Newsom explained.
The idea is simple: train an artificial intelligence model to analyze a video feed for signs of growing fires and alert a human team that can put out the fire before it becomes a disaster.
But as the Palisades fire has shown, it’s not that easy, especially when weather conditions are as extreme as they were earlier in the week. The Palisades Fire was aided and abetted by Santa Ana winds that reached 100 mph spreading embers through brush, trees and ultimately thousands of homes in the area.
“The problem is that in conditions like this, you can only have 60 seconds from the time the fire starts to the point where it’s not controllable,” said Daniel Swain, a climate scientist at the University of California, Los Angeles. Forbes.
Since the summer of 2023, Cal Fire, the state’s fire prevention and response agency, has relied on a statewide network of more than 1,100 publicly available video cameras, known as ALERTCalifornia, to power its AI system. which was created and is maintained by the University of California, San Diego. Over the past year, there have been at least 1,200 cases across the state where the AI tool not only detected a fire, but was faster than traditional 911 reporting during 30% of those incidents. (Data for 2024 has not yet been released.)
“All fires start out as small fires, but when it’s pushed by 60-100 mph winds and the fuel goes from grass and brush to homes, filled with petroleum products, that’s just unsustainable.”
It is difficult, if not impossible, to understand whether this AI has been specifically helpful in preventing the spread of otherwise devastating fires.
“It’s really hard to quantify the number of fires that didn’t happen,” said David Acuña, a spokesman for Cal Fire. Forbes. “But I can tell you that the data from ALERTCalifornia has been extremely helpful, we’ve used it in fire behavior training.”
Experts say that while early detection is generally helpful, this week’s Palisades fire west of Los Angeles — fueled by unusually strong winds and extremely dry conditions — grew too quickly before AI cameras could catch up. a significant difference.
“All fires start as small fires, but when it’s pushed by winds of 60-100 miles per hour and the fuel goes from grass and brush to houses, filled with petroleum products, it’s just unsustainable,” Acuña added.
UC San Diego’s ALERTCalifornia camera network captured this image of the Palisades Fire on January 7.
ALERTCalifornia | UC San Diego
The Palisades Fire has been called the most destructive wildfire in Los Angeles County history, and Wall Street analysts predict it will likely result in tens of billions of dollars in damage. As of Friday morning, Cal Fire was reporting that the fire had burned over 20,000 acres. More than 9,000 homes and other structures were damaged or destroyed, and over 150,000 people were under evacuation orders throughout the Los Angeles metropolitan area.
Josh Wilkins, who worked as a firefighter for the San Bernardino County Fire Department for nearly 30 years, and now serves as an advisor to a venture-backed fire technology startup, BurnBot, worked with early iterations of this system. He said Forbes that fire detection AI is now much better than it used to be, but not enough.
“[The AI cameras are a] tool we can use, but there’s a lot more we have to do to put out the fire,” he said. “There’s no way to erase the wind. You’re never going to change Santa Ana [winds]. We will never change that. There’s nothing a fire truck or anybody can do.”
Cal Fire estimates that both the Palisades fire and the Pasadena-area Eaton fire, in another part of the region, are now in the state’s top five most destructive in recorded history. California alone spends over $3 billion a year fighting wildfires, while federally, billions of dollars are spent annually fighting wildland fires nationwide.
Multiple startups and even a fire-focused venture capital firm, Convective Capital, have sprung up in recent years to support new fire prevention and mitigation technology. In November, a new trade group called the FireTech Innovation Association was created by executives from Convective Capital and a nonprofit called Megafire Action.
However, the state’s artificial intelligence fire detection system is not without its recent successes. The Orange County Fire Authority, a county fire agency south of Los Angeles County, just announced that in December, it used this AI system to detect and suppress a December incident in a remote area known as Black Star Canyon, east of the city of Orange.”
Armed with this early detection, OCFA dispatchers launched a wildland fire response and our firefighters contained the fire to less than a quarter acre – meeting OCFA’s goal of containing 95% of vegetation fires to 10 acres or less,” the agency wrote on January 3. “No homes or structures were damaged, there were no injuries, and no evacuation was necessary.”
MORE FROM FORBES
