With so much going on in the world, Marketplace is here for you. Support public news for public good. Donate Today!
California’s wildfire detection tech was no match for the Palisades fire
Jan 16, 2025

California’s wildfire detection tech was no match for the Palisades fire

HTML EMBED:
COPY
Cyrus Farivar of Forbes explored how the power of nature overwhelmed the AI-powered network of sensors. But as wildfire damage increases across the country, the emerging “fire tech” field is working on solutions, he explains.

California relies on a variety of tools to stop and mitigate wildfires, some as low-tech as dumping giant buckets of seawater on the flames.

But on the higher-tech side is a new, AI-powered monitoring system called ALERTCalifornia, which was developed at the University of California, San Diego. It’s designed to speedily detect and report wildfires using a network of over 1,000 cameras and sensors.

The developers say the network detected over 1,200 blazes across the state during the 2023 fire season, sometimes with impressive quickness. But the system wasn’t quick enough to prevent the current disaster in Los Angeles.

Marketplace’s Kimberly Adams spoke with Cyrus Farivar, a senior writer at Forbes, who explored how the fury of the Palisades fire overwhelmed that human-made system.

The following is an edited transcript of their conversation.

Cyrus Farivar: I think what is worth noting, and what any fire professional will tell you, is that these conditions, in this particular fire, where you had extremely high winds — we’re talking 80-, 90-, 100-miles-per-hour winds — you had very kind of steep, difficult terrain topography, if a fire gets going in those conditions, especially when the brush and flammable materials on the ground are extremely dry, right? Anyone who’s been in Southern California within the last six to 12 months will tell you there’s been very, very little rain, so you have a lot of kind of worst-case scenario conditions, which is what the National Weather Service has said. And so, did the cameras capture the early stages of the fire? Yes, they did. But was that warning enough to mitigate the growth of this particular fire? No, it was not. Mother Nature pretty quickly overwhelmed 21st century artificial intelligence here.

Kimberly Adams: Yeah, I mean, in an ideal world, what would the camera network have done?

Farivar: Yeah so, you know, a blaze would start, the AI would recognize that it’s a fire, right, as opposed to fog, as opposed to something else that maybe looks like fire but isn’t an actual wildfire. And we have actually a pretty recent example of this out over in Orange County, which is, of course, one of LA’s neighboring counties to the south. You know, there was a fire that the Orange County Fire Authority, you know, dealt with in December that they told the public about earlier this month in January, where they said that there was a fire that had been detected by this AI system in a sort of remote canyon east of the city of Orange, and in that case, the system correctly identified it as a fire, alerted the relevant fire agency, Orange County Fire Authority, and they thought, yeah, maybe this is worth checking out. And they sent somebody out there, and they put out what, at that time, was a very small blaze. And so that’s why this idea of having early detection, early warning, is a good one. But as we were saying, in the Palisades fire, right, the conditions were such that the fire just grew too quickly, too fast, too far, too much, too everything, you know. That’s not to say that this system isn’t doing anything. It’s not to say that it’s worthless. Fire officials that I spoke to from Cal Fire would tell you that, you know, they use this and have used it for training and, and many other purposes, as well as some of these smaller blazes like we saw in Orange County. But it’s just worth noting that there are conditions where the fire is just too much, and even the earliest warning just isn’t going to make a difference.

Adams: Outside of this fire, how well was this camera system working?

Farivar: Yeah, so it’s interesting because if we ask the question, right, like, did this AI-driven camera network prevent fires that otherwise would have gotten out of control, that otherwise would have been disasters? The honest answer is it’s hard to know, and fire officials will tell you that, right? They don’t know. They can’t know, right? We can’t know about fires that never happened. We can’t know about fires that were put out early. What they would say, and what ALERTCalifornia has said, in the fire season of 2023 the AI system detected over 1,000 blazes, 30% of those were reported faster than a 911 call. Were all of those blazes going to get out of hand and cause some major disaster? It’s hard to say. But I think that what, you know, environmental scientists, ecologists, fire scientists would tell you is that we are now starting to see not only destructive fires, more and more destructive fires in California. It’s important to remember that within the last five years, we have had some of the most destructive fires across the state, just in our very recent past and more and more and more frequent weather conditions that allow for more destructive fires. We have longer drier seasons, we have, in some cases, hotter weather. We have extreme winds. It’s really important, I think, to heed warnings that our fire officials give us and to think about ways that we can, you know, be prepared because, like it or not, you know, in many parts of California, and, frankly, many other parts of America as well, we have to deal with fire one way or the other.

Adams: Have any other states or communities adopted similar systems to this one?

Farivar: That’s a really good question. I’m not aware of a system that is operating on this scale. I mean, it is worth remembering that California, of course, has dealt with wildfire, you know, from time immemorial. And you know, back hundreds of years ago, when there were many, far less people, far less, you know, human-made structures across the state, that was a different sort of story. And you know, Cal Fire is an agency, a state agency that is well-equipped, well-trained, to deal with fires all over the state. But there are lots of efforts, more broadly, not just in other parts of California, but in other states, as you mentioned, to use technology in other kinds of ways. There’s an entire, sometimes it’s called fire-tech community. There’s, in fact, a new venture capital firm called Convective Capital, founded by a guy in Mendocino County, California, to examine and support and try to use technology in a good way to mitigate fire disasters.

Adams: Let’s talk a bit more about that because you’ve spoken with quite a few experts on this. What are some of the ways to potentially improve this system or other systems to deal with huge wildfires like what we’ve seen in California?

Farivar: So there are other companies now that are trying to help in many different ways. There’s one company called BurnBot that is using essentially a little robot, you know, and it’s designed to be a tool that firefighters can use for what they call prescribed burns or controlled burns, right? So that’s when firefighters are deliberately setting areas on fire under very controlled conditions, typically during the winter, so that you don’t have as destructive fire in typically the summertime or the late summertime. There’s another company that’s using other types of sensors that are chemically looking for smoke. So they hang sensors in the forest, and there’s a chemical sensor in the device that is looking for the chemical signature of wildfire smoke. Another company that I think is, actually it’s not a company, it’s a nonprofit, that has been getting a lot of attention lately in the wake of the LA fires is this company out of here also in the Bay Area in Sonoma County, California, called Watch Duty. Watch Duty is an app, it’s free to download, free to use, and it will closely map and monitor where a fire is burning, but also it will overlay things like prescribed burns, but also, you know, where people need to evacuate, where the evacuation warnings are, where the wind is blowing, where the red flag zones are, which is potentially a fire-dangerous area. And it’s essentially a wildfire newsroom that kind of collects and verifies in an extremely accurate and professional way all of this fire information, so I think that’s yet another example of kind of critical use of technology in the era of wildfire.

More on this

My colleague Kai Ryssdal, host of “Marketplace,” also explored the role tech is playing in fire response with Kate Dargan Marquis, senior wildfire adviser to the Gordon and Betty Moore Foundation. One example is low-orbit satellite systems designed to detect fires just a few meters in size.

The future of this podcast starts with you.

Every day, the “Marketplace Tech” team demystifies the digital economy with stories that explore more than just Big Tech. We’re committed to covering topics that matter to you and the world around us, diving deep into how technology intersects with climate change, inequity, and disinformation.

As part of a nonprofit newsroom, we’re counting on listeners like you to keep this public service paywall-free and available to all.

Support “Marketplace Tech” in any amount today and become a partner in our mission.

The team

Daisy Palacios Senior Producer
Daniel Shin Producer
Jesús Alvarado Associate Producer
Rosie Hughes Assistant Producer