- Title: California turns to AI to help fight wildfires
- Date: 11th August 2023
- Summary: EL CAJON, CALIFORNIA, UNITED STATES (AUGUST 8, 2023) (Reuters) SUZANN LEININGER, INTELLIGENCE SPECIALIST AT CALIFORNIA DEPARTMENT OF FORESTRY AND FIRE PROTECTION (CAL FIRE), LOOKING AT AI CAMERA FOOTAGE ON MONITORS MAP AND CAMERAS LEININGER LOOKING AT AI CAMERA FOOTAGE ON MONITORS AI CAMERA FOOTAGE ON COMPUTER MONITOR
- Embargoed: 25th August 2023 10:48
- Keywords: AI Wildfires
- Location: EL CAJON, CALIFORNIA / SAN DIEGO, CALIFORNIA UNITED STATES
- City: EL CAJON, CALIFORNIA / SAN DIEGO, CALIFORNIA UNITED STATES
- Country: US
- Topics: Disaster/Accidents,North America,Wildfires/Forest Fires
- Reuters ID: LVA001373310082023RP1
- Aspect Ratio: 16:9
- Story Text:Firefighters in California have begun using artificial intelligence to help fight wildfires, connecting machines that learn to a network of 1,036 cameras to more quickly spot fire.
When every minute counts to contain a fire in its early stages, firefighters are typically alerted first by citizens calling a 911 emergency dispatcher. But since the California Department of Forestry and Fire Protection (Cal Fire) launched its pilot program on July 10, the AI has already shaved precious minutes off response times, and has demonstrated it is learning how to reduce false positives, said leaders of the program known as ALERTCalifornia.
Moreover, Cal Fire hopes the technology can serve as a model for other states and countries around the world, a need underscored by unusually devastating wildfires in Canada and the Mediterranean this season.
Developed by engineers at the University of California San Diego using AI from DigitalPath, a company based in Chico, California, the platform leverages 1,036 cameras erected by various public agencies and power utilities throughout the state, each one capable of rotating 360 degrees at the command of remote operators.
"It's 100% applicable throughout anywhere in the world, especially now that we're experiencing a lot larger and more frequent fire regimes and with climate change," said Suzann Leininger, an intelligence specialist at Cal Fire in El Cajon, just east of San Diego.
One recent example showed the platform's potential. Fire broke out at 3 a.m. in the remote, scrubby Cleveland National Forest. With people asleep and darkness concealing the smoke, that type of ignition that could spread into a raging wildfire if left unattended. But within 45 minutes the fire was out, Cal Fire said. Artificial intelligence alerted El Cajon's on-duty fire captain, who called in about 60 firefighters including seven engines, two bulldozers, two water tankers, and two hand crews. The fire was contained to an area of 20 feet by 20 feet (6 by 6
meters), Cal Fire said.
Neal Driscoll, a professor of geology and geophysics at UCSD and the principal investigator of ALERTCalifornia, said the AI has already become more accurate in just a few weeks. As it is, any number of phenomena can trigger a false positive: clouds, dust, even a truck with smoky exhaust. Beyond the video, Driscoll said, the platform is collecting vast amounts of additional information, including an aerial survey to quantify the vegetation that will fuel future fires and map the Earth's surface beneath the canopy.
In all the platform is collecting a petabyte of data - about 1 quadrillion bytes, or enough to fill more than 1,000 modern laptops - each year, Driscoll said.
That can be used to eventually improve all manner of fire behavior models and AI applications for studying the
environment and all of it is being shared openly for any private company or academic researcher.
"We're in an extreme climate right now and we need to leverage our funds and work together to mitigate the impacts of these natural disasters," Driscoll said.
(Production: Alan Devall, Daniel Trotta, Liliana Salgado) - Copyright Holder: REUTERS
- Copyright Notice: (c) Copyright Thomson Reuters 2023. Open For Restrictions - http://about.reuters.com/fulllegal.asp
- Usage Terms/Restrictions: Footage contains computer game or software screenshots. User is responsible for obtaining additional clearances before publishing this clip.