Firefighters in California are turning to artificial intelligence (AI) to preemptively combat wildfires. They use footage from over 1,000 cameras across the state. This cutting-edge system then sends alerts to first responders, guiding when to spring into action.
AI Prevents a Potential Disaster
Recently, the ALERTCalifornia AI initiative demonstrated its prowess. It detected a fire in the remote Cleveland National Forest, located about 50 miles east of San Diego. This fire erupted at 3 a.m., a time when most people were asleep. Darkness hid the rising smoke. Without intervention, this fire might’ve grown uncontrollable. Thanks to the AI’s timely alert, a swift response from 60 firefighters had the situation under control within 45 minutes.
This AI platform stems from collaboration between the University of California San Diego (UCSD) and DigitalPath, a Chico-based firm. The integrated system leverages 1,038 cameras, each able to pan a full 360 degrees remotely. These cameras, installed by various public organizations and power companies, form a vast surveillance network.
Since its inception on July 10, there have been several instances of the AI spotting fires even before 911 calls were made.
However, Neal Driscoll, UCSD’s geology and geophysics professor and the primary brain behind ALERTCalifornia, commented, “The sample size so far was too small to draw conclusions.”
Cal Fire envisions this AI-driven approach expanding beyond California. Recent devastating wildfires in places like Hawaii, Canada, and the Mediterranean amplify this need. Suzann Leininger, a Cal Fire intelligence specialist, believes this system can be replicated globally, especially with the increasing frequency and intensity of wildfires due to climate change.
Training the AI
AI accuracy improves through continuous learning. Part of Leininger’s role involves teaching the AI. She reviews video footage, confirming or refuting the AI’s fire detection. Various factors like dust, clouds, or even vehicle exhaust can mistakenly trigger alerts. Yet, with many specialists fine-tuning the AI across California, its precision has noticeably improved.
A Comprehensive Platform
The AI platform isn’t just about cameras. It gathers vast data, such as aerial surveys assessing vegetation that could ignite future fires. Moreover, it maps the Earth’s surface under the vegetation. Drones and airplanes provide infrared and other wavelength data surpassing human vision capabilities.
During winters, this system assesses atmospheric rivers and snowpack. The UCSD team also studies the aftermath of fires, including erosion effects, sediment spread, water quality, and soil health.
Driscoll emphasized the platform’s open access, stating, “We’re in extreme climate right now. So we give them the data, because this problem is bigger than all of us.” This initiative symbolizes a fusion of technology and nature, aiming to address the environmental challenges head-on.