As reported by The New York Times and The Los Angeles Times, the fire department of the California Natural Resources Agency, Cal Fire, is working on utilizing AI to detect wildfires.
With access to a network of 1,039 high definition cameras placed in strategic locations around the state, AI is being trained to detect early signs of smoke so that firefighters can be alerted as early as possible. Some of the cameras are fixed, some are PTZ cameras, and some feature infrared capabilities for night vision. According to The New York Times, during the pilot program, the AI system detected 77 fires before 911 dispatchers had been alerted.
See also: U.S. judge rules art created by AI cannot be copyrighted
Because the pilot program, which took place at six Cal Fire emergency centers, was so successful, the program is now coming to all 21 centers.
“The success of this project is the fires you never hear about,” Staff Chief of Fire Intelligence with Cal Fire Phillip SeLegue told The Las Angeles Times.
Before this program, the network of cameras has been monitored by humans around the clock, which means that human error and fatigue are factors. If and when the program goes full scale, humans will still need to confirm any fire flagged by the AI system, but Cal Fire officials are impressed by the system’s capabilities. Director of ALERTCalifornia Neal Driscoll says that the system has already become adept at distinguishing smoke from other phenomena such as low clouds or fog from miles away.
