Fire departments turn to AI to detect wildfires faster

In Aspen, Colorado, fire officials say wildfire risk has shifted from being a seasonal concern to a year-round threat. To meet that challenge, they're using AI-powered cameras to detect smoke - sometimes before 911 calls even come in.

Arvind Satyam is the co-founder and chief commercial officer of Pano AI.

"They're getting a full 360-degree picture of the landscape," said Satyam.

High in the mountains, rotating cameras scan the terrain for smoke. If the system detects a potential fire, the alert is reviewed by a human team. Verified alerts are then sent to Aspen Fire via text or email, along with the exact location, Satyam said.

"We run a smoke detection algorithm," Satyam said. "So essentially looking at the imagery to determine is that smoke or not smoke."

Satyam said the devices can also give satellite views of the terrain, and overlay a picture of the weather to understand wind speed and relative humidity.

"It gives us that instantaneous intelligence and not off of, you know, hoping we get someone there quickly," said Jake Andersen, deputy chief of operations for Aspen Fire.

Aspen Fire Chief Rick Balentine said the device especially helps with the terrain in Aspen. Balentine said a recent example of when they used Pano AI was for Aspen's Sunnyside Prescribed Fire. The 900-acre Sunnyside Prescribed Fire was ignited back in April of this year on the south face of Red Mountain, fire officials said.

The system is now used across 10 states: Colorado, California, Oregon, Washington, Idaho, Texas, Wyoming, Utah, Arizona and Montana. Multiple fire agencies can share alerts across those regions and coordinate response efforts.

"Having this type of tool for our department, it makes all the difference in the world to understand the conditions we are sending our firefighters into," said Ali Hammond, Aspen Fire's director of community wildfire resistance.