Smoke-detecting AI camera installed on Lone Mountain

New technology detects Shedhorn Fire, fire remains at 74 acres

By Bella Butler EBS STAFF

BIG SKY – Under hazy skies on Monday afternoon, a new artificial intelligence camera perched atop Lone Mountain detected a plume of smoke rising from a drainage 13 miles away. The camera had been installed in early September as part of pilot program to test its ability to detect wildland fires in the Big Sky area. Unexpectedly, it launched into action when a fire ignited in Big Sky this week.   

After receiving a 911 alert as well as a notification from Pano AI, the company that brought the AI camera to Big Sky, the Big Sky Fire Department responded to what would soon be named the Shedhorn Fire in the Taylor Fork drainage south of Big Sky. As of press time today, the fire is burning at 74 acres, and the fire department is still using the Pano AI camera to monitor its activity.

“We use artificial intelligence to detect smoke as quickly as we can see it,” said Arvind Satyam, Pano AI’s chief commercial officer at a press event this afternoon. “Over the course of Monday and the last two days, we were able to demonstrate this technology.”

Once the AI camera picks up smoke, it sends an alert back to a 24/7 Pano AI intelligence center, where real people analyze the data , determine if it’s a fire incident, and notify the fire department.

“It was really integral as we were down there trying to get a better location on this fire,” said BSFD Deputy Chief Dustin Tetrault of the Shedhorn Fire response. Because video can also be reviewed retroactively, he added, this could be a potential tool for investigating the cause of fires. 

Pano AI launched 14 months ago inspired by the spike in major fires around the world in recent years. As part of their pilot program, the company has more than 20 camera stations across Colorado, California, Oregon and now Montana. While the Lone Mountain camera is the only Pano AI camera in the Treasure State, both Pano AI and BSFD expressed interest in expanding to other locations.

According to Satyam, several factors went into choosing sites for the pilot. “Each of these locations were areas which we’ve identified as high fire-risk areas,” he said, “but also ones that are innovative fire districts and innovative communities that wanted to be on the front end to go test this capability out.”

Funding partners for the camera include BSFD along with Big Sky Resort, Spanish Peaks and Moonlight Basin.

“It’s always fun to partner with the community and there’s a long history in Big Sky of doing that,” said Big Sky Resort CEO Taylor Middleton. “So, when the fire department calls, we’re all in.”

Having additional cameras installed atop nearby vantage points would create a more robust view and help pinpoint more exact incident locations, according to Satyam. In Big Sky, this will require more funding.

“We need more cameras so we can triangulate,” Middleton said. “The beauty of Lone Peak is you can see so much. The bad thing about Lone Peak is you can’t see Lone Peak. So, we need more cameras.”

The cameras themselves are also in a perpetual state of evolution or cumulative “deep learning,” as Satyam puts it, from each incident detection and 360-degree spin.

“We’re training our AI model based on 350 million historic data sets,” Satyam said. “In addition to that, we’re constantly learning with all of our stations that are deployed across the four states.”

Because it’s late in the fire season, Tetrault said the fire department will use open burn season to test the camera further. “We’re really excited about getting this technology up here,” he said.