NAMPA — As the western United States is ravaged by wildfires, Idaho has taken a step into the future of fire monitoring.
Idaho’s first 360-degree fire detection camera was installed on the campus of Northwest Nazarene University in July. Now, NNU students and professor Dale Hamilton are working to develop artificial intelligence technology to work with the camera and pinpoint potential wildfires faster.
The fire detection camera has a nearly 360-degree view of the hills bordering the Treasure Valley, from the foothills around Bogus Basin all the way out to the Owyhees. Its live feed of the landscape is available to NNU, the Bureau of Land Management and even to the public at alertwildfire.org.
The camera can provide images of smoke plumes rising miles out into the hills in the event of a wildfire, Hamilton said.
There are already dozens of the same cameras scattered across California, Hamilton said, a few others in Nevada and one in Washington, all part of a program from University of Nevada, Reno. Ken Smith, associate director of the Nevada Seismological Laboratory, said the program began about four years ago as a way to reduce response times to wildfires, which could save thousands of dollars, and potentially lives, as well.
From the beginning of the program to now, Smith said they’ve monitored more than 200 wildfires across the American West. In several cases, he said it was through the cameras that the fires were recognized in the first place.
Smith remembered one instance in southern California when a wildfire ignited along a highway, and they were able to have one of their cameras watching the blaze within 30 seconds. With the live feed providing immediate information on the scale of the fire, the response team was able to determine the best strategy to stop it, Smith said.
The fire burned about 4,000 acres, but Smith said under the same conditions without the use of the detection cameras, a similar wildfire could have been devastating and burned tens of thousands of acres.
A second fire detection camera in Idaho is in the works near Bogus Basin, said Christopher Simonson, a fire management specialist with BLM. Simonson said the cameras are typically installed in areas with high elevation like Bogus Basin, but Hamilton said NNU was selected as Idaho’s first location because of the campus’ good internet connection and placement in the center of the Treasure Valley. Each camera costs about $3,000, paid for by BLM, Hamilton said.
As Idaho continues to install more cameras across the state, Hamilton said comparing the different viewpoints can help response crews pinpoint the exact location of wildfires.
While the state installs more fire detection cameras, NNU junior Blake Johanson is working on developing an artificial intelligence program for his senior project, which he hopes to pair with the camera to detect early stages of wildfires before the average human eye can identify it. Hamilton said Johanson has started his project a year in advance to get a head start on a task he expects will take multiple years to complete. He expects the technology to be operational by 2020’s fire season.
Johanson said the artificial intelligence works by giving the technology thousands of different images of smoke to memorize and reference. Eventually, the pictures can teach the technology to identify smoke in other images with a varying degree of certainty, he said. Hamilton said similar projects cost about $50,000 a year, and at this point their project is not funded through another source.
For the artificial intelligence to be adopted by agencies that control wildfire response, Hamilton said they need to get the technology to identify wildfires and alert the necessary agency before that agency is alerted now. That will be a difficult task, Smith with the Nevada Seismological Laboratory said.
Simonson said BLM mainly relies on in-person lookouts and 911 calls made by the public. Smith said these calls can alert the agencies within minutes, or even seconds, of a fire igniting.
However, with fire season starting earlier and the wildfires turning more catastrophic every year, Smith said he hopes the technology is successful so it can become a new tool in the firefighters’ toolbox. Simonson said with the combination of other innovations entering the field, such as thermal imaging, the artificial intelligence has the potential to make a huge difference in the future of wildfire response.
“If they can make that consistent and perfect that technology, it could be very valuable,” Simonson said.