Agencies and officials tasked with keeping vigilant watch over remote areas prone to wildfires have been incorporating artificial intelligence into their toolkit.
PacifiCorp, which is embattled in a class action lawsuit for its role in four major wildfires in Oregon during 2020, says its recently invested in advanced cameras and camera networks to monitor wildfire conditions and help detect fires quickly. Since 2024, the company, which owns Pacific Power, has added 18 Pano AI-enabled cameras with five in Oregon, five in Utah and eight in California. Pano AI technology combines high-definition camera feeds, satellite data and AI monitoring. Launching in 2020, the company currently has cameras in 17 states.
PacifiCorp spokesperson Simon Gutierrez says the cameras provide a 360-degree view with a 30x zoom, allowing the power company to monitor a wide swath of land from a single camera. Altogether, the cameras cover 7,462 miles of transmission and overhead power lines in high fire risk areas. PacifiCorp has a Wildfire Intelligence Center which is manned seven days a week. Gutierrez says last year the Pano AI cameras provided 1,612 smoke detections and 64 fire alerts. Of those, he told the Source, 777 of the smoke detections and 27 fires were in Oregon. In 2024, when the cameras were first installed, 723 smoke detections and 24 fires were spotted across the state. Gutierrez says there are 29 other agencies in Oregon that have access to their cameras.
Derek Gasperini, communications manager for the Oregon Department of Forestry, says while ODF does not currently use AI heavily in early detection, monitoring or responding to wildfires, this season the agency will be using a new, multi-mission aircraft that does have some AI-augmented technology. The Twin Otter airframe is replacing the agency’s long-serving Partenavia P.68 Observer aircraft which is retiring after more than 40 years of service. The new aircraft is equipped with AI-enabled wide-area mapping, high-definition thermal imaging, and mission management software with 3D map visualization and night operations capability. It will allow ODF to identify and communicate new fire starts more quickly and track fire behavior with greater precision, both during the day and at night.
“We’re always looking to use new technology when it has proven effectiveness,” Gasperini told the Source. “With AI evolving rapidly, we are looking for opportunities but can’t endanger the successful tools we currently use for early detection and response. The stakes are too high.”
The Oregon Hazards Lab at the University of Oregon operates 70 wildfire cameras in Oregon with more being added this summer. Associate Director of Research and Development Nick Maggio says eight of those cameras are in Central Oregon. “These include cameras in and around high-value viewsheds near Bend, Prineville, Sisters, Pine Mountain, Green Mountain and other high points that provide views of the surrounding wildland and wildland-urban interface.”
Each OHAZ camera can rotate 360 degrees, tilt and zoom up to 40 times with a range up to 30 miles during the day and up to 40 miles at night using near-infrared capability.
“The AI component is used for detection support,” Maggio explains. “The system regularly scans panoramic camera imagery for visual anomalies that may indicate a new ignition, such as daytime smoke or a heat source at night. Those potential detections are reviewed by ALERTWest operations technicians before alerts are sent to fire and emergency response professionals. In other words, AI is not making the final public safety decision. It is drawing attention to possible new wildfire starts so a trained staff member can review them, filter out false positives and alert the appropriate responders when warranted.”
He says the cameras can also be used after a 911 call to confirm whether there is a fire, where it is and how big it is, before sending resources.
“I would describe AI as a powerful tool, not a replacement for people. The game-changing part is its ability to monitor huge volumes of imagery and draw human attention to something that might otherwise be missed or discovered later,” Maggio says. “The limitations are also important. AI can make mistakes. Weather and lighting can complicate detection, and false positives are always possible. That is why we use a human-in-the-loop model. The best system is not AI alone, nor humans alone trying to stare at hundreds of cameras. It is AI helping people focus their attention faster.”
Did you know...
There are over 270 unique wildfire detection camera sites across Oregon owned and operated by utilities, universities, private entities and state and federal agencies. Many of the cameras are available to the public through ALERTWest.live.
Additional Information
The Oregon Hazards Lab is using advanced sensing, automation and AI-adjacent technologies to support the Pacific Northwest Seismic Network and ShakeAlert earthquake early warning system in Oregon. The lab is also involved in research to support the next generation of environmental research using AI in conjunction with the National Science Foundation’s Sage Continuum, which is a testbed that uses AI sensors in the field to collect real-time environmental data.
This article appears in the Source June 25, 2026.
MAKE LOCAL JOURNALISM HAPPEN
Republish our articles for free, online or in print, under a Creative Commons license.
