Maricopa County is using nearly 50 AI-powered wildfire detection cameras across elevated locations, bringing Arizona’s wildfire camera buildout further into county emergency response as monsoon storms raise fire risk across the state.
The county network includes Thompson Peak, the most recent installation, and was built with help from Arizona Public Service, AZFamily/KOLD reported. The cameras are mounted on mountain peaks and monitored from a secure, undisclosed county facility used to track emergencies including wildfires and flooding.
Automated scanning meets human verification
“Wildfires are our most common threat we see in Maricopa County,” Ron Coleman of the county’s emergency management department told AZFamily.The system moves the first detection step from public reporting toward automated smoke scanning and human review.
Instead of waiting for someone to see smoke and call the fire department, the cameras scan high-risk areas and route possible smoke detections for review.
The review remains human. “A live human being then reviews that to make sure that it isn’t just a dust devil and that it is a wildfire,” Coleman said. The design puts AI at the start of the alert process, but leaves confirmation with county staff before a fire response is triggered.
Building on early utility deployments
The county layer builds on a utility deployment already underway. In March 2025, APS disclosed AI fire-sensing cameras operating 24/7 in targeted high-risk areas, with alerts sent to APS fire mitigation experts and fire dispatch centers when smoke or heat traces were detected.
APS said at the time it expected more than 30 cameras across Flagstaff, Payson, Prescott, Sedona, north Phoenix and southeastern Arizona by summer 2025.
“The new AI cameras act as powerful extra sets of eyes,” Scott Bordenkircher, APS forestry and fire mitigation director, said in the company’s 2025 announcement. “When minutes matter, early fire detection provides real-time information so firefighters can respond faster and we can make critical operational decisions about our energy grid to help keep communities safe.”
Statewide deployment has grown quickly since then. The Arizona Corporation Commission said in April 2026 Pano AI had zero detection stations in Arizona two years earlier, 51 deployed stations at the time of the commission’s wildfire mitigation town hall and a projected 88 stations active statewide by the end of 2026.
The commission also said APS and Tucson Electric Power/UNS Energy use Pano AI as part of their wildfire mitigation plans.
State fire agency adds another layer
Arizona added a state fire-agency layer in March. Gov. Katie Hobbs announced AZFIRECAM, a seven-camera deployment for the Arizona Department of Forestry and Fire Management.
The state said the 360-degree cameras detect smoke and notify the Arizona Dispatch Center, with cameras positioned at Sawmill near the Hualapai Mountains, Blake Ranch in Mohave County, Wittmann in western Maricopa County, Dudleyville and Suffering Gulch in Pinal County, Foreman Wash in the Tucson area and Chiricahua south of the Dragoons.
Early detection saves crucial minutes
The operating case for the cameras is clearest in sparsely watched terrain. AP reported in May that an APS-linked camera in Coconino National Forest detected early signs of smoke in March.
Human analysts verified the alert, state forestry officials were notified and crews contained the Diamond Fire before it grew past 7 acres. APS meteorologist Cindy Kobold told AP the technology notifies the utility about 45 minutes faster on average than the first 911 call.
Navigating the constraints of cost and command
The constraint is cost and command. AP reported Pano charges around $50,000 annually per camera, including fire-risk analysis and 24/7 intelligence-center staffing.
Patrick Roberts, a senior researcher at RAND, told AP early detection still does not answer the response questions: “Do you send help right away? Do you monitor? Should you worry about it? Where do you send help? Do you think about evacuation? All this still requires people and decision support systems.”