NMSU engineering professors use AI to detect wildfires
LAS CRUCES, New Mexico (KVIA)-- Engineering professors at New Mexico State University have launched a new artificial intelligence system designed to detect wildfires within seconds of ignition, a technology they say could help prevent widespread damage and costly power outages.
The program, called Grid Sentinel, went live in November. Although it has not yet detected a wildfire, researchers say early results show promise.
“Identifying wildfire or detecting them in the first few seconds when they occur is typically very challenging,” said Dr. Shi, one of the project’s developers.
To address the growing threat of wildfires, two professors and three students developed the AI-powered system, which uses cameras mounted on transmission towers to monitor for signs of smoke or flames. The cameras are monitored around the clock by artificial intelligence and can rotate, pan, tilt and zoom to assess potential threats in real time.
Even small fires can force local power grids to shut down, leading to significant disruptions particularly for vulnerable residents, Shi said.
“We have many elderly people whose lives depend upon home equipment,” he said, referring to electrically powered medical devices.
Shi said traditional fire detection methods often rely on reports from residents after a fire has already grown.
“Typically when you find that, it’s already growing into a pretty large fire,” he said. “Time is super important.”
While the system has not yet identified an active wildfire, it has detected smoke from backyard barbecues and house chimneys, demonstrating its sensitivity and early warning capabilities.
Shi said the technology can do more than detect flames.
“We can identify them. We can predict, we can forecast the path so that the power grid, the people living nearby and the forest department can take actions immediately to minimize the damage,” he said.
Three systems are currently installed in the Las Cruces area, with three more planned.
