As drought conditions continue and temperatures change, wildfires are an increasing threat across the Southwest United States, especially in rural, mountainous, and underserved communities.
But two professors at New Mexico State University’s College of Engineering have come up with a solution that could provide much-needed assistance to areas without dedicated wildfire watchers, advanced detection systems, or robust power grid operations centers.
Di Shi, associate professor in the Klipsch School of Electrical and Computer Engineering and the Paul W. Klipsch and Valerie Klipsch Distinguished Professor, and Fengyu Wang, also an assistant professor in the Klipsch School of Electrical and Computer Engineering and PNM Utility Management Endowed Chair and associate director of the Power Utility Management Program, have developed Grid Sentinel, a low-cost option that provides artificial intelligence wildfire detection and response to the power grid.
“We’ve been doing AI research for a long time, and we tend to use very large, expensive, energy-consuming GPUs (graphics processing units) at a very high cost,” Shi said. “Wildfires are also a huge problem for forests and infrastructure, especially for communities close to these areas. Our idea was to target a different approach and develop a very low-cost, palm-sized, low-cost AI system.”
The system developed by Shi and Wang includes cameras that can analyze video and image streams with low energy consumption, allowing multiple devices to be deployed throughout a single location.
“This acts like a safety net for our infrastructure and our communities. It can be detected within the first seconds or minutes of a wildfire, so people can take action, and our infrastructure and power grid can take action.”
Grid Sentinel allows local infrastructure personnel to shut down the power grid near wildfires and protect other communities on the grid from power loss.
“This works in conjunction with the microgrid controller, so it can trigger control actions and isolate a small portion of the grid from the membrane to keep the lights on,” Shi said. “It sends alerts to power system operations and the Forest Service about when and where a fire occurs.”
Shi also said he hopes the system he and Wang have developed will help protect New Mexico and other Southwest states from some of the devastating effects of wildfires. Their goal is to see Grid Sentinel commercially available, which is unusual for a device based on academic research.
Recently, Shi and Wang worked with a team from Kit Carson Electric Cooperative in the Carson National Forest, installing three units in wildfire-prone areas.
“Wildfires are the most significant risk facing KCEC, and also one of the most significant risks for utilities across other states, including New Mexico and California,” Wang said. “This low-cost, scalable system provides tangible value to the communities it serves and is consistent with NMSU’s land grant commitment to community resiliency.”
There are plans to deploy additional units this month and leverage the data generated to advance digital twin research for resilient energy operations. There are also plans to offer this technology to other utility companies and forestry services.
-30-
Photo Caption: New Mexico State University College of Engineering assistant professors Fengyu Wang (left) and Di Shi work with a team at Kit Carson Electric Cooperative to install the Grid Sentinel System, a low-cost option that provides artificial intelligence-powered wildfire detection and response to the power grid. (Photo provided)
Image description: Two men working with devices indoors.
Photo Caption: New Mexico State University College of Engineering assistant professors Fengyu Wang and Di Shi developed the Grid Sentinel system, a low-cost option that provides artificial intelligence-powered wildfire detection and response to the power grid. The system was installed in the Carson National Forest with the help of the Kit Carson Electric Cooperative team. (Photo provided)
Image description: Four men installing equipment on a wooden pole.
Photo Caption: New Mexico State University College of Engineering assistant professors Fengyu Wang and Di Shi developed the Grid Sentinel system, which can analyze video and image streams with low energy consumption and the ability to deploy multiple devices across a single location. (Photo provided)
Image description: Image of a computer screen.
