OAK RIDGE, Tenn. — Researchers at Oak Ridge National Laboratory are hoping that by using advanced technology, they may be able to tell when a wildfire has a chance of starting and prevent it before it spreads.
They said Monday that researchers were using sensors, drones and machine learning technology to prevent wildfires and reduce their damage to the electric grid. They are using the technology to remotely sense electrical arcing, which could indicate faulty equipment. That way, crews can fix the equipment before it sets something on fire in the mountains.
And if a fire does break out, they said they hoped to be able to tell the direction fires could spread using the same technology. It could make response time faster and help coordinate crews better, improving the way wildfires are extinguished.
ORNL said the largest recorded wildfire years have occurred since 2004, coinciding with many of the warmest years on record. They said many wildfires are started by power lines, and the disaster can cause widespread electric outages.
The Department of Energy chose funding for a handful of wildfire projects during the last wildfire season. Two of them were at ORNL.
First, ORNL planned to put sensors near power lines or in power electronics equipment that would register variations in power flow, which could lead to superheated sparks. Researcher Ali Ekti is building an algorithm to quickly recognize these kinds of changes in power flow.
Once the issues are identified, ORNL said drones could be used to record video and take measurements of the area's audio, heat and radio frequencies. It can be a lot faster than sending a helicopter, and drones can fit into areas helicopters may not be able to reach.
They can also be equipped with heat-detecting sensors to tell where a fire is coming from and will be able to find locations where embers can spark new fires.