ORNL researchers have collected images of damage following extreme weather events such as Hurricane Ian to build a robust damage system that can detect and analyze utility pole damage to aid in disaster response efforts. (Image: ORNL, U.S. Dept. of Energy)

As fossil fuels are phased out in the effort to slow global warming, we will be depending upon a reliable source of electric power more than ever. And that means blackouts caused by weather events such as hurricanes, tornadoes, and snowstorms will have increasingly serious consequences.

A team of researchers at Oak Ridge National Laboratory (ORNL), headed by Dr. David Hughes, is tackling the problem of locating damaged utility poles in far less time than it usually takes with currently available methods. An important guiding principle of their system is to keep its cost to a minimum and to make it easy for non-technical first responders and utility repairers to use.

After tornadoes struck Amory, Mississippi, ORNL researchers, including Dakota Haldeman, collected some 10,000 UAS images that were processed in the field and used to create maps of the damaged parts of Amory in support of disaster recovery efforts. The “Cornelia” system can be seen on the circular red helipad in the background. (Image: ORNL, U.S. Dept. of Energy)

They have modified a commercially available drone (UAV) and fitted it with a complex system that includes a camera and flight controller as well as AI hardware and software for edge processing. In order to identify damaged poles and obtain their precise locations, the AI has to integrate the camera images, the attitude of the drone, and location data as derived from the global positioning system (GPS).

They refer to their system as an unmanned aircraft system (UAS), which includes both the remote-controlled UAV and its operator. As a sort of pun on the Department of Energy’s (DOE) Office of Cybersecurity, Energy Security, and Emergency Response (CESER), which funded the research, they named the prototype aircraft “Cornelia” after the first wife of Julius Caesar.

The ORNL Autonomous Systems Group, led by Andrew Duncan, modified the drones by adding a camera, AI, and the ORNL-developed MAVNet communication system, which enables global command and control of the UAV. Importantly, MAVNet combines multiple communication technologies, so it can function when infrastructure is damaged or unavailable.

For purposes of this research, the ground control system (GCS) interface was modified to support hosting imagery collected by the Cornelia aircraft in real time to display the locations of the damaged utility poles.

The location information is uploaded to the already-existing central processing hub, called the Environment for Analysis of Geo-Located Energy Information (EAGLE-I™) system, which is DOE’s centralized platform that monitors power distribution outages for over 140 million customers — 92 percent of the U.S. and its territories.

In just seconds, the data is uploaded to the MAVNet cloud server and distributed to users anywhere in the world.

“What we want to do is get the ball rolling so that we can detail what the UAS should be like, what the algorithm should be like, and how they should be fielded. And we then look forward to dispersing this information to state, local, and federal governments and say: ‘Let's tackle this problem together,’” said Hughes.

Summing it up. The research team has demonstrated an end-to-end deployable framework including prototype to detect, classify, and geo-locate utility poles on the edge. The proposed framework can be scaled to solve the problem of assessing post-disaster energy infrastructure damage.

Both Hughes and Duncan say that they are committed to this project because it is a great opportunity to use technology for good ends.

This article was written by Ed Brown, Associate Editor, SAE Media Group. For more information, visit here .