American Electric Power (AEP) is using drones and artificial intelligence to spot equipment issues earlier, reduce field exposure and scale inspections with existing staff, according to a recent Factor This and DTECH webinar sponsored by Skydio and Levatas.
The webinar, “Flight to Insight: How AEP Uses Drones and AI to Turn Inspections into Uptime,” was streamed live Jan. 13 and is now available on demand. The session offered a close look at AEP Ohio’s drone inspection program and how the utility is pairing autonomous flights with computer vision to turn large volumes of imagery into actionable maintenance signals.
Jake Reed, the project manager who manages AEP Ohio’s drone program, said the utility piloted drone-based inspections in 2025 and inspected about 4% of its distribution system. Reed said those flights identified more than 150 “tier one” issues, including limbs on lines and thermal anomalies detected with infrared sensors, while also capturing photos of every asset inspected. Historically, Reed said, traditional inspections captured photos primarily when issues were observed.
The increased coverage created a new challenge: the volume of data. Reed said AEP Ohio collected between 400,000 and 500,000 images, and “a single person went through the entire year spending 500 plus hours going over that.” The goal, speakers said, is to automate defect recognition so teams can spend more time inspecting and less time manually reviewing imagery.
During the webinar, Reed shared examples of findings that were not causing outages at the time of inspection, including overheated jumpers and vegetation contact that could have led to equipment failure. In another case, Reed described a pole that appeared suspicious on thermal imaging. After crews replaced it, he said, “when they cut that open, it kind of looked like a cigar. So it was burning from the inside out.”
Skydio’s Christina Park, senior director of energy strategy, framed the discussion around operationalizing technology in utility workflows rather than running isolated pilots. Skydio and Levatas also demonstrated a remote flight conducted from Florida to a drone dock in California and reviewed AI outputs that included thermal anomaly detection and automated readings of analog gauges.
Levatas CEO Chris Nielsen said the integrated workflow is designed to shorten the time between flight and usable information. For post-flight analysis, he said results are typically available “under three to four minutes post flight.”
The full webinar includes the remote flight demonstration, additional inspection examples and audience Q&A. Watch the complete presentation on demand here.
