Team UAV images 500 transformers for NGED test of AI condition checks
The drone specialist built a repeatable capture method and a mapped image database so computer vision models can be tested on spotting corrosion on pole-mounted equipment
- The initial phase of NGED’s wider project is expected to finish by the end of 2026
- NGED owns and operates the local electricity network across the Midlands, South West and South Wales
- NGED is a distribution network operator, separate from National Grid Electricity Transmission
- Team UAV was founded in 2013 and is part of PDG Aviation Services
NGED has split the work between two of its suppliers, image capture from a drone operator and analysis from a separate software firm, rather than from one vendor selling drones and analytics together.
The capture pace works out at more than 35 transformers a day for Team UAV’s crews.
“This innovation project will explore whether aerial imagery, AI, and computer vision can automate aspects of the condition assessment process, generating more consistent condition data to inform asset health reporting and investment decisions.”Favour Eboh, innovation engineer, National Grid Electricity Distribution
Drone operator Team UAV photographed 500 pole-mounted transformers on National Grid Electricity Distribution’s (NGED) network in under two weeks, the company said on 21 October 2026.
NGED will use the photos to find out whether software can pick out corrosion and other wear on the equipment. Before flying the program, Team UAV trialed different equipment, camera angles and capture methods so that shots taken at different sites would be comparable enough for AI analysis.
Crews worked through NGED’s asset list in Cornwall, Wales and the Midlands, swapping in other transformers where vegetation, access limits or missing infrastructure got in the way. Team UAV also built a GIS database so each batch of images uploaded from the field is filed against the right transformer, with its survey notes and metadata.
Mark DeGiovanni, technical lead at Team UAV, said inspecting large numbers of assets across different locations is a real challenge for network operators, particularly where access is restricted, and that the company’s focus was a consistent and repeatable approach so the imagery would suit the next stage of AI analysis.
Softwire, a technology consultancy that also supplies NGED, is writing the computer vision models and a prototype web application to assess transformer condition from the images. NGED said it will test the technology on an initial set of transformers before judging whether it can support asset management across the network.
Sources
- Team UAV, “Team UAV delivers over 500 drone inspections for National Grid Electricity Distribution to support AI innovation”, press release, 21 October 2026. https://www.teamuav.uk/articles-and-case-studies/national-grid-transformer-drone-inspections
