Visual Inspection and Condition Assessment Platform (VICAP3) - Bar by Bar

Completed

Other Cross-Cutting Technologies or Research
Other Power and Storage Technologies(Electricity transmission and distribution)

Applied Research and Development

PHYSICAL SCIENCES AND MATHEMATICS (Computer Science and Informatics)
ENGINEERING AND TECHNOLOGY (Electrical and Electronic Engineering)

Not Cross-cutting

Project Contact
National Grid Electricity Transmission

Network Innovation Allowance

Ofgem

01 December 2024

31 March 2026

ENA months

£488,300

Power

London

Network Innovation Allowance

Project Contact, National Grid Electricity Transmission

In the previous innovation project (NIA_NGET0048), the ability of an AI model to identify and grade corrosion according to NGET standards was demonstrated. This project aims to develop algorithms and data collection methods which enable bar by bar assessments and fittings wear measurements to be done using drone collected images. Images collected from the drone will include internal steelwork surveys. With success, climbing surveys to measure the extent of wear on fittings (amongst other things) can be gradually phased out as they are expensive and hazardous. The first part of the project is the corrosion to bar mapping. Here, Keen AIs corrosion detection and localisation algorithms will combine with positional data from a drone to assess the state of particular steel bars on a tower.The second phase of the project will explore fittings wear measurement. Using known measurements on tower fittings, the localisation algorithms will be enhanced to determine whether there is wear on a fitting and its extent.The third phase of the project will use drones with the latest obstacle avoidance technology to fly inside a tower and gather condition data. This data is then used to complete a full 360 degree tower steelwork assessment.Key deliverables for the project are: Process for mapping images to regions of an 3D model with the ability to calculate extent and measurements. Metrological algorithms for matching of regions of known dimensions in an image and using these to measure wear. Bar by Bar of Corrosion for NGET tower created using bar by bar mapping algorithms. Measurement of Fittings Wear for fittings on a selected NGET tower created using KAI Wear algorithms. Process and guidance for collecting images to standard they can be processed by the Bar by Bar algorithms. Process and guidance for collecting images to standard they can be processed by the KAI Wear algorithms. UAV Collected images for Bar-by-Bar Assessment with requisite EXIF data. UAV collected images for fittings wear measurement. UAV collected images for internal bar by bar measurement. NIA Compliant Completion and Progress Report The objectives of the project are: To determine the state of each steel bar on a tower by combining Keen AIs corrosion detection and localisation algorithms with positional data from a drone. To determine whether there is wear on a fitting and its extent using the localisation algorithms and known measurements of tower fittings. Use drones with the latest obstacle avoidance technology to fly inside a tower and gather condition data. This data is then used to complete a full 360 degree tower steelwork assessment.

VICAP 3 bar by bar will build on the success of VICAP 2. In VICAP 2, a refined artificial intelligence (AI) model was able to assess images of steel tower and automatically grade the level of corrosion of the steelwork. As part of the asset management practice, climbing surveys are carried out to getbar by bar (BB) assessments of corrosion on steelwork towers and to measure the extent of wear on fittings amongst other things. The aim of VICAP 3 is to develop algorithms and data collection methods which enable BB assessments and fittings wear measurements to be done using drone collected images. The expected benefits will be in assessment cost reduction and efficiencies.

09/04/25