Projects

Started

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 May 2025

31 May 2026

ENA months

£326,315

Power

London

Network Innovation Allowance

Project Contact, National Grid Electricity Transmission

Spectral imaging, which offers higher spectral resolution than traditional colour images, can differentiate between the spectral signatures of the surrounding landscape and those caused by rust. Additionally, spectral imaging has the potential to identify varying levels of rust severity by utilising detailed spectral features as indicators of rust type. This refinement will enhance the automated grading system that will be explored in this project. In previous VICAP projects, an AI model demonstrated the capability to detect corrosion and assess historical steelwork imagery to determine corrosion rates for towers across the network over the past decade.Data Quality Statement (DQS):? The project will be delivered under the NIA framework in line with OFGEM, ENA and NGGT / NGET internal policy. Data produced as part of this project will be subject to quality assurance to ensure that the information produced with each deliverable is accurate to the best of our knowledge and sources of information are appropriately documented. All deliverables and project outputs will be stored on our internal sharepoint platform ensuring access control, backup and version management. Relevant project documentation and reports will also be made available on the ENA Smarter Networks Portal and dissemination material will be shared with the relevant stakeholders. Measurement Quality Statement (MQS): ? The methodology used in this project will be subject to our suppliers own quality assurance regime. Quality assurance processes and the source of data, measurement processes and equipment as well as data processing will be clearly documented and verifiable. The measurements, designs and economic assessments will also be clearly documented in the relevant deliverables and final project report and will be made available for review. In line with the ENA's ENIP document, the risk rating is scored low risk. URHICA will deliver a UAV-based hyperspectral imaging system for condition assessment of steel lattice towers. Both hardware and software solutions including machine learning will be employed to allow automated grading of rust with high accuracy and distinguishing rust types. The project deliver comprise of 6 milestones. Milestone 1: Completion of hardware and software adaption to UAV-based hyperspectral imaging system, UAV setup and testing on NTU campus (include test results on daylight variability correction). Milestone 2: Completion of field testing for UAV based hyperspectral/LIDAR solution. Milestone 3: Completion of initial (Machine learning) ML software development/adaptation for processing of hyperspectral imaging data with test results based on data collected on one steel lattice tower. Milestone 4: Completion of fieldwork for steel lattice scenario. Milestone 5: Completion of software for automated corrosion grading and steel lattice post processing for validation.Milestone 6: NIA compliant completion final report along with training manual for the internal stakeholders and dissemination. The key objectives of URHICA project are:Demonstrate that the drone-based hyperspectral system provides accurate automated classification of rust grades, enhancing the quality and consistency of visual assessments.Reduce the time required for data capture and assessment of steelworks.Establish the solution's capability to differentiate between various types of corrosion. ?

Electricity transmission asset monitoring is a time consuming and laborious task, which is often inconsistent and not quantitative or reliable. The URHICA project proposes the use a UAV-based hyperspectral imaging system for condition assessment of steel lattice towers. Both hardware and software solutions including machine learning will be employed to allow automated grading of rust with high accuracy and distinguishing rust types.

24/04/26