Accelerated Condition Assessment at National Grid

Keen AI launched with a small scale of a pilot with National Grid in 2018. This involved developing the first reliable object detection model for extracting a single component, an Andre spacer, from video footage. Enabling the accelerated review of conductor spans between towers.

In summer 2019 we began processing all visual condition data for National Grid’s Overhead Line network. We also improved the reliability and scope of our models – able to identify all core components for Suspension and Tension tower types. Since 2019 we have processed and indexed over 46M images as well as incorporating common defects to the models.

Our AI platform, KAI, is being used at National Grid to support analysts to rapidly identify the objects of interest in recorded footage. Using deep learning, KAI is able to identify the key components automatically and present these to the operator for review.

National Grid were able to achieve an over 60% reduction in the time taken to review over head line footage for condition assessment. As part of a follow on project the input from the user’s assessment is being used to train a machine learning algorithm to allow the model to automatically identify defects.


The Problem Statement

  • National Grid captures video footage of Overhead Lines to identify any potential faults with towers, fittings or conductors.
  • Analysts spend thousands of hours reviewing footage looking for problems and reporting on condition.
  • The process is laborious, has significant bottlenecks and quality is dependant on the engineer reviewing the footage.
We worked closely with National Grid teams to fully understand the end-to-end process
We worked closely with National Grid teams to fully understand the end-to-end process
A mix of images, video and infra-red footage is now collected and analysed
A mix of images, video and infra-red footage is now collected and analysed
First model extracting a type Andre spacer from footage of a span
Detecting Individual Components on Overhead Lines (OHL) Towers for National Grid

Successful Outcomes

  • OPEX Saving: Reduced the time taken to process footage by 66%.
  • Self Sufficiency: Analysts are able to train AI models to identify and extract relevant components (Spacers, Joins, Dampers, Insulators).
  • Risk Reduction: Reduced the risk of asset failure by removing the backlog of footage to be processed.
  • Improved Asset Management: Allows the cataloguing of historic footage to support and maintain asset inventory.
  • Rapid ROI: Entire capability deployed (with models and user training) in two months.

It took just one meeting! Keen AI went away and built a working prototype. We’ve spent months with some external suppliers and not got anywhere.

Mark Simmons Condition Monitoring Team Leader National Grid ETO

Learn more about our AI Condition Assessment solution for Electricity Transmission Operators.

Learn more about how we are using KAI to detect corrosion on infrastructure assets.

Newsletter Sign Up

Stay up to date with our work, and industry-related developments.

Hamzah Reta

Hamzah Reta

Hamzah is excited by the potential of AI to take engineering processes to even greater heights. Following his passion for integrating these two worlds to build a better future, he is dedicated to helping Keen AI grow and achieve that vision.‍

More Posts from 

Hamzah Reta

Sarah Bauroth

Keen AI Book Club

As part of our book club, we read a different book every month. The only rules are, that none of

Read More >
Hamzah Reta

AI and the Justice System

The legal system is close to a crisis of confidence. Notice the phrase “crisis of confidence”. This is not just a challenge of processing a backlog of cases but ensuring the public feels the system is doing what it is meant to do: deliver justice.

In this episode Cari and I discuss what this means. Can AI be used to achieve this goal? What would a framework to evaluate the use of AI in a legal setting look like and why she’s starting with road traffic offences.

You can visit the project page on the King’s College website.

Read More >