Omnivision track monitor

Railway track inspection tasks expedited with machine learning automation

Image credit: Omnicom Balfour Beatty

Rail track inspection processes could be transformed, thanks to artificial intelligence (AI) software developed by Omnicom Balfour Beatty and the University of York, potentially saving the railway industry £10 million per year in track maintenance costs.

The state-of-the-art machine-learning technology is the result of a two-and-a-half-year 'Knowledge Transfer Partnership' backed by Innovate UK, in which the two organisations shared academic understanding and practical, industry-led insights to digitalise and advance the way in which railway line inspections are carried out.

Attached to the front of the train, a camera moves along rail tracks in need of inspection. The technology uses machine vision, which captures high-definition images of the track to generate data. This is then transferred to a system that analyses the data to highlight inaccuracies and faults on the tracks.

In addition, the technology can help in identifying where future faults may occur, allowing problems to be fixed before there is a failure that requires more urgent repair.

The automated technology, which is currently being progressed from proof of concept into commercial-grade software, is set to provide a quicker, more efficient and safer alternative to what is currently a manual track inspection process.

Automation will improve safety by minimising workers’ exposure to live track environments and will also allow inspections to be completed more quickly.

Stephen Tait, head of operations for Omnicom Balfour Beatty and project lead, said: “We are developing digital technologies that are rapidly changing our industry, from ‘predict and prevent’ technology and advanced digital surveying techniques through to data science. All of our solutions are underpinned by a long legacy of design and construction expertise.

“Our collaboration with the University of York has been invaluable. This latest innovation is an excellent example of how Balfour Beatty continues to deliver our commitment to reduce our onsite work by 25 per cent by 2025 as we progress against our commitment to develop technologies to evolve the digital railway for a more reliable, cost-efficient and safe network for all users”

Professor Richard Wilson, lead researcher on the project from the Department of Computer Science at the University of York, said: “These machine-vision technologies for high-speed rail inspection will improve the reliability of the railway network, reduce costs and increase the safety of manual inspection. The computer vision and machine learning technologies provide automated inspection of complex assets such as junctions and crossings”.

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Recent articles