British Steel is exploring the use of artificial intelligence to further enhance the quality of its rail operations

As part of a forensic study to ensure the integrity of its products, the UK’s only manufacturer of rail currently takes 4,000 photographs of each length of rail during the rolling process.

The images – taken at temperatures of 950°C – are studied in detail to find any potential imperfections. If an imperfection is spotted, the manufacturing process is paused while the issue is resolved and the affected rails quarantined.

To reduce the time taken to spot any imperfections, the company is working in partnership with the School of Engineering at the University of Leicester, as part of SUSTAIN and IGNITE projects, to see if artificial intelligence can enhance the process.

Dr Andy Trowsdale, British Steel Technical Director, said: “Our rails are world class, meeting the strictest of standards. To help control product quality, we use 6 cameras to take over 4,000 images of each rail while it is being manufactured, generating over 100,000 inspection images every hour.

“The cameras capture the whole surface area of the rail, giving us our first close-up of the product – enabling us to see if there are any issues with the manufacturing process. The faster we identify an issue, the faster we can resolve it.

“As we take millions of photos of our rails each year, we want to explore if artificial intelligence can reliably identify imperfections quicker than our current systems. If it can, it would enhance manufacturing performance, improve product quality and service delivery, and reduce rejection costs.

“This project is still in its very early days, but we are excited by the potential.”