Cheltenham level crossing chosen for nationally significant high-tech trial

A Cheltenham level crossing has been chosen for a trial which could have national significance.

Artificial intelligence technology is used in high-tech equipment at Alstone Lane level crossing.

The researchers lent their AI expertise to the development of a prototype device designed to improve safety on the UK rail network.

READ MORE: More delays at Cheltenham as busy junction will remain closed for three months

Experts from the University of Leicester’s School of Computing and Mathematical Sciences, working as part of an Innovate UK (KTP) knowledge transfer partnership with systems engineering specialists Synoptix, installed the device .

The OPTIMUS prototype uses machine learning and an AI-based object detection system to identify and quantify different types of traffic using the crossing. The Network Rail operated facility sits on one of the UK’s busiest rail routes, the Cross Country Route handling up to 130 trains a day.

The software will detect road users, including pedestrians and cyclists, on and near the crossing and generate statistical reports.

Get the best stories about the things you love most curated by us and delivered to your inbox every day. Choose what you like here

This locally processed data can then be transmitted in condensed form to a Network Rail control center in real time.

The Leicester team includes Professors Ivan Tyukin and Alexander Gorban, Dr Bogdan Grechuk, Tatiana Tyukina and George Leete, all from the School of Computing and Mathematical Sciences.

Mr Leete, who is leading the development of the machine learning aspect of the project under the supervision of Prof Tyukin, said: “The deployment of the initial prototype at the crossing marks a significant milestone for the project, as well as the safety of the most wide British rail network.

“With the accurate real-time census data collected through the OPTIMUS prototype, Network Rail will be able to accurately assess the risk at the level crossing and adjust safety procedures if necessary.

“The hope is to extend the deployment of the device to all of Network Rail’s approximately 6,000 level crossings, making the UK rail network safer for all.”

Professor Tyukin said: “The importance of this project and its current progress is difficult to overestimate. In addition to addressing a significant technical challenge, we aim to develop a solution to this challenge that is certified as robust, reliable, stable, maintainable and trustworthy.

“This is a great opportunity to demonstrate the impact of our research into reliable, maintainable and stable AI on improving public safety in the UK.”

Figures for 2019/20 show there were 316 near misses with pedestrians on UK level crossings and two pedestrian fatalities.

All level crossings are regularly risk assessed, and although Network Rail has closed over 1,250 since 2009, its strategy also includes improving safety at other sites.

The device is expected to operate for an initial four-month trial period at its current Cheltenham location, with further trials planned in partnership with Synoptix.

We want to hear your thoughts on this story, so log on and leave your comment below.

About Jun Quentin

Check Also

Rail companies are ‘failing miserably’ to help families with young children travel by train | UK News

Rail companies have been accused of “failing miserably” to help families with young children travel …