To tackle the annual autumn challenge caused by ‘leaves on the line’, the Rail Safety and Standards Board (RSSB) is collaborating with the University of Sheffield in a project to develop a tool that uses artificial intelligence (AI) to help predict low adhesion track conditions.

These conditions cost the rail industry around £350 million and are a serious safety and operational issue as they not only cause delays that affect train performance but can also result in trains station over running station stops and passing signals at danger.

The project is investigating how detailed information on local conditions can be used to tackle the problems associated with leaves on the line. The level of adhesion between train wheels and the rails is affected by a number of factors, including temperature, humidity, the presence of leaf layers, and other contaminants.

Artificial intelligence will be used to analyse data and high-resolution video footage to provide more accurate predictions about friction at the wheel-rail interface. An online tool being developed by the project should be ready for use by autumn 2023 and will allow users to enter data that will generate friction predictions for anywhere on the network.

Further information about the project can be found at https://www.rssb.co.uk/what-we-do/key-industry-topics/adhesion/adhere-adhesion-research-challenge

Paul Gray, Professional Lead Engineering, RSSB said: “While people may think of leaves on the line as a joke, or just an excuse used when a train is delayed, the reality is that it’s a very serious issue for the rail industry. Low adhesion causes significant safety risks and operational problems, costing millions of pounds to manage.

“Our new research project will use artificial intelligence and data analysis to predict and identify where and when low adhesion is going to occur on the rail network. This will allow targeted action at these specific locations, to help manage the safety risks and reduce delays.”

Roger Lewis, Professor of Mechanical Engineering, University of Sheffield said: “It is very exciting for the team at Sheffield and RSSB that our fundamental analysis of the causes of low adhesion as well as our extensive collection of data from track is now coming together to enable the development of the AI friction prediction tool that will help the railway industry with performance and safety issues around Autumn.”