Feasibility of integrating operational data with adhesion forecasts

Feasibility of integrating operational data with adhesion forecasts

University of Huddersfield Institute of Railway Research and the Met Office

With support from South Western Railway


SUMMARY

This project will investigate how operational data that is already collected by the railway can be used to enhance knowledge of adhesion conditions and improve adhesion forecasting. This involves validating existing low adhesion forecasts; enhancing forecasts with additional data sources to provide close to real-time information about conditions; and, providing a source of information that could be used to improve real-time operational and safety decision making in the future.


AN INTERVIEW WITH THE PROJECT LEAD


Julian Stow, Institute of Railway Research. University of Huddersfield

How would you describe your project in one Tweet (i.e. 280 characters)? 
"We are investigating how railway operational data can be used to improve forecasts of low adhesion due to leaves on the line to help the rail industry improve autumn performance and punctuality." 

What are the benefits for industry, within in the context of the wider forecasting adhesion landscape?  
"This project aims to investigate what improvements could be made in adhesion forecasting if operational data already collected by the industry were used to enhance knowledge of adhesion conditions on a chosen route. The improvement may arise in several ways; firstly validation of existing low adhesion forecasts, secondly by enhancing them with additional data sources that provide near real time adhesion conditions, and thirdly by providing a source of information which could be used in future to improve real-time operational and safety decision making."

Where did the idea for the project come from and why did you focus on this topic for you entry in the competition
"Low adhesion on the railway is a precursor to SPADs, station over-runs and delays. Understanding when and where low adhesion occurs provides the ability for the railways to plan around the possible disruption and reduces the likelihood of safety-of-the-line incidents. Through its Strategic Partnership with RSSB, the Institute of Rail Research (IRR) has developed a range of tools and capabilities for working with railway safety, operational and adhesion data. The project team has delivered a number of relevant projects such as the Red Aspect Approach to Signals Toolkit (RAATS) and train driver competence performance indicators (DCPIs). These led to the idea that combining a number of operational data sources might provide a means of enhancing low adhesion forecasts."

Team members
Job title and organisation
Interests
Julian Stow
Assistant Director
Institute of Railway Research
School of Computing and Engineering
University of Huddersfield
  • Rail vehicle dynamics
  • Wheel-rail interface engineering
  • Braking systems
  • Railway engineering
Peter Hughes
Principal Enterprise Fellow
Institute of Railway Research
School of Computing and Engineering
University of Huddersfield
  • Railway safety
  • Level crossing safety
  • Big data analytics
  • Machine learning
Dr Rawia El Rashidy

Research Fellow, Department of Engineering and Technology
Institute of Railway Research
School of Computing and Engineering
University of Huddersfield
  • Big Data Analytics
  • Machine learning
  • Internet of Things (IoT)
  • Human Factors
Dr Victoria Chapman


Applied Science Manager
Met Office Surface Transport Team
  • Low adhesion, leaf fall and wind throw forecasting and research
  • Weather and climate impacts on rail
  • Spatial and statistical analysis
  • Model verification
Dr Laura Fawcett

Met Office Surface Transport Team

  • Low adhesion, leaf fall and wind throw forecasting and research
  • Weather and climate impacts on rail
  • Data analysis and model verification


Reports and Research in Brief

More
University of Huddersfield
share