Quantifying the effects of railhead treatments on adhesion

Does big data hold the key to make a step change in how the industry use railhead cleaning treatments to reduce low adhesion?
Class 323 train in sunshine

The GB rail industry invests considerable resource on strategies to manage railhead adhesion, for example, carrying out railhead cleaning and lineside vegetation management. Despite these efforts, the impact of low adhesion to both industry and passengers is estimated at £345m per annum. Better understanding of different railhead cleaning options can improve their effectiveness and enable infrastructure managers and operators to better target their application, ultimately improving train performance and reducing costs.

Combining data routinely collected by industry along with data collected on a one-off basis from several research projects currently being undertaken in partnership with West Midlands Trains and Network Rail LNW Route will create a data-rich picture to assess the effects of railhead treatments on adhesion. We are using the Birmingham Cross-City line, a 32-mile route that frequently experiences low adhesion conditions and supports an intense metro-style service. Working in conjunction with West Midlands Trains, Network Rail LNW Route, we are collecting:

  • Braking performance and operational data
  • Treatment intervention data (Traction Gel Applicators, Adhesion Treatment Using Service Trains (ATUST), and Railhead Treatment Trains and Multi-Purpose Vehicles)
  • Weather data including moisture levels on the railhead.

The data analysis will address the following questions: 

  • What level of train braking improvement is provided by the different treatments immediately after their deployment?
  • How resilient is the improvement over time?
  • What is the variability observed in the improvement and what are the key influencing factors?

The three defining properties or dimensions of big data are volume, variety and velocity. The data, from services across the Cross-City line, will be collected between October 2018 and February 2019. The large size (volume) of the data, the speed at which it will be collected (velocity) and the large number of different sources (variety) means a big data approach is required to handle it and carry out the downstream analysis.

Results from this work are expected in March 2019.

Topics Covered:

  • Rail cleaning
  • Railhead treatments
  • Mitigation strategies
  • Predicable braking
  • Big data
Haven’t found what you’re looking for?
Get in touch with our expert for more information
Aaron Barrett
Tel: 020 3142 5435
Cookies help us improve your website experience.
By using our website, you agree to our use of cookies.