Intelligent rail infrastructure: what’s new at Network Rail?

source: Network Rail

“We need to get smarter. Every railway undertaking is trying to build their own algorithm, but that might not be the best approach. We need to engage the supply chain more, and it would benefit us all if we make more data available”, says Tim Flower, Chief Intelligent Infrastructure Engineer at Network Rail.

Since April 2019, Britain’s railway infrastructure manager Network Rail has been busy with its Intelligent Infrastructure Programme, which is led by Tim Flower. It is a five-year transformation programme (2019-2024) focused on turning data into intelligent information to improve services for passengers.

Which technology or method will change the most in railway maintenance in the upcoming years? “A combination of a really good application of reliability centered maintenance techniques”, says Tim Flower. “It is important to understand degradation rates and where to find these. Also, a big expansion in IoT devices will improve condition monitoring, but it has to be in a way that is manageable”.

There is also room for improvement when it comes to how the development goes in the railway industry, says the Chief Engineer. “We need to get smarter. Every railway undertaking is trying to build their own algorithm, but I’m starting to see that might not be the best approach. It would benefit us all if we make more data available, and we need to engage the supply chain more. They could probably develop algorithms much faster than we can.”


One of the most recent developments are the track elements in the Insight tool. “Insight is going to be our one-stop-shop, where information is provided for engineers and section managers. Along with that, we will have a remote condition monitoring platform that will feed data into Insight as well. Ultimately these two platforms will form the basis of predictive and preventive maintenance regimes.”

Network Rail has now been able to develop predictive models for track geometry degradations. “These models can see when various elements of the track geometry are at an alarm level. Now we’re looking at developing the next stages, better management of defects from the Plain Line Pattern Recognition (PLPR) track inspection system. In January 2019, Network Rail opened a second PLPR facility in Derby to increase their capacity to monitor, inspect and fix track faults.”

The first capability of Insight for long- term planning will be released this year. “That will all be about being able to understand the 3 to 8 year work bank, such as what track access we need and how we can make the right decisions based on different scenarios.”

Text continues below picture

Improved data for maintenance teams


Infrastructure managers are also collaborating in the field of predictive maintenance, for example Network Rail has a Memorandum of Understanding (MoU) with ProRail and rail maintenance company Strukton in the Netherlands. “One project that is exciting at the moment is fibre optic sensing, which we are trialling with ProRail. With this, sensors are fitted on lineside telecom fibers, so they can hear any localised noise to the fiber.”

With clever algorithms, this could distinguish whether a person is approaching a level crossing, a train, or in case of a wheel flat. “The system is in its infancy, but quite successful with train tracking at the moment, and has many possibilities”, says Flower.

What the future of data collection looks like

To collect the data, Network Rail has been fitting sensors for quite some time, says Flower. Mainly in the signalling systems, such as points, track circuits, power supplies, and point heating. “We keep developing new capabilities, and have a prototype for level crossings now. That will allow us to do an almost full maintenance service of the components of level crossings.” It is tested, and going through the final stages with three suppliers. In a few months, it will be decided which are best for the routes on the network.

At the moment, data collecting sensors are mainly retrofitted, but the next step would be for new assets to come sensored already. Network Rail recently expanded forward facing cameras on trains, which now covers about 80 percent of the network. “This system has quite a lot of uses, such as monitoring vegetation, and overhead lines.”
The infrastructure manager also wants to make more use of in-service trains for data collection, with lower-cost sensors. This can be combined with higher quality data from measuring trains. But what is the ideal mix of data collection methods? “We have a special programme that is looking into this. We aim to get a clearer picture of what mix of dedicated fleet and in service-trains will be the best for the future.”

Tim Flower shares more insights at the Intelligent Rail Summit on 21-23 September. Take a look at the programme for more information. Download the free magazine about Intelligent Rail with more interviews here.

Author: Esther Geerts


Add your comment

characters remaining.

Log in through one of the following social media partners to comment.