21.01.20
The age of Intelligent Infrastructure - Network Rail
Source: RTM Dec 19/Jan 20
Predictive technology that informs engineers when, where and why faults will occur on the railway are vital to improving infrastructure performance and driving a better service for passengers.
Jonathan Schofield from Network Rail’s Intelligent Infrastructure programme explains how a new track tool will transform frontline work.
We know the facts: more people than ever are using the railway, demand for rail travel has more than doubled in the past twenty years and passenger numbers continue to grow.
To meet this challenge, the rail industry is facing a seismic change, in what some believe is the arrival of the fourth industrial revolution. Central to this change is the ability to convert masses of data into usable information, combined with improved analytics and technology advancements.
This change will bring benefits to the entire railway industry – from the track teams working to maintain and renew the infrastructure, to the ToCs and FoCs, and of course the millions of people who use the railway every day.
If we are to benefit from the changes now taking place, the industry must do more to turn asset data into useful information to drive operational benefits and improve capacity and safety. And it must do this at a far quicker pace if it is to support the future success of the British economy.
Predict and prevent
This is a time for precision maintenance regimes on the railway, where monitoring – from fixed points on the infrastructure or on trains – analytics, prediction and algorithms are used to drive the next generation of railway engineering.
And it will all be based on data: how we capture data, how we collate it and how we present it to engineers so they can make the right decisions, at the right time, in the right place to keep train services running.
This year marked the launch of Network Rail’s Intelligent Infrastructure (II) programme, a five-year, Control Period 6 programme to deliver information derived from data that will tell the routes precisely where their assets are, how they are performing, how they are degrading and most importantly, when they are going to fail. Simply providing data is not enough, the challenge for Network Rail and the II programme is to turn that data into information that will allow engineers to prioritise and plan their opex and capital work bank and reduce the number of things they need to respond to urgently.
Working directly with the routes, the II programme will help the routes move from time-based, fix-on-fail maintenance, to information-led predict and prevent regimes; capture, analyse and exploit asset data to help the routes prioritise the most critical work and to drive a 10 per cent service affecting failure improvement. This will be achieved through the development of automated network monitoring, advanced data analytics, aerial surveys and integrated decision support tools. The programme’s workstreams include Track, Signalling, Ellipse (NR’s central database) Planning, Civils and Operational Property.
On-track to a digital future
The first significant output from the programme is the Enhanced Track Decision Support Tool (eTDST). Delivered in November 2019 as a prototype, it is now being evaluated by engineers in the routes.
Containing a range of key data sets, it presents information via a maintenance dashboard underpinned by decision support capability for track geometry (TG). It aligns with TG outputs from Network Rail’s fleet of measurement trains, identifying rates of degradation and highlighting sites at risk of reaching Alert Limit (AL) Intervention Limit (IL) and Immediate Action Limit (IAL). This allows engineers to filter information by TG parameters, engineers line references, track ID, start and end mileage, sleeper type and track category. The tool highlights where deterioration rates indicate a section of track will reach an AL within a 90-day window. This allows front line staff to have enough time to plan the right intervention at the right place at the right time allowing the business to shift from reactive to proactive regimes.
An algorithm developed internally within the II programme, aligns trace data from the measurement trains to provide maintenance teams with run-on-run trace that can be used to analyse and demonstrate asset deterioration over time. This replaces the current paper trace (see example) method of viewing and analysing track geometry data. The tool is mobile responsive on both android and iOS devices, allowing engineers access to the data they need from any location at any time across the network.
Go West and digitise
Dan Collins, Western infrastructure maintenance engineer, leads a team of track engineers in one of NR’s busiest routes.
Acutely aware of the current pressures and with a constant eye on the future, his ambition is to see the railway industry reflecting the asset management systems and regimes seen in the aeronautical industry.
He said: “The aeronautical industry has data management systems that tell engineers how long a wing or fuselage component will last. They know exactly when they should intervene and that’s precisely the level of knowledge and evidence the rail industry should work to if we want to meet the current challenge, let alone future rail growth.
“Up until now we’ve relied exclusively on our engineering knowledge to arrive at timescales when building a Risk-based Maintenance Regime, but too often it involves making some assumptions and our current data sources are very difficult and time consuming to use for trend analysis.
“With the enhanced predict and prevent capability the II programme is looking to deliver, and by exploiting the potential of decision support tools like the eTDST, we are reaching a point where we can trust the information given to us to carry out predictive maintenance – to actively fix faults before they impact our customers.
“We use algorithms to convert trace and fault data and by reviewing historic data in the tool, we can then predict at what point we will have to intervene with a repair. Sounds simple but the benefits are huge.
“From a safety perspective it means only sending engineers to undertake work that is required at the right time. When this tool moves into production, I will be able to build maintenance strategies from information the tool provides. We can see earlier what condition the track asset is in and the speed of degradation, therefore act earlier to correct issues before it becomes a major fault – which means more cost-effective work and a greater ability to maintain and improve our track geometry.”
Digital innovation is a priority for Network Rail and the future success of the railway industry relies on it if we are to improve track performance, boost capacity and drive safety. This transformation will drive the digital rail engineer of the future – supporting their expertise with data that confidently informs their maintenance and renewal decisions. Investment in this type of predictive maintenance is critical if Britain’s railway is to fulfil its ambition and commitment of putting the passenger first.
A single platform showing:
- Track Geometry information from measurement trains
- Multiple Run on Run Trace data demonstrating the deterioration rates of assets
- Prediction of when a defect will reach Alert Limit (AL) Intervention Limit (IL) and Immediate Action Limit (IAL)
- History of maintenance work on each asset to show the impact of previous work
- The ability to view degradation for every TG Parameter for every 200mm of track.
The tool will be enhanced and improved with further data streams throughout CP6 via a series of planned releases based on route consultation with engineers.