Switching Britain’s complex rail network over to a digital signalling system is a long and complicated undertaking, but it opens the way to a more efficient and streamlined system. Tim Clark speaks to Scott Kelley, managing director, signalling at AtkinsRéalis, about the challenges involved
It may not be well known, but digital technology has existed in some shape or form since at least the second world war. Designed to help communicate faster than the Victorian-era telegraph, it now underpins most of the forms of communication we know of today.
Except, that is, for the railway. Despite billions spent and countless upgrades, much of the thousands of miles of rail in the UK is still reliant on – if you simplify it down – a system of traffic lights.
This issue is, however, finally beginning to change.
In September, Network Rail unveiled the firm that will work towards digitising parts of the rail network under a £4bn, 10-year framework. The firms include Alston, AtkinsRéalis and CAF, Siemens, and Thales VolkerRail, who will work to use the European Train Control System technology (ETCS).
Interestingly, the framework includes £3bn for digital signalling (Lot 2), and £1bn for conventional signalling (Lot 1).
Conventional signalling currently makes up about 99% of the UK’s rail network. Train drivers rely on track-side signals to determine speed, stops, and congestion.
“Generally the key way a train is controlled under that system is by a driver looking out the window, seeing a signal, and if that signal is red, amber or green or double amber, that tells him some information, or her some information about the congestion ahead,” says Scott Kelley, managing director, signalling at AtkinsRéalis.
While these systems have been computerised, the system is still the same as it was half a century ago.
“We invest an awful lot in conventional signalling, making sure that it is as efficient and as up to date as possible, using a whole host of digital tools along the way,” Kelley adds. “But ultimately, it does rely on a driver seeing a signal in front of them.”
Block party
The main issue with traditional signalling is that the trains run on blocks of track which are designed to give the network enough space to operate safely. Only one train is allowed in a block at a time and a combination of signals and interlocking mechanisms ensured that trains could not enter an occupied block. This significantly reduced the risk of collisions; it also significantly reduces capacity.
Aside from increased capacity, other benefits include improved safety, enhanced efficiency due to real-time data on train positions, and reduced infrastructure maintenance as signalling will be moved into train cabs, rather than track-side equipment.
Kelley noted that the reliance on visual signals remains a limiting factor, particularly in high-traffic areas, as drivers must interpret signals designed for worst-case scenarios.
What digital signalling can do is transition control of how and where trains can go from the direct visual cues of the driver to a screen in the train’s cab. The ETCS system also allows real-time speed and congestion data directly to drivers, allowing for dynamic adjustments based on the train’s characteristics, track conditions, and network congestion.
“Effectively you always want to have a block between trains but with digital signalling, you can change the size of those blocks, in theory, and have more trains running closer together, because you can reduce the line speed of the train if there’s a lot of congestion,” Kelley adds.
“If, similarly, there’s no congestion, you can increase the line speed of that particular train.”
This flexibility increases rail capacity and improves safety by reducing human error, Kelley says.
“It offers the potential to increase the safety of railways where specifically you have mixed rolling stock and so on because effectively, the train control system doesn’t rely on a driver seeing something and taking an action,” he says.
“That is embedded in the system. So if the driver chooses to ignore the information that the train controller is telling him to, the train will take action on its own, so it will not allow the train to go into an unsafe situation. It will take control of that train, and if a train is behaving strangely, then the system allows the trains behind to take action ahead of the problem.
“So it’s a much higher safety integrity level. And it is. It does take a lot of the human interface out of it.”
So far so good. The big issue is trying to adapt a conventional system to a digital age while keeping trains running at the same time. At present routes that use the digital system include the Cambrian Line, which was upgraded in 2006, and also Crossrail. High Speed Two is also mandated to use the digital system once operational.
For digital signalling to succeed, factors such as timetabling, passenger information systems, and staff logistics must align with the new technology.
When it comes to the rest of the route, Network Rail’s plan looks to be to take it slowly; at least for now.
Building digital
Schemes that will include digital signalling include the Midlands Rail Hub, the TransPennine Route Upgrade, the TriLink programme, and the Brighton main line.
The East Coast Digital Programme (ECDP) is another initiative which will replace traditional systems with ETCS from London to York. Siemens has been given the role of replacing signalling on the route. AtkinsRéalis acts as the railway systems integrator for the project.
The roll out of ETCS has been formulated into Network Rail’s Long Term Deployment Plan (LTDP), which was published in 2019. The plan aims to outline how digital signalling will be rolled out from what was Control Period 6 (the name given to Network Rail’s five-year capital investment period from 2019 to 2024), and the current CO7 which runs until 2029.
Analysis undertaken on behalf of the LTDP found that the maximum volume of works deliverable for digital and conventional signalling schemes (so maintenance or replacement along a section of track) is 3,000 SEU, or Signalling Equivalent Unit, per year if it is digital, and 1,800 SEU for conventional signalling respectively. The reason more track could be replaced digitally is because it took up less line-side equipment.
The UK rail network has over 19,000 miles of track and 4,000 trains, meaning that in reality, the complex nature of the UK rail network will take over a decade even if it was at the fastest pace possible.
Costs were also highlighted in the report, with the report noting that “given the relative infancy of implementing digital signalling technology within the supply chain, there is an opportunity for investment in R&D to make a digital plan more affordable”.
It added that across Europe, costs were benchmarked to around £190,000 per SEU, which would work with budget levels.
The new framework will see the rail suppliers adapt existing ETCS systems from Europe for UK standards. This involves rigorous testing and homologation processes to meet Network Rail’s requirements.
“What Network Rail have committed to doing is paying a proportion of our development costs to take these digital signalling systems that exist across Europe, bring them to the UK and effectively try to improve or homogulate them for use in the UK,” Kelley adds.
A large proportion of the work will include simulating the rail line to understand how it performs under the new signalling. The aim is to ensure that the signalling systems work before anything is installed on the actual system.
The idea of the digital railway is something we should be able to provide initially alongside the conventional signal before we turn it off and be confident it works, having tested it in a full synthetic environment, first fully validated it, and at that point it will get turned on,” Kelley says.
“The intention is that all new trains are provided with the infrastructure on board to work with it. The challenge you’ve got is the older rolling stock that’s on the system will require a level of retrofitting, because a driver will no longer be able to look out the window.
“If you imagine you no longer have to install signals and maintain signals that convey that information, and all of that stuff, all of that moves into the cab outside of the railway infrastructure. So it should be a lot less disruptive for the network.”
Scale of transition
Each supplier on the framework has been assigned an initial project with the aim of fostering a gradual rollout of digital technologies across the network.
Infrastructure adaptation is another challenge. For example, densely trafficked areas like East Croydon or Liverpool Street Station demand physical changes to manage congestion effectively, even with digital systems in place. This dual requirement for technological and structural upgrades underscores the scale of the transition.
While initial projects like the East Coast Main Line serve as pilot implementations, scaling these efforts across the network will require careful prioritisation. Lines with the greatest potential for capacity and safety improvements will take precedence.
The UK’s approach differs from European counterparts, who primarily implement ETCS on new lines with closed rolling stock systems, such as the TGV lines in France.
How long could all this take? Network Rail’s report suggests a 30-year time period, however those dates could change as funding allocations change over different control periods.
The LTDP report noted that digital signalling would bring operational benefits “that can be hard to quantify at present but should not be overlooked”. It stated that over a 30-year period, the total benefits could range from £7bn to £18bn, for the worst and best-case scenarios respectively.
The transition to digital rail demands will also require a new skill set, not only for engineers but construction firms working with the rail industry. According to Kelley, AtkinsRéalis has seen a growing need for data analysts, software developers, and systems integrators who can manage complex simulations and integrations. Traditional railway construction skills remain relevant for physical infrastructure upgrades, but digital expertise will play an increasingly central role.
AtkinsRéalis has adopted modular and simulation-driven methods for conventional signalling projects, paving the way for digital implementation. Pre-testing components in controlled environments ensures a higher degree of reliability, reducing risks during installation.
What could 2025 bring? According to Kelley, establishing a unified baseline for digital signalling standards will be a key marker point for the overall programme. Maintaining compatibility with existing ETCS systems used in Europe is critical to avoid unnecessary complexity and costs.
“For us, it’s about starting that development and working with Network Rail to really understand the [project]. At the moment, what we’ve got to try and avoid is making this [signalling system] too specific to the UK to really get the benefits from the system,” he says, adding that if UK firms are tempted to change the signalling systems from the current models used in Europe, they risk over-complicating the project and losing benefits it should gain.
“Our challenge over the next year [is] to really understand with Network Rail how we can cause least disruption and make the development process as low friction as possible,” Kelley says.
The transition to digital will occur over multiple decades and will require strategic collaboration, rigorous testing, and incremental implementation. While the road ahead is long, the promise of a more efficient, safer, and sustainable railway system makes the effort worthwhile.
Main image: Scott Kelley, managing director, signalling at AtkinsRéalis
Read next: Ocean Winds extends contract for Atlantic project
Are you a building professional? Sign up for a FREE MEMBERSHIP to upload news stories, post job vacancies, and connect with colleagues on our secure social feed.
