Digital rail’s barriers are designed in, and GBRX plans to design them out – Toufic Machnouk on ETCS, FRMCS and AI

You have described the rail sector’s technology adoption challenge as being rooted in complexity and inertia. How is your team working to overcome these barriers and accelerate adoption of technologies such as digital signalling?

The role of GBRX is to overcome the barriers to the adoption of strategic technology. Really, there are three underpinnings in the complexity and inertia story. One is that we have a highly fragmented commercial system. Even if you folded the entirety of the industry into one organisation, it would inherently remain fragmented. You have track-train boundaries, major system manufacturers and digital systems that blur the line between operational technology and traditional IT.

The second pillar is what I call designed-in barriers. Procurement regulations, safety validation, product approval and the network code all make sense and are there for good reason. But when you put them together, alongside that fragmentation, you can’t just stand on a pulpit and tell the industry to go forth and innovate. That’s not a system designed for innovation; it’s a system designed for safety and operational consistency.

The third pillar is capability. When you’re innovating and delivering frontier technologies, anything at the edge is, by definition, scarce. You therefore have to build capability very purposefully and carefully.

What we’re doing in GBRX is putting in strategic responses to each of these. For the designed-in barriers, we’re developing different validation processes, creating a more dynamic market to work around procurement constraints, and exploring new mechanisms within the network code. On capability, we’re building networks of expertise, external relationships and more unusual, inspirational partnerships.

You can’t just stand on a pulpit and tell the industry to go forth and innovate. That’s not a system designed for innovation; it’s a system designed for safety and operational consistency.”

The industry’s move towards digital signalling has progressed from the Pathfinder phase on the Northern City Line to the Pioneer phase with the East Coast Digital Programme. As the sector moves through Control Period 7 (CP7), what lessons will shape the wider rollout of the European Train Control System (ETCS)?

The Northern City Line was a key step in stabilising the technical application, and we identified four areas requiring further development: train modifications, safety validation, testing and driver training.

We want to overhaul the testing regime so that much more can be done off-site, rather than running trains up and down to verify authorities, which is time- and access-intensive. Our future aspirations on ETCS unit rates depend on minimising on-site testing.

At the moment, an ETCS scheme is conceived as a traditional project: design it, build it, test it, train drivers and move on. But when you’re delivering route renewals over thousands of kilometres, you need a more strategic approach: building driver competence ahead of deployment, transitioning quickly to no signals, and then extending the footprint.

Those four building blocks are being developed further through the East Coast Mainline deployment. It is a huge undertaking; we are integrating more original equipment manufacturers (OEMs) than any other ETCS scheme in the world. These are now being matured beyond the East Coast into what we call the national portfolio phase.

We have target areas including the Brighton Mainline, Midland Mainline, West Coast North and the TransPennine route. The objective is to modularise how we deliver digital signalling. For example, bringing OEM products to a level of maturity and maintaining a national testing lab means schemes can draw from proven solutions, rather than integrating everything locally each time.

Large-scale ETCS deployment will ultimately rely on the next generation of rail telecommunications as the industry moves beyond GSM-R. How important is the transition to the Future Railway Mobile Communication System (FRMCS) in enabling capabilities like data-driven operations, sensing technologies and artificial intelligence (AI)?

FRMCS is crucial in overcoming the obsolescence and limitations of a 2G GSM network, but it is much more than that. If you look at the broader limitations across the railway – communication, asset positioning, edge data and sensing on trains – each time you try to address one of these, you run into major constraints around how you move data off the train.

If you could click your fingers and deploy FRMCS today, it would create a radically different communications backbone for the rail industry. It is crucial, inevitable and necessary, not just to deal with obsolescence, but to create the telecoms footprint that unlocks wider technological objectives.

The challenge is that such a migration is not straightforward. The technology is still going through certification, and we will need to manage obsolescence extension while transitioning a network spanning over 15,000 kilometres. We are now building the strategic case and plans for how that migration would be delivered.

You have compared the challenge of rail transformation to solving a Rubik’s cube. How will initiatives like the upcoming Xplatform help suppliers, operators and technologists collaborate more effectively?

One of the issues we’ve identified, linked to fragmentation and designed-in barriers, is the disconnect between problem owners and potential solutions. When you look at the supply chain, there is a huge amount of innovation underway, but it takes real tenacity to navigate those barriers.

We’ve found a mismatch between what the industry is trying to solve and what the market understands those problems to be. So, we’ve developed the concept of a platform that effectively creates a dynamic marketplace.

There are legislative and procurement mechanisms within this, alongside a communication channel that allows us to clearly articulate problem statements to the market and for the market to respond with ideas or propositions that we can properly assess.

By creating this model, we can engage earlier and, where something is viable, move more quickly into longer-term commercial arrangements without restarting the process each time.

We’re retiring around 5% of the workforce and bringing back only 1.5%. That’s about half the return rate needed to sustain the industry.”

Migrating to a digitally signalled railway requires new skills. What steps are being taken to build capabilities across the sector, such as the data science and AI apprenticeship programme?

The narrative on skills is quite stark. We’re retiring around 5% of the workforce and bringing back only 1.5%. That’s about half the return rate needed to sustain the industry.

If you redesign workflows using AI agents, you create a suite of tools that still need to be configured, maintained and evolved. It’s not that there are fewer people; it’s that people are doing different things, and being freed up to focus on higher-value activities in areas like safety-critical assets, industrial planning and deep engineering.

We have to build capability in these areas. The apprenticeship programme we’ve launched around data and AI is a prime example. It’s a 10-year horizon. Through GBRX, in partnership with the National Skills Academy for Rail (NSAR), we are building that capability centrally, removing hiring barriers and deploying skills where they are most needed.

If a single operating organisation with 7,000 to 10,000 people hires two or three data scientists, they get lost. This model allows us to build capability at scale and rotate it across the industry over time.

Looking ahead 10 years, what does a successful transformation of the UK’s digital railway look like across digital signalling, AI and emerging technologies?

We want digital signalling to reach a tipping point by the mid-2030s. The East Coast programme starting migration this year is the first step into that phase. The next step is establishing national building blocks and progressing further schemes.

On AI, the UK Government asked us to develop the sector’s action plan. We convened the industry, bringing together an expert advisory group, and developed the plan through a combination of bottom-up input and top-down guidance, assessing maturity, identifying early opportunities and defining pathfinders and enablers.

One key outcome is the development of an AI incubator and accelerator to apply models within the industry, with the appropriate assurance, governance and controls. We’re already developing use cases across pricing and ticketing, business workflows, and asset management. Within five years, I want that to be embedded as core business.

In areas like quantum engineering, we’re aiming for breakthroughs. Quantum navigation systems could fundamentally transform train and asset positioning. We recently completed what is believed to be the world’s first trial of this kind on a national railway. It is now moving from research into application engineering.

It’s about iterating, testing, refining and solving challenges such as device size, sensitivity, interference and drift. Over the next five to 10 years, we want to reach viable application and begin transforming positioning technologies in a meaningful way.