Outputs From The 2026 Conference

Observation: planning discipline becomes more important as uncertainty increases.

Dr Florian Kappler (BLS AG) described the challenge of migrating a live, mixed-use network.

To put that in context:

• 71.5 million passengers annually 
• ~20,000 freight movements 
• A fully operational system, not a clean deployment 

ETCS was presented not in isolation, but as part of a wider system including:

• FRMCS 
• ATO 
• Digital operational processes 

Some of these elements are still evolving, which introduces uncertainty into planning.

Rather than fixing a single path early, different scenarios were tested:

• Rollout speed 
• Corridor strategies 
• Lifecycle considerations 
• Integration with future systems 

These were then narrowed into a preferred approach, balancing cost, risk and impact.

A point that came through clearly — both in the presentation and discussion — was the need to align:

• Rolling stock 
• Infrastructure 
• Communications 

Early in the process.

This is straightforward in principle, but more difficult in practice when not all end states are defined.

There was also a more reflective point around cost.

Short- to medium-term optimisation is relatively manageable.

The more difficult challenge is ensuring that decisions taken now do not lead to repeated upgrades or rework later.

Where that longer-term view is not built in, cost tends to reappear over time.

Kjell Holter of Bane NOR presentation related to a living example of the digital railway in operation. 

The approach integrates ETCS within a broader digital architecture, combining signalling, traffic management, and communications systems. Strong emphasis is placed on system and operational standardisation, replacing fragmented legacy systems with unified functionality.

Pilot and early deployment lines demonstrate high safety, availability, and low maintenance, supporting the case for ETCS Level 2 without signals. However, technical challenges remain, particularly in odometry, where GNSS and radar solutions each have limitations.

The programme also looks ahead to future capabilities, including ATO, FRMCS, and train integrity.

Overall, Norway’s experience shows that ETCS can be successfully deployed at scale, but requires strong integration, standardisation, and ongoing technical adaptation.

Thijs van Steen, Programme Director for ERTMS in the Netherlands, emphasised that:

ERTMS is not just an infrastructure programme.

It is the coordination of an entire ecosystem:

• Passenger rail 
• Freight 
• High-speed services 
• Infrastructure managers 
• Government stakeholders 

— all under conditions of uncertainty.

Governance Model

What stood out was the governance approach:

Not:

• A single centralised authority 

And not:

• A fully decentralised system 

But a structured, multi-layer model, involving:

• Ministry level 
• Executive board 
• Programme directorate 
• Infrastructure manager 
• Operators (NS, freight, regional) 

How Decisions Are Made

Decision-making is:

• Layered 
• Structured 
• Adaptive 

With clear principles for:

• Intervention 
• Escalation 
• Progress tracking 

One of the most striking insights:

“Culture eats strategy for breakfast.”

The implication being:

Success depends less on technical execution, and more on:

• How effectively organisations collaborate 
• How decisions are made across boundaries 
• How uncertainty is managed collectively 

This session provided a useful shift in perspective, moving away from infrastructure and system design, and focusing on what is often treated as a secondary consideration — how drivers and the supporting operational workforce  actually adapt to ETCS in practice.

Rafal approached this less as a training topic and more as a transition in how the railway is operated at a human level.

The central point he made — and it’s one that is easy to underestimate — is that ETCS does not just introduce a new interface.

It changes the role of the driver.

Moving from:

• interpreting trackside signals and physical cues 

to:

• operating through a digital in-cab system, where information is mediated differently 

It is a shift in how decisions are made, how attention is managed, and how confidence in the system is built.

The cost of training came through quite clearly.

Training a single driver can exceed £100,000, depending on the approach taken and the level of simulation involved.

But the more important point was not the headline figure — it was what drives that cost.

In particular:

• time to competency 
• dropout or retraining 
• inefficient sequencing of training stages 

If those are not managed well, the cost increases quickly, and the impact is not just financial — it affects operational readiness.

Simulation as Part of the Programme, Not an Add-On

What SimFactor is proposing is to treat simulation differently.

Not just as a training tool, but as something that supports the wider programme.

In practice, that means using simulation earlier and more continuously:

• at the selection stage, to assess aptitude 
• during initial familiarisation, before formal training begins 
• throughout training, to build confidence progressively 
• and even into operations, as a way of maintaining competence 

There was also a useful distinction made between different types of simulation.

Not everything needs to be a full cab environment.

Depending on the stage, different tools can be used:

• lighter digital environments for early familiarisation 
• more structured simulation for training 
• full cab replication where required 

This is partly a cost consideration, but also about using the right level of fidelity at the right time.

This provides a snapshot of the themes and pressures that shaped the two days—but it only begins to reflect the depth of insight shared across both the formal sessions and the more candid discussions around them.

What emerged is a picture of a European landscape that remains, by global standards, highly advanced in European Rail Traffic Management System deployment. At the same time, there is a growing recognition across different parts of the ecosystem that delivering the next phase at scale will require further alignment—particularly as programmes move from early implementation into repeatable, system-wide rollout.

Rather than being a question of capability or intent, the discussion is increasingly centred on how existing processes, frameworks and commercial models can continue to adapt in line with the realities of large-scale deployment—across both passenger and freight environments, and across networks with very different starting points.

One area that was consistently referenced—albeit from different perspectives—was authorisation.

There was a shared sense among many contributors that, while the current frameworks are grounded in important safety and interoperability principles, there may be opportunities over time to explore how they can operate more efficiently in certain contexts. This is particularly relevant for iterative onboard changes, where some participants noted the cumulative impact of repeated reauthorisation cycles on cost, timelines and programme confidence.

In that light, several constructive lines of thought emerged, including:

• The potential to increase reuse of testing evidence and documentation where appropriate 
• Continued exploration of more proportionate, risk-based approaches in specific scenarios 
• Greater clarity and consistency around how different types of changes are categorised and assessed 
• Ongoing dialogue around how recognition between national authorities could evolve over time 

These are not simple questions, nor are they new—but they appear to be gaining renewed focus as programmes scale.

Alongside this, there was also interest in how assurance models themselves may develop. For example, some contributors pointed to the role that more modular system architectures could play in supporting more flexible and reusable assurance approaches. Others highlighted the possible value of expanding laboratory-based testing capacity as a complement to existing field-based processes—particularly where it can improve efficiency without compromising robustness.

The freight discussion brought a slightly different set of considerations.

Participants highlighted that, in some cases, the commercial drivers for onboard investment can look different from those in passenger programmes. As a result, there may be value in continued exploration of how funding approaches, incentives, or cost-sharing models could evolve—particularly in freight-heavy environments where the wider system benefits are clear, but the individual business case can be more complex.

Finally, there was a noticeable shift in how future upgrades are being thought about.

Several contributors reflected on the potential for more software-led upgrade pathways over time, which could help reduce operational disruption and create more flexible long-term upgrade cycles—although, as always, this brings its own technical and assurance considerations.

What the event ultimately highlighted is not a single conclusion, but a shared direction of travel.

ERTMS and ETCS are progressing—and will continue to do so. The focus now appears to be on how delivery models, processes and collaboration frameworks can continue to evolve alongside that progress, in a way that supports scale, maintains confidence, and reflects the diversity of operational realities across Europe.

But even at a high level, the tone of the conversation was clear:

There is strong momentum—combined with a growing openness to refine how that momentum is delivered in practice.

The full conference output—covering detailed case studies, extended Q&A discussions, and a broad set of recommendations—explores these themes in greater depth and are available to download on this website.

Steve D Thomas 1 April 2026