The railway interoperability challenge

Since the middle of the 19th century, railways have always held an important place in European policy-making. They still do. However, the goals have changed. Whereas in the past, industrial and military national elements were influential in the development of the railways, today’s society is confronted with new problems on a European and global scale. Influence on the environment and changes of climate do not stop at borders between states.

In the past, differences were created deliberately between the railway networks of the various countries in Europe. Today, we look at these differences as an obstacle to a seamless international railway network that can compete with air transport. Interoperability (or the lack thereof) is a modern way to express these conditions. Complete interoperability is a goal that cannot be reached in the short-term because of the huge amount of investment required to replace infrastructure with deviating specifications. But much can be achieved in a short timescale with minimal investment. The challenge is to find the most effective way to improve interoperability.

The current plan is mainly a practical approach, but it does not define a complete introduction of interoperability over all the European passenger and freight network. It is restricted to the high-speed network in as far as new lines are mainly built for high speed, and it incorporates the freight corridors. Another example of this pragmatic approach is the development of cross acceptance principles for existing rolling stock. But what the goal will be after the corridors is something that has yet to be decided.

ERTMS

In the first phase of the movement towards interoperability, a number of legal instruments were developed: the Directives for Interoperability and the Technical Specifications for Interoperability (TSI). Although the set of specifications is not yet complete, considerable progress has been made. Most importantly, development took place in the area of control, command and signaling. Today, the specification for the ERTMS/ETCS system is sufficiently stable to be used in most projects in Europe currently under construction and being put in operation. Of course, the choice for the version of the system specification (2.2.2 with a number of Change Requests or 2.3.0) influences the result of any project, and the recent UIC World Conference on ERTMS that took place in Bern has shown that arguments exist for either choice. Maybe it will be necessary to later update software versions of trackside and onboard ERTMS/ETCS equipment, but the level of incompatibility is substantially reduced with the start of the ERTMS/ETCS implementations.

Lessons from the first interoperable projects

Looking at the progress of interoperability, it may be concluded that ERTMS/ETCS is indeed the cornerstone. More traditional technologies, like track construction and those related to the catenary and power supply, do not limit the compatibility over large parts of the network for modern rolling stock. This is mainly due to technological development in rolling stock, like multi-current locomotives and automatic gear changing equipment. Fitting a locomotive with multiple pantographs is simpler, but certainly cheaper than adapting catenary construction all over Europe.

This brings us to a problem that is not completely resolved and sometimes hinders development of interoperability: the additional cost to achieve pragmatic interoperability is not always evenly shared between the railway operators and the infrastructure managers. Sometimes solutions have been found, but no general model seems to exist.

The European projects that come into operation show another effect of the open market and separation policy. Many projects for new or upgraded lines seem to struggle (more so than in the past) with the fact that the characteristics of the rolling stock that will run on it are not known. It is understandable that the many interfaces between the track-side and the train-side need careful checking, but it is hard to check the line-side of the interface without knowledge of the train-side. This problem can only be solved by a strict application of the uniform interface requirements from the TSIs on both the track-side and the train-side. This is the real interest of the TSIs. It is also the real interest for certification of the track and train-sides.

Safety

The most frequently used motivation to explain national differences in railway specifications and railway operation is safety. Apparently, the judgment about what is safe is different in the member states. The European Commission has issued the Safety Directive to address this problem and the European Railway Agency has the mandate to draw up common safety targets and common safety methods. What still seems to be missing in this approach is common safety validation. The TSIs do not address the railway system as a whole, whereas the member states look at the complete railway system: track, train and operation. Member states enter national safety requirements at the level of the complete railway system.

Related to the safety validation process, an important observation is the difference between the TSI verification model and the verification and validation model described in the CENELEC standards 50126 and 50129. These two standards are widely used in the railway sector. Originally, they where meant for the development of software-based safety systems, but gradually these standards have found usage also for other technologies and other processes. They are not only applied for the development of equipment, but it goes up to the complete railway system and all levels in between. Two measures could improve the progress of interoperability:

• Extending the approval process to cover the complete railway system.
• Lining up the CENELEC standards EN 50126 and 50129 with the interoperability Directives and TSIs.

Interoperability is often approached as a technical problem of compatibility, but this overlooks the importance of operations – demonstrated by the low profile of the TSI operations. As part of the safety approval process, verification of the implementation of the operational rules should be required. In particular, the ERTMS projects make clear that the only way to prove the safety of the control, command and signalling subsystem is to assume that in the operational phase, certain conditions will be met. In practice, this means that much of the risk control is transferred to the operational phase. The locomotive driver still carries an important responsibility for safety, maybe even greater than before, because new processes are introduced. These points must be addressed. The return of experience of the new ERTMS lines must be organized in such a way that it helps to improve the TSIs and limit the amount of safety assumptions and conditions for operation. This will also limit the number of operational rules and harmonise them to a greater extent.

This is where the real challenge of interoperability lies: learn from the projects to introduce and develop harmonisation only where it is needed.