It is now 25 years since the European Union drew up plans for a common signalling and train control system for Europe, and while GSM-R, the telecoms element of ERTMS, has been rolled out successfully across Europe, progress with ETCS has been painfully slow and there are still major issues which need to be resolved.

The release of Baseline 3 software for ETCS Levels 1 and 2 is a major achievement as it should bring stability to the system and give more confidence to infrastructure managers planning to install ETCS. However, several speakers called for a freeze on the specification for Baseline 3 to aid deployment, with one speaker arguing it should be frozen for 10 years. Whilst a temporary freeze would be welcome, it is unrealistic to consider freezing the specification for several years given the need to correct any major faults and the pace of software development. It would also hold back the necessary evolution of ETCS.

Another concern is the lack of off-the-shelf ETCS products, resulting in high costs which in turn jeopardises the business case for investing in ETCS, particularly for private operators who simply cannot afford the huge cost of retrofitting a locomotive, which can range from e300,000 to e1m. This needs to be addressed urgently by manufacturers and perhaps the arrival of Baseline 3 will spur them into action.

Another issue is the failure to develop a high-density ATO version of ETCS comparable with CBTC which is starting to be used by mainline railways on urban lines where a high frequency and sometimes automation is required. Italian Rail Network (RFI) is addressing this by developing its own solution, while in Mexico Alstom is pioneering ATO over ETCS Level 2 on a new commuter line.

Mr Michel Ruesen, director of the ERTMS Users' Group, said this concept is still under development in Europe due to the need to be able to apply it to a whole network rather than a single line.

This requirement to develop ETCS as a fully interoperable system is one of the main reasons why the deployment of ETCS has been so slow in Europe and explains one of the paradoxes of ETCS that it finds greater favour outside Europe than within.

Another factor is the reluctance of networks in Britain, France, Germany and Poland to adopt it. While Spain has installed ETCS on its high-speed network it usually maintains a backup conventional system which increases costs. RFI has sufficient confidence in ETCS to install it without a backup on all of its new high-speed lines. Italy's national SCMT train control system was developed with ETCS in mind which has made it easier to produce a comprehensive strategy for deploying ETCS on different types of line (page 36) as well as a low-cost satellite version for rural lines.

Conversely, the smaller railway networks - Belgium, Denmark, Luxembourg and Switzerland - are pushing ahead with system-wide deployment of ETCS. Here, the rationale is either to boost safety, as with Belgium, or because the conventional signalling system has reached the end of its life and needs replacement anyway.

Hopefully, it will not take another 25 years to see major progress with the deployment of ETCS across Europe in order for railways to reap the benefits of the original concept for an interoperable network.