THE development of new technology does not always achieve the objectives set out at the beginning of the project, and the European Rail Traffic Management System (ERTMS) is a case in point. The original concept for ERTMS when it was launched in 1989 was to develop a train control system which could be overlaid on the multiplicity of signalling systems operating in Europe so that trains would only require their national signalling system and ERTMS to enable them to cross borders with ease.

While ERTMS was being developed and tested in the 1990s and the first half of the 2000s, there was a move in Europe away from traditional locomotive-hauled trains where the locomotives would simply be changed at the border, to fixed-formation distributed-traction multi-system passenger trains such as Eurostar and Thalys and multi-system locomotives for freight trains. These trains and locomotives can be packed with up to half a dozen different signalling systems which makes the installation of yet another system, namely ERTMS, a real challenge for operators both physically and in terms of cost, especially as the deployment of ERTMS has not advanced sufficiently to allow some of the legacy signalling systems to be removed.

European infrastructure managers are national institutions which are funded and regulated nationally, and they are not used to or very good at cooperating with one another. While most European infrastructure managers have developed or are planning ERTMS deployment strategies, these are national plans designed to achieve national objectives and timescales. The result is that plans to implement ERTMS are proceeding according to very different timetables in each country.

The attitude to the European Train Control System (ETCS) element of ERTMS also differs widely between countries. Italian Rail Network (RFI) has embraced ETCS and its new national signalling system on the mainline network is designed with ETCS in mind as it can be upgraded relatively simply. RFI also had sufficient confidence in ERTMS to install it on its new high-speed lines without a backup system, while other infrastructure managers such Adif in Spain have insisted on retaining their traditional signalling, thereby adding complexity and cost. It is interesting to note that RFI was able to get ETCS Level 2 to function reliably with relatively few problems while other infrastructure managers struggled. Perhaps this is due to the realisation that it had to work, which concentrated minds in Italy.

Two of Europe's biggest rail networks - France and Germany - have largely eschewed ETCS up to now favouring their own tried and tested systems such as TVM 430 on French high-speed lines and LZB on German main lines. Conversely some smaller countries have decided to install ETCS nationwide. Denmark and Switzerland are replacing their entire signalling systems with ETCS because their legacy signalling is life-expired, while Belgium is adopting ETCS nationally to improve safety following a major accident.

ETCS II cab-displayUp to September last year European infrastructure managers had contracted to install a total of 38,845km of ERTMS, compared with 29,111km in the rest of the world of which 11,000km is in China. But in Europe this still represents a patchwork of projects with hardly any schemes crossing borders to provide the interoperability needed to make the European rail network more cohesive.

Frustrated by the slow progress to achieve a meaningful ERTMS network in Europe, the European Commission (EC) decided to make funds available to equip the TEN-T corridors with ERTMS as a way to kick-start a Europe-wide deployment. But the results so far are patchy as an EC report on ERTMS deployment published in February makes clear: "Despite the significant provision of EU funding, deployment on the corridors is behind schedule," says the report. "Worse still, delays in one member state affect the entire corridor, which means for the others a delay in realising the benefits of their investment."

The report says notifications received so far show that some member states are keeping to the ERTMS deployment plan, while some forecast delays of up to three years, and a few - in particular larger member states - have major delays. A corridor-by-corridor update on ERTMS implementation illustrates the problem clearly:

• Corridor A running from Rotterdam to Genoa will be completed on time in 2015 in the Netherlands and Switzerland, whereas it will be three years late in Germany, which is in the heart of the corridor, and five years late in Italy at the southern end

• Corridor B linking the Swedish capital Stockholm with Naples in southern Italy should be equipped by 2020, apart from the Austrian section which is ready and the German section, which will not be completed until 2030

• Corridor C from Antwerp to Basle will not be operational in France until 2020-2023 while Belgium and Luxembourg will complete their sections on time in 2015

• Corridor D connecting Valencia in Spain with the Hungarian capital Budapest will not be operational in France until 2021, which represents a six-year delay, whereas the Spanish, Italian, Slovenian and Bulgarian sections will all be on time, apart from the section between the French border and Turin where there is a five-year delay

• Corridor E from Dresden, Germany, to Constanta on Romania's Black Sea coast has various delays of up to five years, and

• Corridor F will be completed by 2020 on some sections in Poland but not until 2027 in Germany.

"In the light of available evidence, the EC is not convinced that delays are beyond member states' control or that appropriate measures are being taken to minimise the delays," says the report, although it does acknowledge that the recent difficult economic situation may be a contributory factor. Nevertheless, such widely differing completion dates make a mockery of the deployment plan.

The EC says that for the next annual update it expects complete and detailed reports with improved information, in particular better justifications for the delays and appropriate measures aimed at reducing them.

Even when ETCS is installed on adjacent networks, this in itself does not guarantee interoperability because of national add-ons which make neighbouring systems incompatible and because of variations in software baselines.

A lively discussion of these topics took place at Global Transport Forum's new Smart Rail event in Amsterdam last month. Mr Libor Lochman, executive director of the Community of European Railways and Infrastructure Managers (CER), set out his priorities to improve the situation. "We have to achieve interoperability in Europe and I hope the new Baseline 3 will bring us to this," Lochman told delegates. "There should be no national add-ons which prevent interoperability, manufacturers must supply what is asked for, and the national safety authorities (NSAs) need to be controlled." Lochman later told IRJ: "The NSAs aren't accountable to anyone. They can demand changes to ERTMS to meet their own national needs, but which then makes it uninteroperable."

Mr Michel Ruesen, managing director of the ERTMS users group, believes that ERTMS is generally successful but acknowledges that there are problems which need to be addressed. "Performance with ERTMS is normally better than what was there before," Ruesen said. "ERTMS has proved much more complicated than envisaged. Yes, ERTMS is expensive, but the older systems are also expensive.

"Baseline 2.3.0d is stable and is good for high-speed projects. With Baseline 3 we will encounter more challenges which need to be solved, but freezing the specifications in themselves doesn't help. Any changes we make must be done in a controlled way." Ruesen said for the remainder of this decade the emphasis must be on deployment and correcting any problems which arise, rather than adding any new functions.

"We don't even have a baseline yet which is interoperable," Lochman added. "We know what to do, but we have to launch it."

Ms Monica Heiming, executive director of the European Infrastructure Managers (EIM) association, was less forgiving in her assessment of the situation. "ERTMS should work much better. We shouldn't still be struggling with interoperability and inter-availability. We will have a serious problem if we don't get the DNA of Baseline 3 right. We have to stand together to get it right now, and not in five years' time."

Despite these problems and the challenges that lie ahead, particularly the replacement of GSM-R, the deployment of ERTMS will accelerate in the future as older signalling systems reach the end of the life and are replaced by ETCS. But it will take a decade or two before ERTMS reaches a critical mass in Europe and we can start to reap the benefits of an interoperable rail network, nearly 40 years after the concept was launched.