Much has been written about ERTMS (European Rail Traffic Management System) since its inception some 10 years ago and much more will emerge as worldwide rollout proceeds. The signalling suppliers message is that a successful system has been achieved, the earlier problems during system design having been resolved. But is this true and are the benefits that were promised being realised? The IRSE staged a seminar on the 15 November to examine the progress that has been made in the UK, in Europe and further afield, and to assess how the system will be steered into the future.

Progress review

Firstly, a brief recap. ERTMS, with its two main constituent parts, ETCS (European Train Control System) and GSM-R (the rail adaptation of the public GSM mobile radio standard), is all about getting a train control system that will work seamlessly across borders. It has three levels of application:

  • Level 1 – a pan-European Automatic Train Protection (ATP) system providing improved train safety as an add-on to existing signalling systems,
  • Level 2 – a complete train control system including Movement Authorities (MA), full speed supervision and train protection with emergency braking. It uses a radio bearer but retains some lineside infrastructure, principally axle counters / track circuits for added position monitoring and control of points and level crossings,
  • Level 3 – a totally radio based control system with only the minimum lineside equipment and with the possibility of introducing moving block.

Evolving Concept

An introductory address by Peter Stanley, who has spearheaded the newly published IRSE ERTMS book, reminded delegates what ERTMS is and, perhaps more importantly, what it is not. Emerging as a set of Interface Standards to create interoperability of onboard and lineside sub systems, ERTMS does not lead to common equipment platforms or common functional control units, nor will it lead to the replacement of route setting and route holding.

ERTMS has already impacted on a number of railway rules and processes:

  • Tasks – How infrastructure and rolling stock is managed including interdependence of activities, data consistency and accuracy, competence assurance and management,
  • Services – Train operations especially in part-equipped areas including capacity gains and junction optimisation,
  • People – Behaviour, procedures and problem resolution,
  • The Future – rolling stock strategies, train control strategy, service and asset planning.

ERTMS Proliferation

It is beyond doubt that ERTMS is making an impact in Europe and the wider world. Jacques Poré from Alstom revealed that since the first pilot scheme in 1999, some 35,000 kilometres of route and 7000 vehicles have been equipped, 50% outside of Europe. Some 70 million kilometres of operation has been achieved, increasing by two million kilometres each month. Whilst Spain, Sweden, France and China are currently the biggest investors, perhaps the progress to date has been dominated by three countries.

  • Switzerland. After early trials on the Lucerne – Olten line, two lines are now equipped with Level 2 for regular usage: Mattstetten (Bern) to Rothrist (Olten) on a newly built infrastructure, and the Lőtschberg Base Tunnel line. These projects have involved 4 suppliers – Alstom, Bombardier, Siemens and Thales – thus testing out the interoperability of equipment. Used by both passenger and freight trains, the lines operate at high speed with 110 second headways
  • Italy. Four high speed lines and 112 vehicles are now equipped for 300kph operation with a 3 minute headway. Alstom and Ansaldo have been the main suppliers.
  • Netherlands. The dedicated Betuweroute freight line, linking Rotterdam through Germany to the rest of Europe, plus three high speed / general lines are now in full Level 2 operation using three suppliers – Alstom, Bombardier and Thales. 116 vehicles have been equipped.

The perceived advantages of ERTMS were summarised as offering improved safety with up to 40% capacity improvement, reduced operational cost through improved reliability and lower maintenance, eco-friendly technology, an open market using the unique standard to give better life cycle cost, and a natural choice for high speed lines up to 500kph. This gives a positive message. But is it just ‘Sales Speak ‘ to promote business for the principal manufacturers?

Swiss role

The experiences of Switzerland have been watched closely by all. A positive message was given by Hansruedi Kaesar and Martin Zűrcher of SBB. Being a country at a crossroads in Europe, it needs to handle much transit traffic and obtaining increased capacity is a challenge. Passenger traffic has risen from 12 million to 17.5 million journeys between 2002 and 2010, with freight traffic also mushrooming.

ETCS is seen as an essential part of getting the needed train paths. The evolving strategy is to equip new lines with Level 2 and existing lines with Level 1, this latter to be completed right across the country by 2018. However, the Swiss do not expect to equip all lines with Level 2 until 2060. This clearly states the perceived longevity of ERTMS, although it is acknowledged that the hardware will need periodic updates from time to time. Expecting present day electronic components to be available in 50 years time is wishful thinking.

The training of drivers has produced useful ‘front end’ feedback on the system characteristics. Training is based around a mix of classroom tuition / simulation and on-train experience. Feedback shows that drivers appreciate a feeling of improved safety, less concern about weather and having relevant information for the journey. Adverse comments are on understanding the complexity of the system with too many types of MMI display.

A more cautious note was given by Martin Sűcher of SBB. Obtaining consistent software versions from four suppliers and delivered to ten train companies has proved to be challenging. Upgrading software in the future looks to be equally problematic. The current position is that if one party wishes to implement a software upgrade, then everyone else has to as well. Backward compatibility will become essential as ERTMS roll out will make it impossible to have every element of track or train borne equipment in different countries and across international borders at the same level of software.

System management becomes all important and a layered model for roles and responsibilities will be vital: a system manager to control European and country standards, an integration manager to oversee the application to train and trackside, an operating company (infrastructure manager or TOC) to control the supply of Products and Equipment. The overall experience in Switzerland has however been beneficial with delay attribution being almost non existent – 115 minutes / month = 0.014 minutes per train.

The UK Experience

The Cambrian early deployment scheme was reported in issues 74 and 79 of the rail engineer (December 2010 and May 2011). Nicola Furness of Network Rail and Peter Leppard of Arriva Trains Wales gave an insight into some of the problems found since full introduction in Spring 2011. The settling-in period saw train service performance plummet: down from 90% with RETB to only 60% with ERTMS. This latter figure has since improved. From an infrastructure perspective, management of speed profile and gradient data, the understanding and categorisation of SPADs / End of Movement Authorities, the control of ESRs and TSRs plus route barring for possessions have all needed new rules to be devised. Having the asset management data in imperial and ERTMS data in metric has not helped. The working relationship between infrastructure and train equipment maintainers was typically always to blame the other, but this has now got much better and an understanding of the total system requirement has emerged with good sharing of data.

From a train operator perspective, the system is more cautious than RETB. Setting up movement authorities out of a loop for two trains at the same time is a problem, as is allowing two approaching trains to occupy a loop at the same time. The lack of incorporation of AHB / AOCL level crossing control signals into the ERTMS system can give conflicting information to drivers and this is a serious concern. When questioned as to whether these shortcomings will be rectified, it seems unlikely in the short term as all efforts are being made to get performance up without the complication of further tinkering.

Retro fitting the Class 158 DMUs has been difficult with portions of ERTMS equipment being distributed all around the train. Getting the DMI display to be readable in both bright sunlight and at night took time to resolve. Cab equipment boot-up takes several minutes, which is a problem when splitting / joining trains at Machynlleth.

It must be remembered that the whole purpose of the Cambrian trial has been to establish how ERTMS would work in a UK environment by using a low risk route. With hindsight, perhaps the Cambrian was not the best choice as the ERTMS rules for main line application are probably too restrictive for a line such as this. Nonetheless, useful data and experience is being gained which will be put to good advantage on the next application – the northern end of the Hertford loop, intended as an integration test line.

The Future of Level 3

Peter Elestedt from Bombardier described the Level 3 trial in Sweden which was reported in issue 82 of the rail engineer (August 2011). This brave attempt to reduce lineside infrastructure to a minimum, using either radio or internet connectivity, should be watched carefully. Sweden intends to roll this out to many secondary lines where a 50% reduction in operating and maintenance costs can be achieved. The testing of the system has been completed and it is now in use operationally.

The vexed question of proving train integrity has to be resolved in the longer term as, without any track circuits or axle counters, the Level 3 system is not capable of doing this. It should be pointed out that signalling was never invented to keep trains coupled together – indeed, for much of railway history, train completeness was achieved by the signaller looking for the tail lamp! Much thought is being given to how integrity can be achieved without the complication of having to put some electronic device on the last vehicle. Ideas will be welcome as until this is solved, Level 3 has a limited future.

European TEN Corridors

One must not forget just how important ERTMS is for improving the viability of the Trans European Network (TEN) corridors. Stefan Wendel, the Programme Director for TEN-T explained that the Rotterdam / Zeebrugge to Genoa route is one such line which is intended to have full ERTMS capability by 2015. Transiting 5 countries, this is seen as achievable except perhaps for a small gap in Germany. An Operations Working Group is setting about harmonising the operating rules, seen as an essential precursor to effective ERTMS operation. Joint co-operation is taking place between the EC, ERA, User Groups, NSAs and UNIFE to ensure the best possible understanding of ERTMS usability and to give a joint commitment on achieving baseline-3 software before ERTMS through-working commences.

Traffic Management

It must be remembered that ERTMS has a third component, the Traffic Management System (TMS), aimed at optimising the throughput of train services. To date, little development has occurred on this. However, a total network roll out of ERTMS is to take place in Denmark based upon a business case of improved reliability and punctuality. Dr Felix Laube from Emch & Berger explained that by rolling it out quickly, and by being prepared to change the operating rules, the result should be a large increase in available train paths as well as a significant reduction in maintenance and staff expenditure. The cost will be €2.6 billion but this is cheaper than the civil engineering work needed to create more infrastructure. The project is big enough to attract international competition so economies of scale will be realised. In short, the railway has to be ‘reborn’ and ERTMS is the ‘transformational force’. As well as the increased capacity, huge improvements in passenger information and customer service are foreseen, with the signalling system being in tune with train company requirements. Certainly a brave commitment and the world will be watching to see how this project proceeds.


It is clear that ERTMS is here to stay. It has taken a long time to become a mature standard and even now there is plenty of on-going development to be done. Much of the implementation so far has concentrated on high speed lines where infrastructure and rolling stock have been part of a single project. The Swiss have striven hard to get ERTMS proven on a mixed traffic railway and the UK (almost by default) has tested the effectiveness of the system on a rural line. It seems that there are three principal concerns that must be resolved in the near future:

  • The provision, control and management of software. Getting new versions in place that are backward compatible with earlier versions will be essential,
  • The capacity of GSM-R. It is evident that the present bandwidth allocation and circuit switched connections will not be capable of handling ETCS commands in high traffic areas. Even if packet switching were to be introduced (GPRS), it would not be a complete answer. The solution is likely to be structuring GSM-R within a future mobile radio standard (perhaps LTE) and to then capitalise on the greater data handling capacity that this would give,
  • The need to have different operating rules for different types of lines. The safety features needed for a high speed line may be inappropriate for secondary and rural routes, where more flexibility in train operations will be required.

This IRSE seminar gave a realistic assessment of ERTMS. The system can now be judged as successful, which is just as well since there will be precious few alternative technologies to purchase in the future.