Commemorating a centenary should always be high profile and the Institution of Railway Signal Engineers did justice to this milestone with its Aspect Conference and Annual Convention in London over the six days, 10 – 15 September.
The temptation is always to reflect on past glories and achievements, and whilst some past events were recalled, the thrust of the conference was to look to the future but learning from the lessons of the last 100 years. This was done with a worldwide perspective, the IRSE being one of the most international of Institutions. Almost half of its membership has a non UK address and both paper presenters and attendees represented many of the countries where the IRSE has influence.
The two events looked at Signalling and Telecommunications technology both in mainline and metro applications, safety in its various guises, capacity and associated signalling initiatives, maintenance philosophy and methods, projects and system engineering, human factor issues plus some visions for the future.
The capacity challenge
Lack of capacity is not a new problem for railways. Oskar Stalder from Switzerland described the changes at Zurich Oerlikon station over the past 100 years where a tenfold increase in traffic has meant considerable growth of the station layout. Signalling has been updated from mechanical frames to relay interlocking and finally to a computer based system to cope with all of this. More traffic demands more information and an IP network provides the data gathered from the surrounding area to ensure train operation is efficiently managed.
Similarly, Charles Lung described the rise of rail transport in Hong Kong over the same period where the impact of war, political change and population growth led to the construction of the Mass Transit Railway Corporation (MTRC) and major upgrades to the Kowloon-Canton Railway (KCR), both of which merged into a single business in 2007. Five more expansions are either underway or planned and high speed train services now link the province with Beijing and beyond. All lines are equipped with both ATP and ATO from a number of signalling suppliers around the world, vital to handle the 4.3 million passengers carried in the 19 hour day period.
Many railways with capacity problems look to the signalling engineer to maximise throughput on existing lines, thus minimising expensive civil engineering infrastructure enhancements. Trevor Moore from the Australian Rail Track Corporation described how computer-based signalling, as developed by Lockheed Martin, was revolutionising non- urban operations. With the radio-based system proven for service in 2012, it will enable a step change in train operation. Using the public 3G network for communication, it requires minimal track equipment, provides bi-directional operation on double track lines, fleeting of trains on single lines, variable length virtual block sections and speeds up to 160kph.
Aimed at both intermodal and bulk mineral trains, train integrity is achieved by accelerometers on the front and back vehicles being constantly compared. All this gives maximum capacity, high speed and enhanced rail safety.
Noel Burton of Invensys informed on the rail resurgence in the suburbs of Auckland. A virtually new railway is being built with electrification, track doubling, re-signalling and two new lines. Fully duplicated interlockings on different sites connected by diverse fibre links give maximum capacity and reliability.
Some pragmatic solutions to capacity challenges are emerging in the UK. The new seven-car S-stock trains being built for London Underground sub surface lines are replacing the C-stock six-car sets. This creates a problem with platform lengths, especially at Edgware Road station where parallel arrivals and departures regularly take place. Until the new Bombardier CityFlo signalling system is commissioned, the old signalling has to be adapted. John Phillips from ARK Signalling Consultants described how moving starting signals outwards, providing extended overlaps and the provision of LED speed signs to drivers as trains enter platforms, enable a minimum cost solution to be achieved.
Lengthening trains has all sorts of implications for infrastructure. Platform extensions are sometimes physically impossible or not cost-effective if passenger numbers are small. As an alternative to selective door opening, Michael Toher from Halcrow told of the Thameslink line study to implement a ‘Double Stop’ process. For a 12- car train, this would involve stopping the first eight cars, then drawing forward to stop the rear eight cars. It was reckoned to only increase the London – Bedford time by seven minutes, but somehow the proposal was unconvincing.
Technology plays its part
Technological advances and opportunities featured a number of topics. The need for a Systems Engineering approach is vital, according to Michael Leining and Bernd Elsweiler from Germany’s DB Netz AG. DB has 80 different types of interlocking, with 20% requiring renewal in the next 10 years. The interfaces are complicated and likely to include bespoke project requirements, all risking a loss of control. The DB solution is to specify interlockings that will connect seamlessly with IP based data networks, thus leading to a standardised architecture for signalling systems.
Six interface specifications detail how the interlocking connects to i) other interlockings, ii) radio block centres, iii) level crossings, iv) axle counters, v) lineside signals and vi) point machines. A cost reduction of 30% per ‘signalled unit’ is predicted. DB claim to have an agreement signed with the major suppliers and hope that it will become a European initiative.
The overlay of Automatic Train Operation on to ERTMS was advocated by Benoît Bienfait from Alstom in Belgium. ETCS as part of ERTMS provides safety and interoperability but not performance and capacity, thus reduced operating costs are minimal. So could ATO be superimposed to yield additional benefits? A feasibility study by Network Rail in 2011 looked hopeful, as did a later simulated trial. Whilst ATO on mainlines is difficult because of the mixed fleet and infrastructure, an incremental approach aimed at urban mainline services appears possible. Firstly, apply ATO as a speed control on ETCS levels 1 and 2; secondly, with additional ETCS data added, ATO can provide accurate stopping at timetabled locations; thirdly, with timetable data provided from a Traffic Management System (TMS), train movement instructions including reversals and optimum speed commands, would be achievable.
Axle Counters remain controversial as to reliability and ease of installation. Martin Rosenberger from Frauscher Sensortechnik in Austria described recent advances in design and application. Significant progress has been made in dealing with the harsh environment that axle counters encounter – climate, temperature, vehicle geometrics, magnetic rail brakes and mechanical loads. Attachment to rails without the need for drilling is solved and an integrated axle counter is now feasible. This would have a serial interface to electronic interlockings using an open vital protocol giving diagnostic information, reset variant requirements, direction of travel detection, level crossing activation and point changeover protection.
Integrated Control Centre design is a hot topic as railways concentrate their operational control into ever larger centres. DeltaRail promoted their IECC Scalable aimed at reducing costs by the use of the IBM ‘message broker’ technology that links to many types of interfaces including legacy interlockings, route setting equipment and passenger information systems. A description of IECC Scalable was given in issue 92 of the rail engineer (June 2012).
Difficulties to overcome
GSM-R data handling limitations present the biggest challenge to the ERTMS programme. This is preventing ETCS being used in busy station areas. Alain Bertout from Alcatel- Lucent in France advocated that the railways collectively adopt LTE (4G) technology as a replacement for GSM-R. 4G exists in both the 700-800MHz and 2.6GHz bands and the specification is capable of handling all the features of GSM-R without a special development for railways. It is claimed not to be necessary to have a dedicated spectrum for rail use.
Whilst a 4G system is likely to be the solution, migration to this from the present GSM-R networks presents a logistics nightmare. Practical suggestions from the cellular radio industry as to how this would be achieved would be welcome.
The signalling of rural routes is a challenge for most countries. Laura Järvinen explained that Finland has many new Bombardier EBI Lock interlockings in place and is committed to introduce ERTMS between 2020 and 2030. These two elements together do not provide a solution for low density lines. ERTMS Level 1 will not deliver capacity improvements and Level 2 will be too expensive. A simpler system to control trackside equipment and set routes is required but this must interface to the existing interlocking equipment. Maybe the Level 3 system (ERTMS Regional) as being trialled in Sweden is a solution.
In contrast, George Raymond described the 30%‘dark territory’of the USA route mileage that has no signals, no track circuits and with train movements being authorised as‘track warrants’given by radio. However‘Positive Train Control’ is to become mandatory in America from 2015 and this will mean some safety enhancements. The use of GPS, end of train devices and turnout position indicators will be part of the package and will be less costly than ERTMS Regional. Perhaps a re-look at the refurbished RETB systems in Scotland could provide inspiration for others.
Human Factors
Signal sighting can nowadays use technology to avoid visits to site. Konstantinos Nikolaidis from London Underground explained how both 2D and 3D pictures can be generated to give an accurate view from a simulated cab. Sightline cones set up for a typical eye provide an unobstructed image of signals at the approach, the sighting point and the close view as a train nears the anticipated signal position. The system takes account of curves, cant and bogie movement and from this the signal positions can be optimised.
The political will to improve train services needs to be matched by the regulatory and licensing bodies being prepared to change legislation so that competence management aligns with the needs of emerging technology. Will Scott from Invensys gave some stark predictions. 30,000 engineers need to join the rail industry every year between now and 2016. The ‘job for life’ scenario that existed before has gone and the average employment of skilled people is now around 10 years. Engineers and HR staff need to understand each other’s position so that multi skilled employees are able to manage the spread of technology that increasingly exists. Competence standards must aim at reducing cost but not at the expense of safety or quality.
A typical example is signalling testing, where Doug Gillanders from Network Rail suggested that current testing techniques weigh far too heavily on past practice. Much scope exists for automated testing, where complete signalling modules can be tested off site, maybe as a complete system inside a single building, then shipped to site and plug coupled together with only the minimum of further testing being required.
Testing of ERTMS kit should be focussed on Control Centre to Train commands, where if satisfied for one train, should lead to a conformance certificate for all other similar trains. In signalling design, don’t incorporate features that will never be used just because the system can do it; they have to be tested and thus waste time. A mindset change is needed if precious testing resources are to be used efficiently.
Modern traffic management technology automates many railway operations under normal conditions. When out-of-course running occurs, control room staff often struggle to cope with the increased workload. Suzanne Heape from Invensys Rail has studied the human factors that can be encountered. A series of best practice recommendations has resulted including clarity of track diagrams, interpretation of alarms, flexibility of control room roles, understanding of individual responsibilities, design of the man machine interface and prioritising the information available.
Safety management
Track worker safety remains a concern. Jos Fries from Movares Nederland described a new process adopted in Holland. This includes dividing up the infrastructure into work zones, each of which is managed by an interlocking(s) under the control of a signaller. Work zones and associated possessions are arranged through a mobile terminal that leads to much shorter take up times and significant annual savings.
Peter Hughes from the Derwent Group in Australia shed new thinking on level crossing safety where, out of 8,000 crossings, only 2,000 have active warning devices with the remainder having just passive signs. To equip the 6,000 with barriers would cost A$10 billion, not thought to be good value for money. The thrust is thus to develop a low cost SIL2 warning device and to apply this gradually across the network.
Xavier Quayzin from Invensys examined the leadership and culture of safety in 10 accidents across seven industries including space shuttles, Piper Alpha, BP oil spills, Chernobyl, Herald of Free Enterprise and the Ladbroke Grove rail crash. Root causes were remarkably similar throughout: lack of top down leadership, cost and business pressures, safety culture, control and enforcement, contractor management and communication. Corporate governance responsibilities with statutory backed enforcement may focus minds somewhat better in the future.
In retrospect
A hundred years is a long time, and S&T engineering has seen many changes over that period. There will be even greater changes to the way trains are controlled in the next twenty years. Many new factors will have to be considered. Peter Symons, the Australian Vice President, looked at the energy and climate change debate. Transport is responsible for about 7% of the industrial carbon emission spectrum. Rail contributes only a small portion of that, with rail infrastructure even less.
Nonetheless, it is worth pursuing and regenerative braking, more efficient diesels, improved battery use to avoid peak loads, automatic traffic management for driver assistance are some elements that should complement the normal safety and performance agenda.
As a stark reminder of past predictions, Andy Stringer from Signalling Solutions Limited, recounted statements from IRSE Presidents of the past. In 1923, proper train detection with associated warning was called for – it was not really achieved until very recently. Also in 1923, centralised control was thought to be well underway but has yet to be universally achieved. In 1948, the adoption of multi aspect signalling on all main lines was thought to be imminent – it has only just been achieved in the UK with the re-signalling at Banbury. In emerging economies such as India, population growth, new technology and poor environment present their own challenges. Demand for new Metros in more and more cities will be the agent of change for city transport. Can CBTC technology cope with this demand?
Signalling in the future will be as much about train-borne equipment as it is about infrastructure. As such, the signal engineer and the rolling stock engineer will need to form a much closer relationship, including improved liaison between their respective Institutions.
Altogether, ASPECT 2012 provided a fascinating five days, with much to be learned by engineers of all disciplines.
