Every two years, the Institution of Railway Signal Engineers (IRSE) holds an international conference to look at signalling practice worldwide, its innovations, technical developments and potential challenges. Known as ASPECT, the ninth conference in the series was recently hosted in Singapore, the first time outside of the UK. It was very well attended and included a lot of delegates who would not have been able to participate in London.

Whilst the event covered many topics, the main themes were Metro Technologies, Professional Development, High Speed Rail and Condition Monitoring, all topics that impact on every country.

Welcomed by the IRSE President, Peter Symons from Australia, his statement ‘Resistance is Useless’ was aimed at those in the industry who believe that the principles for operation and safety are enshrined forever. The perceived ultimate safe state that signalling stops trains is no longer good enough, as it fails to meet the fundamental mission to transport people and goods.

Peter stressed that fail safe must be linked to degradation of service, not stoppage. ALARP (as low as reasonably practical) is not absolute, as residual risk remains, and this risk needs periodic reassessment. New threats are emerging; cyber security needs a much higher profile to assume the same status as safety.

With the world’s ever-growing population, increasing rail capacity is a challenge in all countries. Moving block, control centre consolidation and better braking systems are all part of the solution, but it must also include the use of ‘Big Data’ to sense and monitor every component part and, with clever analysis tools, detect likely failures before they occur.

The cost of signalling is too high; interoperability and interchangeability must become the industry norm, as it is in telecoms. Type approval and safety assurance processes are too onerous and, in a system’s life cycle, the operational cost is now more than the capital outlay, making the business case hard to justify.

Some trends are emerging; design tools are evolving into formal computerised methods, there is increasing use of commercial-off-the-shelf (COTS) products, reductions in trackside equipment look likely, and software development must be portable between products with minimal safety assessment.

Technology convergence between metro, main line and heavy haul would bring huge benefits, as many of the business requirements are common. Above all, communications and radio technology must be recognised as the backbone of signalling, with backwards compatibility being a necessary prescription.

As a scene setter, this was a powerful message and it challenges the status quo. The closing statement was ‘Pay Attention and Keep Up’.

Metro signalling and operations occupied much of the conference agenda, sufficient to warrant a separate article on all the developments and challenges that industry and network providers face. This article considers other elements that constitute a modern railway.

High-speed lines

The building of high-speed lines began in Japan in the 1960s, followed by France in the 1970s and many others, in Europe and elsewhere, over the following decades. The slow progress of the UK to follow suit is disappointing but politics, finance and social issues dominate such projects these days and similar situations are coming to the forefront in other countries.

However, the technology for high-speed lines continues to excite engineers worldwide, and nowhere more than China, where the construction of lines has progressed at an amazing rate. Gao Ling, the director of technical support within CRSC International, described the work to apply ATO to high-speed services.

Firstly, however, she gave details of new and projected lines that have or will be built. In 2016, China had 16,000km of high-speed routes, this is planned to rise to 30,000km by 2020 and 38,000 km by 2025. These will link all the main provincial cities with a population of over half a million people. Signalling employs the CTCS system, which is equivalent to ETCS and comes in similar Levels. CTCS 0/1 are the equivalent of ETCS Level 1, with track circuit based commands, CTCS 2/3 equate to ETCS Level 2, using loop or radio communication, while CTCS 4 is similar to ETCS Level 3, with no track based detection system

In parallel with all this, the Chinese are building inter-city rail networks to serve the urban sprawls out from the cities. Similar to RER operation, these will have a maximum speed of 200kph, a headway of three minutes and CTCS Level 2 with ATO and balise provision for accurate platform stopping. The technology will be compatible with national rail standards.

Eight control modes are envisaged, ranging from full supervision including a default provision if the ATO is disabled, ATO with track based detection, manual driving with ATP, and drive on sight in a degraded mode. 10 cities already have an inter-city network but not all yet have ATO, which will be added incrementally

Shinkansen equipment update

Automatic Train Control (ATC), equivalent to ATP in European-speak, has been a feature of the Japanese high-speed lines since the 1960s. The system uses positioning ground coils to correct any ATC errors, a design which is some 30 years old, uses 64kbps data and is suitable for speeds up to 350km/h. The coil can be adversely affected by snow and, with a distance of 350mm to the train’s antenna, damage from underhanging train parts is not unknown.

The latest Shinkansen train will have a top speed of 360km/h, necessitating a new design of coil, described by Naoki Hashimoto from JR East. This is mounted on concrete slabs, with an increase in distance from train to coil and coated with a hardened material. The downside is that it needs an internal power source but this is a battery with a 10-year life. Initial tests are encouraging and production versions will follow once the full test programme is complete.

Other Far East projects

It is not only the Chinese who are operating high-speed lines in the Far East. South Korea has had such lines since 2004, now extending to 3,861km of route. Updating the train control technology to a radio-based ETCS Level 2 system is underway. This will use a 4G LTE-R system in the 700MHz band with 10MHz up and down links. Information was not given on how LTE-R differs from standard LTE but this will be of interest to all who are involved in the debate as to what follows GSM-R.

In Indonesia, a similar situation exists with a high-speed line being constructed from Jakarta to Surabaya and with other lines planned that will total over 6,000km. ETCS technology will be the framework for the control system but with Indonesian variations that also envisage LTE-R as the radio bearer.

Singapore is a very small country but plans are well advanced for a high-speed line to Kuala Lumpur in Malaysia, envisaging a 90-minute journey for the 350km distance with eight stops en route. 2026 is the likely opening date and confidence exists that this will be met.


Condition monitoring has been around for many years, but are we using it correctly? That was the question posed by Trevor Bradbeer from Balfour Beatty.

Recognising that both over and under-maintenance are undesirable, together with increasing difficulties to get track access as capacity demands grow, a new way of applying remote condition monitoring (RCM) is needed. The objective should be to detect the potential for functional failure and then find the optimum time to act.

Measuring different functions might make things easier. Reducing the fitting of sensors to track side equipment, and instead continually monitoring all asset data back at the control centre, should lead to an indication of the overall network performance. The system intelligence should then give an indication of where and what is likely to cause a future problem.

With ever-more signalling equipment being train borne, establishing the root cause of reliability problems becomes more complicated. ETCS and CBTC often have the infrastructure and train borne equipment provided by different suppliers, both complex in technology terms and with a hard contract boundary. A failure is often attributed to the equipment where the fault occurs, whereas the cause of the problem is elsewhere.

Paul Barnsley from ALIRE Consultants in New Zealand believes that ‘reflected faults’ across the track-to-train boundary need to be identified by much better interpretation of the collected data from all sources, feeding this to intelligence systems that can interpret the situation and make decisions accordingly. A ‘Fault Management Requirements Specification’ is required to pull all of this together.

A pragmatic view from Michael Hamlyn and Bruce McDougall, both with long experience in the signalling profession, considered the reliability implications for ‘brownfield’ resignalling projects, where the introduction of train-borne equipment is being introduced or updated. The transition from old to new often sees a conflict between engineering and operational needs, all in a social and political environment. It is important to understand, quantify and assess the operational risk.

There is no ‘one size fits all’ solution. Two basic options exist – a progressive changeover with some mixed-mode operation, or a direct changeover with both track and train equipment being upgraded together. Some basic rules apply: create a strong client integration team, fully understand the old system, carry out as much off-site testing as possible and then train the operators and technicians before changeover. In all of this, the devil is in the detail.

An example of the detail was given by Hashim Abdullah, from SBS in Singapore, one of the metro line operators. Point machines have no redundancy but have to be fundamentally reliable. Most failures (85 per cent) are attributed to relay failures, often intermittent because of humidity and oxidisation conditions. If these can be replaced with solid-state equivalents capable of carrying high currents, it would be beneficial. Trials with bipolar and commercial field effect transistors did not yield conclusive results but a new metal oxide field effect transistor, encased in the same relay-style plug-in unit to make it interchangeable, looks more promising. It is cheap to produce and is undergoing a six-month trial.

Technical innovation

A conference of this type was bound to attract a number of innovative ideas, some for immediate consideration, some very much ‘blue sky’. Five examples were:

  • Monorail. Johannesburg is considering a system to feed into the Gautrain line, similar to those in Kuala Lumpar and São Paulo. The advantages are seen as being safe (trains are unlikely to derail), low capital cost, the elevated track is ideal for built-up areas, and it is relatively easy to rescue passengers in the case of failure. Downsides are incompatibility with other rail systems, scarce resource knowledge and potential competition from IT-based road transport such as Uber.
  • Use of BIM as an environment for system management. This would lead to eliminating the present difficulties with records, with them becoming digital-data based and not needing multiple drawings. A set of international BIM standards might be required but it is unclear as to who would produce these.
  • Hyperloop. Seen as the ultimate frictionless solution, trains would travel at high speed in a vacuum tube. A test site is established in Las Vegas but challenges would be to establish the proof of concept, acquire funding, obtain regulatory approval and produce the standards. Singapore to Kuala Lumpur would be a 27-minute journey, while Abu Dhabi to Dubai is another potential route. Whilst technically possible, it is unlikely to find favour in the near future.
  • New point operating mechanism. Known as Repoint (Redundantly Engineered Points), this looks at a new way of moving and proving switch rails (issue 131, September 2015 – and see accompanying feature).
  • Fibre-optic sensing. Uses changes in light wavelength when vibration or movement occurs in adjacent infrastructure – see accompanying article.

Engineering resources

It is not just the UK that is hindered by a lack of signal engineers – most other countries face the same problem. The loss of the traditional entry and training regimes, when the railways were vertically integrated and many engineers stayed with the organisation for life, has had an adverse impact on recruitment and retention. The emergence of small companies, whilst good in some respects, has meant most cannot afford to take on trainees. Even those that do have apprenticeship schemes cannot offer the same breadth of training that existed in the past. The result is a significant traditional skills gap. People with other skills (often in IT) compensate to a degree, but with the risk that the basics of signal engineering are not understood and safety and operational errors result.

So how to solve the problem? Daniel Woodland from Ricardo Rail put forward some suggestions. Listing the general trends of business and engineering is a good starting point: digitisation, security, cost efficiency, sustainability, artificial intelligence, gender shift and resource scarcity. Applying these to the railway signalling sector (and indeed other rail disciplines) for the next 20 years gives:

Automation and implementation techniques;

Globalisation with need for standardisation and modularity;

Cyber security and threat to digital systems and Big Data;

Cross fertilisation of technology from other industries, for example autonomous road vehicles.

These are the elements with which young people will identify, but they must be coupled with a better teaching of the basics. Business and individual needs have to converge and employers are slowly starting to realise that the engineers of the future will transfer between industries much more readily than in the past.

Training has to take on a different format than before, with the acceleration and encouragement of apprenticeship programmes, university degrees linked to rail, graduate training placements, support for external training courses, continuance of local in-house training, sideways and upwards moves by planned rotation, self-driven learning within companies, greater emphasis on tested competence, insistence on CPD and professional registration, and participation in conferences and on committees. All this, plus a recognition that there is no single solution and that everyone has a different development need.

The IRSE annual skills survey shows a decline in both numbers and skills from 2012 to 2016. Sam Loveless from Siemens, in recognising that significant technological changes occur every three to four years, also suggested BIM techniques to bridge the gap between signalling and other disciplines. The use of virtual reality to take people “out on to the track” would create a better understanding of the real railway without the safety risks.

Another idea came from Alexander Patton of Siemens UK on the need to re-invigorate STEM – Science, Technology, Engineering and Mathematics – by creating an ATO model railway that school children could build using typical school electronics kits. Packages such as Raspberry Pi, Scratch and Python are readily available in schools and these can be used to simulate interlockings, axle counters and signal panels, all integrated, with the addition of simple LEDs for signals, to form an introduction into object-oriented programming. It’s a great idea, but it will need established signal engineers to participate in the STEM programme.

Even if sufficient engineers were available, there remains the problem of quantifying the skills needed. Alex McGrath and Richard Stephens (picture below) from Australia told of three recurrent problems:

People with lengthy signalling experience often cannot perform a design task;

People with above average competency in a specialist area lack the flexibility to take on new or more complex tasks outside of their immediate experience;

People who are both competent and experienced in engineering cannot produce quality work in pressurised or commercially adverse situations.

This thought-provoking argument was typical of this year’s Aspect conference, and it was a challenge to absorb lots of information in three days. Nonetheless, many topics emerged and it can only be hoped that the delegates will take back the ideas, decide if they are applicable to them, and develop them in to beneficial solutions.

Well done the IRSE for staging this truly international event.

This article was written by Clive Kessell. 

Read more: Birmingham Resignalling