HomeRail NewsWhy is innovation so difficult in railways?

Why is innovation so difficult in railways?

Listen to this article

Railways, and signalling in particular, are generally not well regarded for being innovative. Talented job applicants with a high-tech background, such as software engineering, are predictably surprised when told that mechanical computers (i.e. interlockings) are still widespread in the railway system. Even if not a mainstream technology, relay-based interlockings are still regarded by some as “modern”, and in certain respects are even considered to be superior to ones based on electronics.

To an outside observer who is familiar with the fascinating potential of modern technology, the pace of innovation in railways might well be perceived to lag behind other industries just a bit too much. This observation applies not only to signalling but also to the speed of change in railway telecommunications when compared with commercial and consumer networks.

Mechanical computers

When railways were first introduced, they represented an industry where cutting edge innovation occurred. For instance, when the first mechanical interlocking was installed in 1843 at Bricklayers’ Arms Junction in south London, it was in fact a state-of- the art logic computer, occurring at a time when Charles Babbage was working on his mechanical computing machines.

The mechanical technology for conquering arithmetic problems is long gone, yet mechanical interlockings are still here.

So it is reasonable to ask the seemingly simple question – why is that? Why are mechanical interlockings still being renovated?

If, in fact, there is a business case for such an activity (compared with using lean IT technology with its associated potential for efficiency improvement in operating the overall system), then should we not ask if something has gone wrong with the innovation process in railways, and if so what and why?

Innovation or invention?

To avoid confusion about what is meant precisely by ‘innovation’ for the purpose of this article; we should distinguish between ‘innovation’, ‘invention’, and ‘technological development/improvement’. For the remainder of this article, the following definition and distinction is adopted:

“Innovation is the development of different or more effective products, processes, services, technologies, or ideas that are readily available to markets, governments, and society. Innovation differs from invention in that innovation refers to the use of a novel idea or method, whereas invention refers more directly to the creation of the idea or method itself. Innovation differs from improvement in that innovation refers to the notion of doing something different (Latin – innovare: to change) rather than doing the same thing better.”

No one would claim that there is a general absence of innovation (or inventions or technological improvement for that matter) in railways. The European Train Control System (ETCS), Positive Train Control and Speed Advisory Systems for instance, can clearly be considered as being innovations, based on various inventions and making use of general technological development.

To illustrate the differing ways in which innovation is perceived, some people consider that relay-based and electronic interlockings are just “doing the same thing better” than mechanical ones. Others point out that the range of safety functions implemented in modern software-based interlockings, for example relating to overlap and flank protection, is much more advanced; not to mention the potential for improved efficiency by remote control and automation that they offer. They would therefore claim that these advances are ‘innovative’ according to the definition given above.

A question of scale

There is one factor above all others that governs the speed of introduction of innovation on rail systems, namely the ‘scale’ on which the innovation has to be applied in order to be worthwhile. Thus, for instance, Disneyland (pictured right) had moving block in the 1970s; and2512325489_a80219d0f5_o [online] some metro systems have driverless trains. But these are localised applications.

These advances have occurred not because the engineers in those areas are any better or more innovative than signal engineers working on large railway networks. On the contrary, one could argue that maintaining a large quantity of heterogeneous technology across a large and distributed infrastructure network with such a high level of safety and reliability is an art mastered by no other engineering discipline to the same extent. The longevity of mechanical interlockings could be claimed as proof of the signal engineer’s far-sighted design, rather than being a criticism.

It is however apparent that the scale (size) of a railway network, and the large number of people/bodies that need to be aligned in order to introduce any change, seem to pose more challenges to the innovation process than in other contexts where localised innovation is possible. A further difficulty with innovation may be that railways are a mature industry, so that innovations do not easily offer returns on the investment made.

In addition, there appears to be a number of more subtle reasons for the failure of innovative ideas in our engineering domain, including:

  1. The new idea does not fit with the existing (often aged) infrastructure;
  2. The new idea does not fit with the culture of the corresponding railway/country;
  3. The new idea does not fit with existing regulations and operational procedures;
  4. The idea does not meet a real need, in the opinion of railway experts;
  5. The originators of the idea are not trustworthy and/or do not have the right background, in the opinion of railway experts;
  6. The originators of the idea (or the organisation they work for) are not considered to be likely to be around for long enough to support the innovation through its whole life cycle, right through to obsolescence (50 years or more);
  7. In the opinion of railway managers, there might be no business case for the idea;
  8. There might be a business case on the global level, but local application within a fragmented industry prevents the potential benefits from being realised;
  9. The market potential is seen too small for investment by railway suppliers, because the application circumstances differ too much from country to country;
  10. The idea is innovative at a component level, but there are no standardised non-proprietary interfaces to enable replacement of the old version with the new one, without renovating the systems of several other suppliers at the same time;
  11. Safety approvals appear too difficult to obtain, or there are other liability issues that cannot be overcome;
  12. No sensible roadmap can be constructed upgrading the entire network.

Having established this list of plausible reasons for the failure of new ideas to reach the implementation phase, the fact that innovation appears to lag behind in the railway industry seems less surprising.

However at the strategic level, it should be clear to all stakeholders that any system that consistently lags behind in its application of technology will lose its competitiveness sooner or later and hence either be removed from the surface of the Earth or be banished to the museums at best!

Given the current cost base of the rail industry, one main goal of innovation must be to lower the whole life cycle cost of systems and thereby make change more attractive.

No better elsewhere

As stated earlier, it isn’t suggested that engineers in other comparable industry sectors are better than those in our own. On the contrary, other systems that comprise a large collection of existing infrastructure, such as air traffic control, seem to have similar struggles. For instance, the introduction of new generations of transponders into aircraft fleets takes some 40 years. In comparison, the 20 or so years that it took for ETCS to move from concept to its first reasonably efficient introduction in a project (the Lötschberg Base Tunnel in Switzerland) seems surprisingly fast.

Cleary, no one can imagine a quick technological, “i-phone-like” revolution in railways. On the other hand, it is essential that evolutionary innovation should and must be possible. True innovation needs a clear vision as to how we want to operate our railways and rail transit systems in the future, and needs pioneers/champions committed to take on the challenge of delivering that vision fast enough so that the investments pay off.

Lötschberg_Tunnel [online]Looking again at the ‘12 reasons’ stated above, it should be obvious that we need to distinguish between ‘valid reasons’ that hinder innovation – intrinsic and unavoidable in the system “railway” – and ‘other reasons’ which would cease to obstruct innovation if the right structural changes were made at the strategic level. For example, considerable progress with the standardisation of interfaces in road traffic control systems (see reason 10 above) – another strong competitor of the railway – seems to have been made already. If, as a consequence of such advances, this reason was no longer to apply in the rail sector, it might also remove other obstacles (such as reasons 1, 6, 8, 11).

Based on this example it seems worthwhile establishing a more comprehensive list of reasons for the relative scarcity of innovation in the rail industry and performing a cause-consequence analysis in order to understand the underlying mechanisms better. However, that would be part of the next step – answering the question “How do we make railways more innovative?” – which lies beyond the scope of this article.

Some people might argue that this in nothing new and that there are other underlying obstacles to innovation. For instance, during the development of ETCS, the standardisation of interfaces on the vehicle had been proposed but was declined by the industry, suggesting that the difficulties with innovation may also be attributed in part to conflicts of interests. This may be true, but nevertheless a fundamental review of the mechanisms of innovation in our industry still seems to be a crucial step for long-term success.

Clearly, no single stakeholder in the rail industry would be able to remove a sufficient number of hindrances to innovation. Therefore it would seem necessary for governmental agencies, railway companies, suppliers and research bodies to collaborate and to establish roadmaps for removing obstacles for innovation in the railways, while taking into account the particular interests of each group.

Edited on behalf of the International Technical Committee of the Institution of Railway Signal Engineers (IRSE) by Dr Markus Montigel, CEO of Systransis AG, Switzerland, and published with the permission of the Editor of IRSE NEWS and the Institution of Railway Signal Engineers.

Previous article
Next article


  1. No point in being polite, this is so ignorant it is beyond wrong.

    Railway signalling, and British engineers in particular, have been at the cutting edge of innovation.
    In the 1980s British Rail Research cracked the problem of safety critical interlocking software. The first Solid State Interlocking (SSI) was commisioned at Leamington Spa in 1985. SII is now the standard interlocking in the UK and has sold around the world.

    For lightly used rurual lines BR Research developed Radio Electronic Token Block (RETB) combining SSI with radio to eliminate manned signalboxes and mechanical token exchange for single line working.

    SSI was followed by the Integrated Electronic Control Centre (IECC) where signallers set routes using a mouse or trackerball on VDU screens.

    Not that they have a lot to do because another BR Research innovation was Autometic Route Setting (ARS) which runs the railway to timetable.

    ETCS is taking a long time because it is a political initiative dumped on the railways by the European Union to force cross-boder interoperability on the national state railways. It is not an engineering innovation and is proving very difficult to implement.

    Railway signalling today is all about electronics and software. Your article is on a par with criticising the Royal Aircorce for being backward because the Battle of Britain flight still flies Hurricans and Spitfires.

    Roger Ford .

    Modern Railways

    • Roger has a point as this article implies that there has been minimal innovation with British Railway signalling, clearly this is not the case.

      However this does not detract from the article’s fundamental point that the pace of innovation is too slow and that the underlying causes for this need to be understood and addressed.

      Having spoken to many at the Railway Industry Association’s Innovation conference this week I believe this view is widely shared within the industry. Certainly Britain can learn from other railways. Take the example the Russian KLUB in cab signalling system developed in 1994 with 30,000 cabs so fitted.

      Although there is much to be learnt from Roger’s articles, they do not often address process or cultural issues which are a key issue for innovation and other improvements. For example Coucher’s bullying culture which resulted in under reporting of accidents and much worse was not mentioned in Modern Railways.

      There are currently many good initiatives to support rail innovation which need to be publicised and supported by the railway press. One of which is the 2012 Rail Technical Strategy which Roger described as “mostly harmless”. Perhaps this deserves more support.

      David Shirres
      Engineering Writer
      Rail Media

      • No one would deny that the British rail industry has been and can be innovative. In the Rail Technical Strategy 2012 the industry itself shows that there is no lack of appetite for innovation, it’s more that there has been little to help industry join the dots between the challenges, the funders and the innovators themselves.

        As David Shirres says, the pace of innovation is too slow, compounded, or perhaps caused, by process and cultural issues. It was the recognition of these issues that led to the creation of the new Enabling Innovation Team and the Rail Innovation Fund to pump prime innovation in much the same way as BR did in 1985 by providing development funding and a demonstration site for SSI. We are currently launching new competitions
        and programmes to help join those dots and draw that innovation in – for more information go to the EIT pages on http://www.futurerailway.org

  2. There is another side to this. Vital, in some cases life-critical, infrastructure (the railways are only a little part of it) is dependent on advanced technological networks. When the next big solar flare arrives like the event of 1859, or there is some kind of malicious computer virus attack, there is going to be a re-think.

    An example of this is Selective Door Operation, which uses and therefore relies on GPS. What happens when it goes down? An alternative philosophy for SDO would have been to fit each door with a detector, in series with the door locking system. That would not have had to rely on GPS.

  3. Actually the rail signal industry has been a lot more innovative than it gets credit for. Radar was instituted in the rail industry for automating classification yards as soon as it was released by the military after World War II for civilian use. Electronic track circuits for grade crossing warning systems were begun with vacuum tubes (use for regular signalling was not applied at first because they could not support the longer track circuits needed for regular signalling), but they had problems with the environment found along railroad tracks. Railroads continued to use them until solid state circuitry was available, but the solid state equipment was very subject to lightning damage and had problems with cold weather until someone found that a low wattage incandescent lamp could heat them enough to make the work. Conversion from transistors to integrated circuits also was as fast as possible with the increased susceptibility of that circuitry to electronic interference and spikes on the power supplies. Integrated circuits are now used in “conventional” track circuits now as well.

    Sure mechanical interlockings continued into the 21st Century because it takes a lot of money to upgrade equipment that is doing its job with newer technology that is cheaper at first, but doesn’t provide maintenance savings once it is intalled. Everyone uses microprocessor logic for new interlockings, replacement of worn out mechanical and relay interlockings, and replacement of existing interlockings when new features are required.

    Railroads automated classification yards first with analog computers and then with digital computers when they superseded analog computers as the state of the art. Radio was used by railroads to commuicate between the front and rear of passenger and freight trains in the 1930s. Railroads developed standard colors for railroad signals in the early 20th Century so all railroad signals are the same exact color, unlike the differences one finds with traffic signal colors.

    It sounds mundane, but railroads are still experimenting with the latest steels that provide improved properties over older varieties.

    Railroads have always been in the forefront of making use of computers for both operational chores and back-office chores.

  4. I have to agree with the tenor of the article. There has mean much innovation, and most of it “evolutionary”. But “revolutionary” innovation is lacking. So we’ve added steel sleepers to wood and concrete, but apart from the trough-supported LR55 track (hardly widely trialled or used) we are using the same track as Stevenson. There need to be further innovations in track design. That may result in something ditinctly un-railwaylike, but that’s not the point – it may and up looking like “pods”, but these can become an essntial partner to long-distance rail. I look forward to cable-less, signal-less “movement control systems” (I’m not sure we can call it “signalling” any more) Embarrassingly, it seems we are to see automatically-driven cars before we see automatically-driven trains. There needs to a revolution in train weight, too. I find the treatment of Parry’s lightweight people-mover pretty awful – sceptical delay has just caused the benefits of light trains on branch lines and on potential new light rail lines to be delayed by a decade. Our attitude to tram-trains are no better, though I think further innovation will reduce the total system costs to more reasonable levels.

    • I think you are right Stephen but there is a lot happening to challenge conventional ideas. Look out for the “Inspiring Innovation” article in the next The Rail Engineer


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.