That (Railway) Sinking Feeling

Now that I’m ‘independent’ or ‘freelance’ again (which of these euphemisms do you prefer?) I’m free to spend a bit more time at the sorts of events I’ve often helped organise but can rarely attend*. This week, for example, I unexpectedly was able to enjoy the whole of the two and a half day Institution of Mechanical Engineers Stephenson Conference on research in the railway at Birdcage Walk in London. There were sixty or so papers, many international, including a wide-ranging, ambitious key-note introduction from Graham Hopkins of Network Rail. The conference was preceded the evening before by an excellent free evening lecture from Andy Doherty also of Network Rail on the EU Shift2Rail research programme. This latter alone postulated (it seemed like) hundreds of technically viable great ideas which are not currently deployed on the railway. Attendees at either or both events will now have an enviable overview of the breadth of research and innovation in a sector not noted – at least in the public’s or politicians’ minds - for either. So I’ve been thinking about the adoption of new ideas in rail: is it really lacklustre? If so, why? Certainly the two events proved that there is no shortage of ideas to improve the railway, and thus the country. How can we increase the take-up of viable new ideas in rail?

A starting point is to question the premise of innovation for its own sake - the idea that it is automatically good, and that railways are automatically bad at it. London Underground has traditionally been accused of lacking in innovation, for example, maybe because it is so much in the eye of the London-based media, or maybe simply because it is so old. I remember as a manager at LU being constantly urged by various consultants, review bodies, advisors to the mayor, journalists, suppliers and so on that we were rubbish at innovation, backwards-looking and conservative. Dinosaurs. An illustrative example was the (then novel) LED light. Why did we not fit them to trains and stations? Pointing out that a conventional fluorescent tube on a Central Line train cost 80p, compared to (at the time, and given the small volumes) over a hundred pounds for a bespoke LED fitting, didn’t seem to convince the innovators. It would have taken several centuries for the capital cost to be repaid in energy savings. ‘Aah yes, but what about the labour costs of changing them when they fail? LED lights last forever’, the disciples of innovation would continue. Conventional tubes last around two years and are changed by cleaning staff as part of their routine duties so it wasn’t necessary even to do a fag-packet calculation to see that savings would be minimal and the ‘innovation’ was not, at that time, economically viable. It seemed that ‘innovation’, like motherhood, was more important than a sound business case. Naturally, over time, as LED costs have fallen, the technology has been widely adopted across the industry – but only when it became viable on a case-by-case basis.

To get onto the railway new ideas not only have to pass economic tests but also, rightly, rigorous technical ones as well. The railway environment may be one of the harshest and most variable (least controllable) engineering domains we have. Certainly it is one of the most complex, where the law of unintended consequences applies without mercy. Most engineered things don’t move around – their environment is known and can be controlled. Even things that are as peripatetic as railway vehicles have it easier in some respects: spacecraft don’t have to worry about the weather or the travelling public; cars won’t see 50g accelerations to suspension components; a battle-tank or jet fighter has to be reliable but after a few days or weeks of front-line service its moving parts can expect to be thoroughly maintained or rebuilt. In few branches of engineering is it possible for the failure of a single component to delay thousands, or hurt dozens, of people. Civil aviation is an exception – clearly it does have a comparable safety challenge to rail - and here I note that planes and airports are broadly unchanged during my lifetime. The first 747 was contemporaneous with the last steam traction in the UK (1968) and the model is still just about in production unlike 9F steam locos. Of course air passenger environments, ticketing, traffic control technologies, security, safety, engine efficiencies and so on have all sharply improved – but so it is with rail. If the underlying engineering demonstrably works, and has been perfected over decades, what's the case to change it? Is it reasonable to say that civil aviation is innovative and not rail? Rail’s achievements in moving vast numbers of people and tonnes of freight through and between crowded old cities and towns – and at least in the UK’s case, with improving market share, decreasing unit cost, and impeccable safety – demands innovation and endeavour every day.

So it is, per se, hard to get new things adopted in rail, not because railway managers are stupid or wooden but because the economics and engineering are hard (and many 'new' things have been tried already somewhere or at some time and been found wanting compared to the tried and tested technology). But . . . we could clearly still be doing much more. Many of the ideas in the conference were obviously viable and their adoption would make the railway (and thus the country) better. Governments clearly believe so – many, perhaps most, of the schemes and ideas were being funded and driven by state or quasi-state bodies: Shift2Rail; Network Rail; TfL; RSSB; RDG; TRL; Universities; DB. It was striking how often we encountered during the conference a sinking feeling that I have experienced throughout my railway career: great idea, a no-brainer, but it’ll never happen.

Let’s call this ‘that sinking feeling’ (TSF). To be clear I’m narrowing this discussion down to technically and economically viable propositions (by contrast, for my next LinkedIn article I think I’ll write about the aptly named Hyperloop). At the start of my career causes of TSF were obvious (or so everyone said): the dead hand of the state; monolithic British Rail; a dying industry; no money. Now each of these factors has reversed (supportive government; vibrant, diverse, growing industry; more money than ever before) and yet TSF is still very much a commonplace. Maybe more than ever. Why is it so difficult to do the ‘right thing’ in rail, to improve things? Why is managing within it, or being a front-line railway engineer, so often frustrating and dispiriting?

A simple economic obstacle seemed to thwart many of the ideas, as effectively as a flight of steps in front of a Mk1 Dalek intent on universal domination: the benefits usually went to anyone but the party stumping up the costs and taking the risk. Some simple examples from the conference illustrate the point. We heard a range of great, simple ideas (far from pushing at technological boundaries and solving real problems) including flywheels which could save DMU fuel by acting as an energy store and providing a form of regenerative braking; a basic modification to the problematic doors on heavily-used DLR trains; wider user of OTMR (black box) data from trains for fault-finding and trend analysis; and more use of sanding (invented in the 1860s) to improve adhesion. Not even innovations really; just good applied engineering. The ideas were on the way to adoption in some places but certainly not universally. Even to get that far had taken the proponents a great deal of passion, years of effort and usually special funding. Why, oh why, does not the optimisation of existing technology and the application of ongoing technological improvement not take place organically as it does with, say, cars or computers or phones? Why, if you are a front-line railway engineer or manager, is TSF so prevalent? It is because, in broad terms, we have created an environment where those who invest in innovation (or even basic continuous improvement) are unlikely to benefit.

This of course is a characteristic of railways, a transport mode so capital-intensive and physically invasive, and so frustratingly prone to provide economic benefits mainly to non-users, that governments have always remained in ultimate charge. As the conference showed, this includes sponsoring and directing innovation within it. (That’s not to say of course that state bodies can’t innovate – often their engineers and managers are the most resourceful and innovative – but the relentless innovation we see in, say, mobile phones is driven by private enterprise and it is this sort that governments seem to want.) Even setting macroeconomics aside we do seem to have a simple microeconomic problem which leads directly to TSF: it is difficult for individual industry parties, who may be franchisees with short lives, suppliers, asset lessors, or infrastructure owners, to capture benefits from organic improvement. This is because benefits, despite being very real, pop up in different parts of the system, or take a long time to arise, or maybe even accrue outside the railway at the societal level (to 'UK plc'). They rarely take the form of readily identifiable ten pound notes (savings, say, or increased revenue) and if they do the money appears in somebody else’s pockets. The conundrum was neatly summed up by one speaker, from a new franchisee, who paraphrased his finance director using a rented flat as example: ‘we’re not here to improve the kitchen’. The problem though goes beyond franchising and certainly existed under British Rail or old London Underground. Their kitchens were tatty too, even though occasionally the landlord would reluctantly fork out for an upgrade (sometimes only after a chip-pan fire).

The problem I’ve described here is well recognised and there have been serious attempts to address it from time to time. Andy Doherty mentioned, for example, variable track access charges for lighter vehicles or those doing less damage to track, which has been successful in encouraging lighter more track-friendly trains. There has been a great deal of work to build ‘innovation’ into franchise agreements from the start. Strict contractual regimes penalise those who don’t solve problems supposedly giving yet another incentive to innovate. There is something in me, though, that rebels against forced ‘innovation’ as an axiomatic good. It is the ends that are important, not the means. It would be much better if we could find ways to avoid TSF by encouraging railway people to get on with doing what they do naturally all by themselves: fixing problems and improving things by applying new ideas.

How? Certainly not by another radical industry restructuring. TSF is probably independent of industry structure and nobody wants any more restructuring. Also not, from my experience of the London Underground PPP and mainline track access agreements (schedules 4/7/8 etc) through ever more complex contracts. And not, at least in my view, by building ‘compulsory innovation’ into new franchises. Instead, and simply for the fun of the debate, I propose a new 'railway clearing house for railway innovation and improvement': anyone able to demonstrate that they have implemented a scheme which is generating benefits that they themselves cannot capture because it falls to someone else and they cannot claw it back contractually, or it arises in part beyond a franchise or control period, or perhaps because it generates non-railway economic benefits, would receive a guaranteed share of the benefits from this body, which in turn would collect some proportion of the benefits from the actual beneficiary. Schemes could be agreed, in outline, in advance. This would increase organic investment in the railway simply by making it viable. When and wherever good new ideas arose they would be fundable. Tenants would have an incentive continually to improve the kitchen and not let it run down.

The idea sounds far-fetched but with some thought it could easily bolt on to the existing industry structure - the new clearing house function could readily be taken up by an existing body such as DfT, RDG, ORR or RSSB. These after all already employ engineers and economists. It ought in theory to be self-financing (admittedly, over time and only at the aggregate level of ‘UK plc’ and then only if the scheme were confined to truly viable ideas). It could be designed so that it was not bureaucractic like the original Railway Clearing House (pictured). In fact, if the principle became established government resources could be gradually withdrawn from the state-directed research workstreams - and indeed many of the other complex mechanisms introduced over the years to compensate for various market failures - because it would be much more likely that others were directing their own resources to demonstrably fruitful areas of railway improvement.

Just an idea to avoid that familiar sinking feeling.

 * I have to declare a direct interest, having been part of the IMechE’s seminars and conferences committee for quite a while, although I played no part in organising this excellent event. The credit belongs to an organising committee consisting of Francis How (chair), Simon Iwnicki, Richard Gostling, Sharon Odetunde, David Polhill and Kiesha Lewis, of course ably supported as always by many IMechE staff.