Return to 21st Century Ford Archive.
‘Not Alstom again', readers may sigh, since the Franco British multi-national also provided the subject when this series visited the traction and rolling stock market. Well, it's not favouritism, but the continuing turmoil within the railway manufacturing industry which resulted in Charles Burch Managing Director of Alstom Signalling sitting opposite me as the tape recorder turned at his Borehamwood office in August.
With Alstom Signalling responsible for the controversial West Coast Main Line Train Control System (TCS), Charles was an obvious subject. But, apart from the earlier Alstom interview, the managing directors of the other signalling majors active in the UK have a lot to say. Or rather they would have, had they not become ‘cup tied' by takeovers.
Just as I was setting up the ‘signalling' interview, the news came that Adtranz Signal had been put up for sale. So that was that. Then, Westinghouse Signals was short-listed as a potential buyer of Adtranz. And even then it wasn't finished. Vaughan Harmon's United States parent was acquired by GE to be merged with its Harris subsidiary.
Stymied at every turn, I returned to Charles Burch who was very gracious about it and fitted me in at the last moment before he went on holiday. And naturally, the first topic was the internationalisation of the UK market'
‘The names on all the bids are international, including our own', Burch agreed. ‘Alstom is an international company and the technology that we're deploying is increasingly rooted in Europe because that's where the long term development has take place over the last 20 years'.
It was only afterwards that I realised that 20 years ago British Rail Research was developing its pioneering Solid State Interlocking, leading up to the pilot installation at Leamington Spa in 1985. A lot of electrons have flown through the gate since then.
Even so, the Borehamwood Plant is an active player in Alstom Signals global development plans. For example, it is responsible for a new ‘two out of three' computer based trackside processing centre which will see its first application on the Singapore Mass Transit.
A current example of internationalism is a mini-CTC system in the Republic of Ireland which Borehamwood took over when Alstom acquired the Italian signalling company SASIB. The project management and engineering design is now being handled from the UK while hardware is coming from Italy and the United States .
As with the other railway multinationals Alstom is moving away from vertically integrated companies largely serving national markets, to centres of excellence which provide ‘universal' products and services. This is fine in principle, but railway signalling, in particular, presents difficulties.
And Britain is a classic example. ‘To get a product that works perfectly in Italy , German France, Brazil or North America approved on the UK railway is not a trivial task' Burch comments with studied understatement.
This means that in practice signalling is ‘quite a way away' from being a global industry. Having joined Alstom Signalling at the end of 1997 from the industrial controls industry, Burch is particularly conscious of the gap between corporate organisation chart and reality.
In controls, he was competing in the UK market with products that were freely imported and used from around the world. Signalling, which is just another control system, is nowhere near that situation. But, says Burch, ‘you can see the trend emerging'.
Cross acceptance is making things easier, but even then, ‘the devil is in the detail'. New signalling systems or equipment can't be accepted in isolation, they have to be compatible with the railway environment and national railways have different environments – signalling principles for example. So, while Charles Burch is all in favour of cross acceptance ‘we mustn't be starry eyed about it'.
Interestingly, given that Railtrack has written it off as obsolescent, if not obsolete, Borehamwood has a nice line in exporting SSI. Alstom's version of this classic example of ‘Brit kit' has been in service on Belgian Railways and in Hong Kong for several years and, after successful pilot schemes, has now been adopted by French Railways.
This involved adapting the software to French signalling principles and giving it a proper French acronym. So when you read about PAI in the French railway press, it's good old SSI with added je ne sais quoi.
Which highlights what Burch sees as one of the ironies of Railttrack's demand for ‘standard wizzy new European technology'. ‘If you measure it by the number of countries using it, the most standard interlocking technology in Europe is SSI'.
There is another benefit of Alstom's international sales of SSI. When BR was privatised Alstom did not buy one of the two signalling design offices. This meant the company did not expand its design resources, equally it did not have to lay them off when the market dived.
But, exports meant that Alstom companies in other countries became SSI literate. Thus, the South African company employs IRSE Licensed staff who are busy working on equipment case design for two SSI based projects in the UK and on the basis of this work is double its capacity.
Similar with SSI experienced engineers from Hong Kong and in Australia . And Alstom is currently recruiting in India . ‘We are bringing our global reach to bear', Burch declares, parodying a thousand corporate videos.
After a brief discussion as to whether reach can actually be brought to bear, we decide it sounds to good not to share with the Modern Railways readership.
Of course, despite continuing success in Europe , SSI is an old technology and neither of its UK parents, Alstom and Westinghouse Signal, are putting major development effort into the basic interlocking.
Clearly, the future lies in the Computer Based Interlocking where Alstom's centre of excellence is in Italy . Talking of Italy , Burch cannot resist a mini-commercial. While the Ansaldo ACC interlocking which will go into Manchester South is installed at Roma Termini, Alstom's CBI, acquired with SASIB, is installed 7km down the line at Roma Ostiense.
When will it arrive in Britain ? The answer is 2003. But it won't be unveiled in a dramatic application like Manchester South. Alstom is planning an evolutionary approach, starting with SSI.
First to come will be what Railtrack calls SSI plus.
SSI's big weakness is not the processor but the manpower intensive data preparation – the ‘front end'. CBI systems such as Atranz' Ebilock automate much of this through computer aided design (CAD). You draw the track layout and signals on the CAD screen and the signalling principles are applied automatically in the software – not to mention the production of manufacturing instructions and bills of quantities.
In contrast, with SSI skilled signal engineers and technicians sit at a work station and effectively design the system manually. Not only is this labour intensive, it makes changes or upgrades lengthy and expensive too.
So, at both Alstom and Westinghouse investment is going into developing software that will provide ‘smarter tools' at the front end. Thus, as with the CBI systems, it will be possible with SSI to produce the scheme plan on a computer screen and from this produce the control tables automatically, together with other detailed requirements.
While faintly embarrassed by a platform running at a 2MHz clock speed, Burch sees the new greatly enhanced SSI front end as giving real benefit, particularly when it comes to the industry's key limiting factor - skilled signalling engineers.
‘We tie up good signalling engineers doing data prep' and doing other things manually, checking and rechecking and as a result we constrain the capacity of the industry'. Automating the SSI front end frees this manpower to take on more useful and productive work. And the computerised front end can ‘churn out the work day and night, free from human error and unaffected by tiredness, holidays and sickness'.
So, SSI plus combines new front end with the established processor. And the fact that SSI runs at 2MHz in ‘Turbo' form is only really of interest to technophiles who have to have the latest kit first – as the export demand shows.
Alstom's strategy sees SSI plus migrating to CBI, retaining the new front end. Remember, too, that SSI comes as a system with its own trackside modules and communications equipment which represents a substantial part of the overall costs of a signalling scheme.
So CBI will retain downstream compatibility with the SSI trackside hardware. The final evolution is a fully integrated CBI system.
So reports of SSI's demise as a product are premature. Burch sees SSI being produced for minor extensions and spares for another 10-15 years. And this too presents a challenge, since product support has to be available.
In the European signalling industry there is a ‘big six' of major suppliers. Even with the Adtranz sale there will be a ‘big five' compared with, say, the ‘big three' in traction and rolling stock.
So, is more rationalisation imminent? Probably not, Birch muses, as he consults his crystal ball. Unlike the over-capacity in the European rolling stock market, signalling demand is quite high and the growing internationalisation of products and systems means that the manufacturers can follow the investment cycle round Europe.
All the big groups, says Burch, want to get better at using resources across national boundaries, so that if Railtrack, say, were to enter an investment lull, as was the case until relatively recently, it would not have to worry, because the UK design offices and factories were working on schemes in Italy, Germany and France.
With the signalling industry bearing fresh scars from the last Railtrack investment famine that ended last summer, during which Adtranz, Alstom and Westinghouse all had to lay off staff, you can see why a pan European market appeals.
Another contrast with the rolling stock industry is the continuing presence of a number of what could be called second tier signalling suppliers. Burch sees this as a function of railway signalling's status, even after 150 years, as a new and developing technology.
This could create room for a smart software company, for example, to spot a niche and enter the industry. Burch suggests information systems and information management applied to train control as one possibility. Software, of course, does not require investment in a factory producing hardware backed by design teams and the panoply of safety acceptance.
I think the whole signalling industry, suppliers and customers, has got to get hold of this notion that increasingly, as with other industries that use modern technology, software is going to be the dominant factor. It's what we can provide by modern software techniques that's going to drive a lot of the benefits'. Charles Burch |
So far we have been in overview mode. But Charles Burch is also responsible for has Alstom's highest profile signalling – the West Coast Main Line Train Control System.
Put bluntly, I suggested , it was Alstom's inability to deliver radio based cab signalling by 2005 which wrecked the West Coast Route Modernisation budget and brought the whole project into the public eye. Seeing as how none of the crusty old signalmen of my acquaintance thought it would work anyway, how did Burch plead?
‘At the time we bid, we believed that we could deliver the system' says Burch, adding ‘you have to believe the bid at the time you are making it'.
But what became clear after the bid, and bidding was a very protracted process (see the Modern Railways WCRM feature in the June 2000 issue), was that the complexity and ‘the number of dimensions involved' in delivering a new control system for the WCML were more than anyone had envisaged. ‘Railtrack have been honest about that and we should be too', Burch adds.
Indeed, if Alstom was asked whether it would rather still be delivering the original solution Burch's view would be, ‘what's the point, it's clearly not what's wanted'. The original proposal was, he believes, right at the time given the knowledge available. But then events moved on ‘quite quickly, and it became clear that (radio based) moving block on its own ‘was not going to be a panacea'.
In particular, a significant amount of conventional signalling hardware was going to be retained and replaced which robbed moving block of its cost advantage. That left the capacity and performance advantages.
These are ‘undoubtedly' worth having, but, according to Burch, ‘on a mixed traffic railway like the WCML you don't get the order of magnitude advantage you enjoy on a metro'. Another factor was the number of cabs to be fitted. ‘One of the things that put paid to moving block on the WCML was the prospect of taking 2 000 cabs out of service and physically fitting the TCS equipment'.
Franchise replacement compounded the problem. Most bidders seem to be offering new trains, in which case fitting TCS to existing stock could generate much pain with no long term gain.
Burch thinks that it was these factors, rather than ‘technology fear' that drove the decision to drop moving block for WCRM Phase 2 in 2005.
But he does not write off transmission based moving block on the main line. Sometime between 2010 and 2020, Burch believes that it will have become the accepted form of signalling because of the cost savings on trackside hardware and its maintenance. But to achieve this, a ‘critical mass' of the train fleet will have to be equipped, or ready to be equipped – hence the concept of ‘TCS Ready'.
This is a specification which allows train builders to design in provision for future TCS fitment. It details such things as cable runs, space for equipment cases and brackets on the bogies for the aerials that read the track mounted transponders. The first application is on Virgin's West Coast Pendolino fleet.
Burch believes that every new train should now be built TCS Ready, even if the track it will on currently has conventional lineside signalling. Making it happen could be something for the current joint inquiry into train protection systems by Prof Uff and Lord cullen.
While moving block has gone, the Level 2 European Train control system to be installed on the WCML south of crew will still provide radio transmission based cab signalling taking its data from trackside interlockings.
This requires the Radio Block Centre (RBC), which is really just a clever digital telephone exchange which asks the interlockings for their signal aspects and radios the information to each train where it appears on a cab display.
Initially the RBC was intended to interface with SSI, but now it will also have to be able to talk to two types of CBI. This means that the RBC will have to deal with two types of architecture – the decentralised SSI and the centralised CBI. ‘It's not as simple as falling off a long' Burch concludes, ‘but it can be done'.
Fortunately the viscosity of the treacle is not increasing everywhere. For example, the Network Management Centre (NMC) will now instruct the interlockings to set routes rather than access the central computer that would have managed vital safety in the original scheme. The NMC is a separate project being handled by Union Switch & Signal of the United States .
As I move on to the next question, Burch interrupts to answer a criticism I haven't made, although some have, apparently. When are we going to see something happening on the ground on the WCML?
As we spoke, excavations had begun at Stoke for the foundations of the trackside processing centre – no doubt to become the TPC. The building is due to be completed in January next year and it will then be filled with the radio block centre equipment and telecomms kit. Wireless mast foundations will also be appearing.
Physical things are starting to happen in return for all this money that's being invested Charles Burch |
On the UK market in general, Burch welcomes the arrival of alliances. A recent example was the Sunderland Metro extension where, after being selected as preferred bidder, his company had 16 weeks to work with Railtrack developing the scheme in far more detail than the bid contained. At the end of that period, customer and contractor had a ‘robust plan', understood the risks and knew what to do about them.
This partnership is part of a wider alliance in which Adtranz is responsible for the electrical and mechanical fixed installation and Christiani & Neilson is the civils contractor. ‘It is working well', says Burch. ‘We like alliances', he adds, even though they are ‘no bed of' roses and you have to work at it. But ‘it's the right way to go'.
While we have concentrated on mainstream signalling projects, Burch reminds me that there is considerable scope for the extension of the Group's TBL Automatic Train Protection system on the Great Western Main Line and that at least one TOC manager, who had better be nameless, is looking a various application of cab signalling.
Then, all those franchise replacement bidders who are blithely offering tilting trains are bringing a smile to Charles Burch's face since tilting trains will have to be fitted with Tilt Authorisation & Speed Supervision (TASS) which requires TCS type equipment for each cab and balises every 5km or so along the track.
Burch describes TASS as a ‘small cog sitting at the middle of a lot of wheels'. By the time you have fitted a train with TASS, the addition of radio and more frequent balises gets you close to ETCS. So how train operators implement TASS is important since, it is a precursor to ETCS and, most important, the associated full ATP.
On top of all this activity, Alstom is also a member of the TubeRail consortium short-listed for the JNP infraco in the bidding for the London Underground Private Public Partnership.
JNP stands for Jubilee, Northern, Piccadilly which means that the winner of this competition will have a lot of resignalling to do and the long term funding to make it happen.
This is obviously another challenge to an already stretched European signalling industry. Although London Underground has substantial signal engineering resources which will be transferred with the infraco, this sudden demand will add to an upswing in Railtrack's main line cycle, rather than filling a trough.
Which explains the job advertisements for signal engineers and the enthusiasm, mentioned earlier, of Westinghouse and Alstom to get an automated front-end on SSI plus.
Assuming TubeRail is successful, and possibly even if other bidders without an in house signalling capability take over the infracos, Alstom is looking to supply the French SACEM signalling system for the Underground, long Westinghouse territory. Except that it isn't called SACEM, which was a perfectly good French acronym, but URBALIS, another of those funny names which all the continental transport groups impose on us.
If Alstom gets into the Underground business, the signalling system design will be handled in the UK with the core technology coming from France and the hardware manufactured throughout the group.
So, to conclude, I ask Charles Burch whether he regretted coming into signalling? ‘Absolutely not. It's difficult to talk about the railway industry without resorting to clichés but there's a huge amount of change. Markets opening up, investment is high, all my colleagues in Europe envy me as the guy looking at THE growth market in Europe if not the world. It's a fascinating position'. Phew, all that and the West Coast too.
Return to 21st Century Ford Archive.