On the 9th October, at Kings Cross, TfL unveiled the latest concept for the New Tube for London (NTfL) – the rolling stock design that will serve the majority of the “Deep Tube” lines, beginning with the Piccadilly Line in 2022. For a good percentage of its life, across many lines, this train will have a driver.

The last part of that introductory paragraph may sound like a strange thing to say. It is a topic to which we will return later. First, however, let’s take a look at the design itself.

Taking a leaf from the design process for the New Bus for London, TfL have partnered with design consultancy PriestmanGoode to produce the new concept design, and whilst this means that one should not take it as read that the final product will exactly match the concept, TfL have indicated that they fully expect it to have the same general look, feel, and proportions.

Visually speaking, it is certainly a striking design – one that seems to meld the almost other-worldly concept design put together for TfL by Siemens some years back (when the NTfL concept was still being referred to as the “Evo”) with the colours and curves to be found in the newer rolling stock designs already to be found elsewhere on the network.


The external view of the cab


Concept image from the front


Pictured at a platform


Another angle from a platform


With open doors. Note the green door edges – the concept design features these changing from green to red as the doors are about to close.

Indeed when placed alongside the 2009 Stock (Victoria Line) and the S-Stock (The now-universal rolling stock for the Sub-Surface Lines) there is at least a sense of consistency and design evolution. A nice touch given that together these will form the combined Underground fleet for the next few decades at least.

As was previously confirmed, the NTfL will be future-proofed for full unmanned Automatic Train Operation (ATO) but not run with it out-of-the-box. In its most obvious dimension, this means that whilst it will feature a full drivers cab, the units will be designed in such a way as to allow this to be removed in future should they no longer be required. We will discuss just what automation means and when (if ever) it is likely to happen later.

Overall, carriages are shorter with two, wider door sets per side and walk through access. Inside it feels somehow both retro and futuristic – the result, perhaps, of the combination of curves, lines and grills clearly in part necessitated by the inclusion of air-cooling.


Looking down the train


Seating at the end of the carriages


Seating in the middle of the carriage. Note the air-cooling unit beneath the seats, and the wider window stanchions for air circulation.


A closer view of the air-cooling unit

Here it is perhaps important to pause and note briefly that air-cooling is different from air-conditioning. The latter allows for full temperature control whilst the former makes use of water and air humidity to allow temperatures to be lowered as long as the air itself can be vented regularly.


How air cooling would work, extracting warm air from the bottom and forcing it round and in from the top having been cooled.

The NTfL will be air-cooled, simply because full air-conditioning is not possible to implement on the Deep Tube lines. The difference is important to note as it will likely mean that whilst some element of temperature control will exist once the new trains are in service, those expecting a fully cooled experience in the hottest of temperatures (particularly on the Central Line) will be disappointed. Air-cooling should make things better, but sadly newspaper articles comparing Tube temperatures to acceptable livestock carriage conditions will likely remain an annual feature going forward.

In terms of timescales and train orders, the initial order will be for 250 trains – 100 each for the Piccadilly and Central Lines, 40 for the Bakerloo and 10 for the Waterloo & City. A formal Invitation to Tender is expected to be issued in early 2015 and the contract awarded in 2016, with what one might regard as “the usual suspects” shortlisted for bidding – Alstom, Siemens, Hitachi, CAF and Bombardier. The anticipated delivery dates by line can be found later in this article.

TfL have confirmed that the current schedule for deployment begins with the Piccadilly Line in 2022, to be finished by 2025. This is to be followed by the Bakerloo Line, completing in 2027 and the Central Line/Waterloo & City to be completed by 2032.

Why London Needs a New Train

The fact that new rolling stock was coming, of course, was not in itself news. The trains on the Deep Tube – most notably the Bakerloo and Piccadilly Lines – has been looking increasingly long in the tooth for some time, and despite their proficiency for it there is only so much asset-sweating that London Underground can do. Indeed by the time of its eventual retirement the venerable 1972 Stock that can be found in some form on both the Bakerloo and Piccadilly Lines will be well on its way to qualifying for a freedom pass. Meanwhile the Central Line’s 1992 Stock (also used by the Waterloo and City) has always had its share of issues, most notably with its DC motors, meaning it was never likely to reach the same service milestones as the ’72.

There has also always been the question of signalling and line upgrades. Both the Piccadilly and Bakerloo Lines are slated separately for full resignalling and/or extension soon (that’s soon in the kind of timescales transport planners think in at least – which is anything under 10 years), and neither of those activities is really possible without replacing the rolling stock. New signalling means new signalling kit, and retro-fitting that into old trains is something that any self-respecting transport authority tries to avoid. Even when it’s technically possible, it’s expensive to do and that’s money that generally carries no long term benefit if the rolling stock is shortly to be replaced.

Extending lines, meanwhile, means you need more trains to cover the longer distance they travel – else the gap between services increases and you lose much of the benefit of extending in the first place. Sometimes this can be achieved by “cascading” rolling stock freed up by putting newer rolling stock in place elsewhere and on the surface this may seem to be a valid option for the Bakerloo if the Piccadilly alone were to get new trains. Armchair planners (and Treasury officials) should be aware, however, that in this instance this actually wouldn’t be as simple as it sounds. Although both Lines share the same rolling stock designation, the curves on the Bakerloo are actually tighter than they are on the Piccadilly in a number of places, something current Bakerloo Line fleet is adapted to deal with. Adapting the Piccadilly Stock to run on the Bakerloo would in itself be an expensive activity, if indeed it were possible at all.

Ageing assets and line upgrades, then, mean that new stock is an imperative for all Deep Tube lines, excepting the Victoria (which already received new rolling stock in 2009) and both the Northern and Jubilee Lines, where the relatively sturdy 1995 Stock has already been adapted to the new signalling upgrades both lines have received. Indeed the Northern is actually a rare instance where, to meet the need for extra trains to service the planned Battersea Extension, it has been deemed financially viable to place a small order for what will effectively be replicas of the existing rolling stock.

Moving with the times

These are not the only reasons why new rolling stock is needed, of course, merely the key drivers behind what will be a huge project – one that will run to somewhere between £1 – 2.5bn, depending on how it is finally tendered. Other benefits to updating the rolling stock are also relatively easy to find and understand – engineering, materials and technology have come a long way since the beginning of the century, let alone before. Lighter weight materials and more efficient (and regenerative) drive and brake systems mean operating efficiency savings are there for the taking, and for a network as large as the Underground these could provide a substantial ongoing reduction in costs. Indeed TfL have confirmed that not only does regenerative braking capability remain a specification requirement for the NTfL, but that, as we previously suggested, the aim is to upgrade and configure the traction power supply network to allow the maximum recovery of braking energy for consumption by other trains. More, the goal is in fact to go even further and look at feeding back surplus regenerated power into the grid supply for use elsewhere on the network.

Meanwhile as the Underground has gotten busier and hotter (thanks to both increasing passenger and train numbers and a century or so of heat seeping out of the tunnels into the surrounding earth) the need to make it a more pleasant environment for travellers has also become more imperative. As indicated above, with air-cooling the redesigned rolling stock can help mitigate – although never entirely address – that need.

The (Presumably Driverless) Elephant in the Room

Of course if you were to base your knowledge of both the NTfL and, more importantly, why it was being built on the coverage in the press you could be forgiven for thinking that one of the primary drivers for building the NTfL was… well… something else – getting rid of train drivers.

Revealed: Inside the new ‘driverless’ Tube trains to be phased in on London Underground from 2022 proclaimed The Evening Standard about a train that will absolutely, with 100% certainty, have a driver in 2022. ‘Driverless’ Tube trains: See inside TfL’s new fleet for London Underground, said The Independent relying, like the Standard, on the magical powers of the single quotation mark to protect from the demands of factual reporting. The Telegraph took a similar line with New ‘driverless’ tube trains unveiled by TFL, but at least admitted that the train would indeed have a driver in the lede.

Potentially driverless tube trains unveiled as union suspends strike announced The Guardian who opted for accuracy over that fickle mistress, the single quotation mark, but then promptly undid all their good work by implicitly suggesting a link to a strike that is not about Tube drivers.

There were plenty more, but the point has perhaps been made – to state that coverage was skewed around one particular aspect of future-proofing contained within the design is not hard to see.

The trouble is that whilst it would be ludicrous to claim that the ability to run fully automatic – and potentially unmanned – isn’t part of the design specification, it would be equally ludicrous to claim that it is one of the primary objectives right now.

For despite the occasionally breathless coverage from the media and comments from politicians, the truth is that the NTfL does not signal a tolling bell for the job of Tube driver. Put simply if you are the parent of a child who yearns one day for the pull of the lever and the thrill of the door button, then fear not. We can state with near certainty that by the time they are old enough to enter the work force the Underground will have more, not fewer, Tube drivers than it has now. Indeed should you wish to start a Tube-driving dynasty, then you can be reasonably confident there will be driving jobs on the Underground for your grandchildren as well.

The needs of the modern Tube

Given the amount of hot air the subject garners you could be forgiven for being confused by the previous statement. From a practical perspective, however, it is simply the way things must be.

In the short term (and here, as before, we’re talking in transport planning terms) the number of services run across the majority of London’s Underground lines is going to increase. Meeting those extra service commitments needs extra drivers. The impending launch of the Night Tube also means more drivers will be required for similar reasons.

In an ideal world, introducing new signalling and rolling stock would mean an opportunity to step back and look at how the network operates – perhaps introducing greater levels of automation, where appropriate, at the same time. For an existing metro system like the Underground, however, adapting to fully automated operation (or more) requires extensive (and expensive) infrastructure work as well, something that TfL – an organistion whose business planning is essentially a process of financial triage – just cannot afford to prioritise. New signalling and new trains are critical to meet the needs of the network and thus any changes to operating practices must take a place at the back of the line. Indeed the importance of getting the NTfL in service as quickly as possible has, if anything, only increased in light of the current goal of extending the Bakerloo, for as we highlighted in our look at the Bakerloo Line Extension Consultation the project is effectively dependent on having the new rolling stock ready to go.

Under previous plans the Bakerloo was understood to be at the bottom of the pile when it came to replacement timescales, leading us to question TfL’s planned delivery dates for the Extension. As indicated above, however, the Bakerloo now ranks behind the Piccadilly in terms of the NTfL delivery timetable, underlying just how important to that project getting the new trains in to service will be.

A slippery term

For those only familiar with the “driverless” narrative as it manifests in the press, the idea that it requires more than just the right trains and the right signalling may come as a surprise. In part though this is because when the subject is discussed people rarely stop to define what they actually mean by “driverless” in the first place. Automatic Train Operation is already in place on several Underground Lines, yet ask the average Victoria Line user whether their train has a driver or not and they’ll give you a funny look. This is because the train is still manned by a Train Operator, whose job encompasses a number of key tasks even though the actual movement of the train is controlled elsewhere.

The truth is that all too often the term “driverless” is actually dog-whistle politics for something more extreme – full Unmanned Train Operation (UTO) – not least because for many on the political right it is the removal of the influence of the Unions that is seen as a key objective – and that means taking people out of the service pattern as completely as possible.

And therein lies the extra cost, for whilst controlling a train’s motion is easy to do remotely (whether it is a human being or a computer at the controls) guaranteeing passenger safety during boarding and alighting trains – both planned and unplanned – can be significantly tougher.

Your favourite metro system sucks

For a brand new metro line, with access to modern technology and transport network planning, the issue of safe boarding and alighting is one that is relatively easy to address. By fitting Platform Edge Doors (PEDs) to your nice, relatively straight station platforms you drastically reduce the risk of passengers having issues interfacing with the train. You can also build your tunnels wide enough to have clear, safe escape paths that allow both passengers stuck between stations to detrain to a safe space in an emergency and which allow station staff to reach them quickly and safely escort them out. Indeed you can also try and limit the chance of a train being caught between stations at all by having rolling stock that has batteries to allow them to reach the next station in the event of a complete power failure, one of the prime causes for tunnel detraining.

That last feature is notably something that TfL have confirmed they will be future-proofing the NTfL – it will feature emergency batteries. For an old, complex network like the Underground, however, the other safety elements are very tricky (and potentially expensive) challenges indeed.

Being PEDantic

Due to their obvious safety benefit, PEDs have become something of an implicit requirement for full UTO operation on modern metro systems. Indeed although TfL were (understandably) unwilling to provide absolute confirmation that they consider them to be a prerequisite for any move towards even full ATO, let alone UTO, they were happy to confirm that represents their current thinking at least.

Platform Edge Doors are considered the most suitable solution for managing safety at the platform train interface and a likely to feature in a system solution for fully automatic train operation on the deep Tube lines.

PEDs however, present problems on the Underground. This is a subject we tackled in more detail in our article on the Piccadilly Line upgrade at the end of last year, and those particularly interested in the details problems there are recommended to read that article. In short, however, there are two key problems.

Firstly, PEDs and sharply curved platforms don’t play nicely together. This is because train carriages are, obviously, rectangular and thus stick out over the platform, however slightly, as they pass round any major curve. That makes PEDs difficult to site, as it is imperative that any gap between the PEDs and the carriage doors is kept to a minimum (amongst other things to prevent people getting trapped between the layers) yet anyone who regularly uses the Tube should have no problem bringing to mind the name of at least one station they’re familiar with on the Deep Tube network where a seriously curved platform makes this a major issue.

There are solutions to this problem, certainly. The obvious one is to re-site or straighten the platforms, but that is far from cheap. Interestingly TfL’s concept design for the NTfL highlights one part of another potential solution – make your carriages shorter. Making the NTfL walkthrough is not just to follow current trends elsewhere on the network – such as the S Stock or the 378s running on the London Overground. It allows you to make your individual carriages shorter, giving them a tighter turning circle.

PEDs, however, can do nother about the Underground’s legendary “gap”, another potential risk particularly on curved platforms. We suggested last year that TfL had begun to explore the possibility of extendable, mechanical gap fillers on curved platforms and as part of this announcement they have confirmed that this is very much the case. The practicality of such a solution is being actively explored.


How PEDs might look with the NTfL on the Piccadilly Line

Secondly, and perhaps more critically, the installation of PEDs at a station means all of the trains that use a particular platform must have their doors in the same place. This is a simple, but oft overlooked, basic requirement for PEDs – because if your doors don’t open where the gap is then people obviously can’t board or alight.

For an enclosed network with universal rolling stock design this is obviously a no brainer. Unfortunately, the Underground is neither. Both the Piccadilly and Bakerloo Lines share track and platforms with other lines – indeed in the case of the Bakerloo that shared space is not even with the Underground, it is with National Rail. The simple fact is that as long as those shared interfaces exist it seems we will not see full ATO or UTO operation on either of those lines.

Again, those looking for particular details as to how TfL may address this issue for the Piccadilly Line are directed to our previous coverage of the subject. Bluntly, it is for this reason that Turnham Green services will pass to the Piccadilly, rather than any kind of successful local user campaign (no matter how both the Mayor and TfL themselves have painted it). It is also for this reason that we may actually see something rather radical – the transfer of the Ealing Broadway branch of the District Line to the Piccadilly Line in its entirety. Again, TfL have now confirmed that this is definitely an option, but indicate that a decision or consultation on the subject yet to be made.

On the Bakerloo, however, the options are less clear. For this reason TfL have confirmed that as long as the line shares track and stations with London Overground services between Queen’s Park and Harrow & Wealdstone automation is likely to be limited.

The Need for Realism

Both for those who have long fretted as to whether a more generalised deep tube stock would appear at all, and for the regular passenger at large, the announcement of the current plans for the NTfL is exciting news. Should the final product respect the concept designs on show then it’ll certainly be a striking piece of engineering – one that cleverly addresses many of the subtle difficulties that make the London Underground such a unique (and sadly often expensive) engineering challenge.

Those expecting in 2022 to board a “driverless” train – at least as they understand the term – will almost certainly be disappointed. It is true that the NTfL will be future-proofed for that purpose, but having trains that support full ATO, or indeed UTO, is just one of the many very real barriers to implementing that operational system in the Capital. On some, indeed perhaps on all, of the lines that form the Deep Tube those challenges can be negotiated. But doing so will not be cheap. It will likely require not just changes to the way the Underground is run, but to the way its platforms look and more. Indeed in some cases it may even require changes to where those Lines actually stop.

This is not to say that it won’t happen, simply that right now it is as much an aspiration as a clear cut objective – and very much a long term one at that.

It has become popular in both the media and particular political circles to pretend that the above isn’t the case, as if somehow having the right type of train and making a few changes will make a gigantic shift in the way the 150 year network is setup and operated happen almost overnight. It’s a sort of intellectual sleight-of-hand that does everyone, whether they are a reader or voter, a serious injustice.

Just as importantly, perhaps, the longer that narrative is perpetrated the greater the risk that a future decision as to whether to staff or not staff trains is made for political, not transport reasons. Or, worse, that such a decision is treated as a black-and-white one to be applied to all Underground lines regardless of fit.

All those involved in the debate on the future of London’s transport have a duty to avoid such a mistake at all costs. At the very least every statement about “driverless” trains should be matched with a demand by the listener, or reader, for an explanation of what that word actually means – ATO, full UTO or something else?

The Most Likely Future

Grave warnings, however, are never a good thing on which to finish. So setting mass-media hyperbole aside and looking at both what we knew already and the further information made available over the last week by TfL, what does the current strategy for future automation appear to be?

In fact on current evidence, the answer actually relatively straightforward. With the order in which the NTfL will be rolled out onto the lines changed to allow the Bakerloo Line to be completed second, it seems likely that the first major step towards full automation is likely be seen on the Central Line, if not from commissioning in 2032 then shortly after.

This is because, as is always worth remembering that, like the Victoria line, the Central Line is already fully ATO and has been since 2001. It also has the very obvious benefit of not sharing track with any other lines. With more time now to find a solution to the PED issue and both plan and cost it, and with the NTfL by that point having worked out any implementation kinks on the other two lines, a quick push to full automation might not be far outside the realms of possibility.


Conceptually how a control centre may look

Indeed the above situation is something that TfL themselves are quite happy to acknowledge:

The Central line is unconstrained by inter-working of other services and fully automatic operation is expected to be introduced on the line in the early 2030s.

Of course if the above is looking likely, then perhaps it will actually be the Waterloo & City line which would, officially at least, be the very first line to make the actual transition. Traditionally it has shared both management and staff with the Central Line and thanks to its nature as a short two-station-only design might make a more risk-free test bed for a full Central Line deployment.

Whatever the case, realistically speaking it seems relatively safe to say that when it comes to both the Underground and the NTfL, those looking for big official milestones in both further levels of ATO and in UTO operation should for now perhaps be marking 2032 in their diaries and no sooner.

Circumstances may change, of course, as even for transport planners 18 years is quite a long time. In the worlds of both politics and the media though, it is practically an eternity – and that’s something that those debating the subject in those particular spheres would do well to remember.

More information on the New Tube for London can be found on the TfL website here, as can a brief concept video.

Written by John Bull
John Bull is the Editor of London Reconnections. A transport journalist and historian, his writing often focuses on the political or strategic challenges facing London's transport network and beyond.