The word “driverless” does not feature in the Operational Strategy Discussion Paper once.

Given the amount of times this word has appeared in the press recently, the above statement may seem rather surprising. What it demonstrates, however, is that just as with the topic of ticket office closures, the Paper’s proposals with regards to Train Operators and operation are more complex than one might think.

Automation for the People

Sources indicate that the Paper lays out a brief overview of the current state of automation on the network and how it may progress. It is from here that much of the talk of “driverless” trains has come, yet the actual breakdown seems to be far more nuanced.

The Paper divides methods of operation into three categories, which will be familiar to many of our readers:

  • Manual Operation: the driver controls all aspects of the train’s operation
  • Automatic Train Operation (ATO): the driver retains the “stop/go” element of operation, as well as other aspects such as door control
  • Remote Operation: All motion control is operated remotely, with other aspects such as door operation are operable either remotely or manually.

Sources suggest that the paper identifies a number of stages through which London Underground will pass in the coming decade, which will see the balance between the above methods shift as time progresses:

Stage 1: The Network Now (70% Manual, 30% ATO) – The Victoria and Central Lines have had ATO systems in place for many years. To these has now been added the Jubilee as a result of its upgrade programme.

Stage 2: The Network Post-Northern Line Upgrade (50% Manual, 50% ATO) – The middle point of this decade will see the completion of the Northern Line upgrades, adding that to the ATO fold.

Stage 3: The Network Post-SSR Upgrade (20% Manual, 80% ATO) – the completion of the scheduled signalling and rolling stock upgrades on the Circle, District, Metropolitan and H&C Lines in circa. 2017/18 brings the Sub-Surface Railway (SSR) into the ATO fold.

Stage 4: The “Deep Tube” Programme begins (15% Manual, 80% ATO, 5% Remote) – “Deep Tube” is a topic to which we will return in more detail shortly. Effectively, Stage 4 represents the point post-2018 where elements of a Bakerloo and Piccadilly Line Upgrade begin to manifest.

Stage 5: Deep Tube Completed (80% ATO, 20% Remote) – The completion of Deep Tube in circa. 2020, bringing all lines into either ATO or Remote Operation.

Stage 6: The Far Future (100% Remote) – A hypothetical point beyond the current decade where time, money and technology enabled remote operation system-wide.

As can be seen, of the six stages above, stages 1 – 3 comprise entirely of plans and upgrades currently underway. It is with the addition of “Deep Tube” in Stage 4, that the Paper first starts to head into new territory.

Deep Tube

The Piccadilly Line Upgrade was one of the early casualties of the fall-out from the failure of PPP and the need to tighten the purse strings at TfL. Originally slated for the early-to-middle part of this decade, this upgrade project officially slipped into “tba” territory with the Spending Review.

Various sources have since suggested that London Underground had begun to look seriously at how the project might be recast – both to ensure the necessary upgrades took place and to ensure that they were rescoped and timed in line with other work that still needs to take place elsewhere.

The Deep Tube Programme is the result of this rescoping.

The first official signs of Deep Tube began to emerge in May, when it received a one-line mention in TfL’s Q4 Investment Programme Report. Since then, various details have begun to emerge, some of which this Paper seems to confirm.

Deep Tube focuses on the upgrade of signalling, communications infrastructure and rolling stock on the Bakerloo and Piccadilly Lines. It has long been understood that this includes a completely new rolling stock design for both lines (known to many as the EVO project), with both Siemens and Bomdardier recently confirming in the railway press that they have been undertaking some concept work in this area on London Underground’s behalf.

EVO (and indeed Deep Tube) are both topics which future articles outside of this series will focus on in more detail. In the context of this article, however, it appears that the Paper all but confirms two things, which sources suggest it seems to take ‘as read’ about Deep Tube:

  1. That the result of the EVO project will be a rolling stock solution that doesn’t feature a dedicated cab
  2. That this rolling stock will, out of the box, be designed for Remote Operation as well as ATO operation.

Sources suggest that the Paper proposes that it is with Deep Tube/EVO that Remote Operation be first introduced to the Underground, with the Bakerloo likely to be the first candidate for operation in this style.

Changes to Control and the SSR

Alongside the shift to an ATO/Remote Operated network, sources suggest the Paper includes a number of concepts for reshaping the Sub-Surface Railway.

The Paper, it seems, argues that the rollout of the S-Stock to the SSR, along with the move to a common control centre at Hammersmith as part of SSR upgrade programme, present a perfect opportunity to rethink and centralise its operation.

Sources suggest that the Paper advocates rationalising the SSR into a single SSR unit. This would include redrawing the existing Station Groups into SSR-wide ones grouped geographically, rather than line-specific groupings, and rationalising down Senior Staff into a single SSR management structure. In operation terms, the Paper also appears to suggest making Train Operators on the SSR non-line-specific, and thus available for any SSR line for which they have been qualified. Expanded pool coverage or a more zonal approach are also suggested for consideration.

Beyond the SSR

A more centralised, simplified operational approach is something that sources suggest the Paper argues for beyond simply the SSR. The current system of control centres at multiple locations is, it argues, inefficient and fragmentory. This contributes to bad incident management and a failure to sufficiently learn from mistakes.

The paper seems to suggest the creation of a single London Underground Control Centre within which the currently dispersed control functions would be co-located, including the British Transport Police. Effectively existing Line Service Control Teams would remain separate, but their relationship and responsibilities to the network as a whole would be more strictly defined and a new senior post of “Head of Railway Operations” would be created. The HRO would have formal command of the whole network, replacing the current Rostered Duty Officer role. The Service Control strategy would then be reviewed to take into account the new centralised element.

What this Means For Train Operators

As can be seen from all the above, the Paper highlights that even if only projects currently underway are continued, by 2017/18 the role of the Train Operator will have changed greatly from what it is now. Should Deep Tube proceed as appears to be planned, it further suggests that by 2020 the network will likely see ATO operation over all its Lines, with Remote Operation also beginning to feature.

Sources suggest that the Paper, rather bluntly, asserts that the requirements of the Train Operator role are already outdated in some cases. This, it claims, is largely due to both Union and cultural resistance, which has meant that whilst the technology on the network has begun to change, the roles and frameworks associated with the TOp role haven’t. This disconnect is one that will only grow greater as the various upgrade projects continue.

To address this, sources suggest the Paper advocates a phased plan aimed at rebalancing and refocusing the Train Operator role. The end result (as with the suggested changes to station staffing) would be Operators whose role included both operational and increased on-train customer-focused elements.

The first phase of this process, running concurrent with the upgrades already completed and those underway on the Northern Line, would be a reassessment of the TOp role itself on ATO lines, widening it out to include elements of Incident Management and Customer Service where possible. Minor changes to the current framework – including the option of voluntary overtime – would also be sought in order to increase the flexibility of the role and coverage.

Controversially, the Paper also appears to suggest the introduction of a new job role – Reserve Automatic Train Operators – within some existing support roles, effectively creating a pool of staff with limited “driving” ability outside of the existing Operator pool.

Phase 2, beginning with the rollout of the new SSR upgrades, would see the process of updating existing Operator frameworks continue. Alongside this, a new “Automatic Train Supervisor” role would be created, the occupants of which would be individuals qualified to operate services under ATO and restricted manual (i.e. very slow emergency manual moves).

Finally phase 3, concurrent with the rollout of Remote Operation under Deep Tube, would see the creation of “Train Attendants” for the Deep Tube Lines. These would be far closer in concept to the Train Captains found on the DLR or indeed traditional Train Guards, with the bulk of their role focused on revenue and customer functions both on-train and potentially on-station.

Sources indicate the Paper argues that a successful Remote Operation rollout might theoretically lead to the end of the need to have dedicated Train Attendants on each service throughout the day.

Managing the Trains

Alongside these changes to the TOp role, sources indicate that the Paper argues for a possible phasing out of the Duty Manager grade in the long term. Duty Managers are currently responsible for booking Train Operators on, undertaking various personnel management activities and for reforming the service during disruption. The paper, it appears, argues that a more technical solution should be sought for booking Operators on and off, with that and service reformation possibly become a function of Service Control whilst the other elements of the DTSM moved to a revised Instructor Operator grade.

The suggested staff changes to the Train Operator and DTSM roles above probably represent those within the Paper which will likely be seen as most controversial. Certainly, they’d represent the trickiest both to explain to, and receive buy-in for from, the Unions and staff themselves.

At their root, the changes suggested by the Paper cut to the heart of the Train Operator role, opening a debate that some would argue has been lurking on the radar for some time – as technology and infrastructure changes on the Underground, just what is the role of the Operator?

To a certain extent, it’s a debate that both the Unions and London Underground are at times guilty of oversimplifying in order to try and further their goals. The Unions (notably the RMT) are often guilty of portraying the TOp as the single, crucial point of control within the train – very much the “driver” and heir to the era of steam and early electrics. It’s an image that almost actively dares London Underground to take the opposite stance as ATO spreads across the network – that Operators are “button pushers” with technology and central control now doing the bulk of the heavy lifting.

Both views, however, are ultimately wrong – black and white caricatures of a role that is most definitely grey. Train Operators across all Underground lines are not simple stop/go merchants, but nor are they the drivers of old. “Safety” may be a word that is overused in the eternal grind of the driver debate, but it is very much a fact that the Operator role currently includes various elements related to proactive and reactive safety needs that would be tricky to automate. There are other day-to-day operational elements inherent in the role as well.

At the same time, however, just as technology has increasingly left the average car driver unable to simply pop the bonnet and fix their own car, the same thing is true of Tube trains. The days when an Operator was expected (and trained) to be part mechanic as well as driver are gone.

The simple fact is that, as the Paper rightly appears to point out, the nature of the Train Operator role has changed and will continue to change. Whether the Paper’s initial suggestions as to how this role should evolve are correct is open to debate – but here it is worth remembering that, as indicated in the first article in this series, the Paper is the early foundations of a possible strategy not the final product.

Whatever that ultimate strategy is (and signs do indeed suggest that this will end up being its foundation), for it to be successful London Underground will need to look carefully at not just what is wrong with the current Operator role, but also what is right. Indeed this is true of all the changes suggested in the Operation Strategy Discussion Paper.

Correctly identifying the things that need to be kept as well as those which need to be overhauled will prove a tricky challenge, as will convincing both staff and management that both need to be done (and, importantly, are being done for the right reasons). Sources suggest that this is something that the Paper does indeed acknowledge and accept, and so in the final part of our series, we will look at this topic in more detail, and also some of the barriers to change that London Underground is likely to face.

Next: Managing Organisational Change

Many thanks to the source who confirmed the lack of the word “driverless”

jump to the end
There are 46 comments on this article
  1. ASLEF shrugged says:

    On Changes to Control and the SSR

    The idea of non-line-specific TOps able to work any SSR line rests on them being qualified to do so. We’d certainly need to extend the training period to cover every mile of track, every possible move, every siding, crossover and depot but the real problem would be trying to ensure that the TOps were kept up to date by doing one trip on each line every six months, a bit of a headache for the admin people.

    What this Means For Train Operators

    ATO doesn’t significantly change the TOps role, trains till have to be worked out of depots or sidings manually and on the central Line the slightest drop of rain can bring a train to a standstill halfway into the platform. Or they just stop for no discernable reason in a tunnel.

    To be honest driving is the easy bit, my depot covers the Waterloo and City Line which is manual, doesn’t even have ATP, though I notice it doesn’t get a mention. I’d be delighted if that went Remote and with it’s nice straight platforms just begging for PEDs and it’s relatively uncomplicated track layout I thought it would be an ideal candidate for driverless conversion.

    Speaking of trains mine is between Mile End and Stratford, better go back to work.

  2. Greg Tingey says:

    Fully-auto on LUL is NOT an option.
    Too big, too complicted, too many things to go WORNG …
    100% ATO, is however, possible.
    But, it’s going to need willing co-operation, which, at present, neither LUL’s useless so-called management, nor Bob Crowe’s crypto-marxist union are capable of delivering.

    Oh dear

  3. timbeau says:

    automatic/remote operation is fine where there is a dedicated track and where there is no possibility of a collision with anything else -whether a similar vehicle or a stray animal or person. A lift in a lift shaft, or a cable car, or (just possibly) a line entirely in tunnel and having PEDs. Anything else, and I want a member of staff on board (even the DLR has that). And I want someone to be looking ahead. (I don’t like travelling on the DLR unless I can see ahead).
    So the member of staff on board might as well be the person looking ahead. And whilst he or she is there, they might as well drive the thing.

  4. Anonymous says:

    Mr Greg Tingey, I challenge you to back up your misspelled statements with evidence.

  5. Pedantic of Purley says:

    The thing I am surprised is not mentioned is an incremental change and have unmanned trains when running out of service to and from the depots. The opportunity missed was the Victoria Line Upgrade which has a long run between Seven Sisters station and the depot and is almost entirely in tunnel. I am sure there would be future opportunities such as Morden. This would be a logical “dip the toe in the water” without issues of safety of the travelling public and limit the options for Bob Crow to claim, as he always does, that the public are being put at risk.

    Obviously even this would be fairly drastic when fully implemented as the drivers would become entirely separate from the maintenance facility and would not have an opportunity to check that their train was “prepped” correctly. Furthermore their booking on point would be at a station and not the depot but this is the modern trend anyway. A further issue that I can think of is that without platform edge doors (PEDs) it may be advisable for only the drivers cab to be in the platform on arrival so that the driver can board the train and supervise the full entry into the platform – although the obvious alternative is to have the PEDs there in the first place.

    The benefit would be that almost all the train operator’s working day would be productive in operating trains actually in service and it would gain valuable experience in both in building and operating a more advanced mode of operation before taking this further – albeit now with a member of staff actually somewhere on the train whenever it was carrying passengers.

  6. EW says:

    I think remote operations with train captains on board is a good initial vision to aim for, as it obviously can work as we already have it on the DLR.

    “the Operator role currently includes various elements related to proactive and reactive safety needs that would be tricky to automate” Are you able to elaborate on what they are and how they are addressed on the DLR?

    I have ridden unmanned remotely operated trains in Singapore, Tokyo and the Heathrow T5 transit to T5B. On these systems the platforms are straight and they have platform edge doors which appear to be a sensible pre-requisites for unmanned operations. That would be tricky to implement at “mind the gap” stations where significant curves create a risk of getting stuck between the platform and train doors and were different types of stock serve the same platform.

  7. Ratty says:

    I’m not quite as risk averse as timbeau, but I have to admit that it is extremely comforting to have the person in control of the train (bus, plane) actually present on the vehicle, such that their life is also at risk of their actions. Sorry, but that’s just human nature.
    Regardless of its merits, one big remote-controlled accident would kill the remote operator idea stone dead.

  8. EW says:

    I recently saw a channel 4 documentary (Brave New World) showing a car that can drive itself safely in traffic with no human intervention. Its still in development but the laser and optical sensors and associated software seem to be approaching a point where you can use them to keep a look out and take appropriate action.

  9. ASLEF shrugged says:

    What this Means For Train Operators

    “the requirements of the Train Operator role are already outdated in some cases”

    Last night my “self driving” 92ts came to a juddering halt three times thanks to the rain and I had to drive it manually. Button pusher?

    “Minor changes to the current framework – including the option of voluntary overtime – would also be sought in order to increase the flexibility of the role and coverage.”

    Try major changes which won’t be achieved without agreement with ASLEF and RMT. While there are voluntary overtime agreements on the mainline rail ASLEF re-negotiate these on a regular basis and to be frank negotiation is not one of the LUL management’s strong points.

    “Reserve Automatic Train Operators”,

    This idea has been floated before, the main stumbling block is having adequate cover at their “normal” job to release them every six months to keep their licence up to date. And as they’d either be Station Staff or Duty Managers they would be union members, most likely RMT.

    “individuals qualified to operate services under ATO and restricted manual”

    Does S Stock not have Coded Manual, the intermediate setting whereby the train is driven manually but with ATP signalling? There are some long stretches between stations on the SSR and it would be useful if the driver could go at Line Speed rather than crawling along at 14kph.

    Managing the Trains

    Currently when IOps aren’t involved in training they are just like other TOps, this sounds as if they will be taken off driving duties and simply become DTSMs by another name with the addition of training and monitoring TOps.

    You are correct, TOps are not “drivers of old”, we do the guard’s job as well, that was why the name change came in. The introduction of ATO has for the most part taken out the actually driving element but as ATO, ATP and the trains themselves are less than perfect they have not eliminated it’s need.

    The main obstacle to Remote Operations will be installation. Currently Ligne 1 on the Paris Metro is being converted to driverless trains so it is possible but as we’ve seen with the Jubilee Line upgrade anything undertaken with LUL is likely to be a long, slow, messy and very expensive process.

    Before any new trains arrive the new signalling and control system will have to be installed in parallel with the existing one in order to keep the service running. That will require frequent weekend and/or early closures and on the Piccadilly that is not going to be popular with Heathrow. On the Bakerloo the problem lies north of Queen’s Park where the track is controlled by Network Rail and is also used, I believe, by London Overground who will still be operating old-style coloured light signalling.

    Once everything is in place the new trains will enter service on a one-in-one-out basis with driverless trains operating at the same time as manually driven trains until the last of the old stock is replaced, the lack of space at depots means that delivering the trains in advance and then sending them all out the first day the system is up and running is not an option. Until then the old system will still have to be retained.

    Deep Tube with EVO trains is a long way off and will not be cheap, with TfL’s budget being cut there is no hope of LUL being able to afford this in the foreseeable future. Think 2030 rather than 2020.

  10. ASLEF shrugged says:

    EW – Singapore, Tokyo, Heathrow T5 and all the other driverless rail systems were built that way, with straight platforms and PEDs. To the best of my knowledge the only line to have been converted to driverless is Paris Ligne 1 and that is still in progress.

    Curved platforms form a major problem, there are two vital ones I can think of on the Bakerloo, at Paddington and Waterloo. Not only would it be very difficult to fit PEDs but I have a nasty feeling that there isn’t a point on the platform where a Train Captain or whatever could see the entire length of the train which means that the only safe way to close up the doors is to use CCTV. So where can you put CCTV on a train or platform when there is no cab?

    Has anyone wondered where the Train Captain is going to stand in the cars at 08:30 on a Monday morning? If the train stops in a tunnel and they are half way down the train how would they know if they are being held at a red signal or if the train has developed a fault?

  11. IanVisits says:

    Considering the first sentence of this article, it was slightly amusing to read a press release from the RMT issued this morning warning of dire calamities if driverless trains are introduced.

  12. John Bull says:

    @ASLEF shrugged – to be honest, I’m inclined to agree with you that Remote Operation isn’t really practical on any existing line really at the moment – even within the longer timescales the Paper suggests.

    My instinct is that Deep Tube is likely to be the final step in making the network entirely ATO, with the rolling stock future-proofed for hypothetical RO in the future at some point. Installation issues and gremlins aside, from what I understand the underlying ATO methodology used on the Jubilee is actually rather good. I can see that becoming a model, of sorts, for the approach taken in future.

    I do think its always worth remembering that this wasn’t intended to be a public document – to a certain degree, it’s an aspirational Paper that naturally lacks some of the pragmatism that would inevitably come in were it to become a full strategy.

    The toned down public version of this Paper is, in fact, now out and we’ll touch on that in the final article in the series tomorrow. That’s much more guarded about the future of Remote Operation.

  13. Fandroid says:

    ‘They do things differently there’ – but I have personally experienced the driverless U-Bahn trains in Nuremberg which don’t have platform-edge doors. Not all trains have attendents either. All activities of these RO trains are done remotely, including moving in and out of sidings, coupling units (they are two-car units which can be doubled up for peak services), and door operation. Starting and stopping at stations is also automatic. There has been at least one fatality due to a person on the track at a station, but that was judged to have been unavoidable (even if manually driven) due to the imminent arrival of the train. The tracks at the stations are protected by CCTV and Radio Frequency detectors, which automatically halt incoming trains if triggered.

    If Uncle Bob is doubtful about RO, perhaps he should book a Christmas Market trip !

  14. Wildchild says:

    The sooner the human element is removed from train operation, the better. Never send a human to do a machine’s job.

  15. ASLEF shrugged says:

    For those of you obsessing about Bob Crow and RMT I would give you two points to consider.

    By the time any of this comes to pass Unkle Bob will probably be retired, he’s 50, he’s up for re-election sometime in the next six months and could well go after the next term. There were even rumours he’d be standing down this time but as I’m not an RMT member I wouldn’t like to comment on their accuracy.

    While RMT would lose members under the proposals it would actually have more influence on the Tube if train drivers were removed; they would represent the Attendants as they do on the DLR and they’d continue to be the main union in the Control Rooms.

    John Bull – “lacks some of the pragmatism”? You could say they were expressing thoughts they’d rather not share with their employees. Not much of a surprise; in last years staff survey “Speak Up” around two thirds of staff believed that management was less than honest in it’s communication with us.

    Sometimes I wonder if the main reason behind the push for automation is not so much economy but rather LUL’s inability to manage employees in a competent fashion.

  16. Anonymous says:

    I’d trust a computer system over humans enough of the time to be happy with fully auto. And I’m glad it’s going to more auto – makes sense.

  17. Anonymous says:

    It seems that Stages 1-3 in the main article are the only sure things to happen. Given that the Piccadilly Line interacts with the sub-surface railway at a number of points, and the Bakerloo Line interacts with National Rail, is there really much operational gain from running “driverless” trains? I suspect it will fall into the “nice to have, but too difficult” category.

    On a side-note, while the unions may have a more difficult relationship with LUL than with the TOCs on the national network, this could well be because the TOCs have very little incentive to reduce labour costs when they have such short franchises. For a TOC with only a short period until the end of the franchise, any gain from reduced labour costs is going to be drowned-out by the costs of any strike or other industrial action (a major reason for the salary gains for drivers since privatization). LUL has a longer time horizon. But this is a different discussion.

  18. Peter says:

    The document does’t say driverless, but it does describe train captains working inside the train – if there is nobody at the front of the train with an emergency brake, seeing the road ahead, then what is the train if not driverless?

    In fact the document says Phase 3: remote operation – remove the requirement to staff each train…as technology is proven further increase the level of unattended trains.

    If that isn’t ‘driverless’ then I’m not sure what is.

    BTW for anyone who wants to read the full, original document it is here –

  19. Greg Tingey says:

    Anonymous @ 9.15
    I DO NOT RESPOND to anonymous threats or poison-pen letters!
    I admit my proof-reading could be better …..

    Let’s take it one step at a time.

    ATO with driver/operator at the front works very well.
    We should see this across the whole “tube” network, no problem other than paying for it.
    Provided there are sufficient platform staff (see previous thread for scary comment about dispatching trains “blind”) this should see an improvement in safety and efficiency.
    Train captains, DLR-style – perhaps. But, even a six-car DLR train is considerably smaller and shorter than a 8-car deep-level tube train, never mind something on the Met lines. The most likely incident to cause serious damage is something in front of a moving train. Therefore, someone at the front is the logical place to put that person.

    Full automation, on a syatem as large, complex and diverse as ours?
    I think not. The possible failure modes are simply too many.
    And, as we all know, the automation often either doesn’t work, or partially-fails.
    Ask ASLEF shrugged.

    Or, as happened to me yesterday, case study …..
    GOBLIN line, 06.47 off Barking, about as we slowly rumbled across the ridiculous 20-limit over the Lea, the youngish, thinnish man leaning against the door-pillar oposite me (the train had apprently half-emptied, but I had estimated/counted 119 persons in a 64-seat coach!) either half-fainted or slipped on th floor, and very suddenly sild down with a resounding thud, his feet sticking out. I think he hit his head on the way down, and was concussed. He certainly didn’t move at all for about 20-30 seconds. We managed to help him up a bit, but he was clearly confused. I got the guard’s attention at S. Tottenham, we found he was getting off at Harringey GL, so the guard radioed to the station-staff there, and we proceeded, with vitim made to squat down.
    We passed him to the station-satff at HGL.
    The train was only about 4 late into Gospel Oak.

    Now replay that with no-one on the train.
    Now replay that if you have a “jumper”
    Now replay that if ……
    Truly “driverless is just NOT an option.
    Yes, I know others have it.
    Question… WHEN something goes WORNG on one of these supposedly fully-automated systems, how long does the consequent shut-down last?

    And, the other/same anon at 11/46: I’d trust a computer system over humans enough of the time to be happy with fully auto.
    And MS WIndows is a fully-safe o/s ????
    Nothing can possibly go worng ……
    ALWAYS, always, make sure you have a back-up.
    Simple engineering principle.

  20. Mwmbwls says:

    ASLEF shrugged @ 11.38 on the 3rd November 2011 makes an important point concerning the need to carry all those who have a stake in the future of London’s Transport system along with the change process. He/she points out that a credibiity gap exists between the workforce and the management as evidenced by the employee surveys. This is both good and bad news. The good news is that at least there is a mechanism in place to measure whether such a gap exists and how great it is. The bad news is that the results are in terms of benchmark performance companies astonishingly bad. TfL needs to focus on raising levels of organisational trust if any of its aspirations are going to get off the ground. Trust is a willingness to be dependent on the actions of others. There is a subtle but important difference between not knowing where you stand and knowing where you stand and not liking it. The first condition can be summarised simply as “lost” whereas the second at least means you know what is wrong and therefore can seek to ameliorate the position.
    Without wishing to go into mushy “hearts and minds” mode – the key mantra for both TfL, the unions and the workforce has to be “Engagement, engagement, engagement”.
    In their oversight role perhaps the London Assembly Transport Committee might/should find time to look into the state of employee relations as evidenced by the employee survey. Again without wishing to sound like a management text book – what gets measured gets managed – it really does – that is as long as somebody reviews the data and imbues it with relevance.

  21. EW says:

    Despite the drawbacks being pointed out with the train captain and remote operation models. I still think the document is right to consider them so that they are explored thoroughly and accepted or dismissed for the right reasons. Initially it will come down to if the safety and operational risks it may bring can be mitigated or brought down to a level that is acceptable at a cost that justify the benefits. After that it ultimately becomes a question of getting all the stakeholders to buy into the strategy and as pointed out by many commentators that will be the biggest challenge for these strategies.

    Aslef Shrugged made a good point in pointing out that most full automation systems are on lines that had been designed with that from the start and that Paris is probably the only instance where a line has been converted. LU can learn a lot from that project which has to deal with some of the problems that the Tube will such as curved platforms.

    With tablet computing developing rapidly, is it not possible in the future that the train captains could used them to see the images from the platform cctv and other information that they would have obtained if they were in a cab? Sensors can potentially do a better job than a human at monitoring if the track ahead is clear as they don’t get distracted. Doing away with a cab will also provide a small increase in capacity.

    As a customer I have no problem with with computers running the tube so long as the system is designed to be reliable, resilient, are designed to fail safely and that effective procedures are in place for recovery when failures or incidents occur.

    Humans are often the weak link in any system as we easily get bored, distracted, overloaded and have subjective judgement that varies depending on our emotions. There were 114 Signal Passed At Danger incidents due to train operator error in the last reported quarter (the lowest for 5 years – Although on the Tube these represent a relatively low safety risk due to other safety systems in place, ATO could potentially eliminate these making the system safer and eliminating the 8.5 min of delays that each one on average creates.

  22. PPPPPeter says:

    @pedantic of Purley, your post starts by saying Crow could find no safety issues for customers with trains leaving a depot in ATO on their own, then goes on to say it means the driver won’t know if it has been prepped. You’ve mentioned one of several safety issues.

    Interestingly thought it was once suggested for trains coming out of Stratford Market Depot into Stratford, which would allow LU to cut back on a few drivers had it hapoened.

  23. Pedantic says:

    I suppose Bob Crow could try to argue that way. What I was trying to imply was that LU would have to adapt the procedure. Nowadays this could possibly be some electronic form of self-checking. They got around the problem of needing the guard to do the brake test so I don’t think this would be an insurmountable problem. Indeed an obvious one in your example is to consider enabling the driver to check whatever is necessary in the platform prior to carrying passengers.

    I din’t know of this suggestion but personally I think it is a pity they didn’t implement this at Stratford. No so much to cut out a few drivers but to get some real world experience of what running truly automatic trains entails and whether it really is workable. May be even have some half-height platform edge doors at Stratford to get some experience of installing and operating them.

  24. Fandroid says:

    Sorry to continue to be boring, but the first part of Nuremburg U-Bahn line 2 was opened in 1984. It has been extended four times up to 1999. It was ten years later that the first driverless trains ran on the line. At that stage it was four trains running under remote control, with another four running in the same service under manual control. It was due to go fully remote in January 2010. I travelled on it three times in two days in March this year. I didn’t immediately notice the lack of on-train staff, but realised that I hadn’t spotted anyone in a cab when a local resident told me she was afraid to use it because of the lack of drivers!

    It’s a modern system by LUL standards, but even so, it was not originally built for remote operation, and has proved that conversion is possible and that platform-edge doors are not a pre-requisite. Even mixed remote/manual running has been proven to be feasible. The London system is several orders of magnitude busier than Nuremberg’s but there is no reason why LUL should not give remote operation a really serious study. After all, the Paris Metro is almost as old and complicated a system as London’s.

    ASLEF Shrugged and Pedantic have both made sensible suggestions. That is, why not try a pilot on the Waterloo and City (simple isolated system with no intermediate stations and straight platforms) and also why not try it for depot moves?

    Trains move fairly slowly into terminal platforms, so the safety value of platform edge doors there must be limited.

  25. Fandroid says:

    Apologies for repeating myself, but Paris are not just doing it. They have done it! Todays news is that Ligne 1 is operational.

    See the link

  26. Anonymous says:

    Some of the Brussels metro is going fully automatic by the middle of the decade as well.

    I can’t see the whole tube-network going that way – Paris and Barcelona will only have a minority of their networks fully automatic. However, a mixed system is possible and, imo, desirable to free up cash for either keeping fares stable or future upgrades.

  27. Greg Tingey says:

    You have a “fullt-automatic” and truly driverless train.
    It’s actually underground, not on the surface.
    You’ve got any of:
    1] a jumper
    2] Someone taken seriously ill
    3] A hold-up or a fight

    How do you stop the train / wedge the doors / call for assistance / temorarily stop the system ???
    And how long does it take to re-start?

  28. Anonymous says:

    @ Greg

    1] a jumper

    Platform Edge Doors should prevent this. Sensor technology can look out for trespassers and line obstructions in front of the train and probably react faster than a human could.

    2] Someone taken seriously ill

    Someone pulls the emergency handle and communicates with control room who alert emergency services and staff at next station and bring the train to a halt at the next station. In theory this cuts out the middle man so could be faster.

    3] A hold-up or a fight

    Same as for 2 for a fight

  29. Fandroid says:

    The answers to all of Greg’s questions are just a matter of adapting the way the system is organised.

    One of the stated advantages of remote operation is that trains can be introduced into service to cope with additional passenger loads, without having to have both a spare train and a spare operator available at the same place and same time. It’s possible to imagine that a train might be run with the doors closed to pick up an operator at the station where he/she is available. (Does the DLR shift it’s trains around with no-one on them?)

    For Nuremberg they claim that all operational staff are trained to manually drive trains. So presumably there is now no distinction between station staff and train operators, so providing a fully flexible response to any abnormal situation.

    As for jumpers, I think that anything is good that reduces train operators’ direct exposure to the actual moment of human/train collision. There always will be a need to react to the aftermath, but that should be easier if the staff member did not actually witness the incident.

    For fights/illnesses I’d prefer a Train Captain there inside the train to an Operator up front who has other duties to attend to.

    It’s not really possible to imagine that London will ever go for staffless trains. Talk of Train Captains not being on every train out in the suburbs is just plain daft. It’s out there, on lightly loaded trains at night that passengers want/need the comfort of an official presence.

  30. Pedantic says:

    Interestingly, BBC London reports that LU have told the unions that it expects to begin trials with “cabless” and “driverless” trains on the Waterloo and City in about four years time.

    The clear implication is that this will be the pre-production EVO trains planned for the Bakerloo and Piccadilly lines. Getting the trains down there could be “fun” especially if the adjoining carriages have shared bogies.

  31. Mwmbwls says:

    New stock seems a taj over the top – retrofitting old stock with partially sealed off cab areas and full width corridor connections might be easier unless the articulated protoypes are to be made available in the next two years.

    Not so mich a great leap for mankind in moon boots but a healthy stroll in sensible brogues. A eminently suitable place for a pilot study. Lots of passengers to provide a realistic test of the system integration yet segregated from the rest of the system. Enough potential reputational damage to ensure that everybody keeps their eyes on the prizes ( glittering and booby).What’s not to like?

  32. timbeau says:

    Nice straight platforms, and only four of them to equip with PEDs. (At a pinch, two, since only one platform at a time is used at Bank, and the arrivals side at Waterloo never has anyone waiting on it, for obvious reasons.
    And because of stepping back on such a short line, the drivers per train ratio is very high on that line.

    Even if they do the experiments with converted 1992 stock, rather than EVO, the two unused cabs would make for a significant amount of extra space for passengers.

    And if the whole thing does go belly up, (not an uncommon occurence anyway, as regular users of the W&C will be aware) there are alternative routes available.

  33. Pedantic says:

    To which one can add (if pre-production EVO trains are used):

    a) an opportunity to move the existing trains on to central line depots where, at the very least, they can be cannibalised for spares

    b) an opportunity to increase the stock from the current five to six trains and thus not rely on 100% of the trains being in service to run the timetable. This assumes that the additional required siding space exists or can be created.

    and regardless of the type of stock

    c) an opportunity to extend hours of operation since the marginal cost of finishing later or even running on Sunday would be substantially reduced.

    The drivers per train ratio might be high but the driver time per train time in customer service (as opposed to waiting at Bank or being out of service at the Waterloo end) ratio is even higher than on other lines. Furthermore the drivers are actually provided by a central line depot (Loughton ?) so the non-productive time is even higher. As the drivers are also rostered for central line duties one suspects that the loss of posts could be easily absorbed into the central line without hardship.

    In fact the more I think about it the more this seems to be brilliant idea.

  34. Islander says:

    To address this, sources suggest the Paper advocates a phased plan aimed at rebalancing and refocusing the Train Operator role. The end result (as with the suggested changes to station staffing) would be Operators whose role included both operational and increased on-train customer-focused elements.

    I have picked out this comment and would like to know what the train captains role is with on board customers on the DLR, especially when trains are packed. Does he remain at the front of the train as it enters stations and so can stop the train in an emergency? What I cannot see is a train captain on a future EVO Bakerloo line train at say Oxford Circus at 18.00 being particularity able to carry out customer focused elements very well, would he have to fight his way back to the front of the train at every station, if so they may as well carry on driving them as now.

  35. Jeremy says:

    Are PEDs likely to be a requirement for remote operation? Are they certain to be a requirement?

    Just asking, as this seems like one of the biggest challenges to me.

    I don’t want to see unstaffed tube trains – I think a well-staffed system is important in avoiding problems with vandalism and the fear of crime as well as actual crime against the person. But I see no reason why that person ought necessarily to be ‘driving’ the train.

    Automating the W&C would be a great place to start, for all of the reasons mentioned above. But I do think that if PEDs are to be fitted, you’d need to do all four.

    Q: If the trains are operated remotely, and sealed behind PEDs, would one of the platforms at Bank not make for suitable stabling for an extra set, assuming that it was then the first to be brought into service when things were busy?

  36. Gallions says:

    @ Jeremy
    The DLR is automated and they don’t have PED anywhere on the system, above or underground. Works fine there.

    From a safety/vandalism point of view, having a train captain moving around inside the train is far better than a driver in an isolated cab. I’ve always felt safe on the DLR because of that.

    Automation, provided there is a member of staff aboard to deal with problems and take over the train in the event of a problem (as with the DLR) seems a very sound proposition and its a shame Bob Crow has to be so unconstructive when it comes to tube modernisation.

  37. SimonT says:

    During rush hour(s) train captains on the DLR generally keep an eye on the doors & put them in a semi-automatic state (open button available for use automatically, but closed manually via control key/panel). They also tend to pass through the train by moving along the platform at a station & getting back on at the next set of doors!

    It’s rare to see one being driven manually…

  38. Anonymous says:

    Just to pick up on a point from the other post on the future of the LU, could someone explain why Oyster is considered to have a limited shelf-life? Is its existence getting in the way of further efficiency gains (e.g. as mentioned in the post, many transactions at ticket windows are to buy Oyster cards) or is there something limiting about the technology behind it?

    Also, I take it there is no plan to do away with paper tickets? I imagine that a lot of tourists from non-Western countries won’t have ATM cards with wave and pay.

  39. John Bull says:

    @Anonymous – both really. It’s peaked in terms of efficiencies it can gain, and it’s also slightly constrained by the fact that – like the Underground – it was an early smartcard system that has evolved into its current state, rather than one that has been designed to be like this all along.

    If you haven’t read it already, The Problem with Simples: Why Oyster is a Victim of its own Success was a piece we did a while back on this. Bit out of date now, but might still be of interest.

  40. ASLEF shrugged says:

    timbeau – strangely the W&C trains don’t even have ATO or ATP, we run in manual with old school coloured-light signals down there. The cabs are full of kit, it’s not just a box with a seat and a TBC, all that stuff has to go somewhere, the space you’d gain from losing the two empty cabs would be minimal and converting them could turn out more expensive than bringing in new trains.

    Pedantic – Leytonstone depot cover the W&C, I always give away my “700” duties but I still have go down there once every six months to keep my licence up to date. ‘Orrible place, they can make it NoPO tomorrow!

    Simon T – if a train captain got off a Central Line train at 8:30 on a Monday morning they’d never be able to get back on again.

  41. Greg Tingey says:

    Ah yes, the joys of reliable automation, without train drivers, or captains or crew …
    SE today’s “Standard”
    Umm, err …….

  42. maclondonuk says:

    A number of interesting comments made regarding the rationale for doing this (or not doing it!) and reasons why it has not been done in the past. ASLEF shrugged in particular is right to suggest that the need to keep skills/knowledge fresh has prevented small experiments in the past like operating fully automatic to/from depots.

    However, I am an optimist that a fully automatic/driverless operation will start to happen before 2020 – the technology does exist and the changes to Paris Line 1 which is very similar to what would be required to transform the deep tube lines will show the way (if successful in a city with very similar historic/union/infrastructure problems to London). Furthermore, the piccadilly, bakerloo, central and W&C lines will all need new signalling and trains over the next decade or so which could be delivered more cost effectively by adopting a fully automatic operation. A number of issues will need to be addressed of course including whether PEDs are required, the role and skill level of the on-train attendent, interface with Network Rail infrastructure, etc.

    I do find the glass is half empty (and leaking) pessimism of a number of commentators thoroughly depressing – I suggest we would still be debating whether travelling at faster than 30mph would cause brain damage, miscarriages and other lovely things if we followed their views!

  43. Anonymous says:

    Next Installment…

  44. Greg Tingey says:

    Some of us have seen all the previous cock-ups.
    And don’t trust (especially LUL) British so-called “management” to correctly and safely implement these things.
    It’s NOT the technology (I am myself an engineer) it’s the people supposedly in charge, who can’t be trusted!

  45. John Bull says:

    Aye, sorry – I got unexpectedly stuck abroad without Internet access (shock! horror!).

    Final installment (and more posts) up shortly.

  46. Ben says:

    Huzzah! Not reading LR is like waking up to find youve gone deaf in one ear without any good reason 🙂

Leave a Comment

In order to make LR a pleasant place for discussion, please try to keep comments polite and, importantly, on topic! Comments that we feel do not meet these criteria, or that contain language that could cause some people trouble at work, may be moderated or deleted.

acceptable tags

* (This won't be shown, but you can link it to an avatar if you like)

Recent Articles

Friday Reading List – 17 March


As anyone looking to properly understand London’s transport needs and network knows, context, background and best-practice are important. As readers might imagine, behind the scenes here at LR Towers we thus spend a lot of time sharing links and reading

Read more ›

LR Magazine Issue Five: Overgrounded


With print copies now being prepped for dispatch to subscribers at LR Towers, London Reconnections Magazine Issue 5: Overgrounded is now available to purchase in our online store. Transport is politics, politics is transport You don’t get transport without politics.

Read more ›