We’re approaching the Mayoral election season (already?), and the perennial issue of driverless trains on the London Underground has reared its head once more.

Politicians of most colours and experiences will talk about their plans to improve London’s transport, and some, with more of an eye for a sound bite than real understanding of the issues, will talk about driverless trains.

After all, the DLR is driverless, so it can’t be that difficult. Can it?

Firstly, lets set one thing straight — it’s pretty much a certainty that at some point, the tube will be driverless.

After all, driverless trains have been talked about since the 1930s, and although the tube has become ever more automated as the decades pass, a bit like fusion power, driverless trains have always been “just 20 years away”.

However, no grandstanding by a politician will make it happen. Even with all the pieces in place, it’s a decades long project, and the most optimistic projections will see another six mayoral elections pass before the first driverless tube train carries passengers.

Also, as with the DLR, driverless doesn’t mean unstaffed. 

So if your political goal is to ban tube strikes, then going driverless wont prevent that happening, as anyone who has experienced a strike on the driverless DLR will confirm.

The DLR was built to be driverless, but very nearly wasn’t, and could be driverless thanks to being essentially a brand new railway, with small trains that travelled comparatively slowly. It was also able to push through the driverless aspect as a technology show-off for British industry.

The DLR has been around for 30 years, so why isn’t the tube driverless yet?

To put one issue to bed, yes, the unions would kick up an almighty fuss. A test of a Jubilee line train using Willesden Green sidings under automatic control in October 2016 didn’t go down well with the unions.

But, London Underground’s planning runs into the decades in some places, and the unions concerns about loss of drivers wouldn’t prevent the sort of very long term thinking that would be needed if driverless trains were to be introduced.

And London Underground is moving in that direction already. It doesn’t matter what Mayoral hopefuls say, TfL is already working on it.

The technical specifications for the New Tube for London, which will start to arrive on the Piccadilly line from 2023, includes an option for driverless operations — in the future.

Whoever is elected Mayor in 2020 will see the introduction of those new trains, and there will be a driver in every single one of them. It’s simply not possible for an incoming Mayor to change that.

So when someone says they’ll introduce driverless trains, they mean they’ll do no such thing whatsoever.

They might be able to nudge TfL a bit in that direction, but that’s the best that can happen.

As the new tube trains for the Piccadilly (and other lines) will have the option to go driverless, what is holding them back?

The rest of the tube infrastructure is the main issue, not the trains. The signalling will be upgraded for the new trains, and to replace worn out systems, but it would need an almighty overhaul to bring it even close to the requirements for driverless trains to work.

Back in 2014, it was suggested that if works began in 2016 to add platform edge doors to the Piccadilly line platforms, then along with everything else, it might, just be possible to have driverless trains on that one line in 2027.

As you can guess, we’re already a long way off that target, and if a new Mayor were to demand it, it’ll be their successor who cuts the ribbon, in the mid to late 2030s, at the earliest.

All this costs a vast amount of money, and unless Mayoral hopefuls are sitting on a mystery pot of cash, that’s not going to happen either.

But would it be worth spending that money?

It depends on your aim.

If it’s simply to banish tube strikes, then it’s a total waste of money. The DLR still has strikes after all, and there’s no way that the tube trains would run without onboard attendants, who can go on strike.

To spend billions just to prevent the occasional, if very annoying, tube strike would be the height of folly.

The main reason for doing it though is to increase the number of passengers the tube can carry.

Computers are, generally, more reliable than humans — even in as hostile an environment as a tube tunnel with all that grease and oil and moving parts just waiting to confuse a binary thinker — and one huge advantage of driverless is to run existing timetables more reliably, not to mention a nice feature of being able to ramp up capacity at short notice when needed

A computer system should be able to run trains closer together safely and that means more trains per hour, so they can carry more passengers.

With London’s population surging — that’s not immigration by the way, just London’s young doing what young people do — increasing the transport network capacity is vital.

Hence the imminent arrival of the Elizabeth line and in the 2030s probably Crossrail 2.

So driverless trains are possible, if very expensive, and will probably arrive, in a few decades time, but only because they’re the solution to a capacity problem, not because a politician made an expensive promise several decades earlier.

However, it’s not the trains or the tunnels or the computers that would be the biggest cause of a delay in going driverless.

The stations are.

If you have the trains running closer and dropping off more passengers onto narrow platforms ever more often, you reach a point where the station itself can’t get them off the platform fast enough before the next train arrives.

Even with drivers onboard, by the end of the 2020s an additional 21,000 customers will be able to board Piccadilly line trains every hour during peak times.

If the aim of driverless trains is to increase the number of passengers that can be carried in the rush hour, then the stations need upgrading to cope with the extra people arriving.

Even with platform edge doors, crowds remaining on the platform waiting for space to leave is unwise.

Pretty much every rationale for tube station upgrades over the past decades has been based on coping with overcrowding. The works at Victoria, Kings Cross, TCR, ongoing at Bank, soon at Holborn and Camden are all about having more tunnels to absorb the crowds so they don’t wait on platforms, and more escalators to get them out of those new tunnels.

The recent trial of standing on the escalators at Holborn didn’t get any individual person out quicker, it increased the carrying capacity of the escalators so they could soak up more people waiting at the bottom to leave. That marginally increased the station’s ability to deal with more trains.

On the Piccadilly line, you could need to look at substantial upgrade works at Hammermsith, Earl’s Court, South Kensington, Green Park, Piccadilly Circus, Kings Cross, Arsenal and Finsbury Park. Minor works, mostly to do with ticket hall expansion and additional staircases at the regional stations can also be expected.

There’s even very long term thinking about how to do away with ticket barriers, so that they can remove that bottleneck in getting people out of stations faster.

If planning for that started today, and the billions it will cost was available, it’s still at least a couple of decades of work to carry out.

The trains may go driverless eventually, but it’ll be the stations that determine when.

So if someone says they’ll push for driverless trains when elected Mayor of London… shake your head in despair and move further down the carriage.


Be the first to know what's on in London, and the latest news published on ianVisits.

You can unsubscribe at any time from my weekly emails.

Tagged with:

This website has been running now for over a decade, and while advertising revenue contributes to funding the website, it doesn't cover the costs. That is why I have set up a facility with DonorBox where you can contribute to the costs of the website and time invested in writing and research for the news articles.

It's very similar to the way The Guardian and many smaller websites are now seeking to generate an income in the face of rising costs and declining advertising.

Whether it's a one-off donation or a regular giver, every additional support goes a long way to covering the running costs of this website, and keeping you regularly topped up doses of Londony news and facts.

If you like what you read on here, then please support the website here.

Thank you

  1. Curious says:

    I have two (sets of) questions:

    1) How much of a cost saving (in terms of day-to-day labour costs) would it be if Tube trains could run driverless? Are the costs of paying drivers negligible compared to the other costs involved in running a service (fuel, maintenance)? How would the cost of running a train with a DLR-style operator differ from running it fully unstaffed?

    2) How remote is the possibility of Tube trains becoming unstaffed altogether in the distant future, like on the Singapore MRT? What are the considerations there?

    • Ian Visits says:

      1) Cost savings are negligible in the grand scheme of things – in 2015–16, TfL had a budget of £11.5 billion and by 2020, tube driver wages are projected to be around £250 million. Switching to train attendants might shave £50 million off that bill, so the impact is insignificant.

      2) No idea.

    • Tube guy says:

      2) The main problem with a completely unstaffed Tube is the narrowness of the tunnels. The MRT has 5.8m diameter tunnels (vs 3.6m) which allows space for safe evacuation of passengers. In the event of a fire on the underground passengers would need to evacuate single file from either end of the train and make their way to the nearest station on foot. This could potentially be in the dark (although tunnels have emergency lights there is no guarantee that these would work or be sufficient in the event of an explosion or fire due to smoke or damage) and where several lines intersect there would the threat of other trains colliding with evacuees. There is also the rails, maintenance equipment, points and other track equipment that can be dangerous and may not be instantly recognizable as dangerous to untrained people.

      In short, there is always going to be need for someone on the train with knowledge of the route for evacuation, sufficient safety training and knowledge of the train itself (to cut out equipment on fire or to override the automated systems and move the train). Unless of course we widen the tunnels to include evacuation space, but given that many of the tube lines intersect the foundations of historic buildings already, you’d be looking at £1trillion+ for that sort of project if it’s even possible.

  2. Hacorus says:

    “there’s no way that the tube trains would run without onboard attendants”.
    It is true that removing drivers just to avoid strikes would be utterly ridiculous, however, the presence of a train attendant may not be mandatory depending on the signalling.

    The DLR doesn’t have any door preventing track access and platform control. And the attendant is there to assist with door closing and in case this 30years old system fails.
    There are plenty of cities with highly reliable driverless signalling system. So the need for a back up attendant would be goldplating. Why spend billions on an unreliable system? Of course they wouldn’t (although knowing tfl…).
    And if platform doors are installed then there would be absolutely no need for a mandatory train attendant.

    The main reason to want to be driver free for a train is for service flexibility. A problem on the victoria line? Ramp up the piccadilly. Night tube? No problem. Ramping up the service or winding it down will be a lot easier and that’s where value is.

    A member of staff however adds additional value to customer journey, either onboard or on the platform, assisting passenger, providing customer oriented services, potentially new revenue generating services (such as dry cleaning at the train station as in the Netherlands).

    In short, unattended driverless trains are a possibility for London, it would take time and cost a lot of money but bring a lot of benefits by freeing staff from the cabine, by giving more flexibility to the service, by improving the passenger experience.
    If Paris can do it, so can London !

  3. Paul says:

    I assume one advantage of driverless would be quicker turnaround time at terminals, no need for the driver to change ends ( or take loo breaks!). With other efficiencies, I guess this could mean that the same service level could be achieved with less actual trains?

    • Melvyn says:

      That is already dealt with on some lines by using ” stepping back” where basically a driver brings a train in and is then met by another driver who brought an earlier train in and will take this train out , while it’s driver walks to opposite end to take another train out .

  4. Noah Bowie says:

    I think you’re being too negative I think an upcoming mayor will get started on station upgrades but it will take a long time. When they say they’ll push for driverless trains I think that means they would set the ball rolling and start upgrading stations because that has to be done anyway.

    • Jason Shinn says:

      How are they going to pay for those upgrades? You say that as if money is no object but TfL is already having to cut back on some projects because of their limited budget. Even upgrading a single station is a major undertaking and we’re talking about substantial parts of the network here.

  5. Melvyn says:

    In a TV series about The Tube when Tim O’Toole was in charge he mentioned how you could tun more and more trains but the problem was Stations where passengers have to trudge up stairs to clear the platforms thus making hard to clear a platform before next train arrived with more passengers!

    Solving this problem at some stations will be difficult given how short many of these flights of stairs are and thus not suitable for escalators. One option would be lifts but very expensive unless part of full step free access scheme although in some stations if lift is to escalators that reach Street level then at least stair free access would be created.

    As for driverless trains well the Waterloo and City Line is a prime candidate being a simple shuttle service operating not to deep below the ground.

    It’s doubtful that the sub service network will ever go driverless both because of their complexity and interaction with other lines some main lines some Tube .

    Simplifying tube lines might be first move with Piccadilly Line becoming Cockfosters to Heathrow with Rayners Lane branch switched to the District Line with something similar for Central Line in west London.

    A major problem is many stations don’t have straight platforms creating gaps and bends just think of Central Line at Bank Station to see how fitting PEDs would be impossible without major Station works to straighten platforms.

    Perhaps it’s time to return some outer sections of the underground to the Overground?

  6. Kris Jones says:

    I think one needs to make a distinction between “driverless trains” and “unstaffed trains”. Victoria Line trains don’t have drivers in the proper sense, but they are staffed. I find it difficult to imagine a future where the public would accept unstaffed trains on the Underground. Let’s remember there are already many people who fear using the system. I think they will always need to be reassured there is someone on board who can respond to passengers in times of emergency and have direct communication with the control room. Perhaps we’ll have train captains rather than drivers.

    • Ian Visits says:

      All of which is why I am using the “staffed” example of the DLR in my article.

  7. Edwin says:

    “A major problem is many stations don’t have straight platforms creating gaps and bends just think of Central Line at Bank Station to see how fitting PEDs would be impossible without major Station works to straighten platforms.”
    Paris seems to have cracked that one too – take a look at Bastille on line 1.

  8. Dave says:

    The best place to introduce unstaffed trains would be the Waterloo and city line, with a walk way put in between the running rails and emergency auto ramps (From Trains)to the floor its completely possible especial with extra walk way cross overs from north to east.
    Other then that I would say reopening the Aldwych branch would be a good pilot on the Piccadilly line.

  9. ap says:

    Anyone wanting to understand how hard it will be to introduce driverless trains on the existing lines should look at the work being done in Paris to make Metro Ligne 4 driverless.

    Paris has the advantage over London of mainly straight platforms but despite that key stations get closed for up to 3 months. This year the platforms at Gare du Nord got sorted. Last year it was Montparnasse. The platform doors won’t be going up for a while yet. It will be 2022 before the line is fully automatic.


  10. Chris Rogers says:

    Leaving aside the references to ‘drivers’ (tube trains aren’t driven in any conventional sense of the word – they’re ‘operated’, like a lathe or a lift), it will come but not for decades, as you say.

    Really interested in your point about the Holborn trial – I don’t recall that reason being mentioned at all at the time, though it makes sense.

    Not that many will ever use it unless going right across London, but Crossrail will add many minutes to any journey because the platforms are very very long and very very deep. And the west entrance to Tottenham Ct Rd is actually closer to the east entrance of Bond St than to its own eastern brother!

  11. Andrew Gwilt says:

    Ive heard that the Central Line will not be getting new tube trains for at least 12 years which to be fair is a joke. Why wait 12 years. Why?

    • Andrew Gwilt says:

      People who use the Central Line will still be suffering from the effects of the heat whilst using it underground. But Elizabeth Line will help with commuters travelling from East & West into London without having to use the Central Line when the weather turns hot.

    • Ian Visits says:

      Because there’s a limit as to how fast you can build trains and pay for them and upgrade signalling systems for them.

    • Andrew Gwilt says:

      Right. Now I understand.

  12. Firstwave59 says:

    It is worth remembering that the Victoria line – dubbed ‘the robot line’ when it opened – has been running One Person Operation on an automatic train operation (ATO) system based on block signalling, since the late 1960s. Indeed, the driver’s role was to merely check the doors were safely closed and press the start button when a green light under a Automatic Train Operation system. However, I knew one of the original team of VL ‘regulators’ – the people who managed the trains and services from a remote control centre. What he said to me in 1968 was: “If we really wanted to we could actually run this line without drivers at all, but passengers would be scared if we took the driver away.” The DLR of course has since taken that next step.

  13. Jim Hall says:

    I would suggest that you expand your system to driverless on the grounds of safety and speedier operation. Computors dont have SPAD’s or get fatigued and can operate a more intensive service. Yes you will require an attendent who will not require to be trained to the same standard as a driver or the salary to match. The unions will of course disagree as always to anything that advances technology or reduces the expertise of their members. Nearly all new rail operations and new trains being built around the world are now featuring driverless systems. After all you cannot compare driving a steam train to todays trains, just different skill sets. Computor tech is the future and safer

  14. Jason says:

    Beggars belief that Elon Musk and many others besides are building systems to allow cars and lorries to be self-driven across huge distances and yet we are told here that we cannot build a driverless tube ????

    • ianvisits says:

      It is possible to build a driverless tube in an existing network, just very very expensive. Car owners don’t need to worry about traffic light maintenance, road maintenance, safety walkways, killing more than a dozen people every time it goes wrong.

      Driverless trains need to widen their tunnels to allow for evacuation — costing billions, they need to cover the cost of the signalling upgrades and track maintenance, and when it goes wrong, hundreds of people die.

  15. David Dorrington says:

    I get quite annoyed when people say why don’t we have driverless trains. Automatic trains are not new in this country. In 1969 they opened a brand new tube line in London. Believe it or not it was called The Victoria Line. It was run under automatic train control from the start. The first railway in the world to do so. All the operator in the cab had to do was open and close the doors at stations. Then press two start buttons and the train would drive itself to the next station. We were very proud of these automatic trains. Since then the Central, Northern and Jubilee Lines have gone to A.T.O. They are now slowly converting the District, Circle, Hammersmith & City and Metropolitan Lines To A.T.O. The only reason they put the operator up front at the beginning of automatic operation was because they thought passengers may be nervous about no one being in the cab. Ironically Thameslink is now running A.T.O. Through the central section. Another British first.

Home >> News >> Transport News