Could we have driverless trains on the national network?

[Ken H]

On the uncoupling thread, someone mentioned that with increasing automation in the cab, we could dispense with the driver.
https://www.railforums.co.uk/thread...assengers-on-board.178987/page-4#post-3967994

Do we think this is likely? We will have driverless cars soon, why not trains?

and with the increasing automation, is it a job anyone will want, sitting all day doing nothing, just waiting for the very rare event that needs their intervention? And will the bored driver then read or text or whatver?

When the CL91 came out, i remember a piece in Modern railways describing how the trains wee driven. The driver selected a speed, and selected full power. The train automatically accelerated to that speed. As the allowed speed changed, the driver just twiddled the speed selector and the train did the rest.

Actually, what do airline pilots do once take off has been done, and the auto-pilot takes over?

 

[Bletchleyite]

I'm certain we will, but it's a way off yet. I doubt any drivers need to worry about their jobs for now.

I think the result will be a resurgence of the guard - guard-only operation, with the guard trained to drive manually at low speed in emergency. Same concept as the DLR, basically.

 

[Jorge Da Silva]

I'm certain we will, but it's a way off yet. I doubt any drivers need to worry about their jobs for now.

I think the result will be a resurgence of the guard - guard-only operation, with the guard trained to drive manually at low speed in emergency. Same concept as the DLR, basically.

Isn’t Thameslink between St Pancras and Blackfriars Driverless? (Although the driver is present they do not drive the train).

 

[ForTheLoveOf]

Driverless cars have been there "in 5 years" for the last 20 years (well, not quite, but the principle stands). The technology still has a long way to go before it is good enough that ordinary people trust it, let alone that regulators allow it to be used.

Driverless trains are going to face a similar sort of issue from passengers, even if they are proven to be safer than trains with drivers (although zero fatalities vs zero fatalities doesn't prove a great deal...). And that's completely ignoring the inevitably backlash from drivers.

 

[duffield]

Isn’t Thameslink between St Pancras and Blackfriars Driverless? (Although the driver is present they do not drive the train).

I believe this section is driven automatically, yes, but it's obviously a section of line which is very 'contained' and segregated from the outside world.

I think this sort of part-automation of sections of line where trains run very freqently at close intervals are probably most likely to happen in the near future. At the other extreme bits like single-track branches with physical tokens and open crossings etc. are probably the most diffcult to automate due to it being diffcult to justify the infrastucture expenditure required.

 

[swt_passenger]

Can’t be that long since the last “driverless trains” epic thread, I wish people would think of doing a quick search, because you can say right now that nothing new will be posted.

 

[DanDaDriver]

https://www.railforums.co.uk/threads/driverless-trains.145433/

https://www.railforums.co.uk/threads/advantages-and-disadvantages-of-driverless-trains.156585/

https://www.railforums.co.uk/threads/automated-driverless-trains.116306/

https://www.railforums.co.uk/threads/driverless-trains-when.93480/

 

[swt_passenger]

Isn’t Thameslink between St Pancras and Blackfriars Driverless? (Although the driver is present they do not drive the train).

Not yet done in normal day to day operation. Trial running only. This has been pointed out numerous times in the main Thameslink and Class 700 discussion threads...

 

[jon0844]

Trains may well get more automated as time goes on. Technology, security, cost, policy and union approval permitting...

I can't see there won't still be a driver in the cab. They'll just do less. They may not need the same level of training as a result, and maybe - just maybe - the pay will fall because management will deem it easier and a LOT quicker to replace a driver.

Likewise, cars can get only so far for automation. That last level would need a totally controlled/safe environment, not roads with other vehicles, people, animals, obstructions and so on.

Cars that can steer, keep to lanes and avoid hitting the vehicle in front (even keeping to the same speed) exist today. There are other things that can be automated, but you'll always need (IMO) a driver to step in when something goes wrong, as it will.

Vehicles (and trains) break down all the time, so imagine how you fix an issue with no driver. That and so many other factors are why no train driver should fear for their job as I'm sure even new drivers will have retired before anything major happens. It would however be naive to assume it won't happen one day, or that perhaps you may get some new built services designed to operate like the DLR from day one.

 

[gazthomas]

Time to close the thread mods!

 

[trainmania100]

Signalling systems and signal failures aren't percect they're always breaking down , I feel the technology to bring driverless trains everywhere would be a huge cost and is a long way off

 

[hexagon789]

We can but we won't.

At least not for decades if ever. Next step is probably full in-cab signalling. Then perhaps semi-automation on some lines but still with a fully trained driver in the cab at all times.

 

[ComUtoR]

Cars that can steer, keep to lanes and avoid hitting the vehicle in front (even keeping to the same speed) exist today. There are other things that can be automated, but you'll always need (IMO) a driver to step in when something goes wrong, as it will.

Even TESLA, who seem to be leading the way with automation, state that the human up the pointy end still needs to have their hands on the wheel at all times and step in where needed.

Oh Jesus. Here we go again.. Let me get a bag of popcorn. I’m staying out of this one, I’m just going to watch the carnage instead.

Um yeah.... errrr....

TECHNOPHOBE !! LUDDITE !!!! STUCK IN THE 70s !!! Technology never works, Mk 1 eyeball, Anti-Driver, makes it cheaper/more expensive, tech not good enough yet, GSMR drops out, cars do it so trains can, closed network/open to the elements, signalling failures, crashes/suicides, but they have it in [random country], works on the tube, humans are better, job for life, Unions won't like it, save millions, hundreds of years away, it can be done really cheap using existing tech, software will save the world, less training, additional training, mass unemployment, safety, what if, [scenario a, b, c, d...] /thread

 

[Peter Kelford]

This is something that is already a formally stated goal of other countries:

 

[jon0844]

If France can get the unions to agree, God help the rest of the world!

 

[hexagon789]

This is something that is already a formally stated goal of other countries:

Is it actually a goal?

Also, will they get it past the unions?

 

[Peter Kelford]

Is it actually a goal?

Also, will they get it past the unions?

It's more than a goal. It's a concrete part of the development plans.

 

[hexagon789]

It's more than a goal. It's a concrete part of the development plans.

For a start I'm surprised at that fur mainline rail. For seconds I can see it falling by the wayside as with some other SNCF projects.

 

[deltic]

There are already a large number of driverless trains in operation worldwide and probably have far greater potential than driverless cars in being widely rolled out.

 

[Western Lord]

Actually, what do airline pilots do once take off has been done, and the auto-pilot takes over?

Rather significantly, they have to land the aircraft. There seems to be a belief among some people that aircraft can land themselves, they can't. Autoland systems rely on the aircraft communicating with the airports ILS system. 99% of landings are manual, due to the fact that most airports do not have landing aids of the required capability and in anything other than perfect, no wind, weather conditions it is virtually impossible for the aircraft to remain "connected" to the airport's ILS (so all of those "hairy" crosswind landings that you see on youtube etc. are manually flown).

 

[ForTheLoveOf]

Rather significantly, they have to land the aircraft. There seems to be a belief among some people that aircraft can land themselves, they can't. Autoland systems rely on the aircraft communicating with the airports ILS system. 99% of landings are manual, due to the fact that most airports do not have landing aids of the required capability and in anything other than perfect, no wind, weather conditions it is virtually impossible for the aircraft to remain "connected" to the airport's ILS (so all of those "hairy" crosswind landings that you see on youtube etc. are manually flown).

Landings will generally only be done with autoland in a very limited number of circumstances, which will vary depending on airline policy. There is no issue of aircraft remaining connected to ILS during poor weather, otherwise the very point of ILS would be defeated. The typical weather conditions for autoland are, however, a low amount of wind but with extremely limited forward visibility.

Pilots still have plenty to do in the air at the moment - it is just like Tube drivers on ATO lines. They have to monitor the aircraft, review all systems, and so on. In a way I think aircraft may be the last to become totally automated, seeing as there's no 'hard shoulder' to pull over on at 35,000ft if things go wrong!

 

[Ken H]

There are already a large number of driverless trains in operation worldwide and probably have far greater potential than driverless cars in being widely rolled out.

driverless as in nobody at the front end or driverless in that the 'driver' sits and waits for an alert that needs intervention?

 

[driver_m]

Love how my post gets deleted for adding nothing to the thread, even though we’ve had a ton of these threads before, and yet it’s there in someone else’s quote. Why would that be?

 

[Bikeman78]

When the CL91 came out, i remember a piece in Modern railways describing how the trains wee driven. The driver selected a speed, and selected full power. The train automatically accelerated to that speed. As the allowed speed changed, the driver just twiddled the speed selector and the train did the rest.

Actually, what do airline pilots do once take off has been done, and the auto-pilot takes over?

On the Dutch railways the sprinter units from 1975 have a similar feature. There are several buttons, the driver pushes the button for the desired speed and the unit accelerates up to that speed.

 

[Bletchleyite]

Is it actually a goal?

Also, will they get it past the unions?

Do they need to? It'd be like Southern DOO - "right, DOO tomorrow" - they had everything in place for it and the drivers didn't block it so the guards lost. (Note: this is an example, please don't turn this into a DOO debate, it's been done to death already! )

A DLR style "train captain" (which is what I'd see as the future of on-train staffing once the driving was fully automated) is different enough from a classic guard that I reckon they could press the nuclear button and do it via redundancies and new recruitment.

 

[Peter Kelford]

The order in which automation will occur on a commercial or semi-commercial scale is probably:

1. Metro and Urban Light Rail - Today
2. Rail - Early 2020s
3. Extra-urban autonomous driving systems - Mid 2020s
4. Unmanned Space Navigation - Mid 2020s
5. Street-sharing guided transport (trams, guided buses) - Mid 2020s
6. Fixed-route transport (buses) - Later 2020s
7. 100% Autonomous Personal Road Vehicles - Late 2020s, early 2030s
8. Maritime Navigation - Mid 2030s
9. Manned space Navigation - Late 2030s, 2040s
10. Civil Aviation - 2040s, 2050s

Just to clear up a few definitions:

'Guided transport': transport that is somewhere between light rail (on dedicated railway track and separated from the road network) and road transport.
'Unmanned Space': Research, cargo etc.
Extra-urban autonomous driving systems: An upgrade to cars which enables the driver to sleep or keep their eye off the wheel on long distance journeys (e.g. Leave London at 10pm, car 'takes over' once on the M1, driver rises the following morning around 6am to discover that they are already in the vicinity of Inverness). This is autonomous driving without the crazy minicabs who will ram into any car they see.

 

[DanDaDriver]

The order in which automation will occur on a commercial or semi-commercial scale is probably:

1. Metro and Urban Light Rail - Today
2. Rail - Early 2020s

They need to get cracking then, I still sign s lot of semaphores...

 

[hexagon789]

A DLR style "train captain" (which is what I'd see as the future of on-train staffing once the driving was fully automated) is different enough from a classic guard that I reckon they could press the nuclear button and do it via redundancies and new recruitmen

Yes, I would agree that seems plausible in the future. However I think in-cab signalling is the likely next step in the immediate future here at least.

Ever since the French first developed TVM it was possible to use it to actually drive the train, not only does it not do this but so far SNCF still have drivers...

 

[bb21]

We've been through this many times so no need for another repetition.