Boris demands TfL agree to Driverless Trains

PEDs were only installed on the "tunnel section" stations on the Jubilee Line Extension, not on the three "open section" stations - Canning Town, West Ham and Stratford.  When I was training to work on the JLE (Feb/Mar 1999) we were told that the primary reason for installing PEDS was to improve ventilation by reducing the "tunnel wind" effect (obviously not a problem in open section) and to reduce the amount of combustible materials on the track.  Reducing passenger "intrusions" on the track was a secondary benefit.

Why would "they" say that?  PEDs were installed on eight of the 11 JLE stations, it can't have been that much more expensive or difficult to have installed them on the three "open section" stations.  "One unders" are relatively rare and no more likely at JLE stations than elsewhere, why bother to fit PEDs at all when there's a driver in the cab to activate the emergency brake?

West Ham and Canning Town JLE are island platforms with the DLR (ex-Silverlink to North Woolwich) on the other side of the westbound track.  If they'd wanted better track access on the JLE tunnel section stations they could have built it.

Splitting units is part of driver training, we get regular refresher courses as part of CDP/ATOR/5-day block where we split a train in the depot and it doesn't take that long.  The obvious problem with splitting a train during a "one under" is traction current will be discharged and you can't move trains without "juice".

So in light of the moderation and to bring this vaguely back on track... is there any viable scheme that would deliver automation up to the level of, let alone beyond, the DLR on any underground line before 2050?

W&C should be possible to do something with but may no longer be worth the faff... Most should surely be able to achieve what exists on the DLR with some effort; maybe maintaining the very high frequencies would become more problematic but that seems more an issue of how long it takes people to get on and off a train rather than running a train now.

If we compare the massive advances in automated driving for cars, the railway is a much more controlled environment than the roads are so surely this can also feed in? However having an operator-free train seems a very unlikely proposition to me.

Really, this thread is just going round and round in circles, now.

I don't think anyone here imagines that anything will happen during the current or even the next two terms of the MoL and PM.

If you consider a realistic time frame, you'd probably be looking at something like this at the beginning.


1. Agreement to automate W&C as initial 'proving ground'. [ 1Y ]

2. Draw up detailed list of requirements for changes. [ 1Y ]

3. Pass enabling legislation and garner agreements from unions. [ 1Y ]

4. Generate detailed plans for modification or replacement of signalling, trains, and infrastructure. [ 2Y ]

5. Invite tenders for contracts for above and await submissions [ 1Y ]

6. Select contractors, negotiate detailed contracts and sign. [ 1Y ]

7. Contractors do work [ 2Y ] for most of which time the line would be closed.

8. Thorough initial and safety testing [ 6M ]

Those are generally very optimistic estimates and they take us to 2030, and it's unthinkable that there would not be significant over-runs.

At the end of this period you would have one point to point line with two stations. You would then need to go through the whole process again for the next candidate line - probably the Victoria, which would take considerably longer because it's a much bigger job and the technology would have moved on to such an extent that you would probably not gain too much from re-using existing work.


The bottom line for all of this is: Yes, go ahead and agree to potentially having fully automated lines, but don't expect it to happen over the whole system for several decades.

Glasgow Subway plans full automation with "half height" platform screen doors. But....I would be very surprised if we see fully automatic trains on the  anytime soon though I' guessthere may be a few new trains in service next year operating under the existing signalling/ATO system.

The 2014 Kings Cross exhibition for the New Train for London (NTfL) revealed that these trains could be fully automatic without a driver's cab, although Mayor Boris pledged that a member of staff would remain aboard. The first fleet arrives on the Picc 2024-26, but with resignalling not expected before 2030 they will be driven as the current trains. With a staff member on board, TfL have already shown that costly precautions for driverless trains makes them uneconomic. So accepting a staff member in deep level single bore tunnels, why not remain as current auto lines do with someone at the front of the train. But out in the open where access is easily available, could an unstaffed train demonstration project operate east of Arnos Grove for the 30 minute run to Cockfosters and back? That would be a proving ground for west of Hammersmith, although the local tunnels west of Hatton Cross and interworking west of South Harrow would require local teams of drivers to join trains there. This would mean around 40% of each trip would be unaccompanied, so providing staff savings to compensate for additional equipment costs on outdoor sections. Would this be sufficient to make such conversion worthwhile? New trains planned for the Jubilee, Bakerloo and Central lines would run on current signalling, so probably uneconomic to modify, with no further lines in the coming decade.