Automated Trains Commence On London Undergrounds' Subsurface Railway

[david.hill64]

3 hours ago, Nearholmer said:

This explains the basics reasonably clearly https://en.m.wikipedia.org/wiki/Communications-based_train_control

 

Also, St Simon and David Hill, is there a terminological issue here, in that “in-cab signalling” seems to be used in some circumstances to describe systems up to and including ATO? I notice that Thameslink Core is referred to as “in-cab signalling”, for instance, when I think I might call it “ATO on fixed-block” (at least, that’s what I understand it to be).

Yes there is an issue with terminology.

Communications based train control (CBTC) refers to the use of a communications system - usually Wi-Fi in MRT systems and GSM or TETRA radio in mainline systems for the two way communication of signalling information between trains and wayside controllers.

 

CBTC systems may be either fixed block (eg Bangkok BTS) or moving block (eg Bangkok Purple Line).

 

CBTC systems incorporate Automatic Train Protection (ATP: the supervision system that keeps trains within the permitted speed envelope and within the movement authority) and can interface with Automatic Train Operation (ATO: the system that provides the driving function). ATP is safety critical - Safety Integrity Level 4 (SIL4) in the jargon but ATO is not. ATO usually requires a lot of software tuning with real trains and infrastructure to be fully effective and you would not wish to make changes to SIL4 software unless you have to, as the costs of certification of the changes is very high. ATO is an add-on to the signalling system, not an integral part. It is an electronic driver, getting instruction from the ATP and an internal database.

 

ETCS levels 2 and 3 are certainly in-cab signalling. It is arguable whether the legacy ATP systems on GW and Chiltern, or the original Victoria line count. I would argue that any system that requires wayside signals, even though supplementary information is available in the cab, are not in-cab systems. In-cab systems may provide lineside marker boards.

 

(Supplementary note for those interested: CBTC systems on metros often have a fall back system which may either be a simple Points Locked Indicator system - Bangkok BTS) or a secondary fixed block signal system - Bangkok Purple Line.)

 

Simon has replied as I have been drafting this and gives useful information.

[david.hill64]

23 minutes ago, St. Simon said:

 

Then, on top of all of this, you have Automatic Train Protection (ATP), again, this is sometimes gathered into the general 'in-cab signalling' term, but is totally separate. You can have 'in-cab signalling' without ATP, it just so happens that the systems that have 'in-cab signalling' have ATP integrated as well. I think you might be able to get ATO without an ATP system, but it is almost certainly a really bad idea! Similarly you might be able to have a moving block system without ATP, but it's not a good idea.

 

 Simon, I think the issue here is that around the world ATP is mandatory for new signalling systems. I would argue that it is integral to the CBTC system in all modern systems.  All CBTC systems are in-cab signalling. The actual Driver Machine Interface in the cab is not safety critical - SIL2 usually.

 

[Nearholmer]

Hmmmm ......... 

 

Do LU or other metros use the term "in-cab signalling" when talking about these systems? 

 

I'm pretty certain I never heard them described that way in 20+ years working alongside those who were specifying them, and overseeing their deployment, and until coming across NR use of the term "in cab signalling" for an ATO system had always understood that term to mean something that involved a human train operator taking instructions from a display in the cab, that is a system operating open-loop, usually with a very basic closed-loop system operating in the background to trap exceedances.

 

 

 

[david.hill64]

I have worked with metro systems on 5 continents and have never had a client refer to ‘in-cab signalling’. They all want CBTC. 

[ess1uk]

article here

https://www.railengineer.co.uk/2019/06/10/ato-exploiting-the-technology/

 

Edited July 10, 2019 by ess1uk

[CloggyDog]

On ‎26‎/‎06‎/‎2019 at 00:16, Nearholmer said:

Hmmmm ......... 

 

Do LU or other metros use the term "in-cab signalling" when talking about these systems? 

 

I'm pretty certain I never heard them described that way in 20+ years working alongside those who were specifying them, and overseeing their deployment, and until coming across NR use of the term "in cab signalling" for an ATO system had always understood that term to mean something that involved a human train operator taking instructions from a display in the cab, that is a system operating open-loop, usually with a very basic closed-loop system operating in the background to trap exceedances.

 

 

 

 

Never heard the term 'in-cab signalling' used in my 30+ years with LU.

 

ATO yes for the Vic and then the Central and CBTC for the Jub and Nothern and, more recently, the SSL.

[Ray H]

I don't doubt that the signalling side of things will be readily achievable. However, I can't say the same about service performance on SSL. I'm not certain about now but at one stage the Jubilee was going to struggle to reach 24 tph (trains per hour) because the system had no concept of "recovery" time/additional run time between stations, something that has to be a requirement when interworking trains from different lines/areas together.

 

I believe the proposed system works on the basis of forecasting approaching trains when trying to achieve regular intervals between trains. The Aldgate triangle is often cited as the biggest headache in this respect. I don't often hear tell of the High Street/ Gloucester Road/Earl's Court area where it is necessary to look at trains from five different areas - 3 eastbound District plus Inner & Outer Rail Circles - tied in with trains reversing at High Street Kensington and crew reliefs at Earl's Court.

 

I wouldn't take any bets on whether the train services in these areas will be better or worse than they are now.

[Nearholmer]

I don’t disagree about the bottom LH corner of the Circle, but your post gives a rather long out of date status on the Jubilee system, which now routinely delivers a lot more than 24TPH, and has a now-well-proven service recovery algorithm.

[Ray H]

11 hours ago, Nearholmer said:

I don’t disagree about the bottom LH corner of the Circle, but your post gives a rather long out of date status on the Jubilee system, which now routinely delivers a lot more than 24TPH, and has a now-well-proven service recovery algorithm.

 

I retired nearly 10 years ago and haven't kept up to speed with what the Jubilee have managed to do. I do know that the system as originally presented would have struggled to run the then level of service punctually. There were some significant flaws in the logic that would have caused endless delays in the Stratford area. However, I presume these problems were resolved either before the system went live or since. I would be interested to learn what the maximum tph rate of the current design is.

 

We shouldn't lose sight of the fact that several lines were routinely scheduling anything up to 36tph (and above, I think) several decades ago when there were signallers on the ground so to speak.

 

Apologies for drifting off topic.

[jim.snowdon]

At least the bottom LH corner of the Circle has chords large enough to accommodate a train with comfortable margins. I would have expected the bottom RH corner (Aldgate) to be more of a challenge, given that the two through platforms are only just long enough, and in conventional signalling terms, the overlaps are into the H&C and District Line junctions. Edgware Road has similar problems.

 

Jim 

[CloggyDog]

On ‎11‎/‎07‎/‎2019 at 09:49, Ray H said:

 

I retired nearly 10 years ago and haven't kept up to speed with what the Jubilee have managed to do. I do know that the system as originally presented would have struggled to run the then level of service punctually. There were some significant flaws in the logic that would have caused endless delays in the Stratford area. However, I presume these problems were resolved either before the system went live or since. I would be interested to learn what the maximum tph rate of the current design is.

 

We shouldn't lose sight of the fact that several lines were routinely scheduling anything up to 36tph (and above, I think) several decades ago when there were signallers on the ground so to speak.

 

The Jub currently achieves 30tph in both directions for c2½ hours in each peak Mon-Fri. Even Sat/Sun has 24tph for the bulk of the traffic day.  The method of working at Stratford was changed a while back to utilize just the island platform (14/15) during the peaks and interpeak with train operators 'stepping back' and the single face platform (13) only used off peak (pre-am and post-pm) for trains entering/exiting service.

[Nearholmer]

Yes, I’ve got a little 30mph-type-sign lapel pin, which I think I was given during the Olympics when I worked as a station assistant at Wembley on the Jube, which symbolised the throughput that we had to support.

[Titanius Anglesmith]

On 10/07/2019 at 22:07, Nearholmer said:

the bottom LH corner of the Circle

 

There’s something distinctly paradoxical about that statement! 

 

 

(I know exactly what you’re on about though)