Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 16, 2018 20:46:52 GMT
I was on a Westbound Jubilee line train from Stratford to Waterloo
At Southwark, the train entered the platform at a usual speed but didn't brake on time and slightly overran the platform. On the double doors, the left carriage door matched the right PED. The driver immediately said "dear customers, please be aware that wasn't me, that was the computer. I ask of your patience whilst I contact my line controller". After about 2 minutes, the train departed without opening its doors leaving a few people visibly confused and a few angry faces on the platform (there had been quite a gap in the service). The driver apologised and directed people to change for an eastbound service at Waterloo (what you'd expect) but the whole situation was rather strange.
Any ideas?
|
|
londoner
thinking on '73 stock
Posts: 480 
|
Post by londoner on Mar 16, 2018 21:01:19 GMT
No system is perfect. It may be that the braking system on that particular train, coupled with the natural wear and tear of the track and wheels, led to slightly different behaviour than expected by the computer. I wouldn't be so concerned, unfortunate as it may be that it happened.
|
|
|
|
Post by goldenarrow on Mar 16, 2018 21:02:41 GMT
Usually it's just the computer calculations (VOBC) falling to account for the wear of the train it self, the track standards and adhesions on the JLE are probably some of the most consistent on the network but the trains themselves are subject to much less predictable forces over their service and as such minor discrepancies can build up starting with something as minor as a couple of faded/sluggish brake pads that ultimately skew the calculations of stopping distances meaning situations such as the one you described can occur no real fault of anyone. The fact that Jubilee line trains never identically line up with PED's every time shows just how many possible factors computers still can't account for reliably and still require the human skill set. Not quite on target...  |
|
lt
Posts: 65
|
Post by lt on Mar 16, 2018 22:37:06 GMT
Arriving at Waterloo yesterday from Westminster the train stopped and the moved forward a couple of yards so guess the opposite to the above but at least I didn't have to travel via Southwark! Same thing happened on the district coming into St James Park, In the morning.
|
|
|
|
Post by alpinejohn on Mar 17, 2018 12:08:42 GMT
Once that train had overshot – the acceptable “opening doors” zone, then I suspect the driver had no option but to proceed to the next station without opening the doors. On stations with PEDs I suspect the size of the “acceptable zone” is probably far more restricted than that required to release the doors on S7/S8 stock, and even they encounter a similar problem of passing the mark once in a while.
I am no expert on the Jubilee line TBTC system, however in the main it seems to work fairly well. However I suspect that much like the Victoria line ATO system, the Jubilee presumably still allows the driver to intervene and apply extra braking if he spots the train is coming in a bit too hot. Like this
Undoubtedly variables like passenger loadings, rail conditions and brake efficiency can all contribute to the computer having a hard time stopping precisely at the same spot every time. Train headways and allowed platform dwell times are also factors which make the challenge harder and I guess someone in TFL towers has to decide the balance between the occasional overshoot and running a very intense service.
Obviously if the service was less intense then the approach speed profile could be reduced sufficiently to pretty much guarantee you stop on the mark but with the obvious impact on line capacity. If overshoots occur frequently at that platform then hopefully someone will tweak the approach profile.
|
|
|
|
Post by superteacher on Mar 17, 2018 12:28:55 GMT
The original Victoria Line ATO had quite a big margin for error in terms of stopping.
On the Jubilee and other ATO lines, the only way the driver can attempt to correct an overrun is by moving the manual driving handle. However, this leads to a full EP brake application and disengages the ATO, requiring the driver to reset it in some way. Of course, the other option is to press the emergency stop button on the driver’s console.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 17, 2018 12:33:17 GMT
Once that train had overshot – the acceptable “opening doors” zone, then I suspect the driver had no option but to proceed to the next station without opening the doors. There are options, but they are not normally considered worth the hassle. You can, however, drive it back to the correct position. However I suspect that much like the Victoria line ATO system, the Jubilee presumably still allows the driver to intervene and apply extra braking if he spots the train is coming in a bit too hot. You could put the emergency brake on and bring the train to a stop, but you can't just drop in and put on a bit more brakes. The traction brake controller must be stowed at all times when the train is in ATO, there's no brake handle you can use like they had on the 67s. Obviously if the service was less intense then the approach speed profile could be reduced sufficiently to pretty much guarantee you stop on the mark but with the obvious impact on line capacity. If overshoots occur frequently at that platform then hopefully someone will tweak the approach profile. And with an impact on journey times. The brake rate in the open is already very low, but that's more about wheel slide in adverse conditions. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 17, 2018 12:47:36 GMT
Thanks for all the info, but I have another question. There was already a 6 minute gap in the service during rush hour so when we entered the platform the 'signal' surely would've been green. Why was so much time wasted (about 2 minutes) before the train continued. I understand there must be some sort of procedure but stations West of Southwark must've been getting dangerously overcrowded. It was clear the doors couldn't open so couldn't he have contacted the line controller on the move or is that not allowed?
|
|
|
|
Post by goldenarrow on Mar 17, 2018 12:57:31 GMT
Thanks for all the info, but I have another question. There was already a 6 minute gap in the service during rush hour so when we entered the platform the 'signal' surely would've been green. Why was so much time wasted (about 2 minutes) before the train continued. I understand there must be some sort of procedure but stations West of Southwark must've been getting dangerously overcrowded. It was clear the doors couldn't open so couldn't he have contacted the line controller on the move or is that not allowed? The Jubilee line doesn't operate with Red/Green signals in the traditional sense and as such there are no track based signals in use with ATO on the Jubilee line extension and most of the other parts of the Jubilee with a few exceptions. Since the line uses Automatic Train Operation (ATO) or Transmission Based Train Control (TBTC), the driver will only get authority to move in ATO when a three note tone sounds with the message, "ATO Start Required" and a roughly 10 second countdown for the driver to close the doors, get his pilot light (Door Closed Visual) and activate ATO for the train to start . I'd assume that an overrun of this nature would have to be reported immediately to the line controller in situe because the train won't be able to start up because it hasn't received it's ATO target command as part of the usual sequence of events when a train stops correctly. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 17, 2018 13:13:59 GMT
Thanks for all the info, but I have another question. There was already a 6 minute gap in the service during rush hour so when we entered the platform the 'signal' surely would've been green. Why was so much time wasted (about 2 minutes) before the train continued. I understand there must be some sort of procedure but stations West of Southwark must've been getting dangerously overcrowded. It was clear the doors couldn't open so couldn't he have contacted the line controller on the move or is that not allowed? The Jubilee line doesn't operate with Red/Green signals in the traditional sense and as such there are no track based signals in use with ATO on the Jubilee line extension and most of the other parts of the Jubilee with a few exceptions. Since the line uses Automatic Train Operation (ATO) or Transmission Based Train Control (TBTC), the driver will only get authority to move in ATO when a three note tone sounds with the message, "ATO Start Required" and a roughly 10 second countdown for the driver to close the doors, get his pilot light (Door Closed Visual) and activate ATO for the train to start . I'd assume that an overrun of this nature would have to be reported immediately to the line controller in situe because the train won't be able to start up because it hasn't received it's ATO target command as part of the usual sequence of events when a train stops correctly. Oh Ok, makes sense. I wasn't sure of the correct term to use which is why I put signal in quotations. This is really interesting actually. Am I pushing it or may i ask what happens next? Would the driver restart the whole ATO system (I'm sure that's probably the wrong terminology) or will it be done via the control centre. |
|
|
|
Post by goldenarrow on Mar 17, 2018 13:30:26 GMT
Oh Ok, makes sense. I wasn't sure of the correct term to use which is why I put signal in quotations. This is really interesting actually. Am I pushing it or may i ask what happens next? Would the driver restart the whole ATO system (I'm sure that's probably the wrong terminology) or will it be done via the control centre. I'm not to sure on this as the answer I gave you was an assumption based on the usual procedure when working ATO on the Jubilee, given that the train as you said was stationary for a while, the controllers would definitely need to manually input some form of command or relay instructions to the Train Operator to give authority to move off either by re-entering the train onto the control system or by giving the start command again. |
|
Chris M
Global Moderator
Forum Quizmaster
Always happy to receive quiz ideas and pictures by email or PM
Posts: 19,758
|
Post by Chris M on Mar 17, 2018 16:08:58 GMT
Another possibility (I don't know how likely) is that the driver was given movement authority to drive the train manually to a given point (no further than the next station, and probably not that far)
|
|
|
|
Post by superteacher on Mar 17, 2018 16:50:07 GMT
Another possibility (I don't know how likely) is that the driver was given movement authority to drive the train manually to a given point (no further than the next station, and probably not that far) I think that’s what happens on the Central. The train has to be driven in coded manual until the next station, where ATO can be resumed. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 17, 2018 17:17:18 GMT
Another possibility (I don't know how likely) is that the driver was given movement authority to drive the train manually to a given point (no further than the next station, and probably not that far) I think that’s what happens on the Central. The train has to be driven in coded manual until the next station, where ATO can be resumed. On the Central in the event of an overrun I believe you will indeed have to go in Coded Manual you're right. I think it has to do with the potential that the ATO data has become corrupted. This does in theory require permission from the line controller. On the Jubilee you could similarly select Protected Manual and drive to the next station or wherever, assuming you had a target point. There wouldn't be any particular necessity to get as far as the next station, unlike on the Central where you need to stop over the PAC loops to get your ATO data, however that's obviously the most logical point. It's very important not to get this confused with the RM procedure, however, which would not be used. |
|
|
|
Post by punkman on Mar 18, 2018 11:53:16 GMT
Some general comments in relation to this in addition to the other useful ones already posted. I agree 2 minutes is a long time from a service perpective, but in that time the driver had to make the announcement, call the controller and wait for a reply, then the contoller (or signaller more likely) would have had to reset the computer to allow the train to continue. To get a target point under normal conditions the doors would have to be proved cycled (opened and closed) first, presumably so drivers don't forget to do so. In this case with no door cycling it would have required controller / signaller input to reset the system to allow the train to continue. Possibly in this case the VOBC would also have been swapped over by the control centre to prevent the same thing happening at the next station, i.e. normally the rear VOBC is "Active" controlling the train but the front one could have then been made the active one. (If it was me I would have driven in protected manual - PM - to the next station anyway to make sure the same thing DIDN'T occur!) Fortunately in this instance it was Southwark rather than Waterloo, which would entail more inconvenience as the overcarried passengers would have had to change at Westminster where there is no simple cross platform changeover as the east and westbound platforms are on different levels.
Regarding the possibility of the driver applying an emergency brake, the train is normally coming in at or close to the target speed and there is no knowing it may overrun until the very last split second, by which time taking into account reaction time it would make little or no difference.
I have heard of instances of trains sliding through due to grease on the track after night time maintenance, but I have to say the Jubilee is generally pretty reliable in stopping at the right place.
@it's experience would most likely a train being driven in PM, probably a trainee driver.
|
|
|
|
Post by alpinejohn on Mar 18, 2018 14:15:20 GMT
Hmm - As a passenger elements of this thread are not entirely comforting.
In normal circumstances I am sure it makes sense for the computers to be left to manage the train's progress in line with whatever movement authority it was issued with. But I am concerned by the implication that the system is intentionally set up so that the "man(woman) in the cab" "cannot instantly trigger an emergency brake application" if the circumstances warrant.
The man in the cab may for instance spot something untoward well before any automated system knows there is a problem - just like someone drilling a building pile straight through the tunnel roof. I had previously thought that all the driver needed to do was fully release the control stick (dead man's handle) to apply the brakes. Am I wrong?
It seems strange to think the man in the cab is effectively discouraged to step in even when it becomes clear the micro-chip has got it wrong because of the time required to intervene (stuff being locked out?) or wary of the subsequent hassle to get the train moving again if they intervene.
I suspect that most of the customers who were unable to alight as planned, would have preferred a bit of late/heavy braking rather than getting carried onto the next station and having to retrace their steps. Presumably someone is actively monitoring both the frequency and locations of these events and also which units are involved just in case there is a developing fault or evidence that the approach profile at that location has been set just a bit too aggressively to ensure trains gets stopped where they should each and every time. If that data indicates specific trains (or times of trains) are responsible for a significant number of overshoots then, if nothing else, it hopefully that provides the maintenance folk with something to investigate.
|
|
|
|
Post by superteacher on Mar 18, 2018 14:24:18 GMT
Hmm - As a passenger elements of this thread are not entirely comforting. In normal circumstances I am sure it makes sense for the computers to be left to manage the train's progress in line with whatever movement authority it was issued with. But I am concerned by the implication that the system is intentionally set up so that the "man(woman) in the cab" "cannot instantly trigger an emergency brake application" if the circumstances warrant. The man in the cab may for instance spot something untoward well before any automated system knows there is a problem - just like someone drilling a building pile straight through the tunnel roof. I had previously thought that all the driver needed to do was fully release the control stick (dead man's handle) to apply the brakes. Am I wrong? It seems strange to think the man in the cab is effectively discouraged to step in even when it becomes clear the micro-chip has got it wrong because of the time required to intervene (stuff being locked out?) or wary of the subsequent hassle to get the train moving again if they intervene. I suspect that most of the customers who were unable to alight as planned, would have preferred a bit of late/heavy braking rather than getting carried onto the next station and having to retrace their steps. Presumably someone is actively monitoring both the frequency and locations of these events and also which units are involved just in case there is a developing fault or evidence that the approach profile at that location has been set just a bit too aggressively to ensure trains gets stopped where they should each and every time. If that data indicates specific trains (or times of trains) are responsible for a significant number of overshoots then, if nothing else, it hopefully that provides the maintenance folk with something to investigate. The driver can apply the emergency brake anytime - let’s make that perfectly clear. What was being said was that by the time a driver realises that a slight overrun is happening, it makes little difference whether the brake is applied or not. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 18, 2018 14:38:19 GMT
In normal circumstances I am sure it makes sense for the computers to be left to manage the train's progress in line with whatever movement authority it was issued with. But I am concerned by the implication that the system is intentionally set up so that the "man(woman) in the cab" "cannot instantly trigger an emergency brake application" if the circumstances warrant. The man in the cab may for instance spot something untoward well before any automated system knows there is a problem - just like someone drilling a building pile straight through the tunnel roof. I had previously thought that all the driver needed to do was fully release the control stick (dead man's handle) to apply the brakes. Am I wrong? Oh no, no. Sorry. You're not wrong. The thing is, in the old 1967 stock cab you had a brake handle you could pull that was frequently used owing to certain flaws in the ATO system and the braking systems of the trains it was running on. If a driver spotted a train was coming in too fast, as you described, they could give the handle a quick tug, and get the thing stopped. You can read more about it here. With TBTC and the 1996 stock, things are a little different. The 1996 stock's brakes don't have the same flaw. What's more, what was being described is that the ATO always comes in more or less as fast as it can, it sits basically on the "hand" which is a red line on the Train Operator's Display that marks the maximum speed the system has set for the train. This hand moves steadily to the left - down to 0 - as the train approaches the stopping mark. And the ATO sits on it as far as it can, sometimes a bit over actually, sometimes a bit below, but on it as far as it possibly can. So there's not really a lot of time to spot: oh, this is a bit faster than usual, I'll just put the brakes on now and avoid overshooting by 6 inches. What you can do is to precisely trigger an emergency brake application if the situation warrants, by pushing the dead man's handle out or hitting the emergency stop button. But there's no dipping into the system and applying a little extra braking. There's the full emergency stop, or let it carry on doing its thing. The full emergency stop is useful in full emergencies, but not a practical way of avoiding overshoots - not because it would be the wrong thing to do if you were able to notice it was going to miss the marks by 6 inches and react in time, but because it's really not easy to notice you're going to miss the marks by 6 inches and react in time. If it's possible at all. And obviously the full emergency brake rate is pretty fixed, so you'd have to choose your moment quite well  The PEDs do not afford the same kind of margin for error I suppose you would have had on the Vic in the days of the 67s. |
|
The PEDs do not afford the same kind of margin for error I suppose you would have had on the Vic in the days of the 67s.
