Union claims Northern line collision ‘narrowly avoided’

[class303]

http://www.london24.com/news/union_claims_northern_line_collision_narrowly_avoided_1_2272440

Tube union RMT has traded blows with London Underground over claims a head-on collision between two Northern line trains was narrowly avoided earlier this week.

A train departing Finchley Central was asked to apply the emergency brake when it emerged another train was approaching from Mill Hill East, the union said.

The driver was then forced to reverse back into the station the two train lengths that had been travelled, instead of continuing to Mill Hill East, which is a single-station spur off the Northern line with only one track.

Both trains were in service and carrying passengers.

RMT, which represents drivers on the London Underground, said the incident took place on Monday evening, though it did not provide exact times.

The union’s general secretary Bob Crow said: “If it hadn’t been for the swift and decisive action of a Northern Line driver we could have had a head-on collision on our hands with fatal consequences.”

But a spokesman for London Underground dismissed the claim.

He said the trains were a full kilometre apart and on entirely different sections of track.

He added: “The safety systems worked entirely as they are supposed to work. If the train had gone onto the same section of track the trains would have stopped.”

 

[YorkshireBear]

Unfortunately for Bob his constant talking crap in the media leads me to believe the LU side of the story!

 

[Tomnick]

Hmm, well I'd be amazed if the signalling system allowed a conflicting route to be set - such wrong-side failures are very rare indeed. There's always the possibility of a SPAD though...!

 

[A-driver]

Hmm, well I'd be amazed if the signalling system allowed a conflicting route to be set - such wrong-side failures are very rare indeed. There's always the possibility of a SPAD though...!

Not on a train being driven by a computer there isn't...

And if it was a SPAD then the trip cock would have brought the train to a stand, they wouldn't have had to broadcast an emergancy message.

 

[Tomnick]

That's true - do they still have tripcocks on the ATO sections anyway? It wasn't my intention to speculate - apologies, I did go off on a bit of one, but I am getting rather tired of brother Bob's often entirely misinformed rants that usually do more harm to his members than good.

 

[A-driver]

That's true - do they still have tripcocks on the ATO sections anyway? It wasn't my intention to speculate - apologies, I did go off on a bit of one, but I am getting rather tired of brother Bob's often entirely misinformed rants that usually do more harm to his members than good.

Why misinformed out of interest? Have you any examples of things he has said which have been wrong/misinformed?

He merely writes from his political view point-that is no different from many politicians, journalists, business leaders etc.

One of the original reasons for rail unions was to protect workers safety. If it has been put at risk then he is right to 'rant'. Is it not better he brings things like this to light than goes on about pay rises etc?

 

[Mojo]

Trains in ATO mode can still Spad...

Also, no Tripcocks on trains which operate Automagically or in ATP/Protected manual.

 

[A-driver]

Trains in ATO mode can still Spad...

Also, no Tripcocks on trains which operate Automagically or in ATP/Protected manual.

If the train is being automatically operated then how would it be possible for the driver to spad?

And as none of us have any details of what went on why bother speculating?

 

[Nym]

If the train is being automatically operated then how would it be possible for the driver to spad?

And as none of us have any details of what went on why bother speculating?

SPADs are easilly possible when working in Auto...

 

[A-driver]

SPADs are easilly possible when working in Auto...

How? I'm genuinely asking, I don't know anything about ATO.

Surely the point of ATO is that the train will stop itself at locations or when it gets too close to the train infront-the driver isn't controlling it to signals?

 

[Nym]

How? I'm genuinely asking, I don't know anything about ATO.

Surely the point of ATO is that the train will stop itself at locations or when it gets too close to the train infront-the driver isn't controlling it to signals?

ATO is controlled by software / hardware that assigns a target stop point / target speed point in the future. This has a safety margin to take into account the inherant instability of software controlled systems. Where a precise stop point is required, the control system is able to overshoot...

There is also the possibility of 'Cat B SPADs' that remain as with any system, if the train can't stop in time, it won't...

Either way, that is only one instance of where ATO can overshoot, without any other system failures, there are many instances where something failing will cause a SPAD, mainly due to just how agressive the ATO software is nowerdays...

This is with ATO, ATP drives ATO and for acronym haters, I'll provide details if needed.

It gets hard to explain without going into different types of control system implimentation and open / closed loop stability... But for those who know, ATO is open loop unstable, and very close to marginally stable in closed loop.
--- old post above --- --- new post below ---
Also, Trainstops are removed from sections fitted with TBTC / ATP / Thales S40, and the Tripcocks are also disabled on 1995TS when running in PM - ATP or ATO... It is ONLY enabled in "Tripcock Manual"

 

[Tomnick]

Why misinformed out of interest? Have you any examples of things he has said which have been wrong/misinformed?

He merely writes from his political view point-that is no different from many politicians, journalists, business leaders etc.

One of the original reasons for rail unions was to protect workers safety. If it has been put at risk then he is right to 'rant'. Is it not better he brings things like this to light than goes on about pay rises etc?

Sorry, didn't see that the thread had been moved! I'm certainly in favour of rants that might do something to improve safety, and I know that a lot of union time is spent working towards this. You might recall Mr Crow's recent demands for all level crossings to be abolished. In many cases, there'd be little or no improvement in terms of safety, yet many job losses (largely RMT members too). I remember reading some of the outcry - both ASLEF and RMT - following Moreton-on-Lugg, where it was clear that neither union leader had a very good understanding of the subject - it was claimed that it was possible to raise the barriers with signals clear, for example. As an RMT member myself, I'm very grateful for the support of my union - but significant inaccuracies don't help the cause very much.

 

[A-driver]

ATO is controlled by software / hardware that assigns a target stop point / target speed point in the future. This has a safety margin to take into account the inherant instability of software controlled systems. Where a precise stop point is required, the control system is able to overshoot...

There is also the possibility of 'Cat B SPADs' that remain as with any system, if the train can't stop in time, it won't...

Either way, that is only one instance of where ATO can overshoot, without any other system failures, there are many instances where something failing will cause a SPAD, mainly due to just how agressive the ATO software is nowerdays...

This is with ATO, ATP drives ATO and for acronym haters, I'll provide details if needed.

It gets hard to explain without going into different types of control system implimentation and open / closed loop stability... But for those who know, ATO is open loop unstable, and very close to marginally stable in closed loop.
--- old post above --- --- new post below ---
Also, Trainstops are removed from sections fitted with TBTC / ATP / Thales S40, and the Tripcocks are also disabled on 1995TS when running in PM - ATP or ATO... It is ONLY enabled in "Tripcock Manual"

So we arnt talking a cat A spad then unless I still misunderstand? Ie not 'operator' error but software error - I would class spad to mean driver error in the context of this thread as a spad due to software error is still software error which is what the article from the union is highlighting?

 

[Nym]

So we arnt talking a cat A spad then unless I still misunderstand? Ie not 'operator' error but software error - I would class spad to mean driver error in the context of this thread as a spad due to software error is still software error which is what the article from the union is highlighting?

But...
The lines do get a bit blured when it comes to operating in ATO when the driver is still responsible for the train, if, for example, he/she suspects an ATO fault and realises that the train will not stop in time, then should move the TBC immediately from O/R to Emergency to increase the brake force (by increasing the taget accelaration rate sent to the BGUs, by means of breaking the SBC).
If this has not happened and a SPAD occours, puely due to the fault of the ATO box, should the driver, still being responsible for the train, accountable for the SPAD, or it it entirely the ATO box's fault?
At what point does an ATO fault fail to be an ATO fault that is an accident and come under the accountability of the driver to intervine?

Playing devil's advocate here, but honestly, what do we all think the point is where ATO should be overruled by the driver / Train Op? Also, would failure to do so result in the driver being accountable?

 

[A-driver]

But...
The lines do get a bit blured when it comes to operating in ATO when the driver is still responsible for the train, if, for example, he/she suspects an ATO fault and realises that the train will not stop in time, then should move the TBC immediately from O/R to Emergency to increase the brake force (by increasing the taget accelaration rate sent to the BGUs, by means of breaking the SBC).
If this has not happened and a SPAD occours, puely due to the fault of the ATO box, should the driver, still being responsible for the train, accountable for the SPAD, or it it entirely the ATO box's fault?
At what point does an ATO fault fail to be an ATO fault that is an accident and come under the accountability of the driver to intervine?

Playing devil's advocate here, but honestly, what do we all think the point is where ATO should be overruled by the driver / Train Op? Also, would failure to do so result in the driver being accountable?

Well it could only be driver error if proved the driver knew something was wrong. As far as I have always understood ato, if the driver sees an obstruction on the line, believes they are getting too close to the train ahead etc then they would need to act as soon as possible. It's a very grey area if a collision occurs, as you say, how accountable a driver would be as it would come down to proving they could have acted faster. Possibly where ato makes work harder for a driver as it would be very easy to loose concentration far quicker than when actually driving.

Equally it would be a problem if a driver was over cautious and kept overriding the ato so I guess the balance of blame will be decided after incidents occur.

But back on this topic, when it was suggested it could have been due to a spad I read that to be implying driver error, my point being that under ato a driver error couldn't cause this kind of scenario, unless I still misunderstand ato. The driver would only be able to stop the train once aware that the section was occupied, and if s/he can't see any other trains on the line and the ato moves the train into the section then they would have no reason to believe that it is.

 

[Nym]

OK, I'll answer this directly for the newest ATO and ATP systems I work on...

The train will not move in normal operation under false driver command, providing that...
a) All the safety system circuits are cut in, in the 'normal' position.
b) The ATP is correctly functioning, if this fails, 99.99(and many more 9s)% of the time, it will trip the RTC, so the train will not move...
c) No ******* feeds or failures are present within the safety system circuits, again this is very unlikely, and would usually result in the RTC MCBs tripping, again, resulting in the train will not move... (Even in RM)

Or of course one can move the train to SPAD if...

The train is in Restricted Manual Forward...

PS: I do not work on Thales S40, also known as TBTC, Jubilee Line ATO/ATP or Northern Line ATO/ATP

 

[transmanche]

Playing devil's advocate here, but honestly, what do we all think the point is where ATO should be overruled by the driver / Train Op? Also, would failure to do so result in the driver being accountable?

An interesting point. I guess it would start by having to know what LU's rule book said about such situations. If the Op did everything the rule book says, then surely he/she can't be held accountable for the SPAD?

However I only understand the theory of how the original Victoria line ATO system worked. I don't know anything about the system used on the Northern line.

 

[Mojo]

Cat A Spads are not necessarily the fault of the Train Operator (T/Op), it is merely the train passing a signal (or limit of movement authority) without authorisation, and where the signalling is functioning correctly. There was a Cat A4 Spad on one Automatic line the other week. This was not the T/Op's fault though.

Traincrew depots running wholly Automatic in service still have Spads on their scorecards though; this is typically when not running in Auto (eg. in a depot environment). Parts of the Central line also run Protected Manual on Sundays too.

 

[maniacmartin]

Parts of the Central line also run Protected Manual on Sundays too.

Out of curiosity, why only on Sundays?

 

[transmanche]

Out of curiosity, why only on Sundays?

Presumably to retain T/Op knowledge and skills at a time when the service level is lower - i.e. practise when it's quiet.

 

[A-driver]

Out of curiosity, why only on Sundays?

Many ATO lines do that-we have been told there will be time scheduled when the Thameslink core will be manually driven to retain road knowledge of drivers incase of ATO Failiure.

 

[jon0844]

When the new franchise tries to run 20+ trains per hour through the core, the only knowledge drivers will need is to find new and interesting ways to apologise for the delays when a train fails and/or passengers fail to board and alight in about 10 seconds.

 

[londonboi198o5]

When you say thameslink core running ato do what do you mean

Do you mean fcc thameslink as in FCC NR running in automatic???

I didn't think NR would allow automatic running??

Or am I still jet lagged and mis reading this

 

[A-driver]

When you say thameslink core running ato do what do you mean

Do you mean fcc thameslink as in FCC NR running in automatic???

I didn't think NR would allow automatic running??

Or am I still jet lagged and mis reading this

The core from belle isle/ Kentish town to London bridge/elephant will have to be run by ato once all is open as conventional signalling won't allow the amount of trains that they need to run.

But yes- what is currently the NR/FCC thanes link core section.

 

[Peter Mugridge]

I wonder - would NR and the H&S people ever be persuaded to allow the same operating methods as on the Paris RER in peak times though?

 

[tbtc]

So one side says "head-on collision ... with fatal consequences ... narrowly avoided"

The other side says that the trains were "a full kilometre apart and on entirely different sections of track".

Now, there's usually a little bias/ artistic licence in these kind of Union/ Management arguments, but surely one side must be wrong here?

Worrying if its the Union lying to try to exaggerate "safety" hysteria, but more worrying if its the operator lying about safety by pretending that two trains that could have crashed on a single track line were nowhere near each other.

 

[transmanche]

Now, there's usually a little bias/ artistic licence in these kind of Union/ Management arguments, but surely one side must be wrong here?

Will this incident trigger an RAIB investigation?

 

[tbtc]

Will this incident trigger an RAIB investigation?

I don't know, but I hope that there's some investigation into this - I don't mind a bit of the "half full/ half empty" kind of arguments that you sometimes see between Unions and Management, but the two versions of the story seem so far apart that someone appears to be telling porkies here, and I'd like to know the truth.

 

[swt_passenger]

The only mention of this 'problem' in District Dave starts off by suggesting the 'RMT are wrong on this'.

http://districtdave.proboards.com/thread/21481/northern-line-tbtc?page=7

Given the long debates that usually occur over there, they're a bit quiet on this subject...

 

[paul332]

So one side says "head-on collision ... with fatal consequences ... narrowly avoided"

The other side says that the trains were "a full kilometre apart and on entirely different sections of track".

Now, there's usually a little bias/ artistic licence in these kind of Union/ Management arguments, but surely one side must be wrong here?

Worrying if its the Union lying to try to exaggerate "safety" hysteria, but more worrying if its the operator lying about safety by pretending that two trains that could have crashed on a single track line were nowhere near each other.

Not exaggerating at all. Merely pointing out the inherently catastrophic nature of capitalist-inspired automatic trains designed to remove all drivers and smash the unions. I was on a Jubilee train yesterday that came to a dramatic stop only metres from the train in front, and I really thought that was it for hundreds of passengers. Bob Crow must reverse all the so-called technological improvements and save millions of lives.