Serious Accident in Bavaria

[SpacePhoenix]

Are PoSA signals (in the UK at least), the ones with two white lights and a red light (to whites for proceed with extreme caution - a white and a red for stop)?

 

[MarkyT]

Are PoSA signals (in the UK at least), the ones with two white lights and a red light (to whites for proceed with extreme caution - a white and a red for stop)?

A 'PoSA' is given by two white 'position lights' arranged at 45 degrees in exactly the same form as a 'shunt' or 'call on' from a main signal. Unlike the shunt or call on however where the lights are steady, the PoSA flashes them in unison. Where a position light head for any purpose is incorporated with a main signal (which must have a red), the position light head does not have its own red light indication and is extinguished when any other aspect is displayed, unlike a ground shunt signal where stop is indicated by two red lights arranged horizontally.
--- old post above --- --- new post below ---

Forgot to add: The main red remains illuminated when a call on or shunt is displayed. That will also be the case normally with a PoSA but as it is for degraded working controls allow the PoSA aspect to be displayed without the associated red if the lamp has failed.

 

[Gerald Fiennes]

I see the German Press is reporting that the cause is being put down to two things: the signaller using the PoSA signal to call the East Bound train on and then, realising his mistake, using the wrong radio switch (ie presumably not the GSM-R Red Button) as a result of which he spoke to all the signallers in the area but no drivers.... I'm still not clear how it is allowed that a PoSA signal overrides the interlocking (ie. track section ahead is occupied). This seems to be pretty basic safety requirement.... Or am I missing something?

 

[Groningen]

There is irritation about Joachim Herrmann (member of the government of Bavaria) with police and justicedepartment. It appears he is talking while not all facts are known.

JH says that both trains got permission to depart. The traindispatcher realized his mistake and tried to reach both trains by radio, but maybe stress made another mistake by pushing the wrong button. Shortly a second message was sent, but than it was too late. But even when the message sent through first time prevented the crash is open for discussion.

At the Theresienmonument (see downstairs) a plaque will be placed to remember this event yearly. How it looks like is still unknown; somewhere 6 months more will known.

 

[BRX]

I'm still not clear how it is allowed that a PoSA signal overrides the interlocking (ie. track section ahead is occupied). This seems to be pretty basic safety requirement.... Or am I missing something?

It does seem surprising. But it appears that there are certain procedures that are supposed to be followe before the over-ride signal is used, which the signaller didn't follow. I've not seen it described anywhere what these procedures are. Presumably, they are such that if they're followed, safety is guaranteed.

 

[axlecounter]

I think that we'll have to wait for the EUB report to finally see some technical details about what happened.

 

[Bletchleyite]

It does seem surprising. But it appears that there are certain procedures that are supposed to be followe before the over-ride signal is used, which the signaller didn't follow. I've not seen it described anywhere what these procedures are. Presumably, they are such that if they're followed, safety is guaranteed.

The whole point of interlocking is that humans are fallible and that as such they may not, either deliberately or accidentally, follow procedure.

Personally, this confirms my view that the German-style subsidiary signal led directly to these deaths, and that it should be removed on single and bidirectional lines[1], to be replaced with a UK-style "have a conversation first, and thus involve two people in the issue" approach. Similarly, the UK style PoSA signal should not be extended to any single or bidirectional line.

[1] Driving on sight and expecting a stationary train is fine, you drive at a speed where YOU can stop in advance of any stationary obstacle. Driving on sight and finding a train going towards you at linespeed on clear signals, OTOH...

 

[Groningen]

This brings back memories of the traindisaster north of Groningen in 1980. Secured with radio, 1 too early departure and a foggy day resulted in 9 people dead.

https://en.wikipedia.org/wiki/Winsum_train_disaster

The radio and used from 1967 till it went wrong.

 

[ComUtoR]

...To be replaced with a UK-style "have a conversation first, and thus involve two people in the issue" approach. Similarly, the UK style PoSA signal should not be extended to any single or bidirectional line.

You are still misunderstanding PoSA's There will always be two people involved regardless of seeing a PoSA or not. In both systems the same mistake can happen. It would not change or remove human error. It is not the PoSA that is the problem.

The "single line" argument also shows a lack of understanding as the same issue would exist on a single direction track too but you may rear end a unit instead of a head on. I would say in that instance you would be more inclined to raise your speed and be more complacent because your not expecting a unit in front. It is also potentially easier to introduce an error because there may be less checks done compared to a single line where you have flow indicators and need additional checks.

Removing the human factor is just as detrimental too as having a 100% computer based system means your fully reliant on the computer and what happens when that goes wrong ? I'm not sure I would trust the computer over a human.

Interlocking is also not 100% safe and has problems of its own. I have seen interlocked points set in the direction of an oncoming train. With complicated layouts I would say its next to impossible to set points away from every train movement. Interlocking is very complicated and is more than points and compensating for human error.

If your looking for a solution then the only solution I see is that you NEVER pass a red. With or without authority. PoSA's are the way forward. Our latest new layout will have PoSA's.

 

[Bletchleyite]

The "single line" argument also shows a lack of understanding as the same issue would exist on a single direction track too but you may rear end a unit instead of a head on.

The issue is not quite the same, using the German system at least.

On a single line, a situation can arise that one train is given authority to pass a signal at danger and run on sight, when the other train in the section has passed a green signal and is running at linespeed. Can the PoSA not result in this? It seems the German system could and did.

Two trains in the same section *both* of which had entered on a PoSA or equivalent would, as you say, be much less of an issue as both would be driving on sight.

 

[ComUtoR]

The issue is not quite the same, using the German system at least.

I agree. The German system sets a max permissible speed and for a set distance. Our system of passing signals is much more stringent and has many caveats. Even with the PoSA

On a single line, a situation can arise that one train is given authority to pass a signal at danger and run on sight, when the other train in the section has passed a green signal and is running at linespeed. Can the PoSA not result in this? It seems the German system could and did. Our system really is pretty safe and should always prevent a collision. I run on a PoSA bi directional line and I'm confident that the system will protect my train.

It shouldn't but the fact remains that it did. Our system would be no different as the same mistake could happen too. Its a head on vs a rear end the potential for collision will be there regardless.

Two trains in the same section *both* of which had entered on a PoSA or equivalent would, as you say, be much less of an issue as both would be driving on sight.

Yes. In our system IF the same mistakes were made and both trains entered an occupied section they should both be going at caution so any impact should not happen (due to going at caution) or at least be mitigated.

PoSA or not when you pass a signal at red you 100% expect an obstacle. It doesn't matter the type of line as you can get still get a unit going wrong direction.

Neil et al. I think I previously mentioned that we need to stop using "drive on sight" It is very misleading. In the UK we drive at caution. "the speed in which you can stop your train in the distance seen to be clear" I could just as easily go 60mph at "caution" as I could doing 5mph at "caution"

From my earlier research the German system was set at 40 kmph (iirc)

I remember a situation a few years back where a unit passed a signal at red and almost collided with the unit in front which had failed. I even read an incident recently where there was a runaway bit of track equipment. How do you allow for things like that. You can't. We do the best with what's available. Humans, computers, interlocking, procedures must all work together to have a safe system of work. Any failure of one system can be fatal.
--- old post above --- --- new post below ---

On a single line, a situation can arise that one train is given authority to pass a signal at danger and run on sight, when the other train in the section has passed a green signal and is running at linespeed. Can the PoSA not result in this? It seems the German system could and did.

The German system did have protections in place. Its just that they appear to have been ignored for some reason. From what I read about the German Erstaz signal it is not designed as a PoSA is in the UK. It has a different purpose than ours. We have more protections because our system is designed to keep the trains running and designed with obstacles etc in mind.

The PoSA debate has come about because its the best equivalent system for laymans to understand. I think that was my fault for bringing it up :/

Neil, thinking on some of the points you raise. I have a nagging suspicion that having the PoSA as the authority actually adds another layer of protection as its another thing that would/should be interlocked. If all you needed was verbal authority then its easier to make a mistake. To make your checks and then still have to activate the PoSA for the Driver means the system will be making a final check. I'm not a Signaller so thats my best, but educated, guess, but its been nagging at me because of posts like yours and something inside me just trusts the systems in place because I have read so many incident reports and seen first hand Signallers doing their jobs.

 

[furnessvale]

Two trains in the same section *both* of which had entered on a PoSA or equivalent would, as you say, be much less of an issue as both would be driving on sight.

I have not been following this thread too closely, but I would not be very happy with two trains approaching each other on "line of sight".

With both trains moving at the same speed, the effective distance between them is halved and both drivers would need to be able to stop in HALF the distance they can see.

 

[MarkyT]

Just to be absolutely clear, in UK a PoSA aspect used on a bidirectional line WILL NOT override opposing route locking. If a movement is already signalled in the opposite direction, or an opposing train is still proceeding through the conflicting section after having been signalled by a normal aspect (or indeed a PoSA itself), then an opposing PoSA CANNOT be displayed. The interlocking simply will not allow it. PoSA aspects are designed only to override certain defined equipment failure modes, not to give an unlimited facility for signallers to completely override the interlocking. I do not know what standard interlocking controls apply to a German 'Ersatzsignal' however.

 

[ComUtoR]

Cheers Marky. Every time you post the technical side of things it makes me warm and fuzzy inside. It serves to reassure me that we do have a safe signalling system and that I can continue to go out there with confidence.

I think I posted up-thread that the Erstaz is not interlocked. (its a local override)

 

[edwin_m]

Just to be absolutely clear, in UK a PoSA aspect used on a bidirectional line WILL NOT override opposing route locking. If a movement is already signalled in the opposite direction, or an opposing train is still proceeding through the conflicting section after having been signalled by a normal aspect (or indeed a PoSA itself), then an opposing PoSA CANNOT be displayed. The interlocking simply will not allow it. PoSA aspects are designed only to override certain defined equipment failure modes, not to give an unlimited facility for signallers to completely override the interlocking. I do not know what standard interlocking controls apply to a German 'Ersatzsignal' however.

However a track circuit failure on the single line, of the type a PoSA is normally intended to work round, would presumably prevent release of the opposing route locking even after the train in question has passed through.

This is an example of the tradeoff that exists in backup systems of this sort. The more protection is built into the backup system, the more it is vulnerable to being put out of action by the same fault that made it necessary to use it in the first place.

 

[Bletchleyite]

Cheers Marky. Every time you post the technical side of things it makes me warm and fuzzy inside. It serves to reassure me that we do have a safe signalling system and that I can continue to go out there with confidence.

I think I posted up-thread that the Erstaz is not interlocked. (its a local override)

Yes, that was my understanding as well.

PoSA is probably a red herring in this case, then, as I think you already said.

I believe, FWIW, that there were as late as the 1990s German secondary lines with no interlocking whatsoever. Makes you shudder.

 

[ComUtoR]

I believe, FWIW, that there were as late as the 1990s German secondary lines with no interlocking whatsoever. Makes you shudder.

In the UK I believe the biggest concern is a sideswipe more than anything else. What makes me shudder is the number of signals that aren't fitted with TPWS.

At one location recently there was a incident which only by the quick thinking actions of the Driver managed to prevent a collision. Scarily even TPWS would not have prevented the collision either.

I understand that from the outside looking in it can be daunting and people will always see fault with the system and try to find immediate solutions. The first steps we need to take is to ensure that we have adequate protections in place and are aware of where the system can fail and tighten up where possible.

 

[AlexNL]

I do not know what standard interlocking controls apply to a German 'Ersatzsignal' however.

As I understand it: procedures. If those are followed to the letter you'll be fine, if not... you can find yourself heading for a head-on collision with another train.

I believe, FWIW, that there were as late as the 1990s German secondary lines with no interlocking whatsoever. Makes you shudder.

It's worse...

Back in 2011, 10 people lost their lives in the Hordorf (Saxony-Anhalt) crash when a freight train had passed a signal at danger. The line was not fitted with any sort of train protection, at that time PZB was only required on lines with maximum speeds exceeding 100 km/h.

 

[AlexNL]

The signaller who was on duty when two trains collided head-on near Bad Aibling in february has been taken into custody. According to the prosecutor, the man was distracted by playing a game on his mobile phone. He is being held on suspicion of involuntary manslaughter, negligent assault and dangerous interference in rail transport.

Source: Der Spiegel

 

[Bletchleyite]

Back in 2011, 10 people lost their lives in the Hordorf (Saxony-Anhalt) crash when a freight train had passed a signal at danger. The line was not fitted with any sort of train protection, at that time PZB was only required on lines with maximum speeds exceeding 100 km/h.

Perhaps no PZB (though I think not absolutely every line in the UK has AWS/TPWS, does it?) but I was referring to *interlocking*, i.e. you could quite happily set conflicting routes (not using Zs1, using a regular signal) in a signalbox and nothing would stop you doing so.
--- old post above --- --- new post below ---

The signaller who was on duty when two trains collided head-on near Bad Aibling in february, has been taken into custody. According to the prosecutor, the man was distracted as he was busy playing a game on his mobile phone. He is being charged with involuntary manslaughter, negligent assault and dangerous interference in rail transport.

Source: Der Spiegel

What a complete fool, if found guilty. However, I still take the view that the system was and is not fit for purpose, in that one allegedly incompetent/inattentive member of staff was all it would ever take to cause a collision.

 

[ComUtoR]

What a complete fool, if found guilty. However, I still take the view that the system was and is not fit for purpose, in that one allegedly incompetent/inattentive member of staff was all it would ever take to cause a collision.

That's all it ever takes.

I know you and I disagree over the systems in place being able to protect the trains and I do agree its more mitigation than prevention but no system is perfect and there will always be potential for failure somewhere. The weakest part of any system is people.

This would not be the first and will not be the last time that a mobile phone has caused a crash on the railway.

There is a thread up in GD where there was a recent collision in the UK. It's very hard to accept it happens but what's the alternative ?

http://www.spiegel.de/panorama/just...erspiel-abgelenkt-gewesen-sein-a-1086820.html
(translated with google translate)

It is a surprising turn in the investigation of the railway accident at Bad Aibling with eleven dead: The accused Dispatchers had been remanded in custody, said the prosecutor's office in Traunstein. The man was accused of having played on his cell phone shortly before the collision of the two trains an online game and thus to have been distracted.

Investigators see a direct link between the games of the Dispatcher and the cause of the crash: "It must be assumed that because of the close temporal relationship of that the accused was distracted from the control of the crossing traffic of trains." The 39-year-old played accordingly over a longer period on his smartphone.
The Rosenheim District Court had ordered detention for manslaughter, negligent assault and dangerous interference in rail transport. The Traunstein prosecutors pressed upon request give any information about why the dispatcher was remanded in custody.
According to investigators, the man accused by the new findings on his cell phone use not only instantaneous failure, but a significantly heavier weight falling into dereliction of duty. The dispatcher had the trains been wrong signals because of gambling with his cellphone and pressed later for emergency calls to the radio the wrong key combination. He had been playing, but denied to have been distracted.

During the collision of two commuter trains near Bad Aibling eleven people had died on 9 February on the route from Holzkirchen to Rosenheim and 85 passengers were partly critically injured.

The investigation into the causes of the train crash lasted despite the new findings of yet, said the prosecutor. "To date, however, no evidence of interference has shown, could be the cause or a contributing cause of the disaster."

 

[Taunton]

I suspect that such a signalling setup would have failed the Interlocking Test in Britain at every point of the last 100 years.

Not that more modern designers are immune. The Ladbroke Grove collision occurred on a newly resignalled layout where the track the departing train was on ended, and points were Set Left so any overrunning train was routed head-on into the Up Main (as happened). They should have been Set Right towards the Down Relief to give Flank Protection fro any overrun, a sideswipe (not that there was any movement signalled alongside) being less of an issue than a head-on, but this somehow wasn't part of the designer's thought. However, it was in the mechanical boxes at Taunton a generation ago, designed and installed I understand in the 1930s. And you thus had to throw a couple of extra levers for each such movement. Which all accepted was the Right Thing To Do.

 

[AlexNL]

Under normal circumstances the systems would have prevented the crash from happening, that's what the interlocking is there for. The IXL did its job, because the signaller had to use the Zs1 override to (wrongly) give the eastbound train permission to depart.

If the Zs1 had not been there, then the signaller had probably called the driver and told him over Zugfunk that he had permission to pass the signal at danger - and the driver would probably have done so without questioning the signaller's authority.

Don't forget the importance of culture, as hierarchy is very important in Germany. It's uncalled for someone to question their superiors, and frankly... I can't see this being any different in the railways: signallers set routes, drivers operate trains according to the instructions given to them by the signallers.

 

[WatcherZero]

Playing poker gambling on his mobile while on duty and incorrectly dialling the emergency number, can see why the charges have been upgraded.

 

[edwin_m]

The Ladbroke Grove collision occurred on a newly resignalled layout where the track the departing train was on ended, and points were Set Left so any overrunning train was routed head-on into the Up Main (as happened). They should have been Set Right towards the Down Relief to give Flank Protection fro any overrun, a sideswipe (not that there was any movement signalled alongside) being less of an issue than a head-on, but this somehow wasn't part of the designer's thought.

Not wanting to start an off-topic discussion, but the length of overrun available from SN109 was much longer towards the Up Main than towards the Down Relief. So although the result of an overrun towards the Up Main was more severe it was also less likely. Maybe the balance of the two risks was wrong, in hindsight at least, but that doesn't necessarily mean it wasn't considered.

 

[AndyPJG]

The signaller who was on duty when two trains collided head-on near Bad Aibling in february has been taken into custody. According to the prosecutor, the man was distracted by playing a game on his mobile phone. He is being held on suspicion of involuntary manslaughter, negligent assault and dangerous interference in rail transport.

Source: Der Spiegel

Also being reported by BBC:

http://www.bbc.co.uk/news/world-europe-36025951

Germany train crash: Controller 'distracted by computer game'

A German train controller has been arrested over the February rail crash that killed 11 people in Bavaria, as prosecutors suspect he was distracted by a computer game at the time.
According to prosecutors he was playing the computer game on his mobile phone and made a signalling error, then dialled the wrong emergency number.
He has admitted that version of events, German media report.
Two commuter trains collided on a single-track stretch near Bad Aibling.
Eighty-five passengers suffered injuries, some of them life-threatening.
The man could be charged with involuntary manslaughter and could face five years in jail. ....

 

[2HAP]

Latest from the Süddeutsche Zeitung, in German.

http://www.sueddeutsche.de/bayern/z...rend-der-notrufe-vom-handy-ablenken-1.2947088

 

[LNW-GW Joint]

The signaller who was on duty when two trains collided head-on near Bad Aibling in february has been taken into custody. According to the prosecutor, the man was distracted by playing a game on his mobile phone. He is being held on suspicion of involuntary manslaughter, negligent assault and dangerous interference in rail transport.
Source: Der Spiegel

Certainly looks like signaller error. He's admitted playing games on his phone at the trial.
http://www.bbc.co.uk/news/world-europe-37936198
This appears to be the gist of the prosecution case:

Although the line has a safety mechanism, prosecutors say the controller mistakenly disabled it, sending two commuter trains towards each other. When he tried to warn the train drivers, he then pressed the wrong alarm button, they say

 

[EAD]

I will type up more when I get home - I had just summarised the information from day 1 of the trial today from local German media, but it logged me out so I lost it!

 

[Tim R-T-C]

Wow that is quite a charge sheet, what sort of penalties does that carry I wonder?