TreacleMiller
Member
This.
Modern LED ones can be so bright at range they distort and are harder to read than a feather.
-
Our new ticketing site is now live! Using either this or the original site (both powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!
Lumo overspeed incident at Peterborough (17/04/2022)
- Thread starter TheBigD
- Start date
- Status
Modern LED ones can be so bright at range they distort and are harder to read than a feather.
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R
RailUK Forums
Fleetmaster
Member
The underlying truth of this incident is rather depressingly familiar. A situation arose where by design, the only thing that prevents a multi fatality accident in a rare but entirely foreseeable type of normal operation, is the human link in the chain of safety critical systems. This is the very definition of an unsafe system.
In that scenario, better training and risk assessment at the operator level can only delay an incident. It was always going to happen eventually, and in years not decades. And in that regard, we are surely well past the point that any promises that safer systems are due for delivery in Year X will actually be delivered anywhere close to Year X, if indeed they ever get done at all.
The system is what It is. What else has been overlooked, what risks have arisen in the recent decades as trains and tracks have been upgraded but the manner of signalling has at a very basic level not changed at all really? As we can see from this incident, the assumption that this system cannot be upgrade in certain cost effective ways, is erroneous. A fix existed. It sadly took not one but two near misses for it to happen.
In that scenario, better training and risk assessment at the operator level can only delay an incident. It was always going to happen eventually, and in years not decades. And in that regard, we are surely well past the point that any promises that safer systems are due for delivery in Year X will actually be delivered anywhere close to Year X, if indeed they ever get done at all.
The system is what It is. What else has been overlooked, what risks have arisen in the recent decades as trains and tracks have been upgraded but the manner of signalling has at a very basic level not changed at all really? As we can see from this incident, the assumption that this system cannot be upgrade in certain cost effective ways, is erroneous. A fix existed. It sadly took not one but two near misses for it to happen.
stephen rp
Member
- Joined
- 25 Jun 2016
- Messages
- 241
That sounds more like the greater problem was the corporate link - in the TOCs' training, and in Network Rail's not doing the fix after the first incident. Human choice may still be there but at a level beyond an undertrained driver.
(See "Herald of Free Enterprise")
Signal Head
Member
- Joined
- 26 May 2013
- Messages
- 407
Which is why the approach release point will generally be closer to the signal than with a JI.
Bletchleyite
Veteran Member
With regard to the luggage side of this, LNER have just sent a Tweet being rather more specific about luggage than normal, and linking to a page stating you're now only allowed one large bag, plus an airline sized carryon, plus a small handbag, rather than three of the traditional maximum size. It's also quite specific about where to put them.
Is this as a result of this incident or was it always like that?
The page is here:
www.lner.co.uk
and says:
(The rest of it is a bit too pictorial to quote)
Is this as a result of this incident or was it always like that?
The page is here:
Luggage information
We all a maximum of three bags per passenger on our trains. A maximum of one large suitcase, a carry on bag and small handbag can be brought onboard. Find out more here.
and says:
(The rest of it is a bit too pictorial to quote)
Adam Williams
Established Member
It was like that back in May 2022, seemingly:
And June 2021:
And June 2021:
Bletchleyite
Veteran Member
Ah OK, I've just not noticed, then. I did try the Wayback Machine but it for some reason won't presently load for me. Thanks.
dan4291
Member
My understanding is that LNER's luggage policy has been in place for some time but has never been enforced or even advertised really.
12LDA28C
Established Member
With regard to the luggage side of this, LNER have just sent a Tweet being rather more specific about luggage than normal, and linking to a page stating you're now only allowed one large bag, plus an airline sized carryon, plus a small handbag, rather than three of the traditional maximum size. It's also quite specific about where to put them.
Is this as a result of this incident or was it always like that?
The page is here:
Luggage information
We all a maximum of three bags per passenger on our trains. A maximum of one large suitcase, a carry on bag and small handbag can be brought onboard. Find out more here.
and says:
(The rest of it is a bit too pictorial to quote)
Ridiculous. A 'carry on' bag? This is a train not a plane, all bags are 'carry on'. I don't recall there being a luggage hold on an Azuma.
It is very LNER language lol. I'm guessing it's just an easily understood terms for a size of bag most people would be familiar with. Unless they think the DVT/bike storage on azuma is a hold now.
stephen rp
Member
- Joined
- 25 Jun 2016
- Messages
- 241
Deep Vein Thrombosis? Even more like flying.
12LDA28C
Established Member
I understand the signalling has been modified at this location so that the junction signal can now only show a single yellow with the Junction Indicator lit so that the signalling sequence that was a precursor to this event can now not be given by the signaller, mitigating the risk of a similar incident happening again.
AlexNL
Established Member
- Joined
- 19 Dec 2014
- Messages
- 1,691
Driving Van Trailer, a type of vehicle.
That's a reasonable mitigation, I wonder how many other locations there are across the network where this sort of circumstances leads to misreads.
12LDA28C
Established Member
Driving Van Trailer, a type of vehicle.
That's a reasonable mitigation, I wonder how many other locations there are across the network where this sort of circumstances leads to misreads.
It was a combination of factors that caused the incident, not just the misread but the distance of the junction signal from the pointwork and the acceleration capabilities of the train that led to excessive speed being attained by the train prior to reaching the junction. You'd hope it wouldn't be necessary to mitigate against drivers not noticing or responding to Junction Indicators being lit at every junction around the network.
You could say the same about drivers not noticing or responding to a red aspect at every signal around the network, but...
If the sequence of aspects has been changed, but with no other engineering control of the risk of overspeeding, so that the risk mitigation is ultimately still reliant on a human being then, as Fleetmaster says in their excellent post earlier in the thread (post 542), it's the definition of an unsafe system. Unfortunately that seems to be the underlying culture, that hazards that result from workplace human factors are often placed in the 'too expensive/too difficult/too late now (we didn't think of that during the design phase)' pile, and the final responsibility for risk mitigation is passed to the human being at the front of the train.
12LDA28C
Established Member
You have to have a certain level of expectation that human beings will behave in a certain way though, ie reacting appropriately to signal aspects. If you didn't have that expectation, no trains would run at all as the risk would be too high. Now that a risk has been identified at this particular location, which clearly wasn't predicted at the design and implementation stage, action has been taken to mitigate the risk and rightly so. I assume that the action taken has been relatively low cost in comparison with say installing TPWS overspeed at the junction or moving the signal.
So the problem has been put in the 'too expensive' pile and the responsibility for risk mitigation has been passed to the driver (again).
The one thing that is entirely predictable about human beings is that they can behave in very unpredictable ways (I get involved in a lot of investigations and sometimes the actions taken by a driver, guard etc. are completely baffling unless you see them in the context that human beings do unpredictable things at times).
All the human factors work so far has been focussed on the people at the sharp end (e.g. training them on an understanding and use of NTS etc.). The next big leap I am hoping to see is that the people involved at the design stage will become much more aware of human fallibility and unpredictably. But I'm not holding my breath just yet.
12LDA28C
Established Member
Only in so much as the driver is now expected to react appropriately to a single yellow signal and prepare to stop at the next signal which should be expected to remain at danger. If you believe that this is an unreasonable demand then again, I suggest we should close the entire railway network right now as it's clearly unsafe, according to you.
It's a reasonable expectation that the driver will do so on most occasions, but to assume it will happen every time without fail (and therefore provide no mitigation for human error) shows a lack of understanding of how human beings behave. And progress on that has been glacially slow: it only took us ~150 years to get TPWS across the network, after all. Before that, it was all about blaming the individual: treating them as though they had (deliberately) done something wrong, then disciplining/punishing them.
Unsafe is a relative term. Is the railway 'safe'? That depends on the level of risk, and the number of injuries/fatalities, we (society) are prepared to accept. Could it be safer? Undoubtedly. But we have reached the point where there is limited scope for improvement simply by relying on the human being at the front of the train, hence ETCS etc. to provide ever-higher levels of mitigation. And part of the progress towards an ever safer has to include a greater understanding of human factors by systems/equipment designers, to address the organisational/workplace human factors that have so far not been addressed in any depth (and I know for a fact that the HF experts at RSSB would agree with me on that).
12LDA28C
Established Member
A reasonable level of risk (whatever that looks like) has to be accepted. You can never fully eliminate risk, it's simply not possible where humans are concerned and I'm just not talking about train drivers (eg the Great Heck incident). There comes a point at which a cost/benefit analysis will conclude that chucking unlimited money at a problem will not provide sufficient benefit to justify the expense.
bramling
Veteran Member
It’s okay as a mitigation, but it does have the difficulty that drivers will potentially acclimatise to the single yellow aspect, which then becomes “normal”. Hopefully it’s enough to achieve the objective of not having a massive overspeed over the points, but in the grand scheme of things it’s far from a perfect solution.
12LDA28C
Established Member
Quite possibly but drivers are trained to expect that the next signal will be held at danger and drive accordingly, which in this instance should prevent them applying power up to 80mph. If they fall into the trap of 'when I get a yellow here the next one is always green' then that is no different to any approach-controlled (or similar) signal anywhere else on the network.
But they're human beings who make mistakes, so it is inevitable that there will always be the possibility that they won't 'drive accordingly'. That's precisely what this whole thread is about: a driver who didn't drive according to the indications given by the signalling.
To state that 'drivers are trained to expect that the next signal will be held at danger and drive accordingly which...should prevent them applying power up to 80mph' completely misunderstands the nature of human behaviour. Of course they know not to take power in that situation - they've will have been trained that, and will likely have been confronted with the same scenario dozens of times before with no problem at all. But you have to factor in human fallibility - there is always the possibility that an individual driver on an individual day might not respond to the signalling indications in the same way as they may have dozens of times before, for any number of reasons: fatigue, distraction, overwork, underwork, complacency, conditioning (e.g., 'falling into the trap of 'when I get a yellow here the next one is always green' as you said yourself) etc. etc.
It's no different to a SPAD: of course the driver knows they need to stop at the red, and they know exactly which controls in the cab to operate to do so (the evidence is there in the fact they will have done so dozens of times before), yet SPADs still happen.
That's why 'driver training' is never an adequate mitigation of risk, other than for very low-risk hazards - you can train someone to do a task till you're blue in the face, but you can still never guarantee that they will complete the task as they were trained every single time. Given that knowledge, in any credible mitigation strategy for significant risks, driver training and anything else to do with the human doing the job must come at the very, very bottom of the hierarchy, with appropriate engineering controls coming first.
Yes, of course those controls (and the cost thereof) must be proportionate to the risk, but given that in this incident overspeeding at the junction very nearly resulted in a high-speed derailment with potentially catastrophic consequences, the cost of implementing TPWS OSS protection is negligible in comparison, as is the case for TPWS OSS and TSS at a huge number of high risk signals.
TreacleMiller
Member
Quite right.
That's the risk at every signal showing a single yellow, every one of them.
Two drivers here messed up, one didn't see the signal due to his blind being stupidly low - everything else as to why they messed up is simply just filler. At another TOC they would have been fired.
This seems like an argument that should be taken to its own thread. Completely eliminating the risk of human error from all tasks seems impractical.
bramling
Veteran Member
Yet it seems to be more of an issue in some places than others, hence why we have multi-SPAD signals. There is a definite issue with drivers normalising things which become routine and not acting upon the message being conveyed, even in situations where the sighting is absolutely perfect.
Maybe, however it’s worth remembering LNER have had an overspeed incident in a different part of the Peterborough layout.
Likewise the fact that a second driver messed up in the self-same way on the Up Fast despite the first incident getting a blaze of publicity does suggest there’s more to it than humans simply not paying attention.
anothertyke
Member
I can see the SPAD trap argument in cases where an amber and a feather followed by a clearing signal is frequently encountered. My impression from the report is that in this case, for a non-stop train, this is an unusual move, so the driver will not be preconditioned to expect the next signal to be anything but red. So there is significant mitigation in what has been done.
12LDA28C
Established Member
Indeed, although note that signals are not 'amber' on the railway, but yellow. Might as well use the correct terminology.
TreacleMiller
Member
Totally different incident involving the LNER driver, the fact it was in Peterborough is irrelevant. Overspeed and TPWS activations are more common than is known - it happens.
Peterborough signal box leaving signals at red for no reason other than "being busy" on that occasion didn't help.
Sure not being able to see your next signal increases spad risk but the driver in this instance wasn't following company policy.
- Status