That’s why people here, when they hear the fire alarm they just wait for it to go off: the probability that it is actually a fire is minuscule… but the alarm triggers a lot.
Not sure how to actually solve this… but I’m also not making excuses for what happened. High speed trains is not the same as a fire alarm in some office building, the people in charge should be held responsible somehow.
Guess whats the consequence of that? Less and less people pay attention to the alerts.
Safety culture is blameless, but only if everyone plays along.
This literally happened to me last night, I was having dinner at my apartment, and suddenly I start hearing a smoke alarm, but not quite inside the apartment, I just kept having dinner thinking "bah, someone probably burned a toast".
Then it wont stop, I walk to door and hear people talking, still don't bother enough, wait a few minutes, then go out and my neighbor said "there are three fire trucks down there, I think it's a real one" hahhaha
The hallway was kind of hazy and smelled like burnt food, for sure... so I just grabbed my laptop and went out hahaa.
This isn't a game of Hot Potato where the last person touching it goes to jail for gross negligence. You have to look at it through a Swiss Cheese model: every individual safety layer can and will have holes, but a disaster can only occur when all of those holes align.
Why did the LZB system fail? Why did the train dispatcher use the wrong backup procedure? Why did a backup procedure with zero safety checks even exist? Why did the train operator accept this backup procedure? Why did the train allow high-speed driving under those unsafe conditions?
It does indeed seem like the train dispatcher made a big mistake, but that was far from the only thing that went wrong, and it would be myopic to use him as the scapegoat for a far larger systemic issue.
If I can have reasonably accurate gps readings of my phone while I'm on the train, there is zero reason for train companies to not have systems that are at least as accurate.
No, what I describe has nothing to do with the companies, that's the research state. It's just an indictment of you making bold statement about a subject, you don't now enough about.
> I can have reasonably accurate gps readings of my phone
Would you dare to cross on a rope over a deep valley, blind using only GPS readings of your phone? Because that's kind what you are suggesting, expect you also move a few hundred other people.
There are optimizations possible like make segments smaller etc but the original algorithm is trivial. If we can't figure things out, we make segments longer and we slow down the trains until we can. I am positive for a sufficiently large track segment and a sufficiently slow train, we can always answer the question with a no.
Hang on, this is nothing at all like that - trains run on fixed tracks. The only bit of data we are interested in for collision avoidance is the linear offset along the track. Even the shittiest GPS fix of all-time would give a good-enough intersection with a fixed path to avoid crashing into a second stationary train in the open country-side (do you see any satellite line-of-sight blockers in the crash footage?)
So that is naive, but assuming that would be all that you need, with GPS you would also need the positions of all other trains. But why would you transmit all this information to every train, instead of just computing the collisions in a central node? (That's what is done now.) A better fit for what you suggest would be radar, and this does exist. [1] But that's where your original assumption is naive. You regularly pass other trains within <2m at high speeds, so linear offset is not all you need.
[1] https://www.dlr.de/en/kn/research-transfer/projects/next-generation-train/rcas-railway-collision-avoidance-systemBecause central nodes can fail (as they did here, both human and machine), and providing this redundant data allows both the computer and the human-in-the-loop on the fast-moving chunk of steel to independently validate the decisions coming from the central node.
> You regularly pass other trains within <2m at high speeds, so linear offset is not all you need.
In this collision, both trains were on the same track. So while we obviously care about adjacency issues in the general case, this specific failure mode can be detected purely via a linear offset.
> In this collision, both trains were on the same track.
Yeah, but how do you distinguish that? So when I understand you correctly, what you want is: Find another moving chunk of metal in the radius of 2km, which moves toward you, but not in a direct or ballistic trajectory, but in a arbitrary complicated trajectory, but which can lookup in a database. Also you are operating in the real world, so there are potentially houses, trees and mountains in between. I guess that would be possible, but it would be really expensive. Your best bet would be to adapt a missile detection and targeting system from the military. A system which does exist and is somewhat similar is RCAS (https://www.dlr.de/en/kn/research-transfer/projects/next-gen...).
Throwing up our hands about how hard a problem this is seems like a ridiculous position to stake. If we can design self-driving cars, that operate safely in far less controlled environments, we can figure out the linear offset of a large vehicle traveling along a physical rail running along a fixed pre-determined path.
> If we can design self-driving cars
We can not design self-driving cars to the safety standard that is required for railways. They are way less safe than the current half-manual system for railways. What you suggest is using less automatic safety systems than what we currently do.
Current operation of a train is basically self-driving and supervising driver. So if you suggest "just use self-driving", you make the system less safe, not more, because it means just getting rid of the human supervisor. This is not currently done, because it would be too insecure. It might be done in the future, but rather because less safety is going to be accepted in favor for getting rid of the costly human.
A self-driving train will happily cause a crash, if the target is "drive with high-speed in that segment currently occupied by another train". The issue is that that becomes a target, not that there is a human in the loop.
And some problems are just _open_ to outside looks. My assistant tells me that the room for error in trains on one dimension ('left - right') is mere 10s of centimeters. I understand that problem. But what track you are on, isn't that obvious to the human in the cabin? (Again, simplifying, not flaming. Production ready safety is _hard_.) So could a vision based solution help count tracks from left to right? (Assistant: "US PTC, on about 54,000 route-miles, uses GPS for the lead locomotive plus crew-entered track information and switch position".) So you use the dimension that matters most in crashes (forwards / backwards) from satellite, augmented with vision or human based left-right positioning. Some trains already have sensors to sense things like leaves on the track to adjust speeding and braking... So the sensor based approach is already there.
Perhaps I am read as suggesting that the problem is easy, and that is not my intention. But for cars this was solved decades ago for routing and years ago for self-driving. I use that daily. Flying is organised with a very clear 6 sigma approach. Crashes are nearly always situations where events / risks stack. Trains (in Spain)...?
Tl;dr - Safety is a function of (human; process; technology) and there is clearly something lacking for trains, when compared to other types of transport.
> Would you mind sharing some more of the why it is still an open problem?
Because for train sensors, there is a small drift which accumulates too much, hence it is synchronized by magnetic position markers in the tracks.
> I understand satellite positioning isn't definitive
Yes, and the point is it might be some solution if we would have the safety systems from 1860. But we do not and hence it needs to be more accurate than the current system, and it is not and in addition also more expensive (and subject to remote influence and reliance on potentially adversary nation states).
> Sorry for the bleak introduction of economics
No, economics is basically the primary constraint for safety. We could pay million people to make sure a single person does not stumble if we want total safety, but that's completely pointless, and also these million people would starve, because no one is producing food anymore.
> But even as a tertiary or lower risk mitigating factor
But for that it's too expensive and still unreliable. There is RCAS[1], which I already linked elsewhere in the thread. That's basically cutting edge and too expensive to introduce everywhere. Also previous safety systems are already not rolled out enough, which is causing these accidents, so rolling out yet another expensive system is not going to improve much.
> But what track you are on, isn't that obvious to the human in the cabin?
Yes, but if it is already known, why use GPS to calculate the same information with more error?
> So could a vision based solution help count tracks from left to right?
If the train driver or dispatcher makes use of vision to assess the train position, the error will be higher than the current sensor/electric-resistance based system.
> Assistant: "US PTC [...]
BTW, that seems to be a verbatim quote from https://www.railwayage.com/cs/the-way-forward/
> US PTC
According to Wikipedia, ASFA being introduced in 1975 is rolled out in the entire Spain network. ETCS which is the European standardization seems to be rolled out on all high-speed lines or currently being rolled out.
> So you use the dimension that matters most in crashes (forwards / backwards) from satellite, augmented with vision or human based left-right positioning.
Duh, that just sounds like trains would be operating blindly in space currently and need to assess their approximate position.
> Perhaps I am read as suggesting that the problem is easy, and that is not my intention.
No, the issue is that you are suggesting far more complicated solutions to already solved problems, with simpler systems in operation for decades. The current system is that nobody in the train even has a say on where the train goes, so the problem you are trying to prevent doesn't even exist. This IS why it's way more safer than a car. There are movement authorities, that define from where you are allowed (and have to) drive with your train. The problem is not, that the train driver isn't able to know where he is and hence drives wrong or too far (that would be prevented by the on-board safety system), but that the movement authority itself is faulty.
On an up-to-date signaling box (withing the past 6 decades), even the train dispatcher is no say in safety questions. You could put a monkey in charge of it. There would be chaos, but they would not cause a crash. The current job of a dispatcher in normal operation is to assign resources to prevent deadlocks and to implement vaguely defined priorities. (Basically the equivalent of the public thinking a programmer is a code monkey, while the actual point is clarifying stakeholder interests.)
The actual issue here is, that while these safety system have a very small false positive rate, they do have a too large false negative rate for there to be an acceptable availability. Hence in the latter case, the human needs to take over and then you are back to a much higher false positive rate. The issue would be alleviated by relaxing the constraints on cost per availability. Either by allowing for less availability or by reintroducing more humans, so that the system situation can be assessed by humans directly.
These accidents are caused as follows: the automatic system fails, because the internal state of the system does not match the actual state of the world. Hence the human needs to take over. But the human also needs to rely on some system to know the world state. Which systems can he trust, given that the whole situation is that the systems report that he can't trust them? So he needs to make some decision and in most cases, that decision is still correct. But there is still an error rate to that, and then something like what we are discussing happens.
> But for cars this was solved decades ago for routing and years ago for self-driving.
This is absolutely not solved for the cars, and to the sense that it is, it was solved and rolled out from the 1870s onward for railways using mechanical computing systems, the switch to using electricity happened in the 1910s.
> Flying is organised with a very clear 6 sigma approach.
You are correct, that aircraft safety systems need to be designed to a higher reliability than railways, because on a rail you can always stop in case of an error, while in the air such a choice doesn't lead to a safe state. But if you compare these safety systems, basically aircraft is operated like railways used to be operated in the 1920s. (My peers actually had a task to compare them and find ideas from one that can be applied to the other. While there is a lot of automation to be taken from rail that can be applied to the aerial industry, there is not really much that can be applied the other way.) It is also operated with the amount of personal than 1920s railways. That's because there is a lot of money in aircraft, while railways are mass-transit and need to be cheap. If you would operate today's railway with the amount of human oversight than in the past, it would be a lot safer than aircraft.
> there is clearly something lacking for trains, when compared to other types of transport.
What are you talking about? Railways are so much safer than cars. https://news.ycombinator.com/item?id=50288518 Yes, there are less accidents in planes than in trains, but this is due to what I already explained.
[1] https://www.dlr.de/en/kn/research-transfer/projects/next-generation-train/rcas-railway-collision-avoidance-systemAlso the fallback procedure is also aided by automatic systems and a sizable number of accidents in the past were caused by that information being faulty.
There will be a so-called "train order" allowing passage of a specific section ignoring the safety system warnings, which has strict validity limits both in time and in the area where it applies, usually denoted by specific hectometric mileposts
When cab signaling is involved (LZB is one of these) and it malfunctions, the entire system normally changes to "driving on sight only" - which indeed limits the train to 40, 30, 20, sometimes - 15 km/h (see https://fr.wikipedia.org/wiki/Marche_%C3%A0_vue) or the line gets switched to lineside signaling - if it is available.
I don't know how you could violate all of those rules - both from the driver's side and from the traffic control side - but it sounds absolutely horrible. Like an override of at least 2 (if not 3) very explicit safety margins.
They were the wrong orders apparently.
I remember a long time ago I was stuck in a train in a field a considerable time because of a similar failure and conductor just wouldn't move unless green
I was expecting the same here, conductor is the first to die so just refuse until the idiots fix their system
What exactly?
A conductor always just gets a movement order. They are not able to assess whether the track is clear, that's the reason why the signalling system exists in the first place. How they get that movement order doesn't necessarily tell what the security steps were to back it.
Of course an attentive conductor can notice and mention there something being weird ("Doesn't the other train normally arrive first?"), but that's a rare lucky occasion, nothing to be relied on. And they just need to swallow the answer ("No, you go first today...") anyway.
> and conductor just wouldn't move unless green
The conductor won't ever move unless they get a movement authority, which most of the time is a green signal.
Being a snowflake here has 2 benefits:
1. Getting company managing tracks to fix the lights (they need to deal with angry customers and pay off damages for delays and more), so there's a direct path from unreliable lights to economic damage.
2. Avoid crashing and killing yourself and passengers on the offchance the person on the radio is a moron
Yes, but not because of the light itself, but due to the safety steps done before. Which should also have happened before the movement order in any other form.
As a train driver, you do not have the information required to distinguish these situations. If drivers simple refuse all movement orders, besides "green lights", running trains will become a rare occasion and the drivers will simply get fired for not doing their job. _The train driver is not the one who decides where a train goes._ For this safety-relevant decision a driver is (roughly speaking) nothing more than an innocent bystander.
You can think of a train driver more of the person who makes sure the automatic engine works correctly and also has an eye, that nothing weird happens. Most of your train driver education is knowing how the train works and how to get a broken train back working fast. In addition you also learn the safety protocols how you get the information in which state the engine should operate in. This is what train signals are, it's not really the equivalent of a green traffic light for car traffic.
It cannot be bypassed, if an operator must go through a red they have to manually reset the e-brake on each car once it passes the signal.
I would actually sit in front now. Sometimes you would be able to still get vital information and being actually able to act. But the safe decision would be no to enter the train.
It sucks but that's how it is over here. It's "my" party the one in power when a catastrophe like this happens? It's not the politicians, it's someone else. Is it the "other" party? Surely the politicians are ultimately responsible.
This reflects on the news media as well.
Salary for ministers is around 80k euro per year. Yes significantly more than the minimum salary but hardly a fortune.
Sadly we'll probably watch a lot of individuals being blamed/scapegoated for systemic issues, but hopefully even Spanish politics can organise a proper investigation after this many deaths in a single year.
I don't see how this would fall on the drivers at all.
As we also know from the aviation industry, controllers are human and therefore will make mistakes. A controller shouldn't issue a wrong order, but a driver also shouldn't accept an order they aren't certain about, and there should be technical systems backing both sides of that up.
It doesn't "fall" on the driver, but it doesn't "fall" solely on the controller either. It's a systemic issue, and many pieces have to fail for a disaster to happen.
The aviation industry operates like the railway industry ~1920 (somewhat by necessity).
> You wouldn't drive your car into a lake just because your satnav tells you to, would you?
You are assuming a train driver is equivalent to a car driver. He's not, the analog would be more a repairmen that you always take with you on journeys.
A train driver has zero insight in what is going on and where other trains are. They do need to trust the orders they get, if you want to question every order you will never start to drive.
That's why it's a union job
Also he's the first person to die on accident so it's that and the safety of everyone else
They do not 'need' to follow others, they just need to refuse unless all safe and green.
If they never start to drive it's the companies problem to deal with angry non-dead customers
Every train network has situations where drivers are authorised to proceed against a red signal. Not usually at full speed though.
At most face unlikely discipline.
Here they would even have valid reason to refuse - order was given for full speed, regulation says low speed ready to stop.
So a "No, i'm not doing that, you're moron" would be justified.
The point of this discussion is that people blame the train driver or say "use self-driving", when that is not going to solve the problem, because that's not where the error is. The error is not there, *because* for rail making errors operating the train is not going to cause an accident (most of the time), which is not the case for cars.
I agree. But the decision whether all switches are locked in the correct position and that the track is clear, is not one that he is supposed to make and he also doesn't have the information for it.
> The controller and the driver both went with full speed despite it being completely against regulation on a red signal.
Going full-speed in the case of a red signal is not necessarily against regulation. Also the information of "red signal" or "you can go" is both a signal from the same source. There is no information for the driver to make a different decision over the dispatcher.
I can't comment about whether the current accident was caused by going against regulation, and more important by whom and against which. We need to wait for the proper safety report, not for a BBC article. If you take the latter at face value, it would mean there are clowns operating, which I am sure, is not actually the case. The real error is always way more subtle, otherwise it would not have happened.
A written movement authority is a signal as well in railway terms. It's more like "I don't trust your second traffic light, that you placed next to the normal one, because the normal one is broken".
Potentially, but the point is that signallers should physically be unable to allow trains to proceed, and drivers should physically not be able to override them. British signalling from the 1800s has physical interlocks; this isn't a matter of manual versus automatic train operation. It is clearly systemic.
The current system is basically that the train is driven both automatically and by a professional attentive driver at the same time. Changing it to just the automatic system is going to reduce safety. Also the real world is way to messy to reach acceptable availability rates with only automatic systems.
[1] https://www.itf-oecd.org/sites/default/files/spain-road-safe...
In this case the “object” was another train. That’s not the “hard” part of the problem. This sounds like an apologist comment.
Are there systems that can warn drivers and prevent collisions without external signaling, like ACAS?
If signaling worked properly then it wouldn't matter whether the train was driven autonomously or not, the "problem" becomes trivial. Establishing and continuously verifying that signaling is working properly is the hard part.
Without signaling you're limited to crawling speed just like humans are when they don't have any signalling .