Accident at Alton Towers

[Chris Brown]

I just logged a complaint through ofcom. Was really quick to do and only takes a few complaints for the subject to be reviewed. It only takes a few minutes and is more damaging than signing a petition. Here's the text from my complaint. Was done in a rush on mobile:

Kay Burley used unprofessional, rude and bully like tactics whilst interviewing Nick Varney representing alton towers in light of the rollercoaster accident. She showed no empathy and continued to dismay his points in order to get a reaction. He was constantly interrupted and was not allowed to explain or justify his actions without being interrupted or bombarded with repeat questions which had been answered. To see an interview conducted in this manner was offensive and blood boiling. Kay Burely cannot be allowed to interview with these tactics as she was clearly extremely biased even before the interview process had begun. She tried to pressure out sensitive and private information regarding hospital patients involved. This was simply wrong and immoral.

 

[Howie]

^If this post had a 'like' button, I'd click it.
Might even do as you suggest and lodge a complaint, I really was appalled at the way she conducted that interview.

 

[scw55]

I don't think she should be sacked.
But I think she ought to be disciplined.

If you see her Twitter account (please for the love of God don't abuse her), you'll see her replies to people expressing their disgust at her interview manner are very bitchy. If you didn't know that she was a Newsreader, you'd think she was a troll. I kid yet not.
Corrects people's spelling or grammar without addressing the point.
If she addresses the point (when the Tweet on the INTERNET is free from errors), she plays the #perspective card. (There are homosexual people in countries around the world who are being killed all the time because of the way they were born. If not slain, imprisoned. If not imprisoned, actively abused. If you're going to use a #perspective card, someone potentially losing a limb is minor in comparison to these immense atrocities).

She's in a powerful position to affect how other people think. She needs to take more responsibility with this.

 

[mcg0ughchris]

We're getting a bit off topic here. Maybe we should open up a news reader hate thread instead

 

[MouseAT]

I've talked with a few of my friends at Maurer and Zierer and all three confirm that it's not possible to send a train into an occupied block even in manual mode. So unless Gerstlauer is doing things much differently than either of those two companies, that all build similar coasters, overriding the system that knows there's a stalled train in the block simply isn't possible.
BTW - All three are engineers that do final programming and testing before handing new rides over to parks.

I don't know whether Jerry is still watching this thread closely, but if so, would you be able to ask (and would they be willing to clarify) what would happen if a ride is reset by the engineers, and a train was part way through a block, between sensors? Would the previous block state be retained, or would the ride reset cause the previous block state to be discarded?

I cant have been the only programmer thinking RACE CONDITION on Tuesday. That would be quite serious if found to be the case. Race conditions kill people.

I had considered the possibility of something like this, or some sort of overflow, bounds check or wrap-around problem somewhere in the code. It's possible that we're looking at an Ariane 5 incident all over again.

I still reckon it's more likely that it's something to do with the ride reset process than a straight software bug though. Unless there are some clever safeguards built in to the system to track train locations when starting up from cold, I can (with the benefit of hindsight) see a potential failure state that has been pointed out by several people already in this thread.

 

[SaiyanHajime]

I've talked with a few of my friends at Maurer and Zierer and all three confirm that it's not possible to send a train into an occupied block even in manual mode.

I'm no tech expert, but, with Maurer this is either not true or definitely depends on the blocks in question. A standard mouse, for example, has 4 blocks in the station and you can stack 5 cars bumper to bumper across those 4 blocks. That is standard practise for storing cars. Can we find out how this differs?

It's also possible if a car slips through a block if it's wet, for example. Though the system should automatically estop in the event this happens and have fired numerous "too fast" warnings, from what I understand. But minor "crashes" (more like gentle taps less intense than a bumper car) on multi-car coasters are fairly common, and lots of parks close these rides in the rain for this exact reason.

Obviously this has nothing to do with the Smiler incident, because these things I'm describing are related to when cars are stacking on final blocks or in station blocks - not mid circuit in lift and coasting sections of track.

I'm really interested in finding out what happened on this ride. I cant have been the only programmer thinking RACE CONDITION on Tuesday. That would be quite serious if found to be the case. Race conditions kill people.

And I was thinking even more so, since the interviews with them saying perhaps there was "a unique sequence of events".

Issues with operating software, human input... or both? Is the operating software on rollercoasters properly reviewed and tested?

Hardware failure seems a bit unlikely IMHO.

I am assuming the software has sufficient auditing and logging et cetera... but think i'd feel a bit uncomfortable going to parks again if it transpired it was easy for operators to simply over-ride software safety systems.

I'm really interested that non enthusiasts are, in light of this incident, coming to make accounts on here to talk coaster safety! Welcome!

Rides do have manual override modes - they need to for technicians to be able to do their tests, but also because the systems are designed to be over the top to avoid any potential for accidents. The slightest thing and a ride will estop itself.

A strange series of events led to the Smiler incident, a series unique to this incident.

Rides are safe - or rather, they're as safe as it is reasonable to expect anything to be - especially in large established parks with established, over the top procedures like Alton Towers. But that doesn't mean there is no risk and that things wont, occasionally, go wrong. Getting out of bed in the morning is a danger and so is driving a car to a theme park and so is walking to a ride and so is standing in the queue and so is riding the ride. When you consider how many millions of guests visit theme parks and ride coasters the world over every day, it's remarkable that so few accidents occur. I bet somewhere in the UK, people died in a traffic accident on the same date Smiler crashed, but due to the rarity of theme park accidents and the lack of understanding from the public, the media jump on it.

 

[SpeedyWorm]

Something that has not yet been talked about is why the original train valleyed in the first place. There would have been no crash in the first place without the valleying taking place. I am well aware that valleying in itself is hardly abnormal, that some coasters are more susceptible to it than others and that variety of reasons such as extreme weather can be cause it. But rural Staffordshire is hardly the Arizona desert though and unlike Jubilee Odyssey and PMBO, Smiler is not on the sea front being battered by high winds. So why does it keep stalling at the Batwing element? I can not think of one modern multi-inversion coaster, that is way past the honeymoon period and has been open to the public for longer than its first season and has stalled more than once or lost momentum to the extent where it hasn't been able to complete the circuit. I can not think of one instance, not one that has no instantly explainable reasons anyway.

Do not shoot me down but a question entered my mind that i genuinely think is legitimate, I am not in any way suggesting this is what is happening merely asking if others think its plausible.

The section of track which ''seperated'' in 2013 after the bolt incident is in the first dive loop-between the first drop and the batwing-the batwing being where the valleying has occurred and on more than one occasion. Is it at all remotely possible that an actual flaw in the construction or even something fundamentally wrong in the design stage of the ride itself are responsible for the mechanical (in classical sense) problems Smiler has experienced?

I'm not throwing muck at the walls and seeing what sticks, its just that with all the emphasis on what went wrong with the electronic and human side of things there is an aspect that hasn't been mentioned yet.

 

[coaster2015]

A newbie here and have been reading with great interest. I heard today on the grapevine that they have now closed Karacho in Germany too today... Does this mean that it's more a technical
Fault than human error? X

 

[SaiyanHajime]

Speedyworm - It was quite windy here in the south on Tuesday - anyone live up in Staffs area confirm it was windy also?

Smiler has stalled in that same spot before. It is a complexed track and isn't a particularly fast ride. The stalling shouldn't happen because it's an operational nightmare to retrieve train, but it's not a hazard in itself. Colossus at Thorpe Park is also away from extreme weather, but has stalled a few times, and that is a comparable ride in terms of being a lot of potential to lose momentum in inversions - the heaters under the station at Colossus that warm the wheels are there to reduce the chance of it getting stuck.

coaster2015 - Welcome! Not necessarily - could be a review of procedure and that could mean anything. Might be implementing new emergency training for staff, for example. Or new checks that must be made before dispatching trains. Implementation of new CCTV. Or a coincidental closure related to something else. Could be anything.

 

[coaster2015]

coaster2015 - Welcome! Not necessarily - could be a review of procedure and that could mean anything. Might be implementing new emergency training for staff, for example. Or new checks that must be made before dispatching trains. Implementation of new CCTV. Or a coincidental closure related to something else. Could be anything.

Thanks for the response, yes I guess it's better for them to review all similar rides/safety etc.

I shall keep following this thread with interest, it's nice to read facts and ideas rather than the sensationalist news stories that keep appearing!

 

[SpeedyWorm]

Speedyworm - It was quite windy here in the south on Tuesday - anyone live up in Staffs area confirm it was windy also?

Smiler has stalled in that same spot before. It is a complexed track and isn't a particularly fast ride. The stalling shouldn't happen because it's an operational nightmare to retrieve train, but it's not a hazard in itself. Colossus at Thorpe Park is also away from extreme weather, but has stalled a few times, and that is a comparable ride in terms of being a lot of potential to lose momentum in inversions - the heaters under the station at Colossus that warm the wheels are there to reduce the chance of it getting stuck.

Colossus stalled on its 9th inversion though- right at the very end of its circuit when its at its slowest speed and you could appreciate that the most minute of factors may affect its momentum, such as the heat of the wheels or the track. If it stalled in its cobra roll or in the corkscrew on more than one occasion, now that would be a bit different..

Smiler valleyed when virtually traveling at its top speed not long having exited the first drop- 3 elements in to a (albeit over 2 legs) 14 inversion track. Surely the only thing that would cause this bar extreme winds that they surely would not be running a roller coaster in is something like an axis or wheel coming lose? It seems to me to be very mitigating circumstances despite valleying being common.

 

[MouseAT]

Something that has not yet been talked about is why the original train valleyed in the first place. There would have been no crash in the first place without the valleying taking place. I am well aware that valleying in itself is hardly abnormal, that some coasters are more susceptible to it than others and that variety of reasons such as extreme weather can be cause it.

I'm not throwing muck at the walls and seeing what sticks, its just that with all the emphasis on what went wrong with the electronic and human side of things there is an aspect that hasn't been mentioned yet.

The thing is, whilst you can (and probably should) try to take action to reduce the risk of a train valleying, there's always going to be a risk of it happening, even if it's only under really extreme circumstances. The ride system has to be able to handle a valleyed train safely, 100% of the time. I think that's why avoiding valleying hasn't been discussed much yet, as it's a separate (albeit related) issue. The valleying problem is a convenience, maintenance and financial problem, whereas the block system not working as designed is a major safety one.

That said, if they're not going to modify the track layout in light of this, they should probably think about installing a maintenance/evacuation platform at the bottom of the batwing if there's room to do so, in case a loaded train ever gets stuck there again.

 

[nemesis_guy]

Something that has not yet been talked about is why the original train valleyed in the first place. There would have been no crash in the first place without the valleying taking place. I am well aware that valleying in itself is hardly abnormal, that some coasters are more susceptible to it than others and that variety of reasons such as extreme weather can be cause it. But rural Staffordshire is hardly the Arizona desert though and unlike Jubilee Odyssey and PMBO, Smiler is not on the sea front being battered by high winds. So why does it keep stalling at the Batwing element? I can not think of one modern multi-inversion coaster, that is way past the honeymoon period and has been open to the public for longer than its first season and has stalled more than once or lost momentum to the extent where it hasn't been able to complete the circuit. I can not think of one instance, not one that has no instantly explainable reasons anyway.

Do not shoot me down but a question entered my mind that i genuinely think is legitimate, I am not in any way suggesting this is what is happening merely asking if others think its plausible.

The section of track which ''seperated'' in 2013 after the bolt incident is in the first dive loop-between the first drop and the batwing-the batwing being where the valleying has occurred and on more than one occasion. Is it at all remotely possible that an actual flaw in the construction or even something fundamentally wrong in the design stage of the ride itself are responsible for the mechanical (in classical sense) problems Smiler has experienced?

I'm not throwing muck at the walls and seeing what sticks, its just that with all the emphasis on what went wrong with the electronic and human side of things there is an aspect that hasn't been mentioned yet.

Looking back at a previous stall then according to this:
http://old.towerstimes.co.uk/news/news. ... 2C10%2C117

in this instance weighted dummies were being used. It is believed that the rides computer system caused the trim brakes on the first airtime hill to be applied unnecessarily, causing the vehicle to come to a halt - See more at: http://old.towerstimes.co.uk/news/news. ... 563dm.dpuf

Could there have been another error like that?

 

[SpeedyWorm]

If it was the trims kicking in on the air time hill -and it seems pretty likely considering they are directly before the Batwing itself- that caused the stall, then that would require another computer error on top of the other subsequent ones that lead to the crash itself. If sending out empty or weighted trains in the past has caused the system to behave erratically and differently then it does when you have actually have riders on board then what would even be the point of conducting test runs?

 

[SaiyanHajime]

I doubt the empty train was traveling fast enough to engage the trim breaks.

Smiler valleyed when virtually traveling at its top speed not long having exited the first drop- 3 elements in to a (albeit over 2 legs) 14 inversion track. Surely the only thing that would cause this bar extreme winds that they surely would not be running a roller coaster in is something like an axis or wheel coming lose? It seems to me to be very mitigating circumstances despite valleying being common.

Batwings are an uncommon inversion. Like pretzel loops, there is a fine line between too slow to make it round and too fast, making uncomfortable forces. (That's why there are trims right before it.) Unlike a cobra roll, where the train enters at speed, slows down as it reaches the top and then speeds up upon exit, a batwing enters slow, speeds up, then slows down on exit. It's that slowing down on exit where the potential for Smiler to stall occurs, causing it to roll back down into the bowl of the inversion. And again, it has stalled here before, so it's not that weird. It's the least weird part of the puzzle, which is why no one was discussing it.

 

[Slayed]

Interesting quote from Adventure Island:-

"It is impossible for two cars on Rage to collide as the independent dual computer system prevents a car starting its ride until the previous car is back in the station.

Our first action on hearing of the tragic incident at Alton Towers was to immediately contact Gerstlauer in order to take their advice. They assured us that Rage is perfectly safe and there is absolutely no reason at all why it should not continue to operate as normal."


Obviously Smiler (and Saw) are more complicated rides than Rage, with multiple blocks outside the station confines, but this proves that even a standard Eurofighter goes as far as having redundant control systems in case one fails. I'm pretty confident their error detection will be similarly robust, and I still think the cause is unlikely to be a control system fault.

newfoundglory, as regards your race condition theory, it's an interesting idea but my instinct says probably not - although the PLCs out on track will be communicating asynchronously with the controller, this is typical of industrial control systems and it will have been designed and built to handle concurrent messaging from the outset.

Despite this being a grim topic I'm loving this discussion, well done everyone!

 

[Jake]

Won't be anything to do with the trims - due to the height restriction, and cramming in so many inversions, the cred comes quite close to stalling. Especially when empty.

Stumbled across this on youtube - I know it's from when the cred was new and testing, but it comes nail-bitingly close to stalling several times - particularly on the same inversion it did stall on before the accident. Skip to 1:58.

https://www.youtube.com/watch?v=NLw1Mpx-KaE

With no weight in the trains I'm surprised it hasn't stalled more often during morning testing etc.

 

[Smithy]

Wasn't there talk that they'd just put that train on before it stalled too? Could have been a mixture of empty train, cold wheels and wind that caused it to stall.

Oh and Kay Burley blocked me on Twitter ages ago the grief peddling zombie faced ****.

 

[Slayed]

So the older girl on the front row "is still critically ill although surgeons have managed to save both her crushed legs," which could be where the "lost leg" rumour started originally. Sounds like good news anyway.

Also looks like the press are finally getting more accurate, and are starting to get closer to our discussions: http://www.mirror.co.uk/news/uk-news/alton-towers-accident-three-trains-5836451

I'd have liked more detail on what exactly these "proximity sensors" are, as the article makes it sound like they track the trains around.

 

[Intricks]

Holy hell does it seriously come close to not making it during those inversions. Seriously, I dont know how powerful your gusts of wind can be, but if it were to creep that slowly and have even a 25mph gust at the right moment, I can easilly see it stalling.

Raptor at Cedar Point has only ever stalled once, but now dont they take precautions in winds that are in excess of a certain speed?