The computer is programmed to stop in any circumstance, except when a crane falls on it.
The system requires that an obstacle be in physical contact with the trackbed in order to trigger an emergency stop (in addition to sensors on the train itself that also focus on the track ahead). The crane boom was not in contact with the trackbed or lineside equipment as it was suspended from the noise barriers. Whether this is a true edge-case, or something that should have been forseeable and mitigated is, I suppose, up for discussion, but itās worth noting that all of the signalling, control and safety systems are made by Alstom and are in use elsewhere, not just on the Taichung MRT.
Both the station manager and train attendant were yelling into their radios for the control centre to shut the system down, according to witnesses on the train and the platform. I presume the control centre has a big red button for that purpose, but Iāve not seen any reporting on what the expectation in such a scenario would be.
Wouldnāt there be sensors to detect anything blocking the track in front of the train?
Iād say this is a system failure.
They say its because of āchabuduoā personality.
I wonder whether it might have been a similar situation (or rather the inverse situation) to that power plant issue last year, where the big red button had a ādo not operateā sign and nobody was sure whether they should push it. 
Personally, I always try to avoid riding at the front of the train/MRT and stay close to the back, and Iāve been doing that for years. While obviously rare, accidents like this every so often remind me that I donāt want to be near the front if anything happens. 
Well, again, like I said, thatās up for discussion. The detection system appears to have been built on the premise that an obstruction would necessarily be resting on at least part of the trackbed. Is that assumption a safe one? Clearly, in this scenario, it was not; but itās also a bit of a freak event, in that an object long enough to span the noise barriers fell with sufficient force to partially crush them, but not enough force to quite contact the tracks. I imagine, as well, that there are other considerations in designing systems like this such as not bringing the entire system to a juddering halt on detecting birds flying across the tracks, for example.
The bigger issue, to my mind, was that staff who were well aware of the obstacle before the train even departed did not have the means to bring the train to a stop at will. The fact that the only means aboard to instantly stop the train is kept under lock and key suggests that the SOP is to radio the control centre in the first instance (which is in fact what the staff immediately did). The lack of a timely response from the control centre is probably therefore of greater concern than the specifics of the obstacle detections system.
Unless you were on that ICE train where a loose part took out a bridgeā¦
I reckon you could well be right. Hesitation due to uncertainty about what to do, or not wanting to f*ck up and get fired. I hope it gets properly and fairly looked into - thereās some reason to be optimistic on that front, as Taichung MRT have been fairly impressively forthcoming and transparent thus far. In fact, they have gone out of their way to explain and justify the actions of the train crew and station staff, which is refreshing. There is often an unfortunate tendency to blame the frontline staff after an accident here, and then to look no higher than that. Taichung MRT deserves some credit for that, at least.
I wasnāt.
I donāt doubt there are exceptions, but I still strongly suspect most train crashes/accidents start at the front of the train.
Actually, rear facing seats are going to be a lot safer than front facing seat. So keep that in mindā¦
THIS!
Anyone see that bonkers accident in Taichung?
You mean the one being discussed in the previous posts right before your post?
I hardly use public transport, because I hate the general public, so I donāt read this thread.
I would think it depends on what you consider a timely response to be. From the crane being recognised by someone as being over the line the accident happening was something like < 15 seconds. I donāt know about you but thatās certainly not enough time for me to contact someone, then get them to understand the crazy message I am passing on āA crane on an elevated train track??ā Seriously?
As to people saying that the track detectors are inadequate. Yes they are but also consider this, you have X amount of money to spent on the entire system, Y amount of this can go on safety features. Your train track is evelated up out of the way of 99.9% of issues. You put side guards on the track so the only access to the track is from the top. This elimates 90% of the rest of the issues that might occur. You also invest in a track blockage system that detects items lying on the rails themselves.
After all that pretty much the only thing that can happen now is something has the reach to go over the track and then fall on it i.e. a crane or a plane/helicopter does.
At what point do you stop spending money on safety? Some would say you never do but then if that is the case youād never run an MRT.
I read that it takes 20 seconds for the person in the control center to stop the train. The accident happened about 40 seconds after the debris fell onto the tracks.
Add to that the time it needed for the platform attendant to relay the information to the train attendant who then had to relay it to the control centerā¦ thereās no way that train would have been stopped in time.
The main issue is the lack of any way to pull the emergency brakes from the train itself.
Agree with this - although it does seem there were big red āstopā buttons that could have been pressed - it just that they werenāt.
My understanding is that the red handles bring the train to a halt in the next safe place, usually a station, and are not emergency brakes per se.