A man was dragged to death by a New York City subway train when a bag he was carrying got stuck in the door during the Monday evening rush hour commute, according to police.
Yaw Boahene, 57, had been exiting the train at a Brooklyn subway station around 6 p.m. when his bag, which police described as a shopping bag, snagged in the closing train doors. The Brooklyn resident was pulled onto the tracks and run over by the train as it left the station, police said in a statement.


Excellent points, thank you for explaining
They’re not excellent points, they overstate tradeoffs into a reason not to use platform doors.
That’s just straight up misleading. Most metro infrastructure was built decades before platform doors became practical. Looking at modern systems, it’s a completely different story. The other comment that just says “Singapore” is a great example, they retrofitted platform doors on all their stations.
Now the excellent points:
Sure, it’s an added cost. It also reduces other costs, like the cost of injuries and fatalities, debris related shutdown, cleaning, and there’s also energy benefits when you’re huge stations are air conditioned. But even if they didn’t actually save more than they cost, even if they saved nothing, relative to the whole system cost they wouldn’t make transit “much more expensive”.
All the other points are similarly overstated or incorrect, look into them if you care. Especially the safety one, these barriers are designed with safety standards and they can be pushed from train side with small force, in an emergency even if the train doesn’t align with the gate, you can push the whole barrier down with ease.
Our modern rail system doesn’t shut down when some jackass throws something on the tracks, because they can’t. If a gate malfunctions they rope it off and it takes an extra minute on a busy day to unload/load from one less gate. We have a clean station/tracks, AC in the station, and it’s certainly cheaper than NY subway.
The first platform doors appeared in the 80s, that’s almost half a century ago. Also, there are enough examples where platform doors were retrofitted to existing subway lines. Maybe don’t accuse me of being misleading when you’re writing stuff like that.
Oh, and I never said it makes public transport much more expensive, I wrote it makes it that much more expensive. Two very different statements. Funny you mix them up after accusing me of misleading people. But accusations that are actually admissions, I guess. Never mind that you’re simply looking at the costs of the does themselves, and not the costs of the inflexibility you being into your system by installing these, among others.
It’s still a barrier, which is exactly the opposite of what you want in an emergency. And it can’t be that easy to push in since if it were, the air pressure from any train going through the station would pop them right off. But even assuming it is that easy to push off, it’s still an extra obstacle in the way, which not only risks slowing things down but also causing a panic.
Someone throwing something on the tracks isn’t all too big of an issue though. It’s something that doesn’t happen all too often and is solved fairly quickly. A fairly common cause for delays? Problems with doors. Adding platform doors is roughly doubling the possibility of door problems at every stop.
And interesting you mention the commenter that commented Singapore. Because that person linked to a list of incidents there. And I don’t know if the list is filtered or anything, but it is pretty interesting that over half the incidents listed were either exacerbated or caused by platform doors, including a case where someone got stuck between the train doors and the platform doors and miraculously didn’t get killed when the train left the station.
Appeared != Practical. You’re being misleading again.
I’m not doing this. That’s just insultingly stupid.
It’s exactly what I want when I’m waiting for my train and some psychopath who can’t take criticism decides to shove my on the track. Do you want to lookup the number of fatalities due to barriers vs not having them, or should I do that for you? And when you find the two fatalities ever recorded with barriers, do you want to lookup the updated safety standards on barriers made after to see how they’ve been addressed?
That doesn’t follow. Adding a second set of doors does not “roughly double” door-related failures unless the platform doors fail at roughly the same rate, with roughly the same operational consequences, as train doors. There’s no reason to assume either, and actual operating data says otherwise.
Singapore, for example, reported only 10 platform-screen-door faults that caused delays longer than five minutes across the entire MRT system from 2015 through 2017, and then zero such incidents for the following 19 months. So platform doors clearly aren’t just generating a second copy of the train-door failure rate.
They’re also easier to degrade gracefully. A failed platform door can generally be isolated and kept closed while trains continue operating; that’s explicitly built into operating procedures for systems using them. I already gave an anecdote about my experience, but feel free to look up procedures at your least favorite life saving station.
And “someone throwing something on the tracks” is a pretty narrow description of what the barrier prevents. Track intrusions include people falling, entering intentionally, suicides, objects, and other incidents that can require power isolation, inspection, emergency response, or police involvement.
So yes, platform doors introduce another maintainable component. That’s a legitimate cost. But “twice as many doors = twice as many door delays” is not reliability analysis. If train doors cause 100 disruptions and platform doors cause 5, adding them gives you 105, not 200. And you then have to subtract the track intrusion disruptions the barriers prevent. The relevant question is the net effect on reliability and safety, not the raw number of moving parts.
That page isn’t a list of Singapore rail incidents. It’s five selected TSIB investigation reports from 2024-2025. Singapore’s railway-occurrence investigation regulations only took effect in 2024. Singapore actually has historical SMRT data from before the elevated stations received platform doors. In 2004-2008 there were 102 track intrusions, including 20 cases where someone was on the track with an approaching train. During the 2009-2012 platform door rollout the numbers collapsed, and by September 2012 (after the retrofit was completed) SMRT recorded zero “one-under” incidents and only one unauthorized track-access incident. The Singapore court reviewing those data explicitly noted that one-under incidents fell to zero after the retrofit. So taking 3 out of 5 specially investigated modern incidents and treating that as an accident rate is not a meaningful comparison.
Of course when you look at a report for incidents in a platform door system, you’ll have incidents from platform doors. That’s just selection bias.
https://www.elitigation.sg/gd/gd/2013_SGHC_286/pdf https://laws.sg/court-decisions/court-decision-sghc-2013-286
It’s funny you’re so upset about a life saving system because it may make your ticket more expensive or significantly delay your travel, despite all available data contradicting that. Yes, your points were misleading, perhaps unintentionally, perhaps you’re just repeating talking points from operators that would rather you die than spend any money on additional safety systems, but the data contradicts your argument regardless of how insulted you feel about that being pointed out.