« countering the "empty buses" myth -- with video! | Main | quote of the week: the portlandia streetcar »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Joseph E

The real advantage of platform doors would be to make it politically easier to automate trains. Of course, Vancouver BC runs automated trains (with no drivers) without using platform doors, and Google is trying to convince everyone to allow self-driving private autos on public streets! But in the short term, if platform doors are what is needed to make trains fully automated, the huge savings on reduced unionized labor costs would make it worth the $0.5 billion dollar cost. And if you value those 55 lives lost per year at just $5 million a piece, again the platform doors would be cost-effective if they could prevent even 50% of the deaths.

Andre Lot

Generating employment should ALWAYS be an afterthought on any discussion of how to design and operate a transportation system.

This being said, I don't agree with the "it is rare, hence let's do nothing" chants heard on the transportation blogosphere.

Using that token, most costly measures taken last 10 years to make air travel safer (in terms of operational/technical/aeronautic safety, without dwelling in the terrorism discussion at all) wouldn't be warranted, since they usually means spending millions of dollars for additional fatality prevented.

I think platform screen doors should be fit on stations, period. I also favor sidewalk waist-high fences and isolation of tramways to speed road/tram/bus traffic up and reduce pedestrian collisions.

Matt D

The other technical problem you rarely hear mentioned is that most older subway systems depend on the piston effect to help ventilate stations and tunnels. Platform doors might require ventilation retrofits to existing stations and tunnels. The systems that have doors are generally newer systems and the doors are installed from the get go.

The much larger issue is that this type of overreaction to captivating but exceedingly rare events seems to be the MO for our government. Hence you have to take off your shoes if you want to get on a plane, but feel free to knock back 4 beers before you get in your car to drive to the airport.


Instead of million-dollar-per-station renovations, the MTA should just spend a little money on a PR campaign. The subway is safe right now, the real problem is that people *think* it's dangerous, some advertising and education is probably the best way to address that. (Plus someone needs to launch a public safety campaign explaining how dangerous cars really are, most people have no idea. Maybe they could team up with whoever has money to throw at that problem...)

Andre Lot

@Matt D: you don't need full-height doors that isolate the tunnel. You can use 7ft high doors or even 4ft high doors that don't seal the tunnel.

Morgan Wick

The problem is that, when the media thinks "dog bites man" isn't news but "man bites dog" is, people start to think men bite dogs far more often than the other way around.


I fully agree that the media exaggerates certain types of accidents, but in this particular case some simple number-crunching reveals that platform doors might actually not be a bad idea:

Stations in New York subway system: 468

Cost of installing platform doors: $1m per station, or $468m total.

Accidental deaths per year: 36

Cost per accidental death prevented: $468/36 =$13m

If annual interest and depreciation costs are 10%, this is equivalent to $1.3m per accidental death prevented.

Valuing a life is notoriously difficult, but the U.S. Office of Management and Budget apparently puts the figure at $7-9m, so according to those estimates it’s a cost effective solution.

Of course, if the funds were to be diverted from other transit projects that would reduce driving and/or prevent deaths even more effectively, the analysis could be quite different.



Alon Levy

What Andre said. Some cities, e.g. Tokyo and Shanghai, retrofit older stations with chest-high platform edge doors rather than full-height platform screen doors. These are cheaper.

That said, New York should consider both options. The piston effect in New York stopped working when platforms were lengthened mid-century; at some stations, it actually causes overheating instead of ventilation. The platforms at many stations are extremely hot, and platform screen doors will make air conditioning platforms feasible.


So a person who rides the subway four times a day is actually at as much risk from subways as she is from cars? I hadn't realized these things might be on the same order of magnitude.

Miles Bader

Half-height platform doors do indeed seem much more practical than full-fat platform enclosures (although they're still expensive, etc), but what Andre actually said was waist-high fences on sidewalks—which is pretty much the most insanely awful idea ever.

Martin Barry

I love that in the Singapore photo there are even markings on the platform to show those boarding where to stand either side of the doors so those getting off have a clear path.


My understanding....limited though it is....is that authorities in North America are much more likely to call a subway death accidental than authorities in other parts of the world. This means the number of suicides is likely higher and the number of accidents less. If you are still worried stand back till the train is in the station and your risk should be almost zero. That said even at a million per station it is not too expensive to install and I think it may provide some crowd management to allow shorter dwell times at busy stations if done correctly.

Tim McCormack

You can't discount suicides if you're asking *why* the union is pushing for slower trains, etc. -- they may be irrelevant for accidental-death stats, but the train drivers themselves suffer just as much whether a person falls, is pushed, or jumps in front of a train. The trauma from that can last a lifetime.

Robert Wightman

I have ridden on several systems with Platform Edge Doors, PEDs,and on each system there seemed to be an extra 5 to 10 seconds required per stop because of the necessity for a precise stop location. On manually operated trains the motor person would stop about 3 to 4 m short then creep up to the proper stop.

ATO is supposed to be able to stop a train within 0.5 m every time but the new ones I have used seem to have a 1 to 2 second delay after stopping before the doors open. This seems to also be a problem on the new TTC open gangway Rockets.

I wonder about the actual savings from going to uncrewed trains. In most systems, except for the larger ones, the number of train crew compared to the number of employees in the stations and control centres is relatively small.

The other advantage in having crew is that there is someone there in case of an emergency. If you get a train stuck in a 2 km long tunnel because of a power failure then the crew can tell them what to do. If you are stuck on an elevated guide it is not as frightening to some people as being stuck in a tunnel.


According to the FTA's draft New Starts policy guidance, a fatality costs $6.2 million. So 36 deaths a year is $223.2 million. Multiply that over the service life of a platform screen door, maybe 30 years? Now we're at $6.6 billion in accidental death costs, and we haven't even included injury costs. That's a lot of money available for platform screen doors if you ask me. Seems to me that they would be well worth the cost.

Miles Bader

@Robert Wightman

I have ridden on several systems with Platform Edge Doors, PEDs,and on each system there seemed to be an extra 5 to 10 seconds required per stop because of the necessity for a precise stop location. On manually operated trains the motor person would stop about 3 to 4 m short then creep up to the proper stop.

Maybe that happens in some places, but it's hardly inherent.

In places like Japan, they perform very precise stops, manually, with no perceptible delay or slowdown. [In Japan, they always do so, not just on lines with platform barriers, because people line up at marked door positions on the platforms.]

Computer assistance can make it easier, of course, but all it takes is training and practice to do the same thing manually.


Alon, Andre,

"waist-high fences on sidewalk... is pretty much the most insanely awful idea ever.

I'm with Alon on that, for four reasons:

1. People can't completely keep off the road, because often they need to cross it. At that time, they are empowered to make a choice: Cross right away if it seems safe, or divert to a crosswalk if they don't think they'll be safe. People should be free to make that choice.

2. People drive as fast as the road seems to be designed for. Drivers will move very quickly on a completely segregated road with no pedestrians in sight. That likely means that at the crosswalk, drivers are skipping red lights, or making turns without slowing down and looking out (if the crosswalk is signalised at all), so crossing the road is actually more dangerous.

3. In London, similar fences often exist near crosswalks so pedestrians are forced to use them. There is a grave problem of cyclists being killed as a turning vehicle crushes them against the fences. Safety advocates in London are screaming for them be taken away, not put in.

4. What of deliveries to commercial premises? On a busy urban street where pedestrian traffic is significant enough to merit railings, ever part of the street might need to accomodate a van unloading at some point in the day. Cyclists also need to get off their bikes at the appropriate place on the road.

Now let's never speak of this again.

Miles Bader


"waist-high fences on sidewalk... is pretty much the most insanely awful idea ever."

I'm with Alon on that, for four reasons...

I think you meant "I'm with Miles on that ..." ><


You can try to enter the system with a Metrocard and all of a sudden someone is in your face exiting. This is especially a problem for reverse commuters like myself.

Andre Lot

@Robert Wightman: automated trains are always connected to a central operation center overseeing the whole system. On the rare instances something should happen, they can communicate remotely via radio with the train and give instructions.

That is no justification to keep a unionized well paid worker (or two, as it is the case often on MTA services) on a train to give instructions on the eventual train stoppage in the middle of a tunnel.

You can also put surveillance cameras on all cars at a relatively low cost, and fit all cars with panic buttons for emergencies on the cars.

The fact is that any automation on transit will attract some preemptive unwarranted criticism of powerful unions and certain groups of users that veer to "fandom" status instead of merely users.

Just take a look of removal of staffed ticket booths in NYC and the hysteria that accompanied it, despite ticket machines being reliable, effective and much cheaper. Straw man argument flew around ("OMG what about an old lady that doesn't know how to use computers", "what if someone becomes ill on a station hall", "I know from my experience staffed booths operate faster than ticket machine lines") but the transition to automated machines has been mostly uneventful.

Matter of fact is that unmanned train operation is awesome, it removes most employee costs from operations and allow higher frequencies even on less trafficked hours.

Jack Horner

People's perceptions of risk are notoriously irrational. Are you more afraid of being eaten by bear or choking on a fish bone?

For another transit example: in Newcastle Australia there is a push to close the local train service which runs a block away from the main street. Part of the argument is that the line has several level crossings which delay traffic and are supposedly dangerous for pedestrians.

So a gated, light-controlled pedestrian crossing of a fenced train line that sees a train each 15 minutes is dangerous - but cars and buses flying past inches from shoppers on the footpath of the main street is just fine.

Yet a bus will kill you just as dead as a train if it hits you at 50kph.

I think part of it is that cars and buses are seen as normal, but trains are big scary monsters.


If doors on the subway reduce service disruptions, that could be a major source of savings that hasn't been considered yet-- savings to the agency having to deal with the results (e.g. running replacement buses), savings to the people who aren't delayed, and even greater ridership resulting from improved reliability.


The TWU would have an entirely legitimate and defensible reason to support serious investments in subway safety: their workers and conductors bear disproportionate health risks associated with subway travel (track workers because they can sometimes get hit with trains, operators because the experience of hitting and killing someone with a train is incredibly damaging and traumatic for the operator). To be fair, these things don't happen often. They happen so infrequently, compared to other transpo casualties, that it would be very wrong to impose onerous costs and delays on subway travel to avoid them--slower subways => more cars => more car deaths. But the TWU is doing the right thing in speaking up for the legit health and safety concerns of their membership. We should dump their recommendations, but it's not as if TWU's being unreasonable.

Rational Plan

Re the safety fence issue. In London they have been removing the fences, the result has been that the decline in pedestrian deaths has stopped.

After some digging this was down to an increase in accidents at junctions where railings had been removed. I think the average was an increase of 7%.

Recently there has been call to re-examine the data and determine which railings can be safely removed.

Pete Brown

London has automated train operation (ATO) on the Victoria Line (since opening in the 1960s), the Jubilee Line, and the Central Line. In all cases an operator sits in the cab to operate the doors at stations, and perform reversing moves at termini. The Docklands Light Railway (DLR) is also ATO, but the cars have no cabs. A train attendant operates the doors and gives the starting signal from a control panel above the passenger doors. The attendant also performs revenue protection duties and customer care.

The Jubilee Line has PEDs at all underground stations on the section Westminster to Stratford. These were justified for their air management qualities rather than on safety grounds. On the deep level tube lines the trains really do act like pistons in the small circular tunnels. Wind speeds in cross passageways at stations can be such that one must really lean into the wind if walking against the flow.

Whilst full automation is appropriate for new build metros, I find it a shame that people think its OK to take away gainful employment from existing train crew.


Rational Plan,

I would suppose that that 7% increase in collisions principally involved people that had chosen to take a risk by not using the crosswalk. That, to me, is significantly less disturbing than collisions involving cyclists that were following the rules of the road, typically using the nearside cycle lane and being struck by vehicles making nearside turns.

In other words, there is a difference between achieving safety by restricting people's ability to take their own calculated risks, and by reducing the risks an external agent poses to innocent others.

In the case of the New York subway, Jarrett is probably correct to point out that it's very easy for even an intoxicated individual to keep behind the yellow line. If anything, a case for improved safety measures lies, to me, in preventing the traumatic impact upon train operators of suicides. Platform-edge doors are likely to achieve that more effectively than slow trains and more staff.

Alon Levy

If the TWU pushed for PEDs/PSDs and explained that the issue is important to it because of the psychological damage caused to train drivers, I don't think there would be this much criticism. Instead, the TWU is making demagogic posters with blood splatters and demanding that the MTA slow down the trains. Instead of asking to install newer technology, the union is asking to worsen service for passengers.

Bust the bus

Trains? These killer jack knife buses should not be allowed on streets that ban thru trucking. They can't stop quickly enough. When they are late, they are really very late causing old folk to lose control and stink up. And they take forever to load up, opening us to silly schemes where they ride for free without fareboxes in the back. We need smaller buses like scalpels instead of hatchet jobs.


Platform dynamic doors, which do work with all permissible stopping positions, would actually give drivers a better sense of security. Tunnel access needs to be restricted at every possible opportunity.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name and email address are required. Email address will not be displayed with the comment.)

the firm

Jarrett is now in ...

Related Posts Plugin for WordPress, Blogger...
Related Posts Plugin for WordPress, Blogger...