- The train doors need to line up closely with the glass doors, for obvious reasons. Commuter trains in the US are not capable of stopping so precisely. Usually the trains would have to support some kind of computerized stop-assist in order to be precise enough to pull this off.
- Said stop-assist wouldn't just be a train feature, it would involve communication with the track, and therefore upgrades to the entire signaling system. Many systems have this "for free" by virtue of being built later. The actual track Caltrain runs on pre-dates Caltrain itself by a wide margin.
- You need to run a unified, homogeneous vehicle fleet so the doors are in the same places on all trains. This is the problem with safety doors in large legacy train systems (see: NYC, Chicago) where due to legacy factors there are many vastly different trains all running on the same track.
A lot of non-trivial problems where the only solution is vast capital spending.
For some greater detail: AAR regulations require that trains that run on lines which include freight traffic meet certain weight and crash standards. US freight cars have the highest standard axle loads in the world, so any train that runs where it can hit one must be able to withstand a head-on collision. The sole exception to the rule is when all freight traffic runs at a particular time when no passenger trains are running, with derails (a safeguard) in place to keep any freight from making it on to the tracks outside the window.
Caltrain runs on the old Southern Pacific line, which includes several on-line customers and a few branch lines. Due to the way the line is made, there is no way to have the required separation without significant upgrades.
Finally, stop-assist is now doable via GPS signals and other lineside equipment, removing the need for signalling upgrades. However, it also requires either lighter equipment or disc brakes on all cars along with equipment in the cab. Both are quite costly, and always require a homogenous fleet with the exact same door configuration.
This is very cool. Thanks for the TIL!
It's not as hard as you make it out to be. I live in Japan where these (very effective) safety doors have been rolling out for more than a decade.
I don't think there is any need for computerized stop-assist. On at least some of the trains here, and I think most, it is up to the driver top stop the train. Once in a great while, they accidentally overshoot by a foot or two, apologize on the loudspeaker, and back the train up to align it properly.
There is not need for a homogeneous vehicle fleet, either. In fact, in Japan just about every line gets a unique train design (which does seem weirdly inefficient, but whatever). At first they we in fact designing different safety doors to match every train. But later, to cut costs and make it possible to have safety gates at stations featuring trains with different door configurations stop, they just designed wider safety gate that can open more on one side than the other.
This brings to mind another simple optimization -- if your trains really suck so much ass that the driver is unable to stop them with reasonable precision, just make the safety gates six feet (or ten feet, whatever) wider than the train doors. (Easier if you make them lift up to open rather than opening laterally.) There is no need for the safety doors to be the same width as the train doors.
They also don't have to be expensive and made of glass. A cheaper gate consisting of a few strong bars is nearly impossible for a human to accidentally fall through.
Finally, it doesn't have to be a track feature. Some of the ones here are essentially the same as an American style garage door opener. When the operator/conductor opens the train doors, he then opens the safety gates.
Japan has one of the largest passenger rail industries with massive capital investment. They are quite good at station design and engineering and I definitely applaud their use of enclosed platforms - which ironically came about more out of a desire to reduce wasted energy on air-conditioned subway stations ventilating into the tunnels.
> I don't think there is any need for computerized stop-assist. On at least some of the trains here, and I think most, it is up to the driver top stop the train. Once in a great while, they accidentally overshoot by a foot or two, apologize on the loudspeaker, and back the train up to align it properly.
Lighter trains can do so manually, but on the majority of lines in the JR network commuter trains and shinkansen have a computerized stop-assist that will get the train within a few inches of the desired point. Some lines do not use the computerized stop-assist (Yamanote, Chuo, etc.) as they are primarily above ground with longer platforms.
> There is not need for a homogeneous vehicle fleet, either. In fact, in Japan just about every line gets a unique train design (which does seem weirdly inefficient, but whatever). At first they we in fact designing different safety doors to match every train. But later, to cut costs and make it possible to have safety gates at stations featuring trains with different door configurations stop, they just designed wider safety gate that can open more on one side than the other.
In Japan, the majority of trains use very similar door configurations. They either have end-cap doors, or 2-4 doors with even distribution. Most commuter lines with the enclosed platforms employ a single type of rolling stock on a dedicated platform.
The Keisei railway and other private roads have platform gates, which are a bit different. Those are waist height and are simply to keep people from stepping to the edge when express trains pass. I think you are referring to these, and they are definitely useful, but see my final point as to why they aren't feasible for Caltrain.
> This brings to mind another simple optimization -- if your trains really suck so much ass that the driver is unable to stop them with reasonable precision, just make the safety gates six feet (or ten feet, whatever) wider than the train doors. (Easier if you make them lift up to open rather than opening laterally.) There is no need for the safety doors to be the same width as the train doors.
They don't suck ass, they are just quite heavy. A Caltrain rolling stock set is 160 tons for the locomotive and then 20 tons per car (est.) That is far, far heavier than any JR trainset in operation and much harder to stop. You don't get electro-pneumatic brakes to stop every car at exactly the same rate - instead you have dynamic brakes and air. Try to stop that with some precision :P
In the end, the main sticking point isn't the rolling stock anyway. The problem is that the platforms are open and anyone can walk on at anytime. There is no way in current Caltrain stations to restrict access with any degree of certainty. To change this would require grade separation and massive overhauls of each station at the very least.
Japanese trains exclusively use raised platforms and 99%+ of stations have restricted access (with only rural/flag stops being the exception). This makes it easy to use some sort of safety mechanism to restrict access to the tracks. Trains in the US are far different and don't really fit the same mold, unless you get to the Northeast and ride the NYC commuter trains (MTA, LIRR, ConnDot, etc.)
I have always felt enclosed platforms are necessary to move passenger rail forward in the US, but they are a long ways off - and likely will never happen regardless.
If they're suicides then there's no point spending a lot of money trying to prevent them, at least not at such a late stage.
However I believe there is at least another solution doable for commuter trains in the us. It is not that hard and not that expensive to have a sliding panel who opens in front of the doors.
But what probably matters here is that cost of a dead human < cost of new system to protect human lives. So as long as this is the case they won't change it.
http://www.caltrain.com/stations/systemmap.html
This isn't a subway, where the entry points are clearly defined. Look at this station, for example
The tracks keep going, with basically nothing protecting them. Putting up walls around the entire thing would be a significant undertaking.
It isn't that you can't enclose a long platform - more that it is both costly and requires raised platforms with grade separation.
Caltrain stations are not designed to restrict platform access to ticketed passengers and anyone can just walk to the end of the platform and go around. Subways and many raised-platform stations on high-density lines are designed to restrict platform access in such a way that it is difficult to get around and hop on the tracks.
Of course, not much will stop a very determined person - it is more designed to stop accidents.
If someone gets trapped between the doors, then the train should be disabled from moving. You see it most often - people dashing in at the last moment getting themselves or their backpack getting stuck. The station doors also have to synchronize with the train doors. Else say, the train door opens, you step out but the station door does not open. Now the train is ready to move...