In a few years I think we'll recognize that self-driving vans (in a few different sizes) are the best trains: cheaper than trains (and therefore lower environmental impact), completely flexible routing, and you can even let them use dedicated right of way if you want to increase capacity.
How does that work? For the same amount of capacity, vans are not and probably never will be cheaper. Steel wheels on rails, with massive capacity, is just drastically more efficient, even with 0 driver costs. You'll still have much more maintenance (the tires, breaks, road) to do with a much higher number of vehicles.
Trains are unbeatable for inland freight between limited destinations, but don't make much sense for current day commuting or traffic patterns.
There is a notable exception that might distort techie perception on this issue: BART. It implements many of its own standards, even down to track gauge. But it is absolutely an outlier (standardised track gauge, ironically, being the USA’s main contribution to the development of the railways, which it was otherwise late to).
Almost everywhere else has rather interchangeable infrastructure and a lot of trans-national businesses.
Nah, this is a UK absurdity because they UK decided to privatise their railways in the worst possible way - every single part (rail infrastructure, operations, rolling stock) was privatised in a local monopoly manner. The rolling stock owners could lease to multiple different operators, so there was some limited competition on that front (probably eviscerated by the need for the ROSCOs to have a profit margin, on top of the fact that there were 10 of them, which meant reduced economies of scale both at acquiring but also maintaining those trains).
They're currently trying to nationalise some parts, after already nationalising the infrastructure (because people died due to bad maintenance due to misaligned incentives - maintenance doesn't bring profits).
But other than that, rolling stock leasing mostly exists in freight.
https://en.wikipedia.org/wiki/Rolling_stock_company
> trains are actually not only mass produced (e.g. by Bombardier)
But yes, trains of all types are mass produced by companies such as Alstom (which were already massive and also bought Bomboardier's train arm), Siemens, Stadler, etc. They can be customised (size, features, etc.), but e.g. the Alstom Metropolis line of metro trains can be found all over the world.
Trains can certainly HOLD more people than a bus, and I hate buses with a passion normally reserved for religious arguments, and trains are Choo-choo, but they are expensive as all hell.
If they weren't why would BRT exist at all? https://www.youtube.com/watch?v=3q-UHd9tFNk
The only way out IMO is that we have to stop ripping off the public by giving away unlimited road space for free.
Trains are a bigger upfront investment, but are cheaper in the long run, especially once capacity is factored in. You need a lot of busses to equal moderate sized trains.
Busses have their place, but not as the backbone for rapid transit in even moderate sized urban areas.
BRT trades CAPX for OPEX. In Latin America where BRT is hugely successful capital is expensive and labour is cheap, so hiring a ton of drivers is easy. In high labour costs markets like the US, Canada, and Europe BRT falls apart. It's often all transit agencies think they can get funding and support for so it's pushed, but it's way too easy to cut back BRT attributes like signal priority, dedicated lanes, and all door boarding to end up with just a bus with a fancy livery.
For a given capacity requirement, in a dense city, they're cheaper. The biggest costs of an urban train line are 1) building stations on expensive land 2) driver salaries, and buses are worse on both aspects; you need much more station space to load/unload the same number of people from buses than from trains, and buses carry far fewer passengers per driver.
> If they weren't why would BRT exist at all?
As far as I can tell BRT is a spook, a way for the road lobby to stop cities building rail. Has it ever actually worked out well for passengers?
Until you actually do the basic middle-school math of calculating capacity. Both of vans, and of the roads
seems to be the worst of all worlds but never quite dies
Most suburbs are dense enough to have demand for full sized buses if service is good. The cost of a bus means that cities don't give them good service, but automated bus drivers would reduce the costs enough that they could run good service and this in turn will get so many people to use the bus instead of driving that the van could not work.
For rural areas a van might make sense, though you quickly get to a frequent van is less efficient than just owning your own van.
This structure allows for greater QoS through the system. Almost any sufficiently dense place without regulation limiting it approaches a mix of larger buses on fixed routes with smaller vehicles on slightly less fixed routes.
SF Muni has a 13% absence rate. Without the driver we can have more vehicles, more often, and on-time every time. Once we can do that, we can vary the size of the vehicles to cover different routes and to allow for QoS.
There is a safe minimal distance between trains, in fact, a safe distance for a given speed. Shorter trains are not exempt from obeying it. You can make shorter trains more frequent at the expense of lowering traveling speed.
What is the cap of throughput is due to these speed limitations is an exercise left for the author of the article.
They already did that exercise:
> 3-car trains running at 30-40 trains per hour (a normal peak frequency for automated or even some human-driven metro lines) reach a capacity of about 18,000 passengers per hour per direction, well above the expected demand of any American line that doesn’t go through Manhattan.
What is incredibly expensive, though, is retrofitting a line designed for manual operation to run automatically instead.
DC Metro just recently re-enabled full automatic train operation across all the lines in June.
The fundamental constraint is not technology, but people and physics: you need to decelerate and stop, let people disembark and get on, accelerate and clear the platform. This cycle requires a bare minimum of 90 seconds, although IIRC a few lines in a few places like Paris and Moscow do 85 secs.
That was a world record for a line following modern safety standards, set less than 10 years ago. It's hardly a case of "not bothered", it's just hard.
(But this strategy is orthogonal to the article, because it requires long platforms.)
> They already did that exercise
They took "30-40 trains per hour" out of thin air and exercise was to calculate whether it is even possible to have more frequent shorter trains.
It’s the same with cars. A 2s headway with cars holding 1 person each means that the maximum capacity of a highway lane is 1,800 people per hour, no matter how fast they go (the cars are further apart at higher speeds).
In my experience (on UK roads) this critical speed is around 17 MPH - but it might be a little different elsewhere.
There are roads that regularly suffer from acutely insufficient capacity in many metro areas; specifically, repeatedly at times _the dynamic pricing toll that would discourage enough people from using it to stay uncongested_ would overshadow the price of a rental-with-driver (Uber-style) during off-peak times. It's not that the people shouldn't get through; it's that most people won't need more than a backpack worth of luggage with them and could thus be packed 3~4 passengers for each driver. Splitting the toll would be the reason to do so.
Unfortunately only really dynamic congestion tolls would really fix the concept of rush hour traffic jams. And the necessary surveillance system would bring severe mass surveillance/tracking concerns with it at least in central Europe.
When operating at slightly faster than max capacity, local slowdowns cause a local increase in throughput, allowing bunches to dissipate.
If you maintain the proper following distance at maximum capacity when there is an issue you can momentarily drop to 1 second (while braking) and then expand back to normal and there is no effect on traffic.
But if the average following distance is such that the road is exactly at the peak of throughput, or any smaller, then any momentary dip into tighter following distances pushes the road it into a positive feedback operating mode, which triggers a traffic jam.
If you haven't seen it, here is a classic experiment demonstrating the effect: https://www.youtube.com/watch?v=Suugn-p5C1M
(But obviously the solution to traffic isn't 5 more lanes, it's viable alternatives to driving!)
You see a lot of blanket assumption, in discussions of traffic, that throughput should be maximized, and almost no examination of whether increasing throughput is a goal that makes any sense.
For example, working from home has catastrophic effects on throughput.
Signaling systems used on automated trains know the position, speed and capabilities of every train. Keeping a safe distance behind isn't a problem.
Trains have the capability to accelerate and decelerate faster, we mostly don’t do so for comfort and safety reasons.
Singapore built the orbital Circle MRT line exactly as the author wants: small trains (3 carriages, vs 6-8 on other line), small stations to fit these trains, frequent fully automated operation.
However, the line turned out to be much more popular than planners anticipated, with the result that at rush hour, it's common to have to queue just to enter the platform where you need to wait for multiple trains to arrive before you can squeeze in.
And the tricky bit is that there is essentially nothing you can do now to fix it. There's a hard physical limit (around 90 secs) on how fast the cycle of a train arriving, people getting on and off, and departing can get, and retrofitting all 30+ deep underground stations to be larger would be an insanely expensive and disruptive exercise.
But before this you had no idea that there was so much demand, right?
It's quite a lot easier to sell a huge monetary upgrade on something oversubscribed rather than a huge monetary outlay that may be a complete white elephant.
A better solution that no one has the political will to implement is inferior to every solution that can actually be implemented.
- If the rush hour contines to get more crowded, the transit fee would raise until people avoid commuting on peak hours.
- If the rush hour disappears, the extra premium fee would fall to zero.
It is not unrealistic as it sounds; JR East (the biggest railway operator in Japan) recently introduced "Off-peak Commuter Pass", which is 1) 15% cheaper than a normal pass 2) and cannot be used between 7:30 - 9:00 AM.
So they are beginning to implement a dynamic policy based on how crowded their trains are.
Of course it's fine to give people an incentive to ride outside of peak hours if they can, but demand is not that elastic, because of daily rhythms and how interconnected they are. Raising prices until you have removed the last option a lot of people have to get around, just so that others can enjoy their ride more, will remove a lot of opportunities from the economy.
I'd also be interested to know if the slack introduced by people moving their travel to off-peak times is not immediately taken up by others again, in the way that widening a congested road often just allows more cars to be congested at the same time. (Edit: Which can still be a good thing, especially for public transportation, because I'd argue that giving more utility to more people is its mission.)
Commuting times used to be rigid until COVID and the resulting WFH and hybrid policies showed corporate the world would not burn down if some people showed up at 800 vs 830 vs 930 vs 1000. Being in at 9am sharp it turns out is not a hard requirement for many jobs.
There is a good basis to this. Every new mile and station you add to a system compounds on the size of the system. The 11th station connects you to 10 places. The 51st station connects you to 50 places. Build small and you never get the critical mass needed to see widespread adoption.
A more DC-like BART probably would have prioritized the second Transbay Tube and Geary Subway over the line to San Jose. BART did throw out the Geary baby with the bathwater when Marin County left.
You can build another line. This one was cheap and the demand is clearly there. And two medium capacity lines are generally better than one high capacity line in terms of offering more options to travelers.
Also to add: building a high capacity line (say with 2x the capacity of the Circle Line) doesn't take 2x the land. There are obvious economies of scale.
Building two lower capacity lines has some diseconomies of scale, as the opportunity costs of the land use mount.
Singapore's MRT lines are some of the most expensive public transport projects ever. The Circle Line, fully automated and fully underground, cost S$10 billion[1]. The recent Thomson-East Coast Line, still partially under construction, is projected to cost S$25 billion[2]. It was not 'cheap' by any means.
> You can build another line
Singapore is building another line: the Cross-Island Line[3]. It has planned or is constructing at least three more lines[4][5] to achieve something like 460 km by 2040, thereby exceeding the length of the London Underground. About S$100 billion is earmarked for public transport expansion.
But the Circle Line was, as someone who has used it ever since it opened in 2009, ill-conceived as a 'small line'. It is heavily overcrowded. Because of the immense traffic and somewhat lacklustre maintenance, it has suffered several delays and breakdowns. The ideal thing for LTA to do would be to expand each station's capacity, because it links all of Singapore's radial lines at heartland residential areas.
[1]: https://medium.com/from-the-red-line/was-the-circle-line-bui...
[2]: https://www.mot.gov.sg/news/Details/speech-by-minister-khaw-...
[3]: https://www.lta.gov.sg/content/ltagov/en/upcoming_projects/r...
[4]: https://www.lta.gov.sg/content/dam/ltagov/who_we_are/our_wor...
[5]: https://en.wikipedia.org/wiki/Mass_Rapid_Transit_(Singapore)...
My ideal default rapid transit buildout for midsized urban areas would basically be SkyTrain with value engineering to extend the platforms to 100 or 120m with minimal cost in the future.
Nothing feels slower than standing still. Sitting in congested traffic sucks, and so does loitering on a train platform or a bus stop.
When the vehicle arrives so quickly you don't even consider the timetable or care if you just miss a train transit is an easy choice.
A simple example would be 3 services on the same line that follow a repeating pattern every 3 stations. ==[A]==[B]==[C]==[A]==[B]==[C]==. The service (AB) stops on the stations labelled [A] and [B] (skipping C). The service (AC) stops on the stations labelled [A] and [C] (skipping B). The service labelled (BC) stops on the stations labelled [B] and [C] (skipping A). In this manner, all three services skip over 33% of the stops on the line, but no matter what your origin and destination, it is always possible to travel from an [A] station to a [B] or [C] station (or another [A] station), and likewise from a [B] or [C] station.
To the extent I've worked out the logistics of this, if you allow for trains to catch up but not pass each other at the platforms, you can push this idea as far as only stopping at 2-in-5 stations without sacrificing headways or capacity.
Just a weird thing thats been taking my attention lately.
You will also find long trains that split in half mid-journey, so you need to make sure you get in the right car or you'll go to the wrong place.
Edit: I guess it's not exactly what you're saying, in Europe you will find platforms split into several sections with multiple trains to board, but they'll be for different lines with different destinations.
Consider a metro system with trains 70 m long. With 10 m of space between the two boarding areas, that means the length between both sides of the station is 150 m. If the entrance to the platform is in the center, the walk to the middle of each boarding area is 45 m, taking about 30 seconds. If the entrance is at the end, that becomes 35 m or 115 m (taking about 20 to 90 seconds to walk). I think those figures are comparable to the dwell time of a typical metro system.
I do think it's a very interesting idea though! I think it'd work better for longer trains over longer distances, where the time spent accelerating and braking is often greater, but unfortunately few places are even considering anything shorter than 15 minute headways for such rail services.
So in a 2-of-5 stopping pattern you'd save 3 min every 5 stops, but depending on when arrive at the station you will sometimes get unlucky and just miss your train, in which case you'll have to wait an additional interval than you normally would for the next one (costing you 3 min). So for very short trips (under 3 stops) it actually makes travel time a bit worse since there's very few stops you can skip over in such cases. But for more median trips there is a saving, and for the longer trips it saves well into 2-digit minutes.
In your example I think in the case with the entrance at the middle of the platform there wouldn't be any real effect, because splitting the platform also in the middle doesn't marginally change how quickly you can get to the nearest open door on the train you want. There's no real need to get to the center of either boarding area. In the case of platform at the end of the station, I can accept that adds about 40 m of walking distance that wasn't there before (at least for some people), and that translates into an extra 20ish seconds. Less than that if you do the rational thing of walking briskly when you see you're about to miss your train.
So you're right that the platform split adds some time which isn't completely negligible, and depending on what fraction of the next n stops you want to skip it comes at a penalty of longer wait times for the correct train, but both of these are fixed upfront costs whereas the dividend accrues linearly the longer your on the train.
This lets you schedule them much closer together than the conventional 3 min while still being safe.
Why? If the minimum is 3 min then it's 3 min (at least with a modern moving-block type setup). If it's safe to run them 2 min apart it's (generally) safe to run them 2 min apart the whole time.
As for short trains on long platforms, this is pretty common in NL where the bigger train stations can support both very long international trains and shorter local trains on the same platform; a train can switch to a center track halfway on the platform.
I don't think it would really work for a subway system; people expect it to be hop on, hop off. In some places the stops (or people's final destinations) are so close together they can choose to get on/off earlier/later, but this system makes that less viable. You'd have more people shuffling through train stations trying to figure out which train to get or whether they need to wait for the next one, also putting extra load on staff for confused not-locals. And finally, you'd need extra rails or tunnels so that a train can pass another.
That said, full automation is a major game changer because it enables short trains with high frequency which you would otherwise struggle to afford. Don't build light rail, build light metro which is a small cost more (build a metro with light rail) but now you can have full automation.
Munich is planning on building a second subway route. It just does not have the money nor the space to build one. There has been discussion on building a second tunnel below the first, or on enabling the common train system for overflow by suburban trains -- by installing more signals to run more tightly packed.
[0] https://db-engineering-consulting.com/en/projects/munichs-se...
EDIT: Listen, you can downvote me but it's true https://www.etany.org/statements/impeding-progress-costing-r...
"Facts don't care about your feelings"
It's also a feat of mental gymnastics to put European infrastructure on a pedestal while ranting about NY's taxes and, more broadly, the US' difficulties building infrastructure.
If we think of the station as the head of a Turing machine ...
When you're building high capacity transit for dense cities, yes. You could stop the train twice in the same station, sure, but at that point you have to stop for twice as long (probably a little longer than that as your riders will get confused), so your train takes ~twice as long to get anywhere and can therefore carry ~half as many people.
The moment you have only some cars platforming, then people need to make sure they are in the right car, and if, say, a slow-walking elderly person is in the wrong car, then her getting into the right one to get off wastes a lot of time.
> Wait times are the second-most underrated component in making metro systems effectively faster (right after street to platform time).
Street to platform time made using public transit annoying in several Asian capitals. Wait times are the main reason why I often hate traveling with public transit even in a city with one of the best public transit systems - if you aren't planning your life around the transit schedule, the wait time is essentially an unpredictable part of your travel time (and if you do plan around it, it's one more annoying thing to deal with).
If you leave your doctor's office, and you have a 10 minute train home that goes every 30 minutes, then you really have a 10-40 minute train home - so for practical purposes, you essentially have a 40 minute train. Once unreliability or delays become frequent enough to start being a factor to consider, this could easily turn it into a 70 minute train.
Another big factor is the time between the actual start/end of the trip and the nearest station. Even with near-zero street-to-platform time (e.g. busses/trams) and reliable, reasonably frequent public transit, once you add it all up, a bicycle often ends up being faster than even excellent public transit.
Others have already pointed out some of the practical things that would limit what you can achieve with shorter trains like minimum headway between trains and the congestion that would occur after a delay where you would expect people to wait for maybe 10 short trains to arrive and depart before you could get on.
There are other problems. The cost of the smaller trains means if e.g. 2 carriages each, then every other car is a driving vehicle with motors and control equipment. On longer trains, this could be 1 in 3 or 1 in 4. Each of these not only requires regular maintenance (so that's doubled the maintenance requirements) but also creates massive congestion issues in maintenance yards. Commonly, 1 or 2 full trains fit in each siding so getting most trains out if e.g. one is broken down is usually easy enough. Imagine having to move 3 or 4 separate smaller trains out of the way, they are not automated in the yards.
Most people would be very unhappy knowing that they might be alone on a train, which is a main reason why operators are not quick to get rid of all staff. But currently, a single driver and conductor is close by for, maybe, 8 carriages. This would double if you needed a single attendent on each 2 carriage train.
The signalling is very heavily designed around traditional trains with its delay after passing a signal. 4 x 2 carriage trains would utilise more signalling capacity/time than 1 x 8 carriage train. If these were all coming from the same location and needed to return, that would also need a lot more platforms so that the following trains don't block the first trains in terminal platforms. This is already a major problem at most UK terminal stations so that is largely unsurmoutable.
I also heavily question the idea that station cost is largely dependent on platform length. On most surface stations, platforms are relatively easy to construct and whether they are 4 carriages or 12 carriages long still require a station building or 2, some ticket machines and CCTV etc. For elevated railways that is more likely to be true maybe.
So yeah, like someone else said, we are unlikely to come up with ideas that 1000 other engineers haven't already asked, the easier problems to solve are around planning, design, legal rights for infrastructure projects etc. since these tend to eat up sometimes decades and billions of dollars.
There are two passing vehicles in the background, based on the direction of movement they seem to be driving on the left. Assuming the image is not reversed this would imply London...
of course, this could also be discovered by recognising literally anything else
salynchnew•14h ago
Depending on the system in questoin, this might be impractical or impossibly dangerous.
jrockway•14h ago
You can see whether the problem is being cheap or if it's actually a capacity issue. If weekend service sucks but peak hour service is good, then it's just being cheap. If rush hour can't handle passenger volumes, then you need a signal system / automation upgrade. (Or a parallel line!)
As a New Yorker, I'm very jealous of Vancouver's SkyTrain system. ("But NYC has the best metro in North America!!" Maybe...) The NYC subway has a lot of peak hour capacity. I hate traveling at peak hours. So that means I'm always standing in stations waiting 10 minutes for the next train. If I lived in Vancouver, then that would be 3 minutes. Sounds good to me!
I also agree with the author in that I'm not sure what elevated trains did to hurt people. I lived in Chicago next to the L for years. It never bothered me. It's nice to see out the window and look at something while you're in transit. And it's cheaper than building things underground. NYC apparently got rid of its elevated railways because of snow, so that's something to watch out for. But Chicago gets more snow and it does OK. (Having commuted on both systems in the snow, it's a mixed bag. Chicago doesn't shut down, but it's slow as people remember how to deal with snow. NYC can run on snow-free underground tracks, but sometimes the governor is like "fuck it, I'm cancelling all the trains anyway" and then you're just stuck.)
troupo•13h ago
I've only seen them in Hamburg. They are generally an eye-sore. And the area underneath them is somewhat unusable (like any area under a bridge): you can't really build anything there. So it's an eyesore above an eyesore.
PaulHoule•13h ago
theluketaylor•13h ago
SkyTrain works because of a virtuous cycle of attributes: Full grade separation enables automation. Automation enables many trains per hour. Many trains per hour with short trains still has tons of capacity. Short trains means lower costs for stations, which as the article notes is a huge portion of rapid transit costs. Lower cost means building more transit for the same budget, so more transit gets built. More transit with great headways results in transit being the mode of choice. Take any one of these elements out and you can still have a functioning transit system, but the magic is missing.
Link light rail is so close to full grade separation but not quite there, so headways are limited by the grade crossing. With longer headways, bigger trains are needed to serve the same capacity. Bigger trains mean big stations and beefier, more expensive viaducts. Big stations are expensive.
Link is gaining ridership and offering great service, but it's hard not to think if they had learned the full lessons from nearby Vancouver it would be even better (and cheaper).
mschuster91•13h ago
You need to either build them in right at the start of a new development or you gotta demolish a lot of housing - similarly to what was done in a lot of US cities when the highways were built [1].
[1] https://www.reuters.com/world/us/us-freeways-flattened-black...
crazygringo•12h ago
They're horrifically ugly for pedestrians and city life generally. If you've been around the elevated subway tracks in Brooklyn, for example, they're not pleasant to be around. They block out the sun, they're incredibly noisy, they make the street claustrophobic, they definitely become streets to avoid unless necessary.
Yes, they have a nice view if you're a passenger. But they make the city much, much worse for everyone below them.
kccqzy•12h ago
theluketaylor•12h ago
Melbourne is eliminating grade crossings with new guideways and creating linear park space underneath as they go.
bregma•26m ago
bobthepanda•4h ago
jillesvangurp•3h ago
These days, you should be designing for autonomous operation for new infrastructure. Much easier when you design from the ground up. Not doing that could be an expensive mistake.