Comment by xnx
7 days ago
> Small Train is Good Train
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.
> self-driving vans (in a few different sizes) are the best trains
Until you actually do the basic middle-school math of calculating capacity. Both of vans, and of the roads
https://en.wikipedia.org/wiki/Personal_rapid_transit
seems to be the worst of all worlds but never quite dies
> cheaper than trains (and therefore lower environmental impact)
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.
I've yet to see a convincing argument that trains are cheaper than busses (especially when you consider that most of a bus's route (in the USA) would be on "free" roads).
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
If the bus is using "free" roads, then it's forced to compete for space with general traffic and likely to perform far worse than BRT or trains that don't have to do this. Worse performance/speed/frequency => lower ridership => worse profitability => calls to cancel the "wasteful" bus etc.
The only way out IMO is that we have to stop ripping off the public by giving away unlimited road space for free.
> I've yet to see a convincing argument that trains are cheaper than busses (especially when you consider that most of a bus's route (in the USA) would be on "free" roads).
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?
1 reply →
Bus lifespan is 15-20 years max and needs tons of maintenance during that time. Trains last 40 years and go 100,000 miles+ between failures.
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.
2 replies →
The key benefit of BRT is that they have a dedicated right-of-way without conflicts with other traffic just like train tracks with full grade separation. So to do that, BRT suddenly no longer has free roads. You now need dedicated BRT-only roads.
This will depend on density and what's already built. An extreme case: it wouldn't make sense to replace a large, rural school district's bus fleet with trains.
All trains, tracks, and stations are specially designed and built, which makes them very expensive. There is no mass-produced off-the-shelf solution like with cars and roads.
Trains are unbeatable for inland freight between limited destinations, but don't make much sense for current day commuting or traffic patterns.
This is not really true. There are absolutely things in the rail system that are as “mass produced” as roads are. Railway platforms for example are a highly standardised form that has many other applications (loading and inspection bays, walkways etc.), trains are actually not only mass produced (e.g. by Bombardier) but are actually leased by their operators in Europe; there are five or six, I think, main ROSCOs —- rolling stock companies —- in the UK for example.
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.
4 replies →
There are off the shelf trains, tracks, and stations already. It's one of the key ways many european and asian operators deliver tons of transit for the dollar.
We can call them a bus...
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.
Smaller vehicles allow more dynamic dispatch. SF buses are low quality transport because they have to stop every block to allow for an elderly local population. With vans, you can put more of them on the road and dynamic dispatch allows them to skip stops for non-loading.
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.
That is a horrible idea and shows how little you care or think about bus riders. Bus riders need predictable. If the but has dynamic dispatach you can't trust it will get you there on time. We need predictable routes with minimal variance so everyone gets predictable. Good qos is predictable.
2 replies →
Yep, but I think automated bus drivers change the math. A standard-ish city bus is $1M and seats 50 (though they're rarely full). You could definitely get 5 vans that would seat 6 for $1M. Absolute seating capacity is lower, but total throughput could be greater because of the extra flexibility.
That doesn't work. If you run frequent service more people will use it and so you still need the big bus.
> A standard-ish city bus is $1M and seats 50 (though they're rarely full).
This sounds wrong, or at least your local transport company is being ripped off. Dublin Bus bought 600 plug-in hybrid double-deckers a few years back at a unit cost of 450,000 eur; they take about 90 people. More recently it bought a couple hundred electric buses of the same size for 500k apiece (though of course these are cheaper to run).
Also, define 'rarely full'. At rush hour, most buses will be full for at least part of their route.
Taken as an example, Dublin Bus has 1,100 buses, almost all ~90 person double deckers (homeopathic numbers of smaller single-deckers and giant three-axle double-deckers can be ignored). Let's say 90% are operating at peak time. That's a capacity of 89,000 people on the road; implementing your plan with 6 person vans would require about 15,000 vans. This is obviously completely absurd.
(The original claim that they were good _train_ substitutes is even more absurd; replacing a 1,000 person rapid transit train would require 166 vans. Which is multiple kilometres of vans. On a decent metro line such as train might arrive every 3 minutes. There is just no way you can do that with vans.