- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
You must log in or register to comment.
As much as I agree that decent public transport is significantly preferable to autonomous cars, this video is a bit of a straw man argument.
Self driving cars will never beat trains/busses on efficiency. But that’s not why people like them. The major selling point is the illusion we can sustainably keep living in spread out suburbs.
Trains will not stop at the doorstep of your mcMansion. Trains will not allow you to comfortably carry your biweekly Costco shopping. Trains will not provide you the luxury of traveling from point A to B without ever having to come in contact with the plebs.
Unfortunately, he kind of admits as much in the end of the video. Where says that self driving cars will entrench suburbia and will work in sparsely populated areas.
He’d have been better off focusing on the economics of self driving cars. E.g. how the vegas hyper loop is an insanely expensive money sink, unneeded musk subsidy, and just plain doesn’t work.
I haven’t looked enough into the hyper loop. Why does it not work? It seems pretty straightforward to have a tunnel with self driving cars. I feel like that’s the perfect environment for them, like a sterile lab.
Hear me out: Autonomous Transit
I live somewhere with automated light metro and it’s great.
As much as I enjoy saying Fuck Cars, this person is just wrong. They’re making some fundamental assumptions that are wrong to base their argument on.
Transit doesn’t work in low-density locations.
It’s great in Tokyo or Hong Kong, but if you’ve ever been to one of those cities and the compare it to somewhere like Seattle, you’ll know they’re fundamentally different. The Seattle Metro Area is 15,000km^2, Metro Tokyo is 13,500km^2. Metro Tokyo is 40 million people, Metro Seattle is 4 million people.
That doesn’t mean there aren’t locations in Metro Seattle where a transit application work best, but they’re just extremely limited. Most other American cities are the same. The ones that are better suited to transit(more higher density areas) already have more of it (like New York)
Autonomous cars will improve the transit situation in lower density locations until the density exists to support mass transit.
The reason why American cities are low density is because they were built for cars. Or destroyed for cars. But American car-dependent suburbia is not financially sustainable. Car infrastructure isn’t just absolutely miserable, it’s also extremely inefficient in terms of both cost and space. Self-driving cars or electric cars won’t solve this. America needs to learn how to build cities properly again.
This.
Everyone shrieks that we can’t just tear down cities for public infrastructure and higher density, but completely refuse to acknowledge that that’s EXACTLY what we did for cars.
They weren’t built for cars, they were built for single family homes with yards. Cars are just required for that level of density.
A lot of people don’t want to live in cities, because it means giving up the house and yard.
That isn’t a horrible opinion, but it does drive up prices near cities because of inefficient land use.
Transit doesn’t work in low-density locations.
What’s the minimum density for transit to work then?
It always seems to be “more than we have now” to US naysayers.
Have you checked out the Portland Light rail (MAX) and streetcar map these days? It’s quickly starting to look like an actual European style transit map. Was the original MAX line (Blue from 10th downtown to Gresham) along a terribly dense and special case part of Portland? Nope. It was through pretty normal US city areas. It’s become a backbone of the city’s development and transit since then.
Part of how cities get denser is to stop allowing low density development, and then putting in transit so that developers can build up instead of out.
Another great example is in Seattle. They’ve pushed serious Missing Middle mixed use development outside of the city core. My favorite transformation is at the Fauntleroy Way and Alaska junction area. It’s built up nicely and now they’re running a tram through it and it’s going to just keep getting better for people living there.
The US seems to have developed an idea that once a city it built in a certain way, it can never be changed. There’s a serious lack of imagination, vision, and willingness to look to other templates for how a place can grow and adapt in response to changes over time.
Total cost of a bus is ~$122 USD/hour ($165 CAD)
A car costs $729 USD / $988 per month on average and is used 380 hours per year on average. Therefore $23 USD / $31.2 CAD per hour (no driver cost, unlike the bus). Therefore bus costs the same to run as 5 cars.
It depends of service frequency and stop radius, in North America (while there are different definitions), 106 households per sq mile is the line before it starts dropping to rural. That’s 6 acres per household.
Assuming they have 2 cars (not a big assumption with that density) and there is a single stop that goes to a transit link 30 minutes away (1 hour round trip) you get the following frequencies:
800 yard radius: 0.65im^2 = 69 households = 138 cars = $1.2m per year = 9,880 bus hours per year = 27 busses per day. And a second stop and you’ve got a bus every half hour.
Add some bike racks for get a 2400 yard radius: 5.84mi^2 = 619 households= 1,238 cars = $10.8m per year = 88,770 bus hours per year = 243 bus hours per day = 10 busses per hour. A bus every six minutes all day, every day. For houses with 6 acres of property.
That assumes everyone wants to go to the transit link.
I’ve lived in Japan, and not owned a vehicle. I’ve lived in a small north American city and taken the bus (I lived on the busiest route) It’s orders of magnitude different. I still had to check the bus schedule here in north America. I never even bothered in Japan because the next train was always 2-3 minutes.
Autonomous electric busses are probably the best option for most North American situations, they reduce the per hour operating costs enough that we can absolutely saturate even the least used routes with sub 10 minute service and entice some people out of cars.
Autonomous cars will help though, I think people underestimate how easy and cheap autonomous taxis will become. That combined with buses will probably allow people to give up paying for a dedicated vehicle.
It’s also a very low density scenario. Upping density to a cramped 3 acres per household (barely enough for a single horse), that bus frequency increases to 20 per hour, or every 3 minutes.
And this is all before adjusting for peak/off peak. I’m saying if there is a bus at 03h00 Monday morning, the next bus is 03h06 at six acres a house or 03h03 at 3 acres a house. And this is single stop line.
If you put 3 stops on the line, at three acres per household density, were talking a bus every minute. I’ve chosen a transit link, as that should get people anywhere afterwards, but a suburban stop between two towns might split their hours between the two 50/50, or at whatever frequency works.
As a different example, let’s look at NJB’s favorite: fake London. London ON has 273,000 registered vehicles (2022). That means the population of London spends CAD $3,236,688,000 on personal vehicles per year. That $3.2 Bn can pay for 19,381,365 bus hours. That’s 53,100 bus hours per day.
The city has 9616 roads, but I can’t figure out how many Km of road it has. I do know the city is 437 km^2. It’s pretty square, so we’ll go with 21 km by 21 km. If we slap Chicago city blocks (100mx200m) on that we get 210 roads one way, 105 roads the other way, or 315 roads total for 6615 km.
Going with frequent stops, we normally get 30km/hour but what’s with other traffic; since we eliminate everything but trucks and emergency vehicles, we can use the BRT speed of 52km/h. so it takes 127.2 bus hours to travel one direction the entire road network, or 255 bus hours to cover both directions.
With 53,100 bus hours per day, we can send a bus on every road in both directions 208 times per day. That’s a bus every 7 minutes, both directions, every hour of every day. And that’s the un-optimized solution.
If we “borrow” bus hours from low periods (23h00-04h00) we can increase peak periods. Same with borrowing from weekends or holidays
Autonomous electric busses are probably the best option for most North American situations, they reduce the per hour operating costs enough
Busses cost $122 USD per hour, a driver costs $18 USD per hour. So that’s a savings of 14.75%. But a human is better a dealing with traffic cones, knowing when they don’t have to stop, calling the police if something goes wrong, knowing when to wait a few more seconds for someone running after the bus, letting someone off between stops so they are closer to home, etc. Also, based on you yanks decapitating hitchhiker-bot, you probably want human supervision on your busses. Flexibility on scheduling goes to robo-busses though.
Autonomous cars will help though, I think people underestimate how easy and cheap autonomous taxis will become
I think so too, but I don’t think the robotaxis can take over until most the human driven fleet is off the roads.
I think the key here is focusing on robotaxis, which operate all the time, in contrast to personal self driving cars, that are working <400 hours per year.
You’re missing some intangibles that don’t allow for that calculation with the density you’ve got.
People spend money on cars because it’s more convenient. You wouldn’t get the same amount of money out of a less convenient option, even if it’s more “efficient” overall. You can’t(or rather shouldn’t) have sex with your high school date in the back of the bus.
What about long distance travel? Bus service between towns would be needed too, and at high frequency in order to make up for the lack of cars. It currently takes me 2.5 hours to drive to my in-laws place. On a bus it’s a 4-5 hour trip because of bus switching at key cities in between. Additional frequency doesn’t even fix that. Density would, because then there would be high speed trains between population centers (like in Japan)
Another big one is shopping, a family can’t grocery shop by bus unless they’re doing it multiple times per week to make each shop smaller. This problem is solved by density, where’s there’s a grocery store within a couple of blocks of every house like what I had in Japan, but doesn’t work when you need 15 minutes on a bus to get to the store.
It also doesn’t account for peak commuting. If you have 60% of your population all needing to travel in a one hour period each morning and afternoon, all moving in a single direction, when your bus routes are 30-60 minutes long, you end up with a lot of problems of needing to over-buy busses and over-hire drivers who aren’t needed outside two single runs each day and which are separated by 7-8 hours. A car doesn’t mind sitting there unused and unpaid all day.
These problems don’t exist with autonomous vehicles.
My bigger point is that the density required for busses (and therefore trams, regional, and high speed rail) is way, way, lower than Canadians and Americans think it is.
People spend money on cars because it’s more convenient
Correct, and cars are mostly more convenient because of the lack of transit options.
You can’t(or rather shouldn’t) have sex with your high school date in the back of the bus.
Just have sex in your house? Car sex is awkward and ungainly anyways.
What about long distance travel?
Trains.
Another big one is shopping, a family can’t grocery shop by bus unless they’re doing it multiple times per week to make each shop smaller
Shop every day, or have it delivered. Also, assuming 1.2 pers per car, we get 6 people per bus. Lots of space for your stuff!
It also doesn’t account for peak commuting
Correct, I gave the number of bus services averaged over the year, these can obviously be adjusted to each services points real requirements.
needing to over-buy busses and over-hire drivers who aren’t needed outside two single runs each day and which are separated by 7-8 hours
Busses, yes. Drivers? Split the shifts to cover the two commuting peaks, cross-train with light maintenance. Non-driving is wrenching, cleaning, and admin. Frankly I think finding sufficient drivers is a bigger problem.
These problems don’t exist with autonomous vehicles.
Different problems though.
In 1903, New York Times predicted that airplanes would take 10 million years to develop. Nine weeks later the Wright brothers achieve the first manned flight.
This guy is a fool.
deleted by creator