Space efficient in what way? Parking. I don't think this is a limiting problem. It is the gridlock on the roads. Tunnels can be put underground with a large number of levels if need be. People don't really like the space efficiency of standing packed right next to each other for an hour.
With an automatic system of vehicles you can build a tunnel/track and can put into the system any size vehicle you want. One passenger, two, three, ten, twenty, one hundred. What is the difference between a car, or a bus, or train anyway in this system? Without a driver needed you can have vehicles of any capacity.
No, not just parking. For every passenger that takes a train, that’s one passenger not sitting in a car that takes much much more space. An average subway train can fit around 800 people seated, and takes up much less space than 800 cars. It is precisely how much space vehicles take while being driven that is the issue here.
Likewise, “flexibility” is precisely the issue here. Why should a train carrying 800 people have effectively the same priority as a car that carries one, maybe two or three people?
I know that it’s tantalizing to think that cheap tunneling can solve this issue, but you are just outsourcing the fundamental problem that traffic exists because people want to go places.
A good example is going to the airport. People need to get off, unload their luggage, and there’s only one place you can do it, at the departures terminal. No amount of self-driving is going to solve the problem that the transport vehicle needs to stop, let these people off with their luggage, and start again.
Transit-only lanes are a colossal waste of space. Let's say we have a 5-mile long route with 12 minute headways. In San Francisco, buses move about 8.1mph, so that's 3 buses. (Each bus completes a loop every 36 minutes, and we'll "assume a spherical cow" by saying that bus bunching doesn't exist and they're evenly spaced. We'll also assume that the extreme ultra-luxury decadance of a mere 12-minute wait somehow got past SFGov).
Three buses, 40 feet long, on 26,400 feet of road. Only 0.45% of the road is utilized. 99.55% of the bus lane is being wasted.
It would be smart to keep the road below the point of saturation, where it declines in utility for everyone, but demanding that 99.55% of it sit empty is certainly not an efficiency-minded position.
I'm not sure if you're being satirical or not, but the efficiency of a road (or any transport method) is not, and should not, be the density of vehicles able to be contained on it.
At least go approximately by how many people can use it in a given amount of time, at what cost, to get people from their trip start to their destinations.
Lets not even get started on the additional considerations for transport logistics (where you park, fill up, what you block via each option, who pays), because at the very least you first need to understand the notion of throughput rather than vehicle density :\
>, but the efficiency of a road (or any transport method) is not, and should not, be the density of vehicles able to be contained on it.
Great! Then you will not ever point to parking space to claim that cars are inefficient, and will join me in calling out people that do?
>at the very least you first need to understand the notion of throughput rather than vehicle density
Yes! And although a road packed with buses has higher throughput than a road packed with cars, in practice our choice is between a road packed with cars and a road which sees a bus every 10-20 minutes. Since each bus has 8800 feet to itself to carry 45 passengers, each passenger can actually have 195 feet to themselves at breakeven. The bus route with realistic headways has much lower throughput than heavy private car traffic. Although I agree, it would be nice to see wall-to-wall buses with multiple buses stopping per minute.
(Obviously all of this is crude since at some point stopping and starting at red lights, and waiting for pedestrians before turning, becomes a major concern).
Tunnels can be put underground with a large number of levels if need be. People don't really like the space efficiency of standing packed right next to each other for an hour.
People don't like taxes either :) Boring is expensive, boring more is more expensive. And to achieve the same throughput of people you'll need something like forty times the number of lanes; just imagine trying to get all these cars entering and leaving the tunnel the same time as a single bus: http://farm4.static.flickr.com/3662/3398936283_f32ac7d42a_o....
An alternative would be to have trains with platforms with regular seats, and platforms for transporting cars (like Amtrack's Auto Train, but open). Then people willing to pay for the extra space could do so.
A think we both really agree on the solution, just coming from slightly different perspectives.
Boring is expensive. That is why Elon Musk wants to develop new systems and thinks a factor of ten in cost reduction is not to hard, with smaller tunnels being the biggest easy gain.
With your cars on trains idea you almost get what I am saying, but why do you need the cars to go onto trains? Just get the vehicles to run close together and you have the density of a train. Imagine a tunnel with cars going at 100mph and 40ft of space per car. One gets a rate of 13200 cars per hour. BART at peak does about 60,000 people per hour. If you mixed this tunnel with buses and cars with various amounts of people in them you could get a great throughput. Like you said, charge by the total space taken up by the vehicle. Bus rides pay a $1 each while the single person in a huge car pays $20. You don't need taxes. People will pay for transportation, just not many want to pay for others peoples transportation.
Cities have a lot of their value in allowing millions of people to potentially interact with millions of others. Restricting when and where people can meet really diminishes the vitality and value of the urban life.
With your cars on trains idea you almost get what I am saying, but why do you need the cars to go onto trains?
Trains - or Elon Musk's "skates" - allow you to use existing cars, without having to wait for the whole fleet to go electric and self-driving, but sure, in the future they could not be needed.
An advantage that does hold even for future cars is saving their batteries for outside travel.
Yes, the skates are good for moving non-electric cars with electric cars in tunnels. I think it is a great idea you have to equate and call these closely-packed individual vehicles trains. Trains are a morally good way to travel while using a car is a morally bad way to travel. We all want to be morally good, right?
$20 per car, 13,200 cars per hour, eight hours a day gets you $770,880,000 a year.
The Big Dig in Boston cost $14 billion, but some estimates put it at more like $22 billion when all is said and done.
So if you run your big cars through the tunnel for twenty years, you'll pay for one small tunnel.
But fine, Musk can get it down to 1/10th - but the Big Dig is 3.5 miles, and you need to effectively replace highways. Unless everyone in their big car pays up front for decades of highway use (this is HN, so maybe we can use an ICO, why not), you're going to need cash up front - the kind of cash that governments that want their constituents to live in a nice place and prosper therein are willing to provide.
(Someone double check my math and assumptions, please, I'm neither a civil engineer, nor a city planner, nor an economist.)
Tunneling in the US is really expensive but many projects have been done for a billion per mile (China is like $100 million per mile). The Big Dig is the highest priced tunnel ever done and not the expected cost of tunneling. If we could do $100 million per mile from San Francisco to Palo Alto (~45 miles), that is $4.5 billion. Charge the $20 each and make ~$1 billion a year. This would be a no brainer. Even at $1 billion a mile and $45 billion total it would almost make sense financially in the current low interest environment. For keeping people sane in the Bay Area, it would definitely be worth it.
