String Rail - a low cost, low impact, high speed transport alternative

[Mapleson]

Reliability of computer guided cars is not the only strike against a robocar/car-share paradigm. The vehicles, and infrastructure are inherently heavier and more expensive, and thus more energy and resource intensive. The fact that they operate on a plane in mixed traffic limits safe operating speed, increasing travel time. Operating in mixed traffic with imperfect traction will also lead to increased headways over human operators.

(1) The infrastructure for a robocar is no different from a human-operated car.
(2) Hybrid cars are have additional components compared to traditional vehicles, yet they are more energy efficient, why then must a robocar be less efficient?

(3) All known transportation systems work on a plane. Why is this a special "strike" against robocars?

(4) The condition of mixed-traffic operation is set when creating a system. Compare apples to apples.

(5) Human operators cannot safely operate with the same or lesser headways that automated control. The "2-second rule" on highway driving is 1 second for you to react and 1 second for the car to decelerate. Or do you still believe that human reaction time is superior to machines?

 

[afransen]

(1) Precisely. Roads are hugely expensive, and only exist to the present degree because users don't directly pay their cost. Continuing with (robo)car-oriented transportation will require hundreds of billions in incremental infrastructure maintenance and construction costs over the next few decades. Anything that can reduce car traffic will reduce the wear and tear on the infrastructure and the demand for new lane kms from being built.
(2) PRT doesn't need to lug around batteries or a powerplant. Robocars will always need to. Robocars need to be more crashworthy because they operate in mixed traffic. That's why I said robocar technology is inherently heavier than PRT technology can achieve.
(3) How about airplanes? All kidding aside, PRT makes it cheaper to create overpasses. Overpasses built to carry 40 ton trucks are pretty expensive. See (1).
(4) Are you suggesting we put robocars on dedicated, grade-separated ROWs? If so, congrats, it's a full-fledged PRT system (not necessarily a very good one).
(5) I don't (and have never) believe human reactions are better than a computer could achieve, by any stretch. I'm referring to brick-wall stopping regulations, which are common. I can't imagine any robocar system being allowed to operate with less than this headway, given imperfect traction and mixed traffic. It's a liability nightmare. Consider robocars on a multi-lane highway with 1 second headways. If a car unexpectedly loses control (let's say a tire blowout), instant pile-up involving upwards of 4 vehicles. It's simply not safe enough, and has nothing to do with the quality of computer control.

 

[rbt]

(3) How about airplanes? All kidding aside, PRT makes it cheaper to create overpasses. Overpasses built to carry 40 ton trucks are pretty expensive. See (1).

Presumably a PRT vehicle needs to carry 40 ton trucks too; or do you envision a grid of this system in addition to the existing roadway system?

 

[Second_in_pie]

That's the point. PRT working as a full on replacement to roads would be a huge, huge, huge capital investment for some trivial gains (since we're assuming that you're not building PRT infrastructure all through every little subdivision cul de sac and so people will have to walk or carry their goods to their house.) Using PRT as a type of transit really doesn't work either, because you've got much more economical ways of providing transit to lots of people (I'll bet that in real world costs, PRT would be on the same capital investment level as an elevated subway.)

 

[afransen]

Nope. Roads will continue to be necessary, but to a lesser degree. I suspect a fairly advanced PRT system would see more, smaller freight vehicles.

That's the point. PRT working as a full on replacement to roads would be a huge, huge, huge capital investment for some trivial gains (since we're assuming that you're not building PRT infrastructure all through every little subdivision cul de sac and so people will have to walk or carry their goods to their house.) Using PRT as a type of transit really doesn't work either, because you've got much more economical ways of providing transit to lots of people (I'll bet that in real world costs, PRT would be on the same capital investment level as an elevated subway.)

Of course PRT isn't door to door, but it's certainly feasible to get people within a couple hundred meters of their destination for many of the trips they would make. Of course you're not going to carry a sofa or a refrigerator on a PRT vehicle, but you're not going to put it in your car, robo or otherwise, either.

You're making unsubstantiated claims about how much a PRT system would cost. Claiming it must cost as much elevated subway is patently absurd, from an engineering perspective. How could there be no incremental cost over building infrastructure to carry ~1 tonne vehicles to carry a 40 tonne subway train*. Do you really believe that?

Edit: I made a mistake: apparently each car is about 40 tonnes. So for a 6 car train, that works out to closer to 240 tonnes.

PRT isn't about ripping out our existing road infrastructure, but diminishing the need for new road infrastructure. Nor would PRT replace all transit, but rather it would complement it. The level of road infrastructure investment to provide a similar service level to PRT in a region like the GTA would be astronomical. We simply don't have the land. If we don't invest huge amounts of money in new roads, we will experience crippling congestion over the coming decades.

In terms of real world costs, Heathrow has built a pilot PRT system. The cost has not been unacceptable for BAA to propose significantly expanding the system provided it operates as planned. The system when built out should cost 3 million pounds per km including vehicles. You're suggesting that this estimate is off by a factor of roughly 30 (to be fair, some of the track is at ground level, but a significant amount is elevated). The system is also expected to cover both operating and capital costs, with positive NPV at 6% discounting. Maybe they are optimistic, but I doubt they're off by a factor of 30. This is only baby steps, and far from a PRT system I would like to see rolled out in urban areas--it certainly has plenty of flaws. But as a proof of the technology, and it's ability to provide a high level of service at low total system cost, I think it's pretty valuable.

 

[Mapleson]

The difference is that you need 1 240-tonne subway train to carry 1000 people, while you need 500 1-tonne PRT vehicles to carry 1000 people.

 

[afransen]

The load is spread out, which allows the infrastructure to be lighter. Much like how stepping on onel nail results in a punctured food, but standing on a bed of nails does not. The maximum load on a 30 m span might be, say, 10 tonnes for PRT, but for an elevated subway the maximum load would be upwards of 100 or 150 tonnes.

 

[Second_in_pie]

The difference is that you need 1 240-tonne subway train to carry 1000 people, while you need 500 1-tonne PRT vehicles to carry 1000 people.

Remember though, you're not going to get anywhere near 100% efficiency with car placement, so there's going to be a lot of empty cars going around to connect with waiting passengers. I wouldn't be surprised if it was more like 700 or 800 cars on average for 1k people (in a reasonable amount of time.)

 

[afransen]

You're probably right that average occupancy would likely be in the 1.2 occupant range. Please note that subways do not always operate at crush loads, either. No point in getting overly bogged down on this point. Suffice to say, that elevated heavy rail would be at least an order of magnitude heavier than any reasonable PRT, which I guess explains the 'heavy' in heavy rail.