Thursday, August 15, 2013

Hyperloop and autonomous driving versus betting the farm on high speed rail

We'll eventually find out whether Musk's Hyperloop concept will come to something. It's hard to imagine some of the claims, particularly about costs, will be anything other than underestimates. Just building tunnels to get through mountains north of Los Angeles and maybe other areas will rocket up the costs. I've also been following and peripherally-involved with California High Speed Rail for years, and I know that simply putting your structure up on pylons, which has also been proposed for HSR in places, isn't a guaranteed way to avoid land use conflicts.

Still, for the present, I view Hyperloop as similar to autonomous driving in its effect on our proposed HSR - both concepts have an known unknown chance of making HSR become a huge mistake. Autonomous driving might not be able to go 200 mph, but the cars could pod up together on a dedicated or separated highway lane and drive well over 100 mph. That improved drive time across California combined with the ability to use the time productively because you don't have to drive the vehicle could pull a lot of the market share away from HSR. Or not. Maybe they won't be all that fast or take longer to realize than I imagine.

But if you assigned some number to the risks, like a 15% chance that the Hyperloop will kill HSR before it pays off, 25% chance that driverless cars will do the same, and 30% chance that costs will be so much higher or income revenues lower than anticipated so that it doesn't make sense, that all results in a 55% chance that HSR will be a failure just based on those factors alone. Of course I pulled those risk numbers out of the air but they don't seem unreasonable.

I've been cautiously supportive of California HSR although I don't think it's popular in my area. We need to get people out of planes, soon. Almost everything I hear about it just makes me more hesitant though.

Betting a huge amount of money on a single piece of technology, no matter what it is, strikes me as un-hedged.

Unrestrained criticism of Hyperloop is here, YMMV.

11 comments:

badger badger badger said...

More analysis here:

http://greatergreaterwashington.org/post/19848/musks-hyperloop-math-doesnt-add-up/

As in the critique you post, I'm annoyed why this gets any wide attention at all.

{new, clean URL, thanks Susan}

dhogaza said...

I read it as simply being an attack on HSR. Musk is simply saying "we could build something 3x faster for 10% the cost in the same timeframe, here's my back-of-the-envelope calculations proving it, now let's stop the HSR project and NO, I'm not interesting in building the hyperloop, I'm too busy. Just stop building HSR, please".

Is there something unique about CA (other than automobile worship) that means HSR over a distance similar to Madrid-Seville, with the same time benefit vs. cars as that route (2:30 for HSR, 5-6 hrs driving), and shorter flight times will fail when that similar route is pretty much booked solid in Spain?

One possible drawback is that once you arrive downtown, both Madrid and Seville are easy to navigate without a car, whereas LA ...

dhogaza said...

HIs claims about HSR are also simply lies, i.e. "slowest bullet train in the world", even his reworked "one of the slowest" is wrong. Across Spain, at least, HSR is built out pretty much the same way as the CA proposal, i.e. they run at high speed through the empty parts of the country but thus far the huge investments that would allow them to avoid slowing down to commuter rail speeds within the urban/suburban sprawl zone around cities such as Madrid hasn't really been done.

Still, you get a 2x advantage over driving, arrive downtown rather than a 25-50€ taxi ride (or lengthy subway ride) downtown from Barajas (the airport), have much more comfortable seating, much less security hassle etc than airline travel.

The link provided in the OP points out one thing that surprisingly I've not seen elsewhere (though I noticed it immediatley) - seating in the 7 ft tube looks horribly uncomfortable, one is semi-reclined in a narrow, cramped space.

I wouldn't be surprised if he's underestimated the costs of R&D + construction + deployment by two orders of magnitude, either.

Fixed Carbon said...

Hyperloop in the straight line then Rapid Ultrilight Rail for the local stuff. The latter might be betaed in Davis http://davisvanguard.org/index.php?option=com_content&view=article&id=7548:new-transportation-options-a-paradigm-shift&catid=88:environment&Itemid=205

Steve Bloom said...

California HSR is all about facilitating more sprawl in the Central Valley.

Anonymous said...

Canman says,

While airplanes use a lot of fuel, they are otherwise resource thrifty. All you need are planes and airports. There's very little infrastructure maintenance.

According to UCS, buses appear to be just as "green" as rail and run on existing infrastructure that you have to have anyway.

http://www.ucsusa.org/assets/documents/clean_vehicles/vaca_trav_carbon_guide.pdf

Anonymous said...

"autonomous driving might not be able to go 200 mph, but the cars could pod up together on a dedicated or separated highway lane and drive well over 100 mph."

As safely as high speed rail?

The issue is not whether one can drive "well over 100 mph", but whether one can do it safely -- and as safely as high speed rail.

Brian continues: "Of course I pulled those risk numbers out of the air [or someplace much darker and danker] but they don't seem unreasonable."

Without reference to actual studies done by people who actually know something, statements like yours are essentially vacuous.

But when did that ever stop you from speculating before, right?

Brian said...

Last anon, I expect that dedicated and especially physically-separated lanes for autonomous driving would be very safe, much safer than current driving and possibly safer than flying. Probably not as safe as rail, but that doesn't seem like a primary factor.

As for your demand for studies, go find them yourself. I think you'll find no one can definitively quantify the risk from Hyperloop and autonomous driving to HSR, and the existing economic forecasts are extremely controversial. Sorry I can't help you.

Anonymous said...

"I expect that dedicated and especially physically-separated lanes for autonomous driving would be very safe, much safer than current driving and possibly safer than flying.'

You were specifically comparing autonomous driving to high speed rail, not "current driving".

Remember?

Your title was "Hyperloop and autonomous driving versus betting the farm on high speed rail".

Besides, what you "expect" may not be the same as reality*

Hence my request for studies providing actual evidence.

*I "expect" that evidence of the safety of pods of cars traveling at "well over 100 mph" (even on dedicated lanes) might be a (just slightly) significant issue in getting people to use them.

What the point in even posting on this stuff if you are not willing to provide evidence?

I mean, seriously. This is supposed to be a science blog, right?

THE CLIMATE WARS said...

Accidents will happen. and if you have on the order of 1000 KM of Pneu de grand vitesse each 1000 KPH crack-up may take out several hundred meters of very expensive track.

Apart from the large cost of replacement, repairs may take weeks to months, seriously degrading the load factor .

Subsonic mail tubes do not self destrct if plugged.
Gun barrels are another matter.

David B. Benson said...

I'm an engineering ex-professor. I think the concept is feasible, enough so that DoT ought to pay the national laboratories enough to research far enough to decide about development. I suspect that will take about 5--7 years.

At that point one can adequately estimate the LA<->SF route costs. I'll guess that stopping in Oakland and using BART the rest of the way makes more sense. I opine the cost has been underestimated by a factor of 4--6.