« 3D Print-to-Order | Main | Tuesday Topsight, July 29, 2008 »


Were-Car.pngBrad Templeton wants you to stop driving.

Templeton (Chairman of the Electronic Frontiers Foundation, programmer, dot-com entrepreneur, inventor of the "dot com" domain name structure -- no kidding! -- and more) laments the tens of thousands of people killed every year in traffic accidents, the waste of urban space for parking garages and gas stations, and the various institutional roadblocks to moving to renewable energy systems. But he doesn't suggest that you go get a bicycle, you lazy bum, or spend hours on packed public transit. He wants you to get a robot.

A robot car, to be precise.

Brad Templeton's set of essays, under the collective title "Where Robot Cars (Robocars) Will Really Take Us," explains exactly why robot (autonomous-driver) cars are possible, likely, safer, cleaner, and all-around a good idea. This isn't meant as a nuanced thought experiment; Templeton lays out page after page of statistics, arguments, and data. This is a massively detailed piece. If you think of an objection, chances are he's already covered it.

(Disclosure: Brad sent me a link to an earlier version of this piece, and I sent back numerous comments.)

Templeton doesn't make any claims that this would be easy, or that it could be done soon. As a professional programmer, he's well-acquainted with both the risks arising from relying on computer controls, and the difficulty of putting autonomous systems on the road alongside human drivers. He sees these as solvable issues, though, and points to present-day examples of extremely reliable coding and the "Darpa Grand Challenge" for automated drivers as reasons why. The social (particularly the legal-liability) issues are less-easily solved.

Probably the most provocative aspect of this piece is Templeton's effort to play out some of the consequences of a shift to robotic vehicles. Not only would autonomous vehicles allow for major changes to urban design (don't need downtown parking if your car can come when you call) and major reduction of accident rates (crash-avoidance would be the first form that car automation would take, potentially eliminated tens of thousands of crashes per year, saving hundreds of millions of dollars), we'd likely see the end of mass transit (with a few long-haul exceptions).

(His data on the overall energy efficiency of mass transit, versus standard, hybrid, and ultra-light automobiles, is startling.)

I suspect that both technophile and envirophile readers will find aspects of Templeton's piece to argue with, but I suspect you'll be surprised at how strong and reasonably well-supported most of his claims are. This is the kind of piece you go into thinking that it's all crazy, and come out thinking it's all quite plausible.

Do I believe him? I think he lays out a pretty compelling scenario. I do think he still under-estimates the social, cultural, and legal inertia likely to slow the rate of acceptance of such systems. This strikes me as almost certainly a generation-change issue -- that is, the rate of acceptance will map to the maturation of kids growing up riding in semi-autonomous vehicles. Lots of resistance for longer than expected, then boom, a phase shift.

But I doubt it will happen first in the US. Singapore, maybe Scandinavia, Japan almost certainly... but I expect USians to be watching this from afar.


I had also read Brad's articles on robocars but I don't think that he played up the land use changes very well. LA uses more than 50% of the land to roads and parking if you could cut that percentage to ~35% the cities income should greatly increase because the amount of usable land (and the taxes that derive from it ) has greatly increased, that can be a big incentive for city government to get on board.

As for social inertia if the cost of gas continues to climb like it has been for the last couple of years, 8 dollar a gallon gas will motivate a lot of people into action.

I just hope the software isn't written by Microsoft.

Thanks, Jim, I had meant to write about that, and have added a section to the "notes" page to talk about land re-use.

It will be interesting because as we re-use parking lots for new purposes, this would lead to denser cities, supported by the higher traffic capacity of robocar roads. Though only up to a point of course.

L.A. can certainly get denser without much trouble though. Meaning shorter trips, more walkability, greater transportation efficiency.

It would be interesting to see those adopted in some of China's newly developed eco-cities, where land could be set aside directly for robocars. I think it unlikely we will see those share the road with human-driven cars until they have been deployed a few times in limited environments.

I think a big barrier to this is the illusion of freedom a car gives you. A big part of that comes from being behind a wheel you move, even though driving in traffic is the equivalent of sitting for an hour in a supermarket checkout queue.

This speaks to something I have been thinking quite a bit about lately. Yes, we get our electric/hybrid cars. Yay! We still get to sit in stop-and-go traffic in them. I'll read Brad's post at lunch and comment directly.

One thing that struck me was Brad's graph of energy used by various transport modes.

In particular, the astonishing indication that electric scooters are markedly *more* efficient than pedal power!!?

Does the figure take into account all energy used by the cyclist anyway, or just the additional pedal power?

(Is the corollary that we should save the world by not exercising??;-)

It's a topical point for me: I'm just getting to grips with a newly installed electric motor for my bike (Here, if anyone's interested). So, maybe Brad's data gives me the conviction to persevere?

Is a car the only means to freedom? Spend time in New York or Hong Kong or many cities where car ownership is rare, and many people don't feel too restricted in freedom. Now improve the equation, where getting a vehicle is reliable, fast and cheaper than the subway, and the rid is comfortable with few stops.

Would you feel the freedom, or feel constrained? Never having to worry about parking or refueling or traffic?

As to energy of the electric bike... It is no surprise that the human rider burning food is less efficient as a motor than an electric motor driven by a battery -- in fact they are fairly similar, which is what surprises me. The electric motor gets 3x less efficient when you consider the cost of making the electricity in a coal plant, and the human gets less efficient when you consider the cost of making and shipping the food from an agribusiness.

Brad, as you indicate, the back-story can get quite intricate.
I'm not surprised about the relative efficiencies of a human vs an electric bike motor (at least, not when I think about it!). I was just wondering whether the figure you used for pedal power was for total energy expenditure (ie inclusive of the number of calories a person would burn while sitting in a car), or just for the extra effort of pedalling.

There are a bunch of figures for cyclist. There's the raw food energy (Calories) used to cycle, which is comparable to electric energy use. There's the incremental energy used (which I calculated at 38 Calories per mile at 10mph)

And then there's the energy used in agriculture to generate that food. And if you want to go a bit further, the energy used to make that energy.

As pure converters of food to motion on a bicycle we're pretty efficient, but our food chain isn't.

There's also the issue of how the metabolism works eg a friend of mine, a keen cyclist, was laid up with an injury for a while. His remark was that his food intake didn't vary at all. So, it's all a bit fuzzy.

Thanks for your response to what is, after all, a very minor point to your thesis.


Creative Commons License
This weblog is licensed under a Creative Commons License.
Powered By MovableType 4.37