Dow Constantine & Denis Hayes: Self-Driving Cars Are a Distraction

Comments

1
Everyone likes cars because we are conditioned to like cars. Very few people care about energy, alternative or not.
2
The technology is years away. If transportation is the focus then the money is better spent on infrastructure. It's not as sexy but needed.
3
Each car does not require one Watson computer. If a Watson-class computer can control 60,000 cars, then the energy expenditure is the same. That may seem like a lot, but if it were designed specifically for the purpose of controlling cars then maybe it is not unreasonable.

Cars take their form from their function. If we have self driving, electrically powered single occupancy vehicles that were at least as safe as today's cars (and hopefully orders of magnitude better), they could be designed to be much smaller.
4
The thing is, every time a mass transit measure comes to the ballot, the opponents unearth some shiny, new, half-imagined alternative that's ostensibly supposed to make rail look obsolete, as they say "a 19th-century technology."

Back in the 1960s, when we were voting on a subway system, the alternative was jet-powered buses, if you can believe that.

Back in the 2000s, when the monorail and ST2 were on the ballot, the talk was of PRT (personal rapid transit).

Now it's self-driving cars. Never mind that, while 19 years from now we'll have a second light-rail tunnel giving a fast, reliable route between downtown and Ballard and opening up the possibility of other detsinations, it's hard to imagine that 19 years from now or 39 years from now we'll have some web of self-driving cars that's a breakthrough improvement over our current road system. And never mind the technology aspect but the economics of it and how this is going to happen in a way that doesn't lock out anyone but the most affluent.

But here's the irony in all this. The movie star Paul Newman was once asked why he'd never cheated on his wife Joanne Woodward. He said, "Why go out for hamburger when you can have steak at home?" Even if, miracle of miracles, the full far-fetched vision of self-driving cars worked out as planned and even if, beyond that, it benefited the masses, we still wouldn't have something that provides as good an experience on the major corridors as grade-separated rail or would be as effective at shaping urban development.
5
... and people won't buy them. Teeny cars have been available for decades, but the Minis, Sparrows and Four2s haven't swept all before them.
6
gcm @3, forgive me for throwing cold water on a fun topic, but we're having a public policy discussion. I think what you're interested in is more suited for slashdot or Scientific American. While you're at it, perhaps you can speculate on the feasibility of human space colonization.
7
Also, debating whether the trains will go anywhere useful, versus merely checking boxes off an arbitrary list while being difficult to access and contributing embarrassingly little to mobility, is "a distraction".

Also, questioning whether Sound Transit's modeling can be trusted when it pursues objective worst design practices, yet pledges Seattle will have the 5th-highest rail ridership in the U.S. (ahead of Philly and the entire Bay Area, some of whose trains actually, like, go places), is "a distraction".

Also, discussing whether 54 billion dollars are being well spent is "a distraction".

(The personal attacks that will now come my way for daring to question Dow's facile political framings will not, of course, count as "a distraction".)
8
(p.s. I do not give a shit about self-driving cars, which are legitimately problematic for a host of reasons. I do care about not making the entire political and regional planning discourse revolve around identity politics and canards, which seems to be the chief ST3 debate strategy.)
9
For what it's worth, I do support ST3. The reality is that self driving vehicles are already on the road and will be more widespread soon. It would be foolish for us to write that off as a red herring instead of planning for their existence. Trotting out power usage data in a misleading way doesn't help much.

How about a fleet of self driving buses or vanpools? As the technology matures, we will see it showing up more in the public sphere.
10
gcm @9, OK, I get where you're coming from. No disagreement here.
11

What we should be asking is: What about self-driving buses?

In any case, to fully realize benefits both self-driving cars and buses need not only communication with each other (vehicle to vehicle - V-V) but with roadside infrastructure (vehicle to infrastructure V-I). Putting that roadside infrastructure in place is expensive, it's a public expense, and it's years away.
12
Many say that self-driving car will help transit as it will serve the last mile problem. Sounds plausible to me.
13
How many misunderstandings can fit into such a short article? Let us count the ways:

"The problem is not that there is a driver or not, but that the car will still take up the same amount of space and cram the same streets." First, self-driving cars will not require nearly the same amount of space for parking, which is a huge portion of urban space dedicated to cars currently. Second, they can be much smaller than current cars. And third, they require far less lane space to safely drive than humans (a fleet of self-driving cars could move faster on a 2-lane highway than humans can on a 5-lane highway).

“The brain uses energy at a rate of about 12.5 watts of power," said Hayes, "And Watson, the IBM Jeopardy champion computer, draws about 750,000 watts of power." Self driving technology is not the same as IBM's Watson. For one, it can scale. For two, it's a different technology altogether and why is this comparison even being drawn?

"And why is there more excitement about them than about cars that use little to no fossil fuels or alternative modes of transportation?" Nobody takes "self driving car" to mean "fossil fuel powered self driving car." The two technologies are being developed in tandem.

"Where is the gain? Some say the gains will be found in increased safety and less accidents."
Yes, "some," meaning "all." 30,000 people die in auto accidents in the U.S. each year. Stop for a second. Think about that. Even a 10% fatality reduction, which would be extremely conservative, would save as many lives per year as were lost in the 9/11 attacks.


14
This isnt about self driving cars. Self driving technology means public transportation can become much more personal, available and inexpensive.

It will take cars off the road. Many people will find zero reason to even own a car.
15
Time to cancel that unfounded over budget Bertha Tunnel we don't need and get real
16
"So, what is all of the excitement behind self-driving cars about?"

It's about self-driving cars following the law 100% of the time based on their programming, rather than drivers who say 'fuck it' and create traffic by blocking intersections, inconveniencing pedestrians by blocking crosswalks, endangering cyclist by trespassing into bike lanes, etc.

This doesn't have to be one-or-the-other: I'm excited about the prospects of our entire transportation enviroment being transformed by enhanced transit AND the widespread use of autonomous vehicles.
17
Dow's foot ferry is a distraction, too.
Take the bus, Cancel the costly ferry.
18
@13

"Second, they can be much smaller than current cars."

Sure, and traditional cars can also hypothetically be much smaller than they currently are. Why would the advent of self-driving car technology make Americans want to travel around in Smartcar-sized vehicles? Do you really think people still won't want 8-passenger SUVs or giant pickups to cart their families and stuff around?

"a fleet of self-driving cars could move faster on a 2-lane highway than humans can on a 5-lane highway"

...what? Do you have a citation for that? The idea that autonomous driving technology could increase road capacity by 150% is a pretty outlandish claim. As far as I know, this technology hasn't been tested or implemented on a scale that would make estimates like this remotely possible, and the benefits wouldn't be seen until conventional cars were banned from roads, which optimistically would take decades.

How would this happen? From following more closely? Because of the laws of physics, cars moving 60mph still need 150 feet to stop no matter who's controlling them, which means that minimum spacing would be 200-300 feet, which isn't much less than the 3-second spacing they tell you to use in driver's ed.

There will be plenty of benefits to self-driving cars, including safety and convenience, but the idea that we should shape our urban planning decisions around the hypothetical benefits of untested technology is insane.
19
skemmis @13: 30,000 people die in auto accidents in the U.S. each year. Stop for a second. Think about that. Even a 10% fatality reduction, which would be extremely conservative, would save as many lives per year as were lost in the 9/11 attacks.

The whole context of this post is that self-driving cars are being presented as an alternative to building grade-separated mass transit. So if you're going to go into full 9/11 demagogue mode that somehow there's blood on our hands for our not embracing a future where self-driving cars are the dominant mode of transportation to the exclusion of everything else, let's also keep in mind that those 30K+ lives lost a year to motor vehicle fatalities are just as much a function of our having unwisely invested in an auto-dependent transportation infrastructure. Let's see what the equivalent statistics are in Europe. We could save lives too by investing in infrastructure like Sound Transit 3 and in creating more walkable communities and, heaven forbid, traffic calming. Or does your interest in saving lives end at the water's edge of a self-driving cars agenda of untold price tag or timeline?

#LivesSavedByTransitMatter

Oh, and I'll clue you in on a little something. The American people don't care so much about those 30K+ lives anyway. From a Nicholas Kristof NY Times column from 2004, 117 Deaths Each Day (those deaths being auto fatalities):
''If the United States could achieve Sweden's current standard, this would save 12,500 lives per year,'' the authors say.


And I happen to appreciate the nuance offered by slvtrdlrn @18, nospin @16, and even Not Good Red Herring Either @12. It's not an either-or.
20
I think there is a lot of stuff on point here but I think one point that's being missed, especially wrt to smaller cars, is that automated vehicles will likely eliminate the need for owning a vehicle. On demand cars that can take you anywhere is where we are going.

Also, carpooling will be come much easier and efficient. You summon a car and it can figure out the optimal way to get you and everyone around you to your destinations in the fastest way possible.

All that being said, I definitely support ST3 and don't think one supersedes the other. Both are necessary.
21
Denis Hayes is well-known for his absurd predictions about the future.
22
Let me just state that I do not think "self-driving cars" (or even "buses"...wft?) are not the solution to Seattle's transit issues, grade-separated mass transit is with high "people miles per vehicle" is the solution. Until we develop home teleportation units.

@2 - The technology is years away.
Not really. It's happening faster than you think. A $1000 "add-on" self-driving thingy will be available next year... for any car.
23
I have no idea what the age of driverless cars will bring, but the goal should not be to solve traffic, the goal should be to give people the option to get from one place to another as easily as reasonably possible, and trains—ideally underground or grade-separated—can achieve that.
24
erg.. eliminate the double-negative in sentence 1. Where's my editor? They're fired!
25
@20 - Automated vehicles could greatly reduce personal ownership one day—at least in urban areas (assuming urban living remains fashionable), but cars are so woven into the American psyche as status symbols (almost preceding their utility), it could take generations.
26
@20

Being able to summon a car on demand to take you where ever you want isn't some new byproduct of driverless car technology. You just described taxis (or Uber, Lyft, etc). They've been around for a long time, and they haven't usurped private car ownership or transit because they're expensive, inflexible, and inaccessible to the poor.

I'm sure taxis and ridesharing services will be among the first to use driverless cars on a large scale, but there's no reason that the average person would want to rely solely on taxis for transportation just because they're autonomous.
27
A lot of the benefit from self-driving cars, as far as traffic goes, is that they'll theoretically be cooperating with the other traffic around them, so that stoppages are less likely to happen (platooning is the term I think they use for one effect). Human frailty and impatience cause a lot of the smaller traffic issues these days. Theoretically parking will be less of an issue as well, because your car can go park itself further away than it would have before, while not being any less convenient to get to since it will hopefully come when called.
28
Oh no, not this again. Sorry, but if you follow transit blogs (both local and national) you can find numerous articles about the issue. Here is a quick summary:

There are multiple stages when it comes to automation. We are actually at the first stage. You can get cars that park themselves, or drive for short distances. Big deal. All that does is make cars a bit more attractive (as say, a good car stereo does).

The second stage is automation, as in, you don't have anyone at the wheel. As mentioned earlier, this could also mean bus automation. Like a lot of automation, this probably won't completely eliminate the user. For example, checkout lines at the grocery store still require a cashier, just one cashier for many lines. Likewise, I expect fast food to go this route pretty soon (one or two people in the restaurant, not a dozen during lunch rush). For buses this is tricky, but you simply add remote controls. So now a dozen buses are driving themselves through town, but they get freak out occasionally and come to a complete stop (as do Uber cars). Someone sees this on the monitor, and steers the bus. Same with fare evasion (caught on camera) or security concerns (a bus driver shouldn't be a cop anyway).

That is basically the next step, and it really isn't that far away. From a transit perspective, this is huge. By far the highest cost for running a bus route is the driver. Side Note: Metro, right now is struggling to hire enough drivers, and service is struggling as a result. If you eliminate, or greatly reduce the cost of running a route, then you can add a lot more service. But it gets better. Right now most of our buses are big buses. It costs more to run big buses. But once you pay for the driver, it doesn't cost that much more. That is why, for example, we don't have a huge fleet of vans.

But if that cost goes way down, things change dramatically. If you increase frequency on a bus route, it becomes a lot more popular. Frequency is one of the most expensive, but most important factors in improving service (one often overlooked, because there is no comparable problem with driving). Transfers as well as simply a ride on the bus is completely different if a bus comes every five minutes, instead of every half hour. At some point, in less densely populated areas, you will never fill up a big bus, even with the added riders. But you can fill up vans, and again, if the cost of a rider is minimal, that might actually be less expensive.

The third stage is to have cars wired to each other so that they can manage their speeds properly. This is a very long ways off, since you would need just about (if not all) of the cars on the road to be wired in this way. Besides, if the cars are synchronized, there is only so much you can do. A bus, or a train, is simple more efficient. With zero space between vehicles, a bus or train carries way more people.

Personally, worrying about that third stage (or dreaming about that third stage) seems silly. Who cares, really. As far as Sound Transit is concerned, I think it is meaningless.

But in twenty years, it is reasonable to assume that there will be a lot of automated vehicles, especially automated buses (and trains). So you could argue that investing in a large light rail project of dubious value is misguided. We might be better off investing in hundreds of vans, and the new technology that will make running them far cheaper.

I think that misses the point. Investing in a good subway system is worth the money. You won't here me arguing against the Second Avenue Subway (in New York) or the Millennium Line extension on Broadway (in Vancouver). I voted for (and enthusiastically supported ST1 and ST2). But ST3 is crap. It is a stupid, poorly designed subway that ignores both common sense as well as a hundred years of mass transit history.