Automated Driving
My friend Tim Lee has an article up at Ars Technica on a subject I’ll admit I haven’t given much thought to — what would/will the social consequences of self-driving cars be?
On the second page he makes a number of observations that are relevant to my transportation and planning interests. One is that if cars didn’t need drivers, then taxis would become relatively cheaper. You could imagine quickly and easily ordering a self-driving cab from your cell phone or some such. Tim says that “when self-driving taxis are readily available, many people—even far from dense urban areas—will find renting both cheaper and more convenient than owning a vehicle.” I’d be a little bit skeptical of that, but at a minimum I think what you can say is that you’d see much lower rates of car ownership among people who do live in-or-near fairly dense areas — the sort of places where you could expect such cabs to be widely available.
Perhaps more interesting is the idea that self-piloted vehicles could revolutionize our understanding of parking. To make drivable suburbanism viable, each person needs one dedicated parking space right adjacent to his or her house, plus another dedicated parking space right adjacent to his or her office, and then on top of that each retail establishment needs to be immediately adjacent to a number of parking spaces roughly equal to the amount of parking needed at peak demand times. But obviously most of the time it’s not peak demand at the mall. And most of the time your car isn’t parked at the office. Each car is maintaining a huge space footprint. On top of that, in order to prevent free riding on people who’ve already built parking lots, regulatory schemes are put in place demanding that all new construction involve new parking facilities meaning that space is used less-and-less efficiently.
But a self-driving car could drop you off at work and go back home. What’s more:
[A]s Brad Templeton points out, parking lots would work differently if cars could move themselves at a moment’s notice. We can only park today’s cars in places where we’re sure they won’t be in anyone’s way. But cars that can move themselves could park in lots of places—in front of driveways or fire hydrants, in strangers’ driveways—that ordinary cars cannot. If they find themselves in someone’s way, they can quickly move and find somewhere else to park. The same point applies in parking lots. Self-driving cars can safely double- or triple-park, dramatically improving space utilization. When a car needs to leave, it can signal its neighbors to move out of the way and let it get through.
Then with less space dedicated to parking lots, development will naturally become denser which further makes a model more oriented around some walking plus some taxis rather than individual car ownership look appealing.
I’d want to think about this more, but it’s all very interesting.
55 Responses to “Automated Driving”
Register
RSSLogin
Topic Cloud
ArchivesBlog RollAbout Matt YglesiasContact Me
Donate
October 14th, 2008 at 10:18 am
Matthew,
There are already designs for driverless pods on rail. Individual cars that you treat as taxis essentially though they have predetermined paths.
http://www.motorauthority.com/driverless-podcars-to-revolutionize-urban-transport.html
October 14th, 2008 at 10:18 am
I think this will happen eventually, but there will be tons of pushback. The most vocal will be the first time one of these cars has an accident– imagine the first time a family of four is killed with one of these things. Now, the rational among us will say “these are safer than human driven cars, self-driven cars only kill X people a year, human driven cars used to kill 165 Americans a day, etc.” But people aren’t rational and won’t care.
October 14th, 2008 at 10:26 am
They won’t let cars drive themselves until they’re significantly safer than human-driven cars. So you gotta figure it won’t be long after cars can drive themselves that they won’t let humans drive cars anymore.
October 14th, 2008 at 10:26 am
There’s bound to be pushback, but I think it will scarcely register against the overwhelming demographic need for driverless cars in a graying population. Private automobiles are, from a social justice point of view, a terrible way to move a population around, since they’re only available to healthy, young (but not too young!) people with money. Removing the driver removes two of those three requirements.
Looking at recent advances in the technology, I’ll be very surprised if it’s another thirty years before people are able to buy a car with some kind of substantial autonomy.
October 14th, 2008 at 10:28 am
By the way, Jim–couldn’t agree more. First step will be to outlaw manual driving on the Interstates and other high-speed highways. Give it 20 years or so after the first autonomous car appears?
October 14th, 2008 at 10:30 am
Wouldn’t this at least double our fuel usage? Instead of one trip to and from work, the car makes two trips, one to drop us off and one to pick us up. It would help that half the trips are with no passengers, but all the additional driving would *ahem* drive the number up.
October 14th, 2008 at 10:30 am
We are many, many decades away from automated cars being able to navigate city driving, y’know.
This is a “what are the implications of vacation condos on the moon?” kinda topic…
October 14th, 2008 at 10:32 am
What about the carbon footprint of 4 car trips per day between home and work? Are you presuming that we’ll be beyond gas-powered vehicles when we get driverless cars?
October 14th, 2008 at 10:33 am
Freddie’s right. People bite on the advertised anecdote and ignore the statistics. It’s how the Republicans/AMA/insurance companies keep us in thrall.
October 14th, 2008 at 10:40 am
Traffic will also go up in areas where parking is tight as people set their cars to just drive around the block repeatedly until they’re done running their errand.
October 14th, 2008 at 10:45 am
How is it possible that I’m the first person to cheer the Total Recall clip? “Cohagen, give ‘zem ‘zee air. Nooooowwww!” Well done Matt.
Second, automated cars bring an awful lot to the table. Being able to communicate both with other nearby cars and with some local road network would almost certainly reduce crashes. It would also reduce traffic, as an automated car wouldn’t slow down to gawk at accidents, or drive too slow or drive too fast, but would coordinate with the cars around them to reach the optimal speed vis a vis the number of cars on the road at that time. Moreover, in the event that a road reaches its capacity, an automated car could redirect to another route.
The real drawback here is what happens if there’s a break in the system. What if your car’s wireless antenae breaks? Or if there’s a glitch in the road software? Or if the entire system comes under attack by terrorists? There will have to be layers and layers of fall backs, including a manual driving mode, to account for all the things that could possibly go wrong. Of course, if people get used to not driving but are forced to take manual control during an emergency, they may actually make it worse.
What I’m most interested in, however, is how automated driving will affect the design of cars. With no driver needed, will we have seats which face each other? What other aspects of current vehicle design are dependent upon the current need for a driver? Also, when are we going to get Demolition Man-style impact foam?
October 14th, 2008 at 10:45 am
Kunstler’s right. Listening to people “plan” for the next phase of the Age of Happy Motoring just drives home (no pun intended) how far we are away from understanding what’s coming next.
October 14th, 2008 at 10:45 am
Driverless cars would really lessen people’s emotional attachment to their cars. Many fewer gas guzzling beasts with sentimental value, more ugly but functional cars.
October 14th, 2008 at 10:47 am
“What about the carbon footprint of 4 car trips per day between home and work? Are you presuming that we’ll be beyond gas-powered vehicles when we get driverless cars?”
By the time we get fully driverless cars, they’ll almost definitely be powered by something other than gasoline.
But that said, parking lots can be closer to home than the full distance of a commute.
October 14th, 2008 at 10:47 am
Jessie,
Why would you need 4 trips? No need for the car to go all the way home after dropping me off, it just has to go -somewhere-. As they point out in the article, one possibility is that your car could rent itself out as a cab while you are at work or, more likely, it would be a cab to start with. My guess is that the first change would be moving from the 2-car family as standard to the 1 car and cabs model.
Petey, RTFA. It’s a lot closer than that
October 14th, 2008 at 10:48 am
This sounds like a disaster for a pedestrians and cyclists. If cars are moving about with little warning people will get killed. The kinds of behaviors we should be encouraging would be discouraged. Safety would be non existent if they did not stop for people and all of your efficiency and traffic improvements would go right out if self-driving cars did stop for people. (How long do you think it would take the entire island of Manhattan to become gridlocked, a half an hour? If you peg their behavior to traffic signals what happens to pedestrians in suburbs and exurbs with dangerously short walk signals?)
As imperfect as it is you need a human response to human factors.
It makes me sad to see smart people go in completely the wrong direction. More better technology will not help ameliorate the social consequences of a piece of widespread technology. You have to change the behavior not improve the very thing that enables it.
October 14th, 2008 at 10:52 am
“Kunstler’s right. Listening to people “plan” for the next phase of the Age of Happy Motoring just drives home (no pun intended) how far we are away from understanding what’s coming next.”
Kunstler is an idiot.
Don’t get me wrong. I’m glad he wrote his book. Eschatology is both fun and useful - if you don’t take steps to avoid major future crises, they will indeed come true. But he really is a bit of an idiot, if you take him even vaguely seriously.
October 14th, 2008 at 11:00 am
With drunk driving no longer an issue, what effect will this have on my liver? And will this car clean itself up when I blow-chunks all over the back seat? These are the important questions.
October 14th, 2008 at 11:08 am
Every futuristic movie which features a self-driving car clearly demonstrates the reasons why we will never have them.
As you had an article about error rates, just apply it to these cars. Even if there are 1/10 the accidents caused by these cars, every accident will be attributed to the same “driver”. We will be unable, psychologically, to ignore the error rate, and who was responsible.
And that is just accidents. How many will suffer malfunctions of some sort, where the passenger gets delivered to the wrong location, the computer reboots in rush hour traffic, the route passes through a “bad neighborhood”, a hacker takes over control of the car, etc. Even a tiny error rate will kill the acceptance of this type of technology.
But a scaled down version which will keep the elderly from plowing through crowds of pedestrians…that would be a winner.
October 14th, 2008 at 11:09 am
“Petey, RTFA. It’s a lot closer than that”
I did read the fucking article.
It’s a pipe dream to think fully automated cars will be available for highway driving exclusively in the next 20 years. And you can expect another couple of decades after that for fully automated cars to be available for non-highway duties.
Highway full automation will be relatively easy. Urban street full automation will have to wait for genuine AI tech to evolve and get somewhat refined.
In short, cars will still have steering wheels in 2050.
Assuming a non-Kunstler-ian future, automated cars are indeed inevitable in the long run, but so are vacation condos on the moon…
October 14th, 2008 at 11:10 am
Kunstler is an idiot.
Petey, you’ve never been more right. When talking about Kunstler, it’s always worth noting that before peak oil, his bread and butter was frantic Y2K masturbation. If I had to describe him in eight words or less, it would be:
He has an onion tied to his belt.
October 14th, 2008 at 11:23 am
It’s a weird time to call Kunstler an ‘idiot.’ It’s not merely his peak oil theories, he predicted this economic crisis (along with others) with amazing precision, even down to naming Bear and Lehman as victims many months (maybe even a year?) before they actually fell.
He also asserts ‘technology is no substitute for energy’ and ‘we’ve wasted all our post-war wealth on building ugly/non-sustainable infrastructure’ without having any true productive american industry.
October 14th, 2008 at 11:23 am
“Petey, you’ve never been more right. When talking about Kunstler, it’s always worth noting that before peak oil, his bread and butter was frantic Y2K masturbation. If I had to describe him in eight words or less, it would be: He has an onion tied to his belt.”
Agree, but will reiterate that I’m glad he wrote his book.
There really is a non-zero chance that something close to his vision will come to pass. The coming fossil fuel clusterfuck of scarcity and global warming is both real, and as much of a threat to human civilization as nuclear weapons were in 1960 - aka a non-trivial threat.
But Kunstler is not a reasonable analyst about the things he’s analyzing.
October 14th, 2008 at 11:27 am
On the contrary, they would be safer. There’s no way we get automated cars that can’t sense pedestrians and bikes. If all automated cars are using similar protocals, they will be more predictable. Right now, there is a lot of guesswork between car drivers and pedestrians (and bikes) about which laws get obeyed and which get disregarded.
October 14th, 2008 at 11:38 am
Petey, you’ll have to be more specific about which Kunstler book you read. He’s written several. And he’s not an idiot. He’s a crank. His 90s books on urbanism, “Home from Nowhere” and “The Geography of Nowhere” were truly groundbreaking popularizations of theories of urbanism and sprawl and they are both truly excellent.
Since that time he’s become a good bit more paranoid. He’s got a point about peak oil, but he takes it to an almost religious extreme–like many Christians he simply can’t wait for his own Rapture.
However, I completely agree with Petey that this discussion is beyond silly. It reminds me of 1950s era predictions of a world where kids travel to school by helicopter and mom’s lives are made easy by dehydrated food products that transform themselves into delicious meals in seconds.
Actually, I disagree with Petey about one thing. Vacations on the moon are not inevitable. Sadly, baring the discovery of a massively powerful free energy source (like cold fusion) I think that will never happen. We’re stuck here on Mother Earth for the duration.
October 14th, 2008 at 11:48 am
There’s no way we get cars that can’t sense pedestrians and bikes? Really. How long did it take for someone in this thread to even mention pedestrians and bikes? Go take a look.
If the emphasis in the development and the deployment of the technology is on “efficiency” and “personal convenience” the public factors, the ones not related to how cars interact with each other or with some impersonal imaginary space but with how they interact with the vulnerable bodies of strangers, will inevitably be given short shrift.
October 14th, 2008 at 11:51 am
Sally, by Isaac Asimov
And then they take over. Be very nice to your robotic car.
October 14th, 2008 at 12:05 pm
I have to say I’m disturbed by your take here. You’re known as one of the leading voices warning against the coming robot hegemony. Have you been bought off? Or even…horrors!…replaced by a robotic Matt?
Please, think of your fellow humans.
October 14th, 2008 at 12:07 pm
Really, shouldn’t we be talking about the massive changes that will be brought about when we get transportation by transporter beam?
All I know is, in those Star Trek movies, they do an awful lot of transporting. Imagine all of the infrastructure we’ll be able to do away with then! I’m just not sure whether a transporter beam is carbon neutral or not.
October 14th, 2008 at 12:14 pm
A car that’s self-driving doesn’t have to be self-propelled - not all the time, anyway. “Cars” might become hybrids of existing automobiles and trains: self-propelled until they reach a major thoroughfare, where train-engine-like “tugs” drive back and forth (perhaps only during peak hours?) GPS and decent computer control would identify optimum routes to latch your car onto a nearby tug wherever possible (probably you can tell your car to optimize a given trip for speed or fuel consumption). The tug network would over time improve and adapt to changing traffic patterns, launching tugs and modifying their routes as needed to pull the optimum number of cars.
This would combine nicely with automated taxis, of course.
October 14th, 2008 at 12:26 pm
Did anyone read the linked article in the series, the first part about the technical viability of these kinds of cars?
The Future of Driving, Part 1
I’m a skeptic about driverless cars, so this was interesting for me. I hadn’t been keeping up with the progress in the DARPA grand challenge, and it sounds like they’ve made some impressive strides. However, this is the money quote:
Montemerlo is the head of one of the teams that was among the finishers of the Grand Challenge. This is his work. He’s both a qualified authority and, if anything, incented to be optimistic about the possibility of autonomous vehicles. So he’s more optimistic than Petey’s “many decades,” but it seems safe to say that driverless cars won’t be here soon enough to really impact much in the way of current practical planning for vehicular travel.
October 14th, 2008 at 12:31 pm
To propose a self-driving car is to regard the car as a form of transportation, which it is only incidently. You might as well propose a self-driving penis.
October 14th, 2008 at 12:38 pm
Like the man says, self-driving cars should mostly be smaller and lighter, because the vast majority of our car time is spent alone, with minimal luggage.
One of the implications of this is that most such cars wouldn’t need to be two people wide. You’d be able to cut lane width in half, which would enable twice as many cars in the same bit of pavement - which would enable us to reclaim some of that pavement back from the cars.
Which would allow for more walkability and greater density.
October 14th, 2008 at 12:47 pm
One partial step in this direction is adaptive cruise control. This intrigues me since, being a bicyclist, I a very aware of the benefits of drafting the vehicle in front of me. It greatly reduces the effort (fuel) needed. With adaptive cruise control set up appropriately, it should be possible for a car to safely follow the vehicle in front very closely. This is a feasible technology today and it could have quite an impact on our fuel usage.
October 14th, 2008 at 12:58 pm
45 minutes.
Somehow, I think it will be more than 45 minutes between concept and implementation of automated cars.
October 14th, 2008 at 1:04 pm
Jesse,
It would reduce energy use for those two round trips to about 1/15th of what it is today. If a trip today uses a gallon of gasoline (equivalent), that future double-trip would use a Cup of gasoline.
In the article’s linked essay, the author goes into the details of how much energy is used per passenger mile in the US. In these BTUs/passenger mile figures, cars use about 3500, and mass transit uses about the same (surprising, but in most cities trains and buses aren’t full all the time).
The automated cars are lightweight (because of better materials–think that 200mph VW–and they’re basically impossible to hit) and only as large as the trip requires.
They use 100 BTUs/passenger mile.
October 14th, 2008 at 1:47 pm
“he’s more optimistic than Petey’s “many decades”
Note that I’m positing that there will be a long gap between the introduction of fully automated cars that can handle limited-access highways and the introduction of fully automated cars that can handle everything else.
Automated cars on interstates is far more trivial than automated cars on city streets.
I certainly think it possible to have full automation on limited-access highways in 2030. You can pretty much see how it’d work using existing and near-future technologies. But non-highway full automation is going to need to wait on technologies that aren’t even close to being on the horizon.
Tangentially, it’s worth noting that a normal Space Shuttle voyage is entirely automated, except for the landing, which is done by a human pilot. Similarly, in 2045, we’ll likely have cars that can get from the outskirts of Philly to the outskirts of NYC without human intervention. But you’ll stlll need a human pilot to get from the NJ Turnpike to the destination in Manhattan.
October 14th, 2008 at 2:32 pm
Looking For Auto Shipping??? I haven’t even had to worry about it since i found this company. The Car Shippers are a Veteran owned company with over twenty years experience, shipping all types of cars to destinations nationally and worldwide. I had years of problems with auto transport companies but they’ve taken away that problem. Try them out, you wont regret it. To ship your car contact this auto transport company at 866-429-9677, or visit their website at http://www.dtdtransport.com.
October 14th, 2008 at 2:35 pm
You don’t make neighborhoods more walkable by changing cars you do it by changing the neighborhoods themselves and rendering daily use of a car unnecessary. Better cars don’t improve transportation for people on foot or on bicycles radically as fewer cars do.
There is no mention of the impact on pedestrians and cyclists in the linked article except to say that these developments will inevitably make things more pedestrian friendly. How is it inevitable that these issues will be taken into account when the writer at ars technica, yglesias and most of the commenters here did not even consider the implications. There is a certain amount of magical thinking going on here as if you say it will be better then it must be so.
The reality of being car-less right now in America is that in most of the country being without a car means you have little access to amenities; the things like sidewalks, safe traffic lights, bike lanes and public transit that make transporting yourself safer and easier. If the focus is on making cars better there will not be any effort to make the environment better, just as it has been this entire time (exceptions like bike friendly cities and federal funding for alt transport are precisely that hard fought exceptions). If everyone is assumed to have their own car there are “no” pedestrians, no pedestrian friendly amenities are built, if none are built there will be very few pedestrians, those that are on foot are in danger because nothing is made to take them into account. Pedestrian and cyclist safety is most dependent on the number of pedestrians and cyclists because it increases the caution of the flawed human driver. If the developers of this technology and the admirers of this technology are not including pedestrians and cyclists in the thinking behind it the pedestrians and cyclists will inevitably be ignored and they will not be safe. If they are not safe they will not walk or ride and from a resource and energy perspective that is a complete waste.
I suppose few other people here have ever had somebody lean out of a window as you were riding a bicycle on a public road to scream “Get a car, bitch”, never had to walk a 1/4 mile out of you way through parking lots to get to the entrance of a store that was 75 feet away as the crow flies, been driven off the road by a cab and hit another person, or almost been hit by cars several times in the suburbs when the lights changed when you were in the middle of a six lane highway that barreled through the neighborhood but the idea that cyclists and pedestrians will automatically be considered as an essential part of transportation in the US strikes me as kind of preposterous.
October 14th, 2008 at 7:51 pm
Petey has no clue what he is talking about.
As the DARPA urban challenge showed last year, we already have, or are very close to having, the technology for fully-autonomous cars driving in urban environments. The main obstacles to the widespread adoption of robot cars will be social and legal, not technological.
The clearest implication of robot cars is the death of conventional mass transit. No one is going to use a bus or a train when they can make the same trip much faster, much more comfortably and just as cheaply in an automated taxi.
October 14th, 2008 at 8:27 pm
Matthew,
Then with less space dedicated to parking lots, development will naturally become denser which further makes a model more oriented around some walking plus some taxis rather than individual car ownership look appealing.
Huh? Robot cars will make car travel much more attractive than it is today. People who today, for economic or practical reasons, have to rely on walking or public transportation to get around will instead be able to use cars. This will greatly increase their mobility. The obvious implication of greater mobility is lower density, not higher density. The easier and cheaper and faster it is to travel longer distances, the more spread out housing and infrastructure are likely to become. Sprawl was first created by buses and trains, was greatly expanded by conventional private motor vehicles, and will be expanded even more by automated ones.
October 14th, 2008 at 9:12 pm
Mixner: “The obvious implication…”
It must be nice to not require proof.
October 15th, 2008 at 12:10 am
The ARS Technica article is, as it says, largely based on my recent series of Robocar articles. I think if you delve into those — and I admit they are not a super quick read — you will find answers to many of the questions and issues posed in the comments here. Indeed I have chapters on the various roadblocks, the various downsides, and answers to many of the common objections.
In addition, I recommend to all that they check out the videos of things like the Darpa Urban Challenge before they declare this to be total science fiction. These are just prototypes, to be sure, but they were built with trivial amounts of money and time, and are harbingers of what is possible, if we dedicate ourselves to it.
The rewards — every year saving just in the USA 40,000 lives, 50 billion hours and over a trillion dollars and 50 BILLION gallons of gasoline — seem worthy of such dedication. Multiply out by 10-20 for the entire planet.
Anyway, please check out those sections at http://robocars.net
December 26th, 2008 at 8:25 pm
If you click the second one down at sourcearticle.info there’s some more information
January 18th, 2009 at 5:56 am
laptop battery
laptop batteries
February 8th, 2009 at 11:01 pm
laptop battery
March 2nd, 2009 at 5:23 am
levitraI want to say - thank you for this!
March 12th, 2009 at 11:42 pm
Excellent site. It was pleasant to me.
March 14th, 2009 at 5:31 am
Great site. Good info
xanax
March 17th, 2009 at 2:30 am
Very interesting site. Hope it will always be alive!
tramadol
March 22nd, 2009 at 6:21 am
tramadol
I bookmarked this site. Thank you for good job!
April 3rd, 2009 at 4:18 am
I want to say - thank you for this!
cheap brand pfizer viagra
April 9th, 2009 at 6:37 am
I bookmarked this site, Thank you for good job! viagra
April 15th, 2009 at 4:25 pm
I promised my lawyer I’d pimp his site online ( http://ticketvoid.com ) after he got me off for drinking & driving. All charges we’re dropped today! Kudos
April 18th, 2009 at 8:21 pm
I got a DUI ticket and my lawyers ( http://ticketvoid.com ) convinced the judge that the police don’t know how to administer the drinking test.