Strange that the Economist would catch the diminished need for parking spaces, but miss the real economic impact of driverless cars: many fewer cars would be needed as utilization rates rise far above the current ~4-7%. This would cause a total collapse of the car industry, not save it as suggested by the article.
 7500-15000 miles at an average of 25MPH is 300-600 hours a year or or about 4-7% utilization.
Another way of thinking of it is that there is demand for a certain number of vehicle-miles per year. In the aggregate over the long term, it is no different if we buy 5 cars, each of which lasts 5 years, or if we buy one car every year for 5 years as the old one wears out.
Higher utilization should make fleet replacement a smoother function over time (whereas now, you can see dips and spikes as people decide when to put capital in their cars based on factors like the overall health of the economy).
 It is not clear what the impact of driverless cars would be on mileage. It would eliminate some useless trips (e.g., doubled miles for dropping somebody off). But the convenience may mean that people simply use it more.
 I'm assuming that wear and replacement time are purely a function of miles traveled. This is not entirely true, as age is a factor, but I think mileage tends to dominate in the average car (i.e., one that is being used, not one having its belts dry-rot over many years of neglect). Companies that maintain car fleets probably have data on this.
Besides, who says that the driverless car isn't more like industrial equipment instead of the current consumer crap, made to go on "forever" to justify its high price?
While designing cars to last longer would increase the weight of some components, for example allowing more metal around the cylinders so that they can be re-bored a number of times, this could be at least partially by having a much reduced burden of crash resistance.
Some things could be alleviated simply by avoiding the maintenance / cost cutting shortcuts used today. For example fitting grease nipples to ball joints and greasing them regularly makes them last a great deal longer than the sealed units which are riveted to swing arms today!
But other vehicles are made to last much longer. why not cars? Also, computer drivers can be programmed to be much gentler on the engines and transmission.
The cost of the extra equipment won't be negligible - but it'll be very low once this stuff gets widely adopted.
 http://www.its.dot.gov/research_docs/pdf/cicas_tech_docrpt2.... PDF, see pages 12 and 13
The cost doesn't need to become negligible. The cost needs to become justifiable.
2. If we are treating all cars are like a fleet of autonomous taxis, it doesn't seem really crazy that you could also optimize carpooling where the car picks up other passengers to optimize the trip with respect to where you and they are going.
If you have optimized traffic, you could live further out and the commute wouldn't take any longer, so that could lead to more sprawl. You also would be able to do something while commuting besides watch the road, making the commute more tolerable, which could lead to even more sprawl.
Are you referring to multiple cars in one household becoming superfluous? This would, of course, only be the case if driverless driving would be permitted entirely — I can very well imagine a prolonged transition phase where a human being needs to be able to intervene at all times.
Or are there already plans for automatic car sharing? This could be easily opposed by promoting the opinion that having an own car that is customized to your personal needs and where you can leave your belongings bears an advantage big enough to buy one.
Of course, ecologically, it would be great if both of this would be permitted, but my point is, autonomous cars could be introduced without car manufacturers incurring any immediate losses.
What will change demand is that the car occupancy rate will grow. And since one doesn't need to own a car anymore, the number of cars will be reduced.
Bringing together strangers on the same route will get a lot easier in the future. And if one isn't using one's own car but essentially a taxi, we'll see a lot more ride sharing. Especially, if some of the oil price horror scenarios (at least partly) play out.
Right now, the car occupancy rate in the US is around 1.6. Imagine it going up to 2, which is around the rate of some Eastern European countries a couple of years ago. I think it will be even higher.
Another thing to consider is the demand change, since the usage changes (to some degree) and the buyers change (renting companies instead of consumers). If that benefits domestic car manufacturers, idk.
Even with a pretty decent taxi service I still own a car, as do most people I know.
So I suspect that even when cars become completely self-driving we will still generally end up buying the things for possibly less than completely rational reasons.
The IRS mileage rate is 14 cents per mile driven in service of charitable organizations and 51 cents per mile driven for business.
Taxis haven't replaced private cars because they're radically more expensive.
According to the article, at the end of a 12-hour shift, a driver goes home with $160 after paying $120-$130 to rent the cab & medallion and $8 for credit card processing. There's also gasoline costs, which the article doesn't give a number for, but as a guess, a driver might go through half a tank on a shift for $40, and that's in an inefficient large car with a V8.
So over half of the cost of an NYC taxi is the driver's labor, and a nontrivial portion of the taxi rental goes to pay for the taxi medallion, which has gone up in cost to a million dollars in recent years:
Many other cities have the same problem, like my hometown Milwaukee:
The economics of taxi service in the US are largely driven by labor costs as well as regulations that have been perverted to funnel money to a rent-seeking set of license/medallion owners, not the running cost of the motor vehicle.
I guess a good measure of how corrupt a particular political environment is will be how long it takes to give in to the overwhelming reasons (ecological, economical and social) to allow the true, very low cost of self-driving cars to have its impact on the taxi industry.
After all, why wouldn't they?
London is one place that doesn't have any limit or medallion system regulating the number of cab drivers - although they do have regulations regarding eyesight, criminal record and route-finding ability.
Some parts of the world don't have such silly laws. Arethuza mentions Edinburgh just has a £2k licence fee. That's a small part of the price and cost of operation of a car, particularly one that's self-driving and on the road 24/7.
At normal cab rates, this would bring in shitloads of money. So people would start cutting the rates. And cutting more. Soon, a £40 cab ride will cost £20... then £10... then £4...
When taking a cab for half an hour costs a couple of quid and the cab is there at the push of a button, you need a car even less (most Londoners already don't have cars).
Once this transformation starts hitting, some cities with politically corrupt systems will still maintain their protectionist laws... but for how long? The ones that keep them longest are de-facto the most corrupt (at least in transport policy/laws).
Also, many regions don't bother with the ridiculous monopoly system of "medallions". I can start a mini-cab firm in my area without any such artificial supply restrictions at all.
Should driverless cars become mainstream, criminals of all kinds will quickly find ways to game the system to make vehicles driver-controlled, and (more ominously) will figure out how to hack into other people's cars and use them for nefarious purposes -- vehicle botnets that can cause major economic and physical damage. The police will need ways to override driverless vehicles to keep up with criminals.
Moreover, many individuals who love driving their sports cars don't want driverless vehicles; many law-abiding car hackers and tinkerers don't want to relinquish full control over their vehicles; and civil-rights organizations will likely end up fighting driverless vehicles in the courts.
Instead of an Utopian, smoothly-working, accident-free system that doesn't need police, seat belts, or insurance, I would expect our road-traffic system to evolve into a more complex mix of driverless and driver-controlled vehicles. The unintended consequences of the transition to this system, and its ultimate legal and technological requirements, are really unknowable at this point.
PS. In response to paulsutter below, this is not "dystopian terrorist hacker paranoia!" All I'm saying is that the transition will be a lot less simple and smooth, and will have more unintended complex consequences, than suggested by this piece. That's usually how it goes when society makes fundamental changes to the architecture of a large, widely used, complex system!
Hybrid solutions are also possible: there will be certain roads (controlled access interstates) where driverless is mandatory while certain streets where driver intervention is required. You still get most of the traffic optimization benefits if the driverless roads are congestion bottlenecks. However, the real benefit doesn't come until we have driverless cities, where people don't need to own cars and a roaming fleet of autonomous vehicles serve as pervasive taxis. You could even get rid of parking lots!
The system can easily trace what cars are bad actors and adapt accordingly (shut down traffic in a region with a bad actor until police arrive).
People who want to drive would do so on regular roadways. Within 20 years, virtually all cars would be driverless-capable, and society would have a clear idea of the advantages and whether to allow old school driving anywhere.
I dont buy the dystopian terrorist hacker paranoia. More and more of the physical world will be controlled by software. We will either make that secure or go back to living in caves.
On seeing those you have to wonder what a computer driver would do in some of the circumstances. There are some where there will definitely be a crash (kinetic energy has to go somewhere). Should the computer minimize the number of human casualties/injuries amongst all the vehicles involved even if that means harming the (fewer) occupants of the vehicle more?
Teens aren't getting cars because they are too lazy to take the drivers test. It's because the cost, fuel, and maintenance just aren't worth it to them. Even if it drives itself.
1. A need for status. Driverless cars will first greatly distinguish luxury cars from more ordinary cars.
2. Convenience and comfort. Driverless cars will remove the stress from rush hour commuting. Driverless cars will be much more comfortable for long-distance trips.
3. Safety. Driverless cars will obey the traffic laws. "drivers" and passengers of such cars will much less likely be arrested for driving under the influence of alcohol or drugs.
4. Efficiency. Driverless cars will use less fuel than otherwise comparable self-drive cars. For example, cooperating driverless cars can draft at highway speeds. Additionally, driverless cars can smooth out stop-and-go traffic, coast down hills and otherwise conserve momentum. And once driverless cars are commonplace, intelligent intersections will permit cooperating driverless cars to interleave without stopping - while accommodating self-drive cars with conventional signs and signals.
2. Not convinced anyone really gets stressed driving, and not convinced that "not being in control of the car" would make anyone less stressed. Personally, when someone else is driving the car, I'm always more stressed not less. If I knew it was being driven by some software, my stress level would hit the roof. Everything I own that has "software", crashes every now and then. I don't really want my car to have more software than it needs.
3. Software bugs are plentiful in any system. Simple mechanics, and humans will always trump software for things like this. Surely if this stuff was so easy to automate, we'd have software flying planes already (Which seems a substantially easier problem).
4. Considering that we still can't make automatic cars that change gear at the right point, I'm skeptical.
2. Regarding stress and strain while driving - here is a research paper that might convince you: http://www.ectri.org/YRS07/Papiers/Session-9/Schiessl.pdf
3. Experience with diverse embedded software systems, from medical electronics to military jets, has demonstrated that software can be indeed engineered, and verified, to be more safe than simple mechanical and purely human controlled systems.
4. In fact current automatic transmissions are now more fuel efficient than manual transmissions: http://green.autoblog.com/2010/08/18/greenlings-why-do-autom...
4. I'm sure it's possible to make an automatic transmission better for fuel consumption, that's an easy problem. But make an automatic transmission that knows what you as a driver want (I'm betting that's rarely "fuel efficiency") - that's the holy grail, which certainly hasn't been achieved yet.
Combine those two unavoidable traits of human existence and we get what is called Rush Hour. And it is not going away. At best, driverless cars represent a minor improvement in the pain of rush hour, but it doesn't abolish it...nor does it rid us of the need for public transit. There are simply too many geometric facts in the way.
1.3 million people commute to Manhattan every weekday. At 1.2 cars per person (typical for a morning commute) that is still more than a million cars fighting for some 2000 miles of road space. Even without any need for parking (assuming they would all be taxis, which is unlikely), that is still 1 car every ~10 feet of road length...including intersections. In other words, geometrically impossible, unless you want to argue that no bottlenecks would exist anywhere (such as intersections or bridges).
Driverless cars will be nothing more than a minor improvement in congestion and parking costs.
Where the real improvement comes from is driverless trucks and buses. A driverless truck reduces per-mile costs of road freight by about 40%, while allowing for 24hr days (13 more than the max possible with one driver, and 15 more than the average), making all road freight more than twice as fast. That is transformative for commerce. And for city buses, 85% of the cost is going to drivers. We could literally be offering buses at 4x the frequency at the same cost. That is transformative for cities.
First, there are people who simply will not trust the driverless cars. I was talking to a friend yesterday who said that he wouldn't ride in a driverless car, and he wouldn't want to drive on the same road as one. Uptake is not going to be as fast as people seem to expect once the cars are practical for market.
And second, in the words of Jarrett Walker : "Technology never changes facts of geometry." No matter how small you can make the engines, airbags, and frame, as long as you have each passenger in a separate, self-propelled, hopefully crash-proof shell, they are always going to take up far more space than a bus or a train, and require far more resources to construct.
This will change over time (and quite fast). Ever heard of red flag laws?
> In the United Kingdom, the Locomotive Acts (also known as Red Flag Laws) was a policy requiring self-propelled vehicles to be led by a pedestrian waving a red flag or carrying a lantern to warn bystanders of the vehicle's approach.
> The most infamous of the Red Flag Laws was enacted in Pennsylvania circa 1896, when legislators unanimously passed a bill through both houses of the state legislature, which would require all motorists piloting their "horseless carriages", upon chance encounters with cattle or livestock to (1) immediately stop the vehicle, (2) "immediately and as rapidly as possible... disassemble the automobile," and (3) "conceal the various components out of sight, behind nearby bushes" until equestrian or livestock is sufficiently pacified.
At least with driverless cars, we're getting support from the government (in NV and CA, so far).
I've heard that when electricity came about there were people afraid of touching light switches because they had something to do with electricity. I personally know a person (mentally healthy, over 80 years old) that does not want to be alone in the room with turned on TV because she's afraid of electricity.
 I guess it depends what buses and trains you catch.
When I lived in San Diego, the bus line by my house ran at half the frequency using buses twice as large on weekends, which I can see no explanation for other than saving money on labor by making transit service much less useful.
-) As soon as software controls something it's the question about if you control the software, or if the software controls the user
One of the big arguments rms always had about free software, but think about it:
Let's say i have a self-driving car, how do i know that there isn't a backdoor in there?
How about my car being hacked and being remote controlled?
How about my car being pulled over by the police by remote control if i don't stop?
And what about spontaneity in driving?
Will i still be able to make a U-Turn whenever i want?
Having navigated through big cities with cars i can tell you that it's more than just reading signs and staying on the correct lane...
Once we got self-driving cars then all of the above points will be issues and we should seriously think about all of this in my opinion.
If i could make sure that non of the above is valid i might still fancy it for boring highways. For cities or narrow situations i will always trust myself.
But even i witnessed my fair share of motorcyclists, who just played daredevil on the street.
The way many behave on roads is not acceptable and this goes for drivers of cars and drivers of motorcycles.
Counter question: Would you trust a self driving motorcycle?
So why should i - as a in my opinion perfectly good driver - trust a self driving car at all times?
The question of self-driving motorcycles is an interesting one. Would I trust them? Yes. The reaction time and ability to process multiple streams of information of computers far outweigh that of humans.
Would I want to? Sometimes. There are times when I'm out riding and it's all about the curves of the canyon. Just like any driving enthusiast, I'm there for the experience. But crossing Wyoming on I-80? Bring it on!
I have no doubt that there will be a Honda Goldwing or BMW R1600 that includes a self-driving mode along with a HUD-based entertainment system in your helmet. And maybe I'll be able to afford it by then. :)
I hope Google is investigating these weird edge cases, and testing for them.
This is a bit delusional. Someone will have to pay for accidents and injuries.
And yes, that will probably be a smaller market although you can still sell add-ons, etc.
If this came to be, coupled with high speed wireless, I would go so far as predicting the end of cities as we know them. They would serve essentially no purpose anymore.
Cities are more populous and wealthy than ever before because there is no substitute for human proximity in a service economy, and more humans in close proximity makes for more productive services.
More importantly, I don't trust my brain's thread scheduler to drive with full attention and think about something else productively simultaneously. How many of us can say that they have never been in a dangerous scenario because they were lost in thought while driving?
Many measures that would increase pedestrian and cyclist safety are blocked by members of the entitled minority motorist class: speed cameras because many motorists believe themselves unable to drive at a speed safe for other road users, and bike lanes and wider sidewalks because drivers feel the region of the public roadway they have colonized for the storage of private property is a permanent entitlement. Automated vehicles would better obey speed laws and make nearby street parking less necessary.
I think the article "Why Driverless Cars Are Inevitable--and a Good Thing"
by Dan Neil of the Wall Street Journal, published last month, is a good commentary on why ordinary people will mostly be glad to use driverless cars, and regulators and insurers will be glad to nudge drivers to use them.
"As a mature, postindustrial society, the U.S. has in many ways topped out economically (population growth, consumption) compared with younger competitors on the world stage. Americans are learning hard lessons about the value of their work in a race-to-the-bottom global economy.
"The one brilliant part of the U.S. economic profile is productivity. It turns out, Americans are a little nutty when it comes to work.
"If autonomy were fully implemented today, there would be roughly 100 million Americans sitting in their cars and trucks tomorrow, by themselves, with time on their hands. It would be, from an economist's point of view, the Pennsylvania oil fields of man-hours, a beautiful gusher, a bonanza of reverie washing upon our shores."
Driving here in Minnesota, I can't wait until the typical Minnesota driver's low-grade performance is replaced by the performance of road-certified driverless car systems. And I most definitely look forward to having much more time to think and to enjoy undistracted conversation in a driverless car than I can now safely achieve while driving. That indeed will be a boost to my productivity and to the quality of my family life.
Another subthread here talks about the trade-offs between owning a personal driverless car versus taking rides from a driverless taxi service. My family, like many families in the United States, considers the issues of the one-car-per-adult lifestyle versus the one-car-per-household lifestyle. Most United States families with children would probably make their choices about car ownership with that being the key decision point. I can see a lot of United States families deciding to have one owned family car that is filled with all of their favorite things to have along during road trips (so that the car is one more personal storage unit, in part) that would accumulate a lot of pleasant memories of family outings and so forth. But the family's second car (in some families, the third car) would be replaced with a subscription to a driverless car service, which would reliably pick up family members on a moment's notice for trips to the grocery store, the soccer game, a music lesson, a party while the other parent is still at work, or whatever. If the driverless car services become reliable enough to whisk a child (with or without parent) over to an urgent care center for sudden injuries and illnesses, one major reason parents keep a second "transportation car" besides the main commuting and traveling car would vanish. I fully expect individual rates of car ownership per household to drop when driverless cars become reliable.
Seems like we've been a while since the last one, which was the internet. To me the next two look like driverless cars and advanced 3-D printing. Affordable fusion might be one, but I'm not holding my breath.
There are a lot of other cool things out there, of course: Elon Musk and his rocket program, trans-humanism, progress towards strong AI. But assuming we all actually stay individual people and don't become something like a hive mind (which is a big assumption), being able to go places without effort and make our own goods at home will empower people in ways never seen before. 50 years after these technologies mature, it'll be a completely different world.
What the mainstream media also aren't realizing is that the technology for driverless cars is also the same technology that automates away most factory/farming/mining/retail/distribution jobs. The effect in the west might not be so large, but in the rest of the world it could be devastating, leading to riots and civil wars.
On the other hand, I do have to say I'm a big fan of airships, but I'm not holding my breath for their return. I even doubt people would accept that travelling has to be so much slower in an airship, now that he jet age has arrived.