Greater Greater Washington

Roads


How will self-driving cars change transportation?

Yesterday, I argued that we will start seeing autonomous vehicles operating on our roadways in 7-12 years. But whether self-driving cars hit the roads 5 years or 30 years from now, they will bring major changes in our transportation system and even our society.


Photo by imnewtryme on Flickr.

They'll be more often in use, less often parked: Since most cars are parked for 98% of their existence, a self-driving car can be put into use when it would otherwise be idle. This can kill several birds with one stone. After dropping off its passengers, the car can do double duty as a taxi, delivery vehicle, or just get out of a congested area.

A model like Zipcar becomes an on-demand taxi service with self-driving cars. And a model like SuperShuttle becomes a micro-jitney service with self-driving cars. Now, SuperShuttle only serves airports, and the driver and dispatcher try to create the most efficient routes based on their ever-changing flow of customers. A computerized system could make this work everywhere.

If it's not needed, a self-driving car can park itself at an offsite location, thus eliminating the need to build large amounts of parking at desirable (and expensive) locations.

They'll reduce labor costs: A self-driving car needs no operator, thus removing human labor from the equation. Self-driving cars will put taxicab drivers out of business. What will those thousands of people do with their skillset when a computerized system makes them obsolete?

They'll expand access to transportation: The process of driver training and licensing will be obsolete, and the requirement that people be 16 or 18 to drive a car will be irrelevant since now there are no drivers, only passengers.

This is great news for the disabled, especially the sight-impaired, as well as for adults who have lost the ability to drive. Will we create some new paradigm of age restrictions for being an unattended passenger?

Self-driving cars eradicate the car-ownership paradigm. If you can easily and affordably (remember, no labor to pay) book taxi service from your smartphone, more people than ever will eschew the costs and annoyance of car ownership.

They'll be safer: Self-driving cars likely won't make human errors. Auto crashes typically claim around 38,000 lives per year, and that's been true for decades. Over 80% of these are attributable to human error, either negligence, distraction, incapacitation, malice or other uniquely human quality.

They'll reduce congestion: Self-driving cars can manage congestion as a system, rather than a collection of self-interested units. A lot of congestion stems from the way each driver acts in his own self-interest. For example, changing lanes might (or might not) help one individual driver, but hurts the overall performance of the road. Speeding into a gap and then braking also creates worse congestion overall.

If all cars are self-driving, then they can cooperate to mitigate congestion. For instance, the cars could all slow down to 35 mph past a crash or police traffic stop, rather than allowing the speeding up and slowing down and rubbernecking which lead to traffic and more crashes. Over time game theory and other disciplines will help engineers devise ever more complex strategies to keep the system performing optimally.

They'll make current transit economics obsolete: Self-driving cars represent a major existential threat for current and planned transit systems. Our current transit paradigm relies on capital and operational subsidies. We can't charge riders enough to pay for everything that goes into making transit work. As we raise fares, more riders forego transit and choose the automobile.

If, as I suspect, self-driving cars are handled primarily in the private sector, their operations will not be subsidized, and their relative convenience and utility will call into question the logic of investing billions into the construction and operation of transit systems.

They won't last as long: Automobile manufacturers will have to adapt the volume of vehicles they produce annually. While many fewer cars will be needed across the economy, those that are autonomous will be driving much more frequently. Their replacement cycle would be more similar to police vehicles, which only last around 3-5 years before wear and tear makes replacement a better option than repair.

Most passenger cars today spend around 98% of their time parked somewhere in between single-occupancy trips. Consequently, their average lifespan is between 15 and 26 years.

They can be electric: An electric self-driving car can go to where the charging stations are. DC and other governments are currently embarking on a campaign of spreading electric vehicle charging stations around the urban environment under the assumption that we must cast a wide net of these kiosks around so that they are convenient to an EV owner's origin or destination. But in a few years, it is likely that this will be entirely unnecessary, and rather the car can take itself to a central charging location, like a power substation or electrified parking garage that can efficiently charge hundreds of vehicles on an as-needed basis.

They'll change culture: A self-driving car eradicates a unique part of the American identity, the freewheeling mastery of the open road. We'll wax nostalgic for what we've lost, but everyone will benefit from the gains.

A world with self-driving cars would operate very differently than the one we currently live in. I would say that's mostly for the better. As urbanists, we've often succumbed to a gut reaction that cars are bad, transit is good. However, the reality is that it is not cars that are bad, but the single-occupancy driver paradigm that is so damaging to our environment, urban fabric and quality of life.

We still live in an America where 78% of people drive to their jobs by themselves. I'm convinced that we're about to see that start to change as self-driving cars become a reality. It is time to start having the conversation about how we want this future to unfold in order to best plan for a very different world.

Will Handsfield earned his master's degree in public policy from the University of Denver in 2008, and has worked on transportation projects in Los Angeles, Denver and Washington, DC. Will bike commutes and lives with his wife and son in the Lincoln Park neighborhood of Capitol Hill. His posts reflect his personal views. 

Comments

Add a comment »

"Self-driving cars likely won't make human errors.'

kind of like how my bank no longer makes errors, right?

Or how easy it is to solve every customer issue using a phone to talk to a computer?

I will require more evidence that this can work on a large scale before I advocate changing existing infrastructure plans for it.

by AWalkerInTheCity on Dec 7, 2011 2:14 pm • linkreport

Again, this should not be on GGW.

Technology doesn't solve all problems. It never has and never will.

Hype-o-thetical technology solves even less. In fact the only problem it solves is how to pad the bank account of the huckster who solicits investments in it.

by Cavan on Dec 7, 2011 2:16 pm • linkreport

I'd like to point out that we're still going to need a licensing regime, as these cars will most likely (at least at the early stages) only be self-driving on the highway, wit hthe human operating on secondary roads and side-street - otherwise, we'd have to convert every single road in America to this infrastructure.

Secondly, we need human operators for the same reason we still need airline pilots - breakdowns in the automatic system. If there were a bug in the system - or, more likely, a terrorist of some sort were to hack the system and disable it - we would still need humans capable of picking up where the computer system left off, lest we all die in fiery crashes.

by Matthew on Dec 7, 2011 2:17 pm • linkreport

I concur that self-driving cars will be a huge game changer. And like your earlier article, I think that they are coming soon rather than later with respect to skeptics.

Business models and other aspects of life might change quite dramatically as well. For instance, I suspect that delivery costs will fall. Consider super markets that have little or no store fronts. You place your order and people load your order onto a self driving vehicle which you pick up in front of your home. Also, these vehicles might be much smaller or more efficiently shaped since they need not fit people.

by Geof Gee on Dec 7, 2011 2:17 pm • linkreport

"the cars could all slow down to 35 mph past a crash or police traffic stop"

Why would there still be crashes and traffic stops?

by Theo16 on Dec 7, 2011 2:21 pm • linkreport

Yes! That's the future I want. Self-driving cars will allow us to expand the built environment that the vast majority of us actually prefer: low-density suburbia. I imagine the future of the U.S. will look a lot more like outer Houston and Phoenix and a lot less like NYC or DC. Sorry, urbanists. People don't want to live in a condo in an urban environment fraught with crime, bad schools, no greenery, bad air, noise, and high taxes. The numbers show people are more and more abandoning this failed model for the suburbs.

The urban cores will diminish even more than they have now, especially when wasteful public spending on 19th century modes like rail systems is cut. I imagine we'll find a new use for those subway tunnels, maybe as utility conduits or something. Or maybe we can turn the old city cores into farms or parks or something (in the event that they're completely abandoned, like it looks Detroit will be soon). No one will want to live in overcrowded dense cities and take crowded, smelly buses/trains with odd-collections of dangerous or annoying strangers when they can A). Live in their own detached house in the suburbs B). Take their own self-driving car where they want to go.

Americans like privacy and convenience, and suburbia and cars offer that — urban living and public transit don't. Cars also suit our individualist tastes (collectivism is more of a European/Asian phenonomenon). Find out more about how the suburbs and the autocentric environment are America's future in Joel Kotkin's great book "The Next 100 Million: America in 2050". Other good writer's on this subject are Randal O'Toole and Wendell Cox.

By 2050 when we have these cars, we'll look back and be glad that wasteful pointless programs like High Speed Rail and whatnot were squelched by forward-thinking individuals.

by Larry on Dec 7, 2011 2:22 pm • linkreport

Sorry, urbanists. People don't want to live in a condo in an urban environment fraught with crime, bad schools, no greenery, bad air, noise, and high taxes. The numbers show people are more and more abandoning this failed model for the suburbs.

You crack me up when you just make stuff up with no regard for whether it reflects reality. In reality, those with the resources to choose are disproportionately moving into the cities, which are steadily becoming more and more expensive because of competition for scarce housing.

http://www.brookings.edu/~/media/Files/Programs/Metro/state_of_metro_america/metro_america_report.pdf

Here in DC, that's even more true than it is nationally.

Not everyone is a misanthrope like you. Some people enjoy the company of others and the convenience of being close to amenities and events.

Personally, I can say that having a self-driving car would have no impact on my choice to live in the city.

by dal20402 on Dec 7, 2011 2:33 pm • linkreport

You had me at "Self-driving cars will put taxicab drivers out of business."

by Joe on Dec 7, 2011 2:38 pm • linkreport

Will and Larry, if the future of human settlement looks like outer Houston and Phoenix, we will destroy the planet and the human species. It won't matter if self-driving cars run on electricity. We cannot produce enough renewable energy to run automobiles and a suburban lifestyle for everyone. Even if we could, we would hit hard limits in the amount of rare-earth metals needed to produce the vehicles and the renewable energy plants. Moreover, the suburban development pattern requires infinite growth to sustain, which is just not possible. Promoting any kind of automobile is not and will never be forward thinking.

by Phil LaCombe on Dec 7, 2011 2:40 pm • linkreport

I can imagine the headaches if your car drops you off and goes somewhere to park itself and gets in an accident with no one in the car.

Will it send a text message to the owner telling them to now take a taxi to the crash site so it can exchange insurance info?

Or when my car spins out on an icy patch because self driving doesn't mean that the properties of rubber on ice change.

That said, I welcome the prospect of a self driving car but there are A LOT of issues that need to be worked out still that don't just talk about technology.

by Canaan on Dec 7, 2011 2:41 pm • linkreport

"Transit" just becomes a variation on the service which uses different vehicles and infrastructure as appropriate to capture certain efficiencies. How all this gets financed--including new road and parking systems, and any other new infrastructure designed to make use of the capabilities of driverless/fee-for-service vehicles--is not really easy to anticipate, but it will remain true that the beneficiaries of personal transportation are not limited to the people being transported themselves, therefore making the prima facie case for ongoing subsidies.

By the way, an awful lot of car usage today results from the fact that the fixed costs of car ownership are high in relation to the marginal costs of car usage. In other words, once you have to actually own and store a car for your personal use in order to fulfill certain transportation needs, you will likely add on a bunch of other uses which could not themselves justify car ownership.

Since the combination of driverless cars and fee-for-service models of car usage will change all that, I wouldn't rush to the conclusion that this technology will be bad for, say, walkable neighborhoods and business districts.

by BrianTH on Dec 7, 2011 2:42 pm • linkreport

Just some quick comments:

I believe Google's experiments have already shown that driver-less cars can handle complex environments like cities. And this is very early in the game.

Moreover, there is already a strong argument that he need for pilots is already limited; instead our prejudices that prevent us from taking advantage of the technology.

http://www.overcomingbias.com/2011/12/over-regulated-flight.html

by Geof Gee on Dec 7, 2011 2:43 pm • linkreport

Seriously, Larry, this doesn't address how in the world we're going to POWER all of these vehicles driving all over your suburban paradise, which I'm sure will be built over choice farmland to boot.

Oil isn't getting cheaper, quite the contrary. Electric and hydrogen are NOT going to be panaceas, not by a long shot. A self-driving car doesn't mean much with gas at 12 or 14 bucks a gallon. I'm sorry you have a pathological hatred of cities and it seems your fellow human beings since you seem to prefer to cocoon yourself away from anyone and anything unfamiliar. Unless we build a bazillion new nuke and coal power plants and up the capacity of our electric grid by 500% electric vehicles are not going to be practical as a substitute for gas powered cars.

Also, no one is going to dismantle the N.Y.C. subway or Washington's Metro any more than cars spelled the end of London's tube or Paris's Metro. Keep dreaming!

Detroit is in no danger or being abandoned any time soon.

And get some other references besides some pseudo-libertarian sprawl-worshippers.

by Brad on Dec 7, 2011 2:46 pm • linkreport

If it's like a regular car, but with electronics for sensing, processing, and communicating, it'll cost more than a regular car because it does more. Sure, there are economies of scale and whatnot, but you're adding labor to creation of the car, which adds cost. To offset the cost, you would indeed need to use it more. To make it do double-duty as a cab means that it's not "your" car in the same way that your car is today. Other people will do things you would do in your own car: eat in it, make a mess, maybe even puke; they'll treat it like a cab. Your car would be a taxi (for everyone) that you get first dibs on. Who buys a car to get really good taxi service? Who does that?

As for an auto-car being entirely free from subsidies, well they still have to travel on roads. It's difficult to make transportation infrastructure private because it occupies scarce land and serves such a public role. Transportation infrastructure will always be a monopoly (or at least an oligarchy) and heavily regulated for that reason. So it's very likely that roads will continue to be owned by the public as they have been for millennia. So who pays for the roads that these auto-cars drive on? The government? I smell subsidies.

Transit succeeds because of economics of scale. It's just more efficient, and that efficiency translates well into real money. Cars, in their current form, have held on so long because of artificially low gas prices and because it was easy to feel attached to car because it was your own. Loss of emotional attachment (through sharing) and higher gas prices* won't help an auto-car.

Electric cars are a completely independent issues, because power (gas vs. electricity) and navigation (human vs. machine) are independent issues. That said, if there were a way to make owning an electric car cheaper than a gasoline car, it's likely that electric cars would be selling like hotcakes by now. The sad truth is that electric cars with performance similar to gasoline cars require lots of scarce materials. Perhaps cheaper substitutes can be found, but batteries and electric motors have been around longer than gasoline powered cars; the technology is already quite advanced.

TL;DR: I'm betting against auto-cars (and electric cars) for lots of reasons.

by Amber on Dec 7, 2011 2:50 pm • linkreport

I'm sure there's more depth to the conversation, but so far the presentation (today's and yesterday's articles) seem to be much more focused on what we can do as opposed to what we should do, the latter being where I've historically found GGW's focus to be.

Also, I don't see them revolutionizing transportation quite the way the author suggests. It will always be faster (and more pleasant) to take the train to NY from DC than to drive or ride in a car. It will be even faster with high speed rail, and no amount of car that can prevent congestion will change that. There is still simply a growing volume of people and, even with more multi-passenger trips, public transportation will continue to be a more economical, environmentally responsible, and in many cases faster choice.

by Joe on Dec 7, 2011 2:51 pm • linkreport

@Geof Gee @ 2:17 They already have grocery stores with no storefronts that deliver to you. It's called Peapod.

by Amber on Dec 7, 2011 2:54 pm • linkreport

However, the reality is that it is not cars that are bad, but the single-occupancy driver paradigm that is so damaging to our environment, urban fabric and quality of life. We still live in an America where 78% of people drive to their jobs by themselves. I'm convinced that we're about to see that start to change as self-driving cars become a reality.

Why are self-driving cars likely to reduce, rather than increase, the number, or share, of single-occupancy car trips? They'll make car travel cheaper, easier and more widely available than it is today. The economic and practical incentives to share rides with other people, especially strangers, will be reduced, not increased.

by Bertie on Dec 7, 2011 2:55 pm • linkreport

@Joe @2:38 Maybe the auto-cars could still sexually harrass you and try to rip you off, just for nostalgic purposes. OTOH, that would be one aspect of auto-cars (if they were to catch on) that I'd find both likely and desirable.

by Amber on Dec 7, 2011 2:58 pm • linkreport

The economic and practical incentives to share rides with other people, especially strangers, will be reduced, not increased.

Exactly. People don't to share a ride with strangers. Americans prefer private property to public accommodations, always have, always will.

by Larry on Dec 7, 2011 3:00 pm • linkreport

"Our current transit paradigm relies on capital and operational subsidies."..."If, as I suspect, self-driving cars are handled primarily in the private sector, their operations will not be subsidized,"

The car system isn't handled entirely by the private sector now, we drive on _public_ roads. The car system is already subsidized heavily (gas taxes don't cover road expenses). Why would self driving cars suddenly convince the auto market to self-finance its entire operation?

Is the robot car market going to build an entirely separate and parallel road system or is it going to use the one that has already been built at public expense?

by Paul on Dec 7, 2011 3:00 pm • linkreport

I think the 2nd installment of this makes several nonconvincing arguments.

1)No more taxi driver
2)No need for licenses
3)No more big transit initiatives

None of those seem like great benefits.

The biggest benefit of this (as I see it) is the possibility of car-sharing programs mastering this techonology.

I think it's odd to promote the use of self-driving cars as a better (and affordable)alternative to existing modes of transit.

Are these cars capable of switching from automatic to manual operation?

by HogWash on Dec 7, 2011 3:00 pm • linkreport

Larry said: "Cars also suit our individualist tastes (collectivism is more of a European/Asian phenonomenon)."

You crack me up, Larry! I love the subtle racism that non-Americans (whoever they may be) are an anthill/beehive/Borg people that don't understand freedom or individualism.

by cendra5 on Dec 7, 2011 3:03 pm • linkreport

Again, this should not be on GGW.

Really?!?! I think this is one of the most interesting series of articles on GGW ever. I predict another 75+ comments for this article.

I'm sure there's more depth to the conversation, but so far the presentation (today's and yesterday's articles) seem to be much more focused on what we can do as opposed to what we should do, the latter being where I've historically found GGW's focus to be.

So, what do you think we should do?

**********************

The article seems to miss one of the biggest efficiency gains from driverless vehicles -- huge capacity increase of our roads. Currently, highways can accommodate 2000 cars per hour per lane. If vehicles can be driven much more closely together because they're automated, you could increase that to maybe 10,000 cars per hour.

********************

Why would there still be crashes and traffic stops?

Because there's no such thing as a foolproof system. Especially one designed by corporations that have every incentive to cut corners and governments that can be dysfunctional in oh so many imaginative ways. And, thus, that will be the downfall of driverless systems. I don't think people will tolerate risk (even relatively low risk compared to alternatives) that they cannot control.

by Falls Church on Dec 7, 2011 3:08 pm • linkreport

I'm inclined to agree with Cavan.

Also, self-driving cars are not the same as "robotic" cars.

by selxic on Dec 7, 2011 3:09 pm • linkreport

I'm willing to entertain the optimistic scenario presented here ...

However, I think "driverless car" is a misnomer. Driverless vehicles are more likely to resemble buses or shuttles. It's easy to imagine a scenario where trip management will be handled for numerous vehicles by a central server. The server constantly receives trip requests with an origin and destination from users. It then aggregates clusters together and maps optimal routes. Perhaps you could pay more for a faster time, and the server would be less likely to combine your trip with another. You could save money by being willing to wait while others embark and disembark en route to your destination. You would express this preference when making a trip request.

Urban areas would be privileged over lower-density areas, because they would have the customer base necessary to run the system efficiently. Trips to lower-density areas would be possible but likely cost much more. There would be little congestion at all, so low-density areas would lose that critical advantage they currently have.

Once enough people start using the system, it would function much more like transit currently does than like private automobiles currently do.

by Daniel on Dec 7, 2011 3:10 pm • linkreport

Thinking of these as very cheap taxis and very cheap shuttles can be helpful.

Do very cheap taxis means people will no longer have an incentive to live closer to regular destinations like work? Well, no. The "taxi fare" to work will likely be cheaper if you live closer, because most of your fare will still be determined by marginal costs varying with distance. Again, it is the large fixed costs involved in self-funded car ownership that reduce the relative weight of marginal costs, and so in a "cheap taxi" world versus is personal ownership world, many more people will have a stronger incentive to limit marginal use, including by reducing the length of regular trips.

Do cheap taxis mean people will no longer use cheap shuttles? Again, the answer is likely no. If there are efficiencies to be gained from transporting multiple people in one vehicle in certain situations, those efficiencies will still be there, and available for sharing with prospective passengers through lower fares. And again, the fewer people who are operating on the personal ownership model, the more people who will be making transportation choices on a case by case basis in light of such marginal-cost-driven considerations.

by BrianTH on Dec 7, 2011 3:20 pm • linkreport

Seriously, Larry, this doesn't address how in the world we're going to POWER all of these vehicles driving all over your suburban paradise

We'll power them with some combination of fossil fuels, nuclear and renewables. We have enormous reserves of fossil fuels. Not just conventional oil, but tar sands, shale oil, coal, natural gas, plus technologies to convert coal and gas to liquid fuels suitable for internal combustion engines. Biofuels may also become an important energy source. As the vehicle fleet becomes increasingly electrified, we'll power cars increasingly from nuclear and renewables. There is also enormous potential to increase the efficiency of cars, regardless of energy source. The current vehicle fleet averages only about 20 mpg. The Toyota Prius gets 50 mpg. The Chevy Volt gets 37 mpg on gas, and 94 mpg-equivalent on electricity. The Nissan Leaf gets 99 mpg-equivalent. And the technology to improve efficiency is constantly advancing.

Also, no one is going to dismantle the N.Y.C. subway or Washington's Metro any more than cars spelled the end of London's tube or Paris's Metro. Keep dreaming!

No, we're not going to "dismantle" them. But as travel in those cities shifts to self-driving cars, we can expect the subways to run less frequently. Eventually, they'll be running only at rush-hour, on the densest routes, if at all. Of course, the vast majority of cities don't have a subway. If they have any rail transit at all, it's light rail and commuter rail. And self-driving cars will be an even bigger threat to light rail and commuter rail than they are to subways.

by Bertie on Dec 7, 2011 3:22 pm • linkreport

Bertie, you should read up on EROEI--energy returned on energy invested. It will dash your hopes of running today's society or some techno-future society on anything less than conventional oil as its main power source.
http://www.theoildrum.com/node/3786

by Phil LaCombe on Dec 7, 2011 3:37 pm • linkreport

We'll power them with some combination of fossil fuels, nuclear and renewables. We have enormous reserves of fossil fuels. Not just conventional oil, but tar sands, shale oil, coal, natural gas, plus technologies to convert coal and gas to liquid fuels suitable for internal combustion engines. Biofuels may also become an important energy source.

Biofuels (particularly corn ethanol) take more energy to produce than is able to be extracted from them. Ditto for tar sands, shale oil, etc. There's a reason oil companies aren't eager to extract these, and why they push for deepwater drilling instead.

Not to mention the enormous environmental impact even if it was possible to burn every last molecule of fossil fuel on the planet. No melenge of alternative and renewable fuels will give us the cheap energy of the 20th century oil boom. You are quite mistaken.

No, we're not going to "dismantle" them. But as travel in those cities shifts to self-driving cars, we can expect the subways to run less frequently. Eventually, they'll be running only at rush-hour, on the densest routes, if at all.

Parking would still be an issue, self-driving car or no. And as I pointed out your rosy prognostications about continuing the happy motoring of the 20th century are unlikely to come true. Rail is a far more efficient at moving people than even the most energy-efficient private car. I'll go out on a limb and say that number 1, self-driving cars will most likely not be coming any time soon and that number 2, the coming energy crunch, in part made possible due to economic growth in the developing world, is going to make operating a private vehicle far more expensive than what we're accustomed to.

by Ken D. on Dec 7, 2011 3:38 pm • linkreport

Let's think about subways, and other similar technologies. Why do we run them so much in off-peak hours? Well, because the peak usage justifies the fixed costs, and the marginal cost for additional off-peak service is relatively low.

The crucial thing to keep in mind is that if driverless vehicles allow a transition from an ownership model to a fee-for-service model, then everyone will be making transportation choices case by case, based on marginal costs (transmitted via fees). So if the marginal cost of using the off-peak subway (or similar system) is lower than the marginal cost of using the one-person vehicle system, people will still have an incentive to use the off-peak subway--and again, many more people will be making decisions based on such marginal cost considerations.

by BrianTH on Dec 7, 2011 3:39 pm • linkreport

Likely bikes are not used 98% of their existence too...so let's not get too excited here.

by Pelham1861 on Dec 7, 2011 3:39 pm • linkreport

Exactly how with driverless cars help with congestion?

Google already has access to GPS information on drivers, thanks to Android phones. You can see it doesn't make a damn difference when it gives directions. If anything, it has made things worse by telling everyone to go the same route.

And the argument that driverless cars are more expensive is true. However, I don't see why they would be abused like cop cars. Not to mention most cop cars can last a hell of a lot longer than 3-5 years.

The google model of building up a giant database and comparing the real world to the database isn't going to scale.

Isn't it amazing we are having this discussion and WMATA can't get its trains to run on auto.

by charlie on Dec 7, 2011 3:42 pm • linkreport

It's also amusing from the ever-annoying and unrealistic 'green crowd' ... because they never explain where we are going to get all this EXTRA electric capacity. Hopefully from nuclear, certainly not wind and likely not coal...so this laughable idea of self-driving cars is about to drive into the pipe dream ditch.

by Pelham1861 on Dec 7, 2011 3:43 pm • linkreport

Do very cheap taxis means people will no longer have an incentive to live closer to regular destinations like work? Well, no. The "taxi fare" to work will likely be cheaper if you live closer, because most of your fare will still be determined by marginal costs varying with distance. Again, it is the large fixed costs involved in self-funded car ownership that reduce the relative weight of marginal costs, and so in a "cheap taxi" world versus is personal ownership world, many more people will have a stronger incentive to limit marginal use, including by reducing the length of regular trips

No they won't. Self-driving taxis will be so much cheaper than today's taxis that the economic incentive to keep trips short will be trivial. The higher cost of longer trips will be dominated by fuel costs, which today average about 18 cents per mile for gasoline vehicles (7 cents for the Toyota Prius), and about 3 cents per mile for electric vehicles. People who trade car ownership for getting around by taxi will save so much money in fixed ownership costs that they'll be able to travel much greater distances than they do today and still save money. But I doubt most people will trade ownership for taxis. Most of the market for self-driving taxis will probably come from people who are today dependent on mass transit or walking/biking, or for whom ownership is only marginally attractive.

by Bertie on Dec 7, 2011 3:47 pm • linkreport

I would build upon Daniel's great points,

"I think "driverless car" is a misnomer. Driverless vehicles are more likely to resemble buses or shuttles. It's easy to imagine a scenario where trip management will be handled for numerous vehicles by a central server. The server constantly receives trip requests with an origin and destination from users. It then aggregates clusters together and maps optimal routes."

"Once enough people start using the system, it would function much more like transit currently does than like private automobiles currently do."

If this system will work, it'll be the way Daniel describes, except that there's already a reaction to having every aspect of our lives is solely controlled by electricly powered computers. Without some redundancy though, cyberwarefare will make this scenario impractical.

So rather than 'aggregating clusters together and maping optimal routes', why not build light rail and have the cart follow the donkey?

by Thayer-D on Dec 7, 2011 3:49 pm • linkreport

hasnt pelham gotten the word? this tech is the new weapon against transit, urbanism and the latte sipping enviro elitists, or something.

seriously, if I was an advocate for this tech, I would be concerned at the way it is being hijacked by a particular ideological persuasion.

Of course I would wait for a much better demonstration of it before spouting about its impacts.

remember when the internet was going to make us all intellectuals?

by AWalkerInTheCity on Dec 7, 2011 3:50 pm • linkreport

A driverless shuttle system probably would not replace longer range transit, and certainly not intercity transport like rail or airplanes. In fact, it would work in tandem with these other systems very nicely. Imagine taking a more traditional rail ride from New York to DC. You would arrive with many riders requesting trips at the same time from a single node - Union Station. This would greatly increase the efficiency (and reduce the cost) of distributing all of the passengers to their desired destinations.

I totally disagree with Larry that Americans are so individualistic we will always refuse to be near other humans. We don't currently demand our own elevators? Our own private rooms in restaurants? If the price is right, we'll share. We're actually a pretty pragmatic culture.

by Daniel on Dec 7, 2011 3:52 pm • linkreport

Exactly how with driverless cars help with congestion?

The max capacity of a lane is 2000 cars per hour but at that rate, the vast majority of the road is used for spacing between cars going pretty fast (spacing is required because there's a lag between when a car ahead is slowing down and when we notice and react). Interestingly, if you try to add cars to this system, it actually decreases capacity down to 1200 cars per hour, which is road capacity when the road is congested.

The optimal situation is if you could have cars driving fast but with very little space between them. That's possible through automation because your car would sense when the car ahead of you is slowing down in an instant and slow your car down as well. The car ahead of you would also sense what's going on behind it, and adjust it's speed accordingly.

I found a quote explaining:

When a freeway filled with human drivers is operating at full capacity, Dr. Thrun notes, the cars actually occupy less than 10 percent of the road’s surface area. The rest is empty space between cars. Smart cars could be grouped more closely together, doubling or tripling the road’s capacity, as engineers have demonstrated by running a platoon of driverless Buicks, spaced just 15 feet apart, at 65 m.p.h. down Interstate 15 near San Diego.

http://www.nytimes.com/2007/12/04/science/04tier.html

by Falls Church on Dec 7, 2011 3:54 pm • linkreport

Some of the points that people make about paradigm changing are interesting -- making automobile type vehicles more like transit even could be thought of as guidewayless PRT. PRT makes no sense from a "mass transit" perspective, but as a ground up replacement of individually owned automobiles through this kind of system, like jitneys (e.g., ) maybe so.

The infrastructure cost though for city streets seems like it would be high.

WRT fuel, I agree with the commenters about the increased costs for fuel being a significant issue, and the issue of increased cost for production as harder to pump/refine fuel sources (like tar sands) become a greater proportion of the production mix.

The thing though about biofuels, the chairman of Nestle makes some really important points about biofuels and regular food in terms of calorie production. The target amounts of fuel from biofuels (20% of the fuel stream) are something like 4x-5x the total production of food from the standpoint of calorie production. Meaning the goals are impossible to reach, and set up massive problems in terms of competition between fuel production and agriculture, impact on water, etc.

This is from their webpage: http://www.nestle.com/AboutUs/faqs/Pages/CurrentTopics.aspx#biofuels

but there was a good interview in Financial Times within the past few months.

WRT fuel access, yes, increased prices have led to increases in production, at least on a marginal basis, as new techniques are used to generate greater efficiency in what were thought of as declining fields.

OTOH, the US, with 6% of the world's population, will not be able to consume 25% of the world's oil supply, as it does today, especially vis-a-vis China, Brazil, India, etc.

If those countries began consuming oil at a comparable rate to the U.S., likely production would need to double in order to maintain the same style of life (gasoline-based, automobile-based economy). It's an understatement to say that the likelihood of this happening is remote.

Electric cars (fueled by natural gas, coal, or nuclear energy based power generation facilities) are more expensive to create compared to gasoline powered cars, and the battery materials scarcity issue (batteries require scarce minerals primarily available only from China and Bolivia) is significant as well.

by Richard Layman on Dec 7, 2011 3:56 pm • linkreport

sorry, meant to include this cite of a past blog entry, http://urbanplacesandspaces.blogspot.com/2007/07/idea-for-free-public-transit-within-dc.html, which links to/quotes from an entry by Rob Goodspeed on South Africa, and a cite of a book about jitney transportation there.

by Richard Layman on Dec 7, 2011 3:58 pm • linkreport

Fascinating how we can compute so much raw data so quickly and turn it into a car driving itself. Re-affirms my faith in humanity now and then when I read about how smart we are.

****
However, I'm probably in the "we will see this in 100 years" category. I agree with an earlier poster who noted that this will probably be mostly for highway driving until the technology improves. Keep in mind we already have "lane assist" technologies and cruise control manipulators that keep safe distances in luxury cars (oh, and self-park too).

Furthermore, I don't see Americans relinquishing control of the wheel anytime soon. I, for one, find driving fun - I grew up outside of a small city without a traffic problem. Keep in mind, for a lot of the country, driving doesn't suck like it does around here. A mildly eccentric professor I had in college (classics) put it this way: the ancient Romans had gods for everything (hearth/home, love, war, etc). What are our gods? Our gods are cars (and athletes and other things, not relevant here). They are everywhere - we want the best ones, we are barraged with them in advertising constantly.

I think "driverless" cars will be another piece of the puzzle in getting everyone around in the next 100 years. If this country is going to continue to grow in population, it has the resources and land to do so. However, the best places to live are already getting pretty congested, and are too built out (hence urbanism).

by OrangeLiner on Dec 7, 2011 4:10 pm • linkreport

Here's the thing:

If automation will allow for closer spacing between cars on the road (therefore increasing a road's capacity), doesn't that require universal adoption of that automation? Google's current robocar is just like a human - it doesn't offer the kind of capacity improvement. The rules of physics still apply: stopping distance, accelerations, etc.

Second, as others have mentioned, the same kind of automation would be much more easily applied to a closed system like a rail transit network. And, indeed, we do have some automated transit systems. But it's expensive and not nearly universal.

by Alex B. on Dec 7, 2011 4:19 pm • linkreport

The crucial thing to keep in mind is that if driverless vehicles allow a transition from an ownership model to a fee-for-service model, then everyone will be making transportation choices case by case, based on marginal costs (transmitted via fees). So if the marginal cost of using the off-peak subway (or similar system) is lower than the marginal cost of using the one-person vehicle system, people will still have an incentive to use the off-peak subway--and again, many more people will be making decisions based on such marginal cost considerations.

Where is the MTA going to get the money from to reduce subway fares? It's already in dire economic straits, with a huge unfunded backlog of maintenance and repair needs. As New York's taxi fleet becomes self-driving, taxi rides will become much cheaper than they are today (and safer and easier). This will create a big incentive for current subway riders to switch to taxis. Subway ridership will fall. So subway fare revenues will fall. So the MTA will need to either raise fares, cut services or increase subsidies. Fare increases and service cuts will provide a further incentive for subway riders to switch to taxis. Taxpayers will ask why they should pay higher subsidies for fewer riders when there's a cheaper and better alternative.

The only plausible outcome is a much smaller subway system, if it survives at all. Off-peak services are likely to be replaced entirely by taxis. Peak services will only survive on high-density routes where there remains sufficient demand to justify them, if there are any such routes.

by Bertie on Dec 7, 2011 4:32 pm • linkreport

Another application of autonomous vehicles when the trip is particularly dangerous. Take in an episode of "ice road truckers" to see what I mean -- there are certain delivery route that are quite hazardous. Also, in combat areas -- autonomous fuel delivery in Afghanistan would be a good way to reduce Army IED casualties.

by goldfish on Dec 7, 2011 4:37 pm • linkreport

@ Cavan:Technology doesn't solve all problems. It never has and never will.

Just posting that statement on the internet is already a contradiction.

BTW: Technology never claimed it will solved *all* problems. Just a whole lot. Good luck hunting down your dinner to your cavane tonight with some branches. Or when that sounds problematic to you, perhaps you'd like to use any of the solutions offered by technology such as your house, stove top, supermarket and electricity.

by Jasper on Dec 7, 2011 4:45 pm • linkreport

I don't credit arguments dependent on assertions about broader cultural norms that would require people to act against their individual pecuniary interests. History suggests Americans will use whatever transportation model offers them the best deal given their transportation needs. For many people, that is currently car ownership. But if they are offered an even better deal, I have no doubt they will happily switch.

In other words, our cultural "gods" once included cowboys, but that doesn't mean we are all riding horses to work.

by BrianTH on Dec 7, 2011 4:50 pm • linkreport

Multi-passenger vehicles of various sorts will continue to enjoy efficiencies over single-passenger enclosed vehicles in many cases, including less land usage for a given volume of passengers. Dense cities where land values are very high will likely reclaim a lot of the land currently devoted to automobiles, and adopt land use fees (aka "tolls") that will give multi-passenger vehicles a significant fare advantage.

by BrianTH on Dec 7, 2011 5:04 pm • linkreport

"BTW: Technology never claimed it will solved *all* problems. Just a whole lot. Good luck hunting down your dinner to your cavane tonight with some branches. Or when that sounds problematic to you, perhaps you'd like to use any of the solutions offered by technology such as your house, stove top, supermarket and electricity."

Technology never claims anything. Humans advocating for, or selling a technology, make claims for it. Some are true, some are false, some turn out to way more complicated than one expects.

And of course no one ever said agriculture would be a good way to feed us all. First cause folks didnt talk like that then, but also cause agriculture wasnt 'adopted' as such. It was a supplement to hunter gathering. Which enabled populations to rise. which finally meant hunter gathering was no long viable. At which point human beings had shorter, sicker, more malnurished lives than their hg ancestors, but there were a helluva lot more of them, organized into more sophisticated political regimes. too late to go back. Didnt get back to good nutrition and health till the industrial era, thousands of years later.

That doesnt mean all technology is bad. It means its okay to be skeptical of some claims.

The internet changed our lives. It did not make cities obsolete, as was claimed by some.

If this makes shuttle buses more viable, or improves safety on limited access highways, I will be more than pleased. The "everyone can now live wherever they want" or "we dont need any more of your stinking subways" stuff isnt realistic analysis of a technology, its ideological wish fulfilment latching on to a technology.

by AWalkerInTheCity on Dec 7, 2011 5:05 pm • linkreport

Bertie, have you really been to New York City? There's no way that most "off peak" trips could be absorbed by taxis so that people can have private trips instead of using mass transit--there just isn't the available road capacity.

Not to mention that even without a driver, there is no way that a taxi ride will cost the same or less than a transit ride, especially if you have a monthly pass. Even at only double the rate of a single transit ride, the likelihood of tens of thousands of people switching is remote. Not to mention the capital cost of the vehicle is still what it is.

For NYC, "off peak" still means between 50,000 and 200,000 rides per hour at some points in the day.

I don't know what the typical block size is in NYC. A typical DC block is about 300 feet long. Even with no space between the cars, the lane capacity is only 20 cars, or less, if you include trucks.

NYC would not exist the way it does without the subway system. It would be completely unworkable without transit.

by Richard Layman on Dec 7, 2011 5:08 pm • linkreport

"When a freeway filled with human drivers is operating at full capacity, Dr. Thrun notes, the cars actually occupy less than 10 percent of the road’s surface area"

er wait, what? im sure its more complicated, but that sounds like an average of 9 car lenghts between cars. Thats not what I think of as full capacity around here. On the beltway folks go 65mph at less than 6 car lengths - more like 3 or 4. Sometimes less. At slower speeds (full capacity?) much less. Of course on some highways the rated speed is 55, and on suburban arterials its below that. On a 45 mph arterial, with frequent on and offs, some of them impulsive (related to shopping, etc) its hard to see the same improvement as on a limited access highway at 65 MPH - even more compared to the elusive 65 mph where folks drive with 9 car lengths between cars.

by AWalkerInTheCity on Dec 7, 2011 5:12 pm • linkreport

Multi-passenger vehicles of various sorts will continue to enjoy efficiencies over single-passenger enclosed vehicles in many cases, including less land usage for a given volume of passengers.

So what? People value other things besides efficiency. Things like speed, comfort, convenience and flexibility. If maximizing efficiency and minimizing land use were of paramount importance we'd all live in high-rises and get around mostly on foot or by bicycle. The fundamental effect of self-driving cars will be to shift the incentives even more in favor of cars and car-oriented development than they already are.

Bertie, you should read up on EROEI--energy returned on energy invested. It will dash your hopes of running today's society or some techno-future society on anything less than conventional oil as its main power source.

Fossil fuel energy will probably become more expensive, but not prohibitively so. We already have the technology to produce liquid fuels from coal and gas at large scale at 2-3 times the equivalent cost of conventional oil. And the reserves of coal and gas are vast. The technology will almost certainly continue to improve. The fuel efficiency of cars will almost certainly continue to improve. Switch from the average car today to a Toyota Prius and you cut your fuel bill by more than half. $4/gallon gas effectively becomes $2/gallon gas. And even current, first-generation electric cars cost only about 3 cents per mile in energy costs.

There's a reason oil companies aren't eager to extract these [unconventional oil], and why they push for deepwater drilling instead.

Yes, it's because they're still much more expensive than conventional oil. But as the price of conventional oil rises, or the price of alternatives declines, the alternatives will become more competitive.

Rail is a far more efficient at moving people than even the most energy-efficient private car.

This certainly isn't true.

by Bertie on Dec 7, 2011 5:43 pm • linkreport

Yes, it's because they're still much more expensive than conventional oil. But as the price of conventional oil rises, or the price of alternatives declines, the alternatives will become more competitive.

That doesn't mean "cheaper", sure if conventional gas goes up to $20/gallon then $15/gallon alternative fuel will be 'competitive' with it, but not the equivalent of the cheap fuel we have today. The oil industry nonsense about oil shale and tar sands being the next bonanza is just that.

<1>This certainly isn't true.

It is--take a from me, a transportation logistics engineer!

So what? People value other things besides efficiency. Things like speed, comfort, convenience and flexibility. If maximizing efficiency and minimizing land use were of paramount importance we'd all live in high-rises and get around mostly on foot or by bicycle.

Sounds good to me, and that's a good thing too, because the cheap energy, self-driving car future you're predicting is about as likely to happen as the flying cars from Back to the Future. This discussion is a tempest in a tea cup, self-driving cars are not coming any time soon, whether for technological, legal, economic, or cultural reasons. Let's concentrate on real solutions--eliminating the need for personal transportation by promoting walkable communities and proven methods of public transport (already available); bus, light-rail, subway, bike paths, etc. Leave aside the fact that all of these are more environmentally friendly than either gas or electric vehicles.

by stan on Dec 7, 2011 6:02 pm • linkreport

Although it seems the technology is already good enough to drive a vehicle in a mostly automated fashion, that last step to totally automated machines that we trust to drive around unattended is quite a reach. Lots of situations the vehicle would have to be able to handle, and lots of legal obstacles. Count me among the skeptical that we'll have that anytime soon.

Even after fully automated vehicles arrive (decades from now, I would guess), I don't think the basic dynamics of traffic will change that much. Even if they make traffic flow more efficient in some cases, there will still be congestion -- especially if a hugely expanded fleet of autonomous delivery vehicles is moving about. Folks may be overestimating the extent to which this development would change the calculation of whether to use a car versus other means of transportation.

by Teague on Dec 7, 2011 6:06 pm • linkreport

Bertie, have you really been to New York City? There's no way that most "off peak" trips could be absorbed by taxis so that people can have private trips instead of using mass transit--there just isn't the available road capacity.

This seems very implausible. But whatever the exact amount, there is LOTS of unused off-peak road capacity in New York City, even in Manhattan. Self-driving taxis will poach a huge share of the subway market. And that's in the most transit-oriented city in the country. The transit systems in other cities don't stand a chance. Who's going to bother walking to a transit stop, waiting around for the bus or train to arrive, waiting some more as it stops at all the intermediate stations before you get off, waiting again if you have to make a transfer, and then walking to your final destination at the other end, when for a lower price you can get a faster, more comfortable, direct, on-demand, door-to-door ride in a self-driving taxi? It's a no-brainer.

Not to mention that even without a driver, there is no way that a taxi ride will cost the same or less than a transit ride, especially if you have a monthly pass. Even at only double the rate of a single transit ride, the likelihood of tens of thousands of people switching is remote. Not to mention the capital cost of the vehicle is still what it is.

According to the National Transit Database, in 2009 (last full year reported), the NYC subway cost about 68 cents per passenger-mile. The average retail cost of owning and operating a private car is about 30 cents per passenger-mile. Self-driving taxis will almost certainly be cheaper than this. But even if self-driving taxis are no cheaper, New York could replace subway service for all its current riders with free self-driving taxi rides and still save money.

by Bertie on Dec 7, 2011 6:14 pm • linkreport

Wow! The anti-urban, pro-sprawl folks are having a field day with this. Why is this even on GGW?

As many posters have pointed out, this doesn't address the coming oil crisis or the environmental costs of promoting sprawl, which will make whether or not your car can drive itself a moot point.

by Allen on Dec 7, 2011 6:23 pm • linkreport

Good point, Allen! I fail to see how self-driving cars will make Washington DC a better city for those of us who prefer biking or walking. More than likely it will just promote sprawl and I shudder to think how many people will get flattened by these cars if they ever came to be. Just one "glitch" and you'll get whacked by a robocar going 60 mph through a city street. I'm sure as usual they'd blame the non-motorist.

I agree with Stan, we need more investment in public transportation and not be continually investing in anything that promotes more sprawl and less density

BTW, that Larry guy is nuts!

by terriG on Dec 7, 2011 6:29 pm • linkreport

Bertie, you're deluded.

First, even a self-driving car serving as a Manhattan taxi can't possibly hope to come close to the cost of operation of a normal private car. Taxis, per mile, use about 3x the energy of normal private cars and have 4-5x the maintenance costs, both because block-by-block in the central city is a brutal driving environment. And then there's the cleaning expense. I'd informedly speculate that the cost of operating a Manhattan cab is closer to $1.25/mile (outside of capital expense), and a self-driving car might reduce that to $1.

Second, a full 6 train can hold about 1000+ people. If you have ever spent any time on the East Side, you know they run every few minutes, and are full even during off-hours. There isn't the street capacity to put each of those 1000+ people into his or her own self-driving taxi.

by dal20402 on Dec 7, 2011 7:01 pm • linkreport

dal, I think you're the one who's deluded. Please show me the source for your claims that taxis "use about 3x the energy of normal private cars and have 4-5x the maintenance costs."

But the cost of today's taxis is irrelevant anyway. A typical self-driving taxi of twenty or thirty years from now will almost certainly be smaller and much more energy-efficient than a typical taxi today. It will most likely use a hybrid, plugin hybrid or all-electric drivetrain. These are particularly suited to urban driving environments. Maintenance costs will also be lower. Most taxi trips involve only one or two passengers, and little or no cargo, meaning that most self-driving taxis need be no larger than a Smart Car. This means the capital and operating costs will be much lower than those of today's taxis. In addition, taxi companies enjoy wholesale pricing and economies of scale that are simply not available to individual private car owners, who must pay retail prices for all capital and operating costs. Since the risk of accidents will be much lower, insurance costs will also be much lower. And cleaning will also likely be completely automated, and therefore cheap.

So the idea that self-driving taxis will even cost as much as today's private cars (about 30 cents per passenger-mile) let alone much more is highly implausible. Your estimate of $1 per mile for operating costs alone is preposterous.

by Bertie on Dec 7, 2011 7:47 pm • linkreport

If the cost of these auto taxis are low enough then, public bus systems in the suburbs could disappear. They could be replaced by auto mated mini buses connecting a dozen travelers journeys. It would require a sophisticated travel computing system,but a journey in the future could involve you looking on your phone pressing request cab and then adding a destination from your presets.

So on a Monday morning the system could give three of four prices. Option A for a cab all the way Foggy Bottom could be 15 bucks as the you want single occupancy at peak and the lane rental system only allows so many vehicles on any road.

Option B could be a mini bus that picks up 10 other people in your neighbourhood on the way downtown that could be 7 bucks say.

Option C could be 4 bucks with a trip to the nearest metro station and then switch to Metro.

Formal bus systems may whither away replaced by automated mini buses, but certain highly traffic routes will retain their trains, metro's and even bus lines because they can carry more passengers in the same space.

The denser the city the less likely any private cars could enter the city in peak hours, even small buses could be a bit pricy. Such a system would encourage high capacity transit in near the core.

Also all those parking lots, what to do with them? Well I'm sure land owners would be happy to monetise all that unused space, so more condo's offices and retail.

So your city becomes ever denser as more people can live close to work. At the end of the day an autocab may be cheaper but still won't drive any faster. That two hour commute still won't be attractive.

If there is an energy crunch it won't matter whether the car is electric or not it will still cost a lot more to drive, so the pressures to live close to work won't go away.

by Rational Plan on Dec 7, 2011 7:53 pm • linkreport

Option A for a cab all the way Foggy Bottom could be 15 bucks

But a robocab won't cost $15. It'll cost less than driving today. And less than the bus or metro. Given that ride-sharing is so unpopular today, especially ride-sharing with strangers, I have no idea why you think there will be a significant market for "automated minibuses" and the like, especially if the government subsidizes robocab rides for the poor.

Also all those parking lots, what to do with them?

Turn them into parks and squares and public spaces. Landscape them. Make the city greener.

Well I'm sure land owners would be happy to monetise all that unused space, so more condo's offices and retail.

They won't be able to monetize it with condos, offices and retail unless there's a market for those things (and zoning laws/NIMBYs allow them). Since self-driving cars will allow more people to travel greater distances in a given period of time, and largely destroy the demand for mass transit, they're likely to reduce urban density.

by Bertie on Dec 7, 2011 8:50 pm • linkreport

I agree, Bertie. I think by that point we'll have moved past the point where cities are relevant at all. I think in the future they'll either be bulldozed and used for "central park" areas for the suburbs, or they'll be used as a type of museum to see how people used to live in the pre-car, pre-suburban era. Bring on the self-driving cars! I don't know why these people can't get with the program and see that the future of America is in the low-density suburbs and highways and not in the crumbling 19th century cities.

by Larry on Dec 7, 2011 9:03 pm • linkreport

For once, I agree with Cavan.

by Rich on Dec 7, 2011 9:35 pm • linkreport

I define cars as personal transit. Driverless cars can be seen as a bus that can come directly to your door on demand. Yes, they could go point to point, or in trunk situations, they could be feeders to more heavy duty transit like light rail and HSR.

by John Bailo on Dec 7, 2011 9:41 pm • linkreport

People care about efficiency when and if it translates into lower costs to them personally. Transforming cars into a component in an overall fee-for-service system means people will use cars when, but only when, they are more efficient on a marginal cost basis.

I might note some of the scenarios above clearly depend on failing to account for the same technology penetrating all aspects of this future transportation system. You won't necessarily need to walk anywhere to get picked up by a multi-passenger vehicle, because these technologies will also facilitate flexible routing. And of course as necessary you could spend some time in a solo-passenger vehicle before switching to a multi-passenger vehicle, and then maybe back again to a solo-passenger vehicle. Multi-passenger vehicles will be adopting more efficient propulsion technologies along with solo-passenger vehicles. And so forth. The bottomline is that if there is any extended period in which a significant number of people are trying to travel the same route at the same time, these technologies will make it easier for them to use a lower-marginal-cost multi-passenger vehicle to their mutual advantage.

I also think it is important to note that one of the most important costs to car trips is the reduced value of the passengers' time. Of course relieving many passengers of having to drive will somewhat reduce this cost on average, but it is still unlikely that the time spent in the vehicle will be as valuable as time spent at the origination or destination. The shift from an ownership model to a fee-for-service model will again make this factor increase in relative importance. And only needing to make shorter trips will still be the most significant way of minimizing that cost.

by BrianTH on Dec 7, 2011 10:21 pm • linkreport

@Cavan:

Technology doesn't solve all problems. It never has and never will.

Who said that self-driving cars would solve all problems? By the way, did you comment via personal computer?

by ed on Dec 7, 2011 10:33 pm • linkreport

@ Larry:
smelly buses/trains with odd-collections of dangerous or annoying strangers

stay classy

Cars also suit our individualist tastes (collectivism is more of a European/Asian phenonomenon).

Gotcha: Others.

By 2050 when we have these cars, we'll look back and be glad that wasteful pointless programs like High Speed Rail and whatnot were squelched by forward-thinking individuals.

People will still have central locations where they will work, govern, attend a show, visit museums, see a sports team, and what have you. It is quite possible these will still be called "cities." High speed rail will quite likely still be an efficient way to move between these central locations in the future.

by ed on Dec 7, 2011 10:42 pm • linkreport

But the cost of today's taxis is irrelevant anyway. A typical self-driving taxi of twenty or thirty years from now will almost certainly be smaller and much more energy-efficient than a typical taxi today. It will most likely use a hybrid, plugin hybrid or all-electric drivetrain.
(Bertie)

I can't believe that this discussion hinges on the characteristics of a "typical self-driving taxi of twenty or thirty years from now." It's impossible to argue with conspiracy theorists because those folks dismiss any evidence as falsified. Same thing in this conversation, except here, we're talking about The Future! Where self-driving taxis do anything and everything that the commenter wants them to do.

It's not as if technology only advances for one mode of transportation. Mass transit vehicles are always growing lighter, more efficient, lower-maintenance, and less manpower-intensive.

by David R. on Dec 7, 2011 10:57 pm • linkreport

If automation will allow for closer spacing between cars on the road (therefore increasing a road's capacity), doesn't that require universal adoption of that automation?

As usual Alex B, my brother from another mother, hits it on the head. This is the fundamental issue. Like I said, it's a cultural issue, not a technological one. To say, "Oh, well, the development of automated cars will drive a cultural change" is naive.

Ask the @pelham1861s of the world to explain the problem to you.

Again, you can have their steering wheel when you pry it from their cold dead fingers.

by oboe on Dec 7, 2011 11:10 pm • linkreport

People care about efficiency when and if it translates into lower costs to them personally. Transforming cars into a component in an overall fee-for-service system means people will use cars when, but only when, they are more efficient on a marginal cost basis.

No, it doesn't mean that. I don't know why you can't grasp the obvious fact that people care about lots of other things besides efficiency. Things like speed, comfort, convenience, privacy and flexibility. Cars already overwhelmingly dominate our transportation system. The added capability of automonous operation will make them even more attractive than they are now.

As for "multi-passenger vehicles with flexible routing," they're already a part of urban mass transit, called demand response. Demand response is a tiny part of mass transit, because it's so expensive and slow. It's mainly geared towards elderly and disabled people who have difficulty getting to and from bus stops and train stations.

Multi-passenger vehicles will be adopting more efficient propulsion technologies along with solo-passenger vehicles.

Motor buses will likely be able to benefit from any advances in efficiency that can be applied to cars. There already are a significant number of hybrid buses and alt-fuel buses, and a few all-electric motor buses. Rail transit already gets the efficiency benefit of electric propulsion. No new technologies to substantially increase the efficiency of rail transit are on the horizon. But buses and trains won't disappear because they can't match cars on efficiency. They'll disappear because people will be able to substitute self-driving taxi rides at lower cost.

by Bertie on Dec 8, 2011 3:29 am • linkreport

I love how some assume that auto cabs will be so cheap that everyone will suddenly be able to drive twice as far as they currently do.

Freeways might increase in capacity with closer vehicles but city streets will need the same spacing. Pedestrians need to cross streets etc. Auto cars will not end traffic jams, in fact they could be a lot worse if everyone gave up on transit.

Also everyone is again assuming cheap energy that an electric car can drive for many miles etc. Electric vehicles will work in dense areas where journeys are short but long commutes mean long recharge times, or people in the far suburbs will still need gas.

If people own their own auto car then the same marginal cost decisions will come into play as they do now If you have already paid for it you might as well use it for every extra journey. Therefore if people rely on autocabs each journey they take will not only cost the few cents of electricity(ha)but also need to cover wear and tear, finance and depreciation. So while there is no sunk cost to personal mobility there is a real cost for every journey. While it won't be exactly the same, you would be better looking at the personal travel habits of people who use Zip Car for the future of travel patterns.

If you can't build extra road capacity now, why will you suddenly be able to in the future. Transit is a tiny share of transport in the Suburbs now, so one way or another it won't make much difference to the level of congestion.

At the end of day in busy areas you can't fit everyone into their own autocar and get to work, the roads won't have the space. You will only be able to run cars closely together once you ban self driven cars. Inside the Beltway will still need transit of some sort. The Metro will still form the backbone and some of bus network will still exist.

What will people do about congestion? Will they leave it as now or price it so the roads always flows. Either way it encourages some form of transit. The bus networks could disappear as public owned networks and reappear as private companies running rough routes to major destinations, picking people up as they bid for fares from peoples phones.

Living near a major route may still be an important locational decision. It will be easy to pick up shared rides on major streets at reasonable prices. Someone stuck half a mile up a sub division is not going to popular from other riders or the cab company. You could have other riders outbid you for their mini bus not to pick you up.

Richer people will either own their own autocar or pay for solo rides during rush hour. At different times of the day pricing would change. In the biggest cities some form of transit would operate all day and start to wind down after 8pm, as the roads empty the price for single cars would plummet. Eveneing transit could still exist, if single autocars are still constrained by high energy prices then, shared min buses would still be popular in the evening.

Would single ride trips be actually cheap. I don't see why people expect these cabs to be so cheap, it's not as if the cab drivers are rolling in it. How much would a cab cost if there were no wages to pay? A third less? Half? You think about that a 10 mile trip how much is that going to cost.

You could have a networked bus system. Local circulaters picking up fares and depositing them at local hubs where express buses run to other hubs. During peak times more circulaters will mean fewer deviations from set routes and high capacity links between hubs, at off peak times mini buses could deviate more so they pick up the same number of passengers (there would be fewer running, obviously, off peak). The inter hub links could then run less frequently or be less direct taking in several hubs.

It's also quaint to state that land owners will suddenly leave their land as open space, because it will be worth less because every one will leave the city. If land prices are high in the city now and city will allow them and tenants no longer caring about how much parking is available next to their building, I imagine they will look at other uses. Even in the suburbs that would happen. it will dpend how this will affect the retail market. Big box stores would not need so much parking. Structured automated parking, combined with automatice valet pick up at the loading zone in front of the store could see the footprint required to drastically shrink. it would take time but these strip malls and big boxes are just metal sheds on asphalt, it won't take much to redevelop. Some sites will just see a lot more pads plonked on their lots, but the consequences are clear a lot less land will be needed for retail. even if all that happens is a lot of tract housing with some low rise apartment complexes, that will still densify those suburbs and put yet more automatic vehicles on those suburban roads.

But what if the retailers don't need all that space anymore. You could see two trends. Fewer but massive stores, if people can travel further to shop, occupying smaller footprints as most of their parking is structured. The space between the store and the street is for loading and valet drop off, the rest of the parking can be stuck round the back. Other retailers may have smaller units that act as primarily as exhibition spaces for new products and impulse buys, but are mainly for click and collect internet purchases. Some chains could do both, depending on the size of the local market. Either way these are trends that exist without the need for automated cars.

If gas prices stay high and go higher the price of electricity is not going to stay low either, all those electric cars will need new power plants and barring some magical breakthrough in battery technology, electric cars will just not be as good as gas cars. Under such a scenario travel to work costs will be even more important, all that wasted space with all those single storey metal boxes for retail will be a tempting target for apartment blocks, all conveniently next to major streets, where new local downtowns could sprout. If you want to see the future just look at the proposed change to Tysons.

by Rational Plan on Dec 8, 2011 7:19 am • linkreport

Thanks for the great article Will, and the input Bertie. Hopefully cheap, convenient private transportation will be the death knell of the failed traditional urban community and will herald a new expansion era for the low-density suburban pattern that Americans know and love.

I look forward to a day when trains and buses are something people will see in museums only. Americans rejected urbanism, "walkability", and public transit decades ago, and nothing's going to force us back into these dank hellholes.

I wish the best for Google and their self-driving car! :)

by Larry on Dec 8, 2011 8:23 am • linkreport

@bertie
"According to the National Transit Database, in 2009 (last full year reported), the NYC subway cost about 68 cents per passenger-mile. The average retail cost of owning and operating a private car is about 30 cents per passenger-mile"

Well you this is true if you include all costs for transit and only people's personal cost for cars. You are forgetting things like the cost to provide and maintain roads. That is rather large. You are also ignoring all sorts of economic costs associated with congestion and the economic losses associated with using limited land for transportation infrastructure instead of other uses. But other than that you are correct.

by Nathaniel on Dec 8, 2011 8:31 am • linkreport

This discussion has gone from completely ridiculous to utter fantasy in one night. Reality check:

1. People will still want to own their own cars. For the same reason they don't want to take the bus - your car is cleaner, you know what's been done with it, it's instantly available when you want it, etc.

2. You may eventually be able to double highway capacity (when the tech is universally adopted HAH). You won't be able to do the same for urban or even suburban streets because who's going to allow zillions of cars to scream around pedestrian areas at 60MPH?

3. Vehicle operation costs will not decrease substantially due to this technology. The "30 cents per PM" figure is fantasy, it probably doesn't include parking costs (which can add up to half the total cost) and it's averaged out and includes tons of open-highway driving. It doesn't apply to urban situations and certainly not the cost of a taxi ride in NYC. Costs are 40% higher on urban arterials (http://www.vtpi.org/tca/tca0501.pdf) and go up from there. And if your solution to the parking problem is to have all the cars drive home during the day and back to pick up their owners, then you've effectively doubled vehicle operation costs anyway since half the trips your car isn't actually doing anything FOR you.

The people predicting the death of cities due to this technology are wrong mostly because this technology is not even close to achieving these fantasy ideas. But beyond that, they are claiming that the cars will be able to do THIS and THAT and THIS OTHER THING without thinking that all of those ideas are contradictory. You can't get rid of parking AND quintuple highway capacity AND send all the cars home and back in the middle of the day AND lower everyone's costs. All of those things will increase costs

by MLD on Dec 8, 2011 8:39 am • linkreport

flying car

But how will our flying cars land?!

We've wandered that far into the realm of speculative fiction that I'm not sure that a serious discussion of this issue is possible. Nevertheless, against my best instincts, I'll bring up the notion of the S-curve: that new technologies have very little effect for a long time, and only then does rapid change ensue. Experiments with containerized cargo, for instance, began around 1900, but the real change only came in the 1970s.

by David R. on Dec 8, 2011 9:03 am • linkreport

Self-driving cars will have unlimited room to operate once we get all Fifth Element and have multiple horizontal routes for flying cars to whiz by on at different elevations. Unfortunately, DC won't be able to accomodate this revolution. Damn you, height limit!

by HooShotYa on Dec 8, 2011 9:28 am • linkreport

Platooning vehicles on the highway may increase the number of vehicles accomodated, however, how long will it take until there are no humans allowed to "mess up" the automated travel? I think many years will pass before all humans are banned from the automated freeway. Several generations, to be sure.

by Vstrom44 on Dec 8, 2011 9:48 am • linkreport

Lotsa good comments here, especially from Rational Plan and MLD. The thing about this discussion is a lot of the pro-driverless car folk seem to be arguing the anti-city position.

The reality is that cities are about efficient exchange. Mobility is just what we consume in order to get to and from to conduct exchange.

You can talk all online-telecommuting-etc. that you want but the reality is that the "gestalt" of creativity, innovation, and generating new ideas, products, services, and knowledge requires face-to-face contact. At least it does for me, and I get plenty out of online interaction, but it isn't enough and can't replace face-to-face contact for generating new ideas and business.

The anti-city people are focusing more on how they think everyone wants to (1) live-- in a suburban mcmansion with lots of open space, even though most people can't live that way anymore, unless they go farther out and (2) get around-- in a car, by yourself.

It's true that only a subset of us humans act rationally and only on a certain set of conditions, not everything. People like me prefer living close to our major activities so that we can walk or bike or use transit to conduct activities efficiently without having to own a car and pay for its operation, including parking. If we live close, and to transit that is also efficient, we can conduct these activities relatively quickly, or if the time required may be greater than that for driving, at least we can do other things with our time, like read or use a computer.

Again, the Leinberger research shows that 30% of people want to live in cities, 40% in traditional suburbs, and 30% are happy with either.

That hardly sets the stage for auto-based suburban wonderland, especially these days in a scenario of increased global competition for natural resources of all types, especially oil.

by Richard Layman on Dec 8, 2011 10:07 am • linkreport

For those that are skeptical about how soon this is coming, the DARPA demonstration is very convincing. It was a race, it is quite fun to watch. Nevada is on the legal vanguard: it made an autonomous vehicle lawful. It is first state to do it.

Because the entire fleet will need to be replaced which will take 20-30 years, the adaptation of autonomous vehicles will evolutionary, not revolutionary. There will be no changes to the roads. During this period they will have to interact with human drivers, so there will be no benefits of sharing data between vehicles: that means there will be no computer-driven increase in the density of cars on the highway. Of course there will be a number of legal problems.

Regarding city living, if I were on foot or on a bicycle in traffic, I would prefer that cars are controlled by well-behaved and predictable computers rather than impatient, tired, distracted, and prejudiced humans. Regarding suburban living, if one does not need to drive but can instead do work or watch a moving during a commute, means that living further and further away becomes more viable.

by goldfish on Dec 8, 2011 11:42 am • linkreport

"Regarding suburban living, if one does not need to drive but can instead do work or watch a moving during a commute, means that living further and further away becomes more viable."

I don't think we should promote that! It would waste valuable resources and case even more watershed and habitat destruction. The last thing we need is more sprawl development or its promotion. We've had 70+ years of it already.

by claire on Dec 8, 2011 1:04 pm • linkreport

I don't think we should promote that! It would waste valuable resources and case even more watershed and habitat destruction. The last thing we need is more sprawl development or its promotion. We've had 70+ years of it already.

Who's the "we" here, claire? Big government? Well, I'm opposed to that! I say, let the market lead, and lead it will to where it has for a long time, to suburban living, where people actually want to live. Down with the big cities and their statist masters!!

by Larry on Dec 8, 2011 1:08 pm • linkreport

@Claire: I don't think we should promote that...

I was merely stating the facts as see them: this technology will enable longer commutes (just like the automobile did). There is no doubt that autonomous vehicles are coming, which obviously will affect the urban environment: you might as well try to stop the incoming tide. Given that, I think it is best to prepare for it.

BTW, I am a confirmed city dweller. But I think that those that want to live the exurbs, and are willing to make the sacrifices for this, should be able to. So actually I view this is a good thing, because it will relieve pressure on property prices in DC.

by goldfish on Dec 8, 2011 1:30 pm • linkreport

Larry,
It's "Big Government" that built the suburbs in the first place. Who do you think paid for the highways? As for where the people want to live, if you believed in the market place, you'd know the answer. I look forward to hearing more of your opinions, becasue what they lack in facts they make up for with passion.

by Thayer-D on Dec 8, 2011 1:41 pm • linkreport

@Bertie,

You should get your hands on some of Wendell Cox's stuff. You two would get along like a house on fire...

by oboe on Dec 8, 2011 2:27 pm • linkreport

I believe Google's experiments have already shown that driver-less cars can handle complex environments like cities. And this is very early in the game.

I don't want to nitpick, but Google doesn't have a driver-less car. They have a mostly driver-less car, which as Miracle Max will tell you, is not the same. It is a long way from "needs a driver very rarely" to "never needs one". It's the difference between 3 sigma and 6 sigma or possibly even something higher than that.

In order to get to driver-less taxis (where the passenger isn't asked to intervene every 2 miles) is a whole order of magnitude more difficult than what google is doing. What is the crash ratio people will accept for these?

So what I think we'll see are cars that need less driving, and are perhaps safer, but they won't park themselves, won't do away with taxis, etc... Which means none of the other advantages that Bertie and Larry are promoting will come to be. Making cars that are TRULY driverless and don't crash is very very difficult.

Teague said the same thing I notice....

If vehicles can be driven much more closely together because they're automated, you could increase that to maybe 10,000 cars per hour

I don't think they can be driven much more closely together. You still need an adequate stopping distance in case a child or deer jumps out in the road and the car in front of you slams to a halt. So there may be less reaction time, but I'm not sure that changes the following distance much. And since most people follow too closely behind the car in front of them, robot cars might acutally have larger spacing.

Because there's no such thing as a foolproof system.... I don't think people will tolerate risk

Bingo.

The crucial thing to keep in mind is that if driverless vehicles allow a transition from an ownership model to a fee-for-service model, then everyone will be making transportation choices case by case, based on marginal costs (transmitted via fees). So if the marginal cost of using the off-peak subway (or similar system) is lower than the marginal cost of using the one-person vehicle system, people will still have an incentive to use the off-peak subway--and again, many more people will be making decisions based on such marginal cost considerations.

BrianTH is killing it in this comment thread. Another excellent point.

In NYC, for example, the primary cost of a taxi ride, I believe, is the medallion. So getting rid of the labor will not reduce the price that much.

Bertie's argument that per mile travel by driver-less car will be oodles cheaper just isn't founded in fact. Even if one switches to electricity and gets rid of the labor of driving, that will not cut prices to the levels that Bertie seems to think it will. And the savings will still be offset by the higher price of the cars. And, once again, the technology isn't really there to get rid of the driver.

So what we're really talking about is safer personal vehicles. Which is great, and should make driving cheaper as insurance rates go down, but won't suddenly empty cities or kill railroads. We'll have true driverless trains before we have true driverless cars.

the ever-annoying and unrealistic 'green crowd' ... because they never explain where we are going to get all this EXTRA electric capacity

Solar and wind combined with car batteries and a smart grid are all you need.

I fail to see how self-driving cars will make Washington DC a better city for those of us who prefer biking or walking.

That may not be the argument. IF, and that's a big if, this kind of technology were to come into being - and it doesn't exist now - it would dramatically change the city. And it would be foolish to bury our heads and not think about what that would mean.

self-driving cars will allow more people to travel greater distances in a given period of time

I don't see how that is true. There is no evidence that these cars will go faster than cars today. And since time will still equal money, people will want as short a commute as possible. If land closer to work gets cheaper because of less parking needs, then people will move closer to work.

by David C on Dec 8, 2011 3:36 pm • linkreport

Should not this new technology be easier to adopt to rail transit, including street cars, thus lowering the costs of transit operations?

by Douglas Willinger on Dec 8, 2011 4:30 pm • linkreport

@Douglas Willinger: Should not this new technology be easier to adopt to rail transit, including street cars, thus lowering the costs of transit operations?
Obviously yes. So then robo-taxis will not have lower operating cost than rail. Since the costs remain comparable, the subway, streetcars, and buses of the city will remain.

by goldfish on Dec 8, 2011 4:58 pm • linkreport

It is probably much easier to automate a choo choo than it is to automate an automobile - because the choo choo has a defined system of rails ahead of it, it doesn't need to navigate as much, and with the railroad doing much of the navigation work for the so-called "outdated" technology, choo choos are already ahead of flivvers. Furthermore, if you combine a lot of people in a large vehicle, you will have less material in the vehicle per person - less metal cage stuff, for example, because there is less forward and rear metal in the front and back per person - and that probably means less mass in the vehicle per rider. This leads to less inertia per rider, which translates into greater efficiency. We already see this in the airline industry, where high fuel costs are leading airlines to ditch smaller planes for larger ones. This logic would hold true in the driverless car era, too. Second of all, cities will hold their own, because some people are naturally attracted to city living. New York City has a prestigious reputation, and I doubt folks will give up Manhattan just because their taxis now lack drivers. Taxi costs will probably shrink a bit, but subway costs might also decline. Unfortunately, unemployment might rise, because we have numerous transit workers and taxi drivers thrown out of work. (That fact is likely to spur resistance to self driving cars until we can find a way around the increased unemployment problem, which is costly in and of itself). My prediction is that traditional transit and taxis will hold their own in many places where they are common due to political inertia, while driverless cars (and other vehicles) open up new transportation options and opportunities (such as expanded transit service, particularly personal transit service, in suburbia). Amtrak will become more useful than it is now, because it will be linkable to by driverless rental car service. (Because there are railfans in this country, there is still a market for passenger rail trips. Some of those trips would be made much easier if there were a driverless car service to go to and from various stations, such as the one at Ticonderoga. Thus, Amtrak use may very well go up in the driverless car era, particularly if Amtrak promotes usage of those car services in combination with its trains. The Amtrak traveler is often on Amtrak because he or she wants elbow room and a train ride on the trip, not necessarily speed. So why should he or she want to take a robocar all the way, when he or she can just robocar to the choo choo and then robocar from the choo choo? The same goes for some other rail services. Some rail services may decline in that era because of straighter routes from home to work along highway routes. However, in places such as Manhattan, there'll always be a place for rail.)

by Rickyrab on Dec 8, 2011 5:16 pm • linkreport

The robocar will become a modal choice in its own right and it will ultimately lead to some merging of automobiles and transit, particularly over moderate to long distances, mimicking what the train does on rails. Cars would come together, and might conceivably group around common vehicles and attach to them via passageways. Why might people do this? Because they might want to meet other people, have meetings on the go, eat without having to stop, etc. This would be most effective if both roads and rails were put to work with such an option in mind. We are headed for a transformation of cars and transit, not an end to transit.

by Rickyrab on Dec 8, 2011 5:37 pm • linkreport

Ok, so what some of you seem to be saying is that if I were in the city and took a self driving car to pick up the kids, it would be like hailing a cab and would supposedly be cheaper than the cab. Suppose a lot of other self driving cars were also going about. Then they would get stuck in traffic like normal cars, right? Just like taxis do in Manhattan. They mix with normal cars and get stuck in traffic, even when they are in traffic consisting mainly of taxis (or self driving cars). It takes time for a self driving car to react to changes in the environment, right? Just like a driver's car. Robots need money to buy and operate, right? They use electricity and have moving (and other) parts that can wear out, and programming errors might be a headache. So is a robot necessarily cheaper than a person? So, back to my example. I get into the robot taxi, push some buttons or something, it goes ahead, and gets stuck in traffic, like a normal taxi does, and then there is the security problem. How do I know someone's not going to come into the car midtrip or some hacker isn't going to take over the car? At least a driver might be able to help. Then again, a driver might be dangerous himself, so who knows? Well, after getting stuck in traffic and moving ahead and getting stuck in more traffic and so on and so forth, I finally get to the child care place, only to get billed $20 for the ride, not that much lower than a normal taxi's charging. Why? Because medallions and taxi permits are expensive, robots might be expensive, and the taxi company resisted lowering rates anyhow (not to mention the regulators making up the citywide taxi rates). The fat cats merely cut out the employees to save themselves money and are protected by the medallions and permits from competition. So we wind up with less employment and not very much more in the way of benefits. So why bother? I'll likely continue taking the subway.

by Regina Farkas on Dec 8, 2011 7:23 pm • linkreport

I love how some assume that auto cabs will be so cheap that everyone will suddenly be able to drive twice as far as they currently do.

It's not an "assumption." It's a conclusion that follows from the elimination of labor costs for human taxi drivers, lower insurance costs due to lower accident rates, lower energy costs due to more efficient routing and less congestion, and other money-saving consequences of automated operation.

Electric vehicles will work in dense areas where journeys are short but long commutes mean long recharge times, or people in the far suburbs will still need gas.

The average commute is about 20 miles roundtrip. 80% of commutes are less than 40 miles roundtrip. Even today's electric cars have more than enough battery capacity for the vast majority of daily commutes on a single charge. The electric cars of 20 or 30 years from now will almost certainly have much higher ranges. But this is a separate issue from self-driving operation, anyway.

If people own their own auto car then the same marginal cost decisions will come into play as they do now If you have already paid for it you might as well use it for every extra journey. Therefore if people rely on autocabs each journey they take will not only cost the few cents of electricity(ha)but also need to cover wear and tear, finance and depreciation. So while there is no sunk cost to personal mobility there is a real cost for every journey.

The point is that the cost will be very low, almost certainly lower than the current 30 cents per passenger-mile of privately-owned cars. For a single-occupant self-driving electric vehicle, Brad Templeton estimates the average total cost per mile will be less than 10 cents. But the exact figure doesn't really matter. All of these numbers are much lower than the cost of current mass transit (about $1 per passenger-mile). When people are able to travel much further for a given cost than they can today, and at higher speed, in greater comfort, and with greater safety, they're almost certainly going to travel MORE.

by Bertie on Dec 8, 2011 7:43 pm • linkreport

Now, suppose I live in a nice house in the suburbs. Suppose I don't want to drive to work. I might own a robot car, and do all sorts of button pushing, and the car drives me much of the way into the city. However, then it has nothing to do. I would still need to either park it, which would waste land, or keep it in a holding pattern, which would waste gas. I might lend it out to others, but that might run into the same taxi medallion and permit headaches that anyone breaking into that industry would have. I could hire a robot car instead, but then there'd be the issue of whether the car would pick someone up or not and take that person elsewhere. A lot of people surge into the city for work in older cities in the morning and leave in the evening. That leads to a lot of cars without anyone to transport in the daytime. In more sprawling cities, the going might be easier because of suburban workplaces, but we run into the same problem: a lot of people are at work and not going anywhere for several hours, leaving a lot of lonely self-driving cars out there hoping for passengers. So where are they going to go? Back home? Wouldn't that waste gas? If they just move around, they waste gas; if parked, they waste space. This is the same issue with present day transit and cars; switching to robots doesn't get rid of it. Self driving cars will do much better in off hours and in off periods, where the crowds are light and there are more places to go and people are going all over the place, rather than just between work and home. Sure, I'd like a Herbie the Love Bug, but I wouldn't necessarily want to ride it in rush hour traffic, and parking would still be a headache. Also, in the suburbs, wouldn't you have to call for a self driving cab, like you do a normal cab? Then it'd take time for the cab to get you. So what does that leave for the working mom? Perhaps she'd want to ride a Herbie to the train station, get dropped off by Herbie, which then drives back to the garage, and take the train into the city, where she'd transfer to the subway. At the end of the day, she takes train back while Herbie picks up the kids, and then she gets into the car and goes home with the kids - it's a bit like park-and-ride, only with more flexibility, less parking space (and associated costs) and some more fuel and depreciation costs. In places like Phoenix, she'd take Herbie all of the way to work, while Herbie would either park or go wandering around, and I suppose some greater density would wind up limiting Herbie to a drop-off-at-the-station role. So a self-driving car has some uses, but it's not the end-all and be-all.

by Regina Farkas on Dec 8, 2011 7:55 pm • linkreport

So then robo-taxis will not have lower operating cost than rail. Since the costs remain comparable, the subway, streetcars, and buses of the city will remain.

The costs will not "remain comparable." They're not comparable now. Taxis are currently much more expensive, because of the labor cost for drivers. Eliminating taxi drivers will dramatically reduce the cost of taxi rides. Eliminating bus and rail transit drivers will reduce the cost of transit rides only slightly, because labor costs for drivers are a much smaller share of total transit costs than they are of total taxi costs. Transit users will have a huge incentive to switch to self-driving taxis, because the taxi rides will be so much cheaper than they are today. This will likely lead to a death spiral for almost all transit services. As more and more transit users defect to self-driving taxis, the cost per passenger-mile of transit will rise, and transit fare revenues will fall. So transit agencies will have to either raise fares or cut services. Which will create an even bigger incentive for the remaining transit users to defect, and so on.

by Bertie on Dec 8, 2011 8:00 pm • linkreport

Bertie and David -

Both of you seem to be arguing over the makeup of taxi costs. Where are you guys getting the information from? Also, I am a bit skeptical of Bertie's claims because of one problem: politics, and the fact that there are several taxi drivers out there that would stand to lose jobs to robotic systems. Also, taxi companies might want to preserve medallion and permit systems that overprice taxi service. In the meantime, transit services might get more widespread with smaller vehicles, which would become more affordable to operate than previously with fewer drivers. I still don't see the legacy railroads or bus mainlines disappearing any time soon, because robots, like humans, would still cause traffic. Which would drive up time costs. In the meantime, those mainlines would continue to stick with large vehicles, although park-and-ride lots might dwindle (becoming less necessary), because of less fuel costs per passenger with larger vehicles at high occupancy. (At low occupancy, it's a different story, and dense cities and suburbanized cities will follow their own bent on transportation.)

by Regina Farkas on Dec 8, 2011 8:21 pm • linkreport

Bertie's argument would be fine if a) cost were the only factor driving people's decision making (in which case there's no need for first class, stretch limos, etc.) and b) taxi costs (in time and money) could be driven to lower than traditional transit costs. Perhaps they can, but the math remains to be shown, and this math depends on the regulatory and economic (and political) environment, on the bidding process, and on the marketplace. On a densely populated route, big vehicles cost less per user than small vehicles, because of the need for fuel and energy to move a given mass. Since small vehicles have more mass per user than big vehicles, this consumes more fuel starting and stopping (and with a taxi in the city, as with transit, there's quite a bit of starting and stopping in traffic). On a lightly populated route, small vehicles win because of fewer passengers per vehicle, even when the vehicles are larger. Moreover, Bertie fails to account for robotic and mechanical breakdowns, as well as the welfare costs of taxi drivers being put out of work. What's to say a company won't just pocket the savings from the robots and keep charging the same prices?

by Regina Farkas on Dec 8, 2011 8:36 pm • linkreport

Bertie's argument that per mile travel by driver-less car will be oodles cheaper just isn't founded in fact. Even if one switches to electricity and gets rid of the labor of driving, that will not cut prices to the levels that Bertie seems to think it will.

Why not? What evidence is this claim based on?

And the savings will still be offset by the higher price of the cars.

Another evidence-free claim. The technology required for autonomous operation is not very expensive. Video, radar, GPS and possibly lidar for sensing and navigation, plus relatively modest computer power for processing. Much of this technology is already present in less sophisticated form in current high-end cars, and is becoming increasingly common in mid-range vehicles too. Adaptive cruise control, automated parallel parking, lane departure warning and prevention, pedestrian detection and avoidance, automated pre-collision braking, and more. And of course GPS navigation is already very cheap and hugely popular.

And, once again, the technology isn't really there to get rid of the driver.

Google's fleet of prototype self-driving cars has driven almost 200,000 miles in California and Nevada, in all sorts of environments and conditions, from crowded city streets in San Francisco to rural highways, without a single accident. No one is saying the technology is ready for mass-market commercialization TODAY, but it's getting close. In fact, it's advancing much more quickly than I expected it would just a few years ago.

I don't see how that is true. There is no evidence that these cars will go faster than cars today.

Self-driving cars will reduce travel times by reducing, and eventually eliminating, congestion. Not just through closer vehicle spacing on roadways, but by reducing accidents and maneuvers that disrupt traffic flow, and by optimal routing. Eventually, they'll eliminate the need for vehicles to stop at intersections too, except possibly at pedestrian crosswalks.

by Bertie on Dec 8, 2011 8:43 pm • linkreport

lower insurance costs due to lower accident rates, lower energy costs due to more efficient routing and less congestion, and other money-saving consequences of automated operation.

Ah, but don't forget that transit would also be affected by some of the same economics. Including the lower insurance costs and even (in many cases) the more efficient routing. I could see vehicle size on transit more closely meeting the ridership demands, like they do on the airlines today.

Now let's take this logic to its conclusion. Suppose robots, everywhere, become cheaper and less accident prone than humans. So now nearly everybody, except for businesspeople and merchants and government, is out of work. So how are they going to get money? Hmm? How is there to be a customer market? How are people supposed to live? And where would the market for robo taxis or whatever come from? Perhaps the same merchants and the robot community?

The free market is wonderful, but it gets abused. I am not a Luddite, but we need to find some solution for this unemployment problem.

by Regina Farkas on Dec 8, 2011 8:52 pm • linkreport

On a densely populated route, big vehicles cost less per user than small vehicles, because of the need for fuel and energy to move a given mass.

If transit vehicles could run at high loads on every part of their route at every hour of every day, they'd be very efficient. But they can't. Demand varies dramatically by time of day, day of week and route segment. A bus that is fully-loaded heading into downtown at 8am on a weekday morning may run almost empty when it's heading back out, or in the middle of the day, or at night, or on weekends. On average, transit buses are only about 25% occupied. About 7 or 8 out of 10 seats are empty. This is obviously not very efficient.

In addition, buses and trains are large and heavy relative to their passenger capacity. Much of the floor space must be used for access rather than seating. High-profile design to allow for access and standing passengers adds additional weight. On average, transit buses are no more fuel-efficient than cars.

by Bertie on Dec 8, 2011 9:08 pm • linkreport

Bertie's argument that per mile travel by driver-less car will be oodles cheaper just isn't founded in fact. Even if one switches to electricity and gets rid of the labor of driving, that will not cut prices to the levels that Bertie seems to think it will.

Why not? What evidence is this claim based on?

The free market is like a horse. It goes all over the place. And it's not even completely free, either, because of the various hands that large players and national governments have in it. So it might be cheaper or more expensive, depending on supply and demand. So either side might be correct.

by Regina Farkas on Dec 8, 2011 9:09 pm • linkreport

If transit vehicles could run at high loads on every part of their route at every hour of every day, they'd be very efficient. But they can't. Demand varies dramatically by time of day, day of week and route segment. A bus that is fully-loaded heading into downtown at 8am on a weekday morning may run almost empty when it's heading back out, or in the middle of the day, or at night, or on weekends. On average, transit buses are only about 25% occupied. About 7 or 8 out of 10 seats are empty. This is obviously not very efficient.

In addition, buses and trains are large and heavy relative to their passenger capacity. Much of the floor space must be used for access rather than seating. High-profile design to allow for access and standing passengers adds additional weight. On average, transit buses are no more fuel-efficient than cars.

Larger buses might have smaller buses substituted for them to account for lower ridership at different times of day. Robot cars would have the same problem that transit vehicles do: they either need a place to park, which uses up land, or will need to roam (perhaps empty), which wastes energy. If the vehicle size varied with the ridership, then occupancy would increase. Furthermore, the floor space that is used for access on a bus amounts to one narrow aisle, which passengers might use as part of their seating room and elbowroom during the trip. On trains, such space is sometimes used for standees, and the same often goes for transit buses. If buses were regularly packed, they would be more fuel efficient per passenger than cars, and, at any rate, buses with every seat occupied would have more volume available per unit of fuel used, which means more comfort per unit of fuel used. I noticed that at current usage rates, vanpools were more fuel efficient than other modes, followed by efficient hybrid vehicles, followed by motorcycles. Now, suppose we had a fleet of efficient hybrid robotaxis. The efficient hybrid mode would lose fuel efficiency because they'd be subjected to the same problems encountered by traditional transit vehicles - the need to drive around or park somewhere. And parking does, in fact, take up land, which costs money.

by Regina Farkas on Dec 8, 2011 9:31 pm • linkreport

I take it that "vanpools" are not aboard fuel-efficient hybrid vans. That would mean that a hybrid vanpool would be even more efficient per passenger. So...if efficiency alone is considered... conditions often favor large vehicles in theory (at least in densely populated areas or on heavily traveled routes), and minibuses and vans in practice, particularly if they start and end as vehiclepools like they do in some third-world countries. If one wants space, a large vehicle might be useful. If one wants cheapness, one might look for crowded buses. If one wants privacy and door to door convenience, one might go for a taxi. Of course, in suburbs, the tendency would favor smaller vehicles. In the city (by which I mean places like Miami or New York or other densely populated areas), using robot taxis all the time would be impractical because one can get from point a to point b on foot in many cases, and the traffic would still be heavy enough to warrant slowdowns. But they'd have a niche. So would driving and so would transit.

by Regina Farkas on Dec 8, 2011 9:57 pm • linkreport

MLD,
The "30 cents per PM" figure is fantasy, it probably doesn't include parking costs (which can add up to half the total cost) and it's averaged out and includes tons of open-highway driving. It doesn't apply to urban situations and certainly not the cost of a taxi ride in NYC. Costs are 40% higher on urban arterials (http://www.vtpi.org/tca/tca0501.pdf) and go up from there.

It's your claim that's a fantasy. Parking costs for self-driving taxis will be very low because they'll spend much less time parked than today's private cars, because they'll be able to seek out the cheapest parking, and because they'll be able to park at much higher density. And since they will almost certainly use hybrid, plugin hybrid or all-electric drivetrains, self-driving cars will be much more efficient than today's average car (average car = 20 mpg, Toyota Prius = 50 mpg, Nissan Leaf = 99 mpg-equivalent), and more efficient than cars that mix urban and "open-highway" driving.

You're also ignoring all the other factors that will make self-driving taxis much cheaper than the average privately-owned car today: wholesale pricing, economies of scale, lower insurance costs, lower congestion costs, and lower costs arising from their small size. As I said before, most taxi rides involve just 1 or 2 passengers and little cargo. So most self-driving taxis don't need to be any bigger than a subcompact, with correspondingly low capital and operating costs in addition to the savings I described above.

At 30 cents per passenger-mile, even today's private cars are much cheaper than mass transit, which costs about $1 per passenger-mile. Since self-driving taxis will almost certainly be even cheaper, there's simply no way mass transit will be able to compete with self-driving taxis economically. And even if it could compete economically, taxis beat the pants off transit on speed, comfort, convenience and flexibility. The market for transit buses and trains is simply going to disappear, with the possible exception of a few rush-hour routes with very high demand where congestion is so bad that self-driving cars could not provide a complete substitute.

by Bertie on Dec 8, 2011 10:15 pm • linkreport

Larger buses might have smaller buses substituted for them to account for lower ridership at different times of day.

But they don't. Transit agencies cannot afford to maintain a fleet of large buses that are used only a few hours a day when demand is high, plus a parallel fleet of smaller buses that are used at other times when demand is low. That's why they have to burn so much fuel hauling around all those empty seats at times and places where demand is low. There is no magic bullet for improving the efficiency of mass transit.

by Bertie on Dec 8, 2011 10:32 pm • linkreport

It's not an "assumption." It's a conclusion that follows from the elimination of labor costs for human taxi drivers, lower insurance costs due to lower accident rates, lower energy costs due to more efficient routing and less congestion, and other money-saving consequences of automated operation.

Bertie, if you're talking about what will happen in the far future with the adoption of technologies that don't even exist yet, there is nothing this can be BUT an assumption.

Also, you can't really have it both ways. Either congestion will go down, or people will abandon other forms of transit. But if people abandon other forms of transit, I think the best you can hope for is that congestion remains the same. There are only so many cars you can get into New York City at rush hour - robot or otherwise. That's physics.

For a single-occupant self-driving electric vehicle, Brad Templeton estimates the average total cost per mile will be less than 10 cents.

Well, since we're talking about things that don't exist - and that I suspect are far more complicated than you understand - what does Templeton think the price of Leprechaun chauffered unicorns is per mile?

But the exact figure doesn't really matter.

Oh but it does. A 1 cent per mile savings will be vastly different than 10 cents.

Also, you're talking about a technology that is at least 70 years away, if not farther. But you act as though there will no comporable changes in transit technology that will bring down prices there.

Why not? What evidence is this claim based on?

My own best guess. Same place your evidence comes from. Unless you have some.

And the savings will still be offset by the higher price of the cars.

Another evidence-free claim. The technology required for autonomous operation is not very expensive.

Well how much does it all cost? It's not free, right? So those costs will offest the gains. You made my point for me. I did not claim that they will entirely offset the costs. I'm not good at guessing the price or benefits of non-existent technology. But clearly all this added hardward will cost more.

Google's fleet of prototype self-driving cars has driven almost 200,000 miles in California and Nevada,

No such thing. Those aren't self-driving cars - not really. They mostly drive themselves. But you can't get out and let it go park itself. The driver has to be there to take over in some situations.

Self-driving cars will reduce travel times by reducing, and eventually eliminating, congestion.

What evidence is this claim based on?

by David C on Dec 8, 2011 11:28 pm • linkreport

Bertie's claims are wishful thinking. In the 50's or 60's, he'd be telling us flying cars are just around the corner. He can't back up any of his wild claims with one shred of evidence, not his "self-driving cars are just around the corner" claim or his "driving will become cheaper and cheaper until it's practically free" claim. He just says they're self-evident.

His positions are opposed to the ideas of this board, he basically wants endless sprawl at the expense of urban DC. If he wants to peddle this Wendell Cox/Randy O'Toole nonsense, let him...but I suggest ignoring him. He clearly is enjoying annoying people here.

by Allen on Dec 9, 2011 7:00 am • linkreport

What ever happens Bertie self driving cars will not have lower operating costs than ordinary cars today. There will be no sudden lowering of costs because in case you have not noticed cars are already mass produced items of tremendous complexity. If self driving cars remain gas powered then the only difference will be is that taxis will be cheaper, but each trip will not cost the same as if you owned the vehicle yourself, because:

Each trip will need to recover the cost of capital employed, so that includes the purchase price, depreciation, repair, fuel, insurance and of course profit margin. All automatic driving does is strip out the cost of a cab driver, nothing else.

They will become electrically powered if gas become too expensive. The problem with electric is that it has a slow recharge time, a low range and the batteries wear out by a 100,000 miles.

Those batteries are also incredibly expensive. While people talk of electric cars costing only a few cents per mile, that is true of the marginal cost but not the real cost. The cost of batteries will have to come down drastically for people not to be phased by effectively having to buy a new engine every 100,000 miles.

At the moment electric cars are expensive, can't drive very far or very fast or be refueled quickly. For them to become the norm will require many breakthroughs that might not happen.

The more likely scenario, for electric to take over, is that gas becomes to expensive, people switch to hybrids and electrics. But electrics only become viable for people in cities and close in suburbs, those far flung suburbs will need major restructuring to for them to survive. The most exclusive will survive, as those that live their can afford the fuel. Others that have some cohesive form can develop local downtowns with denser residential cores, but can they afford to add sidewalks, proper pedestrian crossing, cycle lanes and more importantly bulldoze through subdivisions, so there are paths and cycle tracks criss crossing them and linking them directly to main streets, so that is possible to get to transit quickly on foot. The drive to work might now be a drive to a park and drive lot a third of the way to work, where various business parks or large corporates run van pools. You'll just have to hope that you live in the same suburbs as your fellow workers, can you afford the gas this month otherwise? If companies start finding it difficult to recruit people because they can't afford to get to their out of town campus then a slow shuffling of locations will begin, it won't mean that everyone will flee to Downtown, Arlington or Alexandria, but those new suburban town centres will need to become real ones surrounded by high density housing.

So what will actually happen with self driving cars?

At first, if they actually work, is the number of accidents will be cut and insurance rates fall for motorists. All very valuable in it self and worth the cost of the system, but not anything that will change how things are done at the moment.

It's only when you can have a self driving car that has no requirement for human input other than, a shouted home, work or bar will you see real changes.

The easiest one to see is that manual drivers become regarded as smokers in society.
Insurance companies might begin to refuse to insure young drivers unless they either take advanced motoring class course costing thousands, or they only buy cars that can't be driven manually. This can be affected by politics, but I don't see too man votes in subsidising the insurance of young people.

It might become illegal for old people to drive themselves after they are 70. As the years tick by more and more metropolitan areas declare self driving illegal in the interest of public health. Live free and die states fight the other side and both sides fight to switch the interstates to be one or the other.

Attitudes to drink driving will change, it will take a change in the law, but if you own a fully automatic self driving car, why can't you drink and drive? If that is the case, well there goes half the taxi industry anyway.

On the other hand if you can't drive a car (because the insurance is sky high for manual drivers) will you be so attached to it? If you live in the city taxis will be cheaper in the evening rather than more expensive. Cities may decide that after a certain time transit is not worth the cost, if it's cheaper they could subsidise auto taxis in the evening instead. The city can still have it's minimum wage workers get to and from night shifts or early starts, just at a lower cost. Transit could start at 6am and end at 10pm in a biggish city.

Big changes to the way we live but not that many changes to where we live. Until urban school districts improve then the growth of inner suburbs will be held back. On the other hand the best public schools are where the wealthiest people are. Either they start private schools or they capture the local school board and change it to their requirements. If enough middle income people move back, their activism will change the local schools.

Does anyone have reliable figures for the running costs of a taxi including wages? If we know that then we know how much fares could fall. But considering how low american transit fares are they would have to be incredibly cheap to displace public transport, especially in places where they charge flat fares for buses or metros.

Unless there is a series of breakthroughs on green alternatives to gas, then electric vehicles will only take over if gas is very expensive and then any idea that we will see people fleeing ever deeper into the exurbs as fanciful. In that new era of electric cars the roads my be free flowing, not because, of fancy electronics, but because a third of the population can no longer afford to drive a car. Multicar households could shrink and even no car households could soar, these people will not want to be stuck somewhere where there is no transit or the drive to work is long. The can not afford the gas, and the electric is only good for a hundred miles just enough to get you to and from work with a bit spare for the grocery store, but if it takes you a few hours to recharge you need to plan your journeys. The gas powered auto taxi will still cost a fortune and the electric version only makes sense in denser environments.

by Rational Plan on Dec 9, 2011 7:13 am • linkreport

@Allen

His positions are opposed to the ideas of this board, he basically wants endless sprawl at the expense of urban DC. If he wants to peddle this Wendell Cox/Randy O'Toole nonsense, let him...but I suggest ignoring him. He clearly is enjoying annoying people here.

I'll give Bertie the benefit of the doubt and take his chosen alias at face value, but I would point out that every few months, some anonymous contributor shows up hitting the exact same points in the exact same style. I'm not saying "Bertie" ( or Mixner, etc...) is a sock-puppet for Wendell Cox, I'm just sayin' they're eerily similar.

In any case, you'll note that these issues are really the only ones that seem to hold any interest for this clique.

Which brings me to a phrase which I coined last night on my ride home from work: "The Tragedy of the Hedgehog", which essentially describes a person who knows "one big thing"...which is unfortunately wrong.

by oboe on Dec 9, 2011 9:11 am • linkreport

the savings will still be offset by the higher price of the cars.

I predict that--after R&D costs--these autonomous vehicles will be effectively free. Because, rather than physically constructed, they'll be grown in vats. Also, fuel costs will be negligible because they'll forage for biomass in urban parks and greenspaces while they're not actively occupied in transport.

Finally, they won't be subject to the congestion that plagues non-autonomous vehicles because they'll have the power of temporary flight, and will be able to teleport themselves over distances under 1.5 miles.

The real game changer will occur in 2017 when President Gingrich makes good on Science's multi-decade unfulfilled promise to give us jetpacks. These Pan-Dimensional Personal Mobility Packs (PDPMPs) will let their users experience zero congestion, speeds of up to 350 mph, and total freedom of travel (not cosetted by socialistic urban planners with the straight-jackets called "roads").

by oboe on Dec 9, 2011 9:21 am • linkreport

@@Allen

"Bertie's claims are wishful thinking. In the 50's or 60's, he'd be telling us flying cars are just around the corner. He can't back up any of his wild claims with one shred of evidence, not his "self-driving cars are just around the corner" claim or his "driving will become cheaper and cheaper until it's practically free" claim. He just says they're self-evident.
His positions are opposed to the ideas of this board, he basically wants endless sprawl at the expense of urban DC. If he wants to peddle this Wendell Cox/Randy O'Toole nonsense, let him...but I suggest ignoring him. He clearly is enjoying annoying people here."
-----

If opposing views are unwelcome on this blog, I'm certain David would have told us so by now.

BTW, much of what Bertie is saying makes sense - as does some of Randall O'Toole. If that bothers you to the point of implying their opinions are "annoying" and "unwelcome" to everyone here, you might wabnt to consider starting your own blog so you can have total control over what other people express.

Just a thought.

by ceefer66 on Dec 9, 2011 12:58 pm • linkreport

ceefer,
It's clear that Bertie wasn't listening to any counterarguments here. I think that was the point Allen was trying to make.

by delmar on Dec 9, 2011 1:34 pm • linkreport

Bertie, if you're talking about what will happen in the far future

I'm not talking about the "far future." I'm talking about what's going to happen over the next few decades. I think Will Handsfield has the timescale about right. The first commercial fully self-driving cars will likely appear within 10-15 years, if not even earlier, and by 2040 or 2050 they will be extremely common, and may have almost completely displaced conventional cars (and mass transit) because of their overwhelming advantages.

Also, you can't really have it both ways. Either congestion will go down, or people will abandon other forms of transit.

No, congestion will go down AND people will abandon other forms of transit. Transit's share of the urban transportation market is already so small, especially at off-peak times when there is very little road congestion, that it could be completely absorbed by self-driving cars at off-peak travel times in the vast majority of urban areas in the country without any increase in road capacity or congestion. And could be completely absorbed at peak times in almost every urban area with only a minimal increase in road capacity. Self-driving cars are likely to at least DOUBLE road capacity.

And the savings will still be offset by the higher price of the cars.

You haven't produced any evidence to support this claim. You haven't even offered an argument. It's just wishful thinking that ignores all the evidence on transportation economics and the cost of electronic technology.

Also, you're talking about a technology that is at least 70 years away, if not farther.

Huh? Google has prototype self-driving cars NOW. They've already driven 200,000 accident-free miles, day and night, in all sorts of weather conditions, on crowded city streets and rural highways and every other kind of driving environment. All the major automakers are also working on the technology. Mercedes is expected to debut a new option on its 2013 S-class models that will allow full autonomous operation in heavy traffic at low speeds. GM has announced it expects to have a similar option for its cars by 2015. Volkswagen recently demonstrated its "Temporary Autopilot" system that allows semi-autonomous operation up to 80mph.

But you act as though there will no comporable changes in transit technology that will bring down prices there.

There won't be. What changes in transit technology do you envisage that will bring its costs down even remotely as much as eliminating human drivers will bring down the costs of taxis? Eliminating bus and train drivers will reduce transit costs only slightly, because the cost of a transit driver is spread over so many more passenger-miles than the cost of a taxi driver.

Well how much does it all cost? It's not free, right? So those costs will offest the gains.

No, they won't. Self-driving technology will add at most a few thousand dollars to the cost of a vehicle. This is just a few cents per passenger-mile. It's trivial. Computers and wireless communications are already cheap and getting cheaper all the time. Cars are already substantially computerized. Video cameras are already cheap and common in new-model cars. GPS is already cheap and common. The benefits of self-driving cars in faster travel, greater comfort, reduced congestion, reduced accidents, higher productivity, etc. will swamp the small costs of the technology.

What evidence is this claim based on?

I already told you. For a more complete treatment, read Brad Templeton's discussion of Traffic Congestion & Capacity

by Bertie on Dec 9, 2011 7:17 pm • linkreport

I'm talking about what's going to happen over the next few decades....Self-driving cars are likely to at least DOUBLE road capacity.

You haven't produced any evidence to support this claim. You haven't even offered an argument. It's just wishful thinking that ignores all the evidence on transportation economics and the cost of electronic technology.

I looked at Brad's website and it's all speculation. It sounds like a popular science article from the 50's about how by 1990 we'll all be flying to Mercury for the weekend. There are no cars in the world that can do any of the things he's talking about.

What changes in transit technology do you envisage that will bring its costs down even remotely as much as eliminating human drivers will bring down the costs of taxis?

What share of a taxi ride fare goes to pay the driver, because for WMATA over half the operating costs are related to labor. And they also have casualty and liability losses. They also would save money from cheep electricity. Buses would also benefit from smoother moving traffic and less traffic jams. Wouldn't hopping on a bus that only makes smart stops be cheaper than hopping in my own car?

You haven't produced any evidence to support this claim.

I don't have to. You did it for me "Self-driving technology will add ... a few thousand dollars to the cost of a vehicle. This is ... a few cents per passenger-mile."

Google has prototype self-driving cars NOW.

Again. They are not self-driving. They have to have a passenger in the car who can take control. And they have to do this several times per trip. That's not self-driving - not enough to do half the fanciful things Brad thinks they'll do.

I already told you

No I don't think you did. And if your evidence is Brad's oh-so-1993-looking website, I'll need more. But even he notes that it won't work so well when there are a lot of pedestrians. By the way, this does not mean that I endorse him as an expert, I'm merely using your own source against you, to discredit your position. I do not think Brad knows what he's talking about.

Because robocars will always yield to pedestrians, it could actually encourage more jaywalking. This could reach a level where the constant need to slow down for pedestrians causes road congestion, and you can't meter the pedestrians, only make rough predictions on their numbers.

That seems like a pretty big monkey in your wrench. He also notes that transit will still be needed - and talks about how it also can be made better. Another monkey.

But in the end, Brad doesn't give any evidence that "Self-driving cars will reduce travel times by reducing, and eventually eliminating, congestion." He gives a lot of conjecture and no evidence. And he doesn't consider induced demand or the fact that making roads safer - which should be the primary goal - will almost surely mean slowing cars down (Actually he does and notes that people might be willing to accept that).

He does tell us how the first super-beings may descend from apes instead of humans, and if we're lucky they'll keep us as pets. So I'm a little more worried about that than the future of American cities.

So no, you didn't already tell me. But I'm waiting.

by David C on Dec 9, 2011 11:16 pm • linkreport

I looked at Brad's website and it's all speculation

Huh? He describes in detail the many specific mechanisms by which self-driving cars will reduce congestion. If you think his facts and arguments are wrong, then you'll have to do more than just dismiss them as "speculation." All predictions are "speculation."

There are no cars in the world that can do any of the things he's talking about.

Wrong again. Adaptive cruise control, which is already available in production vehicles, maintains a constant distance from the vehicle in front with far greater accuracy and far faster reactions than any human driver, reducing congestion by allowing such vehicles to more closely and reliably follow the vehicle in front. Automated parallel parking systems allow vehicles to parallel park in tight spots that the driver would otherwise be unable or unwilling to use. Some GPS navigation systems already provide real-time traffic data allowing drivers to avoid congested roads and areas. Smart parking systems already provide real-time parking availability information, reducing congestion caused by vehicles driving around looking for parking. And so on. Automation is ALREADY providing some reduction in congestion. Self-driving cars will massively increase these benefits.

What share of a taxi ride fare goes to pay the driver, because for WMATA over half the operating costs are related to labor.

You didn't answer the question. I repeat: What changes in transit technology do you envisage that will bring its costs down even remotely as much as eliminating human drivers will bring down the costs of taxis?

A single WMATA metrorail driver drives a train with dozens of passengers, perhaps hundreds of passengers at peak travel times. A single taxi driver rarely drives more than one or two passengers. Eliminating drivers will therefore have only a small effect on the cost of metrorail, but a huge effect on the cost of taxis.

Again. They are not self-driving. They have to have a passenger in the car who can take control. And they have to do this several times per trip.

You're wrong yet again. Google's prototype vehicles are ENTIRELY self-driving. The human passenger is only there for safety and to comply with the law.

by Bertie on Dec 10, 2011 12:03 am • linkreport

As usual this is just devolving to a point where you play dumb and repeat yourself and make stuff up.

Brad has no facts on his website. He just imagines what life will be like if these magical cars come into being. This is one guys imagination. It is not fact.

The robocars we have now do not do the things you're talking about. They don't find their own parking. They don't find their way home. They don't adjust to traffic to find alternate routes. They don't use a hive mind to reduce congestion. They don't move out of parking spaces at rush hour. They can't safely drive without a driver.

What they can do is a very small thing which is help a driver drive more safely and, in one prototype, allow them to take their hands off the wheel for most of the time. That is very different from what you're talking about. You're talking about a technology that you're still only HOPING will exist. It doesn't yet. We don't even have it for trains yet and that is an infinitely easier situation to handle. When we have a real prototype, one that can drive 100,000 of miles without any monitoring by a person. Just say "go to Disneyworld" and it goes, then I'll be worried. But that is a long way off. I once helped build a car and I've been in the tech business almost my whole life so I'm talking from experience here. These last steps are much more difficult than the first ones.

What changes in transit technology do you envisage that will bring its costs down even remotely as much as eliminating human drivers will bring down the costs of taxis?

These same robo-driver technologies you have pointed to. The ones that don't exist.

Google's prototype vehicles are ENTIRELY self-driving. The human passenger is only there for safety

So without the human passenger, the car isn't safe. That does sound entirely self-driving. And self-crashing.

I get it you want a vehicle that is cheap to use, doesn't require much parking, doesn't cause congestion, doesn't cause pollution and can be shared among people to reduce storage needs. And you're hoping this magic car will be that thing. But it isn't, because there isn't one. Maybe some day there will be something that fits the bill.

by David C on Dec 10, 2011 12:47 am • linkreport

We don't even have it for trains yet and that is an infinitely easier situation to handle. Oh? I've been on automated trains. JFK and Newark Airtrain. (Not that those things actually go that quickly.) London Heathrow has automated personal rapid transit. Heck, elevators have been automated for decades.

They don't use a hive mind to reduce congestion. That would be fun to watch if they did. All that jerking, starting, and stopping, and quick negotiating with other cars in order to find a way through the streets. I suppose it'd make a nice amusement park ride. I also suppose people will prefer robocars that move more slowly and smoothly in a heavy traffic situation...even if some gridlock, jams, and slowdowns result.

by Rickyrab on Dec 10, 2011 4:22 am • linkreport

Robocars, if they are to become welcomed after people's honeymoon with them wears off, must follow the Asimov Three Laws of Robotics:

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.

2. A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.

3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws.

by Rickyrab on Dec 10, 2011 4:26 am • linkreport

I can a few additional robot laws.

4. A robot car will always finds me a sweet parking spot, near the entrance to the Mall, or what is the point, I don't want to wait 10 minutes as it queues with the others, at the auto valet pick up point.

5. A robot car will pay attention to what I say. When I say 'go to mom's' I don't mean my god dam mother in laws.

6. A robot car should tell me that I've left my briefcase in the trunk before it drives off to get serviced. And it should also not nag me all the time about remembering to take what I need from the car.

7. It won't be my fault if the kids are better at overriding the security code than me, and end up mall hopping all day, rather than going to school.

8 When I'm outside the bar it should be able to find me and not end up giving lifts to other drunks and then drive off leaving me behind.

9. It should let me run over those god dam kinds who keep playing chicken in the street.

by Rational Plan on Dec 10, 2011 5:05 am • linkreport

The advent of the self-driving car will lead to a new beastie being seen on some public transit systems: the podcar on rails. And I suspect the easiest place to put such beasties will be on already automated systems that run frequently but apparently get few people per trip, such as the JFK Airtrain, but even New York's Interborough could wind up using podcars in the off-hours (or wee hours of the morning). This is, essentially, a form of personal rapid transit, and it would be used in much the same manner as the current subway/bus system, with a twist: maps would be more commonly used. When this happens, the efficiency and economics of the transit systems improves, because podcars would come when the riders demand it, or nearly so. In high frequency, high ridership situations such as the daytime Lexington Line and other Manhattan subways, they will use traditional trains, because there would be hardly any change in the basic economics of the system.

Now let's look at the situation out in suburbia and edge city land. Systems like NJ Transit and taxicab companies vie to put robocars in place. However, there's one small problem: it takes time to respond to someone's request for a cab. True, robocars might get there more quickly, but they still have to come from somewhere. So robocar x will leave from parking lot y, probably somewhere close to the customer, but not necessarily so. This will require parking spaces and hookups near the customers, but, because of where the customers are, population density matters. In all likelihood, you're going to have to wait a few minutes before a robocar comes, unless you're at, say, a personal rapid transit station with robocars on hand. There's also option b, own your own robocar, in which case you pay for the costs and you pay for your own damages. You could rent the robocar to others, true, but that requires finding customers, and there will be costs associated with that, too. Nonetheless, there will likely be public transportation services, and robocars will make them more reliable and efficient per passenger mile. Which will make "transit" (actually, a mixture of transit and paratransit) cheaper to taxpayer and perhaps user. But we will still need some kind of transit map, and we will still need some kind of station, for some robocar services, because it's more efficient to have customers within easy reach of parked robocars then to have robocars roam far and wide to get to the customers.

In exurbia, residents will likely own their own robocars, but some taxi company or regional transit agency will still need to deal with incomers and outgoers. In this case, some base near where most of the activity is (say, a shopping mall, or a community clubhouse) will be the home base for robocar taxi services, with wait times of up to 10-20 minutes or even longer. (It still takes time to go from point a to b, remember.)

We will see new robobus and robotrain services making use of the robocar, effectively making a new form of transit. Much travel uses corridors, such as highways and roads, because it's efficient to corridorize transportation, due to infrastructure costs. (If travel was all by air over short distances we'd probably see a somewhat different dynamic, but, even then, travel would clump around points and activity centers, producing natural corridors of travel. Going to and from major urban areas is cheaper than going to and from minor ones today by air, after all.) You build main roads and people set up camp in the vicinity of those roads. So you're going to see clustering and clumping, which is efficient in and of itself because it makes use of aerodynamics, and some of those clumps will wind up connecting to central vehicles (hey, people want to eat on the go, and some might want to meet and shop on the go, so why not?), producing a new form of transit, at least on longer corridors. This will be especially true of intercity trips, which can be several hours in length. People still have a need to get up and stretch and walk around, after all.

by Rickyrab on Dec 10, 2011 3:17 pm • linkreport

But they don't. Transit agencies cannot afford to maintain a fleet of large buses that are used only a few hours a day when demand is high, plus a parallel fleet of smaller buses that are used at other times when demand is low. That's why they have to burn so much fuel hauling around all those empty seats at times and places where demand is low. There is no magic bullet for improving the efficiency of mass transit. Regulatory changes can take care of that. Besides, transit agencies can conceivably afford a mix of vehicles, with a lot of robocars to go around, if the political will is there to have a lot of robocars in a transit fleet. The same is true with robovans and robobuses. Even today, agencies use minibuses on some of their smaller routes, and no doubt they would expand their use in the robot era. At Rutgers University, we have a mix of articulated buses, nonarticulated buses, and minibuses, all of which have human drivers, and we do run a paratransit service after hours called the Knight Mover using minibuses. Granted, it's not perfect. But robocars could conceivably be used for the after hours service, robovans could come into play in the evening and while classes are in session, and the big buses during classes. Furthermore, companies could pool their resources to rent vehicles for use during times of greater demand, while outsourcing vehicles during other times (or just leaving them parked somewhere). For example, Rutgers hires buses from Academy Bus Company for use during football games.

by Rickyrab on Dec 10, 2011 3:55 pm • linkreport

The robocars we have now do not do the things you're talking about.

I just gave you several examples of "the things I'm talking about" that some cars can already do. It's the same old pattern with you. You make some absurd claim. Someone points out to you that your claim is false. But instead of acknowledging your error and reevaluating your position you simply ignore it and move on to your next false claim.

These same robo-driver technologies you have pointed to. The ones that don't exist.

How will robo-technologies will bring down the costs of mass transit even remotely as much as they will bring down the costs of taxis? As I have already explained to you, twice, eliminating drivers from buses and trains will reduce transit costs only slightly, because the cost of transit drivers is spread over so many more passenger-miles than the cost of taxi drivers.

So without the human passenger, the car isn't safe.

No, as I already told you, Google's prototype self-driving cars are extremely safe. They have now driven almost 200,000 miles without a single accident. The human occupants are only there in case something did go wrong, and to comply with the law. As Google continues to enhance the technology, the cars will become even safer than they already are.

by Bertie on Dec 10, 2011 4:41 pm • linkreport

taxis beat the pants off transit on speed, comfort, convenience and flexibility.

Flexibility, yes.

Convenience, depends on the location, as I've said.

Comfort, dubious. Legroom and elbow room account for much of the comfort thing, and this is going to depend heavily on the particular vehicle. I crammed into a NYC taxi with two other people and had less legroom and elbow room than I might have had on a full bus.

Speed, probably. But a lot of taxis would probably wind up getting stuck in their own traffic, so this is not quite certain. And don't forget that taxis are going to have to come from the back of beyond in some urban areas, which takes time, robotized or nonrobotized. Transit needs to make multiple stops to pick up and drop off people. Robotaxis might behave in a similar way to economize on vehicle-miles traveled, which would lead to an effect similar to that of transit.

Transit agencies cannot afford to maintain a fleet of large buses that are used only a few hours a day when demand is high, plus a parallel fleet of smaller buses that are used at other times when demand is low.

One example of a fleet with multiple sizes of bus is the WMATA Metrobus/MetroAccess fleet. Current practice uses smaller vans for paratransit service for people with disabilities and the larger (30 feet to 62 feet) buses for everyone else.

On the roster (for both "fleets" combined): 86 30-foot buses, 20 37-foot buses, 1165 40-foot buses, 309 42-foot buses, 43 60-foot buses, and 22 62-foot buses --- and also over 600 paratransit vans. (Subcontractors provide the vans, but they run under the MetroAccess label.)

While smaller bus fleets do tend towards standard sizes, there's no actual need for this: small agencies can conceivably rent buses from larger holding companies, such as First Transit and Academy Bus Inc., which have a mix of buses in reserve. (Academy already offers "54, 38, 28 and 15 seat vehicles", according to its charter bus website.) And so a regime can be set up that runs different size buses at different times of the day.

by Rickyrab on Dec 10, 2011 4:50 pm • linkreport

In fact, a robot transit system will be easier to run than a driver oriented system because you don't have to do run cutting (i.e., matching drivers to vehicles for runs and adhering to union-based driver work rules).

by Rickyrab on Dec 10, 2011 4:53 pm • linkreport

Nonetheless, there will likely be public transportation services, and robocars will make them more reliable and efficient per passenger mile.

No there won't. No one is going to bother walking to a bus stop or train station, waiting around for the bus or train to arrive, waiting again as it stops at all the intermediate stops/stations before they get out, waiting again if the trip requires a transfer, and then walking from the stop/station at the other end to their final destination, when they can get a fast, comfortable, on-demand door-to-door taxi ride for less money. Demand for mass transit will collapse. The only places where some buses and trains might survive are very high volume routes at peak travel times where congestion is so bad that self-driving cars would not be able to completely substitute for mass transit.

Regulatory changes can take care of that.

How are "regulatory changes" going to conjur up the huge amount of money transit agencies would need to maintain two parallel sets of vehicles, one set for peak travel times and the other for off-peak? Transit agencies are chronically short of money. They can barely afford to run their existing services. When self-driving cars appear, and transit users start switching to robocabs, transit ridership will fall, fare revenues will fall, and transit agencies will need even bigger subsidies than they already receive just to maintain their existing services.

You don't seem to grasp just how disruptive self-driving cars will be to buses and trains. As Will Handsfield says, it'll be like the transition from horse-drawn vehicles to motorized vehicles. Over the course of just a few decades, horses essentially disappeared from urban streets, because they just couldn't compete with motorized vehicles. Buses and trains will suffer the same fate.

by Bertie on Dec 10, 2011 5:09 pm • linkreport

No there won't. No one is going to bother walking to a bus stop or train station, waiting around for the bus or train to arrive, waiting again as it stops at all the intermediate stops/stations before they get out, waiting again if the trip requires a transfer, and then walking from the stop/station at the other end to their final destination, when they can get a fast, comfortable, on-demand door-to-door taxi ride for less money.

I disagree. Some people do, in fact, like walking as part of their travel. It gets them exercise, and sitting after walking is important to their lives. There are also busfans and railfans, and some people just like the sheer amount of space they get aboard a train or bus, and so they'll pay a premium for that.

waiting again as it stops at all the intermediate stops/stations before they get out,

I've seen passengers on both local and express trains and on local and express buses. Express services have fewer stops than local services. I admittedly prefer express services, but local services are also fine.

Demand for mass transit will collapse. The only places where some buses and trains might survive are very high volume routes at peak travel times where congestion is so bad that self-driving cars would not be able to completely substitute for mass transit.

In smaller areas and less dense areas, maybe, but the demand might also increase. People congregate around corridors, and combining smaller vehicles into larger vehicles can increase efficiency of travel (and so can combining smaller groups of travelers into a larger group on a larger vehicle).

How are "regulatory changes" going to conjur up the huge amount of money transit agencies would need to maintain two parallel sets of vehicles, one set for peak travel times and the other for off-peak? Transit agencies are chronically short of money. They can barely afford to run their existing services.

As I've said before, holding companies can hold a pool of vehicles of different sizes, and transit agencies can rent from them. Even with the distribution of driver costs over more passenger miles, there is still a benefit to transit agencies for cutting the driver out of the equation. Moreover, regulations resulted in plenty of oversize buses on various lines, due to optimistic ridership estimates. Furthermore are also restricted by union work rules at the moment. Many taxi companies probably don't have union work rules for their drivers (a fact which makes the impact of cutting drivers out of transit more and the impact of cutting drivers out of taxis less).

Transit agencies are short of money because they use overbuilt structures, rely on union work rules, overpay drivers, and use oversized vehicles. A bank would notice that. A venture capitalist might say, "Hey, wait a minute. If you used and maintained a mix of different vehicle sizes, like the airlines do, then cost per passenger-mile will come down, occupancy would go up, and transit service will be profitable on route x, y, and z". Subsidies may well get SMALLER, as transit agencies supplement their buses and trains with robo-rail and robo-bus and robo-van and robotaxi services.

by Rickyrab on Dec 10, 2011 5:36 pm • linkreport

The only places where some buses and trains might survive are very high volume routes at peak travel times where congestion is so bad that self-driving cars would not be able to completely substitute for mass transit.

Ahh, but if that logic were true, Amtrak would've vanished long ago, and so would the Vegas monorail and some other services. Transit agencies and services will stay alive, even if it means politicians wringing hands and pointing out how the use of a mix of different vehicles will reduce per-passenger-mile cost.

by Rickyrab on Dec 10, 2011 5:41 pm • linkreport

I think people can prolly agree on the following:

-Anything with a wait time of more than 10-20 minutes will run into trouble (oh wait, such services already have, but they're still around).

-Anywhere where there's a crowd will likely still have some form of transit, because of the sheer difficulty of moving a crowd.

-Legacy transit and tour systems will survive because they are important to both tradition and tourist industries. San Francisco never gave up its cable cars, and it ain't about to start now.

-Some new forms of transit will arise because people want to get up while still traveling. And meet other people while still traveling. And go on tours. And shop.

-Political inertia will keep transit systems running in many places, and political processes may even result in new transit systems, even if their per passenger mile cost is high. It has happened before and it'll happen again, especially in good times when people are more worried about what the kids are up to than about money going down the drain.

-The Old Order Amish never gave up their hosses and buggies.

-It will be easier and faster to catch a robotaxi in some areas than others. Especially in places where people live.

by Rickyrab on Dec 10, 2011 5:56 pm • linkreport

horses essentially disappeared from urban streets, http://www.centralpark.com/guide/activities/horse-drawn-carriages.html http://www.capemaycarriage.com/schedule.htm http://www.bostonbridalcarriage.com/

Only "essentially", I guess.

by Rickyrab on Dec 10, 2011 6:01 pm • linkreport

People who trade car ownership for getting around by taxi will save so much money in fixed ownership costs that they'll be able to travel much greater distances than they do today and still save money. But I doubt most people will trade ownership for taxis. Most of the market for self-driving taxis will probably come from people who are today dependent on mass transit or walking/biking, or for whom ownership is only marginally attractive.

(from an earlier Bertie argument)

So, humankind is only economic when it comes to transit taking and not auto ownership?

But I doubt most people will trade ownership for taxis.

Even though it costs less money?

No one is going to bother walking to a bus stop...when they can get a fast...door-to-door taxi ride for less money.

Tea Party lazy-libertarian logic detected here.

by Rickyrab on Dec 10, 2011 6:32 pm • linkreport

No one

There are always exceptions to the rule, you know.

by Rickyrab on Dec 10, 2011 6:34 pm • linkreport

I just gave you several examples of "the things I'm talking about" that some cars can already do.

Those examples are not things that you need to destroy transit and cities. Those things make driving easier and safer.

You make some absurd claim. Someone points out to you that your claim is false. But instead of acknowledging your error and reevaluating your position you simply ignore it and move on to your next false claim.

Are you talking about me or yourself? It sounds like you.

As I have already explained to you, twice, eliminating drivers from buses and trains will reduce transit costs only slightly, because the cost of transit drivers is spread over so many more passenger-miles than the cost of taxi drivers.

You haven't explained it so much as stated it. Two questions you repeatedly refuse to answer.

1. What percentage of taxi fares go to pay for the driver?
2. What percentage of transit fares go to pay for the operators? Include all the costs associated with the issues Rickyrab mentioned.

No, as I already told you, Google's prototype self-driving cars are extremely safe. They have now driven almost 200,000 miles without a single accident.

But they have never been driven on city streets without a driver who can and does take over time to time. In the Times article linked at the top, the driver took over 2 times in one trip. That is not a robotaxi. Not even close.

BTW, the Google car has been in at least one accident.

As it turns out repeating the same untrue thing over and over again doesn't make it any truer. So though you "keep telling me" things. These things are false; which is why I push back.

Look, in the end I don't really care if something comes along that is so awesome that transit dies. It turns out that more than I like transit, I like better even better. So if it can achieve all the goals of transit for less cost or achieve them even better, great. Bring it on.

But it won't kill cities. That's ridiculous. As Layman pointed out a lot of people like living in the city. That won't change. If it helps people who live in the city but want to live in the suburbs (and vice-versa) move to where they long to be, all the better. But I don't see a lot of people who live in DC who long to move to the suburbs but only lack a cheaper way to commute. Maybe they are there, but I don't buy it.

As an exercise ask people where they would live if they could live anywhere. You'll get a lot more Paris, Manhattan and London then you will Oak Cliff, Elizabethtown and Redondo Beach.

We need cities. That's where ideas come from. People interact more. (See Richard Florida for more on my opinion on this)

And we really can't afford sprawl anyway. It's killing the Chesapeake for instance. I don't think this transit system will mitigate those impacts.

So replace transit with something better - great. Replace cities with suburbs - no. That would be awful.

by David C on Dec 10, 2011 6:58 pm • linkreport

One more thing. There are many billionaires who are so rich they could choose to live anywhere in the world they want. Many choose to live in high rises in cities. Trump and Bloomberg both live in Manhattan, for example. I suspect most billionaires live in cities.

by David C on Dec 10, 2011 7:01 pm • linkreport

I disagree. Some people do, in fact, like walking as part of their travel.

Yes, perhaps there are a few eccentrics who would genuinely prefer, as a matter of course, to spend an hour making a trip by bus or train than fifteen minutes making the same trip by robocab, even with all the waiting involved in using transit, the exposure to bad weather and extreme temperatures while walking to stations and waiting for transit vehicles to arrive, the uncomfortable seats, the lack of privacy, the potential for noisy and disruptive behavior from other passengers, and all the other common annoyances of using public transit. Perhaps there are some people who would genuinely prefer it despite all that. But I rather doubt there are enough of them to matter.

As I've said before, holding companies can hold a pool of vehicles of different sizes, and transit agencies can rent from them.

I don't understand how you think this would solve the economic problem. You'd still need two sets of vehicles, a set of large vehicles that are only used a few hours a day when demand is high, and a set of small vehicles for off-peak use. The holding company is going to have to charge higher rents to cover the costs of purchasing, storing and maintaining two sets of vehicles instead of one.

Ahh, but if that logic were true, Amtrak would've vanished long ago, and so would the Vegas monorail and some other services.

Amtrak has withered away to virtually nothing, because it cannot compete with cars and planes. Americans travel about 100 miles domestically by plane and almost 1,000 miles by car for every 1 mile they travel by Amtrak. I doubt self-driving cars will make mass transit COMPLETELY extinct. Some cities will likely maintain a few heritage systems for tourists and local color, like the cable cars in San Francisco. But as practical, everyday transportation, mass transit doesn't have a future.

The Vegas Monorail seems to be a huge boondoggle. $5 to travel even just one stop. And a long walk from the casinos, especially the ones on the other side of the strip. The couple of times I rode it it was virtually empty. It doesn't seem to have made any difference at all to congestion on the strip.

by Bertie on Dec 10, 2011 7:29 pm • linkreport

Look, in the end I don't really care if something comes along that is so awesome that transit dies. It turns out that more than I like transit, I like better even better. So if it can achieve all the goals of transit for less cost or achieve them even better, great. Bring it on.

If you can call it a death rather than an improvement. Methinks transit will just move onto newer goals. It has happened before, too. When airliners outcompeted ocean liners, they became cruise ships, and one became a hotel and museum. And that's one example.

by Rickyrab on Dec 10, 2011 7:33 pm • linkreport

Those examples are not things that you need to destroy transit and cities.

I didn't say they were. I was responding to your false and absurd claims that "there are no cars in the world that can do any of the things [Brad Templeton is] talking about" and "the robocars we have now do not do the things you're talking about." Not only do you fail to acknowledge your errors when they are pointed out to you, but you pretend you said something different than what you actually said.

You haven't explained it so much as stated it.

The statement is the explanation. Do you dispute it? If so, on what basis?

But they have never been driven on city streets without a driver who can and does take over time to time.

Yes, they have. You don't know what you're talking about. In almost 200,000 miles, the human occupant has intervened only a handful of times, and none those interventions was necessary to prevent an accident.

But it won't kill cities. That's ridiculous.

No one said robocars will "kill cities," so this is yet another of your strawman responses. Robocars will change cities. In particular, they will destroy the market for all or almost all mass transit, and they will reduce the already small incentives to build things close together.

by Bertie on Dec 10, 2011 7:52 pm • linkreport

Amtrak has withered away to virtually nothing,

LOL

http://www.amtrak.com

Americans travel about 100 miles domestically by plane and almost 1,000 miles by car for every 1 mile they travel by Amtrak.

But Amtrak has not withered away to nothing. It has had a growing ridership over the years, too.

I don't understand how you think this would solve the economic problem. You'd still need two sets of vehicles, a set of large vehicles that are only used a few hours a day when demand is high, and a set of small vehicles for off-peak use. The holding company is going to have to charge higher rents to cover the costs of purchasing, storing and maintaining two sets of vehicles instead of one.

And what about all of those robocars that will sit idle for all except a few hours of the day? Where will they get their riders from? The need for capacity varies in a day. The need for robocars fluctuates, like the need for transit vehicles. In the end, there will be a need for surplus vehicles, no matter what happens, and so the opportunity cost for maintaining two sets of vehicles might not be all that great as compared to the current (supposed) single set (if it can be called a single set). When vehicles sit, they don't guzzle electricity and they aren't wearing out as much as when they are moving. That cuts maintenance costs. So the higher rents will not really be that much higher. Economies of scale will help, too.

Yes, perhaps there are a few eccentrics who would genuinely prefer, as a matter of course, to spend an hour making a trip by bus or train than fifteen minutes making the same trip by robocab, even with all the waiting involved in using transit, the exposure to bad weather and extreme temperatures while walking to stations and waiting for transit vehicles to arrive, the uncomfortable seats, the lack of privacy, the potential for noisy and disruptive behavior from other passengers, and all the other common annoyances of using public transit. Perhaps there are some people who would genuinely prefer it despite all that. But I rather doubt there are enough of them to matter.

LOL negative advertising. Besides, some of those "noisy and disruptive behaviors" are fun to watch. And people like to gawk at other people.

even with all the waiting involved in using transit, This is really the main problem with transit, along with walking through the weather (and some people like the exercise anyhow). This is where robocabs can cut into transit (or even help transit by augmenting it during off hours). The rest can be designed around (seats can be made more comfortable and stations can have roofs over people's heads, and robocabs can even be used as transit waiting areas).

Yes, perhaps there are a few eccentrics

Yes, there are, and some own private railcars and buses. And other groups are not eccentric but communal, and those groups have their own transit vehicles and services.

Furthermore, there are high density areas that are simply not going to go away, and the marginal cost of transit is not that much higher than that of robocab or ordinary cab usage.

by Rickyrab on Dec 10, 2011 7:55 pm • linkreport

marginal cost of transit

I meant to say "opportunity cost of transit".

by Rickyrab on Dec 10, 2011 7:58 pm • linkreport

I also meant to say "there are places where the opportunity cost of transit is not that much higher than that of robocab or ordinary cab usage". Dense urban areas, population centers, job centers, anywhere people go to and crowd upon.

by Rickyrab on Dec 10, 2011 8:06 pm • linkreport

The Vegas Monorail seems to be a huge boondoggle. $5 to travel even just one stop. And a long walk from the casinos, especially the ones on the other side of the strip. The couple of times I rode it it was virtually empty. It doesn't seem to have made any difference at all to congestion on the strip.

That didn't keep people from building it - or building it in the wrong place, for that matter. Had they done it properly, it should have moseyed up to casino entrances, which it doesn't do. No wonder it's in deep doo doo. But folks built it anyway. The Deuce (a bus system) does better.

by Rickyrab on Dec 11, 2011 4:01 pm • linkreport

I was responding to your false and absurd claims that "there are no cars in the world that can do any of the things [Brad Templeton is] talking about" and "the robocars we have now do not do the things you're talking about."

It's not absurd of false, because it's true.

Existing prototypes do not do any of these things, each of which is a direct quote:

* make car travel cheaper
* make car travel more widely available than it is today
* provide on-demand, door-to-door rides in a self-driving taxi that is cheaper than current taxis
* automated cleanig of taxis
* Cars that reduce congestion

BTW, 200,000 miles (of closely monitored driving with a driver who takes over a couple of times a trip) is far too small a sample to say it's even safer than a driver. The biggest advantage is probably that the car never speeds.

Do you dispute it? If so, on what basis?

Yes. I do. Because I think that labor makes up about the same percentage of the cost of a taxi as it does for a bus, if not less. You can prove me wrong by answering the questions you keep dodging.

Two questions you repeatedly refuse to answer.

1. What percentage of taxi fares go to pay for the driver?
2. What percentage of transit fares go to pay for the operators? Include all the costs associated with the issues Rickyrab mentioned.

Yes, they have. You don't know what you're talking about.

No they haven't. You're lying. These cars always have a driver. That driver frequently takes over. That's not true driverless.

No one said robocars will "kill cities,"

kill cities = promote car-oriented lifestyles and car-oriented urban development.

In part because there is no such thing as "car-oriented urban development."

Also, if it reduces "incentives to build things close together" how is that diferent than "killing cities?" Cities are defined by density.

by David C on Dec 11, 2011 4:23 pm • linkreport

Existing prototypes do not do any of these things,

You're trying to move the goalposts yet again. You claimed "there are no cars in the world that can do ANY [emphasis added] of the things [Brad Templeton is] talking about." I just gave you a list of some of the things that Brad Templeton talks about that cars can already do. Not merely prototypes, but actual production vehicles already on the road. Your claim is flatly false.

[There are no] cars that reduce congestion

Wrong yet again. I just gave you several examples of the automation technologies already available in production cars that reduce congestion. Adaptive cruise control reduces congestion by reducing spacing between cars. Automated and smart parking systems reduce congestion caused by vehicles driving around looking for a place to park. GPS navigation with real-time traffic data reduces congestion by rerouting vehicles around congested areas. And every automation technology that reduces the risk of accidents reduces congestion. Accidents are a major cause of congestion. Again, your claim here is flatly incorrect.

Yes. I do. Because I think that labor makes up about the same percentage of the cost of a taxi as it does for a bus, if not less.

You have offered absolutely no evidence to support this wildly implausible claim. Please explain why you think it's plausible that eliminating drivers from vehicles that carry ten passengers on average would reduce total costs per passenger-mile by the same percentage as eliminating drivers from vehicles that carry less than two passengers on average. For rail transit, your claim is even more implausible, since the ratio of passenger-miles to drivers is even higher for rail than it is for buses. Light rail trains average dozens of passengers, and subway and commuter rail trains may average more than a hundred passengers.

But your claim here would still be irrelevant to the point even if it were true. Taxis cost more per passenger-mile than buses. Therefore an equal reduction in the percentage cost means a greater reduction in the dollar cost for taxis than for buses. Eliminating drivers would reduce the cost of taxi rides more than it would reduce the cost of bus rides, even if the percentage reduction was the same.

Also, if it reduces "incentives to build things close together" how is that diferent than "killing cities?" Cities are defined by density.

More nonsense. Cities are defined by political boundaries, not density. There is enormous variation in the density of cities, and cities in general have become much less dense over the past 50 years, as I already showed you. Self-driving cars will make them even less dense.

by Bertie on Dec 11, 2011 5:26 pm • linkreport

In case no one has yet mention this- self driving automobiles could "de-frag" parking spaces, which will be increasingly relevant with the replacement of marked spaces with individual meters, with muni meters.

by Douglas Willinger on Dec 11, 2011 5:59 pm • linkreport

In case no one has yet mention this- self driving automobiles could "de-frag" parking spaces, which will be increasingly relevant with the replacement of marked spaces with individual meters, with muni meters.

Yes indeed. Every driver knows the frustration of trying to find street parking and seeing lots of wasted space in the form of gaps between parked cars that are much bigger than they need to be, but not big enough to accommodate an additional vehicle. Automated parking will greatly reduce this wasted space and greatly increase the capacity of street parking. It will also allow much higher parking densities in parking lots. Automated vehicles will be able to park at valet or supervalet densities. They will also relieve drivers from the chore of parking, like human valet services do today. Your car will simply drop you off at your destination and then go park itself. As Brad Templeton says, automated valet parking is likely to appear in production vehicles before full-fledged automation, and will help familiarize people with the technology and show them just how beneficial it will be. Semi-autonomous parallel parking systems are already available. Like most new auto technologies, they first appeared on expensive, high-end vehicles, but are now trickling down to mid-range vehicles. Ford's automated parallel parking system, for example, is now available on even low-priced models.

by Bertie on Dec 11, 2011 6:26 pm • linkreport

I just gave you a list of some of the things that Brad Templeton talks about that cars can already do

Nonsense. He doesn't talk about the things you listed and you didn't list the things he talked about.

I just gave you several examples of the automation technologies already available in production cars that reduce congestion.

Can you show any evidence to back up the claim that these technologies reduce congestion? Or do they just do it because you will it be so.

You have offered absolutely no evidence to support this wildly implausible claim.

Right back at you.

Please explain why you think it's plausible that eliminating drivers from vehicles that carry ten passengers on average would reduce total costs per passenger-mile by the same percentage as eliminating drivers from vehicles that carry less than two passengers on average.

I think it's possible because I'm not an idiot and you haven't proven your case.

Bus and train drivers are skilled, unionized workers. They get paid more per hour. They get paid time and half for OT. They get paid vacations. They get pensions. They get mandatory paid breaks. Etc... Taxi drivers are often independent contractors, in general, do not get these things. It's harder to get bus driver and train driver jobs, which is why immigrants often drive taxis, but I never see them driving buses.

So the reason is that bus and taxi drivers are paid much more than taxi drivers. In fact it may be so much more that getting rid of them will save more money then getting rid of taxi drivers will to taxi users.

Now, can you answer the question you are obviously terrified of answering.

1. What percentage of taxi fares go to pay for the driver?
2. What percentage of transit fares go to pay for the operators? Include all the costs associated with the issues Rickyrab mentioned.

If it is so clearly true, it should be easy to prove. And if it's so easy to prove then what are you scared of?

Cities are defined by political boundaries, not density.

You're half wrong. One definition is "An incorporated municipal center". But another is "a large and densely populated urban area." Here I was using the second definiton.

Way to argue on the least important point btw, and one you were - again - wrong on.

Furthermore, this is what we were talking about. But you keep getting the terms mixed up.

For example, you just wrote that "cities in general have become much less dense over the past 50 years, as I already showed you." but actually NYC has become much more dense and so have many other cities - as I showed already showed you. But then you said "No Manhattan has gotten lese dense, I was only talking about the dense part". So why don't you get your terms straight.

[BTW since 2000, Manhattan's population has actually risen, so even your claim that Manhattan's population is dropping isn't correct]

by David C on Dec 11, 2011 6:42 pm • linkreport

He doesn't talk about the things you listed and you didn't list the things he talked about.

Wrong yet again. He talks about every one of the things I listed.

Can you show any evidence to back up the claim that these technologies reduce congestion?

I already described the evidence. Reducing distance between vehicles reduces congestion by increasing road capacity. Reducing accident rates reduces congestion by reducing the probability that lanes will be blocked or closed. Rerouting vehicles away from congested areas reduces congestion by reducing the number of vehicles in the congested area. Smart parking systems reduce congestion by reducing the amount of time drivers spend driving around looking for a place to park. What part of this don't you understand?

Bus and train drivers are skilled, unionized workers. They get paid more per hour. They get paid time and half for OT. They get paid vacations. They get pensions. They get mandatory paid breaks. Etc... Taxi drivers are often independent contractors, in general, do not get these things. It's harder to get bus driver and train driver jobs, which is why immigrants often drive taxis, but I never see them driving buses.

You have produced no evidence whatsoever regarding wages and benefits for transit drivers vs. taxi drivers. The ratio of passengers to drivers is on the order of 5-10 times higher for buses than it is for taxis, and 20-100 times higher for trains that it is for taxis. You're seriously suggesting that bus drivers make 5-10 times as much per passenger-mile as taxi drivers and that train drivers make 20-100 times as much per passenger-mile as taxi drivers, are you? Your "possibility" here isn't simply implausible. It is preposterous.

Now, can you answer the question you are obviously terrified of answering.

No, I don't know the answer to those questions. The questions are irrelevant. Your hypothesis that eliminating drivers would produce a comparable reduction in costs per passenger-mile for transit and taxis is absurd, for the reason I described above. No remotely plausible difference in driver labor costs between transit and taxis could even come close to compensating for the fact that transit vehicles carry much higher numbers of passengers than taxis.

But another is "a large and densely populated urban area." Here I was using the second definiton. ... actually NYC has become much more dense and so have many other cities

Now you're contradicting yourself. Since you claim you're using the word "city" to refer to an urban area defined by size and population density rather than political boundaries, you can't appeal to a political definition to claim that a city has become denser. What minimum size and population density must an urban area be for it to qualify as a "city" as you are using the word?

Cars have already drastically reduced the average density of cities as defined by the Census Brueau. Self-driving cars will continue that process. So the amount of urban land that meets your definition of "city" will continue to shrink. And, by your definition, any urban areas that fall below your minimum qualifying density for a "city" will be -- to use your word -- KILLED.

by Bertie on Dec 11, 2011 9:09 pm • linkreport

I already described the evidence.... What part of this don't you understand?

First of all, you don't "describe" evidence. You show it. You're telling me how these things should reduce congestion in theory. I'm asking you if it actually works in practice. Just become some drivers are following the car in front of them more closely than they did before, that may not actually reduce congestion. I was unaware that leaving too much space was a cause of congestion. That certainly isn't a problem for DC drivers. I also thought automated parking made parallel parking easier, I didn't realize it somehow created new parking spaces that stopped drivers from driving around.

Perhaps you have some evidence (please look this word up if you don't know what it means - it's the only way to learn) to back up these claims. I eager to learn.

You have produced no evidence whatsoever regarding wages and benefits for transit drivers vs. taxi drivers.

No, I don't know the answer to those questions.

Uhh...neither have you. So I see how this works. I have to produce facts, but you can just make assertions about things you absolutely do not know. Is that correct?

But yes, I assert that bus drivers make much much more than taxi drivers. They earn about ~$50,000 base salary in DC. And many earn overtime. It is not uncommon for some to make $100,000 in a year. Plus they earn a 401K, pensions, vacation, health benefits, etc... All of which can make their cost closer to $80,000 a year. Which is about $40 an hour.

Taxi drivers earn about $11 an hour. No benefits.

But there is significant overhead for bus drivers and train drivers as well, so that may not even tell the whole story.

Do you have any actual facts and figures about the relative costs of robotaxis compared to robotrains?

Because looking a the list of jobs, I see a bunch that robots can do. Why not have them do inspections and repairs? They could even do repairs in between trains and communicate with the trains so as to move out of the way when the next one comes.

And you need a lot fewer managers if you don't drivers. You don't need safety inspectors if trains are perfectly safe.

And while we're at it you can build the trains with lighter, cheaper materials because they won't ever crash. This will also reduce energy use. You can run them on tighter headways and stop perfectly every time. Have robots repair the trains. Of course the trains will need a lot fewer repairs since they'll always be operated perfectly.

There's a lot of savings in there. Surely you've calculated the relative future prices. What are they I wonder?

And it doesn't stop there. Want to build more tunnels? No problem tunnel boring robots will do that for you.

The ratio of passengers to drivers is on the order of 5-10 times higher for buses than it is for taxis, and 20-100 times higher for trains that it is for taxis.

You have produced no evidence whatsoever regarding the ratio of taxi passengers per driver to the ratio of train passneger per driver.

Since you claim you're using the word "city" to refer to an urban area defined by size and population density rather than political boundaries, you can't appeal to a political definition to claim that a city has become denser.

I can when I talk about a specific place - i.e. NYC as opposed to general one ie a city.

Cars have already drastically reduced the average density of cities as defined by the Census Brueau.

You can't possibly prove that. We have cars - true. And center cities are less dense - true. But you can't prove that one caused the other. There are just too many variables at play here.

But when have you been in the business of proving things?

by David C on Dec 11, 2011 10:58 pm • linkreport

And adding more to this.

I don't think the $40 a hour captures pension cost.

As anti-transit folks like to point out, transit is subsidized - sometimes as much as 90%. So in those cases a 10% reduction in costs will make transit free. You have to give excellent service to beat free.

And transit service could be improved. Instead of running 1 8-car train every 8 minutes, you could run 1 1-car train every 1 minute because the trains drive themselves meaning that there is no larger labor cost. So that means less waiting. In fact you could shorten headways more and get down to 1 1-car train every 45 seconds or something.

So, there is quite a bit of savings to be found in transit using the same technology that could keep transit very competitive. And since 30-40% of Americans want to live in an urban area, that means it will always have a sizable constituency.

The future looks very bright for dense, urban transit-enabled living. Very bright in deed.

by David C on Dec 12, 2011 9:22 am • linkreport

Automated transit would allow transit service to be even more convenient, as David C points out. If operation is automated you can run trains every five minutes even off-peak, like they do with the Vancouver SkyTrain. That makes people even more likely to rely on transit for their daily travel since it's convenient all the time.

Driverless tech is way closer to implementation on fixed-guideway transit than on the car you drive around on city streets. It already exists on some transit systems.

Labor costs are about 2/3 of public transportation operating costs.

Fuel is 20+% of the cost of driving and and only 8% of transit operating costs (and less for electric operation) - transit agencies will not be hurt as badly by rising fuel prices.

by MLD on Dec 12, 2011 9:59 am • linkreport

One thing that I don't think has been mentioned is Congestion charges. They can be applied in a market fashion. After all street space is a very finite commodity.

It's unpopular and hard to implement right now but will be much easier if every vehicle is networked. The revenue could be used for transit and other de-congesting services making them even more accessible than they are now, maybe even fare free.

by Frank on Dec 12, 2011 4:23 pm • linkreport

Frank, you don't get it; these robocars exhaust magic fairy dust from their tailpipes that makes congestion disappear. Why would we want or need congestion charges then?

by MLD on Dec 12, 2011 4:27 pm • linkreport

You're telling me how these things should reduce congestion in theory. I'm asking you if it actually works in practice.

Yes, it actually works in practise. Reducing the distance between vehicles ACTUALLY DOES allow more vehicles per lane-mile of road. Reducing the rate of accidents ACTUALLY DOES reduce the incidence of blocked lanes of traffic. Knowing where parking is available ACTUALLY DOES reduce the need to drive around looking for parking.

But yes, I assert that bus drivers make much much more than taxi drivers.

You're not listening. Since the ratio of passengers to drivers is around 5-10 times higher for buses than for taxis, and around 20-100 times higher for trains than taxis (and even higher in, for example, New York), higher wages for bus and train drivers could not possibly offset more than a small fraction of the huge economic advantage taxis will acquire over buses and trains from the elimination of driver costs.

You have produced no evidence whatsoever regarding the ratio of taxi passengers per driver to the ratio of train passneger per driver.

Average number passengers per NYC taxi: ~1.4
Average number of passengers per NYCT transit bus: ~18
Average number of passengers per NYCT subway train: ~250

So, compared to taxis, there are about 13 times as many passengers per driver on a bus, and about 178 times as many passengers per driver on a subway train. You're seriously claiming that NYC bus drivers make 13 times as much as NYC taxi drivers, and that NYC subway drivers make 178 times as much as NYC taxi drivers, are you? As I said, your claim is not merely implausible. It's preposterous.

I can when I talk about a specific place - i.e. NYC as opposed to general one ie a city.

No you can't. You just claimed that by "city" you mean "a large and densely populated urban area," not an area defined by a political boundary. Therefore, you can make no claims about density changes in the "city" of New York as defined by a political boundary. You still haven't told us what minimum size and density an urban area must have in order to qualify as a "city" as you are using the word.

by Bertie on Dec 12, 2011 7:52 pm • linkreport

As anti-transit folks like to point out, transit is subsidized - sometimes as much as 90%. So in those cases a 10% reduction in costs will make transit free. You have to give excellent service to beat free.

No, a 10% reduction in the cost of transit would make the cost of transit "only" about 90 cents per passenger-mile instead of a dollar. Eliminating taxi drivers will reduce the cost of taxi rides to the cost of buying and running the taxi vehicles. The average cost to buy and run a private motor vehicle is about 30 cents per passenger-mile. This should be considered an upper limit on the plausible cost of a self-driving taxi. 30 cents is much less than 90 cents. The government will be able to substitute free self-driving taxi rides for every current transit user and still save money compared to what it currently spends on transit subsidies.

And transit service could be improved. Instead of running 1 8-car train every 8 minutes, you could run 1 1-car train every 1 minute because the trains drive themselves meaning that there is no larger labor cost. So that means less waiting. In fact you could shorten headways more and get down to 1 1-car train every 45 seconds or something.

Which would shorten the average train trip time by.....a whopping 4 minutes. Assuming your proposal is even practical. And you seriously think that would stop people from switching to a faster, more comfortable, more convenient, more private, on-demand door-to-door taxi ride, do you? Marginal improvements in the experience of using mass transit aren't going to save it. Nothing is going to save it.

by Bertie on Dec 12, 2011 8:38 pm • linkreport

Labor costs are about 2/3 of public transportation operating costs.

Which might be remotely relevant if the only labor involved in transit operation was drivers (no sales staff, no security staff, no mechanics, no cleaners, no engineers, etc.) and if buses, trains, stations, tunnels, railtrack, etc. (i.e., transit vehicles and infrastructure) magically appeared for free out of thin air instead of having to be paid for.

Fuel is 20+% of the cost of driving and and only 8% of transit operating costs (and less for electric operation) - transit agencies will not be hurt as badly by rising fuel prices.

The average automobile today only gets about 20 mpg. The Toyota Prius gets 50 mpg. The Nissan Leaf gets 99 mpg-equivalent. There is enormous potential to increase the efficiency of cars, and to diversify automobile fuels.

by Bertie on Dec 12, 2011 9:10 pm • linkreport

Because looking a the list of jobs, I see a bunch that robots can do. Why not have them do inspections and repairs?

Because they would be hugely expensive to develop, especially for a specialized market like mass transit vehicles and infrastructure. Robot mechanics would make more sense for cars, because the costs would be spread over a much larger market. But cars have already advanced to the point where maintenance costs are low anyway. They go 100,000 miles without needing much beyond oil and filter changes and a new set of tires. Future cars, especially electric ones, will be even cheaper to maintain. Transit infrastructure, by contrast, is crumbling. The Federal Transit Administration says there's already a $78 billion backlog of transit maintenance and repairs. Most of this is for rail transit.

And while we're at it you can build the trains with lighter, cheaper materials because they won't ever crash. This will also reduce energy use.

The replacement cycle for rail transit rolling stock is on the order of 20-30 years. Transit agencies can't afford to replace trains any more often than that because they're so expensive. Cars are going to benefit significantly from new, lighter, cheaper materials long before trains could.

Want to build more tunnels? No problem tunnel boring robots will do that for you.

Subway construction already uses automated tunnelling machines. It's still enormously expensive. The new Second Avenue Subway in New York is projected to cost $2 BILLION per mile.

by Bertie on Dec 13, 2011 12:09 am • linkreport

Bertie, I ask for evidence and you just give me more uncited claims. It's sad that you need to win so badly that you're willing to make things up. Can you PROVE that existing technology is reducing congestion. PROVE it. Don't just say it. Prove it. Evidence. Proof. Not just "Bertie says so." Facts. Do you have any?

by David C on Dec 13, 2011 12:51 am • linkreport

The self driving cars that Google are currently developing are designed to run in existing traffic, so it won't be possible for cars to run closer together until every single one is automated. Can you imagine trying to be a manual driver when all the other cars are three inches from your bumper.

These Google cars are not networked they are just object avoiding robots that are also linked into GPS. So in the first instance these cars will not solve any congestion because they will driving the safe distance between vehicles.

Now plenty of people have postulated that such cars will be able to platoon themselves on the freeway, saving space and drag. That requires all vehicles to be networked as well as each car needs to know when the others around it need to move. Also such platooning will never happen on city streets, despite their best efforts to banish them there will be pedestrians and you will need to leave space. Currently the law does not seem to care that much how many pedestrians and cyclists that are mown down, but once large corporations are responsible, well I assure the lawyers certainly will.

Also once the number of people that are killed on roads plummets each remaining death will be more and more tragic and the pressure to protect people will be even higher.

The main things is the automated car will not be cheaper to run than the current car, that is the most important metric to consider for the American city. The reality almost no one takes transit in America today. What will shape cities is how cheap it is to run cars. Automation does not effect that equation, it's either gas or electric or some green fuel.

If the price of travel increases then people will have to travel less, and the only way to maintain the number of trips they take and for it to remain affordable, is for them to live in a denser area, where facilities are closer together.

In a high cost of travel scenario, we won't have to worry about congestion on the roads, as half the people won't be able to afford to travel as much. There will all these six lane roads with one bus lane stuffed to the gills with bus passengers while the top 30% whizz past in their BMW X class hybrid in the other empty two lanes.

Meanwhile every congressman is screaming for a subway line out to their suburb so their constituents don't have to endure bus commutes. But even in a scenario where there are plenty of funds, as a whole for new transit lines, with every city wanting some it will take decades for a comprehensive network to be built. Meanwhile the wealthy move inwards and next to existing transit, while the poor have to endure multiple bus connections.

All automated travel does is get rid of the low paid from the taxi industry. If taxi fares fall by half we will be doing well. That is not going to change transit very much, it just makes living in a city easier for those who want to do without a car.

Where there is no pressure to densify land use then developers could be lazy and continue to build every thing the same, except buildings will have a automatic valet drop off zone. Some people might design buildings closer to the street with parking shoved behind, but others won't bother.

It's only where fuel and land costs are high will you see buildings near the street to be near transit, with parking in multi storey structures around the back. Suburban clusters will grow denser as the surface parking is recycled.

by Rational Plan on Dec 13, 2011 5:52 am • linkreport

We already have a transportation option that has most of the characteristics of "robot cars" as regards an urban environment (e.g. cheap, convenient, no parking issues, there when you need them, gone when you don't, etc...): they're called "bikeshare". Add a pair of rainpants and a jacket and "Hello World of the Future!"

:)

by oboe on Dec 13, 2011 10:50 am • linkreport

How will they interact with bicycles and pedestrians?

by AR on Dec 13, 2011 1:14 pm • linkreport

@MLD I know you're comment was tongue in cheek but I agree with Bertie that they will reduce congestion to a large extent, but not eliminate it since many more will be on the road.

We will eventually get to the point of exclusive urban autocars and if they're all networked the increased efficiency could be very significant. And if street parking disappears, many streets will have two extra lanes allowing higher volume, or if we get the politics right, more bike lanes.

To echo Rational Plan, this will not kill the city or the new urbanism model. The biggest villain of new urbanism is the surface parking lot. This will be gone. Everywhere. There is no need to park within walking distance if the car can drop you off and either go park a mile away, or go pick someone else up.

I believe that in time the vast majority will opt out of car ownership. The main incentive to own a car is convenience, not cleanliness. It's the convenience of having it there and ready to leave whenever you want it. With auto-taxis this incentive will no longer be tied to ownership. An auto-taxi can be parked at every urban street corner ready to go at 3am if needed.

by Frank on Dec 13, 2011 2:57 pm • linkreport

The biggest villain of new urbanism is the surface parking lot. This will be gone. Everywhere. There is no need to park within walking distance if the car can drop you off and either go park a mile away, or go pick someone else up.

The biggest villain of new urbanism is cars, not parking lots. Parking lots are just the symptom. The faster people can travel, the lower the need to build things close together. That's why, as cars have become increasingly common over the past 60 years, cities have become progressively less dense, and why the vast majority of new urban development is low-density.

I doubt that robocars will eliminate dense development entirely. It'll just become even less common than it is now. Cities will probably retain a few high-density clusters that serve as arts / entertainment / cultural / dining / high-end retail districts. Think of them as large open-air malls. The "downtowns" of many cities are already evolving into this function. A few people will also live in these districts, in small, expensive condos and apartments.

by Bertie on Dec 13, 2011 11:31 pm • linkreport

The biggest villain of new urbanism is cars, not parking lots. Parking lots are just the symptom.

Saw this coming. Potato potato. Parking lots are a serious "symptom" that hurts walkability which can and will disappear. Cars never will, and planners know this. In transit oriented development the car is still a major consideration, otherwise there would be no streets.

I agree that robocars will make dense development less common, but not to a huge extent. Cars have existed for almost 100 years now. Nothing changed in car availability or technology to fuel the urban renewal we're in now. It's just a re-balance. Some people want privacy and solitude, and others are willing to give that up for better access to arts, goods, and people. Robocars won't change this.

by Frank on Dec 14, 2011 11:43 am • linkreport

One more thing,

The faster people can travel, the lower the need to build things close together.

This may be a bit too simplistic. People want choices and variety. That's why they live in cities. That's why offices reside in cities. They may not be able to travel as fast, but they have access to much more in the same travel time.

by Frank on Dec 14, 2011 11:57 am • linkreport

@Frank: Nothing changed in car availability or technology to fuel the urban renewal we're in now.

To the contrary: the development of emission controls did a lot to improve the urban environment. I submit this is partly the reason for current urban revival. During the 60s, most people wanted to escape to the suburbs because the smog in the city was so bad. Get next to a well-maintained car from that era and smell the fumes, if you don't believe me.

by goldfish on Dec 14, 2011 12:18 pm • linkreport

the reality is that it is not cars that are bad, but the single-occupancy driver paradigm that is so damaging to our environment, urban fabric and quality of life.

I don't see how self-driving cars change this. But that's OK - where we win is that instead of having millions of 4-6 passenger cars carrying just one person, we'll instead have millions of 1-passenger cars carrying just one person. These cars will take up less space on our roads, so we will be able to go a long time without needing more roads. And they'll burn a lot less fuel simply by virtue of being much smaller and lighter than today's cars.

Why would this be?

It's an evolution.

First, once self-driving cars are the overwhelming norm, if accident rates drop as predicted, the perceived need to have a big car for safety purposes will go away.

Second, Zipcar-style car-sharing will become more extensive. As things stand now, a Zipcar needs to have a sufficient number of customers within walking distance of where it's usually parked to make it profitable. But if the Zipcar shows up at your door at the time you've reserved it for, it can be parked a few miles away, and you can still be part of its customer base.

Third, the overwhelming majority of car trips involve just one person. As Zipcar use expands, they'll be able to create their own market for cars, and automakers will build lines of cars to meet their specs. If they're mostly selling rides for one person at a time, they'll want to order and buy the most inexpensive car that can provide that service. Hence small, light, fuel-efficient as possible: one-person cars.

Zipcar and its competitors would still have multi-passenger cars, but the economic logic would be that most of their fleets would consist of single-passenger cars.

Fourth, once they're generally available, people who continue to own their own cars (a gradually diminishing slice of the population) will first buy them as their second or third cars, then find that they're the cars they use most, because most of the time they're the only person on a given trip.

Once people realize this, the automobile market overwhelmingly becomes a one-person car market, and multi-passenger vehicles become a specialty market. That market won't ever go away, because people will want the companionability of sharing a car when they and their family/friends are all going to the same place at the same time.

But the ratio of single-occupancy cars to multi-passenger cars will likely converge to the ratio of single-person car trips to multi-person car trips, because you don't need to keep all that capacity in your driveway just for the infrequent occasions when you need it: when you want to share a ride with someone, you'll reserve a Zipcar.

by low-tech cyclist on Dec 14, 2011 12:50 pm • linkreport

@goldfish
Certainly pollution had a part in the suburban flight, and the de-industrialization of cities cleaned up the air just as much as emissions control making them more desirable to move back into, but these are minor factors compared to the lifestyle incentives. And for the sake of this discussion, pollution will not play a huge role in the future of autocar networks, assuming technology continues in it's current trajectory.

@low-tech cyclist
I like your prediction of the evolution. One could even see a system arising of road hierarchy.
Neighborhood Roads: used only for light single/double occupancy vehicles like the MIT City Car; maintain slow speeds; easily mix with pedestrian and cyclists.
Boulevards: for both light and heavy such as freight and busses; limited ped/cycle interaction.
Grade Separated Highways: only heavy vehicles; perhaps even rail can be included at the center of highways merging in at grade - vehicles on the highways can "part the red sea" as it's merging.

It all depends on how much politics play a role in the planning. With so many developed countries each with different levels of government interaction, it will be interesting to compare systems when the time comes. The US will probably not be the first to do this. And even if we are, government interaction will be minimal so the evolution will take longer.

by Frank on Dec 14, 2011 2:16 pm • linkreport

Makes me think of "better living through chemistry" stuff from the '50s, like this:

by Geoffrey Hatchard on Dec 15, 2011 10:31 am • linkreport

In a world of self-driving cars, the Zipcar model becomes absorbed into the taxi model, or simply disappears. You don't need a Zipcar strategically parked within walking distance of your home, because the car can come to you. There'll be no distinction between a Zipcar rental fee and a taxi fare.

Eventually, most robocars will probably be single-passenger or two-passenger, but larger vehicles will still be popular. The need and desire to travel together in the same vehicle with family members or friends isn't going to disappear. What is going to disappear, or at least become very rare, is urban travel with strangers.

I doubt that most people who can afford their own car (i.e. most people) will give up the benefits of private ownership. The primary market for robocabs will be people who would otherwise be dependent on slower modes of transportation -- walking, bicycling and mass transit.

by Bertie on Dec 15, 2011 11:09 pm • linkreport

I agree that robocars will make dense development less common, but not to a huge extent.

Well, dense development isn't very common even now. They won't have to reduce it much to make it pretty rare.

Cars have existed for almost 100 years now. Nothing changed in car availability or technology to fuel the urban renewal we're in now.

Yes, "urban renewal" doesn't have much to do with cars. It's mainly just gentrification of certain inner-city neighborhoods, often fueled by lavish government spending. As I said, I doubt this will go away. There'll still be some yuppie urban playgrounds. But middle- and lower-income people will likely continue to flee from cities to suburbs.

by Bertie on Dec 15, 2011 11:34 pm • linkreport

Driverless cars won't be allowed on the roads until they can safely deal with pedestrians and cyclists, and by that I mean 2-year olds running out into the road from between parked cars, and cyclists taking the lane while travelling 15km/h in a 60km/h zone.

So I see them as being a great boon for urbanism. Without human drivers to worry about, people will feel comfortable cycling down the middle of the road at a relaxed speed, and walking across the street whenever and however they feel like. The advent of driverless cars will therefore drastically slow vehicle speeds, calming the road environment and making it pleasant for all. Long-distance suburban travel patterns will become unviable. Cycling will boom.

The streets will become a little like they were 100 years ago - filled with a very diverse range of slow-moving vehicles and pedestrians.

Bring it on.

by Colin on Dec 16, 2011 4:37 am • linkreport

@Colin
While I hope that many roads will allow bikes and peds along with slow moving cars, a case can be made that self-driving cars will monopolize roads even more than cars do now since they could serve as a digital witness to any alterations. They'll obviously contain elaborate sensing mechanisms like video cameras or lidar, allowing them to assemble a "black box" of detailed information leading up to any incident. If this can be examined by law enforcement, it will serve to discourage people from taking advantage of a self-driving car's conservative avoidance algorithms.

Hopefully this will provide a much more ordered coexistence but the nature of it will depending on the speed of the road. Faster roads will need to have greater legal restrictions on peds and bikes creating obstacles to non-motorized connectivity. This will likely be more widespread in suburban environments where pedestrian flow isn't as high a priority.

by Frank on Dec 16, 2011 3:19 pm • linkreport

How different is this "self-driving" car from a PRT system; especially the proposed system for Masdar City?

by Ryan on Dec 17, 2011 9:51 pm • linkreport

How different is this "self-driving" car from a PRT system; especially the proposed system for Masdar City?

Radically different. The self-driving cars under development by Google and the automakers don't need special tracks or guideways or sensors built into the road. They're not limited to a fixed set of stations or destinations. They're not controlled by a central computer. They will operate just like regular cars, except that they'll be able to drive themselves, much better than any human driver.

PRT that requires new road/track/guideway infrastructure is going to be far too expensive to be more than a niche product.

by Bertie on Dec 18, 2011 4:29 pm • linkreport

From a http://www.wired.com/magazine/2012/01/ff_autonomouscars/4/">Wired article:

"For as long as anyone, even Google, is willing to predict, cars will by necessity be semiautonomous; human drivers will still have to play some role."

'When a car ahead slows to turn, for example, the Mercedes fails to recognize that the vehicle will soon be out of my way, so we brake too much for my taste and then accelerate from a dead stop. The car’s lane-departure warning feature, which alerts drivers when they drift out of their lane, doesn’t work if the lane and edge markings are worn away—a common phenomenon in our infrastructure-challenged country. Then there are the technical issues that still plague sensors. Ice bedevils radar; snow challenges cameras. If, while cresting a hill, I were to encounter a car that was parked in the middle of the road, the Distronic Plus would treat it like any other stationary object—a building, a billboard, a mailbox—rather than a vehicle that might move soon. And radar doesn’t like bends. “If a curve is sharp,” Delgrossi says, “it cannot follow objects in front of you.”'

by David C on Jan 25, 2012 2:41 pm • linkreport

THE ITS INITIATIVE MAIN FLAW: The future of automated freight delivery and personal transportation is neither a delivery truck nor a car, the ITS initiative main flaw is assuming it will!

Why is anyone trying to figure out how to make a truck and a car drive themselves after they are designed and built to be driven? I suppose the reason is that we already have the vehicles and the roads; but, has anyone considered developing and implementing a fully automated freight delivery and personal transportation system using vehicles no one has to drive from the start? I did and submitted a U.S. Patent Application for it.

Please go to my post to read more about it... Thanks, AZ

by Alberto Zayas on Jan 30, 2012 8:58 pm • linkreport

Add a Comment

Name: (will be displayed on the comments page)

Email: (must be your real address, but will be kept private)

URL: (optional, will be displayed)

Your comment:

By submitting a comment, you agree to abide by our comment policy.
Notify me of followup comments via email. (You can also subscribe without commenting.)
Save my name and email address on this computer so I don't have to enter it next time, and so I don't have to answer the anti-spam map challenge question in the future.

or

Support Us