How will autonomous driving vehicles change our cities?

[the urban politician]

I think this belongs here instead of the transportation forum because the focus is on how cities will change as a result of this fast arriving technology.

I'd like to speculate how self driving cars will alter our cityscapes over the course of the next 50 years, assuming humanity doesn't destroy itself.

Here is what I see happening:

1. Less car ownership and greater use of ride sharing
2. Decline in transit usage in all but the densest areas
3. Correlated with the above, reduction in the property value advantage that properties near transit presently enjoy
4. Less development pressure to create off street parking for residential and commercial projects
5. More walkable streetscapes being built as a result of #4 above
6. Creation of more housing farther from existing transit networks due to #3 above, and an increase in property values close to major roads
7. Creation of huge "car-share stations" which deploy large numbers of autonomous vehicles which go out to pick up and drop off people
8. Reduced congestion on roads, increase economization of existing road networks, and increased desirability of suburban office space
9. A widening of the already existing gap between the booming metros and floundering small towns and rural areas, since as a result of economies of scale larger metros will have much more well established autonomous car sharing networks for their residents.

[dave8721]

It seems all of the below depend on #1 but why would #1 be so? Assuming they were able to make automous cars relatively affordable (maybe a tenuous assumption), why would people abandon their own self driving car for a sharing service that they would have to wait for?

[brian_b]

I think that in the short term, the vehicles are most likely to be used to assist those with disabilities, old age, or prohibitions on driving.

I think you are going to see much stricter enforcement on the privileges of driving. States are going to be much more likely to take away driving licenses because you will no longer claim that you can't get to work without one. DUIs, speeding, running red lights, etc.

[JManc]

They will evolve into sentient beings and demand equal rights and that we design our cities to be even more car friendly. Slippery slope...

[chiphile]

The biggest implication is for transit. Smart cities will intertwine self-driving vehicles into their systems so that transit can give 100% point to point options for everyone, thereby making transit a very appealing option.

Metra in Chicago is making a deal with uber or lyft for this very purpose.

In the meantime, I hope they make driving tests stricter and force all the slow, lack of spatial intelligence morons who currently clog our roads, into self-driving cars, and allow the better drivers to drive their own cars.

[The North One]

There's a big dilemma in this car sharing idea.

Not only is your vehicle now doing twice the driving, it has to do more driving on top of it to get from the people renting out the self-driving car back to you, essentially the opposite of efficiency. The self-driving car is doing more driving than two separate cars would be combined. There is still the same amount of cars on the road, it poses no benefit to the environment and is in no way a replacement for mass transit.

[relnahe]

Bars and clubs will see a big uptick from people being able to indulge into booze with no worries.

[the urban politician]

Originally Posted by dave8721

It seems all of the below depend on #1 but why would #1 be so? Assuming they were able to make automous cars relatively affordable (maybe a tenuous assumption), why would people abandon their own self driving car for a sharing service that they would have to wait for?

To the contrary, I think car ownership will decline. Stringent standards and regulations of vehicles will only increase, and thus many people may see no value in owning a vehicle.

Instead, a network of self-driving vehicles will exist which can take you to all of your needed destinations for a cost, I envision some sort of subscription.

we may reach a point where there are no human drivers on the road.

[eschaton]

The predictions I have seen suggest that urban car ownership (as opposed to car usage) is going to begin collapsing by 2025. Basically people who only use a car a few times a week for shopping and socialization will move over entirely to ride share, going carless. Two-car families will be likely to downgrade to only one car as well, as rideshare can fill in the gaps - plus if they own a self-driving car, one partner can send it to the other one to use once they've reached their destination.

That said, subscription models will never replace other modes of transportation in one area - peak travel during rush hour commutes. It won't be profitable for rideshare companies to offer enough cars to cover rush hour, because those cars will sit around all day not collecting fares. Hence if you need a car for your work commute, chances are you will still need one. Similarly, self-driving cars will not replace high-ridership commuter transit lines, as they are far more efficient at getting large groups of people in and out of job centers than individual "pods" would ever be. Actually, self-driving cars can solve the "last mile" issue for park and rides, which means rapid transit lines should see an uptick in usage.

The biggest difference in terms of cities will be the death of proximate parking, and the rise of remote parking. City parking authorities may have to declare bankruptcy, for example, as even most people who own self-driving cars will instruct their car to go find somewhere free to park. The upside of this is downtown parking garages and lots will be pretty quickly redeveloped, leading to a much denser and more walkable urban core. Podium parking will be hard to un-engineer, however - in a lot of cases these garages may become the resting spots for portions of rideshare fleets.

Inter-urban travel (Greyhound buses, Amtrak, etc) will totally collapse. The schedules are simply too inconvenient compared to longer-distance rideshare rental, even if the costs can be brought to parity.

It will be interesting to see how the suburb is ultimately retrofitted say 30 years from now in the post-car ownership culture. I expect creative reuse of garages (particularly the bigger ones) will be a big thing.

A lot of small rural communities are going to go into steep decline. Particularly the ones which sprung up to cater to truck drivers around highway exits. Trucks may still need to refuel, but with no humans behind the wheel, there will be no need for the convenience stores, fast food places, and cheap motels which have proliferated.

[the urban politician]

^ A lot of great thoughts there.

If city revenues from street parking plummet, that would ironically turn Chicago's notorious parking meter deal from on of the worst to one of the best municipal deals in recent times.

[eschaton]

Originally Posted by dave8721

It seems all of the below depend on #1 but why would #1 be so? Assuming they were able to make autonomous cars relatively affordable (maybe a tenuous assumption), why would people abandon their own self driving car for a sharing service that they would have to wait for?

Rideshare is most attractive to people who don't use cars as part of a regular work/school commute, but need them a few times per week to a few times per month for other reasons, like shopping trips and socialization.

Essentially add up all the costs for owning a car right now, including car payments, insurance, repairs, gas, oil changes, registration, drivers licence renewal, and potentially property tax depending upon if cars are taxed in your state. That is the annual cost for you to own a car. In 2015 this was roughly $6,000 for the average household. Now, the average American drives around 280 hours per year, meaning the hourly cost for owning a car is around $21.43. But if you have similar annual costs, but you drive 1/5th as much, your cost per hour of driving is over $100 per hour.

Traditionally speaking, it's already cheaper for many of these occasional drivers to use other systems, like taxis, rental cars, and transit. The problem historically with such services is they are a hassle. Traditional taxis can take over an hour to arrive, and may not come at all. Transit systems for off-peak travel between non-core neighborhoods are weak and have spotty schedules. if you want to rent a car, you need to first travel to a certain locale usually far from your house. The hassle makes the lower cost not worthwhile in terms of metal energy to many consumers.

Ridesharing is already taking much of the hassle out for people, even when human drivers are included. Once there are no more rideshare drivers, the cost should fall further, and the hailed rides should get there faster. At that point, there is no logical reason to own a car any more.

Do I think everyone will ditch their cars at once? Absolutely not, if for no other reason for the fact that used cars won't be worth crap. But I do think that young people who had no history of driving cars will see no reason to begin to drive, meaning within a generation or two mass car ownership will collapse.

[eschaton]

Originally Posted by The North One

There's a big dilemma in this car sharing idea.

Not only is your vehicle now doing twice the driving, it has to do more driving on top of it to get from the people renting out the self-driving car back to you, essentially the opposite of efficiency. The self-driving car is doing more driving than two separate cars would be combined. There is still the same amount of cars on the road, it poses no benefit to the environment and is in no way a replacement for mass transit.

While some models, like Tesla's, call for consumer-owned cars to be rented out on their off hours, the subscription models of rideshare which are being promoted by Uber and Lyft do not. They should be optimized to have as little time spent driving to the next customer as possible, which should reduce time spent on the road, when you consider the car will not be driving to a parking location.

That said, lots of people will begin using cars who don't now. The elderly and disabled for sure. Eventually I'm guessing parents will feel okay with having their kids in self-driving cars alone, provided they can only go to certain destinations (reminiscent of parental controls on web broswers), so it's possible there will be more cars on the road.

On the third hand, self-driving cars will not look like cars do today in the longer run. There's no reason for a rideshare company to send out a larger car than you need, so we'll see a proliferation of 1-3 seat cars which take up less road space and have lower emissions. Once there are no more human drivers on the road, safety will climb dramatically, allowing for cars to be made from much flimsier materials (like plastic and styrofoam) again allowing for lower emissions. And there are arguments that AI will be able to use things like wind wakes from other vehicles to further cut down on fuel emissions, along with never having to engage in hard breaking. So self-driving cars should ultimately be much more energy efficient.

[The North One]

That still sounds nowhere near as efficient as transporting passengers in bulk like a bus or a train. Buses would be going through those changes as well if these "safe roads" are realized making them more efficient and driverless, but I don't see that happening because you would have to get everyone to stop driving their own cars in order to eliminate all human error. people are always going to have/want their personal cars and good look passing legislation preventing citizens from driving their own cars in the freedom obsessed nation that is the U.S. or even democratic European nations for that matter.

[dave8721]

Originally Posted by the urban politician

To the contrary, I think car ownership will decline. Stringent standards and regulations of vehicles will only increase, and thus many people may see no value in owning a vehicle.

Instead, a network of self-driving vehicles will exist which can take you to all of your needed destinations for a cost, I envision some sort of subscription.

we may reach a point where there are no human drivers on the road.

How is that any different from a system like Uber or taxis? My point is that self-driving vehicles (in some far off future) would become cheap enough for everyone to own one (if price points approach what you currently pay for a car now) people who own cars now would probably want to own a self driving one as a replacement to their existing one.
Yes if a self driving car would cost like ~$100,000 than only large car sharing companies could afford them which maybe they would be at first, but eventually they would be mass produced and fall to a price point where they are affordable to everyone (like big tv's, computers...etc).

[eschaton]

Originally Posted by The North One

That still sounds nowhere near as efficient as transporting passengers in bulk like a bus or a train. Buses would be going through those changes as well if these "safe roads" are realized making them more efficient and driverless, but I don't see that happening because you would have to get everyone to stop driving their own cars in order to eliminate all human error. people are always going to have/want their personal cars and good look passing legislation preventing citizens from driving their own cars in the freedom obsessed nation that is the U.S.

I'm 37. I think it's entirely possible cars will be banned from at least some U.S. streets by the time of my old age, if you presume the following things.

1. Self-driving technology becomes totally ubiquitious (in every new car) by 2035.

2. By around 2050, virtually all of the old manual cars have aged off the road.

3. Insurance rates for manually driven cars begin shooting up through the roof, because they remain an order of magnitude more dangerous.

4. Kids in the next generation (say under 10 now) never bother learning how to drive, because self-driving is going to be perfected in their teenage years. Fast forward another 20 years and between this group and slightly older Americans who ditched car ownership may be a voting majority.

I do agree it will never be as efficient as a bus or train. However, it will be more convenient, as it will not work on a fixed schedule.

[mhays]

Lots of good points in this thread.

Car sharing is already paying off. I haven't seen the latest stats on car ownership, but it seems to be going down at least in urban locations. New buildings with little or no parking used to be rare in my city, but they've become very common, for this and other reasons, and this is happening at higher rent levels. The same seems to be the case anecdotally in other cities. This is allowing a lot more density than would otherwise be possible.

Self-driving cars are another topic. People can buy them and use them in the current infrastructure in less-urban locations in ways that will never work in city cores. In the latter, car sharing should prevail along with continued use of transit for the major routes.

One topic will be damage to the cars and who's responsible. There will be an underclass who's kicked out of car sharing because they ruin cars.

[tjp]

Originally Posted by dave8721

It seems all of the below depend on #1 but why would #1 be so? Assuming they were able to make automous cars relatively affordable (maybe a tenuous assumption), why would people abandon their own self driving car for a sharing service that they would have to wait for?

I'm wondering this, too. Why will self-driving cars reduce car ownership more than Uber / Lyft already are? Will they make rides cheaper?

[eschaton]

Originally Posted by tjp

I'm wondering this, too. Why will self-driving cars reduce car ownership more than Uber / Lyft already are? Will they make rides cheaper?

Self-driving cars should be cheaper in the longer-run than having rideshare drivers - otherwise Uber wouldn't be investing so much money in perfecting the technology.

Look at it this way. Say right now an Uber driver working full time makes around $40,000 annually. There is of course also the cut that Uber takes in terms of profit off the top, along with the costs that the driver must have to maintain the car.

Now presume you're dealing with an Uber-owned, self-driving car. Say the technology costs $100,000 to implement per car. Amortize that out over ten years, so that $10,000 in revenue per car must be put toward paying back for the self-driving system. Also presume the regular maintenance costs for the cars are roughly similar, but the cost is now borne by Uber instead of the individual contractor. There is then around $30,000 left to play with it would seem, which means Uber can both cut prices and have a lot more profit per vehicle as well, as it is directly capturing the revenue rather than just charging commission to an independent contractor.

But the real bonus is above and beyond that. The self-driving car can drive for 24 hours straight, with no piss breaks or sleeping breaks - only stops to refuel. This means that a self-driving car should always have more hours driven than a human-driven car would. It also means more wear and tear, of course, but it's more efficient to have a smaller fleet constantly in use versus a larger fleet in use half the time or less.

[tjp]

Originally Posted by eschaton

Self-driving cars should be cheaper in the longer-run than having rideshare drivers - otherwise Uber wouldn't be investing so much money in perfecting the technology.

Look at it this way. Say right now an Uber driver working full time makes around $40,000 annually. There is of course also the cut that Uber takes in terms of profit off the top, along with the costs that the driver must have to maintain the car.

Now presume you're dealing with an Uber-owned, self-driving car. Say the technology costs $100,000 to implement per car. Amortize that out over ten years, so that $10,000 in revenue per car must be put toward paying back for the self-driving system. Also presume the regular maintenance costs for the cars are roughly similar, but the cost is now borne by Uber instead of the individual contractor. There is then around $30,000 left to play with it would seem, which means Uber can both cut prices and have a lot more profit per vehicle as well, as it is directly capturing the revenue rather than just charging commission to an independent contractor.

But the real bonus is above and beyond that. The self-driving car can drive for 24 hours straight, with no piss breaks or sleeping breaks - only stops to refuel. This means that a self-driving car should always have more hours driven than a human-driven car would. It also means more wear and tear, of course, but it's more efficient to have a smaller fleet constantly in use versus a larger fleet in use half the time or less.

Makes sense -- thank you!

[JManc]

Originally Posted by eschaton

I'm 37. I think it's entirely possible cars will be banned from at least some U.S. streets by the time of my old age, if you presume the following things.

1. Self-driving technology becomes totally ubiquitious (in every new car) by 2035.

2. By around 2050, virtually all of the old manual cars have aged off the road.

1) If you're referring to total autonomy where the "drivers" ride around able to do other things like in the Jetson's then no way. That's only 19 years out though I could totally see this with assistive technology where the car takes over at green lights to ease bottlenecking/congestion, prevent imminent accidents (My 2016 Acura brakes if I'm too fast behind another car and corrects steering if I drift out of lane) or goes into some sort of autopilot on the open highway. Total autonomy entails a lot of human, infrastructure and political factors not even remotely ready for this. We don't even have laws proposed or on the books to deal with autonomous car related issues. Which is why that fatal Tesla accident in the Netherlands is making news.

2) Not if we keep producing them and there's no timeline as to when they will be phased out...likely by federal mandate. Which is unlikely anytime soon.