High Stakes for Cities as Feds Start Regulating Self-Driving Cars

Last week as part of his State of the Union Address, President Obama announced a $4 billion investment over the next 10 years to test autonomous vehicles and get them ready for the market. Two days later at the Detroit Auto Show, Transportation Secretary Anthony Foxx announced that federal regulators would begin to develop coherent safety regulations for autonomous vehicles — something industry leaders have been pushing.

How long before we start seeing self-driving cars in cities? What kind of change will they bring? Photo: Smoothgroover/Flickr
How can the emergence of self-driving cars be shaped to benefit cities? Photo: Smoothgroover/Flickr

Before you dismiss these developments as just another sop to the car industry, consider the huge implications that autonomous vehicles could have for cities. There are upsides — NHTSA Administrator Mark Rosekind has said that self-driving cars “can eliminate 94 percent of fatal crashes involving human error” — and there are downsides as well. The ease of operating autonomous vehicles could lead to supercharged sprawl, for instance.

The emergence of self-driving cars raises a host of questions about issues ranging from liability in the event of a crash to the potential for shared autonomous vehicle fleets to free up huge amounts of street space.

Right now there’s a patchwork of state laws regulating the self-driving prototypes that companies are testing (and many states have none). Last month, California released the first state rules governing autonomous cars for public use.

I’m not saying this is all good. I’m saying it’s inevitable, so we should be shaping the way it happens.

Federal regulators say they will work with a group of states, car makers, and other interests to establish model legislation for states. Meanwhile, the National Highway Transportation Safety Administration will develop performance standards for self-driving cars.

So the next six months will be a critical time in shaping how self-driving cars are adopted. What should people who care about city streets look for during this process?

In his book Startup City, former Chicago and D.C. transportation director Gabe Klein touches on the emergence of self-driving cars and the potential consequences for cities. We spoke to Klein (who also serves on the board of OpenPlans, the organization that publishes Streetsblog USA) about why these regulations matter and what to look for as they’re developed.

What should urbanists be watching for in this regulation process?

Around the turn of the last century we saw the invention of the combustion engine and the automobile dramatically change the landscape, I would say for the worse.

We are on the cusp of a major technological change. I think this time around we should be thinking, “What kind of life do we want?” And kind of fitting the technology into the kind of life that we want.

We could face a much more utopian urban future as a result of good policies that embrace technological change but fit it into the context that we want in healthy, sustainable, vibrant, economically growing cities. Or it could continue to allow the bad planning decisions of the past to continue. To see sprawl sort of justified.

Are there examples of specific regulations we should be watching? 

I think the [Obama] Administration is very cognizant of the fact that this could have huge consequences. Not just the regulations but the way that it’s sort of rolled out.

You’re talking about kids being able to walk to school again. You’re talking about being able to reallocate up to 30 percent of a city’s space to people again.

Let’s say we incent the adoption of these vehicles into shared-use fleets. That’s a very different outcome than incenting them in terms of personal vehicles.

I think we should be looking at deploying them in shared-use fleets first. We have heard these people from GM saying they should be part of fleets first.

Now that we have another generation of people coming along that very much view their connectivity as technology and a smartphone versus a car, there’s a big opportunity to really commoditize point-to-point transportation.

That’s sort of like what they did in the new California regulation. It only allows companies like Uber to operate self-driving cars. 

Yes, but overall I think the proposal is pretty limiting. I think it’s going to push car companies to test their vehicles in places like Texas. It sort of struck me as a proposal to say, “Hey, maybe we should slow this down a little.”

What GM is doing strikes me as very smart. They say, we’re going to be selling things in different contexts. They’re saying, in cities we’re probably not going to be selling people cars any more.

People are interested in service on demand. It used to be everybody owned a cow. I would own a cow and you would own a cow. At a certain point people said, wait a second I just want the milk. People are not interested in buying the cow, they’re just interested in buying the milk.

So how do you imagine this might impact city streets?

It will be interesting to see how some of the early cities like LA or Seattle and how they will start testing things and the signals they send. For example, how do we allocate the right-of-way? How do we incent delivery at night?

But don’t the auto companies have a clear self-interest in promoting a personal ownership model? They’re going to be key in driving this process.

We only have one time when we can really engage in this and shape the changes.

I’m not saying this is all good, I’m saying it’s inevitable, so we should be shaping the way it happens.

What happens in terms of autonomous transit and freight vehicles? Why do we have tractor trailers blocking bike lanes at rush hour? All of that should happen at night.

What we really need to do is focus more on the outcomes we want. Right now, typically we have three out of four seats in every car empty — it could be very efficient. Cars are not inherently inefficient, it’s the way we use them that’s bad.

What could be some of the good outcomes, in your opinion? 

There are so many consequences, if we get rid of 85 to 90 percent of the parking in cities as a result of the convergence of some of these technologies and people no longer needing to own their own cars.

You’re talking about kids being able to walk to school again. You’re talking about being able to reallocate up to 30 percent of a city’s space to people again.

We only have one time when we can really engage in this and start to really shape the changes to roll out. I would encourage folks who care about active transportation and equity to be engaged. It may be the most consequential change in our lifetimes in terms of cities.

44 thoughts on High Stakes for Cities as Feds Start Regulating Self-Driving Cars

  1. The key thing which needs to happen in terms of regulation is once self-driving cars prove themselves then we should make it illegal to have a human-driven motor vehicle on public roads as rapidly as possible. That might include retrofitting existing cars with self-drive technology. We don’t want to have to wait until most manually-driven cars wear out before finally being rid of human drivers. That could take 20 years or more. If self-driving cars prove themselves over the next few years, we should aim for the early 2020s as the date when the days of human driving are over. Yes, people will object but safety is more important. The carrot here might be that self-driven cars should allow much higher highway travel speeds than now. Most people would trade the ability to drive themselves for 100 to 125 mph travel instead of 70 mph travel.

    What’s interesting is if we make this happen on the time frame I mentioned, then there would not be any need for anyone born in the 21st century to bother obtaining a driver’s license.

  2. I think that problem will take care of itself through rapid increases in insurance premiums for meat-driven cars.

  3. Speaking as a pedestrian, I’d rather take my chances with self-driving cars than with human drivers. I see human drivers making bonehead moves that endanger my life practically every day. Including today — bonehead did U-turn on West End Ave., far from the first time I’ve seen that. Bring on the robots.

  4. In my opinion, the biggest social impact that will result from fully autonomous cars is a loss of how many tens of millions of driver jobs in T&L. I’m not saying this is a reason to oppose self driving vehicles, I’m saying this will be a consequence very quickly, and something we as a society will need to deal with.

  5. Eh, we’ll deal with it the same way we generally deal with structural unemployment is the US—we’ll do nothing.

  6. …make it illegal to have a human-driven motor vehicle on public roads as rapidly as possible.

    This is America—you’ll have to pry that steering wheel from people’s cold, dead hands. I’m sure there’ll be some arguments made that driving is constitutionally-protected right.

  7. One interesting, and not often talked about impact of a self-driven future is sprawl:

    If you have to choose between 45min on the subway or 45min sleeping in a self driven car, which would you choose?

    I expect commuting suburbs and sprawl in general to significantly increase once everyone basically has their own private robot chauffeur.

  8. There will be “disparate impact” lawsuits that will stop this, on the basis that it discriminates against those who can’t afford self-driving cars. Normally, we in the US don’t care when something negatively impacts the less well-off but driving is seen as a right and a moral good.

  9. I do as well. I know a lot of urbanists are whistling past the graveyard on this one, but self-driving cars are going to reaffirm the low-density, auto-oriented, suburban sprawl model. I would argue they would in any case, but as we haven’t shown any inclining to de-prioritize auto travel, letting self-driving cars totally take over is the path of least resistance for auto and development planners. The idea that traffic engineers will suddenly start using excess road capacity for non-auto purposes (bike lanes, increased pedestrian space) is laughable.

  10. Ultimately, automation in general is probably going to eliminate the vast majority of jobs. What this might mean is what the robots produce will not necessarily belong to those who own them. It’s a very complex problem, but ultimately we’ll probably have to move in the direction of everyone being entitled to some minimal amount of goods/services/shelter, with the option to obtain more if you can find some sort of work. The alternative would be letting everyone who loses their jobs due to automation starve to death.

  11. This feels like the whole point of the article… it’s inevitable, it’s not all good, but policy can shape things if we address this now.

    I too worry about the implications of the private automated car model, including massive sprawl. Heck, if you could choose between 90mins in your private automated car or 45mins on transit…

    But I think the shared ownership model mitigates this. If the AV is not in your garage to start the day, sprawling home ownership means long wait times for your car to come get you. And if we plan ahead, a shared model can be far more affordable than private ownership while still facilitating great usefulness, including providing access to many who cannot drive today.

  12. How about the costs associated with infrastructure to enable sprawl?

    I agree that without policies like urbanization boundaries, sprawl may increase some. However, I think there is a limit to how far out you want to build roads, piping, and other connections.

  13. Exactly, I highly doubt that driven automobiles will be off our streets anytime soon. Not until the driving only generations die out.

  14. I’m really interested how autonomous driving will influence cities. Existing infrastructure and collections of people (the latter being vital in a social context) are highly valued so cities have a place in the near future. I wonder how this technology will affect mass transportation and the allocation of space?

    Being optimistic, I feel that autonomous vehicles will inevitably lead to less space allocated to automobiles, and more [hopefully] public spaces for alternative uses. In cities.

    In cities with very high population densities, it may still likely be more efficient to utilize high capacity vehicles to move most people between key areas. However, I see sharing services with autonomous personal rapid transit type vehicles replacing driven automobile ownership for non-traditional trips within these areas. Even now, car share makes a heck of a lot more sense than ownership for most in dense cities.

  15. Self-driving cars will be cheaper. Right now, the extra hardware for self-driving cars is not really expensive (cameras, computer and controllers). And marginal costs would go down quickly with mass production. The challenge now is more on the software side.

  16. I think Interstates, parkways and other controlled-access roads will be the first ones to ban human-operated cars. It is far easier to fit a ROW meant exclusively for cars with the sort of operational parameters necessary to make it fully self-driven.

    Just remember freeways impose some speed limits because of their curve radii, but with an all-computerized driven fleet it should be possible to travel at 100-110mph in most places where speeds are currently limited to much less (think NK Turnpike).

  17. I think we are living at pivotal times where the planning/urban improvement community will, collectively, make a decision. It can fully engage with the process that will roll out self-driven cars, and help frame the debate and the outcomes, or it can play make-believe that cars are going out of fashion anyways, so why bother, leading to worst outcomes.

  18. More car dependence cannot make it safer and more convenient to walk. And people will still want parking close to their destination because they don’t want to wait 20 minutes to go anywhere, just for their car to get to them from a remote lot.

  19. In dense cities where mass transit is adequate. Which unfortunately excludes 99.9% of cities here in the US.

  20. I agree. DOTs will still be focused on making cars travel as fast and efficiently as possible, whether they are self driving or not, pedestrians and bicyclists be damned.

  21. True, but if we keep on building low density, car-dependent sprawl, there is nowhere to walk to in the first place…

  22. Self driving cars will be cheaper because you don’t have to own it. Pay per trip is primarily expensive because of the driver. And you pay for the sort of trip you need – pod car, Truck, etc…

  23. Nope. All you have to do is increase the cost of insurance, and be very strict about enforcing penalties for traffic violations.

  24. One thing to think about – what it the number one reason people don’t bike? Because they are afraid of being hit by a car. Take that problem away completely and you increase cycling. Especially in a world of self driving car fleets where each incremental trip has visible cost.

  25. Even if the whole autonomous car thing never materializes, I could see some interesting safety technology coming out of this program. For example, what if the V2V idea was repurposed to enforce speed limits across a city via a 2-way communication system and smart vehicles. The top speed of a car on a city street could be governed based on the speed limit of the street it’s on. PS The movie Hot Tub Time Machine 2 envisions a future where everyone uses a car-share service of autonomous Smart Cars.

  26. But autonomous vehicles have the potential to become a new kind of mass transit. The best way this can work for cities is to design them to move more people using less space than cars currently get and to design them to allow more freedom of movement for people using other modes — i.e. use the technology to eliminate traffic signals and to give pedestrians and cyclists priority. Is all of this even possible? I hope so.

  27. Note the driver will have the ability to take over control. It’s estimated that drivers will use manual over-ride possibly a majority of the time to avoid being taken advantaged of by other manual vehicles. Therefore it’s not a self driving car, it’s a human driven car which is self driving only sometimes.

  28. For many larger cities, the sprawl is already there, the only question is which neighborhoods will grow, and which will shrink. Take NYC (the example I’m most familiar with):

    The core will do fine regardless: there will always be people that want to be real close to everything, but what about the outer boros? Plenty of people in southern brooklyn, north bronx, and eastern queens face 50min+ commutes on the subway. Why not move out to the suburbs and get the same via personal driverless car, while also getting a bigger house and more nature?

    In the future, it’s likely the less fortunate will be living in these areas, as anyone that could afford it would either live in the center, or within driverless car commuting distance out in the burbs.

  29. Self-driving cars operate as ground-based surveillance drones collecting images and data on both drivers and the public at large, according to World Economic Forum insiders.

  30. “You are going to be able to be tracked like you’ve never been tracked before. That has good uses like charting traffic flows. And then there is the issue that most of the data will be collected by private corporations and they will seek to use that for their benefit.”

  31. If you thought all the tracking, surveillance and data profiling from people’s cell phones is creepy, just you wait.

  32. I disagree with some of the assessment in this article. First, of course 94% of crashes are due to driver error, because the driver is responsible for 100% of the driving. What this statistic says is that all in all, cars are pretty reliable from the standpoint of mechanical failures. While we expect a net decrease in crashes, the percentage due to systems failures will go way up. Second, I think we need to remember that California only proposed regulations, they have not yet adopted them, and the regulations that were proposed will have a limited duration. It would be wise to think of them as regulations for limited pilot testing over the first 3 years of automated vehicle deployments because the DMV has repeatedly stated that they expect multiple regulation packages changing rapidly over the next few years. Unless NHTSA does a 100% about-face from their original 2013 policy, I would expect whatever new policy that they come up with over the next six months to mirror that approach, especially given that the stated purpose for requesting $4B over the next 10 years was to help conduct pilot deployment projects.

    Finally, I highly doubt that we will get rid of 85-90 percent of the parking in cities. Where do you think the automated cars are going to recharge? Where are all the extra capacity vehicles needed during the rush hour but not in between rush hours going to go during the non-peak hours? Given the peak commuting hour demand imbalances, having no parking anyone in a city the size of SF would result in a traffic disaster with 0-passenger trips in and out of the city 4 times a day. And if every city created policies to eliminate 90% of parking, there would simply be nowhere for the vehicles to go and charge.

  33. The european projects referenced in that article are all part of the CityMobile project, and they run at low speeds on dedicated lanes shared with bikes and pedestrians. Unfortunately, there was no exemption for Low-Speed driverless shuttles like this in the proposed California regulations. If the proposed regulations pass, these types of system will only be allowed on private property, not public roads.

  34. Why would there be a rapid increase in premiums for human driven cars? That would assume that human driven crashes would go up from current levels. More likely, premiums would stay the same for human driven cars. If automated cars end up being safer, then the premiums to operate those cars would go down.

  35. I”m not sure I’d worry about reduced parking in cities leading to nowhere to charge one’s vehicle as it would be way too expensive to wire the majority of parking spaces for charging anyway. Most likely, over time our vast network of gas stations will morph into express charging stations of some sort.

  36. I don’t think it will decimate an industry overnight, but over time livery jobs will likely drop. But that happens as technology progresses … and presumably new industries and jobs will emerge as the evolution continues.

  37. I doubt it will take very long. As soon as someone does it, everyone must do it to stay competitive. I’d give it less than the lifespan of a vehicle. That’s an awfully short time to loose millions of jobs. New industries will replace that, but not so quickly. I would be better to have a plan that was more substantial than a shrug.

  38. I’m not sure what you’re considering as “the lifespan of a vehicle” (10 years? 30 years?), but I don’t think all these jobs will dry up overnight. I’d expect to see a gradual but consistent reduction of livery jobs, but I think the pace of the reduction will depend on the many local and state laws that regulate automated vehicles and how these regulations are implemented. (With any luck, hopefully the job loss will roughly parallel the retirement rate.)

  39. 10 years max. Most buses, cars, small vans barely last 10 years.

    I think you’re wrong about the slow gradual decline of jobs, and I’m certain it will vastly exceed the retirement rate. Yes, it depends on specific laws, but just think of who writes the laws in the US. Never in history has technology been stopped from supplanting jobs, I don’t expect it to happen this time.

Leave a Reply

Your email address will not be published. Required fields are marked *


How the Self-Driving Car Could Spell the End of Parking Craters

Here’s the rosy scenario of a future where cars drive themselves: Instead of owning cars, people will summon autonomous vehicles, hop in, and head to their destination. With fewer cars to be stored, parking lots and garages will give way to development, eventually bringing down the cost of housing in tight markets through increased supply. […]