University of Oregon guide helps cities manage self-driving cars on their streets

Some U.S. cities are contending with the rollout of self-driving cars and other autonomous vehicles on their roads. Others are still considering how to approach AVs and whether they’re a good fit for the community. A new guide from the University of Oregon’s Urbanism Next Center aims to help leaders navigate these challenges.

Nico Larco is a professor of architecture and urban design at UO and the director of the center. He joins us with more details on what cities should be thinking about as autonomous vehicles continue to roll out.

This transcript was created by a computer and edited by a volunteer.

Dave Miller: This is Think Out Loud on OPB. I’m Dave Miller. Some cities, like San Francisco, Miami, and Pittsburgh, are contending with the roll out of self-driving cars and other autonomous vehicles on their roads. Others are still considering how to approach AVs and whether they’re a good fit for their communities. A new guide from the University of Oregon’s Urbanism Next Center aims to help leaders navigate these challenges. Nico Larco is a professor of architecture and urban design at the University of Oregon and the director of the center. He joins us once again with more details on what cities should be thinking about as autonomous vehicles become more common. Nico, welcome back.

Nico Larco: Thanks so much. So great to be here.

Miller: We last talked with you about autonomous vehicles (AVs) a lifetime ago, seven years ago. How has the AV landscape changed in those seven years?

Larco: A lot has happened. And if you would have asked me last year, I would have a totally different answer. The whole of 2023 was a completely crazy year for the development of AVs and the roll out of AVs. I remember the last time we talked, it was a long time ago in 2015, there was an article that had come out where someone had talked to the top 10 or 11 car manufacturers in the world. And 10 of them had said that they would have AVs on the road by 2023. We’re in 2024 and we’re still not there. So you can see there was a big difference in what was projected and what happened.

Miller: Well, there have been some on the road, but I guess the idea was they would be super common and that they would be replacing delivery vans or passenger cars?

Larco: Right. Exactly. They would be widespread and some of the projections were how many millions of cars there would be on the road. That was the question.

That has not come to pass. AV roll out has gone a lot slower than many anticipated. It turns out that the technology is pretty hard to make work well. But there’s been a lot of advancement. In 2017 we’re still mostly thinking about things like “Level 3″ autonomous vehicles. These are cars where there’s someone at the driver’s seat waiting at a moment’s notice to take over the wheel because the vehicle is seeing something they don’t understand or can’t really safely function through.

And since then, we’ve had a lot more development in what’s called “Level 4″ autonomous vehicles. And these are the vehicles where you have nobody in the front seat at all. So there’s no one who can recover or needs to recover in the front of the car. And that kind of leads us to what’s been happening in San Francisco, which has been more or less the epicenter of all things AV for the last year… more than a year. In late 2022 and early 2023, Waymo and Cruise, Waymo is Google’s autonomous vehicle arm and Cruise is a company owned by General Motors primarily, they started operations in San Francisco with Level 4 vehicles. These are vehicles that had no one at all in the front seat.

Miller: These are operated largely as taxis, right?

Larco: These were operated exactly [like] what are often called robotaxis. They’re pretty much like taxis or like an Uber or Lyft-type service. You call it on your phone, it shows up, you get in, it takes you somewhere, you get out and it goes on its way. And they had been running Level 4.

I actually rode in one of these last summer, where you call up the vehicle, it shows up with nobody in it at all, like no person in it. You get into the back seat and you close the door. It reminds you that you need to have your seatbelt on and then it takes off and takes you through the city and drops you off somewhere, and then it goes on its way.

Miller: Can I ask… maybe you’re the first person I’ve talked to who’s experienced this. I guess it’s more common in San Francisco than it is in most of the world. What was it like?

Larco: At first, it’s bizarre. I mean, it’s a little bit exciting and kind of crazy to think that this car with nobody in it is coming to pick me up and you get into it and drive around. And there are kind of strange things that happen. For instance, I was in San Francisco riding in the car, and at some point, there was construction going on. And you wonder what it’s going to do because it needs to cross the double yellow line to get around this thing, which any of us would know that we had to do. We’d kind of slow down and make sure there was no one coming on the other side and then [kept] on going.

And the vehicle kind of comes and stops and there’s a sign. I don’t remember exactly what it says, but something along the lines of, it’s connecting to the mother ship back where GM crews, the technicians, are. And then it is approved and the thing crosses over the double line and keeps going on its merry way.

Miller: Do you think that it was communicating with a human being?

Larco: Yes, to my understanding, it communicates to a human being who doesn’t tell it, go exactly here. But it does what’s called breadcrumbs. It puts a little point out in a place that it might not be comfortable going and says it’s OK to go in that direction and the car decides to go and keeps going.

Miller: But is it your understanding that that level of human intervention is a stopgap before things are even more autonomous? They’re not yet sophisticated enough to know how to cross that double yellow line. And eventually it would be? Or might there always be a human back at headquarters to step in?

Larco: My sense is that there will be a human for a good long while, but there will be less and less need for intervention. But eventually the goal is that these things can operate, definitely, in environments that are, at some point, much more familiar. And

you’ve seen a lot of what’s called edge cases, things that aren’t really all that common. And they’d be able to operate mostly on their own. So that’s definitely the direction that it’s going.

Miller: It seems like your ride was relatively smooth besides that. But, as folks who saw headlines over the last year, the last six months probably know, Cruise in particular had a lot of corporate tumult and on-road violence or fatalities. What happened with Cruise this year?

Larco: So as I was saying, they’ve been operating for the last year in San Francisco and they ran into a lot of problems. Some of the things were more minor, like the cars just stopping in the middle of traffic which didn’t hurt anyone or anything. [It] just created congestion problems. It was a fender bender where one of the Cruise vehicles ran into a bus that was stopped in front of it. At one point, Waymo’s vehicles were having trouble moving in really dense fog, so problems that you would expect of a new technology in a complicated kind of environment, trying to figure this out.

Cruise then also had some problems where, for instance, a vehicle went into a block where people were fighting a fire. There were fire firefighters who were active, and the vehicle went into this area, not understanding that you’re not supposed to go in there, as we would normally understand. And [the AV] actually ran over hoses and the firefighters ended up having to break the glass to get it to stop. So there’s been a lot of these kinds of events that have been happening.

At the same time, there’s a whole lot of pressure from the AV companies to really start accelerating the roll out and making sure that they can make a profit here. Interest rates, as you know, have gone up which has put a lot of pressure on making sure that people can actually make money off this. It’s more expensive to not be making money. And so a big move happened on August 10th where Waymo and Cruise were both pushing for San Francisco to let them have vehicles out 24/7, charge for trips, no limit on where they were going within the city, and no limit on their fleet sizes. So this was a big push to open up the market for AVs in San Francisco. I think there were six hours of hearings that the California Public Utility Commission had about this. And it ended up, in a surprise actually to me, agreeing, allowing for these AV vehicles to be able to drive on the roads as they’d been asked.

And then what happened after that, you almost couldn’t have written it. It was a series of problems that the vehicles ran into. So that happened on August 10th. On August 11th, the very next day, 10 vehicles stopped in one of the neighborhoods in San Francisco because a concert had ended. Everyone got on their cell phones at the same time and jammed the signals for the cell towers and 10 vehicles just stopped. That was August 11th. Four days after that, on the 15th, a Cruise vehicle got stuck in wet concrete while driving in San Francisco. Two days after that, a Cruise vehicle crashes into a fire truck that’s got its lights on and the sirens going. And all that’s been causing problems.

Then something really serious happened. Maybe those things weren’t tremendously serious, but this is a more serious injury. On October 2nd, a Cruise vehicle is going down the road in San Francisco. There’s a vehicle next to it, what we call “legacy vehicles,” a normal car. And that vehicle hits a pedestrian, which throws the pedestrian in front of the Cruise vehicle. The Cruise vehicle, unfortunately, runs over this woman and drags her a little bit under its back wheel. It stops and then continues, stops for a bit and then continues on. It repositions itself, moving 20 feet further ahead, with the woman underneath. So the vehicle must have understood that something was wrong and then said, “I need to go to a safe location,” or some type of thing and then moves over, so drags this woman another 20 ft. So that was on October 2nd.

Miller: Just to fast forward, after that, there was a cascade of people basically being forced out of Cruise and that story is not done yet. There are still a lot more questions about even the future of this particular company.

I want to zoom out a little bit because the whole point of your report is that, whether cities like it or not in some ways, this technology is moving forward. And you’re offering a kind of road map or guidelines for how local leaders can and should be thinking about this technology, and whether or not it’s right for them. And you know that there are bigger issues than purely safety. But to stick on safety for a second.

How do we, as humans, make the most rational decisions about it when it comes to driving - ones that take into account the current situation, the totality of traffic violence that, as human drivers, we perpetuate on ourselves, but is also clear-eyed about the current and maybe future limitations of this new technology?

Larco: Well, safety is an enormous question. I think the figure is something around 40,000 people die every year on US roads and another 100,000+ are injured on US roads due to crashes. And so the idea of being able to do better is a tremendous motivation to have vehicles have us move in a way that doesn’t cause these things.

Miller: Waymo argues that they have figured that out. I shouldn’t say figure it out but that in 7 million or so miles traveled in a handful of cities, they say their vehicles are significantly safer than human drivers. Then again, that is the company crunching those numbers. Have there been independently verified studies that can show the same thing?

Larco: Yes and no. So there was a report that Cruise and Waymo put out together which said, as you mentioned, that they’re doing a ton better than human drivers. There’s a huge question in how you’re comparing their driving to other people’s, driving the same conditions. So, for instance, if I was to say driving down my local street, “I’ve never had an accident, so therefore I’m much safer than everyone else.” Well, everyone else doesn’t always drive down fairly safe local streets. So that’s where some of the issues come in. But it seems to me that they are moving towards a safer alternative than individuals driving their own cars.

I think the question is how close they are right now, and maybe the question that you’re asking as well is how close will they ever be to zero injuries, based on an AV driving? I’m not sure about that, but my sense is that we’re getting closer and closer to that. If we’re not there yet, then closer and closer to being safer than a human driver, if we’re not there yet.

Miller: One of the first questions that you say that city leaders should be asking when they’re contemplating whether or not to work with autonomous vehicle companies is, do we have the same goals and desires? What are examples of potentially differing goals?

Larco: Well, the companies, obviously, are private companies and they are looking for a profit which I don’t say in a negative way at all. I understand that companies need to make profits. Cities have worked with a lot of companies that make profits so that’s not really a detriment. But cities have a number of other goals - for instance, helping with inequities, improving access for people generally and then making sure that everyone has greater access to locations, decarbonizing transportation, making sure that we’re not creating more problems in terms of greenhouse gas emissions. So I think equity, safety, and environmental issues are absolute goals that most cities have. And the question is, how well does that align with what AV companies are able to offer right now?

Miller: Let’s say that city leaders, ideally with input and blessing from residents, decide that the interests are aligned enough that they would like to work with AV companies to have some say into the operation of autonomous equals within city limits. How much legal authority, in general, do cities across the country have to actually regulate the vehicles that are on their roads, as opposed to regulations at the federal or state levels?

Larco: That is a great question, and the truth is that most cities have fairly limited abilities to regulate these vehicles. Most of the regulations right now surround safety. And safety is really regulated mostly at the federal level to make sure that the vehicle is safe. That’s the way, traditionally, we’ve done this. And at the state level, mostly dealing with operations. So when I buy a car, the federal government has said that’s safe enough to be on the roads. And then when I want to drive that car, I need to go get a license from my state that says that I’m OK to be able to drive it. And so as safety has been the primary consideration for regulation around AVs, most of that has happened at the state and federal level.

Now, at the local level, we do have examples where cities have more say in regulating vehicles. Obviously, the operating environment exists within cities and local traffic laws exist. But we also have examples for things like taxis, fleets of vehicles, that exist within cities. And cities have been able to say, “we want to make sure our taxis, for instance, pick up any passenger.” They’re not going to discriminate or they’re going to make sure that they serve all these different areas and they won’t discriminate in those areas. Or even sometimes, [at the local level there are] emission standards, based on those things.

So it’s possible to have some say over how the deployment of vehicles is going to happen in your city. But generally, if the conversation is only about safety, that has happened mostly at the state and the federal level.

Miller: You mentioned taxis there and the kind of tech disruption that’s happened in the last 10 or so years. It does remind me that cities like Portland, and many other cities around the country, have had enormous challenges in regulating companies like Uber, when it comes to those kinds of ride-hailing app-based companies, or places like Airbnb. Wealthy tech companies at times, in various ways, have figured out ways around local regulations. Are there current examples of autonomous vehicle companies developing or even deploying their vehicles against the wishes of local governments?

Larco: Well, against the wishes, feels a little bit too strong. I don’t know if many cities have come out and said, “absolutely, we don’t want these anywhere at all.” I think maybe the way that they’re deploying, cities would often be looking for more of a collaborative approach - sometimes more collaborative, sometimes less. But they would like to be working with companies to make sure that the deployment vehicles, as I mentioned before, really align with the goals of the community, finding ways that they’re deployed that are helping get to these outcomes that cities want to see.

Miller: How might a city accomplish that, both in terms of the leverage they have that they can bring to bear and the mechanisms that you’re actually suggesting for how cities can be? Maybe “partner” is too strong a word, but just some kind of a seat at the table?

Larco: Well, actually the things that have been happening in San Francisco are, I think, helpful. Because as I was mentioning before, Cruise has run into a lot of problems. And they’ve actually, based on the things I was saying before, shut down operations completely. And that has been an enormous shock, I will say, to the entire industry. I think one of the messages that has come out from this is that a lot of the frustrations and a lot of the, I’d say, acrimony that has existed between cities and companies have created not a great environment for those companies.

So I think there’s a real shift in attitude. Cruise has definitely had a substantial shift. There were actually articles just out today talking about how the company is changing its stance. And Waymo, as well, has been a better actor with cities and really has a renewed interest in working directly with cities on figuring out how to make these things work. The AVs need to work in this operating environment, which is the city, an operating environment the city controls. So finding ways that the cities and these companies can work together, actually makes a whole lot of sense and can be helpful to both.

Miller: Ten years ago or so, it was common for people who are thinking about the future of cities and the future of transportation to say that now, in 2024, there would be tens of thousands or millions of these cars on the road. If that obviously hasn’t come to pass, what are those same future-oriented urbanists predicting for the future now?

Larco: You get a mix. I mean, some people are still very much, I’d say, boosterish, about this technology. It’s going to be everywhere very soon. We’re just kind of hitting a few bumps in the road. I think others have started to understand that this technology is probably going to exist. It’s going to continue. It does have benefits, absolutely. But that maybe the use cases are not as widespread as we’d originally thought it might be, or that the companies might have thought it would be. There might be more limited places where they make sense, both in terms of geography and in terms of the kinds of trips or the kinds of vehicles that are going to be using the AV technology. So more, we’ll say, humility and thinking about how quickly this will happen or the extent of it, I think, makes a lot more sense.

Miller: Nico, thanks very much.

Larco: Thank you so much. If I could also mention, the report that we put out was done with the Knight Foundation and with Citify and really is a guide for cities to help them understand how it is that they might be able to approach these emerging technologies as they arrive in their cities. And hopefully it will be helpful to people as they think about these topics.

Miller: Nico Larco is a professor of architecture and urban design and the director of the Urbanism Next Center at the University of Oregon.

Contact “Think Out Loud®”

If you’d like to comment on any of the topics in this show or suggest a topic of your own, please get in touch with us on Facebook, send an email to thinkoutloud@opb.org, or you can leave a voicemail for us at 503-293-1983. The call-in phone number during the noon hour is 888-665-5865.