Autonomous vehicles never speed, drive while intoxicated, nor text-and-drive, which gives them the potential to make vehicle transportation safer and more accessible. But seven experts say this high-tech solution to San Francisco’s transportation problems brings what transit-first cities are trying to avoid: More cars on the road.
“The main risk with AVs, whether privately owned or ‘robotaxis,’ is that their convenience seduces us into driving far more often,” wrote Carlo Ratti, a practicing architect and professor of urban technologies and planning at the Massachusetts Institute of Technology. Not unlike other ride-hailing vehicles, such as Uber and Lyft, he added.
Ratti and his colleagues at MIT’s Senseable Lab researched in 2014 how Uber and Lyft would impact the number of vehicles on the road. They predicted that 40 percent of New York City’s taxi fleet might be made redundant.
Instead, however, with cheaper car travel, many people opted for Uber or Lyft instead of the bus or subway.
In fact, a 2018 study by Schaller Consulting, a firm specializing in transportation policy, found that ride-hailing services — underwritten by huge investment from venture capital — were so cheap that they increased the mileage driven by cars across the city. That’s counter to what Ratti and his colleagues had predicted, and it led to more traffic and congestion.
This phenomenon is no different in San Francisco, which sees the highest number of ride-hailing trips per square mile than any other place in California, according to the San Francisco County Transportation Authority. The agency found that in California, between 2010 and 2016, ride-hailing vehicles accounted for 50 percent of the increase in congestion.
The phenomenon also moved some riders away from public transit, though just how many remains unclear. The pandemic and remote work continue to contribute to the city’s decline in transit ridership levels: As of October, Muni ridership levels are 68 percent of what they were before the pandemic.
Muni ridership is still far below pre-pandemic levels
Average weekday boardings, in 1000s
800
700
600
500
400
300
200
100
2017
2018
2019
2020
2021
2022
2023
Muni ridership is still far below
pre-pandemic levels
Average weekday boardings, in 1000s
800
600
400
200
2017
2019
2021
2023
Chart by Kelly Waldron. Source: SFMTA.
The impact of robotaxis remains unclear
Though some experts predict worsening congestion, others are unsure of the impact that robotaxis which, like Uber or Lyft, will send riders a car at the touch of a button, will have on city streets.
“There’s no parallel for automated vehicles,” said Billy Riggs, a professor of urban planning at University of San Francisco. With ride-hailing vehicles, everyone who had a car could suddenly become a taxi driver, Riggs said. That’s not the case with driverless cars that rely on expensive hardware, instead of an app.
“Automated vehicle companies can’t just introduce 1,000 vehicles into the market overnight,” Riggs said.
Though that is, in fact, what Cruise CEO Kyle Vogt vowed to do, saying cities like San Francisco could absorb several thousand vehicles — and quickly. “As for what it would take to blanket a city like San Francisco, our goal … is to make sure we ramp up manufacturing capacity,” Vogt said, in an earnings call for General Motors, Cruise’s parent company, in July.
In August, the California Public Utilities Commission gave autonomous vehicle companies Cruise and Waymo the right to operate driverless robo taxis 24/7 across San Francisco. A week later, the DMV asked Cruise to halve its fleet after “concerning incidents” involving the vehicles. The company cannot exceed 50 cars on the road during the day and 150 at night. Waymo, on the other hand, has announced an expansion of its fleet across the city.
Cruise and Waymo have pressed their sales pitch on the premise of safety: That driverless cars can get anyone from any point A to point B more safely than having a human behind the wheel.
“Humans are terrible drivers,” said Riggs. He said this technology brings opportunities to make vehicle travel safer. “The number one thing is that they obey the law, and go slow,” he said. And, to the annoyance of some other (human) drivers on the road, they will stop fully at every stop sign.
But any safety benefit we may see from autonomous vehicles is not here yet, said Giancarlo Valdetaro, who works as a Policy Outreach Associate at the advocacy organization Transportation for America.
Autonomous vehicles are still involved in accidents, and Valdatero stressed that emphasizing future safety detracts from the investments cities could make in road safety today, with methods that we know work, like speed limits.
As of August, the San Francisco Fire Department had registered 55 “unusual occurrence” reports documenting times when autonomous vehicles meandered into fire and emergency scenes.
Cars, not people, as the problem
But other experts emphasized that cars are what make streets more dangerous, even if people are not driving them.
Driverless cars push a flawed premise: that “you and I, as drivers, are the problem,” said Valdetaro.
Peter Norton, a professor of history at the University of Virginia who teaches engineering students about the social aspects of their technology, said the most logical solution may be to use automation technology for other forms of transit, namely trains. As many places already have — Canadian cities such as Montreal and Vancouver run fully automated trains.
The refinement of the programs that operate driverless cars, he said, means “unbelievable amounts of tech AI, sensors, radar processors … and then it will probably not even do very well,” he said. “Or, just put the thing on rails.”
Bruce Appleyard, an associate professor in urban planning at San Diego State University, echoed those concerns, saying that people, not cars, should come first.
He calls driverless cars “a considerable threat to the humanity of our streets” that might push cities to prioritize land for car use, for instance, or further criminalize jaywalking.
Do driverless cars benefit public transportation?
For all the opportunities and problems driverless vehicles may pose, each expert noted that their effect will boil down to the policies that regulate them — and whether those policies prioritize the public transit networks we already have.
“There tends to be a willingness to agree that robot driving will make public transportation obsolete,” said Norton, referring to how politicians view robotaxis. Not so much because people actually believe that, but because politicians try to appeal to suburban voters — and it’s easier to promise investment in roads rather than public transit, he said.
“The best way to achieve accessible, effective and eco-friendly urban transport is to go back to basics. Buses, subways, bicycles and walking are not only cleaner and more cost-effective but also more efficient than any innovations Silicon Valley has envisioned,” Ratti wrote.
It seems to me that the main benefit of these vehicles is to make a handful of people very rich.
Does anyone actually believe that these high-investment but low-profit vehicles, with hundreds of thousands of dollars worth of electronics on board, and tens of billions of dollars in R&D investment to recoup and pay off with a profit, will ever be affordable for the typical car buyer, let alone for the housebound seniors for whom we’re supposed to believe we must sacrifice our transit system?
Of course they won’t be affordable, and moreover, once they put human drivers out of work, the livery companies that can afford them will jack up their rates beyond the ability of anyone but a privilege urban tech elite to afford. Of course, everyone could then go back to public transportation, except that, beholden to the neo-liberal mandate of profitability for public services, it’ll all have been sold off, privatized, cannibalized, dug up, and paved over for more vacant condos for foreign money laundering.
Once again, well done, tech fraudster billionaire welfare pirate sociopaths! (golf clap, French kiss, etc).
No matter how you look at it, 1,000 more cars on the streets of San Francisco is hugely aggravating the traffic problem, whether or not those cars are safer. How can we get the City to nullify the PUC’s decision? Then there is the issue of the “refueling” stations!!
I’m not against these autonomous vehicles but I did see one that was blocking a right turn only lane with it’s blinker waiting to get in the next lane to the left. Apparently the ai thinks it’s okay to be in the wrong lane waiting to be in the correct lane. It must have infuriated people trying to turn right. It was that intersection of north bound duboce before market street–very busy stretch. The vehicle should have turned right and reconfigured it’s directions. These cars need more ai.
If I’m standing at a cold, dark, ominous-looking corner, I will still choose to ride the bus, right up until the moment when Next bus tells me it’s going to be a 45 minute wait.
Same. How many times have I missed the bus though, when the estimate changes (or 2, even 3 come in a row with a long gap)? This seems pretty easy to fix
After these corporations destroy public transportation (Uber’s expressed aim), you’ll no longer have that choice.
Sure, the AVs drive like your grandma, but there’s a positive tradeoff from fewer pedestrian and bike injuries. Is that worth it? To pedestrians and bicyclists, probably worth it.
They go into their cruise-control PR-speak and gloss over or ignore details. They’ve established “disruption” as something inherently good. They proclaim advantages of AV’s by making simplistic statements–like they’re safer than human-driven cars, assuming that all cars will be robo-cars and that they operate in a vacuum. But like Uber/Lyft, they ADD MORE vehicles to the streets, creating longer queues at lights and often drive empty. That WAYMO lot (in which a huge fire occurred) on 14th St. should be a huge affordable housing with no car parking and a frequent electric shuttle for residents who are unable to walk, take MUNI or bike/scoot.
They go into their cruise-control PR-speak and gloss over or ignore details. They’ve established “disruption” as something inherently good. They proclaim advantages of AV’s by making simplistic statements–like they’re safer than human-driven cars, assuming that all cars will be robo-cars and that they operate in a vacuum. But like Uber/Lyft, they ADD MORE vehicles to the streets, creating longer queues at lights and often drive empty. And they boast about it! That WAYMO lot (in which a huge fire occurred) on 14th St. should be a huge affordable housing with no car parking and a frequent electric shuttle for residents who are unable to walk, take MUNI or bike/scoot.
As long as Sf continues and expands giving street priority to public transit, I think the safety potential AVs have far outweighs any concern of their negative effects.Concerns about road congestion are moot if rads are configured to ensure they’re not congested for public transit. In fact congested roads are a incentive for people to use public transit. The negative sentiments I read in the article and comments are essentially promoting using your car as the default and go to for mobility in SF.
A driverless car stopped for me in Redwood City while I was walking across the street but much too slow to stop too close to me for comfort and seemed to keep slowly rolling forward
I was very creeped out
I don’t think the technology is there and I want the taxi to hear me and respond to me if I say something
If folks have money they aren’t going to wait for a bus or go out of their way to take public transportation.
Just like with regular taxis.
Was almost hit at Folsom and Precita jogging last night by a human driver who ran a red light (and then screamed “**** you” at me as if I was the problem), so glad to wake up to see ML acknowledge that human drivers are unsafe. (Reminder: they’ve killed two dozen this year, with so many other injuries that the city doesn’t bother to count.)
There’s definitely a huge potential congestion problem. I hope ML does a follow up on what world-class cities are doing to address congestion: congestion pricing that applies to all cars (regardless of who drives-owner, taxi/uber, AV) and funds mass transit. NYC is finally rolling it out, and SF should be beginning to plan for it too.
It seems that the argument against automated vehicles is that people will want to use them because they provide a convenient transportation solution. If that is the case, then the issue is not the cars, but the lackluster alternatives.
I am no fan of these cars, but as a bike rider, autonomas vehicles are much safer and more courteous than the crazy Lyft, Uber, and taxi drivers out there on the city streets. Make inner city public transportation free and intercity public transit much cheaper, and the problem is solved.
I have to respectfully disagree about the “courteous” part, if that term can be extended to machines: the complete lack of communication between “driver” and rider is really a problem for me. I can’t tell if the vehicle detects me, unlike most human drivers. There’s no eye contact, hand gestures (polite and impolite) etc to convey intent or understanding. There’s a large social element to driving, despite everyone being inside their cars.
Totally agree, if you are approaching a driverless car on a bicycle, you have no choice but to slow down and yield to it, because you’ve got no clue whether it will see you, or turn in front of you.
Except for the robo-cars that have trouble identifying bike lanes. This cyclist says no way!