Pittsburgh Reveals the Downside of Self-Driving Cars

A post-crash photo of the self driving Uber car that killed a Tempe pedestrian. Photo:  NTSB
A post-crash photo of the self driving Uber car that killed a Tempe pedestrian. Photo: NTSB

Wider inequality. More sprawl. Worse transit. Those are some of the outcomes we could see from self-driving cars in cities, according to a Pittsburgh-based transit advocacy group.

In its new report, Pittsburghers for Public Transit argues that public concerns about equity, the environment and job security aren’t playing a larger role in the conversation about autonomous vehicle deployment.

“The introduction of AV is presented as a panacea to our transportation, environmental and economic woes,” the report says, as it cautions about all that can go wrong for the public, especially lower-income people.

Pittsburgh has been a key testing ground for the technology. With the support of Mayor Bill Peduto, Steel City is currently allowing five companies to test driverless vehicles on public roads. The public has been exposed to risks associated with being guinea pigs in an AV lab, yet not a single public meeting has been held to address public concerns, says PPT.

“The hype from the industry is really dominating the discussion,” said Laura Weins, director of the group. “We have literally no regulatory framework on this. They just do whatever they want and use our public right of ways.”

Pittsburgh does ask the companies to abide by a voluntary agree, but there is no enforcement mechanism. (This is very similar to the way federal regulators are handling the issue.)

Their report questions some of the public benefits of AVs. Here’s a few of the possibilities they are raising in the meantime:

Less transit and more sprawl

Proponents of self-driving cars have argued that they could expand access by reducing operating costs (by eliminating drivers), and the savings could be used to expand service to currently underserved areas. But there are good reasons to be skeptical, says the report.

“The lack of transit in underserved communities often results primarily from a lack of political will to prioritize mobility solutions for underserved areas,” not from lack of money, says PPT.

In addition, experiments with small vehicles providing door-to-door service — “micro-transit” — have been disappointing in the U.S. “These pilot projects saw unsustainably high costs per rider and low ridership (less than four boardings an hour).”

Self-driving cars could help produce an environment that is hostile to transit overall.

“Unregulated AV adoption potentially worsens urban sprawl and increases consumer appetite for personal transportation,” PPT writes.

Because this technology lowers the real costs of driving by freeing passengers for other activities, researchers have predicted it will increase driving miles and promote commuting to distant locations in sprawling or rural.

The group also says AVs “create the possibility of new forms of social segregation.”

Self-driving cars will be expensive, at least for a while. And those who can’t afford them “will experience a decrease rather than an increase in transportation options, fragmenting communities based on their ability to purchase AV.”

Bike and pedestrian safety risks

IN the first pedestrian death by a self driving car last year, the self-driving system detected the victim six seconds before impact, but Uber had tuned the emergency braking feature to be too insensitive to respond in time. Image: NTSB
In the first pedestrian death by a self driving car last year, the self-driving system detected the victim six seconds before impact, but Uber had tuned the emergency braking feature to be too insensitive to respond in time. Image: NTSB

Finally, AV testing presents risks to pedestrians and cyclists because automated driving tech still isn’t perfect. We saw the worst of it when a pedestrian was killed in Tempe, Arizona by a self-driving Uber car last year.

In addition, pedestrians and cyclists risk losing funding for essential infrastructure like sidewalks and crosswalks to special street treatments for AVs. The city of Pittsburgh, for example, allocated $23 million from its “Department of Mobility and Infrastructure” for testing and deployment of an “autonomous micro transit shuttle,” which Weins, the PPT director, called “not really mass transit.”

“In these same communities, residents have called for better sidewalks, crosswalks, dedicated bus and bike infrastructure and expanded transit service to encourage safe and accessible transportation,” the group says.

Although AVs have been sold as a potential big safety improvement, there’s much more the city could be doing right now to reduce traffic fatalities, PPT says.

“Lower speed limits in cities, sidewalk bump-outs to shorten pedestrian crossing distances, and protected bike and bus infrastructure all exist as proven solutions for enhancing human safety,” they write. “Simply prioritizing buses over private vehicles as the form of mobility promoted by our infrastructure is an effective strategy to reduce the number of accidents on our roads.”

Green raw deal

The environmental benefits of AVs rest entirely on whether they are electrified. But almost all the credible research predicts they will increase driving miles without some regulation.

“AVs may introduce empty vehicle travel, which further increases vehicle miles traveled,” the report says. In other words, what’s worse for the environment than solo-occupancy cars? Zero occupancy cars.

“Creating a more resilient world requires a radical shift in our transportation practices away from reliance on personal vehicles, which AV technology does not inherently do, and could potentially worsen.”

Job losses

About 10 million Americans work as professional drivers. Studies have estimated three to four million jobs could be eliminated by the introduction of self-driving cars. And Pittsburghers for Public Transit says the options for these vulnerable workers are limited.

“The average education attainment of those workers is low, at 7.6 percent with bachelor’s degree or higher (compared to 33.4 percent for all occupations), and the average age of drivers is high, on average 52 years old,” the group writes.

Many Pittsburgh workers still have not recovered from the loss of steel jobs to automation and offshoring in the 1970s and 1980s, making the prospects for a “just transition” seem uncertain at best.

Pittsburghers for Public Transit is hosting conversations across city beginning on Thursday about the issue.

11 thoughts on Pittsburgh Reveals the Downside of Self-Driving Cars

  1. Curious why no one is mentioning adapting the technology to buses and trains. Everyone hates taking trains late at night because they come once an hour and lack of express service. Well, cost would be much lower if drivers and conductors weren’t needed.

    Imagine if buses could switch directions upon reaching the end rather than waiting for a union mandated 30 minute rest.

    Transit really needs to adopt autonomy or be a victim to it.

  2. Good points except for the job loss concern. Almost every technology that mechanizes or automates a task will result in a short term loss of the jobs. While the transition is stressful, society adapts and the lost jobs are replaced with new opportunities.

    Millions of farming jobs were lost when the diesel tractor came to the fields. Few people will say that we were better off tilling fields behind oxen. Computers wiped out millions of low level accounting and clerical jobs yet the people who would have filled those roles today are employed in other jobs, almost all of which use computers.

    The same transition will occur when human drivers are made obsolete by robot drivers.

  3. Headline – PIttsburgh “reveals” downsides…

    Actual article – A lobbying group says that…

    Also @Martin – the reason you don’t see self-driving transit investments is that there are large unions that would prevent any US city from adopting self driving cars.

  4. Taxpayers would be better off if some drivers were cross trained to service autonomous buses which still provide same number of jobs, but also provide more frequent service

  5. My biggest concern with self driving cars is the lack of acknowledgement between peds/cyclists and the AI.

    Seeing that a driver sees me while I’m walking or biking is usually the only way I’ll cross in front of their car. With AVs, there’s no feedback indicating the AI detects people in the street.

    This issue will most likely compound pedestrian and cyclist discomfort on streets which are already littered with distracted SUV drivers.

  6. @Martin – while I agree, AV for transit would be a game changer. That doesn’t mean we don’t need human operators on transit. But I’d like to see people doing more customer facing work. Like concierges for transit would be great, and improve the user experience.

  7. @Mark – not only do they hardly recognize bicyclists and pedestrians in general – recognition is even worse for darker people of color. Who have been rendered invisible in basically all the training algorithms. I am worried I won’t even be seen at all. 🙁

  8. @Jamie,
    I think it’s clear that public transit unions are not interested in customer facing work. WHere as in other countries, you see people near fare gates helping out, US workers prefer to sit locked up in a booth. Given that preference, I see little reason to hold up better service from autonomy.

  9. With the ethanol in gasoline (15%?), wouldn’t the automobiles be impaired while they self-drive?


  10. Advocates of self driving vehicles say more cars will be able to travel on a given area of road. That means an endless stream of cars a couple feet apart. An absolute nightmare for anyone who isn’t driving.

    Since these cars supposedly stop automatically for obstacles, cities may have to even further restrict pedestrian and other non vehicle traffic. This will kill cities everywhere.

    It’ll be so easy to send an empty car to pick somebody or something up, lots more trips will be made than we have today with regular cars. Not good.

    Another point: will the cars be programmed to minimize costs by driving around between jobs instead of paying to park? That would also be awful.

    Pittsburgh and other cities should charge meaningful congestion fees based on where and how long these cars are driving, with an even higher fee if the cars are driving around empty as they could be if sent on an errand to pick something up.

  11. Self driving vans or buses need some sort of track system in case the computer mess up. This lane should restricted to self driving and city and emergency cars and trucks only where possible. A new subdivision would to have a 2 14 drive lanes one in each direction. A 11 foot center self driving lane one to be shared for both directions and for left corner turning no parking within 40 feet of intersections and a 40 inch bike lane in each direction plus a 10 foot commercial parking lane on one side of the street with a minimum of 4 floors apartment buildings to get the density of people needed for this level of transport ideal city blocks of 800 feet by 400 feet

Leave a Reply

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


Transit Opponents Fight Ballot Measures With the Language of Tech Futurism

Today’s a huge day for transit on local ballots. Ballot measures in Detroit, Indianapolis, Raleigh, Seattle, and Los Angeles, among other places, could significantly expand transit access for millions of people. In many of these cities, transit opponents have seized on the same feeble argument: Investment in transit isn’t needed because soon Uber and Lyft and self-driving cars will […]

St. Louis “Beat Congestion” and Now Commute Times Are Longer

St. Louis is every highway planner’s dream. Consistently ranked among the least-congested cities in America, the region’s car commuters spend a smaller share of their trips to work sitting in traffic than all but two other cities. That means St. Louis car commuters aren’t encumbered much by other car commuters, just like in those car commercials. But […]

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 […]