It’s an Automatic – Next City

Help us reach 40 new sustainers in September.

9
40 goal

It’s an Automatic

The Road to a Future of Driverless Cars, Dense Streets and Supreme Mobility

Story by Jonathan Geeting

Illustrations by Tim Pacific

Published on Feb 10, 2014

Before the driverless car, there was the carpool. A 20th-century mechanism for car sharing, carpooling was a way for children, coworkers and those without ready access to a vehicle to nod off in the back seat while someone else drove. Maybe it was a friend’s mom. Maybe it was the guy in the cubicle next to you. No matter. What was important was you got where you needed to go and shared the bill on the ride.

Now Google driverless car and countless fantastical imaginings are popping up, many of them spawned from the corporate brain of the search engine wunderkind itself. Once you get past the magazine-ready photos of antenna-topped Priuses, there are Jetsons-esque drawings of bubble cars gliding down narrow future highways and illustrations of dense urban grids dotted with steering wheel-free taxis and zero parked cars.

Human Transit author Jarrett Walker refers to this as the “complete imagined future,” a place so far removed from lived experience that the scenarios residing there often fail to tell a plausible story about how they transcended political and economic constraints. Yet the story of driverless cars is anything but implausible. In the three years since Google cars were first spotted along Silicon Valley’s Highway 1, the technology has become one of the world’s most anticipated consumer products. The driverless, or more accurately, self-driving car is widely predicted to revolutionize mobility by knocking humans out of the driver’s seat as soon as 2030. The technology offers the possibility of infinitely safer travel. Human error — a mistimed turn, a heavy foot on the gas pedal or any one of countless other driver mishaps — caused or contributed to more than 90 percent of car collisions, according to a landmark study done by Indiana University. With automated acceleration, braking technologies and crash-avoidance technology, driverless cars could make highways exponentially less deadly.

Yet there is another opportunity at stake: The chance to dramatically reshape the relationship between public space and the car. For the last 100 years, urban planners have designed cities to accommodate personal vehicles. Every home comes with a driveway or curb for your car. Asphalt seas of parking spaces or costly multistory garages surround schools, shops and office buildings like carbon-spewing moats. What if instead of driving our own cars, we relied on 21st-century carpools — sharable autonomous vehicles?

Recent studies suggest that car use is on the wane in the U.S., with or without driverless vehicles. One study by Michael Sivak at the University of Michigan’s Transportation Research Institute found that driving rates for passenger vehicles have been on a downward slope since 2004. Other studies by the U.S. Public Interest Research Group have confirmed this, with evidence that young people are mainly responsible and that barely any statistical relationship exists between reduced driving and high unemployment rates. While the automobile industry has been quick to blame car use decline on the recession and high gas prices, available evidence suggests that economic hardship is not behind the trend.

Sivak’s hypothesis is that we are witnessing a larger structural change driven by factors like “increased telecommuting, increased use of public transportation, increased urbanization of the population, and changes in the age composition of drivers.” Indeed, the average American car sits unused 23 out of 24 hours a day. Is there any other contemporary technology we spend so much money not to use?

If these vehicles are perceived to contain the seeds of the auto industry’s long-run stagnation, it may fall to Google to finish the job.

Last August, Google invested $258 million in Uber, an app-based ride service company that allows people to order black cars online, pay for them using their smart phones, and track their arrival. It was Google’s largest-ever venture capital investment. As the company continues to develop the technology to pilot drone taxis, it also, presumably, wants to popularize the interface for hailing them, and develop the algorithms for dispatching the fleet. With automated vehicles in the picture, services like Uber, already quickly growing in popularity, could become significantly less expensive and thus accessible to more people. Already, researchers at Columbia University’s Earth Institute have estimated that a fleet of shared autonomous vehicles, providing the same level of service as New York’s yellow taxicab fleet, could slash the cost from $4 per trip mile down to 50 cents per trip mile — an eightfold cost reduction.

There is much to recommend in this imagined world of cheap, ubiquitous drone taxis. This is a world where senior citizens, the disabled, young people and the inebriated can enjoy safer mobility and be freed from the enormous costs of owning and insuring a car. It’s also a world wherein parked cars have a smaller footprint. In 2010, researchers at the University of California, Berkley released a study on the country’s parking infrastructure. While conceding that assessments of the number of parking paces have varied significantly, they cited a series of estimates, from which we can draw a reasonable range. Following this data, current parking codes have made it so there are between three and eight parking spaces for each of the 250 million cars in America. Think about it this way: If the number of personal cars dropped and parking provisions were loosened, as many as 675 million parking spaces, particularly those in urban cores, could be turned into something else — housing, parks or a million other uses more desirable than street-deadening parking lots.

“If you could push all the land currently devoted to parking in Manhattan and Boston and other big cities 10 minutes outside the city limits,” says Anthony Townsend, author of Smart Cities, “you’re potentially talking about trillions of dollars worth of real estate that could be developed, and tens of millions of new city-dwellers who could be accommodated.”

The Eno Center for Transportation, a think tank in Washington, estimates that, at 90 percent market penetration, autonomous vehicle (AV) technology could save the country $201 billion annually and a whopping $447 billion in comprehensive benefits from savings related to reduced congestion and fewer crashes. Eno predicates its estimates on the expectation that a single shared autonomous vehicle could make five times as many trips as a non-shared vehicle. The center assumes 10 percent of AVs will be shared.

But the legal replacement of human drivers is by no means a given, as the millions of workers now in the business of transporting people and goods are unlikely to back down quietly. Indeed, whether fully autonomous vehicles are ultimately ever allowed to come to market will hinge on a high-stakes, behind-the-scenes political struggle just now getting underway.

Regulating the Robot Car

Last fall, Congress held its first-ever hearing on autonomous vehicles. Closely watched by the auto and insurance industries, the next-day headlines focused on a 2020 delivery date for driverless cars announced by Nissan executives and more uncertain timelines from other carmakers, notably GM. While the corporate countdowns made for good headlines, the bigger takeaway was lawmakers’ mounting anxiety over the nascent technology, still unregulated and largely untested by government. From D.C. Delegate Eleanor Holmes Norton’s opening comments about a fatal 2009 failure of the D.C. Metro’s automatic train control system, to later comments aired by Rep. Albio Sires about the danger of roboticizing driver jobs, lawmakers gave clear signals of the rocky regulatory road ahead.

“I was not happy with your testimony when you talked about [how] it might put fewer vehicles on the road,” joked Rep. Roger Williams, who owns an auto dealership, in response to a suggestion by the Eno Center’s Joshua Schank that shared autonomous vehicles might reduce car ownership. The interests of favored district businesses like auto dealers, as well as personal financial interests, will influence lawmakers’ tolerance for disruption of the status quo.

The hearing in Washington came about 10 months after California Gov. Jerry Brown stepped into a white Toyota Prius and hitched a ride to work with Google co-founder Sergey Brin. Neither man drove. The two took a self-driving car to Google’s Mountain View headquarters to sign the first piece of California legislation addressing automated vehicles. The landmark law signed that morning allows Department of Motor Vehicle-approved AVs to hit California streets by 2015.

“This self-driving car,” Brown said, “is another step forward in the long march of California pioneering the future.”

In the past two years, Nevada, Florida and Michigan, along with California, have put state laws on the books regulating the testing of AVs on public roads. By and large, these laws established special licensing and liability rules, with each state requiring a human driver (or two, in Nevada’s case) in the car at all times.

The developing mishmash of state laws has Washington scrambling. As evidenced by the hearing in the fall, lawmakers on all sides of the issue are eager to develop standards and get answers to their constituents’ questions. In May, the U.S. Department of Transportation’s National Highway Traffic Safety Administration released its first set of policy recommendations on the nascent technology. The non-binding recommendations, directed at state transportation agencies, said that driverless cars should not yet be allowed on the open road, but that testing should. The recommendations went on to acknowledge “enormous safety potential” and other benefits to AVs, including reduced fuel consumption and commuting times.

“To help ensure that these economic, environmental, mobility, and safety benefits are more likely to emerge from the current streams of innovation, all interested parties need to work cooperatively,” the recommendation stated.

The specifics of what cooperation looks like to U.S. regulators is what keeps carmakers like Volvo awake at night. The Swedish carmaker plans to do the world’s first large-scale test of driverless cars in 2017, putting 100 on the road. For a venue, the company chose its native Gothenburg. It is working with the Swedish government, a collaboration that Volvo has said comes out of a common interest in reducing traffic fatalities. In the U.S. there has been no such collaboration, and if current patterns continue it could end up less hospitable to the new cars than European nations that have already embraced the technology.

“In terms of regulatory hurdles, we are very concerned about a state-by-state approach that would lead to a patchwork of varying state laws,” says Katherine Yehl, Volvo’s director of government affairs for North America. “There has to be some kind of federal overlay because transportation falls under inter-state commerce.”

The feds use a classification system to define what exactly constitutes an autonomous vehicle — like those being piloted in Google’s backyard — and what is just a regular car with cushy auto-enhanced capabilities, like lane centering controls and automatic braking. Levels 0 and 1 refer to different types and combinations of features, but the really juicy political questions lie in the distinctions between Level 2, 3 and 4 vehicles.

With Level 2 vehicles, the car’s primary functions, like braking or steering, can be shared between the system and the driver, but also allow the driver to turn over “active primary control.” At this level, the driver is still expected monitor the road and be ready to take the wheel at a moment’s notice. Level 3 vehicles are capable of fully automated driving, but require a human in the driver’s seat who can take over in tricky situations. In those instances, the automated system senses the unusual navigating demand and signals that the driver will need to take over soon.

Level 4 cars are completely self-reliant, and do not require human monitoring or intervention at all. In its recommendation, the National Highway Traffic Safety Administration reminds readers that for now, all state laws stipulate that a human be present in these vehicles on public roads. (Of course, it will be up to the states to decide what happens next, unless the feds act first.)

In a world where shared-authority vehicles dominate the streets, all we really get are marginally safer highways and faster traffic speeds. Humans ultimately still pilot taxis and delivery trucks, and nothing jolts car ownership trends off their current path. But if Level 4 cars are given the green light to ride driver-free, infinite unknowable changes, large and small, could reshape the economy and built environment as intelligent software replaces human drivers across the transportation sector.

But while the federal government’s AV policy has set up these classifications with thorny regulatory issues in mind, it stops short of tipping a hand in how these rules will be crafted. And for good reason — things are likely to get ugly. Politicians on state and federal levels could face intense pressure from drivers unions and other auto-stakeholder industries to permanently delay legal deployment of Level 4 vehicles.

Volvo’s worst nightmare would be if every state adopted somewhat different safety and liability standards, throwing a wrench into their efforts to develop a single product that can be sold nationwide. Liability laws present a particularly potent threat. While today’s collision lawsuit awards are often fairly puny, limited by human drivers’ ability to pay, that will change when deep-pocketed owners of autonomous software — the automakers, or Google — are ultimately to blame for collisions.

Google’s Anthony Lavandowski has predicted that automakers won’t bring full AVs to market under current liability laws, and added that the company envisions the possibility of one day retrofitting regular cars with autonomous conversions. A KPMG study found that consumers are already primed to trust Google more than the major automakers to deliver a safe AV. It’s an open question as to how much legal exposure Google’s shareholders would tolerate, but in the medium run, liability remains one of the most powerful levers for the political foes of automation.

In the long run, automakers are likely in no rush to move beyond shared-authority vehicles, and arguably have every reason to oppose drone cars now classed as Level 4. Since level 2 and 3 cars would always require a human in the driver’s seat, they would do little to nothing to change the norm of one car per person. Only level 4 would enable the type of widespread carpooling that could gut auto sales and reshape the built environment in ways that diminish the usefulness of cars in general. If these vehicles are perceived to contain the seeds of the auto industry’s long-run stagnation, it may fall to Google to finish the job.

Undoing Parking Lot Nation

When a young Ian Lockwood arrived in West Palm Beach, Fla. in 1996 to serve as transportation engineer, he found a downtown that been hollowed out. The city’s main drag, Clematis Street, was struggling with suburban flight and vacancy. In 1993, according to the Institute of Transportation Engineers, only 30 percent of building space on Clematis was occupied. Problems were manifold for the once-idealized downtown. But for Lockwood, one problem that could be easily engineered away had to do with the streets.

“In West Palm Beach, we tore down about half of the buildings in the downtown to make room for surface parking lots. It looked like there’d been a good carpet bombing during the blitz of Dusseldorf in the Second World War,” jokes Lockwood, now a principal at multidisciplinary design firm AECOM. “We obliterated our downtown so people could drive in quickly from the suburbs and park their cars. The damage that was done to the city was catastrophic, but was presumed to be perfectly acceptable to the people in those cars.”

Others in the city agreed that something had to be done. Documentarians had chosen West Palm Beach as the setting for a 1989 film about drug abuse called Crack USA: County Under Siege, and the city’s reputation was suffering. In 1992, the city council approved a $12 million bond issue for downtown revitalization. It enlisted Andres Duany and Elizabeth Plater-Zyberk, the husband-and-wife team known as leaders in the New Urbanist movement, to lead the creation of a new master plan. Duany and Plater-Zyberk, together with a team of planners and architects, drafted a new design for the city with input from the mayor, local politicians and community leaders, changing zoning ordinances and making way for denser development.

With the political backing of Mayor Nancy Graham and overall goal of a more vibrant downtown, Lockwood and others set out to reverse the damage wrought by auto-centric planning by doing what was, at the time, unthinkable — grabbing street space from cars and giving it back to pedestrians.

The city went on a “road diet.” Roads were narrowed, curbs bumped out at intersections, roundabouts installed, and buffers of curb parking restored between travel lanes and sidewalks. The goal was to drape a big “slow blanket,” as Lockwood puts it, over the entire city.

To do this, the city had to get an exemption from Florida’s “concurrency” law, which effectively requires all local plans for public land use to adhere to state regulations. Without the exemption, it would’ve been illegal for the Graham administration to allow any new buildings even in the sparsely populated downtown, and especially on the scale that would eventually be built.

In the beginning, Lockwood says, local transportation engineers — to say nothing of suburban commuters and politicians — were resistant to these ideas, having been trained conventionally to create streets that maximize car travel speeds, even though it was clear that the reigning approach had weakened the city.

“The idea of taking away car-carrying capacity was un-American at the time. It violated this emotion that a lot of people shared that equated car capacity with freedom and independence, in a really skewed way,” Lockwood says. “A few of the leaders at the county summoned me to their offices to say ‘you gotta stop the madness.’ They thought it was crazy what West Palm Beach was doing.”

The county had no jurisdiction over city streets and property, so Lockwood went ahead and narrowed the roads. He says congestion today is no different than congestion before the road diets, but the city is much healthier.

In the two decades since the initiatives for downtown renewal began, the city has seen remarkable private retail investment. By 1999, according to the Institute of Transportation Engineers, occupancy on the city’s main street had increased to more than 80 percent. Raphael Clemente, executive director of West Palm Beach’s Downtown Development Authority, says that 100 new businesses have opened in the downtown district in the last four years and that occupancy on Clematis Street is higher than 90 percent today. The city is virtually unrecognizable from the place that once served as the face of Crack USA.

But getting there was not easy. Across the country, state and local policies are lodged in place by that politically bulletproof combination of inertia and powerful interest group politics. Despite a full decade of declines in driving, few zoning codes outside a handful of major cities have become less obsequious to the prerogatives of motorists.

“Urban planners have known for a long time that off-street parking requirements are stupid and that they cause great harm to our cities,” says Alan Durning, executive director of the Sightline Institute, a Seattle-based sustainability think tank. “But they haven’t had a prescription that had a chance in hell of actually changing the political constituencies supporting parking requirements.”

What happens if autonomous taxis collapse the demand for parking, but decades-old zoning codes keep cruising on autopilot?

Analysis of data from a 1997 study by Mark Delucchi of the University of California, still considered one of the best comprehensive evaluations of the costs of automobile use in the U.S., reveals that motorists pay just 1 to 4 percent of the cost of the parking supply in user fees. The rest of those costs are hidden in the prices for everything else. The Sightline Institute recently found that off-street parking requirements cost Seattle renters $246 a month on average.

In theory, shared autonomous vehicles present an opportunity to undermine the political foundations of these secret parking taxes and other anti-density land use controls, and facilitate more plentiful construction of cheap, multi-family housing. But as Durning warns, even large shifts in transportation preferences won’t make the parking subsidies vanish automatically.

“There are a number of European cities, like Stockholm and Vienna, where U.S. urbanists would give their eye teeth to adopt their transportation policies,” he says, “But those cities actually still require off-street parking.”

Pro-parking policy is a tough nut to crack. While some believe that simply improving the convenience of transit and on-demand mobility options like autonomous cars will organically lead to an evolution of parking policies, politicians will have to step up if we want real change to happen. Adding to the complexity of the land use question is another potential unanticipated consequence of adding driverless cars to the urban transportation mix — more crowded streets.

Even if the total number of vehicles dropped, more car sharing could mean more cars in motion at any given time. By the same token, development on land now reserved for parking could mean more densely populated streets, intensifying debates over street space as a flashpoint in city politics.

“Congestion’s not going away,” Lockwood predicts. “This is something you’re going to hear a lot of people say — that Google cars can get five times more cars through intersections than now, since they’ll drive closer together. How’s that going to feel for pedestrians and cyclists? It’s going to be like a machine gun spray of cars down the street.”

The nature of congestion may be different and less painful for those inside the cars, as vehicle-to-vehicle communication negotiates the traffic flow. But for those outside, Google’s claim that it could double road capacity, as reported in the New York Times, could mean trading in parking lots for streets that feel like them, with bumper-to-bumper traffic stewing in the open road.

Yet as always, questions about the allocation of public space and the visual pollution of automobiles and their accessory uses aren’t technical, but political. The option to speed cars through the city will still have to be balanced against service levels for transit, cyclists and pedestrians. Autonomous vehicles may increase street capacity by obviating the need for many curbside parking lanes, but should that space automatically go to cars?

Transportation consultant Jarrett Walker argues that transit, which has also been targeted for automation, will still be necessary and competitive with drone taxis in the driverless city, due to this intensifying competition for scarce road space. “Driverless transit (in fixed separated guideways, mostly rail) will keep growing anyway, because it’s the only way to deliver extreme frequency and capacity cost-effectively,” he says.

Well-defined rights of way will be crucial for the viability of autonomous transit because, like a gas expanding to fill whatever sized container it’s given, AV traffic could otherwise inundate every space it’s not specifically forbidden.

As ever, the amount of public space allocated to different modes will depend on the relative status of their users. Will transit options continue to win exclusive rights of way, or will an exodus of wealthy transit riders to drone taxis restore transit’s stigma as the modal redoubt of low-income others, weakening its claim on precious street space?

Then there is the biggest question of all: Will the rise of the self-driving car break or boost the great American tradition of sprawl?

What happens if autonomous taxis collapse the demand for parking, but decades-old zoning codes keep cruising on autopilot?

Suburban growth is inextricably linked to the personal automobile. There are signs that as love for the car fades, so goes affection for the cul-de-sac culture it enabled. Take last year’s Consumer Preference Survey from the National Association of Realtors. Sixty percent of respondents favored a neighborhood with a mix of houses and stores over one requiring more driving between home, work and recreation. Yet most respondents also said a detached home was most important, with 57 percent reporting they’d prefer a detached single family home with a longer commute over an apartment or condo with a shorter commute. With real estate in desirable urban areas more expensive on average than suburban square footage, the logical choice for most Americans is to choose to live outside of the city. How would the driverless car reshape the scenario?

Well, if the self-driving auto and its ripple effects on the built environment loosen the market for urban real estate and encourage people to give up owning their own car, the costs of living in the city could drop considerably and make the move into town worth it for many families. Alternatively, if the time and sanity costs of commuting fell, that could intensify the appeal of suburban housing, especially if combined with faster and cheaper delivery options for food and retail purchases. Again, it will be up to policymakers — as well as the desires and lifestyles of the people of 2030 — to determine which scenario takes hold.

“You can make huge changes to quality of life and health and so forth through policy,” Lockwood says, “People here don’t like that idea because they think of it as social engineering. But we were socially engineered to give up our transit and our walkable cities. Public policy was changed purposefully to dismantle that system, not just because it was profitable, but because there was this idea that a car-based system was a better way of life. If we still have that idea when driverless cars come along, there’s going to be a lot of driving going on.”

If, that is, self-driving cars are allowed to enter the system.

The Labor Angle

The glummest voice at the Congressional subcommittee hearing on autonomous vehicles last fall was Albio Sires, the Democrat who fretted that an AV revolution would mean job losses for mechanics. “I can’t imagine anybody doing any work on these cars that are so sophisticated,” he said. “To me, I think that’s just going to put people out of work.” Job protection policies are nothing new in Sires’ native New Jersey, where motorists are still forbidden from pumping their own gas for the express purpose of conserving redundant gas station attendant jobs.

The traditional view among economists that automated jobs are replaced manifold in the long run sounds to Sires and a great many voters like glib magical thinking. In fairness, some economists have recently challenged the fallaciousness of what’s been called the Luddite Fallacy, in light of the weakened connection between productivity growth and employment growth observed since the early 2000s. Included in this skeptical bunch are drivers in the trucking and delivery industries. The advent of the drone truck could easily threaten the livelihood for millions in a sector that is one of the few remaining bastions of good-paying blue collar jobs.

Barring an unexpected breakthrough, autonomous vehicles are not expected to become affordable for household ownership for more than a decade, but commercial ownership is a different story. Even at $100,000 — the upper estimate of the Google car’s autonomous platform — AVs will become cost-effective investments for certain business uses, like delivery vehicles, long before they become affordable for households. Ümit Özguner of Ohio State University’s Center for Automotive Research told TechHive that the first AVs to hit the road would most likely be long-haul trucks.

Already, AVs can perform the essential tasks of a long-haul trucker — driving on the highway, exiting at interchanges, and navigating wide local arterials to nearby industrial parks. Drone trucks would offer major advantages over human drivers. They could drive through the night between cities without getting sleepy, which would translate into big savings for households and businesses. Within cities, deploying larger fleets of smaller vehicles (think electric vans, rickshaws and even flying drones) without needing to hire more drivers would further slash direct delivery times.

Strong union representation from the Teamsters, combined with dangerous working conditions and high turnover, give truckers considerable power to negotiate more generous wages and benefits. Labor costs eat up 61 percent of the budget at unionized UPS and 43 percent at non-union FedEx. The highest cost of all may be dead time. A controversial Federal Motor Carrier Safety Administration (FMCSA) rule went into effect this summer limiting truckers to 70 hours of driving per week. Even that was too much for safety advocates, with good reason, but that’s less than half the 160 hours a week or more a drone truck would happily drive for no pay.

Despite much speculation that organizations like the Teamsters, transit unions and the freight industry might begin lobbying state lawmakers for protections from automation, these efforts have so far failed to materialize. (The Teamsters declined to comment for this story.) Even though U.S. labor unions aren’t actively lobbying yet, a look at some related labor disputes provides a window into how the fight over automated shipping could play out, and why state regulations requiring human chaperones in AVs would be so consequential.

Trains running without a uniformed conductor are a rarity in the U.S., but the technology has been in use for decades in quite a few European and Asian cities. Its use does not, however, come without controversy. Transit unions have put up fights against fully automatized systems, playing on public safety concerns to keep human drivers on trains.

A number of incidents have reinforced these fears, one of the most frightening which occurred in March 2012 at the Finchley Road station in London. There, a driver rescued a child who had fallen between the train and the platform. The automatic system didn’t detect the child and gave the signal to start moving again, when the driver saw a tiny hand reach up from beneath the train on a final safety check.

Arguably the lesson is that a fully autonomous system requires platform doors to prevent riders from approaching the tracks before trains arrive, but a rescued child obliterates these kinds of technocratic arguments in politics, and the episode has provided valuable negotiating fodder for unions.

Transit unions in the U.S. aren’t facing automation pressures yet, but some analysts like Jarrett Walker believe developments like driverless cars will lead more people to ask, why not trains?

“I think it’s more likely that driverless cars will make it easier to insist on driverless transit, given equal technology,” Walker says. “France is one of the most union-friendly countries in the world, but they’re still building driverless metro and converting an existing metro line to driverless.”

The Paris Métro’s autonomous Line 14 faced scant labor opposition, and the recently completed conversion of Line 1 to fully automated status was also notably light on labor strife. That’s possibly because rather than lay off Line 1’s 250 workers and redistribute the savings to riders as lower fares, Métro promoted 40 workers and reassigned the rest to other lines. By contrast, the London Underground still has human workers on all lines, even though its first automated train debuted in 1967, largely because unions have rejected the idea of shifting people around.

“Google cars can get five times more cars through intersections than now, since they’ll drive closer together. How’s that going to feel for pedestrians and cyclists? It’s going to be like a machine gun spray of cars down the street.”

It will likely be a heavier lift in the U.S. In my home region of Philadelphia, 19th-century ticket punching technology prevails on SEPTA’s regional rail network, not because no one is aware of the money-saving potential of proof-of-payment ticketing, but because the United Transportation Union Local 61 is hyper-aware that ticketing reforms and level boarding platforms would mean the end of the line for 396 conductors.

As much as the unions may not buy this, there is evidence that job loss fears could be unfounded, as long as the right employment replacement strategies are put in place. In a comprehensive analysis of the economic literature on productivity and employment, published in 1994, the Organisation for Economic Co-operation and Development (OECD) determined that technology can create new jobs as fast as it kills old ones.

“Historically, the income-generating effects of new technologies have proved more powerful than the labor-displacing effects: technological progress has been accompanied not only by higher output and productivity, but also by higher overall employment,” the organization wrote.

But for the 3.8 million people whose jobs Level 4 automation could send the way of the telephone operator, including truck and taxi drivers, subway and streetcar operators, train engineers, and material movers, our advances come at the expense of great personal tragedy. At its core, the debate over autonomous vehicles is a zero-sum conflict between those who drive for a living and the growing ranks of car-free Americans.

The Mobility Revolution

My grandmother, Katherine Pochodylo, lived a five-minute walk from a Giant supermarket in a small apartment in Palmer Township, Penn., a second-ring suburb of historic downtown Easton. Though she was in good health into her early 80s and loved to go for walks around the neighborhood, she still needed my mother to drive her to the supermarket each week to do her grocery shopping — a half-mile away across a busy four-lane road.

She never learned how to drive, and after my grandfather passed away much earlier in her life she became heavily dependent on my mother for running errands. Sometimes she would walk to the supermarket by herself to pick up miscellaneous items, but schlepping home several days’ worth of groceries on her own would be unthinkable.

My grandmother’s inability to drive and the layout of her neighborhood put real limitations on her independence. Until driverless cars entered the conversation, that same fate loomed for hundreds of millions of Americans in suburban and urban areas.

“People for whom it’s a physical challenge to wait on a corner for a bus or taxi, being able to call up an inexpensive vehicle that pulls up right to your building, takes you where you want to go, and then goes off to pick up the next passenger, the potential is really quite amazing,” says Jody Holzman, AARP’s senior vice president for thought leadership.

Ever since Google’s 2010 announcement, AARP has heralded AVs’ potential as a life-changing invention that will allow retirees to age in place in their homes, and extend their independence well beyond the time when they are unable to safely operate a car. Even in cities with good transit options, driverless cars would potentially provide a more affordable and reliable alternative to services like Meals on Wheels or taxicabs.

“I think they would really augment the push for livable communities,” Holtzman says, “particularly in places that don’t have the greatest public transportation systems.”

For now, AARP is taking a wait-and-see approach.

“Will it be slowed down? Will the regulations say no driver is required in this case, but a driver is required in that case? We don’t know,” Holtzman says, “but once the genie’s out of the bottle, it’s not going back in.”

Our features are made possible with generous support from The Ford Foundation.

Jonathan Geeting is a freelance writer based in Philadelphia, where he writes about land use and public space politics. His work appears at Next City, This Old City and Keystone Politics.

Tim Pacific is an award-winning graphic design student entering his senior year at Rutgers University in Camden. In addition to his schoolwork, Tim is an active freelance illustrator. His work can be seen in AIGA Philadelphia’s SPACE, which features a recently completed series of hand-lettered postcards. Among his design philosophies, Tim believes strongly that concept comes first and you should absolutely judge a book by its cover design.