Will Self-Driving Cars Usher in a Transportation Utopia or Dystopia?

Innovation visionaries say electric, self-driving, shared cars will soon revolutionize the way humans move about. But experts caution that unless this paradigm shift is guided by sound public policies, the future of transportation could lead to more pollution, more emissions, and more gridlock.

The news sounds almost too good to be true.

“How the U.S. Transportation System Can Save $1 Trillion, 2 Billion Barrels of Oil, and 1 Gigaton of Carbon Emissions Annually,” proclaimed the headline of an article published by the Rocky Mountain Institute, an environmentally minded, innovation-focused Boulder, Colorado think tank. The institute’s prescription is a technological trifecta: electric, autonomous, shared cars.

Propelled by the ongoing digitization of just about everything, notably including cars, the thinking trumpeted in that 2015 article has been percolating in the transportation sector for the last several years. All three components of this vision are already expanding. Sales of electric vehicles (EVs) are slowly growing and should increase greatly as EVs become cheaper to own than combustion-engine cars — something that Bank of America Merrill Lynch analysts believe will happen by 2024.

Automated cars, often referred to as “autonomous vehicles” (AVs) — whose passengers determine their routes without having to drive them — are being widely developed and tested, and probably will be used commercially in controlled settings within a few years. Lyft, Uber, and others have introduced ride-sharing, in which customers agree to travel with strangers in return for reduced fares. Put all three concepts together in one vehicle, posit that within a few decades this shared EV-AV technology will take over the nation’s automobile fleet, and the outcome seems environmentally irresistible, verging on fantastical.

Shared, electric, autonomous vehicles seem environmentally irresistible. But it’s equally plausible that the vision may turn out to be a mirage.

But it’s equally plausible that the vision may turn out to be a mirage. Automated vehicles may eventually be widely adopted, but if the fleet is not electrified using renewable energy, or car sharing fails to take off, greenhouse gas emissions and air pollution could actually increase. A study last year by University of California, Davis researchers projected that if vehicles are automated but not electrified or shared, greenhouse gas emissions from the transportation sector would go up 50 percent by 2050 compared to business as usual. But if shared, electrified, automated vehicles flourish, greenhouse gas emissions could plunge by 80 percent, the study concluded⁠.

“We’ve learned enough to know the future of [EVs and AVs] could be utopia or dystopia for the environment,” Christopher Grundler, then head of the U.S. Environmental Protection Agency’s Office of Transportation and Air Quality, said in 2016⁠.

Advocates of the shared EV-AV concept see single-file lines of self-driving trucks traveling down highways in “platoons,” each separated by no more than a few yards, achieving reductions in aerodynamic drag as high as 55 percent. that result in significantly lowered fuel consumption. Traffic congestion would fall away, as cars drive closely to one another without danger of colliding, circulate smoothly without aid of traffic lights and signs, and carry the number of passengers that matches their size. “Right-sizing” alone could cut vehicle energy consumption nearly in half⁠, and eliminating congestion could decrease it by another 2 to 4 percent⁠8.

As envisioned by champions of the shared EV-AV idea, passengers would take advantage of their liberation from driving by turning automated vehicles into moving offices, bedrooms, or dens, eliminating the wasted time of commuting. People with disabilities and the elderly would no longer be constrained by an inability to drive. Car accidents, which each year cause 1.2 million deaths globally⁠, would drop by as much as 90 percent⁠, since automated cars, unlike humans, don’t fall asleep, exchange texts and emails, or operate while drunk or drugged.

A demonstration of how Tesla’s self-driving “Autopilot” system works.

The need for parking places and parking garages would plummet, as driverless cars go somewhere else instead of parking. Because roughly 80 percent of toxic air pollutants are released in the first five minutes of driving, the resulting reduction in cold starts would cause a huge drop in air pollution. In a place like Los Angeles, now home to 18.6 million parking places — three per car — the decline of parking would usher in vast land use opportunities that could change the nature of the city.

Most heartening of all, greenhouse gas emissions from the transportation sector, which in 2016 surpassed the power sector as the United States’ biggest emitter of CO2, would drop precipitously, and air pollution would decline.

Daniel Sperling, director of the University of California, Davis Institute of Transportation Studies, is one of the chief proponents of the shared-EV-AV vision. Because of it, he writes in his book, “Three Revolutions: Steering Automated, Shared, and Electric Vehicles to a Better Future (to be published in March), “For the first time in half a century, real transformative innovations are coming to our world of passenger transportation — with the promise of huge energy, environmental, and social benefits.”  A 2016 Rocky Mountain Institute report predicts that as a result of these developments, “peak car ownership will occur around 2020 and will drop quickly after that.”

Of the many ways this vision could fail to reach fruition, one stands out: It requires a paradigm shift, delivering sharp blows to the predominant car culture and its corollary, private car ownership. For it to work, many Americans must give up owning cars and instead rely on companies such as Lyft and Uber that offer “mobility services.” The money such car travelers could save might induce them to do that: By 2030, they could be paying an estimated $4,000 per year less for transportation than private car owners, according to the Rocky Mountain Institute. 

For this to work, many Americans must give up owning cars and instead rely on companies such as Lyft and Uber.

But many car travelers would probably be tempted to own automated vehicles to preserve their privacy and avoid sharing, despite the increased cost. That option might be particularly appealing to commuters, who could perform many of their office tasks en route, dropping the lost opportunity cost of traveling by car to near-zero. How much savings would be necessary to persuade car travelers to give up driving by themselves or with family, going exactly where and when they want to?

The history of ride sharing is not reassuring. Urban policies such as high-occupancy vehicle (HOV) lanes and car-sharing programs have existed for decades but “haven’t been all that successful,” Nic Lutsey, electric vehicle research director at the nonprofit International Council on Clean Transportation, told me. “Autonomous vehicles have some of the same difficulties as non-autonomous car sharing has. In reality, a lot of people will have the same inclinations as they do today, to own a private auto and use it the way they want, without compromises.” Indeed, vehicle sales are still setting records in the U.S. and globally, and U.S. vehicle miles traveled continue to climb.

Some shared EV-AV proponents find reassurance in widespread evidence that car ownership has declined among millennials, but it is uncertain whether that drop reflects a real preference or just strained finances. Indeed, as the economy has regained momentum in the last couple of years, so has car ownership among the young.

And if sharing doesn’t catch on but automated vehicles do, the result could be more traffic congestion, more vehicle miles traveled, and more emissions, even shared-EV-AV advocates such as Sperling agree. After arriving at their work destinations in the morning, AV-owning commuters could avoid paying high parking fees by sending their cars home, then summoning them back at the end of the day. AV owners going to appointments could avoid parking fees by directing their cars to continue circulating until the appointment ends.

Freight truck travel might become far cheaper after being freed from its most expensive component — drivers— prompting an increase in truck shipping. Such developments, a 2015 study published in Transportation Research said dryly, “would have large total energy and carbon implications.” They would also put about 3.5 million U.S. truck drivers out of work, with conceivably major social consequences.

If car sharing doesn’t catch on but automated vehicles do, the result could be more vehicle miles traveled and more emissions.

Because the range of electric vehicles is still limited, it’s not even certain that early automated vehicles will be electric. In fact, Uber announced in November that it was ordering 24,000 plug-in hybrid — not all-electric — Volvo SUVs that it plans to convert into automated taxis. And Ford revealed last month that its first AVs, which it hopes to release in 2021, also will be hybrids. That prompted Mary Nichols, longtime chair of the California Air Resources Board, to tweet, “Earth to Ford: what part of sustainability do you not understand? Driverless hybrid vehicles running 24/7 delivering pizza and passengers means more tons of pollution/GHGs [greenhouse gases] in cities!”

“It’s not obvious that autonomy is a good thing,” Hal Harvey, CEO of Energy Innovation, a San Francisco-based energy think tank, said in an interview. “Autonomous is the most dangerous as a social disrupter, a polluter, a congestion machine.”

On top of this, the transition to shared EV-AV cars is not likely to be smooth — and it will take many decades. Until then, automated vehicles will have to contend with human-driven cars, which, it turns out, behave far more unpredictably than automated ones do. 

“Hardware and software will need to figure out how quickly cars are merging into freeway traffic, anticipate when someone is about to pull a U-turn or make an illegal left turn, and myriad other actions that can result in deadly crashes,” Sperling writes in “Three Revolutions.” As a result, automated vehicles will be forced to give a wide berth to human-driven ones, lowering overall road speeds and increasing congestion.

Despite all this, Sperling is optimistic about the future of “three-revolution” cars. “I think we’re ahead of the curve on this one,” he said in an interview. “Usually when you do policy, you’re lagging behind the disruptive technologies, but in this case we have an opportunity to lead. And companies like Uber and Lyft are well-aligned with the public interest — they see pooling as important to their business model and important to their future.”

Uber started a driverless car pilot program in Pittsburgh in 2016. The company now plans to convert 24,000 plug-in hybrid Volvos into automated taxis for tests in other U.S. cities.

Uber started a driverless car pilot program in Pittsburgh in 2016. The company now plans to convert 24,000 plug-in hybrid Volvos into automated taxis for tests in other U.S. cities. Uber

One reason for quick action is that cities can help influence whether early automated vehicles will be fueled by gasoline — with its attendant climate change and pollution problems — or electricity. “Getting autonomous vehicles to plug into a clean electric grid for power is perhaps the only sure way for them to contribute to environmental goals like improved air quality and climate change mitigation,” Lutsey said. “So it is certainly worth it to develop policy and business solutions to ensure that autonomous vehicles are increasingly electric.” 

Policy solutions involve ending or limiting what are effectively subsidies for private car ownership and replacing them with subsidies for electric and shared vehicles. For example, toll-free roads and abundant curb space for parking provide valuable resources to car owners, while the maintenance cost is spread across the entire population in taxes. Policymakers could begin reversing those incentives by taxing privately owned vehicles and using the proceeds to upgrade public transit. They could subsidize EV sales and revamp building codes to require wiring for EV charging stations.

To discourage private driving, some curbside parking could be turned into passenger loading zones where shared mobility vehicles could pick up and unload passengers without blocking traffic, as they so often do now. The plentiful data that Lyft, Uber et al gather on the routes and number of passengers of their vehicles could be tapped to require mobility service companies to drive without passengers no more than 15 or 20 percent of the miles they cover.

Perhaps most important, cities could declare a maximum number of cars that its streets can support and install a form of congestion pricing potent enough to ensure that that number isn’t exceeded. That, too, would discourage AVs and mobility service vehicles from circulating while empty, an absolute necessity if the low-emission shared-EV-AV vision is to have a realistic hope of success.

Charging for road use is often a political nonstarter, but the alternative may be more and more congestion. Will enough people get sick of gridlocked, unhealthy, climate-damaging transportation systems to make a fundamental shift in how they use cars? The answer will play a major role in determining whether the “three-revolution” vehicle comes of age.