The Autonomous Vehicle Revolution Will Be Underwhelming

We still need to prepare for it today.

A Lyft car drives along Powell Street on June 12, 2014 in San Francisco, California.
When your first Lyft arrives with no driver, you’ll probably be underwhelmed.

Justin Sullivan/Getty Images

Autonomous vehicles now dominate discussions about personal mobility in many research communities, regulatory circles, and the popular press. (Case in point: this series here in Future Tense.) There are impressive demonstrations already: Prototype self-driving cars from major automotive manufacturers, tech giants, startups, and universities now roll through the streets of Silicon Valley with regularity. Ever-bolder predictions of the imminent demise of the human driver seem to follow each major technology conference. I found myself half-expecting to walk out of the Consumer Electronics Show this year and find Johnny Cab from Total Recall waiting for me at the curb

We’re currently cresting the peak of the hype cycle. Viable autonomous vehicles are still in early phases of development and won’t arrive overnight. There are tough technical challenges and thorny policy issues yet to be overcome. The U.S. Department of Transportation is now working on guidance, expected this summer, for how autonomous vehicles will eventually be regulated when they do arrive.

For now, the closest products that consumers can buy are “highly automated vehicles,” such as Tesla’s autopilot function, that control speed and road position but still require vigilant human supervision. Evidence suggests that some of these driver assistance technologies are effective at avoiding or mitigating collisions, and that even partial automation can reduce energy consumption. Unfortunately, American consumers have shown themselves historically unwilling to pay for safety, and depending on when and whom you ask, they may or may not care about efficiency. As a result, we’ll probably need to rely on federal regulatory activity to continue to push vehicle efficiency and driver aids in the near future.

What consumers do pay for is convenience. For an autonomous vehicle to be truly convenient, it wouldn’t need any supervision. It would allow the user to work, read, or even sleep while en route to a destination. And as has been widely discussed, truly autonomous vehicles would provide mobility for the disabled, impaired, or those otherwise unable to drive. But guess what? The seamless, end-to-end customer experience of a perfectly engineered, fully autonomous vehicle is available today, and it’s only as far away as your smartphone. Here’s a clue: It will still have a driver. And it might have a “U” or a pink moustache on the windshield.

The convenience of Uber, Lyft, and their many competitors is addictive: A few taps on your phone and minutes later a vehicle appears and whisks you to your destination. The stress of traffic and search for parking are replaced by a precious few minutes of calm to send an email or catch up on the news. You don’t even have to calculate a tip (unless you really want to). It’s understandable, then, that the ride-hailing business has grown at an astounding pace, with hundreds of thousands of drivers and more than 1 billion rides delivered after only five years. It took McDonalds nearly two decades to sell that many hamburgers.

From a social perspective, the problem of all this convenience is a concept called induced demand: Make something cheaper or easier and people will consume more of it. It took traffic engineers decades to wrap their heads around the idea that adding roads made traffic worse in the long run, not better. We’re facing exactly the same challenge today when we increase the accessibility and convenience of travel in cars, whether they be autonomous or mustachioed: We should expect that demand for car travel will increase.

While ride-hailing companies generally keep data close to their chest, results from early surveys already suggest that the ride-hailing phenomenon is inducing demand. It shouldn’t come as a surprise that increased travel demand is also flagged in a recent study as a key driver of enormous uncertainty in the environmental impact of autonomous vehicles. The authors find that automation could double or halve energy consumption, depending on whether systemic efficiencies can offset the burden of additional travel.

One way to manage additional demand is to enact policies that ensure that if car travel becomes easier, so do more efficient modes of travel. The struggle to maintain this balance is being played out at the curbside of urban centers around the world, where cars, bicycles, buses, and pedestrians all fight to use the same precious few square feet. For nearly two decades Enrique Peñalosa, the mayor of Bogotá, Colombia, has explained that the use of road and curb space is not just a tool to protect the environment, but to protect democracy, and that “an advanced city is not one where even the poor use cars, but rather one where even the rich use public transport or bicycles.” Yet the distinction between public and private transport is increasingly blurry. Ride-hailing companies now offer rides shared with other passengers at lower prices, and shared-van services like Bridj operate larger private vehicles.

So how can we fairly evaluate the environmental impact of today’s ride-hailing services? From a societal perspective, is a private, multipassenger vehicle better or worse than a city bus? The challenge in answering this question is twofold. First, to understand the counterfactual: What would these passengers have done had Bridj, Uber, or Lyft not existed? Would they have taken a bicycle? A private car? Or maybe not taken the trip at all? Second, we need to fairly treat the full cost of the trip, accounting for traffic and emissions generated between passengers, just as we would count emissions of a private car circling for parking. And while we’re at it, ride-hailing trips with a single passenger aren’t really “sharing” and shouldn’t qualify for the carpool lane.

If there’s one thing that ride-hailing services can do effectively, it’s to quickly put lots of miles on a few cars. Ask your next Uber or Lyft driver how many miles they drive. If they work full time, my anecdotal evidence suggests they’re driving 60,000 miles per year or more—enough to wear out a car in just two or three years. Autonomous vehicles, able to run round the clock, could rack up more than 100,000 miles a year. These numbers are a stark contrast to traditional motorization statistics in the U.S., where the average car is 11 years old and driven only 12,000 miles per year.

Researchers are quick to point out that fewer parked cars free up curb space for other modes of transport. But underused vehicles aren’t just a parking nuisance; they’re an impediment to change. An older vehicle fleet will dilute the benefit of forthcoming connected vehicle technologies that depend on a network effect. And an older fleet also limits our ability to introduce smaller, lighter vehicles, since crashes between vehicles with disparate weights and sizes are more violent.

Sharing vehicles is a change agent that permits technology to move more quickly, and all users benefit rather than just those who can afford to purchase a new car. Years from now when viable autonomous vehicles make a beachhead, they will quickly be incorporated into sharing systems. When your first Uber or Lyft arrives with no driver, you’ll probably be underwhelmed. You’ll still send emails in the back seat. It will be cheap and convenient. But hopefully not as convenient as a bicycle.

This article is part of the self-driving cars installment of Futurography, a series in which Future Tense introduces readers to the technologies that will define tomorrow. Each month from January through June 2016, we’ll choose a new technology and break it down. Read more from Futurography on self-driving cars:

Future Tense is a collaboration among Arizona State University, New America, and Slate. To get the latest from Futurography in your inbox, sign up for the weekly Future Tense newsletter.