Emissions May Rise if Self-Driving Cars Become Too Convenient
If a road trip from Seattle to Portland meant three hours of focused work rather than gripping the wheel and gritting your teeth through traffic, would you be more likely to go by car than take the train, plane, or bus?
That’s one scenario that could cause a transition to fully automated self-driving cars to significantly increase greenhouse gas (GHG) emissions from road transportation. In other scenarios, self-driving cars—with their touted improvements to safety and efficiency—could cut emissions.
One big factor is whether we continue to individually own autonomous vehicles or transition to a system of shared, on-demand vehicles, says University of Washington professor Don MacKenzie, who co-authored the paper “Help or hindrance? The travel, energy and carbon impacts of highly automated vehicles” with Zia Wadud of the University of Leeds, and Paul Leiby at Oak Ridge National Laboratory.
They examined several factors that could result in a wide and divergent range of emissions in a future of autonomous vehicles: “Automation might plausibly reduce road transport GHG emissions and energy use by nearly half—or nearly double them—depending on which effects come to dominate,” they write in the paper, published Thursday in the journal Transportation Research Part A.
MacKenzie acknowledges that the wide range of potential effects is frustrating, but says those people who claim to know precisely how a future with autonomous vehicles will unfold are fibbing or not thinking clearly. He and his colleagues undertook this research after observing “a lot of enthusiasm and what we characterize as somewhat utopian thinking about the opportunities and the promise of automated vehicles. We felt that what was really needed was more dispassionate analysis and less rose colored glasses,” he says.
The authors cite the 2012 debut of Google’s fully automated car as the “herald of a new era of automation.” Carmakers and technology companies alike are investing heavily in this new era, and vehicle automation experts predict vehicles capable of autonomous freeway driving will be available by the end of this decade—albeit with drivers at the ready to assist. (Indeed, last fall Tesla began equipping new Model S vehicles with hardware such as radar, ultrasonic centers, cameras, and braking systems to enable “Autopilot.” The goal is not to make the car fully autonomous—yet. “Tesla’s Autopilot is a way to relieve drivers of the most boring and potentially dangerous aspects of road travel—but the driver is still responsible for, and ultimately in control of, the car,” the company said.)
Cars capable of self-driving on urban and rural streets and roads are expected by the mid-2020s, again with a human driver at the ready as backup, according to a survey of vehicle automation experts cited in the paper. By the 2030s, human drivers might be taken out of the equation.
The main arguments in favor of automation center on improving the safety and efficiency of the transportation system, says MacKenzie, who leads the UW’s Sustainable Transportation Lab. But these benefits could also lead to reductions in transportation energy use and associated emissions.
For example, in the near-term, vehicles on the freeway could be programmed to “platoon” together, driving in tight formation and drafting to reduce aerodynamic energy consumption. Also, computers would be expected to better adhere to eco-driving practices—things like coasting to a stop and not stomping on the accelerator—leading to greater efficiency there. “These are things you can train a human to do, but they tend to forget and go back to their old habits,” MacKenzie says.
Autonomous cars are expected to be better at avoiding collisions, which could allow for lighter, and therefore more efficient, vehicles. But for consumers to accept this vehicle downsizing, they’ll need to see “not only impeccable safety from your own vehicle, but from all of the vehicles around you on the road that might run into you,” MacKenzie says.
“What that suggests is that some of these benefits are going to scale non-linearly, so that you’ll probably hit some critical point where safety becomes so good that you can finally make this switch to these much smaller, much lighter vehicles,” he says. “But it’s not likely to be just a straight line from here to there.”
The main factor that could cause widespread adoption of fully autonomous vehicles to increase emissions is their convenience.
“Right now, on an average basis, the driver’s time is the single largest cost of operating a vehicle,” MacKenzie says. If you reduce that cost, demand for travel by personal car could increase enough to “ultimately swamp these energy efficiency improvements,” he says.
Given the wide range of potential futures, I asked MacKenzie if there’s a single, influential factor to watch as the autonomous vehicles industry takes shape in the coming decades:
“One of the key things that’s going to drive the overall outcome is whether we end up with a self-driving car in every garage, or shared self-driving vehicles and people giving up their personally owned cars,” he says. “The shared, automated system opens up a lot more opportunities for energy savings, and … you would probably see less of an increase in travel demand.”
One example is already in our midst through ride-hailing services like Uber and Lyft that let the passenger specify the size of vehicle for a given trip. In the future, a single-occupant pod could be dispatched when you need to get yourself to work, and an SUV for taking the family out for dinner.
“By right-sizing, we can cut energy, it looks like, 20 to 40 percent on a per-mile basis,” MacKenzie says.
But if we each have an autonomous car or two that enables us to endure longer commutes because we don’t have to drive, or to send the car home to park while we’re at work, or to make a bunch of individual trips rather than grouping errands together—that’s a future likely to have significantly higher emissions.