Don MacKenzie, University of Washington – Driverless Cars

Self-driving cars could have some positive and negative environmental factors.

Don MacKenzie, assistant professor of transportation engineering at the University of Washington, explores how making transportation easier could lead to even more people getting on the road.

Don MacKenzie joined the department as an Assistant Professor in 2013. His research focuses on the interactions of emerging transportation technologies and public policies, and their effects on energy consumption. He is currently working in three related areas: (1) modeling the charging decisions of electric vehicle drivers, and implications for electric grid loads and generation emissions, (2) assessing prospective energy efficiency and travel demand implications of vehicle automation, and (3) evaluating the effects of services such as car-sharing and online shopping on total travel demand.

Dr. MacKenzie previously did research and advocacy work on clean vehicles for the Union of Concerned Scientists in Washington, D.C., and worked as a researcher on biofuel technologies at a startup in Vancouver, Canada. He holds a PhD in Engineering Systems and a Master’s in Technology & Policy, both from MIT. He also holds a Bachelor of Applied Science in Chemical & Biological Engineering (Environmental Option) from the University of British Columbia.

Driverless Cars

In just three years, self-driving cars have gone from a curiosity to a seeming inevitability. Detroit and Silicon Valley alike are promising safer travel, less congestion, and improved mobility for groups like the elderly and the blind. But what will vehicle automation mean for energy use and climate change?

Working with Zia Wadud and Paul Leiby, we quantified potential impacts of automation on vehicle energy efficiency and the amount of driving. We found self-driving cars create substantial opportunities to improve energy efficiency: driving in tight, aerodynamic formations on the highway; optimizing acceleration and braking; and reducing consumer demand for horsepower. Each of these could reduce energy per mile by 5-20%.

By far the biggest opportunity is “right-sizing” our vehicles. Today, we buy cars for our peak needs, but mostly drive around alone. Now imagine an on-demand service, like Uber without the driver. A vehicle is dispatched with size and capabilities tailored to the specific trip, reducing energy per mile by an average of 20-40%.

But there are risks, too. The single largest cost of driving is the value of the driver’s time, and automation will reduce this cost. When something is safer, cheaper, and more convenient, people will consume more of it. In this case, vehicle travel could increase by up to 60%.

Putting the pieces together, automation might reduce on-road energy demand by 50% – or double it. This uncertainty arises because the changes don’t come from automation per se, but in how, and how much, we use our vehicles. We should not blithely assume that left to its own devices, vehicle automation will save the climate. In fact, it could make things a lot worse.