Technology Policy Challenges of Electric and Automated Vehicles

This body of work covers some important technology policy challenges that will be posed by the deployment of electric and automated vehicles in personal mobility, ridesourcing, and long haul trucking. Substantial electrification of road transport is critical to deep decarbonization. Self-driving technology continues to inch towards wider commercial deployment, despite well-publicized setbacks. Policymaking must grapple with a deeply uncertain technology landscape and trade-offs between multiple societal objectives. I attempt to help inform public policy by carefully navigating the uncertainty present in the deployment of these technologies and highlighting how trade-offs between multiple objectives might materialize. 

First, I look at the potential energy impact of automation on electric vehicle range. Will electrification and automation of light vehicles develop in sync? I find that automation will incur expected penalties of 10-15% on electric vehicle range, due to the energy consumption of computing and sensors. This suggests that deliberate technology choices towards efficient computing and aerodynamics will be needed to drive electrification of automated vehicles. In long haul trucking, I investigate the potential job impacts from automation of highway driving. I draw on detailed shipment data to show that replacing all highway driving with automated trucking will lead to more than 90% of operator-hours lost to automation. I find limited evidence to support industry claims that increased short haul jobs will compensate for the operator-hours lost on highways. Third, I present a novel agent based model that can answer a series of research questions related to electric vehicle deployment in ridesourcing. Companies such as Uber and Lyft have committed to fully-electrify their fleets by 2030 in the United States. Should policy incentivize larger battery pack sizes or greater investments in dedicated fast-charging infrastructure for ridesourcing? I instantiate the agent based model to study the lifecycle externalities of fully-electric ridesourcing in the city of Chicago, across different levels of battery size and fast-charging infrastructure. I estimate that greenhouse gas externalities are in the range of 10-13¢ per trip, with lower externalities associated with smaller battery packs. Compared to today, fully electrifying ridesourcing can cut greenhouse gas externalities in Chicago by 50-60%. Solely targeting reductions in greenhouse gas emissions, however, comes at the cost of increases in traffic externalities such as congestion