Data-driven Quantification of Societal Impacts from Emerging Transportation Technologies

Emerging transportation technologies like online grocery delivery services, transportation network companies (TNCs), and Electric Vehicles (EVs) have the potential of changing how people interact with the transportation network, offering more opportunities and ease in terms of the movement of people and goods. However, with these technological disruptions come potential environmental and socio-economic impacts. In this dissertation, I quantify the environmental and socio-economic impacts of these emerging technologies in the U.S., as well as provide insights into understanding how these technologies interact with people and society (e.g., CO2 emissions, congestion, and equity gaps in service). These analyses provide critical information that can be used by policy makers in the transportation space to better understand these technologies and to enhance decision-making.

In Chapter 2, I quantify the impact of online grocery delivery services in Seattle, WA, on the environment, measured in terms of CO2 emissions, and the transportation system, measured in terms of vehicle miles traveled and vehicle hours traveled. I use Montecarlo simulation to simulate different scenarios of online grocery delivery penetration, batching and customer substitution behavior, and then use the outputs of this simulation in a static traffic assignment model to translate these scenarios into the resulting changes in transportation network congestion. This analysis provides insights on what policy levers can effectively curtail emissions and congestion impacts, both from the supply side and the demand side.

In Chapter 3, I evaluate the role of TNCs (e.g., Uber and Lyft) as a transportation modality during rain in Chicago, IL. Making use of the causal inference framework, I use the augmented inverse propensity weighted estimator to obtain statistically valid confidence intervals for the effect of rain on TNC ridership. In order to tie these effects to the underlying demographics of Chicago I perform an ordinary least squares regression, which explores the relationship between these effects and variables such as population, income, transit availability and race. These estimates and their relationships to v demographics will help policy makers understand how disruptive weather events like rain change the way this technology interacts with the city.

In Chapter 4, I propose a framework to site level 2 electric vehicle supply equipment (EVSE) that directly considers both the near-term demand for EV charging and the need for public EVSE to spur EV adoption. I develop a bi-objective location allocation mixed integer linear program, for both Pittsburgh, PA, and Seattle, WA. The resulting optimal sitings for different levels of stakeholder equity preference are then evaluated in terms of user experience tradeoffs for both income and race demographics. This framework can be used by urban planners to evaluate which EVSE siting strategy fits their preferred goals. I conclude the dissertation with general policy implications and future work directions.