Investigating Fuel Consumption and Pollutants Emission of Autonomous Vehicles Using the Waymo Open Dataset

Autonomous vehicles are conjectured to be environmentally friendly, but their impact on environmental sustainability is largely hypothesized or assessed via numerical simulations that do not consider driving behavior—an essential component of microscopic emission models. As such, the existing literature is devoid of any concrete evidence of autonomous vehicles’ impact on the environment using real-world data. Using the recently released Waymo open dataset, this study investigates the environmental impacts of autonomous vehicles. Specifically, microscopic fuel consumption models are employed to estimate fuel consumption and pollutant emissions (i.e., carbon dioxide, nitrogen oxide, volatile organic compounds, and particulate matter) for autonomous vehicles and compared with human-driven vehicles obtained from the same dataset. Results indicate that autonomous vehicles consume relatively less fuel (about 8%) and emit lower pollutants (about 66%) when compared to human-driven vehicles. The findings of this study highlight the benefits of integrating autonomous vehicles into the current traffic stream.