How will self-driving vehicles affect US megaregion traffic? The case of the Texas Triangle

This paper anticipates the impacts of self-driving or "autonomous" vehicles (AVs), shared AVs, and Atrucks on travel across the Texas Triangle megaregion using year 2040 land use (and network) forecasts. A statewide travel demand model forecasts changes in trip generation, mode and destination choices, and thus vehicle-miles traveled (VMT), congestion, and travel patterns across the megaregion. Results suggest travelers' shifting to more distant destinations, with average person-trip distance rising from 14 to 16 miles. Within-region airline passenger travel is predicted to fall by 82%, as travelers shift to self-driving ground transport options. Without travel demand management (like credit-based congestion pricing and mandated tight headways between AVs), congestion issues will grow, due to an average 47% VMT increase, especially in the region's major cities (Houston, Dallas, San Antonio, and Austin). Automobile travel is anticipated to rise across all distance categories, with increases most evident between suburban and urban zones. Almost 9.6% of link flows will exceed capacity, relative to 4.6% of segments in the no-AV case (for year 2040). Four of the 15 freight industries are predicted to experience an increase of more than 100 million ton-miles per day, due to the introduction of Atrucks, with rising truck trades largely between Houston and other major cities.