Modeling the Travel and Greenhouse Gas Impacts of Shared Mobility and Automated Vehicles 

Abstract
Cities, regional agencies, and States throughout the U.S. are increasingly developing energy and climate action plans to meet greenhouse gas (GHG) reduction and renewable energy objectives. The City of Boston, Massachusetts and stakeholders worked through the Carbon-Free Boston Initiative to chart a path towards carbon neutrality by 2050. Transportation emission reduction strategies are a critical part of the city’s plans, as about 2 million metric tons of GHGs are emitted as a result of travel starting and/or ending in the City. The rise of shared mobility services, and the potential of connected and automated vehicles (CAV) in the future, has added another layer of uncertainty to local energy and climate planning efforts. Key questions include (1) the extent to which unmanaged shared mobility and CAV services will either reduce or increase GHG emissions, and (2) what policy levers can be used to manage emissions from these sources and work towards achieving overall GHG and congestion reduction goals. The project team modeled the impact of pricing strategies to increase vehicle occupancy in shared mobility services and manage CAV vehicle-miles traveled (VMT) for travel starting or ending within the City of Boston. The team developed the Future Mobility Tool, a spatial model implemented with Python scripts, that combines travel patterns, travel times, and a mode choice model from the regional travel demand model with other off-model adjustments to provide quick-response analysis of geographically specific policies. The mode choice model was expanded to include ride-alone and shared-ride services in addition to the traditional drive, carpool, transit (walk and drive access), walk, and bicycle modes. Scenarios were developed comparing GHG emissions with conventional vs. electric vehicle technology. The impacts of potential changes in auto ownership on mode choice were also modeled, and the impacts of “empty miles,” circuity of shared mobility services, and improved efficiency of CAVs were all considered.
Unmanaged CAV and shared mobility scenarios showed overall VMT increases in the range of 3 to 6 percent compared to the 2050 baseline. However, pricing of CAV trips could help hold VMT roughly at constant levels. Shared mobility cross-subsidies ($1 per mile surcharge on ride alone trips, used to subsidize shared ride trips at $1 per mile) could reduce VMT by at least 3 percent. GHG emissions could also increase slightly with unmanaged CAVs and shared mobility, but could be reduced more significantly, at least 20 percent, if these vehicles were required to be electric, and electricity was provided from a grid much cleaner than today’s. The results of the model suggest that fairly aggressive pricing policies are needed to shift substantial numbers of travelers into shared-ride services and reduce congestion in addition to reducing emissions. However, these policies can be structured to be revenue-neutral or revenue-positive depending on local objectives, benefitting travelers who make environmentally friendly choices.