Regional travel demand models are an institutionalized element of the transportation planning process and require multiple years to develop, calibrate, and deploy. Because transportation planners are being asked immediately how a new technology, i.e., driverless vehicles(DVs), may affect travel demand, this study, using a case study approach with one particular region, identified how such models can be modified to incorporate the potential impacts of DVs and to answer related questions of interest to stakeholders. The DVs described in this report are presumed to be completely autonomous and are what SAE International refers to as Level 5 vehicles.
A key finding is that it is possible to address some impacts of DVs in the model,such as changes in capacity, mode share, travel by age groups that traditionally have had less access to vehicles, trip length, and sharing of such DVs. Execution of the regional model using these modifications provided answers to some questions of interest; for example, a decrease in capacity during the transition period to DVs could lead to a substantial increase in delay (a 46% increase in vehicle hours traveled [VHT] in the case study area, whereas greater access to DVs by groups that traditionally have not had access to a vehicle suggests only a modest increase in delay (a 3.3% increase in VHT). Incorporation of such impacts into the model can also inform policies; for example, it is possible that the advent of DVs could encourage commuters to seek to avoid parking fees by either sending privately owned vehicles back home or sharing subscription-based DVs. Both situations increased zero occupant vehicle trips in the case study model, but the former increased nitrogen oxide emissions by an estimated 11.64% whereas the latter increased them by 2.08% to 6.65% depending on the manner in which the sharing occurred.
The study recommends that language indicating how DV impacts may be incorporated into existing regional models be added to VDOT’sTravel Demand Modeling Policies and Procedures manual when the manual is next updated. Draft language for implementation of this recommendation is provided in Table ES3 in the Executive Summary and in Table 29 in the body of the report.