Project description

In our recently completed PSR project, we developed the first integrated model that combines both the morning and evening commute with ride-hailing (e.g., Uber and Lyft) and ridesharing (e.g., carpooling) trips, and showed that one can capture the mode switches occurring between morning and evening commutes. Moreover, the coupled model revealed that more travelers are able to use carpool than if only the morning or evening commute was modeled alone, and the net result is 8.4% less vehicle miles traveled (VMT) in the studied network. Since the primary focus of that study is to examine how the coupling between morning and evening commute affects mode choice and VMT, it made several simplifying assumptions. In this project, we relax some of these assumptions and extend our integrated model to include deadheading of ride-hailing vehicles and allow for more flexible ridesharing, namely people from different O/D pairs can carpool together. Such enhancements to our integrated model would capture more complex interactions between various modes and provide a more realistic estimate of VMTs. It will also allow us to study the factor combinations that lead to lower VMTs in a road network. This collaborative proposal combines the expertise of domain researchers in transportation science and engineering and leading researchers in mathematical optimization from the two institutions: University of California at Davis and University of Southern California. Their joint effort is a new chapter in the collaboration between these two institutions with the common goal of addressing an important change of events in today's mobility systems that has significant impact to the well-being of urban transportation.