By Zulqarnain Haider
Tuesday’s session on Mobility on Demand systems with transit integration was all about using the modern shared mobility options, like car share, shuttle share, among other options, and effectively integrating them with the existing transit systems. The session started with an insightful talk by Pascal Van Hentenryck of University of Michigan, about a case study from the city of Ann Arbor, which focused on using shared-shuttle service to solve the first mile/last mile problem for public transit. Dr. Van Hentenryck especially underscored the importance of cities reimagining the future of their transit by incorporating multiple options. The goal must be to complement the higher utilization of transit options in high ridership areas with ready availability of small-scale shared mobility options with lower wait times for people in the peripheries of the service area.
The second speaker was Marco Pavone (photo) from Stanford University. Marco focused on the futuristic scenario where autonomous vehicles become readily available. The goal of his research was to integrate the autonomous mobility on demand systems with public transportation. He presented models and algorithms to that effect and presented a case study for New York City.
The third speaker, Crystal Wang, Ford Motor Company, started by introducing the Ford analytics initiative and the mobility on demand options that Ford is developing. She emphasized the need for mobility on demand options for the first mile/last mile problem of public transit. She shared findings from deploying mobility on demand services in different parts of the world including Shanghai, China and King County in the Seattle area. She enumerated various challenges faced by the deployed systems and suggested ways to design them better to increase their adoption.
The final speaker of the session was the session chair Samitha Samaranayake from Cornell University. He stressed the importance of using mobility on demand services in conjunction with the mass transit options. He also emphasized the need for efficient and fast models to match riders with the mobility on demand services. He presented several models and some compression techniques to make the models tractable for large problems on an urban scale. The session was very well attended and the audience members continued interacting with speakers after the session ended.