Planning on-demand services beside public transit

Designing an integrated mobility system is complex, especially when adding shared and on-demand services. A study in the U.S city of Portland, supported by PTV Group modeling, shows how technology helps mobility planners and operators to evaluate the impact of various scenarios.

The study’s goal was to explore the feasibility of on-demand services, and their impact when integrated with mass transit. Using a demand model from the regional planning authority, the study evaluated different scenarios of on-demand fleet operations.

Using PTV Visum software, the writers of the study analyzed travel patterns and travel delays in the Portland Metro area. It turned out that the longest delays were caused by eastbound traffic on bridges across the Willamette River in the afternoon peak hours. The study assumed that 50% of these trips could be replaced by an attractive combination of on-demand services and public transit.

Another aim of the study was to determine the necessary fleet size for on-demand operations in Portland. Using PTV Visum software, a model was created with the following conditions: maximum wait time of 10 minutes, pre-booking time of 5 minutes, and a maximum detour time of 30 minutes. The outcome was a fleet of 249 vehicles with 8-seat capacity.

PTV models also demonstrated that although 90% of passengers will experience up to 18 minutes increase in trip time compared to car travel; but they will still save hundreds of dollars per month in car parking and operating costs. The study highlighted the importance of sustainable policies to drive modal shift.

With the PTV Visum demand models, authorities and operators can test various mobility concepts before implementing them. PTV Visum provides analytical visualizations that are needed to understand city transportation and impacts at the macro level. Other PTV software provide simulations, so operators can choose the right business model. Integrating these tools empowers stakeholders as they plan on-demand and MaaS.