Urban microtransit cross-sectional study for service portfolio design

The uploaded set of codes and datasets are for urban microtransit cross-sectional study for service portfolio design. The research team proposes a novel method to upscale the limited microtransit data available using simulation and a scenario generation process which can be further used for decision-support analysis and microtransit modeling. The method is tested in a case study in collaboration with Via Transportation based on data they shared for Salt Lake City, Austin, Cupertino, Sacramento, Columbus, and Jersey City. The implementation includes simulating microtransit performance data (including both first/last mile access trips and direct trips) using a day-to-day adjustment process for a set of cities (with microtransit data availability) calibrated to match occupancy data. This uses a within-day microtransit simulator (developed for the Federal Transit Administration) that is enhanced to be more parametric in design to be calibrated to different cities, and city wise mode choice model data that is estimated and calibrated to match the Via ridership data. For the service portfolio design, ridership and vehicle miles traveled (VMT) forecast models are estimated using surrogate (upscaled) data on multiple scenarios generated from a scenario generation process. This dataset includes population and built-environment characteristics of the scenarios for which the microtransit performances are obtained using the calibrated day-to-day simulation model. The microtransit portfolio forecast models are developed based on the upscaled data using multiple linear regression with second order polynomial (interacting) features and validated based on observed Via data.