Urban Microtransit Cross-Sectional Study for Service Portfolio Design

University Transportation Centers Program2021PDFTech ReportChow, Joseph Y JZimmerman, RaeRath, SrushtiLiu, BingqingYoon, GyugeunConnected Cities for Smart Mobility toward Accessible and Resilient Transportation Center (C2SMART)United States. Department of Transportation. University Transportation Centers (UTC) ProgramUnited States. Department of Transportation. Office of the Assistant Secretary for Research and TechnologyConnected Cities for Smart Mobility toward Accessible and Resilient Transportation Center (C2SMART)United StatesDemand responsive transportationForecastingMode choiceRidershipUrban transportationMicromobility69A3551747124Due to transportation technologies having such heterogeneous impacts on different communities, there needs to be better tools to evaluate the deployment of emerging technologies with limited data. Microtransit is one such technology. The authors propose a novel methodology to \u201cupscale\u201d the limited data available so that further decision-support analysis and modeling can be achieved by microtransit companies working with cities around the U.S. where none existed previously. The methodology involves simulating data using a calibrated day-to-day adjustment process for a set of cities in which data are available. The day-to-day adjustment process simulates both first/last mile access trips and direct trips with the adjustments made to match occupancy data. A within-day microtransit simulator developed for the Federal Transit Administration is enhanced to be more parametric in design to be calibrated to different cities. A scenario generation process is developed to come up with the scenarios from which the data are generated. 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, and Columbus, as well as publicly available data from Jersey City. For those cities, public data is collected to estimate mode choice models that include Auto, Bike, Transit, Microtransit, and Walk for each city. Public data include U.S. Census Transportation Planning Products, American Community Survey, Transitfeeds, Smart Location Database from EPA, OpenStreetMap, and OpenTripPlanner. The models are estimated initially using maximum likelihood without the Microtransit mode since the data do not include it; afterwards, the Microtransit alternative specific constant is updated to minimize least squares from ridership data shared by Via. The average microtransit ridership error over the 6 cities with the estimated constant is 0.004 while the error with a constant of 0 is 603.59, showing that the method fits quite well.127