This project was funded by the National Institute for Transportation and Communities (NITC), with additional support from Portland State University, the University of Utah, and community partners Assist Inc. and Unlimited Choices.
Machine learning -- Applications to local transit planning, Twitter, Choice of transportation -- United States --Planning, Travel time (Traffic engineering)
With today’s profusion of open data sources and real-time feeds, transit agencies have an unparalleled opportunity to leverage large amounts of data to improve transit service. Thanks to NITC researchers, there is now an open-source tool for that.
The new Social-Transportation Analytic Toolbox (STAT) for Transit Networks, developed by researchers at the University of Utah and Portland State University, is a dynamic platform that combines Twitter, general transit feed specification (GTFS), and census transportation planning products (CTPP)—in this case, job density data—to help agencies evaluate overall system performance and identify connectivity gaps. It can also act as a decision support tool for recommending service improvements.
Liu, Xiaoyue Cathy, Wei, Ran, Aaron Golub and Liming Wang. Leveraging Twitter and Machine Learning for Real-Time Transit Network Evaluation. Project Brief NITC-RR-1080. Portland, OR: Transportation Research and Education Center (TREC), 2019.