Wayne State researchers receive National Science Foundation funds to study potential of AI-powered microtransit systems in at-risk communities
Microtransit is a service intended to complement existing public transportation options. It exists somewhere between traditional fixed-path transit and ride-hailing technology, with routes and schedules kept flexible based on user demand. Detroit and other cities have begun to adopt microtransit as a means of increasing coverage and reaching more people, particularly in low-density or low-income areas.
A shortcoming of microtransit, however, occurs when there is a mismatch between the locations of affordable housing and employment opportunities. A team of Wayne State University researchers recently received funding from the National Science Foundation (NSF) for a study that aims to ease this burden facing at-risk communities.
"With the rise of artificial intelligence and increasingly available smart mobility data, the vision of this research project is to create a dynamic routing-prediction system based on learning the hourly mobility patterns between jobs and housing," said Dongxiao Zhu, associate professor of computer science in the College of Engineering and the project's principal investigator.
At the heart of the project is the design of an AI-assisted microtransit system that transportation officials and other civic partners can deploy to better adapt to spatiotemporal variations in the mobility patterns of hourly workers. Geocoded socioeconomic data can be leveraged to identify and reduce mobility disparities.
"The research innovation is expected to provide immediate, low-cost, effective public transit solutions that benefit vulnerable communities in Detroit by significantly reducing transit risk, commute time and distance, and trip cost," said Zhu.
Wayne State was among 12 teams in the U.S. to receive nearly $50,000 in the first stage of the NSF Civic Innovation Challenge, a competition that targets ready-to-implement, research-based pilot projects with the potential for scalable and sustainable community impact. The team can apply these funds to testing and other pre-development activities while preparing to submit a full proposal for the second stage, in which four teams will receive up to $1 million to execute and evaluate their projects.
Collaborators on the project include Daniel Grosu, associate professor of computer science; Tierra Bills, assistant professor of civil and environmental engineering; and Marco Brocanelli, assistant professor of computer science.
The award number for this NSF grant is 2043611.