ANN ARBOR—On-demand, driverless public buses. Data-driven accident avoidance systems. "Smart" traffic systems that dramatically reduce emissions and congestion. Two new data science projects at the University of Michigan are working toward making these systems a reality.
Supported by the Michigan Institute for Data Science (MIDAS) Challenge Initiatives program and UM-Dearborn, the projects bring together interdisciplinary teams of researchers from both campuses to tackle the grand problems of the future of transportation using massive amounts of data being produced by automated and connected vehicle testing sites, as well as in conventional driver-directed settings, in Ann Arbor and around the country.
MIDAS recently awarded the two projects $1.25 million each under the first round of its Challenge Initiative program, with another $120,000 each contributed by UM-Dearborn. The funding is part of U-M's Data Science Initiative, which was announced in September 2015.
One of the projects, "Reinventing Public Urban Transportation and Mobility," led by Pascal Van Hentenryck of the College of Engineering, will help design and operate an on-demand, public transportation system for urban areas in which a fleet of connected and automated vehicles are synchronized with buses and light rail, using predictive models based on high volumes of diverse transportation data. The goal is to begin testing on the U-M campus within a year, and then expand the experiment to Ann Arbor and Detroit.
Van Hentenryck said one of the goals is to make public transportation a viable option for getting to and from work, health care, and other services for people who can't afford to own a car.
"We're trying to revolutionize mobility for entire population segments with poor access to transportation," he said. "On-call, affordable public transportation that can get you to and from work or the doctor's office efficiently would increase employment opportunities and result in better health care outcomes. The potential for improved quality of life is huge."
The project will use the University of Michigan, Ann Arbor and Detroit as testing grounds for these innovative transportations models, and will include collaboration with the U-M Parking and Transportation Services Department involving real-time data collection on driver behavior. The project will leverage massive amounts of mobility data to design and operate an innovative, on-demand transportation system that will address the "first-mile/last-mile" problem—that is, the challenge of getting people from their homes or final destinations into the transit system.
The system will be multimodal, using light-rail, shuttles, cars and bicycles, and will be well-positioned to add connected and driverless vehicles when they become available. U-M will become the first testing location within a year, with planned expansion to the rest of the city of Ann Arbor, and then to Detroit.
Van Hentenryck's collaborators on the project include researchers from the School of Information, College of Engineering, U-M Transportation Research Institute (UMTRI), Emergency Medicine, Architecture and Urban Planning, and Computer Science.
The other project, "Building a Transportation Data Ecosystem," led by Carol Flannagan of the U-M Transportation Research Institute, will create a system allowing researchers to access massive, integrated datasets on transportation in a high-performance computing environment. Research by Flannagan's team and others will support future transportation research and development.
Flannagan said creating a common repository of transportation data—including data on driving, traffic, weather, accidents, vehicle messages, traffic signals and road characteristics—will inform the development of connected and automated vehicle systems of the future.
"For example, real-world and simulated data on vehicle accidents will be invaluable to federal regulators developing regulations and guidelines for crash avoidance technology in the new generation of automated and connected vehicles," she said.
Flannagan's project includes researchers from the School of Public Health, College of Engineering, U-M Dearborn, College of Literature, Science and the Arts, UMTRI, and the Institute for Social Research.
"These interdisciplinary projects will push innovation in data science and transportation research in ways that will have long-term impact on the way people and goods will move for years to come," said MIDAS co-director Brian Athey, professor and chair of computational medicine and bioinformatics. "This includes studying the social and behavioral side of the equation. For example, will drivers and riders accept automated and connected vehicles?"
Ilir Miteza, associate provost at UM-Dearborn, said the MIDAS Challenge Initiatives program provides a valuable opportunity for collaboration: "These projects show the combined strength of our campuses to use emerging data science techniques to address society's grand challenges."
The goal of the multiyear MIDAS Challenge Initiatives program is to foster data science projects that have the potential to prompt new partnerships between U-M, federal research agencies and industry. The challenges are focused on four areas: transportation, learning analytics, social science and health science.