Research Overview

Let’s imagine what a public urban transportation system for the future will look like: There will be a network of bus stops and transit stations strategically positioned connecting each neighborhood to the city’s business districts, hospitals and shopping centers.  There will be a fleet of buses, light-rail, shuttles, driverless cars and bikes, all controlled by an intelligent management system.  The system analyzes real-time data on the number of passengers at each location and their destinations and traffic conditions, and deploys the fleet to take every passenger to his/her destination in the most efficient and economical way.  No more walking for two miles before getting to the nearest bus stop.  No more congestion at the end of a workday.  No more loss of employment opportunities because two communities are not connected by public transportation.  We are at the brink of the next revolution in public urban transportation and mobility which needs to mitigate congestion, decrease environmental impact, reduce cost, expand service, reduce inequality in access to transportation and bring health and social benefits to inner cities and distressed neighborhoods.  Such a revolution is now made possible by ubiquitous digital connectivity, significant advances in autonomous vehicles, intelligent transportation and asset management systems that benefit from state-of-the-art data science methodology.

Aiming to enable a vision for future transportation, the research team carries out pioneering work in designing novel data-driven urban transportation systems and building the descriptive, predictive, and prescriptive technologies to power them.  The optimal design of these urban transportation systems will be informed by models for travel demand, accessibility, driver behavior, and transportation networks. These descriptive and predictive models will be derived by mining and fusing the rich and large data sets newly available and calibrated through interventions and machine learning. The envisioned transportation systems will be operated using real-time optimization algorithms and innovative coordinated traffic assignments to mitigate congestion, maximize network capacity utilization, and improve safety. The project will not only optimize costs, greenhouse emissions, and convenience, It will also strive to boost mobility for entire population segments and transform how to plan and manage a transportation infrastructure optimized for 20th century notions of human mobility.  We aim to start testing our system within a year on UM campus, in collaboration with the UM Parking and Transportation Services Department.  We will then expand our experiment to Ann Arbor and Detroit.

This is a multi-disciplinary team from four UM colleges, UM Transportation Research Institute (UMTRI), and the CDC Injury Center. The team has a history of innovation and deployment in intelligent transportation and asset management systems. The team has significant expertise in data science, from descriptive analytics to predictive and prescriptive analytics and interventions, and in the underlying economic and social mechanisms that are critical to successful deployments.

Recent Presentations

Pascal Van Hentenryck, project Principal Investigator, speaks at the 2016 MIDAS symposium:

Slides  from Pascal Van Hentenryck’s presentation at the 2017 MIDAS Transportation Symposium (PDF).


February 2018

  • RITMO-Transit: the team has designed algorithms for a multi-modal, on-demand mass transit system, including network design, trip splitting, zone allocation, fleet sizing, driver rostering, as well as the real-time dispatching, routing, and car-sharing. The first deployment (on U-M North Campus) happened in January, 2018. Collaborating with AAATA, the team is now designing a new bus network for Ann Arbor and Ypsilanti.
  • RITMO-Commute aims to design a massive trip-sharing architecture. The team has shown that real-time optimization is essential to this effort.
  • RITMO-Privacy concerns the release of sensitive datasets for benchmarking, competition, or procurement settings. The team has proposed a “constraint-based” differential privacy policy that leverages the post-processing immunity of differential privacy to obtain orders of magnitude improvements in accuracy for mobility and energy applications.
  • RITMO-Survey focuses on evaluating traveler responses to the RITMO-Transit system upon its implementation on the U of M Ann Arbor campus. The team has received 4,473 survey responses, which will provide information on respondents’ current travel behavior and their likelihood of using RITMO-Transit vs. driving, walking and biking under different deployment scenarios.
  • The team has filed five patents, and its mobile apps have been deployed through Apple and Google stores.
  • RITMO has gathered a large-scale dataset on the travel behaviors and travel preferences of the U-M population for their commute trips and inter-campus trips, which may be used by other U-M researchers.

Previous Updates

July, 2017

The team has developed and simulated an on-demand, multimodal transit system for Ann Arbor and is ready to deploy it.  The system improves convenience, cost, and accessibility.

Research Team

  • Pascal Van Hentenryck, Seth Bonder Collegiate Professor of Industrial and Operations Engineering
  • Ceren Budak, Assistant Professor, School of Information
  • Patrick Carter, Assistant Professor, Emergency Medicine
  • Amy Cohn, Arthur F. Thurnau Associate Professor, Industrial and Operations Engineering
  • Tawanna Dillahunt, Assistant Professor, School of Information
  • Robert Hampshire, Assistant Research Professor, U-M Transportation Research Institute
  • Jerome Lynch, Professor, Civil and Environmental Engineering and Electrical Engineering and Computer Science
  • Jonathan Levine, Emil Lorch Collegiate Professor of Architecture and Urban Planning, Taubman College of Architecture and Urban Planning
  • Louis Merlin, Dow Sustainability Postdoctoral Fellow, Taubman College of Architecture and Urban Planning
  • Luis Ortiz, Assistant Professor, Computer and Information Science, UM – Dearborn
  • Jim Sayer, Director, U-M Transportation Research Institute
  • Michael Wellman, Lynn A. Conway Collegiate Professor of Computer Science & Engineering