A free virtual digital twin of its physical AV testing site, providing a safer, faster, and cost-effective platform for testing autonomous and connected vehicle technologies.
The University of Michigan’s Mcity Test Facility has unveiled the first open-source digital twin of its renowned test track, enabling researchers worldwide to simulate and test connected and autonomous vehicle (CAV) technologies. This groundbreaking tool, freely accessible, offers a safer, faster, and cost-effective alternative to real-world testing.
A digital twin is a virtual replica of a physical environment that interacts with real-world data, enabling advanced simulations. The Mcity digital twin represents a significant milestone in mobility systems testing, including autonomous driving. Researchers can now virtually access Mcity’s comprehensive testing environment, featuring diverse road materials, markings, intersections, and traffic signals, without traveling to Ann Arbor. This takes our nearly 10-year-old track and puts its digital counterpart directly over it. It’s a living, breathing representation where mixed reality testing and development thrive.
The digital twin integrates with TeraSim, Mcity’s open-source traffic simulator, to create dynamic testing scenarios. It introduces pedestrians, cyclists, and vehicles while generating safety-critical events, such as potential collisions. Using traffic models calibrated with real-world data, it simulates both everyday and high-risk driving scenarios.
Experts view CAV technologies as key to safer, more efficient transportation, though substantial research is still required to achieve widespread benefits. Virtual testing accelerates progress, enabling autonomous vehicles (AVs) to log “millions of miles” in simulations before real-world deployment. “With the digital twin, we control variables like simulated pedestrian traffic, focusing and accelerating testing,” said Mcity software engineer Darian Hogue.
Opened in 2015, Mcity was the world’s first purpose-built proving ground for CAVs. Its physical features include urban streets, a traffic circle, diverse road surfaces, and accessibility ramps. In 2022, Mcity debuted remote-use capabilities, enabling researchers to control vehicles and infrastructure on-site via cloud-based systems. Using 5G connectivity, real-time data flows seamlessly between remote researchers and Mcity’s physical environment.
“As an open-source tool, the Mcity digital twin reduces barriers to entry, allowing developers to refine systems before transitioning to on-site tests,” said Mcity Director Henry Liu.