A collaborative effort between Prof. Kei Sakaguchi's team at Tokyo Institute of Technology and Prof. Walid Saad's team at Virginia Tech has led to the development of a Smart Mobility Digital Twin. This innovative digital replica mirrors real-time traffic conditions, facilitating the seamless integration of self-driving technology and remote vehicle operation. Their findings, published in IEEE Transactions on Intelligent Vehicles, mark a significant advancement in mobility technology.
While digital twins—cyber replicas of physical systems—have been employed in industries like manufacturing and construction, their application in the dynamic field of mobility had been limited until now. The research leveraged Tokyo Tech's Smart Mobility Education & Research Field, which was built specifically for this project, to construct a digital twin that powers a hybrid autonomous driving system.
The research team successfully demonstrated hybrid autonomous driving, where real-time data from the digital twin was used to guide vehicles, identifying safer, more efficient routes. This confirmed the feasibility of combining local autonomous control with remote guidance, making autonomous driving both safer and more efficient.
This system merges local path planning, driven by the vehicle’s sensors, with global path planning from the digital twin’s broader perspective. Through V2X communication (vehicle-to-everything), vehicles can share information with infrastructure and other vehicles, leading to improved traffic safety and efficiency.
Digital twins have expanded from manufacturing to sectors like healthcare and agriculture, offering advantages such as real-time monitoring, prediction, and anomaly detection. However, implementing digital twins in mobility is challenging due to the highly dynamic nature of traffic environments.
Despite these challenges, Tokyo Tech and Virginia Tech, supported by Japan's NICT and the U.S. NSF, have been working since 2022 to build this Smart Mobility Digital Twin as part of a broader initiative to realize Society 5.0—a future society where the physical and digital worlds converge.
The Smart Mobility Digital Twin system consists of autonomous vehicles and roadside units (RSUs) equipped with advanced sensors such as LiDAR and cameras. This infrastructure creates localized digital twins of the environment, while the cloud aggregates data from multiple vehicles and RSUs to create a global, wide-area digital twin.
The Ookayama Smart Mobility Digital Twin provides real-time collision prediction and route planning, supporting safe driving through infrastructure coordination. With edge servers and cloud computing working together, the system can process real-time traffic data, allowing autonomous vehicles to make informed decisions and adjust their routes accordingly.
For the first time globally, a hybrid autonomous driving system has been practically demonstrated. The system combines the autonomous vehicle's local observations with the digital twin’s global view, offering real-time traffic insights from a bird's-eye perspective. In the demonstration, the system detected a parked vehicle and pedestrians, prompting the autonomous vehicle to adjust its route in real time for a safer and more efficient path.
This breakthrough shows the potential of integrating digital twin technology with autonomous vehicles, opening the door to future smart mobility solutions that enhance both safety and efficiency. The ongoing collaboration between Tokyo Institute of Technology and Virginia Tech promises to push the boundaries of what’s possible in mobility.
