TUM and BMW Develop Digital Twin Against Traffic Hazards Using Artificial Intelligence
An artificial intelligence system developed through the collaboration of researchers and the automotive sector enables drivers to react faster to potential hazards.
A new artificial intelligence system developed in partnership with TUM researchers and BMW combines infrastructure and vehicle sensors to create an instant digital twin of traffic conditions, allowing drivers to respond more quickly to hazards.
AI-Powered Sensor Integration
As a BMW test vehicle approaches an intersection, infrastructure sensors and equipment on the vehicle collect surrounding data.
Cameras on the vehicle roof, radar and lidar units, and sensors on sign bridges feed the artificial intelligence system.
Real-Time Positional Accuracy
The screen inside the car shows road users from all directions to the driver with a positional accuracy of up to ten centimeters.
The artificial intelligence system continuously predicts the next movements of other road users in the vicinity.
Solution for Blind Spots
Professor Alois Knoll, head of the Chair of Robotics, Artificial Intelligence and Real-time Systems at TUM, evaluated the issue.
Knoll stated that thanks to this real-time digital representation, drivers can see areas where their field of vision is restricted.
Industry and Research Collaboration
Ralph Rasshofer from the BMW Group stated that this ongoing partnership highlights the mutual benefits between academic research and industry.
The developed technology constantly updates the system by blending data from vehicles and traffic lights with environmental conditions.
System Updated Ten Times Per Second
By combining collected database information and sensor inputs, the artificial intelligence updates the digital twin every hundred milliseconds.
This high update rate offered by the system allows instant tracking of sudden changes in traffic.
Work Areas and Patents
This innovative technology holds strong application potential, especially in construction sites where traffic patterns frequently change.
Researchers have already filed patent applications for this newly developed system and related innovations.