Digital Twins represent a digital mirror of a given physical reality, leveraging heterogeneous big data, machine learning algorithms and communication interfaces.
Design of a Digital Twin invokes a structural information modelling. Heterogeneous data necessitate usage of ontology for conceptualizing and formalizing a standard vocabulary to enable interoperability, maintenance and reuse of information. The scope of the information model determines the structure of the ontology, while the corresponding knowledge graph utilizes inference services for robust decision making in autonomous driving.


Digital Twin for Smart Services:

• Autonomous and Connected Driving
• Cloud Offloading of Decision Making using Machine Learning and Big Data
• Predictive Maintenance of Transport Infrastructure
• Efficient Route-Planning for Electrical Vehicles
• Health Monitoring of Roadbeds
• High Precision Localization

 
Segments of the Digital Twin are deployed to the vehicle by the digital infrastructure to improve the localization and real-time path planning. A further application of our Digital Twin concerns the detection of damages in roadbeds using data from a ground penetrating radar (GPR). When combined with a high precision positioning system, this system allows to locate damages on a three-dimensional map in order to extend the digital representation of the road substructure.