HOW does a Digital Twin work?
The SAP Digital Twin implementation will at least comprise the following parts:
- Edge capabilities for observing key aspects of the real asset’s state and behavior. This typically implies sensors with corresponding edge processing capabilities for data quality enhancements, such a calibration, filtering, time synchronization, etc.
- The Digital Twin core runtime, using the input stream from the edge to render a (near) real-time digital reflection of the asset’s state.
- The Digital Twin consumption layer, that subscribes to selected data streams from the Digital Twin for various applications. This can be specific end-user applications for monitoring and control, it can be legacy applications for maintenance and asset management, or the data stream from the twin might feed into data analytics and machine learning stacks for pattern recognition and decision support.
General Digital Twin concept description
The technology concept used in the SAP Digital Twin is based on the principle of observations and stimulations. Observations are made by physical sensor measurements on a real system. Actuator motions governed by sensor measurement data are applied in a simulation model by stimulating the Digital Twin of the real system, as illustrated in Figure 1.
Figure 1: The general Digital Twin concept used in SAP Leonardo