Digital Twin (DT) is a cutting-edge technology that promises to greatly reduce waste, increase safety and reliability. A DT is a virtual representation of a physical entity (e.g. component, equipment, manufacturing line, or entire production facility) that can be monitored and compared against the real thing. A DT can be used in real time, collecting data from physical equipment, to detect anomalies and predict failures before they occur, and it can also be used at an accelerated pace (and with many copies in parallel) in a completely virtual environment. When a DT is used in a virtual environment, it enables testing “what if” scenarios much more efficiently and safely (without the risk of harm to people or damage to property) and can provide insight that greatly improves quality, efficiency and safety. And since a DT is virtual, it can be tuned and improved over time to provide better feedback and more effective predictions.
The concept of a DT shares much in common with virtual V&V applications. Both virtual V&V and DT represent physical systems as realistically as appropriate, a true Digital Twin compares inputs from a real physical system being “twinned” and compares against the virtual twin to look for differences in the data and from those differences glean knowledge about the system. The Digital Twin will continuously or periodically update its knowledge base by feeding maintenance data, and other gained knowledge about the system, back into the Machine Learning (ML) algorithms parameters. With the explosive growth of Internet of Things (IoT) technologies, the Digital Twin can be interconnected to the physical system or process in ways not previously possible. This real-time connectivity is a significant differentiator between a Digital Twin and a virtual V&V application. However, to be highly-effective, both a virtual V&V capability and a Digital Twin capability will leverage detailed functional, logical, and physics-based dynamics and software-in-the-loop dynamics to include some of the most complex and time consuming verification efforts.
There are three distinct types of DT:
Type 1: Digital Twin for MBSE Systems Development
Type 2: Digital Twin for What-If Analysis
Type 3: Digital Twin for Live Monitoring
Significant economies-of-scale cost and schedule benefits can be achieved by developing a comprehensive strategy to make use of multiple types of DT capability, potentially across both product and process, that leverages and shares modeling and data assets across each capability and development initiative.
For example, a Type 3 live monitoring DT, data is taken from a physical asset (such as an in-service aircraft), a virtual asset (such as a virtual aircraft), a set of aircraft subsystem and environment models. For the aircraft example, as the pilot flies, pilot control inputs are fed in real-time to the virtual plane models. The physical plane is compared to the virtual plane in real-time and the differences are used to better understand the system, e.g. predictive maintenance, predictive performance, anomaly detection.
Digital twin technology is going to be a competitive advantage. Digital twin is a relatively new concept in digital engineering and digital transformation initiatives, but it is quickly evolving and expanding in breadth and reach. Demonstrating that digital twin can be backward integrated into existing development and manufacturing operations including engineering, operations, maintenance, and supply chain logistics in a cohesive architecture using unifying work processes will provide the evidence needed for confident implementation in production environments.
A functional digital twin allows the organization to more quickly evaluate operations relative to market conditions and test assumptions for innovating and improving capabilities, therefore improving market position in the competitive global environment.
