What is a Digital Twin?
We often get asked the question – What is a digital twin, and more importantly, why is it important?
Let’s start with the Wikipedia definition: A digital twin is a digital replica of a living or non-living physical entity. Digital twin refers to a digital replica of potential and actual physical assets, processes, people, places, systems and devices that can be used for various purposes. The digital representation provides both the elements and the dynamics of how an Internet of things (IoT) device operates and lives throughout its life cycle. Digital twin definitions emphasize two important characteristics. Firstly, each definition emphasizes the connection between the physical model and the corresponding virtual model or virtual counterpart. Secondly, this connection is established by generating real-time data using sensors.
As can be extracted from the definition, there are two relatively distinct concepts in the Wikipedia definition, and in practice with two very different use cases.
In the first part of the digital twin definition, a virtual model is for purposes of computer simulation of the real-world system or device. The model is created to run as an ‘executable model’, driven by stimulus or various types of tests. The output of the simulation produces data which replicates the corresponding real-world device. While this approach provides infinite possibilities to testing, what if scenarios, and behaviors, it turns out that the output is only as good as the fidelity and accuracy of the (executable) model. The primary use model for this approach is when you want to test numerous design options to see which is the most optimal.
The second definition utilizes sensors to represent the device, asset, or system. This definition is more suited to the way most think about IoT. The challenge here is having enough sensors to properly represent all aspects of the asset. The subtle part of this second definition, but the most critical, is the organization of data into an object orientation. The primary use case for this second approach is where you have established and well proven designs, but have a requirement to monitor the performance in the field. Sensors with digital twin object orientation is the enabler for increased IoT driven optimizations.
The Preddio Digital Twin Approach
At Preddio Technologies, our proven approach to digital twin implementation is enhancing existing designs with augmented analytics. Starting with a well thought out data object, the digital twin analytics transforms maintenance and management team’s ways of work, providing an extremely useful tool for comparisons, analytics, what if scenarios, and meaningful traceability. Compared to simply collecting unorganized raw sensor data, our approach abstracts various sensor outputs into a digital twin framework making your assets come to life in meaningful ways. While there are many trade-offs to consider, the key benefits are proven to be:
- Predictability – unlike real world devices, implementation of the digital twin enables real world machine learning algorithms to aggregate operational and performance data, learn from it, and predict the future.
- Analytics – the digital data from the digital twin allows traceability, history, and data analytics of the real-world device, thus making decision making less obscure, and more data driven.
- Optimization – with more (organized) data comes more opportunity to enhance the surrounding process where the digital twin is installed, enabling optimizations that were previously hidden.