By Anil Sawhney
Another interesting ClubHouse room discussing Digital Twins. During the discussion, once again there was a lot of debate over the definition of what is a Digital Twin.
Especially for our built assets, the terminology is used to describe many different, often opposite, outputs which keep firing long debates among professionals and cause disappointment by clients who are, more than often, not getting what they thought they were asking for.
Anil used a simplification to explain the concept (pictured above) which was very well received by the audience but also followed but a long discussion over Is this a Digital Twin?
There are many definitions of Digital Twins, Anil decided to test the example with the definition used by the Digital Twin Consortium (DTC):
A digital twin is a virtual representation of real-world entities and processes, synchronized at a specified frequency and fidelity.[1]
First, we can look at the foundational elements of this definition: virtual representation, real-world entities and processes, synchronization, [synchronization] frequency, and [synchronization] fidelity.
1. Virtual representation: The app is a virtual (in other words, digital) representation of the car (door locking mechanism) for the given use case. But it’s not a 3D model. Do we need a 3D model (BIM) for this to be a digital twin (DT)? Strictly following the DTC definition, the app is a representational model of the car door representing its state (locked or unlocked) and process (locking or unlocking). So, in this way, the example passes a part of our assessment for a DT.
2. Real-world entities and processes: The car door (and its locking mechanism) is the real-world entity, and the locking and unlocking of the door is the process. So, the example passes on this foundational element as well.
3. Synchronization: This element of the DTC definition is crucial to our test. In some definitions, this is also called “connection” between the digital and physical. This distinguishes a digital twin from any other digital model or replica (e.g., a BIM model). At the same time, DTC provides flexibility in the actual definition of synchronization, leaving the connection unidirectional or bi-directional. In our example, the connection or synchronization is bi-directional, i.e., the app senses the door being unlocked.
The human in the loop is alerted. If the decision by the end-user is made to lock the car, the app can actuate this action and lock the car door. It seems a unidirectional connection could also work, i.e., the vehicle sending a notification to the app that it is unlocked, and the end-user walks to lock it. The car example is compliant with this foundational element also.
4. Frequency and Fidelity: The app is synchronized with the state of the car door periodically (with some delay) so that its state can be detected and relayed to the end-user. The app also meets the fidelity characteristic of a digital twin, as explained by DTC. The degree of precision applied to the virtual representation (the app), and its synchronization mechanism seems appropriate for the given use case.
Now, let's consider the crucial sub-parts of the definition:
● Digital twin systems transform business by accelerating holistic understanding, optimal decision-making, and effective action.
This example does not meet all the items on this part of the definition, even though it can lead to “effective action.”
● Digital twins use real-time and historical data to represent the past and present and simulate predicted futures.
The example fails this sub-part as no historical data is available or used, and predictions cannot be made.
● Digital twins are motivated by outcomes, tailored to use cases, powered by integration, built on data, guided by domain knowledge, and implemented in IT/OT systems.
The car example provided here seems to be in line with this subpart.
While the example passes the test of the foundational elements (the what and the how), the example does not do well on the sub-parts (the why). Based on our test we therefore should declare that the example above is not a digital twin? Or are we just using the wrong definition to validate our example?
Join the next ClubHouse room and tell us what you think! and keep an eye on the next blog posts by Casey Rutland discussing, Do we need a digital twin test?