Digital Twin Definition

[LaurenMcG]

Hello all, 

Is there a resource that I might have missed that covers definitions? My question is somewhat obvious 'What is a Digital Twin?'

I have been exploring various resources to understand the concept of a Digital Twin. I know it is a well discussed topic, but I still see conflicting definitions. Some have their own definitions, some of you prefer not to get bogged down in the definition, and some who do not think it is a crucial aspect. I do believe it is important, as it helps those to spot those improperly marketed products and drive innovation in the field. In terms of getting bogged down in it, I do not think that a well understood definition is counterproductive. 

I come from an aerospace/mechanical engineering background, and my go to DT definition is the one in  ISO 23247:1 (Digital twin framework for manufacturing): 

"<manufacturing> fit for purpose digital representation of an observable manufacturing element with synchronization between the element and its digital representation"

The wording of 'with synchronization' here makes me under stand that there is a physical element established that has synchronization with its digital counterpart.

Now there is the upcoming ISO/IEC 30173 (Digital twin — Concepts and terminology), which is to offer an understanding of DT across industries. I believe it is still under development/review. The current definition of DT in it is 

"digital representation of a target entity with data connections that enable convergence between the physical and digital states at an appropriate rate of synchronization"

Note 1 to entry: Digital twin have some or all of the capabilities of connection, integration, analysis, simulation, visualization, optimization, etc.

The wording of 'enable convergence' here makes clear that the digital entity has capability to communicate with the physical entity, whether it is established or not. I.e., you can have a simulation but the capability to connect to the physical asset must be there. I've come across the Digital Twin Hub's Digital Twin toolkit, which includes a definition that might be seen as accepting a digital model without connectivity capability as a digital twin:

"A computer model as a digital representation of a physical asset (e.g. a 3D model of a housing block);
A digital asset where the physical twin may exist or may not yet exist (e.g. a model of an unbuilt tunnel);"

I am aware of other standards related to digital twins, but the question is the same: Is the Digital Twin Hub therefore accepting that a digital model, without connectivity capability, is a digital twin? I am of the understanding that a digital model facilitates a digital twin, can be eventually integrated with the physical model. Would stating that the digital model/asset is a digital twin not undermine the capability, purpose, and benefit of a digital twin? Would this not undermine the goals of the The National Digital Twin programme?

 

I also want to note that I am not worrying about digital twin maturity in terms of digital twin representation of an asset, facility, city etc., as the question regarding digital model is still the same. 

Please let me know your thoughts.

Thanks,

Lauren

Edited August 2 by LaurenMcG
editing

[Melissa Zanocco]

There is a definition in the Gemini Papers in the What Paper.

"A digital twin is a digital representation of a physical asset, process or system1 . It is distinguished from any other digital model by its dynamic connection to the physical twin. A digital twin unlocks value by supporting improved decision making."

It then expands on this.

[LaurenMcG]

@Melissa Zanocco Great, thanks. Good to see the comment about the dynamic connection. I think the ISO definition ISO/IEC 30173 covers the basis, and so perhaps it can be accepted that as long it has the ability to connect to the physical asset is a digital twin, but the real purpose and value of a digtial twin is when it is dynamically connected. 

[Robert Prince-Wright]

Good post Lauren.

I don't think we can brush off the poor epistemology and ontology that is routinely being used to describe the assemblages of technologies in today's toolbox. It has, and will, lead to a waste of public funds, and worse still poor implementation. Given you work in aerospace I presume you are aware of the depth of thought being given to ontologies by Airbus and Boeing.  

Robert 

[DRossiter87]

Hi Lauren, I must say my preference also goes to the ISO 23247 definition.  Both 'fit for purpose' and 'with syncronization' do a lot of (positive) heavy lifting.

 

[Muthu]

What is the difference between  Digital Twin and  Simulation

[LaurenMcG]

@Muthu Digital Twin draft standard "BS ISO/IEC 30173 Digital Twin - Concepts and terminology" defines

Simulation: use of a similar or equivalent system to imitate a real system, so that it behaves like or appears to be the real system [SOURCE: ISO 16781:2021, 3.1.9]

Digital Twin (DTw): digital representation of a target entity with data connections that enable convergence between the physical and digital states at an appropriate rate of synchronization. Note 1 to entry: Digital twin have some or all of the capabilities of connection, integration, analysis, simulation, visualization, optimization, etc.

So I would understand that you can have a standalone simulation, e.g. discrete event simulation using a software of choice, but if that software of choice enables convergence to the physical world, then it can be classed as a DTw. For example, Visual Components is a 3D simulation software for manufacturing. You can create a simulation of a manufacturing process. But it also has the capability to connect to PLCs and robots and OPC UA, e.g. Doosan and UR, and with this you can establish convergence with the physical world.