What is it, where is it and what does it do?

[Samuel A Chorlton 33]

Can and should digital twins always be able to answer the question above?

[DRossiter87 72]

Is what were is the physical asset, or where is the digital representation?

[Steven Zhang 46]

If the purpose of digital twin is to represent something (either physical or non-physical), it should be able to answer what, where, and what for. I am visualising digital twin as an infrastructure of digital world, it need to have clear definition of what, where and why, so that other digital objects/ subjects/ agents/ systems etc. can interact with it. I guess, all of the concepts in the digital world need to be defined by people in some way. 

[GiovanniB 1]

I think it should always answer these three questions.

A digital twin could be created for testing or for a belief that could be used in the future for "something" even if it does not have an exact meaning at the moment it is created.

I would be curious to hear why it does not have to answer these three questions.

[Tom Hughes 47]

I would say it needs to. Answering these and a few other practical questions is what the base data requirement in COBie does and is then extended to cover other plain language questions via project data requirements. More info from the UK BIM Alliance - Project Data Requirements – UK BIM Alliance

Purposefully trying to think of an example where the "always" part of the statement might not apply... I could imagine scenarios involving mobile assets where always knowing location isn't practical or cost effective. A twin might know that something has left one place on route to another but hasn't arrived, has left a place but hasn't returned yet, or entered an environment where reporting on position isn't possible. The location isn't immediately unknown but a prediction that decreases in confidence as the time between known locations increases.

I think I have thought myself into the position that it might be OK for twins to lose things! in scenarios where the cost of knowing where everything is all of the time is greater than the cost of occasionally losing things.

I probably wont remember this logic next time I can't find my keys 😀

[Tom Hughes 47]

I thought of another example over coffee this morning - A digital twin that might start not knowing where things are and work to find them. It was reading an article about mandatory bike marking and the creation of a national bike database in France, and having had an expensive bike stolen in a scenario where I wasn't coved by insurance. 

The more I think about this the more I am moving to an on the fence answer that in most cases the answer to the question will be yes, but there are specialist cases where it could be a no.  

[Nigel Moore 1]

Yes I also think it should address / answer all three of these points, or at least be working towards answering all of them. A DT could be built initially to model and test process before the physical twin is finalised or even built maybe? 

[Andy Parnell-Hopkinson 8]

Like it, Sam. This is about the sanest/simplest description of a DT yet.

What is it - yes.

Where is it - yes, notwithstanding Tom's quantum asset. Schrodinger's bike?

What does it do - I contend that knowing what something is already answers this so I would substitute with "what is it doing". Caveat that the recency of "is" depends on the asset (ie how currently you need to know its state).

I'm now going to try to think of when this would not adequately describe a DT.