Is now the time to stop using the words ‘Digital Twin’?

[Robert Prince-Wright]

I would like to be provocative for a moment but please bear with me. I am wondering if now is the right time to bury the words ‘Digital Twin’ despite the fandom and homage. Let me explain why. I spent a year supporting a Digital Twin Consortium (DTC) Working Group and observed how the Taxonomy team struggled to create a glossary that all could agree on. Part of the problem stems from the use of the word twin. It suggests there are two entities, and that one is digital. A digital twin is therefore, wait for it, the digital sibling, and not the actual thing (or actual twin.) Others have brushed this complaint aside, however, this loose use of language results in CAD, BIM, Lidar scans and the rest being called digital twins when in fact they are nothing more than virtual representations (assuming you use DTCs definitions.) There is so much more I can add but now is a good time to hear what you all think, my only request is that those who are compelled to comment be thoughtful and substantive.

Edited December 9, 2022 by Robert Prince-Wright

[Brian Fugeti]

Hi Robert

Ive just joined the Hub and after looking at the community messages i was compelled to reply to this post. I have small experience in this field but have always tried not to use the word "Digital Twin" especially in the Build Environment where i work. It does look like everyone has their own definition and this seem to make the term lose its value.

I have looked at the following paper on Science Direct - Kritzinger, Werner; Karner, Matthias; Traar, Georg; Henjes, Jan; Sihn, Wilfried (January 2018). "Digital Twin in manufacturing: A categorical literature review and classification" where they have defined the terms according to the level of data integration (between the physical device and its digital object) as the following -  an interesting concept.

Digital Model (DM) - Manual data flow both ways

Digital Shadow (DS) - Automatic data flow to the digital object but manual data flow back to the physical device

Digital Twin (DT) - Automatic data flow both ways

This is also subject to different opinions and interpretations. I'm quite keen to find out what other professional think and wat other terms they have used to replace Digital Twin.

I have used "Digital building and Analytic Platform"

[Dan Fowler]

Misuse of technology terms is common, and Digital Twin is another example. The term Digital Twin is now embedded and unlikely to be dropped. However, it must be used correctly. Computer simulations have existed since the birth of the digital computer. A computer's reason for being is to run a model, from a simple alogrithm to a complex simulation. Even using email and writing a document in a wordprocessor is a model of previoulsy manual tasks. A Digital Twin is a step up in fidelity and feedback, it needs to be a live model of a cyber-physical systems, at any scale. It is more than a simulation or model. It can be used for real-time querying and analysis and running what-if scenarios without impacting the system it is representing until changes are pushed out. The twin part is refering to the properties of the real system that the digital system can provide and mimic. Yes, it is not a twin in the true meaning of the term, but it is a virtual representation of a physical system and is enough of a twin to be useful. It is up to those who know the difference between an algorithm, model, simulation, and twin to point out when terms are used incorrectly.

Dan

[Robert Prince-Wright]

Hi Dan, I would recommend you read the terminology developed by Digital Twin Consortium. It attempts to create robust definitions, but fails on deeper inspection. While it's true to say DT is widely used terminology it makes no sense to promote something which fails the most basic grammar test. Your post makes my point for me, e.g. "Yes, it is not a twin in the true meaning of the term, but it is a virtual representation of a physical system and is enough of a twin to be useful."  You're incorrectly conflating virtual representation with the concept of DT which is, at its heart, the synchronization of something actual with something virtual. 

Personally I would like to see DT die on the vine, and at some point be replaced by a sound and more meaningful terminology. I dabbled with the idea of referring to 'Digital Twinning' while a member of DTC and note Grieves has used that term recently - it's still not good enough tho.  

FYI - I opted to make the original posting after noting now DT had dropped off the 'radar' for two of the most influential groups in this arena. I am now increasingly concerned the emperor has no clothes and, worse still, is distracting us and wasting a lot of time and money. E.g. conferences costing millions of dollars that achieve nothing, and government grants that waste taxpayers money. The hard part of twinning is being starved of funding whilst the tens of millions are spent on fluff and eye-candy. Perhaps this is the reason CDBB was closed down?

[Dan Fowler]

It reminds me of the rise in the use of "architecture" in the software and business world. I'm sure true architects still bemoan the rise of "software architects" and "business architecture" and numerous other corruptions. The metaphorical use of architecture, as Wikipedia calls it, has stuck, as I think the metaphorical Digital Twin will stick. I think the Digital Twin Consortium definition is reasonable on the concept, but as you rightly point out, there is a lot of fluff that needs to disappear and more done on the real engineering of DTs. I think we can all try and point out when something is merely an algorithm, model, or simulation when someone is trying to say it is a DT.