Our collective understanding of digital twins is rather nascent. To ensure that we operate under the same base information there is a need to periodically reflect on the concepts and principles we have outlined. This blog post is one in a series which reflects on previously published concepts to consider whether our collective thinking has advanced.
As we develop the thinking, tools, and resources relating to digital twins, a lot of discussion is taking place regarding their scope, scale and accuracy. Within the Gemini Principles it stated that a digital twin is:
A realistic digital representation of something physical
I want to reflect on this statement. In particular, the use of “realistic”.
For something to be realistic, according to the Oxford English Dictionary, it must represent something in a way that is accurate and true to life. For example, for something to be “photo-realistic” it must appear as if it was a photograph.
However, the Gemini Principles state that a digital twin must represent physical reality at the level of accuracy suited to its purpose. Interestingly, while undertaking discovery interviews with DT Hub members we saw this issue realized.
"Several members commented on how people in their organizations would try to extend the use of their digital twins beyond their intended purposes."
This was seen as both a positive and a negative outcome. The positive being that members of these organizations saw the value in these digital twins and wanted to harness their insight. The negative being that these digital twins did not have the information or, when available, did not have level of accuracy required to be used for these extended purposes. For these extended needs, these digital twins were not realistic.
Amongst DT Hub members there appears to be a shared view that digital twins are, fundamentally, purpose-driven. Therefore, digital twins might not be “real” representations, but instead the “right” representation to support a purpose.
Consider an example. An air traffic control system utilizes a “digital twin” of runways, aircraft and their flight paths along with sensor information (e.g. weather and radar) to assist with preventing collisions, organize and control the landing and departing of aircraft. In this example while real-time information and analytics are used, none of the physical elements (planes, control towers) have realistic representations, they instead use basic representations to support the air traffic controller. Instinctually an air traffic control system does everything we want a digital twin to do, it is a digital representation of physical assets which also includes sensor information where the physical assets provide a link back to the digital twin. Given this, it should be fairly clear that an air traffic control system would be considered a digital twin. However, this does not appear to be the case.
A poll was placed on twitter asking “would you consider an air traffic control system a digital twin”. After 62 votes were cast, the result was exactly 50:50. What does this tell us? Perhaps public messages on what a digital twin is aren’t sufficiently defined? Perhaps the question was poorly worded? Or perhaps, for some, the lack of a realistic representation is the reason they said no? Unfortunately, context for each vote isn’t available. At the very least we can be sure that our shared view may not be shared by everyone.
In an age where many consider data to be the new oil perhaps we should consider using our data sparingly. So long as the data provided is sufficient for its intended purpose, a realistic representation may not always be required.
And there we have it, realism and its place within Digital Twins. Do you believe that a digital twin has to be realistic? Can something be a digital twin without being a realistic representation? Had you voted on this poll, would you have considered an air traffic control system a digital twin?