Digital Twin vs Systems Model?

Digital Twin

While this is a term with many different definitions across industry, Ron Giachetti states, "The digital twin is a high-fidelity model of the system which can be used to emulate the actual system. An organization would be able to use a digital twin to analyze design changes prior to incorporating them into the actual system." -SEBoK

The digital twin can be thought of as the "complete" digital representation of the "specific" physical system.  The term "complete" is relative; it can be argued that you will never have a complete digital twin because you could always add another aspect to the model or a higher level of fidelity.  The term "specific" is used because each system serial number is supposed to have it's own specific digital twin.  

This would be helpful on a long-term continuously running system like a factory where all of the data could help optimize the continuous maintenance frequency.

Systems Model

The systems model is referring to the family systems, an abstract model.  Which is incomplete but hopefully still useful.  As George E. P. Box stated, “All models are wrong, but some are useful…the practical question is how wrong do they have to be to not be useful.”

The goal of the systems model is to create a digital environment and system which you could accurately run new scenarios, new configurations,  and new environments, to understand how your system would perform.  If a good model has been made and accurate parameters are input, then there can be some confidence over time that the model output trends can tell a story and help direct engineers and managers toward the optimal solution faster.

Primary Takeaway

For most cases, the intent is to make a model and not a digital twin.  It is only beneficial to make the model just "good enough" to solve the problem you have.