Just like real-life twins, every digital twin is different. That’s because a digital twin is a virtual replica of a physical asset, a living replica.
Digital twin as a concept does not reflect one universal definition — and we get into trouble when we assume our audience shares our view.
I’ve discovered at least 11 different types of digital twin applied to three distinct phases — design, operation and maintenance — and that equals more than 30 possible use cases, just in the manufacturing space.
For example, some users rely on digital twin to optimize the design of a product or manufacturing process, while others use it to optimize production of a product or maintenance of a production line.
My goal here is to improve clarity around digital twin – how it’s referenced and applied in different scenarios – so users can more quickly and uniformly realize the value.
When we use simulation, we focus on the things that are important to us, the KPI we want to improve. For example, you might simulate the operation of a car to predict its miles-per-gallon (MPG).
It’s the same idea with digital twin, but it is a living “digital replica,” so that means its learning and changing. So with a digital twin, not only could you predict the car’s MPG when it drives out of the showroom, but you could predict how and when the MPG will degrade.