Virtual twins — Intersection of the real and digital worlds
Imagine being able to try out something beforehand so you know how it is going to work afterward. That’s the power of the virtual twin — a digital replica that lets you validate a design so you get it right the first time, saving time and money and speeding innovation. Virtual twins help us find the best solution, and reaching optimal solutions to the challenges of our society — whether they lie in medicine, industry, smart cities, or other areas — serves to “raise all boats.”
A virtual twin is a digital duplicate of a physical object, system or process that lets you examine all the variables and run through all the possible changes to optimize a solution. This digital representation shines a clear light on both the detailed elements and the operational dynamics of the entity that has been duplicated over its entire life cycle, from creation to retirement.
Throughout history, engineering and technological development has been based on physically creating something before you can test it for performance. The virtual, or digital, twin lets you study things without creating the “real” thing. You can then explore alternatives in search of the best solution — not feasible, time- and cost-wise, in the real world. And when the thing is operational in that real world, you can feed back performance data to update its digital twin.
Two technologies underpin the virtual twin concept: computational power and simulation platforms. Computational power has increased by 10 billion times since 1970 — the computer in your smartphone today is equivalent to the Cray-2 supercomputer of 1985. Similarly, simulation platforms now let you “sculpt,” in bits and bytes, digital twins based on the laws of mathematics, physics, chemistry and biology. One instance is in pharmaceuticals, which are now being developed by digitally representing and manipulating molecular structures and reactions.
The Living Heart Project is a good example. Cardiovascular researchers, educators, medical device developers, regulatory agencies, and practicing cardiologists are using a digital simulation platform from Dassault Systèmes to develop and validate personalized digital human heart models. These models are the basis for “in silico” medicine and can be used for training, designing medical devices, testing, clinical diagnosis, and regulatory purposes. “In silico” — wordplay for the silicon in computer chips — refers to a computer simulation that, in this case, speeds up innovation and how quickly advances are rendered into better patient care.
At Purdue, we’re using the virtual twin strategy at our Composites Manufacturing & Simulation Center. Composites — combinations of two or more different materials in wide use in industries like aerospace, automotive and construction — are coveted for their low weight, high strength, and durability. But their applications are limited by a time-consuming, costly process of trial and error. We’re creating digital representations of both composite manufacturing processes and end-product performance, “knitting” together multiple digital twins to explore conditions, materials, geometries and processes prior to developing an actual manufacturing system.
The digital twin has a bright future in industry and society, but it will require a substantial investment to reach its full potential. A good analogy is the decision to create the interstate highway system in the 1950s. The enormous benefits that society has reaped from this rapid transportation system are well documented, and the virtual twin has the same potential.
However, challenges also lie ahead in digitization. Perhaps first among them is the security of digital data — knowledge flows extraordinarily fast in digital form, and those who capture and control the virtual twin achieve real power with that acquisition.
Still, I am convinced that these issues will be addressed, and that the great benefits of the virtual twin will bring prosperity to those who develop it further.
R. Byron Pipes
Member, the National Academy of Engineering
John L. Bray Distinguished Professor of Engineering, and
Executive Director, Composites Manufacturing & Simulation Center,
Indiana Manufacturing Institute,
Purdue University
