What is a Digital Twin?

Since then, the question of what is or is not a Digital Twin has been the subject of some debate within the scientific and engineering communities. At its simplest, however, a Digital Twin can be described as a virtual copy of a real object. Importantly, it is linked to its “physical twin” through a dynamic data feed; as a result, the Digital Twin is continually “learning” to be a better and more accurate replica of its physical counterpart. And because it’s virtual, you can do far more with the Digital Twin than you could with the physical thing.

In a Digital Twin truck, for instance, a fleet manager can monitor fuel levels and engine wear much more precisely—while the truck is on the road. They can get an evidence-based view into driver performance—while the driver is behind the wheel. And they can test the Digital Twin based on real-time data to predict potential breakdowns long before a Diagnostic Trouble Code (DTC) is triggered.

The history of Digital Twins begins more than 50 years ago, with a famous application of the still-unnamed concept during the Apollo 13 crisis. In 1970, the NASA spacecraft was on its way to the Moon—more than 200,000 miles from Earth—when an explosion disabled its propulsion and life support systems. Behind the scenes, NASA engineers fed real-time data into its mission simulators and used computers to model different strategies for maneuvering the craft and, ultimately, getting it and the crew back home.

The Apollo 13 case illustrates that Digital Twins do not necessarily need high-speed data networks or advanced computing capabilities to be effective. The computer network that NASA used to run its simulations had available common memory of only 256 kilobytes; the random access memory, or RAM, in most smartphones today is more than 15,000 times that.

That gives a sense of how much more powerful Digital Twins are now, decades after Apollo 13 was safely brought back to Earth. Advances in data transmission and storage, computing power and artificial intelligence (AI) have dramatically expanded the potential of Digital Twins and the range of their uses. Today, in a growing number of industries, it is not only possible but also, in many cases, economically feasible to digitally re-create machinery and test a very large number of scenarios, empowering the development and application of ever more advanced and accurate predictions.

Even 10 years ago, with the exception of spacecraft, airplanes and military equipment, relatively few assets merited the investment in time, money and IT required to develop a Digital Twin. Now, however, the democratization of data and computing power, along with broader access to AI, has made Digital Twins a viable tool for a much wider range of sectors and businesses. Meanwhile, many “connected” products, from household appliances to assembly line machines and transport trucks, can generate and transmit vast amounts of data that can then be used for Digital Twin development and use.

As advances in AI and computing technology continue, it’s clear that Digital Twin Technology—as powerful as it already is—is still in its infancy. As we look to its potential in the years ahead, the sky is literally the limit.