How it works
Think of digital twins as models that duplicate physical objects to help you visualize things not yet built, track the performance of existing items, and train workers on how to use equipment safely and efficiently.
For existing parts or products, digital twins work by integrating connected sensors into the physical component and collecting real-time data on its performance. The data collected by some sensors is analyzed by AI and machine learning algorithms to identify patterns and make predictions about the item’s behavior. For example, in manufacturing, an engineer or operator based in California might remotely collect data about the temperature, pressure, and vibration of machinery located in Pittsburg. She then could run a variety of simulations on its digital twin to identify ways of optimizing performance, minimizing equipment downtime, and making real-time adjustments to keep everything running smoothly.
Digital twins can also simulate the performance of product prototypes. By integrating data from various sources, such as CAD (computer-aided design) files, you would create a working model for testing various product attributes. Automakers, for example, might simulate the aerodynamics of a car and optimize its design for maximum fuel efficiency. Similarly, an aerospace company might digitally replicate the behavior of a new plane in various weather conditions.
The a-ha moment
The history of digital twins can be traced back to the US space program of the 1960s when NASA used something called pairing technology for remotely improving the operations and maintenance of its space vehicles and systems. In fact, NASA used this technology to simulate solutions for returning its crippled Apollo 13 module and crew to Earth in 1970.
The specific concept of digital twins was later articulated by David Gelernter, a Yale computer science professor, in his 1991 book Mirror Worlds: or the Day Software Puts the Universe in a Shoebox. But Dr. Michael Grieves, formerly of the University of Michigan, is widely credited with applying the concept of digital twins to product lifecycle management (PLM) in 2022. He theorized it was possible to have all information related to physical objects residing “within a digital representation.”
What it means for everyday life
Digital twins are already improving design, maintenance, and training in just about every industry. Engineers are using it to improve jet engines. Doctors are using it to map activity in the human heart and simulate how patients might react to various treatments. The technology has been employed to optimize production in oil and gas fields and has even improved the performance of Formula 1 race cars.
How it might change the world
As Industry 5.0 technologies continue to evolve, it’s likely most — if not all — parts and products will have a computer doppelganger telling people how to make improvements or use them more efficiently. Digital twins are also seen as invaluable tools for urban planning (for example, smart cities). Heck, some people think human beings will have digital twins connecting with technology we wear to help doctors, clinicians, and even physical trainers remotely model and track what’s going on in our bodies at any given time.
