1. Sales & Marketing
  2. Finances
  3. Your Team
  4. Technology
  5. Social Media
  6. Security
Product and service reviews are conducted independently by our editorial team, but we sometimes make money when you click on links. Learn more.
Grow Your Business Technology

You've Heard of IoT and AI, but What is Digital Twin Technology?

digital twin technology
Credit: Sergey Nivens/Shutterstock

Topics like artificial intelligence (AI), the internet of things (IoT), and machine learning are getting lots of hype, but digital twin technology might just be the real game-changer. Digital twin software uses aspects of all the trending tech mentioned (AI, IoT, ML) in a unique way that's changing the way businesses optimize production and investment, and the big boys are already heavily invested.

A digital twin is a highly advanced simulation that's used in computer-aided engineering (CAE). It's a digital duplicate that represents a physical object or process, but it is not intended to replace a physical object; it is merely to inform its optimization. Other terms used to refer to digital twin technology include virtual prototyping, hybrid twin technology, and digital asset management, but digital twin is quickly winning out as the most popular name.

Both NASA and the United States Air Force are planning on using digital twin technology to create future generations of lightweight vehicles that are sturdy and able to haul more than their current counterparts. Goldman Sachs recently examined digital twin technology in their series "The Outsiders," which seeks to identify "emerging ecosystems on the edge of today's investable universe." IBM is already on the digital twin train, combining AR with digital twin optimization and visualization. And SAP recently launched SAP Predictive Engineering Insights, a software product that enables organizations to use digital twin technology for digital asset management.

Independent research companies are steadfast in their belief in digital twins, too. Gartner predicts that by 2021, "50 percent of large industrial companies will use digital twins, resulting in those organizations gaining a 10 percent improvement in effectiveness." So, while digital twin technology isn't getting the same level of media attention as 3D printers or voice assistants, there is little doubt that it will be a game-changer and become an integral part of how all businesses of the future (not just manufacturing) will optimize processes, products, and communication.

Some industry experts argue that a true digital twin must have a physical counterpart in the real world. Indeed, for most of the (admittedly short) history of digital twin technology, this has been the case. Such twins (that have physical counterparts) are sometimes referred to as data-driven digital twins, because they rely on connected devices that make up IoT technology (often in the form of sensors and integration with other tracking systems and databases) to collect past and current data, test new configurations and processes, and inform future decisions.

There's also a crop of digital twin technology that's relatively more affordable, which uses digital simulation in a predictive way without a physical counterpart. This secondary type of digital twin is sometimes called a model-driven digital twin, but it can also go by the standard digital twin moniker.

The reason the second type of digital twin is less expensive is twofold: First, there are no IoT sensors or setup, since all the testing and tracking is happening digitally. Second, businesses can try out different processes and view outcomes prior to investing in actual infrastructure, technology or resources.

Imagine a factory that produces automobile tires. Within that factory, there are several pieces of highly expensive machinery, each one responsible for a different task related to making tires. Now imagine you own that factory, and you want to figure out how to optimize the production process.

Creating a digital twin allows you to do that without disrupting daily operations and with far lower overhead than previous testing and optimization methods. Rather than using trial and error, hiring an outside consultant to implement general best practices, using traditional statistical analysis, or investing in new methods with the assumption that the quoted results are accurate, digital twin technology can be employed.

Using digital twin software, you could create a digital version of your factory, with every piece of machinery included and functioning identically to the IRL version. You could run virtual tests to see how production would be affected by repositioning the machinery, updating a machine or part of a machine, updating half the machines, replacing several older machines with one newer machine, hiring more staff, firing half the staff, etc. You could also run a cost benefit analysis to determine if investing in the latest Tire Master 5,000 (not a real product, but you get the picture) is worthwhile. Digital twins let you test and optimize in a virtual world that has built-in AI and ML.

Simultaneously, while running test scenarios for optimization purposes, digital twin technology (through IoT, AI and ML) records how things are working in the factory in real time. In short, digital twins connect the digital world and the physical world. Of course, as with anything that utilizes AI and ML, the more data a model can draw on, the better its predictive capabilities become.

As a digital twin system becomes more complex and runs for longer, it becomes better at identifying inefficiencies and offering solutions. Hypothetically, an advanced digital twin system could even predict problems before they occur.

There are other use cases for digital twin technology, including in the health and medical fields, prototyping, insurance, investing, risk analysis, research, transportation, customer service and more, but the manufacturing example is among the easiest to explain and visualize, and it's likely that manufacturing will be the first industry to adopt digital twins en masse. The manufacturing example also clearly illustrates the way in which the IoT, AI, and ML merge to promote efficiency in the real world while testing and tracking in the digital world.

New business technology tends to follow the same pattern, assuming it is successful from a functional standpoint. First, the creation or invention phase occurs among a small group of elite developers; such developers might work independently or within an elite government organization or corporate environment.

Once there's a workable solution, high-level organizations begin gradual investment and adoption. Typically, the first groups to adopt advanced tech are firmly in the enterprise, research and big government sphere. If the technology is viable, and there's a demand, large business products are created, followed by lower-cost, user-friendly SMB products.

Thus far, digital twins have not trickled down to the SMB set, but enterprise products are already out there, which means it's only a matter of time before smaller-scale businesses solutions are available. In addition to the major players discussed in the introduction, companies such as GE and Predix already offer digital twin software, and Siemens and Microsoft have been vocal about investing in the development of digital twin solutions.

For enterprising business owners who are in the target fields for early digital twin adoption, such as prototyping and manufacturing, keeping an eye on the changing landscape of this emerging technology is smart, if not downright essential.

We'll keep you updated as enterprises and large businesses adopt digital twin software, and when SMB solutions start hitting the market, we'll give you all the details.

Mona Bushnell

Mona Bushnell is a New York City-based Staff Writer for Tom’s IT Pro, Business.com and Business News Daily. She has a B.A. in Writing, Literature, and Publishing from Emerson College and has previously worked as an IT Technician, a Copywriter, a Software Administrator, a Scheduling Manager and an Editorial Writer. Mona began freelance writing full-time in 2014 and joined the Purch team in 2017.