As The Forecast reported in May 2020, yet another example is the railroad industry. In Italy, for instance, digital twins help railroad operators streamline processes, improve predictive maintenance and monitor the health of their infrastructure, Maurizio Lombardi, chief operating officer of the Transportation and Logistics Business Unit at Italian IT consultancy AlmavivA, told The Forecast.
On the strength of these and countless other potential use cases, digital twinning is expected to explode in the next decade. In fact, a 2020 report by Research and Markets forecasts that the market for digital twins will grow from $3.8 billion in 2019 to $35.8 billion by 2023. That’s a nine-fold increase.
Digital Models, Real-World Impact
While engineers have been designing products using specialized software for decades, only recently has graphics technology been able to aptly mirror real life, according to Michael Grieves, now chief scientist for advanced manufacturing at the Florida Institute of Technology, who conceived the idea of the digital twin in 2002 while engaged in product lifecycle management (PLM) research at the University of Michigan. That, plus the ability to store and transmit data via the cloud, has created a moment that’s ripe for digital twin adoption.
“Right now, my perception is we’re in the conceptual stage of digital twins,” Grieves told the American Society of Mechanical Engineers (ASME) in a recent interview. “We have this information that we can bring together to create this virtual version of real-world environments based on models and behavioral aspects and modeling and simulation. The next step is to have all this information be pulled together automatically and intelligently, and we’re starting to see that occur as the software capabilities begin to arise.”
The convergence of digital twinning with automation and artificial intelligence will yield benefits in a vast spectrum of industries, many of which already are creating and using digital twins.
Here, digital twins mirror everything from factories’ assembly lines to their loading docks. By analyzing factory performance in a digital environment, manufacturers can uncover problems and ideate potential solutions. If production is too slow, for example, a manufacturer might use a digital twin of the production line to locate potential bottlenecks and model proposed changes to the line in order to determine how those changes would affect production. If the manufacturer wants to eliminate waste, it can model changes to the production line in a similar fashion to see whether they will produce the desired results. Running “what if” scenarios like that allows manufacturers to test solutions to see if they’ll work before they invest resources into actually building them.
Digital twins also can be useful in maintenance scenarios. If a machine part breaks during production, for instance, manufacturers can duplicate the incident in a digital twin to get a better picture of its impact on the real-life production line. In doing so, they can find the source of the breakage sooner and develop workarounds if the part can’t be immediately replaced.
Engineers can use digital twins to build digital simulations of systems they’ve designed. An automotive engineer, for example, can run a simulation of a vehicle crashing into a wall to determine how the car would react, then make design modifications to improve its safety performance—all without spending money to build and wreck a car. An aerospace engineer can do the same thing with a rocket to test its performance without endangering astronauts who might otherwise need to be on board.
Oil and gas
Oil and gas companies use digital twins in the same way manufacturers do: for system design and monitoring and for predictive maintenance. In the latter, internet-connected sensors continually monitor assets—an oil rig, for example—and report readings back to their digital twin, which flags underperforming parts that need to be replaced or otherwise maintained. In this way, engineers can remotely monitor the performance of faraway equipment without putting inspectors on the ground, which can be expensive and dangerous.
Digital twins and smart cities make good bedfellows. Using digital twins of cities and infrastructure to visualize data and run simulations, urban planners and civil engineers can solve problems like traffic congestion and air pollution. They can envision the impact a new building would have on its surroundings, for example, and isolate areas of opportunity for reducing waste or saving energy. Singapore’s Virtual Singapore initiative, for example, uses digital twins to determine the best locations for solar cells, cell phone towers and stop lights.
Doctors might one day be able to create digital twins of patients using their medical information. Alongside artificial intelligence, such twins could help them diagnose diseases, prescribe medication and monitor wellness by learning patients’ normal biological rhythms and flagging deviations that might require intervention.
A Better, Safer World
Digital twins’ benefits aren’t lost on business leaders. In its aforementioned report, ResearchandMarkets found that while only a third of leaders (36%) across industries understand the benefits of digital twinning, over half (53%) plan to incorporate it within their operations by 2028.
That number will only grow as digital twins themselves improve, which they will continue to do as technology advances, Grieves said during a keynote speech at ASME’s 2020 Digital Twin Summit. Consider digital twins’ accuracy, for example. Today, digital twins utilizing artificial intelligence perform with an accuracy of about 75%, according to Grieves, who says that number eventually will grow to 98% as technology gets better.
That kind of accuracy will create even greater opportunities for digital twins, which will suddenly become as useful in emergencies as they are in the course of normal operations.
“With the  Air France Flight 477 crash, the pilots continued performing the same [incorrect] action over and over. With a digital twin to show what would happen 30 minutes later in flight, they might have been able to correct their course,” Grieves said in a 2020 interview with Insight magazine, a publication of Oakland University’s School of Business Administration. “Another area where predictors could have life-saving consequences is surgery.”
Or consider the next big pandemic: With digital twins of cities, hospitals and patients—and even the virus itself, or potential vaccines with which to fight it—the next time a lethal illness threatens humanity, it might be possible to model its transmission, predict its spread and engineer public health interventions that stop it in its tracks. Similar models could be useful in the aftermath of a natural disaster or in the midst of a humanitarian crisis. In that way, digital twins can touch individuals on a very personal level by enabling a better and safer world.
“[Digital twinning] is dominating our capabilities,” Grieves told industrial software company Cognite in a 2020 interview. “As long as we have increasing computing capabilities, which we are having, I think the future is … bright.”