By Brian Anzaldua, Principal Global Architect, Dell Technologies
The concept of making a duplicate or ‘twin’ of an asset to enable simulations and predict outcomes based on changes in the operating conditions finds it origins in the 1960s with the Apollo space program. It came to the forefront in April of 1970 when NASA engineers used copies of the spacecraft systems originally intended for training to support the rescue of Apollo 13 after a disastrous explosion of its oxygen tanks. As the astronauts slept in the lunar module, the command module had to be powered down to preserve the batteries. NASA engineers used what amounted to a twin of the command module’s electrical system to devise a series of steps to power up the frozen space craft without draining what little power remained in the batteries. A process that normally took two days on the launchpad was completed in under two hours. It was an extraordinary feat. And yet, the technology has only matured since then.
Digital Twins started as a model of a physical device and have evolved into an invaluable tool, encompassing representations of logical and even biological systems that can shorten design times, reduce costs and increase safety.
However, while they have been around for a while, albeit in different forms, they’re still regarded as novel. And like many emerging technologies, they’re often misunderstood. Understanding the difference between a model and a Digital Twin would help demystify the technology and its application.
Microsoft’s Flight Simulator 2020 embodies the distinction between a model and a Digital Twin, in the sense that the Flight Simulator is a typical model. It reproduces, in detail, the flight characteristics of a specific model plane and how it interacts with the environment in a way that mimics the real world.
It enters Digital Twin terrain however, by including a feedback loop from the real world that influences how the model behaves with every user. By integrating real-time weather updates including wind speed, temperature, humidity, pressure, rain, snow and other data–all the way from ground level to the stratosphere–Microsoft effectively created a Digital Twin.
In some respects, it’s a limited twin. It doesn’t take into account the health of the airplane’s engine or the stresses that may lead to failure of the wing, electrical system or other components of the plane. Is that a problem? It all depends on what the business is trying to deliver with the twin.
It starts with a conversation
From the outset, a business should clearly define what it is trying to learn or optimize with the technology. If the goal is to keep the plane flying for the most hours, then predictive maintenance will require not only an accurate simulation of the engine and its parts but also data from its maintenance records, as well as the flight conditions it was used in.
If on the other hand, its raison d’être is to provide users with the ability to understand the implications of flying in adverse conditions at night then the current system would suffice.
To build an efficacious Digital Twin, it’s important to first agree what problem needs to be solved or what opportunity needs to be explored and how accurate do the predictions need to be. Understanding the goal of a Digital Twin determines which data and sensor feeds are required to achieve predictive value within a defined confidence interval.
These conversations are now happening in boardrooms large and small, across sectors–thanks to the democratization of sensors and processing power. In the past, it was unusual to see Digital Twins beyond the four walls of an Industrial Manufacturing plant–because they were prohibitively expensive to build. Early work with Digital Twins was reserved for very expensive assets which accepted the price tag, or assets that were too expensive to fail (i.e. would threaten health and safety). Prolific and affordable new technologies have lowered the barrier of entry, expanded the use cases for this innovative technology and made Digital Twins more accessible.
Bringing Digital Twins to New Industries
Successfully digital twinning an engine, airplane or any other asset no longer needs be the end state. Advances in Edge computing and in-memory processing enabled by scalable compute delivered though containers, along with new pervasive network technologies like 5G supporting streaming data, make it possible to interconnect these twins. The results can be seen in real-world projects that have twinned full manufacturing lines and complex interconnected processes like supply chain integration. Today Digital Twins are becoming an essential part of everything from smart city planning to improving healthcare.
For example, Singapore has created a complete digital twin of the city-nation to track traffic, pollution, climate, and city layouts so city managers can test accessibility options, see the potential impact of new construction, manage emergency responses, and monitor city health. Meanwhile, doctors are creating patient-specific Digital Twins of lungs to help make decisions about ventilator use when treating COVID-19 patients.
Thanks to all the potential use cases that exist, the digital twin market is projected to rapidly grow within the next few years. The market was valued at $3.1 billion in 2020, but is expected to reach $48.2 billion by 2026. New standards like the Gemini Principles and ISO/DIS 23247 will help ensure new digital twin applications are compatible, secure, and scalable. As a founding member of the Digital Twin Consortium, Dell Technologies is committed to its mission of driving consistency in vocabulary, architecture, security and interoperability of digital twin technology to advance its use across industries and maximize its benefit.
The Future Is Now
For the Future of the Economy Report, Dell Technologies partnered with Institute for the Future to explore how emerging technologies will reshape our economy over the next decade. The report notes that Digital Twins will empower the shift to automated, iterative manufacturing that can meet demand on the fly.
Certainly, the potential for Digital Twins in almost every industry is endless. Previously siloed departments across operations, maintenance, finance, sales, and marketing can all use digital twins to access a unified source of real-world data to predict maintenance, improve design, understand usage, and adjust pricing.
While still an emerging technology today, Digital Twins are the future. That means now may be the perfect time for CTOs to start collaborating with other leaders across their enterprise to understand how to put Digital Twins to work.