There is a long lineage of technologies and tools used to model the physical world, including drawings, diagrams, and CAD models. There are also many ways to use technology to model real-world systems and make predictions, including financial trading simulators, weather predictors, and traffic pattern models.
When users put these two capabilities together — combining a digital representation of a physical-world system and a model that simulates output conditions based on inputs drawn from the physical environment — they get a digital twin. A digital twin allows users to validate the system against a wide array of real-world situations.
Engineers use digital twins in manufacturing, construction, energy, transportation, medicine, science, and other industries to develop products and validate real-world systems. This may sound like science fiction, but with advances in machine learning, systems modelling, Internet of Things (IoT) sensors, data streaming platforms, simulation technologies, and cloud infrastructure, digital twins are becoming more prevalent every day.
To separate fact from fiction, I reached out to several experts to share insights on digital twins and how business and engineering teams use them today.
What is a digital twin?
Prith Banerjee, CTO at Ansys, defines digital twins as “a connected, virtual replica of an in-service physical entity, such as an asset, a plant, or a process. Sensors mounted on the entity gather and relay data to a simulated model (the digital twin) to mirror the real-world experience of that product.”
Beyond a replica, digital twins receive the same real-time data streams as physical-world systems. Simon Crosby, CTO of Swim, focuses on this aspect of digital twins in his definition.
He says, “A digital twin is a live agent that continuously analyses streaming events from the real-world ‘thing’ as they are received in context and delivers the results in real time to other agents, applications, and user interfaces. These digital twins always accurately reflect the current state of the real world.”
What types of problems do digital twins solve?
Crosby shares two ways to use digital twins: augmented reality and real-time views of whole systems. Augmented reality applications have several practical use cases. He says, “Digital twins were conceived as design-time digital overlays for use in augmented reality applications: for example, an engineer fixing a jet engine.”
Augmented reality can help train engineers or simulate procedures before a person implements them in the real world. Augmented reality and digital twins have applications in manufacturing, medicine, energy, and whenever complex training and procedures are performed on expensive equipment, or when human safety is a critical factor.
Crosby shares a more expansive way to think about digital twins. He adds, “Applications can link together digital twins to build powerful models that deliver real-time system-wide views, for example, the current and predicted state of traffic in a city.”
In other words, a smart city’s digital twin is an aggregate formed by linking the digital twins from buildings, transportation, government services, and other systems.
Banerjee adds that engineers use digital twins to model future behaviours and scenarios. He says, “Digital twins enable tracking of past behaviour of the asset, provide deeper insights into the present, and most importantly, they help predict and influence future behaviour.”
Engineering teams also use digital twins to evaluate design trade-offs and deliver production systems faster. Robin Yeman, a strategic advisory board member and director of cyber-physical advisory practice at Project and Team, says, “Building digital twins for cyber-physical systems allows companies to validate multiple design trade-offs in the digital environment before implementation, reducing rework and allowing them to deliver faster.”
How are digital twins created?
Andrew Clark, founding CTO of Monitaur, shares insights on a digital twin’s modelling and development process.
He says, “To create a digital twin, a representative environment of the object or ecosystem must be constructed, which entails deep domain knowledge of the underlying behaviours and mechanisms of the system in question. After input signals are incorporated into the digital twin and a model is created through systems identification, accurate extrapolations or predictions of future behaviour of the system can occur.”
Examples of constructing a digital twin might include building information models in construction that have object-level details on all the building components, such as doors, windows, or materials. In manufacturing, a digital twin can simulate the full production process, including connections to manufacturing execution systems to feed in live data.
Digital twins cannot be one-time models, and they must reflect changes made to the real-world system. Clark adds, “Creating accurate digital twins is a very complex endeavour that requires deep domain expertise, otherwise you end up with non-representative and inaccurate models. To combat this gap, digital twins are often configured to be online learning systems, meaning that they constantly update and retrain off of new input data."
For users thinking about building a digital twin, Brent Pookhay, executive vice president and CIO of Nutrien, suggests working directly with the people in operations with deep expertise in how the systems work.
He says, “Building a digital twin is as much about people as it is about technology. Who are the subject matter experts (operations, engineers, field crews) that run those assets? Bringing their deep understanding, experience, and practical knowledge of operating those assets can be as important as the data feeds coming off of your OT and SCADA systems.”
What are some use cases for digital twins?
Banerjee shares several digital twin examples. “Digital twins are used in various phases, including design, manufacturing, and operations, and across industries such as aerospace, automotive, manufacturing, buildings, infrastructure, and energy. They typically impact a variety of business objectives, including overall equipment effectiveness, predictive maintenance, yield, and budgets.”
Here’s a sample of digital twin projects.
- Thirteen experts share practical digital twin use cases that can be applied in multiple industries, including calculating product performance, simulating complex manufacturing scenarios, and facilitating hybrid teaching.
- My colleague Thor Olavsrud recently shared digital twin success stories from Rolls Royce, Mars, The Teachers Insurance and Annuity Association of America-College Retirement Equities Fund (TIAA), and Bayer Crop Science.
- Las Vegas is creating a digital twin to help it move to zero carbon emissions.
- A public-private partnership is developing a digital twin for the Brooklyn Navy Yard to reduce its energy footprint.
- In urban planning, a Texas A&M University professor is developing a digital twin of Texas coastal communities to study their resilience to natural hazards.
- In manufacturing, digital twins are being used for product design, supply chain management, predictive maintenance, and customer experience analysis.
- CenterLine, a Canadian industrial automation process and technology company, used a digital twin of the factory floor to reduce tool-related issues by as much as 90 per cent and programming time on the factory floor by as much as 75 per cent.
- Digital twins will transform healthcare as scientists develop virtual organs, improve the caregiver experience, classify drug risks, and more.
I expect to see many more examples of digital twins, especially as companies consider developing and supporting more complex products, processes, and other physical environments.
How can developers and devops engineers enable digital twins?
Digital twin platforms such as Ansys, Autodesk, Bosch, Dassault Systems, Siemens, and other vendors provide modelling and simulation capabilities. In addition, the public clouds have extended their IoT platforms with digital twin capabilities, such as Azure Digital Twins and Google’s Supply Chain Twin. AWS digital twin architectures may include Amazon Kinesis Data Streams, Amazon SageMaker, AWS Lambda, and other services.
IT teams must consider the infrastructure required to operate digital twins. Yeman says, “The need to deliver product faster continues to grow; however, lead times for hardware and firmware can slow companies down.”
Developers must also consider that IoT and other real-time data streams may feed multiple systems, including digital twins. That means configuring the data streaming technology to share real-time data between production systems and the digital twin development and test environments.
Digital twins are an exciting emerging technology that demonstrates the convergence of many different technologies, including machine learning, IoT, data streaming, and augmented reality. It will bring a new era of innovation, safety, and efficiency to many industries.