Learn why the ‘brave new world’ of digital twins is an understatement.
BY PeterBilello on Nov 22, 2022
Like any maturing technology-enabled phenomena in a fast-moving and competitive marketplace, the digital twin is being pushed into new tasks, a wider variety of industries and, consequently, is rapidly moving beyond its original applications. From its beginnings in discrete industrial products in automotive and aerospace & defense, digital twins have become common in architecture, engineering and construction (AEC) or engineering, procurement and construction (EPC) projects and even the decision making processes for buying/selling, commissioning and operations management.
(Image courtesy of Bigstock.)
Simple Internet searches reveal that digital twins now exist far beyond their traditional core markets, applications and tasks. Into 2023 and beyond, digital twins are expected to be attached to and represent an endless variety of processes and products. This is evidenced by the fact that already in 2022, digital twins have been turning up in a wide range of applications:
- Supply chains: for packaging, robotics and logistics.
- Hospitals and medical facilities: for better treatments of critically ill patients and to represent individual organs such as hearts, brains or lungs, and even entire human bodies.
- Autonomous/self-driving vehicles: for safety, navigation, AI training and more.
- Insurance: for insuring and underwriting risks and fraud.
- AEC and EPC: for zero-emissions and energy-efficient buildings.
- Banking: for modeling borrowers.
- Wastewater management and flooding: for safety, predictions and more.
Given the growth of digital twins, let’s dig into where they are now and where they are going from an engineering perspective.
Let’s start with the question, what is a digital twin?
To get a better idea of where digital twins currently stand, let’s look into how they are being defined. CIMdata’s definition is “a virtual representation (i.e., digital surrogate) of a physical asset or collection of physical assets (i.e., physical twin) that exploits data flow to/from the associated physical asset(s).” Digital twin capabilities can improve operations by generating valuable insights, guiding the automation of complex processes, and monitoring and controlling industrial assets through specialized software.
A digital twin is also a data repository that is equipped with the latest information technologies. Like every other repository, it gathers and manipulates information needed for specific tasks, for analyses such as failure modes, spare parts needs, suppliers’ and competitors’ status, security threats, scale-ups and much more. The variety of repositories—digital twin or otherwise—is limited only by curiosity and ingenuity.
In contrast to conventional data repositories, the information in a digital twin focuses on a single object, e.g., a serialized product, or a collection of real-world objects such as a fleet of buses of all the same model. These are sometimes called avatars or digital clones. Digital twins often have far fewer variables than conventional, heterogeneous repositories. A tight focus distinguishes the digital twin from predictive analytics.
For users and managers new to digital twins or newly tasked with an implementation, benefits come from highly detailed representations of specific physical objects (e.g., products, assets, systems and processes) that are continually refreshed and communicated.
Once a digital twin implementation decision is made, more crucial decisions are necessary:
- What specific physical object (or asset, or system, or process, or project) is to be modeled, and why?
- Which fundamental physical properties of the object—material, shape, dimensions, physics (stress, oxidation), chemistry, thermodynamics and others—are to be modeled, along with their level of precision? This is best done in a Model-based Structure (MBS).
- What varieties of data are to be gathered by, fed into or accommodated by the digital twin? Choices include real-time and historical data; inputs from asset management systems; inventories, spare parts and substitutes; financial data and cost estimates; electrical connections; electronic components, software, sensors (embedded or attached) and operability (safety, potential hazards).
- What security is needed to address any potential risks posed by the digital twin itself or its end-to-end digital connectivity?
The correct answer for each varies with the industry, product family, customer and intended use of the digital twin.
Digital Twins In 2022
The unprecedented growth of digital twins makes the expression “brave new world” seem like an understatement. However, in some ways “risky” is also an understatement.
For instance, if you are concerned about your personal information escaping from your smartphone, digital twins may be a big worry. One example, according to a Gartner Q2-2022 report “Planning for the Never Normal,” is retailers creating “digital personas” or representations of customers that detail behavior, purchases and spending using artificial intelligence (AI), machine learning (ML) development techniques and visuals built on augmented and virtual reality (AR/VR). You don’t want that getting into the wrong hands.
Given the risks that some digital twins pose, a 2022 survey report by the Capgemini Research Institute (CRI) titled “Digital Twins: Adding Intelligence to the Real World,” found that 69 percent of respondents plan major changes in their cybersecurity. The survey reached over a thousand organizations, and over 80 percent have ongoing digital twin programs—the rest are planning to implement one. Additionally, as many as 55 percent of respondents consider the digital twin a strategic element of digital transformation.
Perhaps the most ambitious digital twin to date, Destination Earth (DestinE), is being built by the European Centre for Medium-Range Weather Forecasts (ECMWF). DestinE will simulate the atmosphere, oceans and even human activities. Working on DestinE with ECMWF are the European Organization for the Exploitation of Meteorological Satellites (EUMETSAT), the European Space Agency (ESA), and the U.S. National Oceanic and Atmospheric Administration (NOAA). ECMWF is an independent intergovernmental unit. Destination Earth will be a successor to Europe’s Copernicus sensor system.
In many cases, the explosive growth of digital twins is being driven by the interactions of social media, climate change, net-zero industrial emissions, the Fourth Industrial Revolution and 5G telecom technologies. This is all facilitated by digital transformation reaching beyond its roots in industry and business.
As digital twins multiply, huge changes will continue to pop up. A big one is the standalone digital twin, newly independent of its longtime role in PLM and often implemented in the Cloud rather than on-premises. Standalone digital twins are offered by AVEVA, General Electric Research, IBM, ScaleOut Software, Wind River and others.
Digital twin capabilities are also being embedded into tools and systems for product design, engineering, construction, manufacturing and supply chains. OEMs are placing digital twins directly in the hands of the users of their products. Built-in dynamic simulations of digital twins are now being used to measure and predict hard-to-quantify phenomena such as sustainability.
Some much-needed help for digital-twin users has appeared in the form of configurable user dashboards. Because so much information can be extracted from a digital twin, users may temporarily get lost in it all. One answer is enhanced user interfaces such as Microsoft Corporation’s Azure Digital Twins 3D Scenes Studio_._This 3D immersive low-code environment can help users visually monitor, diagnose and investigate digital twins.
Digital Twins In 2023 and Beyond
As digital twins and their physical counterparts grow in complexity, visual communications will become increasingly important. Digital twins and their avatar representatives of the physical world may begin using or appearing in 3D metaverse implementations. Popularized by massive multiplayer online video games, a metaverse is a network of virtual 3D worlds built atop social media connections in which players and users appear as avatars.
Despite a dependence on VR devices and connections, metaverse interest is growing as a way to interactively boost workplace productivity, learning environments, e-commerce, real estate transactionsand fashion. In 2017, Microsoft acquired AltspaceVR (now Microsoft Teams) for VR meetings using avatars. Other players include Pixar, Blender, Apple, Autodesk, Qualcomm and NVIDIA. In 2021, the parent company of Facebook changed its name to Meta Platforms.
On the business side of digital twins, new partnerships are forming among well-established solution providers using information from the Internet of Things and the Industrial Internet of Things (IoT/IIoT). For instance, mechanical-engineering and simulation company Ansys Inc. teamed up with Microsoft Azure using Azure’s IoT Hub. In fact, many PLM solution providers have married their digital twin initiatives to Amazon Web Services (IoT TwinMaker and IOT FleetWise), Google Cloud (Supply Chain Twin and its Pulse interface) or one of the big tech consulting companies, including Accenture, AVEVA, Deloitte and Tata Consultancy Services.
In Europe, new digital twin ventures looking to expand the craft include MEVEA Ltd. (in Finland) with Siemens; Spirent Communications plc (in the U.K.) with Amazon Web Services; and Sewio Networks (in the Czech Republic) with SICK AG.
In AEC specifically, Bentley Systems has integrated its offerings with MS Azure and NVIDIA Omniverse.
Additionally, at least three digital-twin-focused industry standards groups have appeared to help guide the technology’s future. One is Twinify, a joint venture between ASME and Black & Veatch (an infrastructure-focused EPC) using nDimensional AI-focused digital twin platforms. A second is the Digital Twin Consortium, part of the Object Management Group, teaming up with the Smart Manufacturing Institute. A third is the Industrial Digital Twin Association in Germany.
Given all these innovations and upheavals, implementing digital twins and even understanding them is not easy. To start an appreciable flow of benefits, a digital twin must be capable of converging information from data architectures, knowledge graphs and many different processes, along with business and social cultures. All of this is, of course, in dissimilar formats.
As time passes, new challenges will arise as parameters, tolerances, design principles and even basic digital twin assumptions evolve. Many digital twins rely on rules-based approaches for simulations and condition monitoring built with code that was 20 years old when installed. As time passes, objects represented in digital twins will change, and processes will evolve, requiring periodic offsets, updates and adjustments.
Today’s sophisticated digital twins have built-in AI solutions that use existing and evolving asset and environment data to help keep its parameters current. Benefits show up as more true positives, fewer false negatives, fewer missed anomalies, and better root-cause tracking. These capabilities point to a potentially huge new role for digital twins—putting an end to physical prototyping.
Where Do We Go from Here?
To summarize, perceptions of the digital twin are changing radically. Not long ago seen as a PLM repository or simply as-maintained BOM structures, digital twins are now often visualized as waystations on the enterprise’s endless bi-directional circular flow of information. They have morphed into standalone elements of digital transformation, many of them in the Cloud. Standalone or otherwise, digital twins are transforming the creation and lifecycles of products and assets with continually gathered information and updates, dynamic simulations and assured on-demand access by an ever-changing spectrum of developers and users.
All of this adds up, of course. According to Allied Market Research, estimates of the global digital twin market spending for 2021 are already in the low billions of [U.S.] dollars. Compound annual growth rates are generally projected at nearly 40 percent, which points to a digital twin market in 2030 of about US$126 billion.
With fast growth and big numbers, the conclusion should be obvious. If your business hasn’t gotten a start on digital twins in 2022, your competitors surely have. If you have started, step on the accelerator in 2023 by asking for more resources and spreading the word to expected users about the big, promised benefits of digital twins.