Digital twins have evolved from a tool into a connected, intelligent and autonomous universal concept
In just a few years, digital twins have gone from being static representations of individual products to intelligent, connected systems – a transformation that has gone largely unnoticed. They are enabling new efficiency improvement, business model and value creation opportunities in areas as diverse as mechanical engineering, urban infrastructure and global climate modeling. In this article for TRANSFER, bwcon research’s Joachim Reinhart outlines how digital twins are developing from static representations to (semi-)autonomous control systems, their main current areas of application, and the strategies that can help enterprises to successfully deploy and add value with this technology.
Digital twins have become a key driver of the digital transformation, not only in large corporations but also in small and medium-sized enterprises. By linking real objects and processes to a virtual counterpart, they enable better analysis, simulations and decision-making and even (semi-)autonomous control. Initially inspired by ideas in the field of product life cycle management in the early 2000s, digital twins have – thanks to sensor, cloud and artificial intelligence technology – now matured to a point where scalable applications are possible. Steinbeis Network member bwcon research has been contributing to these developments for several years through its research and transfer projects, and helping enterprises to leverage digital twins’ potential from the earliest possible moment.
Same same but different: the original incarnation of digital twins
Michael Grieves came up with the concept of digital twins when writing about product life cycle management in 2002 [1]. The idea was to create a virtual model of a physical system that is continuously linked to the real physical system throughout its entire life cycle.
The traditional definition of a digital twin is a digital representation of a real object that is linked to this object through continuous data exchange [2]. One of the key challenges is to maintain digital twins for each individual product instance throughout its entire life cycle. In the case of major automotive manufacturers, for example, this can quickly lead to the creation of millions of digital twins that generate huge volumes of data due to the products’ many sensors and long lifespan. They allow OEMs like Volkswagen to carry out targeted recalls of thousands of vehicles [3]. Well-maintained digital twins can also use artificial intelligence to detect problems and suggest timely fixes.
The digital twin concept is no longer confined to individual products – far from it, in fact. Twins are now also being used for product systems, entire factories, supply chains, urban infrastructure and energy systems [4]. Small and medium-sized enterprises can also create virtual representations that enable optimization of entire production lines. The digital twin has thus evolved from an engineer’s tool into a universal concept.
Digital twins for everything: areas of application
In their more advanced, modern-day incarnation, digital twins can be described as a virtual, dynamically linked representation of a specific real object, process or system. They use a combination of technologies such as sensors, data platforms, simulations, artificial intelligence and cloud or edge computing. In terms of their current development, there are three key trends [2, 5]:
- More objects: Digital twins are being created for more and more things, systems, and even people.
- More connectivity: There is a shift from isolated twins towards “systems of systems”, for example across supply chains.
- More autonomy: Instead of simply representing something’s current status, digital twins are increasingly becoming (semi-)autonomous and capable of independent decision-making.
These trends are being driven by technologies like standardization, cloud and low-code solutions, artificial intelligence, edge computing and blockchain technology, as well as by growing security requirements. The multi-stage model illustrates how the trends are evolving. The more objects there are with digital twins, the greater the potential for analysis and forecasting. As they become more connected, simulation and optimization become increasingly important. And the more autonomous digital twins become, the more they start to take on active control tasks. While the benefits increase with each stage, so too do the data quality and integration requirements.
Real-world examples illustrate the wide range of applications. Rolls-Royce uses digital twins for its usage-based “Power-by-the-Hour” business model ([2], p. 25), BMW’s “BMW iFACTORY” uses them to optimize entire production lines [6], the Digital Twin Earth project is modeling the complete Earth system for more accurate climate and weather forecasting [7], while Urban Mobility Twins control traffic flows in Singapore in real time [2].
Looking ahead: mobility in 2030
A mobility scenario for 2030 describes how these trends might combine in people’s everyday lives. Tom, who works as a sales manager for an engineering company, is traveling to visit a customer in Shanghai. His digital assistant, Lara, organizes everything – the autonomous vehicle that takes him to the airport, a flight tailored to his personal requirements, and a self-driving car for when he arrives in China. Lara informs him that the surprisingly short journey time in Shanghai is down to the high proportion of autonomous vehicles there and the AI-driven traffic management system.
This seamless journey is made possible by several digital twins working in concert. Twins are used for vehicles, traffic management, flight operations and personal service profiles. These formerly separate services now combine to deliver an integrated customer experience, enabling new business models and placing the platform operator in the key role of integrator.
No longer merely a tool for engineers, digital twins have thus become a strategic enabler of new business models. Organizations that are quick to explore how they can harness the three key trends in their own field can seize the opportunity to optimize their processes and open up new markets.
The four steps to success
A systematic approach is key to the successful deployment of digital twins. The projects supported by bwcon research have shown that a structured approach is vital throughout, from the ideation phase right up to the operational phase. The four-stage model, which combines the technology, customer and economic perspectives, has proved a valuable means of achieving this in practice.
The first step involves identifying the potential. Customized use cases are identified in innovation workshops with the aid of customer journey mapping, for example. The next step is to carry out a feasibility analysis. This involves using methods like design thinking to find a balance between technical feasibility, user wishes (desirability) and viability (financial return). The implementation stage draws on traditional or agile methods such as scrum, supported by change management. Finally, the focus during the operational stage is on performance management of the digital twin, i.e. monitoring and optimizing indicators such as system availability and energy consumption.
Focus on adding value
Digital twins have evolved into a strategic enabler in virtually every industry. “This development is characterized by three trends: more objects, more connectivity and more autonomy. As the technology matures, its benefits increase, but so too do the technological, organizational and cost-effectiveness requirements”, concludes Joachim Reinhart. Real-world examples ranging from industry to global climate modeling show that digital twins are now much more than just a virtual representation. They link real systems with a virtual counterpart, carry out analysis, simulations and forecasting and – depending on how advanced they are – can actively affect the physical world. Enterprises that are quick to explore how they can harness this silent powerhouse can achieve efficiency improvements, gain a competitive edge and access new market opportunities in a data-driven economy.