Digital Twins (eBook)

Internet of Things, Machine Learning, and Smart Manufacturing

Yogini Borole, Pradnya Borkar, Roshani Raut, Vijaya Parag Balpande, Prasenjit Chatterjee (Autoren)

eBook Download: EPUB

2023
255 Seiten
De Gruyter (Verlag)
978-3-11-077896-0 (ISBN)

Infos

Y. Borole, Raisoni Institute of Engineering and Technology, Pune, India; P. Borkar, R. Raut, V. Balpande and P. Chatterjee, India.

Textauszug

Chapter 1 What Is Digital Twin? Digital Twin Concept and Architecture

Abstract

The digital twin concept is an emerging technology used in various industries. An overview sheds light on this emerging technology, architectural construct, and business initiative. This chapter explains the digital twin concept using the cyber-physical system architecture, business use cases, and value proposition.

A physical model that can be reflected accurately as a virtual model is called digital twin. Gaining a better knowledge of the performance characteristics of tangible parts, processes, or systems is feasible with the aid of digital twins. Industrial operators can use digital twins to remotely visualise, forecast, and optimise the performance of a system, a process, or the entire process.

A virtual object called a “digital twin” precisely copies the structure, content, and status of the related real-world thing. Consider it a test version before continuing with product development or using a real-world device. Implementing digital twins and the Internet of things (IoT) go hand in hand. For instance, manufacturing systems can be simulated and tested by industrial organisations by creating models using precise real-time data. Using digital twin technology can aid numerous other organisational objectives [1].

The different definitions of digital twin are as follows:

A representation of an organization’s physical resources that aids in outlining its overall business operating model.

A computer programme that uses actual data about a physical system or item as inputs and delivers simulations or predictions of how the inputs will impact the physical system or object.

A digital representation of a procedure, item, or service…

Consider an example of the object being examined, such as various sensors are connected to the wind turbine, which acts as key functioning regions. These sensors generate information about a variety of performance characteristics of the physical device, including energy output, temperature, and environmental conditions. The processing system then applies this information to the digital copy. It is possible to utilise this virtual model to run simulations, investigate performance problems, and create potential improvements after obtaining such data, all in an effort to produce insightful information that can then be applied to the real physical thing.

1.1 Difference Between Digital Twin and Simulations

Both digital twins and simulations employ digital models to replicate a system’s numerous activities, but a digital twin is actually a virtual environment, making it more valuable for research. A digital twin can perform as many useful simulations as necessary to examine numerous processes, whereas simulation often only studies one particular process. This is the major difference between a digital twin and a simulation.

There are still more differences. For simulations, for instance, real-time data is often not helpful. Digital twins, on the other hand, are designed around a two-way information flow that starts when object sensors provide the system processor with relevant data and continues when the processor exchanges insights with the original source object.

Because they have better and more up-to-date data about a wide range of fields and the additional computing power that comes with a virtual environment, digital twins have a greater potential to ultimately improve products and processes because they can study more problems from a much wider range of perspectives than standard simulations.

1.2 History of Digital Twin Technology

With the release of Mirror Worlds by David Gelernter in 1991, the concept of digital twin technology was first presented. But Dr. Michael Grieves, who was then a professor at the University of Michigan, is recognised as having introduced the concept of digital twin software and first employed it in a production setting in 2002. Finally, the term “digital twin” was invented by NASA’s John Vickers in 2010.

The fundamental concept of utilising a digital duplicate to inspect a physical object, however, can be seen far earlier. It is accurate to say that NASA was the first organisation to use digital twin technology during its space exploration mission in the 1960s. At that time, each moving spacecraft was painstakingly replicated in an earthbound version that NASA employees serving on flight crews used for training and simulation.

Digital twins begin with the existence of a physical component, even before a prototype, and continue up until the end of the product’s useful life. Twins for designs, manufacturing, and operations might be considered the three main stages of the digital twin life. Since the 1960s, physics-based models that employ numerical techniques, such as finite element analysis, have been utilised as digital twins for the design phase. Today, they are a commonplace design tool for quickly identifying the best ideas. The same set of tools have also been used to forecast how a part will react to a manufacturing process, enabling engineers to take manufacturing effect into account during the design process and prevent design difficulties later in the product life cycle.

Although the idea of twins for processes like machine learning has been known since the 1960s, it wasn’t until recently that data pipelining and data science advancements made their use more common. The diagnostics (anomaly identification and root-cause investigation) and prognostics (remaining usable life prediction) of engineering systems are performed using these data-based predictive models, which reduce scheduled maintenance and eliminate expensive breakdowns. The two types of predictive models also complement one another in that the real-time data, such as the discovery of crucial load cases that are lacking in the operating environment, can be used to enhance physics-based models. Similarly, for circumstances not covered by real-time data, data from physics-based models can be used as a supplement [1].

1.3 Various Types of Digital Twins

There are various types of digital twins, depending on how magnified the product is. These twins differ primarily in the area of application as shown in Figure 1.1. Different types of digital twins frequently coexist in a system or process.

Digital twins come in a variety of forms depending on how magnified the product is. These twins most significantly diverge in the area of application. It is typical for various kinds of digital twins to coexist in a system or process.

Figure 1.1: Types of digital twin.

• Component twins or parts twins

Component twins are the core unit of a digital twin and the most basic representation of a functioning component. Parts twins are virtually the same thing as identical parts, despite the fact that they refer to far less important parts.

• Asset twins

An asset is formed when two or more components perform well together. Asset twins allow you to investigate the interactions between these factors, generating a wealth of performance data that can be analysed and turned into insightful information.

• System or unit twins

System or unit twins, which show how diverse assets work together to form a complete, usable system, are the next degree of magnification. System twins offer visibility into how assets interact and may make performance suggestions.

• Process twins

Process twins describe how several systems interact to construct an entire production plant. Are all of those systems coordinated for maximum efficiency, or will delays in one system affect others? The specific timing schemes that eventually affect overall efficiency can be found with the use of process twins.

1.4 Pillars of Digital Twin Technology

Three important pillars for digital twin technology are discussed as follows:

1. There are actual physical objects in the real world.

2. In the digital realm, there are virtual object profiles.

3. Physical and digital worlds are connected by a bridge that facilitates information sharing and data interchange.

The digital twin provides substantial advancements in the production of things and materials, how to develop services that contribute to the maximum level of customer delight in the modern world, as well as key turning points in human history like the development of industry and agriculture [2].

...

Produkt-Details

• Erscheint lt. Verlag: 18.9.2023
• Reihe/Serie: ISSN
• Reihe/Serie: Smart Computing Applications
• Zusatzinfo: 6 b/w and 77 col. ill., 1 b/w tbl.
• Sprache: englisch
• Themenwelt: Mathematik / Informatik ► Informatik
• Themenwelt: Technik ► Bauwesen
• Themenwelt: Technik ► Elektrotechnik / Energietechnik
• Schlagworte: computer simulation • Digital Twin • internet of things • machine learning
• ISBN-10: 3-11-077896-3 / 3110778963
• ISBN-13: 978-3-11-077896-0 / 9783110778960