Techne

Abstract: Nowadays, Industry 4.0 is widely known in the industry. Many large companies have pushed and adapted to Industry 4.0 to increase their business competitiveness, market shares, and capture new business opportunities. So small and medium-sized companies must adjust to improve their production capabilities to keep up with the technological development of Industry 4.0. With the recent wave of digitalization, the latest trend in every industry is to build systems and approaches that will help it not only during the conceptualization, prototyping, testing and design optimization phase but also during the operation phase with the ultimate aim to use them throughout the whole product life cycle and perhaps much beyond.

A digital twin can be defined as a virtual representation of a physical asset i.e. product or a process enabled through data and simulators. Recent advances in computational pipelines, artificial intelligence, big data processing and management tools bring the promise of digital twins and their impact on society closer to reality. One such manifestation of digital twining is mass customization, a marketing and manufacturing technique that combines flexibility and personalization offering products tailored to each individual customer’s needs, at a low cost. The mass customization paradigm allows customers to select, order, and receive a specially configured product, often choosing from a multitude of product options. These product options often require real-time prediction, optimization, monitoring and controlling for improved decision making.  In the design phase, one of the main challenges in fulfilling custom orders is to generate bills of materials and design specifications quickly and accurately from the thousands of possible combinations.  In the operational and maintenance phase, one of the main challenges is supervisory, diagnostic and control, and predictive maintenance of the product. In this work, we review the recent status of methodologies and techniques related to the construction of digital twins mostly from a modelling perspective. Our aim is to provide a detailed coverage of the current challenges and enabling technologies along with recommendations and reflections for various stakeholders by designing and developing digital twin systems combined with existing production techniques such mass customization.

Keywords: Industry 4.0; Digital Twins; Internet of Things; Mass Customization; Design Optimization; Predictive Maintenance.

Introduction

In today's hypercompetitive environment, there's pressure to develop and bring to market products faster, to identify efficiencies and how they are built, and to incorporate innovations in design and features. This is particularly true of highly sophisticated machinery and other complex systems, including buildings and cities. In the new product introduction process, the luxury of spending months and even years designing an improved or new system seldom exists anymore. Being able to create relatively quickly an accurate, functioning virtual replica of a future system and test it under all manner of scenarios is highly desired and beneficial. This is the role of a digital twin in the design phase. A digital twin is a virtual replica of a physical thing. This virtual twin exists only as software rendered by computing power. Having a digital replica of a physical thing can significantly improve on one or more of the following processes, design, simulation, planning, building, operating, maintaining, optimizing, and disposal.

For example, isn't it useful to first build a virtual skyscraper, then put it through a variety of earthquakes to accurately understand what might happen? Based on the data provided, the design can then be modified. In the build phase, a digital twin can be used to provide the construction specifications or what we call parametric estimates to different providers. In this way, a digital twin can be an asset in streamlining the procurement process. In addition, and importantly, during build, sensors are applied to the physical object to collect and transmit data back to its virtual replica. This is what enables diagnostic and control during the operational and maintenance phases. At this point, with enough sensors, the virtual twin is providing all relevant data about the state of the physical twin. For example, an operational machine can render accurately in its digital twin its temperature, vibration, speed, and so much more. All of this becomes possible because of increasingly better digital technologies that include faster computers, better telemetry, that is, the communication of measurements from a collection point to receiving equipment, smaller, more accurate sensors, data management, and artificial intelligence.

An ontology approach seems to be an effective methodology for product lifecycle management (PLM). Ontology is a formal and explicit specifications of shared conceptualizations, representing concepts and their relations that are relevant for a given domain of discourse and serve as a means for establishing a conceptually concise basis for communicating knowledge for many purposes. In this paper a framework for representing a digital twin based on an ontology approach is proposed. It consists in the management of a number of ontologies, the product functionalities ontology, product configuration ontology, data pipeline ontology and operations and control ontologies. These ontologies represent the requirement and configuration knowledge, operational, control and predictive maintenance for a real customization of the product.

Literature Review

During the design phase, it is possible to virtually create the optimum solution and accurately render it operational before a single physical action is taken. Then we can simulate that solution under different types of real scenarios. It is the result of exhaustive simulations and rich data that specify areas such as best architecture, configuration, materials, and cost.  During operations, an abundance of data is being collected and fed back to its digital twin over a digital thread. Think of this as a data pipeline that enables analytics of various states and stages. Backed by artificial intelligence, the digital twin can identify and even predict maintenance issues before they happen. It has become a data-informed model of a physical system as Figure 1 demonstrates. This compelling feature reduces cost since it is typically cheaper to proactively conduct maintenance than to repair it after it is broken. Finally, this continuous real-time feed of data can help with optimization. That is, improve its performance by enabling the system to either automatically modify its own behaviour or by prompting the manual intervention of a human.

Digital twins have become particularly ubiquitous in the Internet of Things or IoT world. IoT devices are everywhere now, in our homes, across our cities, and in our factories, where we call them industrial IoT devices. These internet-connected electronics collect and produce data and services and interact and communicate with each other and central systems. The data collected from these devices creates detailed knowledge, enabling capabilities. The use of digital twins in the context of IoT will likely be one of the defining qualities of the future of this topic. With billions of new IoT devices being deployed and managed each year, it must be clear by now that digital twins have a remarkable future ahead. More advanced digital twins can produce specifications for the manufacturing process, including the production line, the metrics that must be captured during build, and testing procedures for parts and the product. As you can imagine, it is helpful to understand how a production line might perform for a given product before that production line is built. A digitally enabled manufacturing plant, called a smart factory, may have digital twins from design through to fully manufactured products, and even to post-deployment performance and maintenance in the field. This is the end-state of comprehensive product lifecycle management or PLM.