Digital twins are one of the most fascinating expressions of digital transformation, particularly the one that refers to the 4.0 paradigm, both in the industrial field and in other civil applications. The concept on which the technology is based dates back to the 1960s when a telemetry system developed by NASA managed to bring Apollo 13 back to Earth after it was the victim of a dangerous failure during its journey into orbit. The story had enormous media coverage, captivating the whole world. It was the first demonstration of how a diagnostic simulation on a similar model could suggest solutions to a real problem on the actual model sent into space.
After the first applications in the manufacturing field, digital twins have gradually expanded in other industries, integrating into the most widespread CAD software in the design field and the product lifecycle management pipelines. Thanks to the evolution of hardware and software systems, the technologies related to the digital twin have progressively become democratized. What was initially available only to large enterprises can now be found in methods currently used by SMEs and consumer applications, thanks to the explosion of the Internet of Things (IoT) and the 3D technologies used to create virtual worlds.
This large symposium of emerging technologies, in addition to artificial intelligence and the field of physical-numerical simulations, is forming the basis for building ever more advanced and realistic immersive applications. It is an interactive context in which the digital twin constitutes one of the reference assets to ensure, among other things, the correspondence between the natural world and the virtual worlds of the metaverse.
What Is A Digital Twin?
In its broadest definition, the digital twin is a digital representation of a physical asset characterized by multiple instances to simulate a plurality of variants through virtual simulations. This concept is applied today to the manufacturing industry, technological systems, buildings, and infrastructure. There is no limit to its application scale, and three fundamental elements characterize it:
- The presence of a physical entity in the real world
- A software-defined digital twin
- A linking system that allows data to be exchanged between the two entities, making them interactive
For these reasons, the digital twin assumes a univocal nature between the physical and digital assets, which acts as a proxy. We must distinguish the digital twin from the 3D model of a specific object, which is equivalent to only one of the components involved. The CAD data, if lacking the data that allows the two entities to relate, would not be able to go beyond the geometric representation. In other words, the digital twin is not a simple digital copy of the actual model but a data system that establishes a relationship between the two counterparts, allowing, for example, to digitally simulate the possible effects of a change on the physical asset.
A digital twin is considered such only if there is a functional data relationship between the two parties involved, albeit highly variable in terms of synchronization, frequency, and level of accuracy, with methods of exchange on demand or in real time. Digital twins are typically created using CAD-based platforms that provide digital continuity throughout the product life cycle (PLM). This pipeline allows you to have all phases under control, including those that go beyond design and production. Think of the aspects related to the management and maintenance of an asset.
Special IoT sensors generally ensure the connectivity between the physical model and the virtual twin, and it is precisely the rapid diffusion of these systems that we have witnessed in recent years that allows the progressive democratization of a technology that until recently was considered highly specialized in the manufacturing sector. Through the IoT sensors, it is possible to establish a two-way data exchange between the physical model and the virtual twin, which helps carry out the required analyses and simulations.
What Is It For?
The digital twin creates a connection between an actual entity and a potentially infinite series of digital instances to know the current state of the physical part and to carry out analyses and simulations useful for decision support. The digital twin interfaces with all effects through business intelligence and business analytics software, as well as physical-numerical simulation systems capable of producing incredibly accurate forecasts about a given scenario.
Digital twins are therefore composed of a 3D CAD component and a set of data integrated with the various company systems, in particular ERP (Enterprise Resource Planning) and MRP (Material Requirement Planning), which are essential for managing what is defined as the BOM (Bill of Materials), i.e., the list of materials and components necessary for the realization of a product.
The digital twin can drastically optimize this management process, ensuring high visibility into the status of all the units involved. In the manufacturing context, where the first applications were created, the digital twin data is also used by the MES (Manufacturing Execution System), the system called upon to manage and monitor the actual production process on the factory floor.
The Benefits Of The Digital Twin
The benefits of a digital twin are easily understood because it can digitally simulate a series of operations that, in the traditional context, were carried out, for example, through physical prototyping or directly on the assets involved, with a decidedly higher expenditure of time and costs.
Digital simulation, which is realistic both from a visual (3D photorealism) and physical (physical-numerical simulation) point of view, makes it possible to evaluate in detail many aspects of a product’s behavior in various scenarios that would otherwise be physically impossible to predict. This aspect allows companies to achieve reduced time to market and general savings in development costs.
The digital twin, again thanks to the physically realistic 3D simulation, allows for the acceleration of research and development processes, significantly reducing the need for physical prototyping, which is usually limited to the advanced stages of the design. Similarly, thanks to IoT sensors, knowledge of the current state of a physical asset allows the digital twin to carry out real-time diagnostics and implement predictive maintenance procedures, which are much more convenient than scheduled maintenance and avoid unnecessary interventions without risking breakdowns and crippling service downtime on production lines.
Structure And Models
Over the years, as we have announced, digital twins have moved from the manufacturing context to a series of applications in very different fields. Each further evolution has therefore led to the need to follow as many structures and models as possible, which can be traced back to a series of standardized categories:
- Components (parts): The basic level of the digital twin, consisting of the essential elements, which are subsequently functionally associated
- Assets (products): combinations of two or more components that give rise to the interaction between the physical and digital assets
- Systems (units): sets of two or more assets that make up a real functional unit
- Processes: two or more systems that interact to satisfy one or more set objectives.
If, at a hierarchical level, the structure of a digital twin appears very simple and intuitive, also due to its being implemented on specific software platforms that are increasingly user-friendly, several obvious critical issues must be prevented and resolved effectively and consciously:
- Data management: digital twins involve large and numerical varieties of data, which must be managed in an agile and efficient manner to prevent chaos from taking over, which would make the digital model unmanageable to all intents and purposes and unable to communicate profitably with the counterparty physics.
- IoT systems: these are technologies that are, in some ways, still decidedly complex to configure, especially as regards the connectivity of the edge architectures that infrastructure them. Despite the evident progress in this direction, implementing an IoT system remains outside everyone’s reach.
- Data security: the digital twin, like all information systems, has a critical component of the interests of cybercriminals: data. These can be violated thanks to ransomware attacks, which aim to exfiltrate the most sensitive information and block the operating systems that guarantee companies the production and delivery of their services.
- System integration: the problem arises above all from two points of view. Firstly, the poor interoperability between different CAD and PLM solutions is because each manufacturer tends to use its technological standards. Further critical issues can arise when PLM provides obsolete systems, even if they are handy and functional for their purpose. Replacing them could represent an unjustified expense, but integrating them into a modern and efficient pipeline could be complex and entail obvious safety risks.his aspect mainly concerns IT and OT systems.
- Supply chain complexity: when many stakeholders are involved in a PLM, relating them and making them interact on a single database can become very complex due to the different technological and organizational solutions used.
Some Examples Of Use
The digital twin found its first applications in large enterprises and large public and private entities in the following application areas:
- Energy (Oil and Gas)
- Smart Cities
- Large civil infrastructures (bridges, dams, etc.)
The democratization of technologies that enable the digital twin has also made it possible to significantly expand the range of applications, with innovative home applications also involving SMEs and typically consumer contexts that do not strictly depend on large economies of scale.
Possible Future Developments
In the first place, the developers of technologies and systems linked to the digital twin will have to produce concrete answers to the previously mentioned critical issues, especially regarding data management and security, as well as the connectivity necessary for their flows to occur correctly and functionally. It is also foreseeable that digital twins will be extended to increasingly varied business sectors involving systems of varying complexity, thanks to the progressive reduction in costs associated with enabling technologies, especially IoT systems.
This democratization, in addition to making digital twins increasingly accessible to SMEs, will gradually allow them to be adopted in a more significant number of processes without limiting themselves to so-called mission-critical conditions. In the medium and long term, it will be reasonable to expect the digital twin to enter the orbit of immersive metaverse applications, given the high harmony of arguments in the relationship between the real world and 3D virtual worlds.
For this to happen, it will be necessary to implement digital twins of increasingly advanced cognitive functions involving artificial intelligence techniques and real-time data analysis. The result to which the current research aims consists in the desire to design an increasingly dynamic, active, and intelligent relationship between the assets that make up the digital twin to ensure a series of increasingly advanced and performing functions capable of simulating the increasingly complex and responsive nature of the real world.