Wen 丨 Internet jianghu
God created human beings in their own image, and human beings wanted to create digital twins based on the way the physical world looked.
Based on the definition of digital twins by industry and academia, the Interpretation of Intelligent Manufacturing Terminology defines digital twin technology as: it is to make full use of physical models, sensor updates, running history and other data, integrate multi-disciplinary, multi-physical quantities, multi-scale, multi-probability simulation processes, and complete mapping in virtual space, thereby reflecting the full life cycle process of corresponding physical equipment.
I believe many people have heard of digital twins. Over the past few years, the popularity of digital twin technologies has been rising, attracting much attention inside and outside the industry. According to the Professional App of Tianyancha, there are nearly 600 digital twin-related enterprises.
However, in addition to some obscure and crooked concepts, most people do not have a clear and comprehensive understanding of the digital twin, and how to recognize and master it has become a problem for enterprises in the digital age.
"Crazy" digital twin
Digital twin, known in English as Digital Twin. Also known as digital mapping, digital mirroring.
In 2002, Professor Michael Grieves, who is engaged in product lifecycle management PLM, proposed a mirror space model in a product lifecycle management course: a virtual digital representation equivalent to physical products. (Although there is no documentary evidence, this is still widely considered to be the earliest source of digital twins.)
By 2010, NASA first proposed the concept of "digital twin", through virtualization, simulation technology and real-time status of aircraft, historical maintenance, health management and other data, the use of digital technology to replace a variety of physical twin objects, in order to meet the needs of the current stage of deep space exploration. Do you think this is the origin of the digital twin?
No, the source of the digital twin is actually the Physical Twin.
Nasa built multiple identical space vehicles in the Apollo program in the 1960s and 1970s, like "twins." During flight preparation, twins are heavily used for training; during missions, they are used to simulate alternatives on Earth models, where valid flight data is used to more accurately reflect flight conditions, thereby assisting astronauts in critical situations to make correct judgments.
With the development of ICT technology, the original "physical twin" entity is replaced by more and more digital model components and extended to the whole life cycle of the product, until a digital twin model that is completely consistent with the physical entity is formed. The advent of the digital twin has given the physical entity a super new stand-in. Since then, industrial manufacturing has become an incubator for digital twins.
Siemens can be said to be a loyal support pump for digital twins, using digital twins to run through the data model between all aspects of the product life cycle. Use digital twin simulation to simulate some of the actual operation processes of the factory, from product design to production line design, to the mechanical design of the equipment manufacturer and the planning and scheduling of the factory, to the final finished product. Examples include digital self-optimisation tools, predictive maintenance tools to reduce and plan downtime, and intelligent assistance systems for machine safety and operation, increasing machine productivity and reliability.
Based on the strong association between physical entities and virtual images, digital twins bring a whole new perspective to problem solving in industrial scenarios. By combining various explicit and implicit knowledge, structured and unstructured knowledge, activating the silent knowledge and data assets accumulated by the industry for many years, superimposing real-time and quasi-real-time dynamic operation data records on traditional industrial models, building an industrial digital twin, and helping people to re-understand and manage industrial manufacturing. It can be said that it is the industrial Internet that activates the digital twin.
Like dandelions, digital twins grew on the fertile soil of industrial manufacturing, after which seeds were blown by the wind to various industries and fields.
In the field of infrastructure engineering, for example, digital twins provide instant access to and real-time synchronization of building models that have already been designed and built, so that progress can be monitored in real time according to the timelines outlined in the 4D BIM model. You can also use model predictive control to verify the percentage of effort and clarify the project implementation process, making decisions based on the building forward simulation.
In the field of smart cities, unlike the "product life cycle" of manufacturing, the city as a complex giant system, the "life cycle" is longer, and its "life cycle" is always generating multi-dimensional massive data. As a result, urban digital twins are also more difficult to deploy in terms of data collection, processing, computing, storage, and management.
There are also potential opportunities in the challenges, inspiring forerunners to keep climbing. Alibaba Cloud's City Brain, Tencent's "Digital Government", Baidu's AI City, Huawei's "Urban Intelligent LifeForm", Haier's Haier Cloud's BIMCloud Digital Twin Platform, Ping An Technology's Smart City Cloud, and Inspur's Smart City Brain... All of them add content for digital twins to the smart city.
Even the metaverse has climbed the ladder with the digital twin. Of course, the two are inherently closely linked, and both are born out of the real world and interact with the real world. From the final orientation, the metacosm is more inclined to the always-on virtual world, while the digital twin emphasizes more emphasis on the simulation of the real world.
The metacosm is a larger and more complex system than the digital twin. In 1992, Metaverse was first proposed as a science fiction concept, and now it is gradually emerging, and it is even further away from maturity. The digital twin was born nearly a decade later than the metaverse, but it has blossomed and even borne fruit in various fields, and it is likely to grow in the metaverse in the future.
The world is made up of matter, information, and energy, and people tend to care more about the value of material entities and energy than to ignore information. In industry, for example, the valuation of a device is almost a world away from a set of software, and software is often bundled with machines, looking like "free gifts".
Professor Mr. Grieves of the University of Michigan, in his book Virtual Perfect Model: Driving Innovation and Lean Products, once said that "information is a substitute for wasted physical resources."
The emergence of digital twins has made people really begin to recognize the value of information and numbers, foresee the correctness of physical entities in advance, and avoid unnecessary risks and high-cost waste in the real world. Essentially, a digital twin is a rehearsal in which digital information replaces physical entities, allowing digital value to be truly reflected.
Digital twin: The intermediate state of the twin digital process
In fact, we still can't completely replace the technology with the word "digital twin", but should see it as the intermediate state of the process of twin digital technology, a transitional stage. In the view of the Internet jianghu, the digital twin has roughly five stages of development, namely physical twin, replication twin, digital twin, decision-making twin, intelligent twin and so on.
In the physical twin stage, the use of physical "twin" entities is costly. In addition to the practice of NASA mentioned above, before large-scale production in the industrial field, the physical prototype produced needs to accurately express the model appeal and move the production predictability forward, which is also a kind of physical twin.
The replication twin stage begins to shift the "twin" to the digital virtual level, at which time it is necessary to be able to make full use of historical data and real-time operation data to achieve accurate and comprehensive mapping of a product or system, and all of them are converted into information storage. At this time, it is equivalent to creating a physical body of a digital twin, but it still lacks the most critical soul, which cannot be used to control analysis, but can only be displayed.
The digital twin stage, which is currently being studied, can be dynamically simulated based on the physical design model of the physical ontology, as well as the real-time data fed back by the ontology sensor, and the historical data of the previous operation of the ontology. For example, the smart city is fully aware and updated in real time to form a real virtual holographic city.
In the decision-making twin stage, it is necessary to refine the scale perception of the data accumulated in the past, simulate the development scenarios of different external factors and different environmental backgrounds in the future, learn, analyze and summarize the operation rules of the physical ontology, and provide suggestions and references for decision makers. The advice provided at this time may be more ideal and deviate from reality, but there is also some reference value.
In the intelligent twin stage, AI is used to accurately analyze and plan and design responses, and make the most favorable judgment and autonomous control. It could also provide participants with new perspectives and even help shift from traditional ways of thinking to big data thinking, but this is still a distant goal.
Back to reality, the current digital twin stage is still in the research and exploration stage, and the specific application scenarios can be roughly divided into two categories: virtual simulation and monitoring and analysis.
Virtual simulation is often through the combination of fluid mechanics, thermodynamics, optics, electronics and other principles, the use of digital computers for auxiliary design and simulation, the real world of physical ontologies (equipment, buildings, cities, etc.), in the digital space to map the twin. It is possible to predict and verify the system performance of twins in a virtual environment, thereby ensuring design quality and robustness. This application is mainly based on design, so as to support engineers to carry out product innovation or system improvement, representative of Nexperia Asia Pacific, AspenTech, Ansys and other manufacturers.
Inspection analysis involves the monitoring and evaluation of the real operating conditions of physical entities. It is necessary to use a large number of dynamic data of the real environment, in the real operation process, through the sensor data collection, judge the environmental changes, system status, to achieve the monitoring, analysis, evaluation and early warning of the future operating state. Representative ones are Helishi, Jiyun Technology, PTC Thingworx, etc.
Essentially, virtual simulation and detection analysis represent two different centers of gravity for data flow. The focus of virtual simulation is that the physical body outputs sufficiently accurate and comprehensive data to the twin to minimize the errors caused by the virtual simulation. The focus of the detection analysis is that the data information fed back by the twin to the ontology should be accurate enough, and then the correct intervention is made to the physical ontology.
In the future, as digital twin technology matures, the difficulty in developing digital twins may no longer be the digital twin itself, but how well the combination of digital twin technology and the industry can be.
This requires companies to have "inverted T-shaped" capabilities. On the one hand, expand horizontally, master the knowledge and experience system of the industry, and lay a good foundation for the industry. On the other hand, master digital twin technologies and look for opportunities to combine digital twins and industries. If you compare the digital twin to a nail, and the industry that lands is like a plank, the key is whether you can find the right hammer to let the nail penetrate deeply.
Write at the end:
In the next few years, the industry with digital twins as the core will mushroom, and digital twins will become the standard for enterprises. However, outside of technology, digital twins still face challenges in terms of system, management, talent, etc., and behind these factors, it is essentially an evolution of a way of thinking, and the development of digital twins requires thinking first.
At the same time, there is a need to be wary of the conceptual pitfalls of digital twins. After all, the digital twin itself does not actively release value, nor can it exist alone as a productive force, and if it is only to sell the concept of digital twin, it will also face the wind of disillusionment at any time like a dream bubble