Abstract:

The study of creating an Atlas of the Industry is a sub-project within our center's research on the transformation of scientific and technological achievements. The objective is to gain a comprehensive understanding of the core technologies behind cutting-edge innovations, the technological competitiveness of pioneering enterprises, and the progress of researchers, thereby enhancing the efficiency of technology transfer. This research will be conducted in a series, focusing on commercial application scenarios in key national strategic scientific and technological fields, with a strong emphasis on timeliness.

This report is part of the Computer Science series within the Atlas of the Industry research: The China Digital Twin Atlas. Digital twin technology, as a cutting-edge innovation, enables the precise mapping of physical entities in virtual space, creating a "digital twin." This technology leverages the Internet of Things (IoT) for real-time bidirectional data exchange, allowing virtual entities to comprehensively reflect the lifecycle of their corresponding physical entities. It supports simulation, prediction, and optimization decisions based on integrated underlying data information. Originating in the U.S. aerospace and military sectors, digital twin technology was first applied to industrial production by General Electric (GE). Driven by giants like Siemens and Dassault, the technology spread from the U.S. to Europe. With advances in artificial intelligence, IoT, virtual reality, and the rise of the metaverse concept, digital twin technology has been continuously refined, demonstrating significant potential in urban management, smart industry, autonomous driving testing, and healthcare, thus becoming a crucial technology for driving digital transformation and upgrades across various industries.

Based on technological complexity, digital twins are categorized into five levels. The technology is currently in a rapid growth stage, driven by the development of the digital economy, industrial internet growth, policy support, technological advancements, and increasing market demand. Key technologies of digital twins include modeling, rendering, and simulation. Modeling technology creates digital models of physical entities through methods like 3D scanning, parametric modeling, and reverse engineering, with representative software including SolidWorks and CATIA. Rendering technology achieves realistic visual effects through Physically Based Rendering (PBR), real-time rendering, and cloud rendering, with mainstream platforms like Unreal Engine and Unity. Simulation technology uses Finite Element Analysis (FEA), multiphysics coupling, and real-time simulation to model entity behavior, with major software including ANSYS and ABAQUS. These technologies require collaboration in fields such as computer graphics, Computer-Aided Design (CAD), Computer-Aided Engineering (CAE), and physical simulation. Industrial software giants like Siemens, Dassault, Autodesk, and GPU manufacturers like Nvidia and AMD are driving the development of digital twin technology, with future technological integration expected to mature digital twins.

The main participants in the digital twin industry are divided into technology service providers and integrated solution suppliers. Technology service providers include Computer Integrated Manufacturing (CIM), Building Information Modeling (BIM), and visualization platform manufacturers, while integrated solution suppliers include operators and large internet companies. The competitive barriers for digital twin companies consist of technology, business, and resources. As technology develops, digital twins face challenges such as business model maturity, high technical support requirements, lack of unified standard systems, and incomplete data capabilities.

Digital twin technology has shown significant application value across multiple industries, promoting urban management, smart industry, autonomous driving testing, and smart healthcare. In urban management, digital twin cities use IoT and Geographic Information Systems (GIS) for real-time monitoring and optimization, applied in traffic management (e.g., Siemens and IBM's smart traffic systems in Singapore), zero-carbon smart parks (e.g., Microsoft's Redmond campus), and emergency management (e.g., GE's solutions). In smart industry, digital twins cover all stages of discrete and process industries, improving production efficiency and equipment reliability, with major application companies including iSoftStone, GE, Siemens, and IBM. For autonomous driving testing, companies like Waymo, Tesla, Baidu, and Aptiv use digital twins for large-scale virtual simulation testing, enhancing system performance and safety. In smart healthcare, platforms like IBM's Watson Health and GE Healthcare's Edison realize equipment management, surgical simulation, and personalized treatment, significantly improving healthcare quality and efficiency. Overall, digital twin technology provides powerful innovation drivers and efficiency enhancement methods for various industries.

In the field of digital twin technology research in universities, progress between Chinese and foreign institutions is generally on par, but research focuses and directions differ. Internationally, especially in the U.S. and European universities, the emphasis is on fundamental theories, high-precision simulations, and multiphysics coupling research, focusing on complex system modeling and interdisciplinary applications, with leading universities including MIT, Georgia Tech, and Johns Hopkins University School of Medicine. Chinese universities excel in application-oriented and large-scale system integration, particularly in smart cities, traffic management, and infrastructure construction, with leading universities including Tsinghua University, Beihang University, and National University of Defense Technology. Chinese universities also promote the practical application of digital twins by combining cloud computing, IoT, and 5G technology. International universities focus on technical depth and interdisciplinary integration, while Chinese universities focus on application breadth and system efficiency, creating a competitive yet cooperative research ecosystem.

Digital twin technology faces multiple challenges in its development, including immature business models, high initial investments, weak user demand, difficulty in replicating customized solutions, and high costs, limiting its promotion and implementation. The standardization dilemma is also significant; currently, there are no unified standards for data collection scales, parameters, formats, and cycles, leading to difficulties in data integration and interface docking. The lack of uniformity in technical frameworks and protocols also complicates project integration and docking. Technical support-wise, digital twin construction involves large models and data volumes, requiring powerful computing and processing capabilities in computer hardware, and poses challenges for high interaction, high immersion, and high-definition display on terminal devices. Insufficient data capabilities also restrict the development of digital twin technology, including low data quality, incompleteness, data format and quality differences, and data security and privacy protection issues. Addressing these challenges requires the concerted effort and coordination of the entire industry. In the future, industry participants need to work together to build an open and win-win digital twin ecosystem, promoting the healthy development of digital twin technology.

Full Text: Atlas of China's Digital Twin Industry