PitchBook-新兴空间简报：数字孪生（英）-2023-6页

Digital twins are virtual replicas of physical assets created using sensor data, analytics, and other technologies, enabling real-time monitoring, modeling, and optimization across various sectors such as manufacturing, healthcare, and energy (e.g., simulating production lines or predicting vehicle maintenance).

Originating from NASA's simulators in the 1960s and formalized in the 2000s by Michael Grieves, digital twins bridge the virtual and physical worlds through continuous data exchange, supporting advanced applications like predictive maintenance.

Key technologies underpinning digital twins include big data (from IoT sensors and software), artificial intelligence (for data analysis), multimodal interactions (e.g., AR/VR for intuitive communication), secure connectivity (using networks like blockchain), and high-performance computing, which together enable a feedback loop for system optimization.

There are several types, such as component twins (replicating individual parts), asset twins (physical objects like machines), system twins (entire networks), and process twins (operations like supply chains).

Digital twins find applications in diverse industries: manufacturing for process simulation, healthcare for personalized medicine and drug development, smart cities for infrastructure management, automotive for testing and optimization, energy for renewable asset control, supply chain for logistics prediction, retail for consumer behavior analysis, and agriculture for resource management.

Despite potential, limitations include high resource costs, slow adoption with lengthy sales cycles, data privacy and security concerns, and lack of standardization, impacting widespread use.

The market is projected to grow significantly, valued at $15.6 billion by 2025 with a 33.8% CAGR, driven by investments from tech giants like NVIDIA and AWS, focusing on smart buildings and other use cases, backed by rising funding in startups.