Digital twin technology faces persistent challenges in physical property annotation, modeling automation, and real-time interactive capabilities that limit practical deployment. Recent advances in physical world generation models demonstrate automated annotation of fine-grained physical attributes and real-time simulation capabilities, yet remain disconnected from established digital twin methodologies. This paper proposes a novel framework that systematically embeds physical world generation models into digital twin architectures through a three-layer design encompassing physical sensing, automated generation, and enhanced digital twin modeling. The generation layer integrates PhysXGen’s five-dimensional physical annotations including absolute scale, material properties, functional affordances, kinematic constraints, and semantic descriptions with Genie 3’s real-time interactive environment synthesis and Dynamic World Simulation’s motion-reinforced consistency mechanisms. This integration enables automated construction of physics-grounded digital twin models with 40-50% improvement in physical property accuracy whilst potentially reducing modeling timelines from weeks to days in certain scenarios depending on system complexity and data availability. The framework maintains rigorous model assembly, fusion, verification, and management capabilities whilst adding automated physical content generation and interactive simulation at 24 frames per second. Applications across manufacturing, infrastructure, energy systems, and healthcare demonstrate the framework’s potential to enhance digital twin accessibility and fidelity for complex cyber-physical systems.
Li et al. (Wed,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: