The digital twin paradigm represents a significant shift in engineering and science, establishing high-fidelity virtual models that operate in real-time. However, the immense computational cost of complex simulations and the uncertainties inherent in the connection between physical systems and their virtual models continue to pose significant challenges. This minisymposium addresses those challenges by focusing on recent advances in key enabling technologies. The scope includes the incorporation of AI-powered methods for integrating operational data and managing uncertainties, with topics including data-driven approaches and physics-informed neural networks. Concurrently, the minisymposium will cover advanced large-scale computational techniques for the efficient and scalable simulation of complex systems. These techniques include, but are not limited to, recent developments in domain decomposition methods, reduced-order models, and advanced interface mechanics for multi-physics analysis. The primary objective is to provide a collaborative platform for researchers from academia and industry to discuss both fundamental and applied aspects. This forum will facilitate the dissemination of recent advancements and foster new collaborations for creating the next generation of digital twins.