Over the last decade, the digital twin paradigm has emerged across various scientific areas for online monitoring, control, and decision support, enabling diagnostic and predictive capabilities that are not achievable with computational models alone. A digital twin is a set of virtual information constructs that mimic the structure, context, and behavior of a natural or engineered system or process. It relies on computational models, is dynamically updated by assimilating data from its physical counterpart, has predictive capabilities, and guides decisions tailored to realize value in the corresponding physical setting.

Whether using physics-based simulators, data-driven predictors, or a hybrid of both, computational models serve as platforms for analysis, simulation, and optimization. At the same time, a synergistic connection between the physical and virtual domains is critical for effective digital twins. This bi-directional interaction forms a feedback flow comprising dynamic data-driven model updating and optimal decision-making. Computational efficiency is also crucial for handling the continuous assimilation of noisy and large-scale data, as well as for accommodating uncertainty quantification and propagation.

This session aims to gather contributions highlighting the impact of physics-based and data-driven computational methods in realizing digital twins of natural or engineered systems or processes. Contributors are invited to discuss topics such as, but not limited to, structural health monitoring and predictive maintenance, simulation and reduced-order models for digital twins, adaptivity in digital twins, verification and validation of digital twins, decision support using digital twins, data assimilation techniques for parameter and state estimation, integration of physics-based with data-driven approaches, multi-fidelity methods, and surrogate modeling.