MS019 - Digital Twins and Their Enabling Technologies for Infrastructures
Keywords: digital twins, infrastructures, real-time monitoring, scientific machine learning, surrogate modeling
Over the past decades, computer simulations have become indispensable in engineering, enabling deeper insights into physical phenomena and improving the design and operation of various processes. Within the field of computational mechanics, extensive research has been dedicated to developing efficient forward solvers to replicate physical systems virtually—these models are nowadays often referred to as virtual twins.
The field’s focus now gradually shifts towards digital twins, which extend virtual twins by incorporating real-time data, such as sensor measurements, to continuously update and calibrate the underlying models. This (often bi-directional) coupling between simulations and real-world data represents the next challenge in computational mechanics, offering unique opportunities that have already proven beneficial in various applications, including production engineering, life sciences, meteorology, and infrastructures.
In the realm of civil engineering and infrastructures, the digital twin concept offers novel capabilities for real-time monitoring and predictive analysis. Through digital twins, engineers and operators can optimize the performance, maintenance, and sustainability of critical infrastructure assets, ensuring resilience, longevity, and efficiency in the face of both routine and extraordinary challenges.
A key enabler for achieving these goals is the application of machine learning methods, which can process vast amounts of data while providing near-real-time predictions after an initial offline training phase. By leveraging these methods to create fast-to-evaluate surrogate models based on high-fidelity simulations, engineers can effectively tackle complex, multi-query tasks such as optimization and control.
Creating reliable digital twins of infrastructures is an interdisciplinary challenge requiring collaboration across a wide range of fields, including sensor technology, numerical algorithms, and machine learning, as well as virtual and augmented reality (e.g., for immersive maintenance training environments) and even the humanities (e.g., to study the acceptance of digital twins by political decision-makers).
Therefore, this minisymposium aims to bring together experts from these diverse fields to engage in interdisciplinary discussion and foster scientific exchange and collaboration.
The field’s focus now gradually shifts towards digital twins, which extend virtual twins by incorporating real-time data, such as sensor measurements, to continuously update and calibrate the underlying models. This (often bi-directional) coupling between simulations and real-world data represents the next challenge in computational mechanics, offering unique opportunities that have already proven beneficial in various applications, including production engineering, life sciences, meteorology, and infrastructures.
In the realm of civil engineering and infrastructures, the digital twin concept offers novel capabilities for real-time monitoring and predictive analysis. Through digital twins, engineers and operators can optimize the performance, maintenance, and sustainability of critical infrastructure assets, ensuring resilience, longevity, and efficiency in the face of both routine and extraordinary challenges.
A key enabler for achieving these goals is the application of machine learning methods, which can process vast amounts of data while providing near-real-time predictions after an initial offline training phase. By leveraging these methods to create fast-to-evaluate surrogate models based on high-fidelity simulations, engineers can effectively tackle complex, multi-query tasks such as optimization and control.
Creating reliable digital twins of infrastructures is an interdisciplinary challenge requiring collaboration across a wide range of fields, including sensor technology, numerical algorithms, and machine learning, as well as virtual and augmented reality (e.g., for immersive maintenance training environments) and even the humanities (e.g., to study the acceptance of digital twins by political decision-makers).
Therefore, this minisymposium aims to bring together experts from these diverse fields to engage in interdisciplinary discussion and foster scientific exchange and collaboration.
