The Faculty of Engineering (FE) at Notre Dame University-Louaize (NDU) recently held a seminar titled, “Surrogate Modeling and Machine Learning in Engineering,” organized by the Department of Civil and Environmental Engineering. This engaging event brought together experts and attendees from diverse backgrounds to explore the exciting applications of surrogate modeling and machine learning in the field of engineering.
The first speaker, Dr. Gérard-Philippe Zéhil, FE Associate Professor, delivered an enlightening talk, “On Surrogate Modeling in Engineering,” on the significance of surrogate modeling techniques in engineering, providing an overview of prevalent approaches such as analytical models, regression-based models, numerical model simplification methods, machine learning models, and model order reduction techniques. Zéhil showcased practical application examples, emphasizing the importance of incorporating principles of sustainability, optimality, and resilience into engineering projects.
Next was Dr. Christine Saab, FE Lecturer, explored the specific application of machine learning techniques in water quality monitoring in her presentation, “Machine Learning in Water Quality Monitoring”. The talk highlighted the challenges faced by water utilities in ensuring sustainable and safe water quality. Saab underlined the role of remote sensing technologies in efficient monitoring, showcasing various machine learning techniques in this area. Real-world examples and the potential for modeling non-deterministic phenomena were also presented.
Finally, Dr. Chady Ghnatios, Full Professor of Mechanical Engineering at the Arts et Métiers Institute of Technology, Paris, presented his lecture, “Physics-Informed Machine Learning with Engineering Applications,” wherein he focused on a hybrid modeling framework that combines classical scientific models with data-driven techniques to create digital twins. Ghnatios displayed multiple use cases, including improving simulations, developing stochastic modeling frameworks, and constructing time-dependent, data-driven models for forecasting applications.
The seminar fostered interdisciplinary discussions and offered valuable insights into the advancements in surrogate modeling and machine learning within engineering. The exciting possibilities in this field hold several implications that can change the landscape within these relevant applications. By harnessing the power of surrogate modeling and machine learning, the applications can propel innovation, enhance sustainability, and drive advancements in engineering.