Physics-informed machine learning (PIML) is a technique that combines traditional machine learning models with the principles of physics [1]. Unlike standard machine learning approaches, which rely solely on data, PIML integrates physical laws, such as conservation of energy or fluid dynamics equations [2], into the neural network training. This integration allows the models to be more accurate, generalizable, and interpretable, particularly in scenarios where data is scarce or noisy. PIML is further useful in engineering and scientific applications as a novel computational approach where the usage of traditional computational techniques is challenged.

This mini-symposium will feature in-depth presentations covering recent advancements and future directions in PIML for modelling and simulation across engineering and science. We invite talks on a wide range of PIML topics, including methodologies, numerical implementation techniques, computer programming, and various applications in engineering and science. The objective is to foster discussions on the application of PIML in tackling complex engineering challenges including nonlinearities, multi-scale modelling, multi-physics and more. The symposium will provide a platform for field experts to share their insights, promote interdisciplinary collaboration, and explore future directions in PIML modelling within computational mechanics.