Machine learning has emerged as a powerful tool for tackling various problems in engineering and science. Typically via the use of large volume of data, deep neural nets can be trained for this end. However, for engineering and science problems, big data is not enough, and is not always available. This course will introduce the newly emerged paradigm and research trend called “physics-informed machine learning”, where physical laws or physical prior knowledge can be enforced into the architecture of machine learning models, to boost the training and promote the trained models to be more physically consistent and generalizable. At the end of the course, students are expected to understand the principles and methods of physics-informed machine learning, and to have hands-on implementations with Python.