The Blended Intensive Programme on Physics Informed Machine Learning Applications to Geotechnical Engineering is organised by the Institute of Geomechanics and Underground Technology of RWTH, led by Professor Raul Fuentes, and has been designed exclusively for students of the ENHANCE Alliance.

The program consists of two parts: an introductory course that will take place online on July 18 and 19, 2022, and a hackathon workshop that will be hosted on-site at RWTH from July 25 to 29, 2022. The program is aimed at Master’s students in engineering, geophysics, computer science, and applied mathematics. Students can earn 3 ECTS for the course and receive an Erasmus+ grant from their home university. The application deadline is June 13, 2022.

The course objective is to gain an understanding of the principles of physics-informed machine learning and explore different situations in which these methods can be applied. The program is structured as follows:

  1. Online Introduction (July 18 and 19, 2022)

The online course consists of lectures, drop-in sessions, and self-study. Two weeks prior to the on-site workshop, reading materials are provided. These include literature on data preparation, optimisation, and an introduction to machine learning. A particular focus will be on neural networks and their relationship with gradients and differential equations. This will also feature an introduction to the tools the students will work with, mainly Python code written in Jupyter Notebooks. On July 18 and 19, lectures regarding the topics will be held online. After the lectures, drop-in sessions are hosted where the materials and questions can be discussed.

  1. Workshop at RWTH (July 25 to 29, 2022)

During this full week, the students will apply their newly gained theoretical knowledge in a hackathon-style workshop. Different tasks and problems will be provided, which will be solved by the students in a collaborative and informal setting. The goal of this session is to deepen the skills gained in the theoretical seminar and share knowledge among participants. The practical nature of the workshop will also lead to a better understanding of the implementation, chances, and limitations of physics-informed machine learning methods. Last but not least, the hackathon is an excellent opportunity to crack interesting modelling problems and network with students and lecturers from different countries. Besides the coursework, social events such as BBQs and local tours will be held.

Photo: Jan von Allwörden/DAAD