PhD - Course 2022: Aveiro, Portugal

Given that new paradigms such as deep learning and explainable AI are increasingly permeating quantitative studies, parameter identification, and pattern recognition demands on PhD students are growing with the complexity of research topics and application requirements. Unfortunately, current developments in these fields are fast and cannot be easily incorporated into standard courses in Engineering or in individual courses by partner institutions. Therefore, there is a common need and high demand for specialized training courses on these topics.

To address these issues in the framework of the PARFORCE project, special training courses are carried out At the University of Aveiro, Portugal by using a blended teaching approach with the ultimate goal to create online courses, while allowing to bundle resources and expertise to impart knowledge in a very comprehensive and goal‐oriented way and providing students with the opportunity to create their own networks in a European environment.

Conduction of the course

The course was divided into two parts: An online part which took place between (21st February – 10th March 2022) with two lectures per week, and a one-week hybrid part which took place (14th to 19th March, 2022), where selected number of students traveled to Aveiro, Portugal and were taught in a classic presential environment while a second group followed the lecturers online. This allows using these courses as a testbed for a comparison between online and in-class training and teaching methods. To this end, the online part of the course were given by a combination of synchronous and asynchronous training to boost students’ background for the hybrid part which can then be efficiently used for the project work and discussion/reflection of knowledge. Furthermore, participants composed of a combination of students from civil engineering with different specializations as well as students from Mathematics. This allows them to follow teaching activities and project work in form of interdisciplinary teams and to experience modern scientific knowledge in realistic training settings.

Participating students worked together on realizing of civil engineering experiments in a virtual environments, the first project involved a force vibration experiment on the multistory RC frame structure to investigate its dynamic properties, here students recorded a real experiment with 360 cameras, and created a virtual video with scientific information regarding setup, procedure and analaysis of obtained data in a VR environment, participants also contributed with a virtual tour of the experiments setup. (VR tour)

The second project involved fire resistance experiments, the virtual experiment created by the participants aims at teaching students different procedures for testing fire resistance, integrity, and insulation, as well as understand their differences in terms of testing systems. Based on the results of the experiment, the student should be able to determine the fire resistance classification of the tested specimen. Students should know the difference between the resistance (R), integrity (E), and insulation (I) requirements (MRI) in order to accurately identify a specimen, according to EN 13501-2. Furthermore, students should understand how to classify a system when its resistance, integrity, and insulation are not the same (e.g., if the class of the integrity of the system is E90, but the insulation falls into class 60, the whole system is classified according to the less favourable criteria, i.e., EI60).

Student learning outcomes

1. Participants were taught:
   • to choose and implement correct regularization algorithms for a given inverse problem;
   • to pre-process (denoising/deblurring/etc.) and analyze experimental data;
   • to evaluate the quality of numerical methods/algorithms for a given mathematical model, make a correct choice and be able to choose the correct method in commercial programs;
   • to analyze errors and their propagation in mathematical models;
   • to implement and train a machine learning algorithm;
   • to setup and properly train an artificial neural network.
2. Students had the opportunity to work in an international environment with students from 4 different countries and practice their management, presentation, and language skills