This spring, the SWN (Smart Water Networks) research group at TU Berlin and Einstein Center Digital Future (ECDF) organised the second edition of the annual “Data Science and Artificial Intelligence for Urban Water Management” Summer School. Twenty-nine students from TU Berlin, FU Berlin and Politecnico di Milano and some other guest listeners gathered online to participate in the school.

The students expanded their knowledge with lectures about Modelling and Control of Urban Water Systems, Leakage Detection in Water Distribution Networks, as well as Deep-Learning Techniques in Urban Water Management. Tutorials and group work revolved around Time Series Analysis, Data-driven Leakage Detection and Water Distribution Networks Modelling with Python.

With lessons learnt from last year, when the school took place digitally for the first time, there were some changes in the structure of the programme. The school took place in two sessions à 3 days, 19th-21st and 26th-28th May 2021. Lectures took place in the morning, while tutorials and group work were hosted in the afternoon sessions. Another novelty this year was the introduction of the Journal Club – aimed at guiding students through reading and analysing scientific papers – which took place in two sessions throughout the school.

Some things have not changed since the last digital Summer School, however. Just like last year, lectures took place on Zoom while lecture materials were made available and homework tasks were coordinated using TUB’s own Moodle platform (ISIS). The whole course was awarded with 6 ECTS and the exam consisted of three parts: a presentation of a paper in groups, individual oral questions on course content and an individual written project report with theory and applications.

According to the course coordinator and main lecturer Prof. Dr. Andrea Cominola, the main challenges this year were the same one faced in the previous edition, with the transition to a digital format and the ongoing limitations posed by the COVID-19 restrictions: “Creating a classroom feeling, fostering interaction and networking among students from different universities is not easy, as long as it’s not yet possible to complement the activities of the summer school with social events and field trips”. However, the main advantages with dividing the school into two separate sessions “consisted of allowing the students to better focus on the different activities and process the contents of lectures and tutorials with better timing, without the stress of digesting all content in one single week”. Moreover, he adds that “the bigger variety of activities and course content in this edition was possible thanks to the active contribution of all research assistants in the Smart Water Networks group and two amazing guest speakers from NTNU (Norway) and TU Delft (Netherlands)”. Speaking of Journal Club, he noted that “often, students do not get trained on critically reading scientific literature and face problems in doing it properly when they have to write their Bachelor or Master thesis. I was very glad to see that already after two sessions of the Journal Club, most of the students were able to identify the main relevant points from a scientific paper and critically analyze its content”. Overall, this initiative proved to be successful and will likely be repeated in the following years: “In line with what we are developing in the ide3a project, we also thoroughly collected feedback from students on each activity to identify room for improvement”. Prof. Dr. Cominola concludes with an outlook on the next edition: “I wish we could organize the next edition of the course in a blended format, with both online materials (maybe also podcasts and recorded videos), as well as some in-person activities. I am confident that this mix would enable overcoming the current challenges and enable a better learning experience for all students involved”.

All in all, the 2021 edition of the summer school “Data Science and Artificial Intelligence for Urban Water Management” turned out to be just as informative and engaging as the previous NWN schools, and the coordinators look forward to organizing the next edition in spring 2022!