ARTIFACT SUMMER SCHOOL 2026

06. July – 10. July 2026. DELFT, NETHERLANDS

Modelling Urban Floods and Nature-based Solutions with Physics & AI

The ARTIFACT Summer School 2026 brings together students and researchers in the field of water management to develop knowledge and practical skills in flood modelling. The programme combines physics-based 1D–2D modelling with machine learning and deep learning methods, offering a balanced mix of lectures, hands-on exercises and group work. Over five days, participants will learn how to collect and prepare data, build and run flood models, apply AI techniques for urban drainage and flood prediction, and take part in a concluding technical visit. In addition to educational program, special Early Stage Researcher program will be organised, along with social opportunities for relaxation and networking..

Event Details

  • Dates: July 06 – July 10, 2026
  • Location: Delft, Netherlands
  • Venue: IHE Delft Institute for Water Education, Westvest 7, 2611 AX Delft, and TU Delft, Mekelweg 5, 2628 CD Delft

Delft

Discover the city!

Program

See the program oveview!

Registration

Book our Place!

Program Overview

Preliminary AGENDA

Lecturers: Prof. Dr Zoran Vojinović, Prof. Dr Zoran Kapelan, Dr Milan Stojković, Dr Veljko Prodanović, Dr Roberto Bentivoglio

Day 1 (Monday July 6th), IHE Delft: FUNDAMENTALS OF PHYSICS-BASED 1D-2D FLOOD MODELLING

  • Introductory lectures on flood modelling, data processing and 1D modelling, followed by practical MIKE+ exercises and group work.

Day 2 (Tuesday July 7th), IHE Delft: FUNDAMENTALS OF PHYSICS-BASED 1D-2D FLOOD MODELLING (CONTINUED)

  • Focus on building 1D–2D models, preparing data for 2D modelling, running coupled models and visualising results. Includes group work and short presentations.

Day 3 (Wednesday July 8th), TU Delft: INTRODUCTION TO MACHINE AND DEEP LEARNING

  • Covers ML fundamentals, deep learning basics, and hands-on PyTorch exercises for modelling urban drainage flows.

Day 4 (Thursday July 9th), TU Delft: ADVANCED DEEP LEARNING METHODS FOR FLOOD MODELLING

  • Participants learn advanced AI applications: RNNs, CNNs, Graph Neural Networks and Transformers, followed by exercises and group work.

Day 5 (Friday July 10th), Engineering Agency, Rotterdam Municipality: TECHICAL VISIT

Who Should Attend?

This Summer School is for:

  • Students, researchers, and professionals interested in computer science, AI, water management, urban planning, and environmental science.

The program offers

  • An introduction to physics-based 1D–2D flood modelling, including data preparation, model setup and result interpretation;
  • Hands-on experience with professional tools, such as MIKE+ for hydraulic modelling and PyTorch for machine-learning exercises
  • Practical work on real flood scenarios, supported through guided group tasks and daily presentations;
  • Training in advanced deep-learning approaches, including RNNs, CNNs, Graph Neural Networks and Transformers applied to flood prediction;
  • Young researcher social and work program: Details to be confirmed early 2026

Registration Details:

  • Application Deadline: May 16, 2026
  • Available places are limited, hence early registration is recommended.
  • Registration is free and includes the full academic program and lunches on days 1-5 of the summer school. Travel and accommodation are self-funded.

Have questions or need more information?

Feel free to get in touch:
artifact@ivi.ac.rs