Short Courses

As part of the activities of the conference, a number of Workshops / Short Courses have been planned. The Short Courses will be held on the morning and afternoon of Monday, 18 May 2026. The structure of these courses will be:

Typical duration: 2-3 hours.
Some courses will run in parallel. It will only be possible to attend one of them.
The Short Courses can be attended by any registered conference participant, attending in person.
The number of places on each course is limited. In case of exceeding the limit, preference will be determined by the order in which the complete registration process (including payment) is completed.
Once the places for each course have been filled, the course leaders will contact the participants to inform them of the course requirements (hardware, software).

* Registration for the Short Courses takes place at the same time as the conference registration.
** Participants can register for one Morning Session Short Course and/or for one Afternoon Session Short Course.
*** Short Courses are only available to registered conference participants that are attending in person.

Short Courses Program

Monday, May 18, 2026

	Time	Parallel Session 1	Parallel Session 2	Parallel Session 3
Morning Sessions	9:00-9:30	Registration
	9:30-11:00	SC1: Modern Optimisation for WDS	SC2: AI for Water Professionals
	11:00-11:30	Coffee Break
	11:30-13:00	SC1: Modern Optimisation for WDS	SC2: AI for Water Professionals
	13:00-14:00	Lunch/Registration
Afternoon Sessions	14:00-15:30	SC3: Physics-informed DL surrogates	SC4: Water allocation and infrastructure planning	SC5: Human-centered leak detection (2 hour)
	15:30-16:00	Coffee Break
	16:00-17:30	SC3: Physics-informed DL surrogates	SC4: Water allocation and infrastructure planning

Short Courses Description

Course ID	Course Title	Course Organizer
SC1: Modern Optimisation for WDS	From Best-Guess to Better Decisions: Modern Optimisation for Water Distribution Systems	Prof. Dragan Savic FREng and Lydia Tsiami
SC2: AI for Water Professionals	AI for Water Professionals	Dr Elad Salomons University of Haifa, Israel
SC3: Physics-informed DL surrogates	A Practical Introduction to Physics-informed Deep Learning for Surrogate Models of Drinking Water Systems	Dr. André Artelt Bielefeld University, Germany
SC4: Water allocation and infrastructure planning	Optimizing transport and treatment infrastructure to balance water demand and supply on a regional scale.	Prof Mirjam Blokker KWR and Delft University of Technology, the Netherlands
SC5: Human-centered leak detection	Co-designing integrated, robust, and human-centered leak detection technology	Prof. Andrea Cominola, PhD. and Dr. Ella Steins Technische Universität Berlin – Einstein Center Digital Future, Germany

SC1: Modern Optimisation for WDS

Short Course Title: From Best-Guess to Better Decisions: Modern Optimisation for Water Distribution Systems

Duration: 3 hours

Short Course Organizer: Prof. Dragan Savic FREng and Lydia Tsiami (PhD Candidate)

Instructors:

Prof. Dragan Savic FREng
Mark Morley, PhD
Lydia Tsiami (PhD Candidate)
Dennis Zanutto (PhD Candidate)
Christos Michalopoulos (PhD Candidate)

Description of the course:

Optimisation is a powerful ally for water professionals. It can deliver substantial savings and help balance multiple objectives in complex problems. It can support how utilities design networks, plan reinforcements operate their networks, but most of these approaches quietly assume a single “best guess” future. What happens when that future never arrives? Demand is uncertain, energy prices jump, cities grow, regulations tighten, assets deteriorate, and suddenly yesterday’s “optimal” solution starts to crack.

This short course explores how optimisation and modern decision-making strategies can move beyond single-scenario thinking. We look at how a few key modelling choices, namely the length of the planning horizon, the staging of decisions, and daily operations, interact with one another. We will explore how modern planning frameworks, including staged, robust and flexible approaches, can help systems perform better under uncertainty. We also discuss how optimisation can find its place in water utilities practice, rather than living only in research papers.

Participants will get an overview of optimisation in the context of water distribution networks, and see a range of formulations in practice, from classical approaches to more recent methods that explicitly consider uncertainty, staged decision-making and the role of operations. These ideas will be explored through simple, hands-on modelling examples in Python. We will also demonstrate a software platform tailored for water professionals and utilities, designed to make these optimisation approaches easier to apply in practice. Finally, we will briefly touch on modern developments such as agentic AI and how they might support decision-makers in exploring options and stress-testing plans.

Intended audience:

The course is aimed at both newcomers and practitioners with some optimisation experience who are curious about where the field is heading and how these ideas can be applied in real-world water distribution planning. Basic familiarity with Python and water distribution modelling is helpful.

SC2: AI for Water Professionals

Short Course Title: AI for Water Professionals

Duration: 3 hours

Short Course Organizer: Dr Elad Salomons, University of Haifa, Israel

Instructors:

Prof Mashor Housh, University of Haifa, Israel
Prof Lina Sela, University of Texas at Austin, US
Dr Elad Salomons, University of Haifa, Israel

Description of the course:

This short, hands-on course introduces water professionals to the practical use of artificial intelligence (AI) tools in everyday utility and engineering tasks. In just a few hours, participants will see how AI can make routine work faster, easier, and more insightful, without needing to be a programmer or data scientist. Using familiar datasets and examples from real-world water operations, the session demonstrates how AI can help analyze pump and system performance, detect anomalies in SCADA or AMI data, design a simple water distribution network, and its operation. Throughout the session, participants will use AI tools (mainly ChatGPT) directly to complete small, realistic exercises that mirror the daily challenges faced by water utility engineers, analysts, and managers. This introductory session is perfect for water-sector professionals curious about applying AI to data analysis, planning, and operational problem-solving, with no coding experience required.

The course is a condensed version of the full two-day AI for Water Professionals Workshop, which has already been delivered several times successfully to water utility professionals. The complete workshop provides in-depth, hands-on training in applying AI to data analysis, hydraulic modeling, and decision support. This abbreviated version focuses on selected examples and demonstrations, offering a practical introduction to how AI tools can be used to analyze water system data, support operations, and enhance efficiency. It’s designed for participants who want a focused, accessible overview of what AI can do for the water sector.

Intended audience:

This short course is designed for a diverse audience interested in the intersection of water management and artificial intelligence:

Water professionals with limited AI experience: engineers, analysts, and utility staff who want to understand how AI tools can make their daily work more efficient. The course focuses on practical, easy-to-use examples, with no programming or data science background required.
Students and early-career researchers: those studying water resources, environmental engineering, or related fields who wish to see how AI is applied in real-world water utility contexts. The course bridges the gap between academic learning and professional practice.
Trainers and facilitators: individuals or organizations interested in running their own full AI for Water Professionals Workshop. This short session provides an overview of the structure, materials, and teaching approach used in the extended two-day version, serving as a foundation for those who want to deliver similar training programs.

General notes:

Each participant must bring a laptop.
Participants are required to have a ChatGPT Plus subscription or an equivalent product. Most services allow you to register a premium account for a monthly fee and cancel it at any time.
The workshop is limited to 25 participants.

SC3: Physics-informed DL surrogates

Short Course Title: A Practical Introduction to Physics-informed Deep Learning for Surrogate Models of Drinking Water Systems

Duration: 3 hours

Short Course Organizer: Dr. André Artelt, Bielefeld University, Germany

Instructors:

Dr. André Artelt, Bielefeld University, Germany
Inaam Ashraf, Bielefeld University, Germany
Luca Hermes, Bielefeld University, Germany

Description of the course:

Rapid urban population growth, among others, makes drinking water system analysis an increasingly challenging task. Often, simulators such as EPANET are used in various tasks, such as planning, state estimation, event diagnosis, etc. However, such simulators are computationally expensive and scale poorly in time and space, challenging the development of digital twins for improved system analysis. In this context, recent work highlights the potential of AI methods for building efficient surrogate models of the simulator that can be used as part of a digital twin.

In this training course, we provide the participants with the necessary knowledge of using physics-informed Deep Learning to develop high-fidelity surrogate models. The course will include many hands-on examples in Python and be split into three parts.

Recap of Machine Learning (ML) basics and foundations of (physics-informed) Deep Learning. A special focus will be on pre-processing strategies, evaluation metrics and procedures, incorporation of different types of physics, and hyper-parameter tuning.
Data generation for training and evaluating ML models, such as deep neural networks. We will discuss and demonstrate the selection and creation of appropriate benchmarks for building and evaluating the ML models. Most importantly, we will discuss scenario generation in Python, including how to break correlations and incorporate parameter uncertainties to ensure high fidelity of the final ML models.
A case study on a physics-informed Graph Neural Network (GNN) based hydraulic surrogate model. After an introduction to Graph Neural Networks (GNNs), we present and explain a hydraulic surrogate model based on physics-informed GNNs. In this context, we will discuss and demonstrate how such a surrogate model can be used in a digital twin for a variety of different tasks, such as state estimation from sparse sensor readings, and network rehabilitation.

Intended audience:

Practitioners and researchers that are interested in exploring the potential of AI (i.e., Machine Learning and Deep Learning) for building surrogate models that can be used in digital twins.
Basic knowledge of Machine Learning is required.

SC4: Water allocation and infrastructure planning

Short Course Title: Optimizing transport and treatment infrastructure to balance water demand and supply on a regional scale.

Duration: 3 hours

Short Course Organizer: Prof Mirjam Blokker, KWR and Delft University of Technology, the Netherlands

Instructors:

Dr. Joeri Willet, KWR, the Netherlands
Dr. Mark Morley, KWR, the Netherlands
Prof Mirjam Blokker, KWR and Delft University of Technology, the Netherlands

Description of the course:

The Water MatchMaker (WMM) is a design methodology to optimize water infrastructure on a regional level to achieve a sustainable balance between water supply and demand. WMM was developed to answer the following questions:

What is the optimal water allocation within the research area?
Which transport and treatment (and in the future storage) infrastructure must be used, constructed, or expanded to make this water allocation possible?

During this workshop we will discuss the suggested approach and work on a hypothetical case study to show and experience how the design methodology works.

By working on a small-scale hypothetical water allocation problem for which infrastructure decisions must be made participants will experience the complexities introduced by the mismatches between water demand and water supply. We will focus on two of the three possible mismatches between water demand and supply: a mismatch in place (the locations of demand and supply do not necessarily coincide) and a mismatch in quality (the quality of the available water is not everywhere the same, and not every function requires the same quality). A spatial reconfiguration of demand, creating new transport options within the water system, tapping the potential of alternative sources (including treated effluent), and applying purpose-specific treatment are all part of the puzzle. The temporal aspect, a mismatch in time (demand and supply vary over time) is part of the WMM methodology but is not part of this workshop.

The mismatches introduced above go beyond the (artificial) institutional boundaries within which most water systems are currently managed. If these institutional boundaries are transcended, it becomes possible to design an alternative system that accounts for the system wide effects of measures. Participants are encouraged to think about the way methodologies like WMM can be used to facilitate collaboration across institutions, both within and beyond the water sector.

Intended audience:

We encourage participants to form teams of 2-4 people. A mix of hydraulic modellers, optimization and water utility experts may yield the best score. A background in optimization is not needed.

SC5: Human-centered leak detection

Short Course Title: Workshop: co-designing integrated, robust, and human-centered leak detection technology

Duration: 2 hours

Short Course Organizer: Prof. Andrea Cominola, PhD. and Dr. Ella Steins, Technische Universität Berlin – Einstein Center Digital Future, Germany

Instructors:

Prof. Andrea Cominola, PhD., Technische Universität Berlin – Einstein Center Digital Future, Germany
Dr. Ella Steins, Technische Universität Berlin – Einstein Center Digital Future, Germany

Description of the course:

Explore the iOLE platform for smart leak detection! Learn, test, and shape new digital tools for resilient water networks.

Efficient water management is becoming increasingly important as utilities face growing challenges from aging infrastructure and resource scarcity.

Within our BMBF-funded project iOLE (intelligent Online LEakage detection), we are developing intelligent methods to detect and locate leaks in water distribution networks quickly and reliably, helping operators minimize losses and improve network resilience.

This workshop will showcase the project’s results and provide participants with an introduction to the integrated, robust, and human-ceneterd iOLE platform for intelligent online leak detection. You will discover how our award-winning algorithms (DualModel and LILA) work in an integrated platform to detect and localize leaks, how the reliability of the detection results is ensured, explore our platform’s user interface, share your feedback and actively co-design its further development.

The event offers a unique opportunity to exchange ideas between researchers, water suppliers, water distribution network operators, and young water professionals promoting collaboration between science and practice in smart water management.

Preliminary agenda:

10 min: Welcome and opening remarks
20 min: Why leakages matter. The iOLE project.
30 min: Discover the iOLE interface for hybrid leak identification & localization
30 min: What is affecting leak detection? Insights from sensitivity analysis
15 min: Feedback round and Q&A
15 min: Next steps in iOLE and future collaborations

Intended audience: water suppliers, water distribution network operators, researchers, young water professionals.