PhD: Intelligent decision support system for water demand prediction

Code	i3WaterSDC5
Host Institution	Universitat Politècnica de Catalunya Intelligent Data Science and Artificial Intelligence Research Center
Location	Barcelona, 08014
Supervisor(s)	Main Supervisor: Dr Javier Vázquez Salceda (UPC, Spain) Co-supervisor: Dr Tatiana Mañunga (UC, Colombia) Industrial Mentor – Experts: Dr Joana Tobella (AGBAR Spain)
Research Field	Artificial Intelligence
Contract type	Fixed term contract
Application Deadline	September 15th, 2026

Description

Research Objectives

The main research goal of iWaterS is to provide, for the first time, rational analyses and explainable intelligent decision support of WDS to increase resilience to day-to-day incidents and to extreme events in the context of climate change such as floods or droughts, through new interdisciplinary and integral approaches for exploiting datasets (on-line, off-line), intelligent models (data-driven, numerical) and simulation results (digital twins, multiagent systems).

The specific research objective is developing innovative Artificial Intelligence methodologies to improve short-term water demand forecasting in Drinking Water Distribution Systems (DWDSs). The project aims to exploit individual consumption data and consumer behaviour modelling to support more efficient, adaptive and sustainable water network operation through an Intelligent Decision Support System (IDSS).

The research objectives are to:

Collect, curate, analyze and exploit anonymized individual water consumption data from real drinking water distribution systems, ensuring data quality, privacy and interoperability.
Characterize the consumption through Intelligent clustering techniques and automatic conceptual interpretation for consumer profile creation.
Design semantic models and consumer typologies that capture different demand patterns and support the interpretation of consumption dynamics.
Investigate and advance agent-based simulation models capable of reproducing consumer behaviour and generating profile-driven short-term water demand forecasts under different operational scenarios.
Integrate data-driven and knowledge-based AI techniques into an Intelligent Decision Support System (IDSS) to support demand forecasting and assess the impact of infrastructure modifications and operational interventions on water consumption.
Validate the proposed methodologies using real-world datasets from European water utilities and evaluate their robustness, scalability and transferability across different operational contexts.

Expected Results:

1) Dataset/s from WDSs gathered, filtered and analyzed.

2) Ontology of types of consumers and normal consumption patterns per type of consumer

3) Research progress in the use of profile-driven agent-based simulation models for short-term water demand prediction.

4) Deployment of a IDSS for short-term water demand prediction

5) Deployment of an intelligent recommender for personalized alerts to consumers.