AI-based spatio-temporal forecasting for water demand and leakage
From reactive to proactive water management: AI for smarter, more resilient water networks
People:
Partners:
Media links:
Technical approaches:
Machine Learning; Bayesian Modelling; Time-Series Analysis; Network Science; Uncertainty Quantification; Explainable AI
Challenge areas:
Development of AI-based spatio-temporal forecasting systems that learn how water demand and leakage behaviour evolve across both time and interconnected network locations.
By combining deep learning, representation learning and network-aware modelling, the work aims to forecast demand, identify emerging leakage patterns and support earlier, more targeted operational decisions for reducing water loss and improving network resilience.
Find out more about the AI-forecasting for water demand and leakage
