WATER-SENSE | AI-driven water efficiency and anomaly detection

Water analytics for tracking events and resource efficiency with secure, enhanced networked sensor environments 

Water scarcity and increasing consumption pressures require smarter approaches to water management. To address these challenges, WATER-SENSE develops AI-driven methods to monitor, analyse and optimise water consumption in buildings, helping organisations detect inefficiencies, reduce water losses, and improve resource efficiency through real-time insights and actionable recommendations.
 

Target group

The project is relevant for:

• Facility and property managers
• Water utilities and network operators
• Industrial companies seeking to optimise process water consumption
• Public authorities and municipalities responsible for water management

Context

Water scarcity is becoming an increasingly important challenge due to climate change, urbanisation and growing demand for resources. At the same time, significant volumes of water are still lost through leaks, abnormal consumption patterns and inefficient usage that often remain undetected.

Current water management approaches are largely reactive and rely on limited monitoring capabilities. Existing solutions typically provide only aggregated consumption data and rule-based alerts, making it difficult to identify the root causes of inefficiencies or detect anomalies in real time.

While large volumes of consumption and contextual data are becoming available, organisations still lack scalable tools capable of transforming this information into actionable insights. 

Objectives & results

WATER-SENSE develops AI-based methods to help organisations better understand their water consumption, detect anomalies more quickly and optimise resource use.

The project aims to:

• Improve visibility into different water uses through detailed consumption analysis
• Help organisations optimise water consumption through analytics and decision support tools
• Provide actionable recommendations to optimise water consumption
• Develop robust, scalable and privacy-preserving AI solutions

Key expected results

• AI models capable of analysing water consumption patterns in detail
• Pilot demonstrators validating the performance of the solutions in real-world environments
• Actionable recommendations to improve water use efficiency
• Best-practice guides and publications supporting the adoption of these solutions
• Scientific publications on water consumption analytics, anomaly detection and explainable AI

These results will help organisations move from reactive to proactive, data-driven water management.
 

Approach

The partners develop and validate the solutions through four complementary workstreams.

1. Data understanding and requirements

• Definition of use cases and key performance indicators
• Collection and preparation of water consumption, contextual and metadata sources

2. Water disaggregation and modelling

• Development of AI and machine learning methods to decompose water consumption into specific usage categories
• Analysis of consumption patterns using advanced time-series techniques

3. Context-aware analysis

• Identification and modelling of contextual factors such as occupancy, weather and operational conditions
• Development of adaptive and explainable AI approaches

4. Real-time detection and decision support

• Integration of models into a real-time analytics framework
• Development of anomaly detection, alerting and recommendation mechanisms
• Validation in pilot environments and real-world use cases

Funding

• Funding agency: Innoviris
• Project type: Joint R&D Project Call 2024
• Contract number: 2025-JRDIC-2b
• Total budget: €432,805.48
• Funding level: 100%

Internal link

MIRAI – Machine Intelligence Techniques for Smart and Sustainable Planning and Operation

MIRAI investigated lightweight AI approaches for IoT and edge computing applications, including water consumption monitoring. WATER-SENSE builds further on this experience by focusing on context-aware water consumption disaggregation, anomaly detection, behavioural change analysis and the generation of actionable recommendations for water efficiency optimisation.