Programme

This presentation will showcase the fusion of heterogeneous data streams to produce accurate, high-resolution spatial rainfall fields and forecasts. The approach is based on Graph Neural Network (GNN) data-fusion framework capable of integrating multiple modalities: rain-gauge point data, radar grids, coarse satellite grids, CML line measurements, and CCTV image-based estimates. The approach involves both Convolutional GNNs (CGNNs) to model spatial correlations across the sensing network, as well as Heterogeneous GNNs (HGNNs) to handle varying data sources. The resulting framework is scalable, data-driven rainfall estimation and supports critical decision-making for climate resilience and urban infrastructure management.

This presentation will explore the digital transformation of urban sanitation in developing regions, focusing on how data-driven systems address gaps in traditional infrastructure. It will identify significant service challenges caused by rapid urbanisation and a historical reliance on manual tracking and external reporting rather than local operational needs. To illustrate potential solutions, the presentation will compare three major platforms—IMIS, DIGIT, and EDAMS—which vary from open-source municipal tools to commercial enterprise software. These systems enhance waste management transparency, improve revenue collection, and allow for real-time monitoring of faecal sludge across the entire service chain.

In the past decade, Fiber Optic Sensing has emerged as a promising and potentially game-changing technology for pipeline monitoring, providing real-time insight on leaks, intrusions and various other events impacting pipeline condition. DALI Pipeline Monitoring has been moving this promise beyond a theoretical concept and is turning this into reality. We will present several pilots and larger-scale implementations across the globe, to provide an update on the current state of technology, the technical and economic feasibility and capabilities of DALI Pipeline Monitoring.

Successful decisions are grounded in reliable data. This is an accepted fact; however, when it comes to the challenging conditions of sewerage and drainage networks, gaining that data is a significant hurdle. Cities require multi-useable data to perform network operations, plan future extensions, and generate numerical models to forecast several scenarios. To be effective, measurement systems must be resistant to harsh physical and chemical environments while providing high-accuracy readings for flow rates and water levels. Furthermore, because temporary measurements are often sufficient to reflect varying hydraulic conditions, these systems must be adaptable and easy to deploy. The City of Munich, serving 8.2 million inhabitants, faced these exact challenges. A project implemented by NIVUS GmbH overcame these obstacles by delivering a reliable data foundation for the city’s extension plans and heavy-rain management. Across more than 240 measurement points, the project utilised long-lasting, battery-operated ultrasonic systems. These digital, calibration-free Cross Corelation sensors which implement simultaneous velocity measurement at multiple depths, ensuring accuracy where conventional sensors fail. Reliable flow measurement in environments characterised by high turbulence, debris, or heavy sediment presents a barrier to data integrity. NIVUS’ advanced technologies—including robust open-channel sensors and adaptive signal processing—ensure continuous data flow even under these extreme conditions. Through automated data validation and remote diagnostics, this technical approach provides authorities with the comprehensive hydraulic data necessary to ground their critical decision-making tools. In this presentation, we will review the project’s technical characteristics and the solution implemented to showcase an example of a challenging measurement campaign.

Detecting and localising leaks remains a primary challenge for water utilities. Pipe breaks or hidden leaks can lead to significant water losses, property damage, water supply service disruption, and environmental hazards. Traditional methods of detection often involve reactive measures, leading to delays in response and increased costs. The Bentley’s solution for reducing real losses provides near-real-time cloud-based anomaly detection and localisation for water utilities to leverage AI predictive analytics, IoT Field sensor data, and calibrated hydraulic models. This session will also cover other innovative technologies offered by Bentley that leverage AI-driven tools for the detection of cracks in water storage facilities like Tanks and Dams, and AI-driven tools that support designers to automate the annotation of drawing plans, a process that commonly consumes significant time on design projects. This session will share outcomes and lessons learned from these technology implementations, highlighting the role of AI in improving decision-making and efficiency across the lifecycle in water infrastructure projects.

The study presents Distributed Acoustic Sensing (DAS) Technology as an effective real-time monitoring solution for urban water supply networks, using fibre optic cables to detect acoustic vibrations along pipelines and identify anomalies with high spatial precision. By sending laser pulses through the fibre and analysing backscattered signals—enhanced with AI and machine learning—the system can rapidly detect, classify, and geolocate events such as leaks, tampering, and external disturbances. A 5.6 km pilot implementation by Canal de Isabel II in Madrid demonstrated the technology’s ability to recognise simulated incidents including leakages, excavations, equipment manipulation, vehicle movement, and even train passage. The results confirm DAS as a powerful tool for improving hydraulic anomaly detection, strengthening infrastructure security, and advancing the digital transformation of water supply management.

ACCIONA has been selected by the Water Supplies Department (WSD) – the entity that manages drinking water supplies in Hong Kong – to carry out the design and commissioning of a digital twin in the Tseung Kwan O desalination plant. This award represents a milestone in the desalination industry and makes ACCIONA a pioneer in the implementation of this type of system in the water sector, as it is the first digital twin that covers all the stages of the desalination process. It represents progress on several fronts, as it forecasts the evolution of production processes, optimises costs by anticipating what happens in processes, detects membranes fouling in real time and optimises energy consumption, among other advantages. These factors are key elements in improving efficiency in the operation and maintenance of the plant. It is in this context that the WITNESS solution arises, with the aim of developing new indicators capable of diagnosing water preventively, in order to reduce membrane fouling and optimise the overall operation of the desalination process. By leveraging advanced machine learning techniques, WITNESS seeks to use process monitoring data (SCADA) to provide accurate predictions and estimates of critical variables. In this case, the main variable to predict is the normalised differential pressure ΔP of the reverse osmosis racks, a key indicator of membrane fouling. This approach will not only enable a better understanding of process behaviour but will also open new possibilities to improve the efficiency and sustainability of desalination plants. The solution is fully operational in the Tseung Kwan O desalination plant.

Reliable online sensor data is the foundation for achieving continuous compliance with strict effluent standards and enabling the digital transformation of wastewater treatment plants. However, sensors operating in harsh, high-biomass environments are inherently vulnerable to drift and data quality issues, which often undermines their use for real-time process control.  This study details a systematic Quality Assurance (QA) framework piloted at the Paranaque WRF that utilizes the Shewhart Control Chart methodology to statistically validate sensors for Nitrates, Ammonia, and pH. By objectively measuring the variance between real-time readings and laboratory reference methods, the system shifts maintenance from reactive troubleshooting to a proactive, evidence-based validation protocol. This approach fundamentally promotes operator trust in instrumentation, ensuring the high-quality data necessary to optimize chemical dosing, reduce energy consumption, and guarantee regulatory assurance across the facility.

From Silos to Streams: How De Watergroep Is Future-Proofing Water Management – In order to become a data-driven company De Watergroep started in 2023 to transform fragmented, regional data silos into a unified, scalable ecosystem powered by its Delphi platform. This strategic shift bridges the gap between IT and Operational Technology, integrating near real-time data from reservoirs, monitoring wells, and third-party providers to enable faster, more responsive decision-making. Beyond core infrastructure, the initiative democratises data through a governance model that empowers employees with self-service capabilities and continuous learning. By acting as a central “water carrier,” Delphi fosters innovative collaboration while streamlining operations through deep integration with ERP and a variety of specialised niche applications used for water production and monitoring.

In this presentation, we will showcase Rheticus Network Alert, a satellite-based monitoring solution for the predictive maintenance of wastewater networks. By analysing ground motion across the network, it pinpoints sections that require the highest priority for inspection, enabling utilities to optimise maintenance activities proactively. First piloted in Italy in 2017, Rheticus Network Alert is now applied across Europe, Latin America, and Asia, delivering measurable improvements in operational efficiency. The technology has reached Technology Readiness Level (TRL) 9, demonstrating full deployment and proven performance in real-world conditions.

The solution is a new production planning tool which has been created from scratch between Anglian water services and Optimatics/Suez UK. It has been trialled in one of our supply systems in Colchester, UK which has some initial very positive results and findings and a case study which can be shared. It is being used to transform Anglian water’s whole operating model, way of planning maintenance, balance abstraction licenses and integrating our biggest capital delivery scheme known as the strategic pipeline, which connects our whole region as one supply system (one of the biggest construction’s ever done in the UK). The presentation would include both an insight into how this project has been rolled out across Anglian water, a live demo of the tool and its capabilities, followed by a specific case study and results of the Colchester region.

NPM AI is a unified software platform that digitises, monitors, and optimises the entire water cycle from source to consumer. By integrating SCADA, IoT, GIS, and hydraulic models into a single operational hub, it gives utilities real-time visibility across the network, enabling predictive and informed management at every level. With built-in NRW and DMA management modules aligned to IWA standards, the platform provides trusted KPIs and regulatory-ready calculations, helping utilities benchmark performance, reduce losses, and operate at optimal pressure. AI-powered anomaly detection, resource forecasting, and dynamic simulations transform daily operations from reactive responses to proactive and data-driven decision-making, supporting smarter distribution strategies and more efficient use of energy and assets. At its core, NPM AI delivers Operational Intelligence & Insights through customisable dashboards and digital-twin maps, giving stakeholders a clear view of operational health, investment priorities, and emerging risks. Its modular design ensures scalability and continuity, whether deployed for a district, a city, or a national network. Proven in large-scale deployments, from Istanbul to Dubai, NPM AI has delivered measurable results: less water lost, fewer failures and a smarter use of the water network. In short, NPM AI transforms complex water systems into predictable, optimised and more sustainable networks supporting utilities on their journey toward data-driven, low-loss and future-ready operations.