FORE

FORE Smart Monitoring is the data acquisition and visualisation layer of the FORE platform. It integrates heterogeneous sensor networks — flow meters, spectrometric probes, data loggers — into a unified interface that gives water operators a real-time view of their infrastructure. The system handles the full data pipeline: collection, transmission, cleansing and contextualisation. Raw signals are automatically filtered and validated, removing noise and drift before they reach the analyst. What operators see is not raw data — it is actionable information, structured around the physical reality of the network. Deployed on the Mar Piccolo monitoring network in Taranto, FORE Smart Monitoring operates in complex environmental conditions, managing multi-site, multi-parameter campaigns across BOD, COD, TOC and DOC indicators. The platform is sensor-agnostic and infrastructure-independent: it connects to existing hardware, adapts to incomplete network documentation, and scales from a single monitoring point to a full territorial campaign. It is the foundation on which FORE's AI and modelling capabilities operate.

The FORE Water Intelligence Engine is the analytical and predictive core of the FORE platform. It combines Physics-Informed Graph Neural Networks (PIGNN) with hydraulic simulation to build accurate digital models of water infrastructure — even when network maps are incomplete, sensor coverage is sparse, and historical data is limited. Conventional AI tools require clean, abundant data to function. FORE's approach is different: by embedding physical laws directly into the learning architecture, our models understand how water networks behave from first principles. This means they can calibrate, predict and optimise where standard machine learning cannot operate. The Engine delivers hydraulic model calibration up to 10x faster than traditional methods, identifies inefficiencies across distribution, sewerage and treatment assets, and supports operational decisions on energy consumption, pressure management and process optimisation. Built on water networks — the most complex and data-poor infrastructure class that exists — the same methodology applies wherever physical networks meet incomplete data: gas distribution, district heating, industrial utilities. Water is where we started. Intelligent infrastructure is where we are going.