Your water networks and treatment sites already produce the signals needed for earlier warning. Sentinel-for-Water learns normal behaviour across pumps, mains, reservoirs, DMAs, and treatment processes, detects deviation early, flags likely events, and shows operators and engineers where to look before bursts, outages, leakage, or water quality issues escalate.
It sits above existing SCADA, PLCs, RTUs, and telemetry. Sentinel-for-Water does not replace control systems. It adds earlier warning, clearer evidence, state-aware anomaly scoring, and operational impact metrics.
Sentinel-for-Water also checks whether instrument channels remain trustworthy, helping teams distinguish process drift from unreliable sensing across flow, pressure, level, and online water quality instrumentation.
Over time, each deployment builds structured fault intelligence that can improve operations, maintenance, leakage control, water quality performance, and fleet benchmarking across zones and sites.
This prototype adapts a proven manufacturing monitoring approach into a water-sector innovation for the Ofwat Water Discovery Challenge. It is an early-stage demonstrator for engagement with water companies, not a live plant control system.
Water Assets Monitoring
Download water results (JSON) — live computed water payload for integrations or offline review.
Your water assets already produce the signals needed for earlier warning. Sentinel learns normal behaviour across treatment works, pumping stations, reservoirs, DMAs and network assets, detects deviation early, predicts likely events, and shows operators and engineers where to look before leakage, outages, bursts, or water quality incidents escalate.
It sits above existing SCADA, telemetry, historian, and controls environments. Sentinel does not replace PLCs, RTUs, SCADA/HMI, or ICS. It adds earlier warning, clearer evidence, state-aware comparison, and operational impact.
Sentinel also checks whether online instruments remain trustworthy, helping teams distinguish process drift from unreliable sensors.
Over time, each deployment builds structured fault intelligence that can improve operations, maintenance, leakage reduction, investment planning, procurement, and cross-region benchmarking.
Field devices → PLC / RTU → comms → SCADA / HMI / historian → Sentinel analytics layer
Sentinel is supervisory and read-only. It does not write to PLCs or RTUs.
Shows how much earlier Sentinel picked up a developing problem than a conventional alarm would have done.
Compares fixed alarm thresholds with Sentinel’s earlier warning based on changing behaviour across the asset.
Shows the live signals Sentinel is reading to understand what the asset or zone is doing now.
Summarises which parts of the system look stable, which are drifting, and which appear to need attention.
Sets out the strongest current evidence for what may be happening and where investigation should start.
Checks whether the instruments still look trustworthy, helping distinguish real process change from bad data.
Shows which alarms are useful, which create noise, and which alarm combinations may point to a larger issue.
Shows the operational effect of abnormal behaviour in energy, carbon, and water-loss terms. Figures shown here are demonstrator estimates based on simulated operating conditions and stated conversion factors.
Shows how behaviour changes across normal supply, peak demand, night flow, maintenance, and other operating conditions.
Compares the normal operating profile and stability of different zones, sites, and assets.
Shows where one site or zone is performing materially worse or better than comparable assets.
Translates earlier warning into practical operational value, such as avoided outages, leakage, and excess energy use.