Catch leaks as they start, not on next quarter's loss report.

The challenge

Most utilities run with 15–30% non-revenue water (NRW). Bursts get spotted when a customer calls; slow leaks survive for months because the daily SCADA report averages them away. Energy spent pumping leaked water is real money — and so is the regulatory exposure when distribution losses breach thresholds.

The data needed to catch these events early already exists. It just isn't aggregated, baselined, or watched in a way that turns it into an alarm someone can act on.

What success looks like

How WaterTwin delivers it

• Real-time monitoring — flow rate and pressure for every metered point, sub-second updates. The night-time minimum flow that signals a slow leak is visible the moment it shifts.
• Multi-condition alarms — "trigger when flow rises > 20% above the same hour last week and pressure drops > 0.5 bar for more than 5 minutes". Fewer false positives than single-threshold rules.
• Historic baselines from the time-series store — the platform retains every reading at full resolution, so "above normal for this DMA at this time of week" is a real comparison, not a guess.
• Geospatial asset register — when an alarm fires, the location, the pipe ID, and the nearest isolation valves are one click away for the crew.
• Tickets & mobile capture — the alarm turns into a dispatched job; the crew records what they found on phone or tablet; the closing report attaches to the asset.

A typical detection workflow

1. A DMA's night-time minimum flow climbs above its seven-day baseline by more than the configured threshold.
2. WaterTwin correlates with pressure in the same zone; if pressure has also dipped, a higher-severity alarm fires.
3. On-call gets one notification — not one per affected meter. The notification deep-links to the DMA dashboard.
4. The duty engineer opens the digital twin, identifies the most likely segment from flow propagation, and dispatches a crew through the same screen.
5. The crew records the find (photo, valve action) on mobile. The asset's history reflects the event.

Where this lands first

• District metering areas (DMAs) — most utilities already meter at this granularity; turning that data into alarms is the lowest-effort win.
• Pump stations and pressure zones — pressure transients are leading indicators for hydraulic failures.
• Distribution mains — once basic detection is working, pattern analysis on historic data identifies repeat-failure segments for proactive replacement.