Back to Portfolio
Live in ProductionWater Utilities / Smart City
๐Ÿ’ง

Smart Water Management IoT System โ€” Leak Detection & Pipe Network Monitoring

A LoRaWAN-based water distribution monitoring system for a municipal utility โ€” detecting leaks within 15 minutes across a 250km pipe network, reducing non-revenue water from 35% to 12% in 18 months.

35%โ†’12%
NRWโ†“
15 min
Detection
400
Nodes

โš The Problem

A municipal water utility serving 500,000 residents had 35% non-revenue water (NRW) โ€” water pumped into the distribution network but never billed. This included physical losses (pipe leaks) and commercial losses. Leaks were identified only when residents called to report water on streets โ€” by then, significant water and revenue had been lost. The 250km pipe network had no instrumentation.

๐Ÿ’กOur Solution

We deployed 400 LoRaWAN monitoring nodes across the pipe network โ€” ultrasonic flow meters at district metering area (DMA) boundaries, pressure transducers at strategic points, and water quality sensors at storage tanks. All data routes through 8 LoRaWAN gateways to AWS IoT Core. A Python ML service applies minimum night flow analysis and pressure gradient analysis to identify leak candidates. When the model flags a probable leak, it calculates the most likely pipe segment from pressure data and pins it on the GIS map. Field crews are dispatched to that location โ€” average leak confirmation time dropped from 2 weeks (passive) to 15 minutes (active detection).

๐Ÿ”—System Architecture

LoRaWAN Flow/Pressure Nodes โ†’ Gateways โ†’ AWS IoT Core โ†’ InfluxDB โ†’ Python ML โ†’ React GIS Dashboard

Tech Stack

Hardware
  • Ultrasonic flow meter (DN50โ€“DN300)
  • Pressure transducer (0โ€“16 bar)
  • Water quality sensors (turbidity, pH, chlorine)
  • ESP32 + LoRa module
  • Solar-powered remote nodes
Communication
  • LoRaWAN (city-wide coverage)
  • MQTT to cloud
  • NB-IoT backup for dense urban zones
Cloud
  • AWS IoT Core
  • InfluxDB (flow + pressure time-series)
  • Python ML (leak detection)
  • PostgreSQL (pipe network GIS)
Frontend
  • React GIS dashboard (Leaflet.js)
  • Pressure zone heatmap
  • Leak probability map
  • Water quality trend charts

Key Features

Real-time flow balance monitoring per district metering area
Pressure transient analysis for pipe burst early warning
AI leak probability map with suggested pipe segment
Water quality (turbidity, chlorine, pH) at storage tanks
Night flow analysis automated report (minimum night flow)
GIS integration with pipe network topology
Demand forecasting for pump station scheduling
Mobile app for field crew leak verification

Results Delivered

  • Non-revenue water reduced from 35% to 12% in 18 months
  • 250km pipe network fully instrumented with 400 nodes
  • Leak detection time reduced from 2 weeks to 15 minutes
  • 85 leaks identified and repaired in first 12 months
  • Water savings equivalent to 2.5 million liters per day

Technologies

LoRaWANFlow MetersPressure SensorsReactGISAWSAI Leak Detection

Who This Is For

Municipal water utilities, industrial water treatment plants, large commercial buildings with water management needs

Need a similar solution?

We've built production IoT systems like this across 15+ countries. Let's talk about yours.

Let's Build Together

Got an IoT challenge?
We've shipped it.

Whether you need a fleet to track, a factory to monitor, or a farm to automate โ€” our team has done it before and we'd love to build it with you. Typical response time: under 24 hours.

No upfront commitment99.9% uptime SLANDA on requestFixed-price options