---
title: "LeakBySat, satellite water leak detection"
description: "Water leak detection: LeakBySat analyses satellite images with AI to map leaks and fraud on water networks."
canonical: https://steps.tn/en/satellite-water-leak-detection/
last-updated: 2026-11-16
---

# LeakBySat, satellite water leak detection

Water leak detection: LeakBySat analyses satellite images with AI to map leaks and fraud on water networks.

URL: https://steps.tn/en/satellite-water-leak-detection/

Reference | Water, satellite and AI

LeakBySat, detecting leaks from space

LeakBySat is data analysis software that studies water networks, detects leaks from satellite imagery and predicts fraud using artificial intelligence. Risk areas are mapped to guide field teams.

LeakBySat × STEPS Reference | Water, satellite and AI 

## LeakBySat, detecting leaks from space

 LeakBySat is data analysis software that studies water networks, detects leaks from satellite imagery and predicts fraud using artificial intelligence. Risk areas are mapped to guide field teams.
 Satellite imagery AI detection GIS mapping 
 Book my scoping session ↗
 
 AREAS TO INSPECT: 3 01 - The context

## What you see, and what gets lost

Much of the water lost escapes underground, out of sight. Networks are extensive and teams cannot inspect everything.
ABOVE GROUNDExtensive networksKilometres of pipes to monitor, across urban and rural areas.
Costly inspectionsSending teams out at random takes time and resources.
UNDERGROUNDInvisible leaksMany leaks never reach the surface and stay active for a long time.
Fraud that is hard to spotUnauthorised consumption blends in with network losses.
 02 - How it works

## From orbit to the field

SPACECOMPUTEGISSOLORBITSatellite imagerySatellite images cover the network over large areas.
ANALYSISAI detectionModels spot abnormal moisture signatures along the pipes.
MAPRisk areasSuspicious areas are matched with the network layout and mapped.
FIELDTargeted inspectionTeams go straight to the flagged areas.
 03 - The interface

## Leaks, visible on the map

Network files (shapefiles) and satellite images are overlaid to show leak areas.
1Analysis area2Leak signal3Base map 04 - The functions

## Four functions, one map

Image analysisSatellite images are processed to measure ground anomalies.
Leak detectionAI flags the areas where a leak is likely.
Fraud predictionSuspicious consumption is spotted on the network.
MappingAreas to inspect are published on a map for the teams.
 05 - The benefits

## What LeakBySat changes for water managers

Search less, find moreInstead of covering the whole network, teams receive a map of priority areas and focus their resources where water is being lost.
Targeted inspectionsTeams go where a leak is likely.
Fewer lossesLeaks are found and repaired sooner.
Wide coverageSatellites cover hard-to-reach areas.
Less fraudUnauthorised consumption is detected.
Documented decisionsEvery intervention is based on a map.
A preserved resourceEvery leak repaired is water saved.
 06 - Our work

## From satellite image to inspection

PHASE 01DataCollection of satellite images and network files.
PHASE 02ModelsTraining of leak and fraud detection models.
PHASE 03SoftwareVisualisation, layer overlay and area mapping.
PHASE 04FieldReporting to teams and inspection follow-up.
 07 - Architecture

## Under the hood, stacked layers

Satellite, AI and GIS: each layer adds information, and overlaying them reveals the leaks.
BASE MAP
NETWORK (SHAPEFILES)
SATELLITE IMAGERY
AI LAYER: RISK AREAS
 FAQ

## Frequently asked questions

Do sensors need to be installed on the network?No. The analysis relies on satellite images and network files.
Do the results replace field teams?No. They guide them to priority areas, where inspection confirms the leak.
Can the solution detect fraud?Yes. AI models spot suspicious consumption on the network.
 Is your network losing water and you don't know where? Let's talk about your project.
Book a scoping session ↗

---

Book a free scoping session: https://steps.tn/reserver-une-session-de-cadrage/

Site guide for agents: https://steps.tn/llms.txt
