South Africa · Durban · See the South Africa profile · See the Durban profile
Evidence: Observational / pre–post Top 62% 53/100 · Ask Evidence Copilot about this practice
eThekwini Municipality uses a satellite machine-learning system to flag suspected drinking-water leaks across Durban's network, tracking over 1,000 leaks. Despite the tool, non-revenue water rose to 65% by mid-2026 — a caution on detection without repair capacity.
/file/dailymaverick/wp-content/uploads/2023/08/Tony-durbansatellites.jpg)
Since around 2021, eThekwini Municipality (Durban) has used a satellite machine-learning system analysing multispectral imagery to flag suspected drinking-water leaks across its network.
The system has tracked 1,000+ suspected leaks, but municipal non-revenue water rose from 58.7% to 65% by June 2026 with about 12,000 reported leaks still outstanding — detection has not yet reduced losses without matching repair capacity.
National / regional programmes
Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.
Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.
Where this practice's information was retrieved from, and when.
Peru
Peru's Ministry of Environment has run Geobosques, a weekly satellite-based deforestation early-warning system, since April 2017; a PNAS randomized controlled …
Singapore
PUB's Anomaly Leak Finder uses AI, machine learning and a daily-recalibrated hydraulic digital twin to detect and localise hidden pipe …
Israel
Israel's state-owned Mekorot uses an AI anomaly-detection platform with IXDen to monitor thousands of water sites, process ~300 million sensor …
Croatia
Croatia's DHMZ launched a nationwide 3-day AI/ML air-quality forecast (PM2.5, PM10, NO2, O3) on 17 Jan 2025, adapting Belgium's EU-funded …
Open full copilot Grounded in cited practices — always check the sources.