Water utilities adopt AI to cut leaks and reduce sewage outflows
Updated
Updated · Financial Times · May 4
Water utilities adopt AI to cut leaks and reduce sewage outflows
9 articles · Updated · Financial Times · May 4
In England and Wales, where about 20% of treated water is lost, Northumbrian, Southern and Yorkshire Water are deploying sensors, satellite data and predictive systems.
Northumbrian says AI using weather forecasts and 750 in-pipe sensors could prevent up to 80% of storm overflows, while Southern says 34,000 radar sensors helped avert 4,000 sewage flooding incidents last year.
Singapore, Japan and China are presented as leaders, with Tokyo and Osaka losing about 3.8% of treated water and Singapore about 5%, far below British leakage rates.
As AI saves water from leaky pipes, are its thirsty data centers creating a bigger drought?
Is AI a true fix for aging water systems, or just a digital patch on a crumbling foundation?
Transforming Water Utilities: How AI Reduced Storm Spills by 17% and NRW by 2% in Two Years
Overview
Between 2024 and 2026, AI transformed water management with major successes. South West Water in the UK deployed 12,000 smart sensors in 2023, preventing over 4,800 storm spills in 2024 and cutting spill frequency and duration further in 2025 despite heavy rainfall. In Sweden, VA SYD introduced an AI leak detection system around 2024, reducing non-revenue water from 10% to under 8%. These advances rely on AI technologies that detect leaks precisely and predict blockages, enabling proactive maintenance. However, challenges remain, including workforce shortages causing loss of expertise and cybersecurity risks due to lack of mandatory protections. Globally, the AI water market reached $1.41 billion in 2024, signaling strong growth and adoption.