Sensors and AI Help Water Companies Detect Sewer Fatbergs

Fatbergs, accumulations of fat, oil, grease, wet wipes, sanitary products and other debris that harden in sewer systems, continue to form in cities around the world. These blockages can cause sewage backups into homes and streets as well as pollution in rivers when sewers overflow.

Water companies are deploying sensors and artificial intelligence to identify developing fatbergs before they reach large sizes. The mass required nine weeks of manual removal using pickaxes and high-pressure water jets. By the end of 2025, a similar fatberg had regrown in the same Victorian-era sewer to more than 100 tonnes.

Richard Martin, head of treatment improvement at Southern Water, described the material as "a magical mixture" that hardens like concrete. Southern Water, which manages sewers in southeast England, tackles an estimated 300,000 accumulations of fat, oil and grease each year across its network.

Southern Water has installed around 34,000 sensors attached to manhole covers. The sensors use radar to measure water levels in the sewers below and transmit the data to a machine learning system. That system combines the readings with rainfall and weather information to establish expected water levels for any given day.

"If it pops outside of that band, that's when we react," Martin said. The approach allows crews to address blockages before they solidify and grow. Historically, companies relied on customer reports or scheduled inspections. The technology has also reduced the need for workers to enter sewers, where they face exposure to hydrogen sulphide, methane, carbon dioxide, viruses, bacteria and parasites.

Southern Water cleared 700 blockages using the AI system so far in 2025. In the prior full year the company handled between 3,500 and 4,000 blockages in total.