The Ghost in the Pipe: AI Finds Hidden Water Leaks

A new multi-sensor platform uses electrical currents and AI to find hidden water main leaks that traditional acoustic tools simply cannot hear

17 Apr 2026

US water utilities are adopting multi-sensor platforms to identify leaks in pressurized mains that have long evaded traditional acoustic detection. The technology, which integrates electrical resistance testing (ERT) and artificial intelligence, allows for precise defect mapping within live networks regardless of pipe material or ambient noise.

The system works by emitting a low-voltage electrical current through the water column. In a sound pipe, the wall acts as an insulator. When a breach occurs, the current escapes through the leak path, allowing the platform to pinpoint the location and estimate the volume of water loss. Unlike acoustic methods, which are often muffled by plastic piping, this electrical approach remains effective across all distribution materials.

Institutional adoption is driven by the need for more granular data. Once ERT identifies a defect, onboard cameras use AI to analyze the movement of suspended particles. By tracing these particles toward the pipe wall, the system provides visual confirmation of the leak. This creates an auditable dataset that aligns with American Water Works Association standards, helping regulators and companies transition toward condition-based asset management.

A recent application by the City of Cleveland demonstrated the limitations of older methods. During an inspection of a 30-inch transmission main located beneath a railroad corridor, traditional CCTV failed to observe any signs of leakage. However, ERT identified significant water loss through the pipe wall. At a separate site, the platform’s data correlated exactly with a known surface leak under adjacent tracks where surface-level surveys were impractical.

The shift toward in-pipe sensing comes as utilities face increasing pressure to reduce non-revenue water and optimize infrastructure spending. By quantifying leaks that were previously invisible, providers can prioritize repairs based on actual severity rather than the age of the pipe. As water scarcity and regulatory oversight intensify, these high-resolution tools are becoming a standard requirement for maintaining national distribution networks.

While the initial cost of multi-sensor inspections exceeds that of basic acoustic surveys, the long-term utility lies in capital efficiency. Reducing the frequency of unnecessary excavations allows municipal budgets to be allocated toward the most critical structural failures.