Anglian Water: visual evidence to verify spills and fix flow issues at source

From Q3 2025, Anglian Water’s Water Recycling Flow Technology Team deployed iDefigo AI visual sensors at remote Water Recycling assets to test one question: can rapidly deployed visual intelligence give teams reliable evidence where people cannot be present around the clock? Its July 2026 report recommends adoption.

“The proof of concept has demonstrated that visual evidence can materially improve confidence in spill verification, asset condition assessment and site intervention decisions.” Anglian Water, Water Recycling Flow Technology Team

• 8 months: trial analysis period ending July 2026

• 11: sites monitored for dry day spills (DDS)

• 20: spill events correctly attributed, across 9 spills at 4 sites

• 5: site interventions driven by visual sensor evidence

• 1–2 hours: typical install, including configuration

The challenge

Many existing spill and event duration monitors still rely on probes and other contact devices rather than the non-contact instruments future specifications require. Dry day spills and flow to full treatment (FFT) issues can occur at any hour, at sites with no permanent staff, power or Wi-Fi. QA teams needed accessible, high-quality evidence to verify what instruments were reporting.

The solution

iDefigo’s iDCam Pro battery and solar visual sensors, with up to five sensors per cellular bridge, were installed at the lowest spill point for DDS and at the FFT measuring point for flow investigations. Routine images were combined with AI-triggered bursts, and imagery was reviewed against EDM activation timestamps on iDefigo’s secure cloud platform.

“In the trial, solar-powered deployments performed reliably throughout the year, including darker winter periods.”

The results

• 20 spill events across 9 spills at 4 sites correctly attributed using visual evidence

• 6 site investigations evidenced, 5 leading to site interventions

• Root causes of false EDM activations identified, including operators, rodents and wildlife

• Evidence typically captured within 3 weeks to 3 months

• Formal recommendation to adopt

What the visual sensors found, and what changed

• Evidence captured: Balancing arm rising too high, stopping the site achieving FFT

• Action taken: Balancing arm adjusted and modified

• Evidence captured: Humus returns are causing dry day spills

• Action taken: Stricter manual returns process, timer and automation/inhibit proposed

• Evidence captured: River encroachment triggering the EDM

• Action taken: Attributed to river encroachment; EDM probes recommended to move from behind the bar screen

• Evidence captured: Level sensor poorly placed over storm weir

• Action taken: Sensor head repositioned to ensure more accurate spill capture

• Evidence captured: Probe and chamber conditions affecting FFT

• Action taken: Probes moved, chamber maintenance schedule changed; FFT improved

“Adoption of the iDefigo AI visual sensor solution is recommended because it provides a practical way of giving teams reliable ‘eyes on’ at remote, difficult-to-monitor assets where personnel cannot be present continuously.” Anglian Water, AI Camera DDS/FFT Evidence Capture: Summary of findings and recommendations

Source: Anglian Water, Water Recycling Flow Technology Team, AI Camera DDS/FFT Evidence Capture: Summary of findings and recommendations, author Brett Harwood, July 2026. Figures and quotations are taken from that report.