Yorkshire Water scales AI monitoring across 4,000 assets

TL;DR:

• Yorkshire Water is adding 1,200 wastewater pumping assets and 300 clean water assets to its AI-driven condition monitoring programme with Samotics over the next 18 months.
• Since the partnership began in 2022, the SAM4 system has detected 726 developing faults, prevented 40 pollution events and delivered £12.87 million in confirmed detection value — a 571% return on investment.
• The expansion will take the estate past 4,000 monitored assets, the largest single-customer deployment of the technology to date.

Yorkshire Water is scaling one of the more thoroughly measured AI deployments in UK infrastructure. The utility will extend its predictive condition monitoring programme with industrial AI firm Samotics to a further 1,500 pumping assets over 18 months, taking its monitored estate beyond 4,000 assets as part of the £8.3 billion AMP8 investment programme agreed with the regulator for 2025-30.

The technology, SAM4, uses electrical signature analysis: it reads current and voltage signals remotely from motor control cabinets, with AI classifying deviations into specific fault types — blockages, airlocks, misaligned components — often weeks before they would surface. That approach suits submerged sewage pumps and screw pumps, critical assets where conventional vibration sensors are impractical. The platform is independently validated at 97% accuracy with under 1% false positives.

What distinguishes this from most enterprise AI announcements is the outcome ledger. Since a first 50-site pilot grew into a £10 million contract in 2022, the programme has logged 726 developing faults detected, 40 pollution events prevented and £12.87 million in confirmed detection value — a 571% return on investment across monitored assets. The decision to expand rests on three years of tracked operational results, not projections.

A benchmark for the sector

UK water companies are under sustained pressure over pollution incidents and ageing assets, and AMP8 obliges them to show measurable improvement. Yorkshire Water’s numbers give the sector a concrete benchmark for what asset-level AI can return — and Samotics expects other utilities to scale similar approaches during the period. For industrial firms outside water, the lesson is the evaluation discipline as much as the technology.

Looking forward

Yorkshire Water plans to integrate SAM4 insights into its strategic data and telemetry platforms for long-term deterioration analysis, informing future investment decisions. The 18-month rollout will test whether the return holds at nearly double the asset count.