Israel's state-owned Mekorot uses an AI anomaly-detection platform with IXDen to monitor thousands of water sites, process ~300 million sensor readings daily, and flag failures and cyber threats weeks in advance — though cost-savings estimates are vendor-reported, not audited.
300 million
Sensor readings processed daily (by Dec 2023)
70
Sites in the initial rollout (2022)
~3,000
Sites covered nationally
up to 60 days
Advance warning window for equipment failure or sensor drift
Details
Maturity
Established
Promoter
Mekorot Water Company Ltd. (with IXDen)
Period
2022-2025
Keywords
water utilities, critical infrastructure, predictive maintenance, cybersecurity
Context
Mekorot is Israel's national, state-owned water company, supplying the large majority of the country's drinking water through a network covering roughly 3,000 sites across 10 regions, including some 700 pumping stations and 20 desalination facilities. Since 2022 Mekorot has worked with the Israeli startup IXDen to deploy an AI-based anomaly-detection and predictive-maintenance platform.
Activities
By December 2023 the system had expanded from an initial rollout of about 70 sites to feed roughly 300 million sensor readings a day into a single monitoring dashboard, with plans to cover thousands of sites nationwide. In March 2025 Mekorot opened a National Water Control Center consolidating previously separate regional control rooms into one 24/7 hub using the same technology to detect faults and trigger backup systems automatically.
Results
IXDen has reported detecting equipment failures and sensor drift up to 60 days before they would otherwise surface, and potential cyber-related sensor tampering roughly a week in advance. The company has also estimated that the platform will help Mekorot save 'millions of dollars' a year.
Conclusions
These savings and detection figures come from the vendor rather than an independently published audit, and Mekorot has not released external verification of savings or detection accuracy. Detailed technical documentation of the AI models is not public, typical for critical national infrastructure but limiting outside scrutiny of alerting logic and false-positive handling.
Implementation
Indicative cost
High (€500k–€5M) — Multi-year vendor partnership (IXDen) plus capital investment in a new National Water Control Center opened March 2025; no public cost figures disclosed.
Time to results
Long (> 3 years) — Piloted at about 70 sites in 2022, scaled to around 300 million daily readings by December 2023, consolidated into a national 24/7 control center by March 2025.
Staffing & skills
IXDen (AI vendor), Mekorot National Water Control Center staff (24/7 hub, opened March 2025)
Conditions for success
Sensor instrumentation across pumping stations and desalination facilities
Central monitoring dashboard aggregating around 300 million daily readings
Vendor partnership for anomaly-detection model development
Common failure modes
Vendor-reported savings and accuracy figures are not independently audited
Underlying AI models and false-positive handling are not publicly documented
Where it fits
Governance type
state-owned national utility
Scale
national
Income level
high-income
Commonly funded by
National / regional programmes
Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.
Do you run this practice?
Claim it —
verified implementers get a public contact pathway and can propose corrections.
Data sources
Where this practice's information was retrieved from, and when.
In a 2023–24 operational pilot ordered under a presidential AI directive, CISA compared AI/LLM-based vulnerability-detection tools against its existing methods …