How Con Edison Uses C3 AI to Monitor 5 Million Smart Meters and Identify $854M in Benefits
Con Edison, one of the largest integrated utilities in the United States serving 4.4 million customers in New York City, deployed the C3 AI Platform and C3 AI AMI Operations to manage its 5.3 million smart meter rollout. By consolidating data from 13 source systems and applying machine learning to 180 billion rows of annual meter data, the utility identified $854M in annual customer benefits and flagged over 2,300 deployment issues within the first four months.
Impact
$854M+
Annual customer benefit identified
2,300+
Deployment issues identified
5.3M
Smart meters managed
13
Source systems integrated
Challenge
Con Edison’s 5.3 million smart meter rollout generated up to 1 petabyte of data per year across 13 siloed source systems, with no unified platform to monitor deployment health, detect meter anomalies in real time, or scale operational analytics across 4.4 million customer accounts.
Solution
The C3 AI Platform and C3 AI AMI Operations were deployed to aggregate 13 source systems into a unified data image, apply machine learning algorithms and 50 analytics to monitor smart meter deployment health, and surface real-time operational status at any level of aggregation from individual meter to system-wide.
Tools & Technologies
What Leaders Say
“C3 AI is a small company that acts in a big way. I think we’ve got a great future.”
“The experience working with C3 has been eye-opening.”
Sign up to read complete case studies, access detailed metrics, and unlock all use cases.
Full Story
Con Edison serves more than 4.4 million customers across Manhattan and the New York City metro area, operating one of the most complex urban energy distribution networks in the country. As part of a multi-year smart meter rollout targeting 5.3 million Advanced Metering Infrastructure (AMI) devices, the utility needed a data platform capable of turning massive sensor telemetry into operational intelligence—at a scale where even small inefficiencies translate into significant costs.
Before deploying the C3 AI Platform, Con Edison’s metering data was fragmented across 13 separate source systems, covering 5 million customer accounts with no unified view. Identifying installation problems, network health issues, or meter anomalies required manual investigation across disconnected data stores—a process that was neither scalable nor responsive enough for a rollout of this magnitude. The team needed a platform that could ingest historical and real-time data, surface problems automatically, and support analytical queries without requiring deep data science expertise.
Con Edison and C3 AI worked together to aggregate two years of historical data across all 13 source systems into a unified data image. The C3 AI AMI Operations application then applied two machine learning algorithms and 50 custom analytics to monitor meter deployment health, flag installation and configuration issues, and track network performance in real time. The platform supports aggregation from individual meter level to system-wide views, and five external systems were connected to the unified data image for downstream analytical use.
The results were immediate and financially significant. Within four months of deployment, the application identified over 2,300 meter deployment issues. The platform calculates $854 million in annual customer benefit by optimizing meter performance, reducing service interruptions, and enabling faster detection of network anomalies across 5 million endpoints generating 180 billion rows of data per year.
Con Edison plans to build on this enterprise data analytics foundation for additional customer insight applications and distribution and transmission automation. The platform’s architecture is designed to expand to other C3 AI applications over time, positioning the utility to layer additional AI capabilities onto the same unified data infrastructure.
Amazon SageMaker