Enterprise Data Analytics Platform and AMI Operations

In tandem with its 6 year-long smart meter rollout plan, Con Edison sought to implement Advanced Metering Infrastructure (AMI) operations on top of a comprehensive enterprise data analytics platform for improved operational insight and customer service for its base of more than four million customers. In order to improve customer service and operations across its region, one of the largest integrated utilities in the United States has rolled out the C3 AI Suite and C3 AMI Operations application on AWS. Con Edison’s project objectives were to deliver on the utility’s commitments for presenting customer data, establish AMI operations across 5 million smart meters to ensure operational health, and build a federated data image platform for analytic capabilities. 

The utility’s smart meter deployment will generate between 100 terabytes and 1 petabyte of data per year, so choosing a platform that could scale and continue to perform analytics on an ever-larger data set was vital.

To put the foundational enterprise data analytics platform in place, C3.ai and the utility worked together to aggregate two years of data from 13 source systems covering 5 million customer accounts. The joint team then designed requirements for third-part data use and configuration of the basic AMI Operations application. Once the integrated data image was in place, the team configured two machine learning algorithms and 50 analytics to identify deployment and installation issues and determine meter and network health.

The utility can now monitor smart meter deployment to identify any installation or configuration issues. The application also provides real-time status at any level of aggregation—from the individual meter to the overall system—and a prioritized list of meters that require attention. In future phases, the company plans to build on its enterprise data analytics plan for additional customer insight applications and distribution and transmission automation capabilities.

Project Highlights:

- 10 Month Project

- 13 Source Systems

- 5M Customer Accounts

- 180B+ Annual Rows of Data Integrated

• [Efficiency Improvement - Asset Monitoring]
  

  2 machine learning algorithms and 50 analytics to identify deployment and installation issues and determine meter and network health.

• [Data Management - Real Time Data Analysis]
  

  Real-time status at any level of aggregation—from the individual meter to the overall system—and a prioritized list of meters that require attention. 

• $854M+ Identified annual customer benefit

• 2300+ Deployment issues identified in 4 months

• 5 External systems accessing unified data