Overview
Con Edison is one of the nation's largest investor-owned energy companies, with approximately $12 billion in annual revenues and $48 billion in assets. Founded in 1823 as the New York Gas Light company, Con Edison’s electric, gas, and steam service now provides energy for the 10 million people who live in New York City and Westchester County.
With the three priorities of safety, operational excellence, and an outstanding customer experience as its beacon, Con Edison is confidently moving toward a more sustainable future for today's world and for geenerations to come.
Supernova Award Category
Data to Decisions
The Problem
In 2017, Con Edison initiated a rollout of 5.3 million Advanced Metering Infrastructure (AMI) meters across greater New York City. The new “smart” meters would generate massive new data volumes including reads every 5 to 15 minutes, ultimately intended to improve operational efficiency with actionable insights. This rollout would deliver improved energy visibility and savings opportunities, ensure public safety, and provide increased visibility for regulators and other stakeholders on the operation of the distribution grid.
Through this AMI deployment, Con Edison is making the grid more resilient and flexible, and protecting neighborhoods by investing in core grid infrastructure. Con Edison is advancing New York’s clean energy goals and slashing carbon emissions nationwide by committing to renewables and ground-breaking clean-energy technologies. And with every action, Con Edison is forging new ways to give customers more choices, more control, and more convenience.
The Solution
Con Edison decided to build an Enterprise Data and Analytics Platform to integrate, process, and correlate data from smart meters and other enterprise systems, analyze the resulting data image using advanced AI algorithms, and operationalize the results in applications that scale as data volumes grow.
Con Edison began by establishing with C3 AI a unified, federated image of all relevant data to serve as a foundation for development. Together, they built data integration pipelines to ingest, normalize, and correlate at least two years of historical data from 27 source systems covering 3.4 million customer accounts. As new meters are added, data is automatically added to the unified data image via these pipelines.
The team then delivered the analytical requirements for Con Edison's initial AMI operations and management application. This involved configuring dozens of analytics to identify deployment and installation issues and to constantly monitor meter and network health.
The results
Since the initial application configuration, this solution has scaled with Con Edison’s smart meter program and now integrates 3.4 million smart meters generating hundreds of billions of data records annually. The utility can now track smart meter deployment progress and direct field maintenance teams to inspect high-priority meters. The team configured APIs to publish results to third-party applications such as work order, asset management, and customer presentment systems to facilitate customer communications and maintenance.
Con Edison and C3 AI have since worked to address over 30 use cases by rapidly integrating new data sources, configuring new machine learning models and analytics, and extending user interfaces, all of which are facilitated by the extensible, modular, model-driven architecture of the C3 AI Suite.
Over the following three years, Con Edison has turned to the C3 AI Suite to solve increasingly complex and important problems. Today, the utility uses C3 AI technology to address meter integrity and maintenance, customer fraud detection and prevention, load forecasting, and 20 other high-value business use cases. Over 300 Con Edison data scientists, developers, and business analysts use the combination of code and no-code tools provided by the C3 AI Suite to conduct AI-based application development, prototyping, and ad-hoc analysis.
Metrics
Con Edison uses data-driven insights from their Enterprise Data and Analytics Platform to reduce operating costs by optimizing work and increasing productivity. EDAP integrates 15-minute interval data from AMI with data from 27 enterprise source systems, creating 800 billion new rows of data per year across a total of 3.4 million electric meters installed at Con Edison and Orange and Rockland Utilities to create a 415 TB enterprise data and analytics platform, growing at 200 TBs per year. Each day, EDAP ingests over 2.2 billion records and processes over 100,000 API calls to provide actionable data, analytics and insights throughout the organization.
Five Con Edison business units (AMI Operations Control Center, Rate Engineering, Energy Management, Customer Operations, and Electric Operations) leverage applications built on top of EDAP. For example, the Conservation Voltage Optimization (CVO) application allows Con Edison to reduce energy generation, operating costs, and CO2 emissions by avoiding excess voltage when delivering electricity to customers. The application uses 5-minute interval AMI voltage measurements and distribution grid hierarchy data elements to determine the voltage optimization opportunities at various distribution levels including substations and individual structures. Con Edison estimates a 3% reduction in voltage resulting in 1.5% in energy savings and $346 Million NPV cost savings over 20 years through improved voltage management.
The Technology
The C3 AI Suite, Con Edison’s enterprise data and analytics technology platform, integrates, processes, and correlates data from smart meters and other enterprise systems, analyzes the resulting data image using advanced AI algorithms, and operationalizes the results in applications that scale as data volumes grow. Business/data analysts and data scientists use C3 AI Ex Machina’s drag and drop interface to rapidly develop advanced analytics without writing a single line of code.
Disruptive Factor
Con Edison’s Enterprise Data and Analytics Platform has modernized operations and driven efficiency while also improving services for customers. By working with C3 AI, Con Edison has been able to accelerate speed-to-value across multiple business units on its comprehensive enterprise AI platform.
The most disruptive factor to date has been Con Edison’s use of the low-code/no-code C3 AI Ex Machina tool to solve business problems without requiring a deep understanding of the complex algorithms required to solve them. This initiative reduced the burden on the Con Edison IT application development and data scientist staff to efficiently prototype solutions before a full application is deployed in production at scale. Con Edison utilized C3 AI Ex Machina to then integrate those prototype applications directly into the existing platform, such as “Hot Socket” Monitoring, AMI Communication Reports, & Meter Health Reports. This ability to solve problems with minimal intervention from data scientists and application developers aids Con Edison in accelerating the design, development, and deployment of AI/ML applications.
Shining Moment
When Con Edison had an urgent need to meet conservation voltage optimization requirements, existing BI tools could not process, analyze, and present the large volume of data. Con Edison leveraged the C3 AI Suite to rapidly develop a new CVO capability within the C3 AI AMI Operations application. Now over 100 users access the operationalized CVO application, updated daily, to see voltage trends, heat maps for easier investigation, and savings realized from efforts within the application.