Join the Customer Analytics Community on Tuesday, January 9, 2024 at 11:00 AM CT for a presentation with Q&A and open discussion on, "Gas Meter Anomaly Detection Using Monthly Usage Data at National Grid" led by Hemanya Tyagi, Senior Data Scientist at National Grid.
Revenue loss is a big challenge faced by the utility industry. Meter issues like stopped meters, ERT (Encoder Receiver Transmitter) issues, and theft of service are major contributors to revenue loss. Each of these types of meter issue creates a pattern in customer usage data, and identifying these patterns is key to proactively identifying problematic meters for investigation . It is well known that machine learning can be used on smart meter or AMI data to identify usage anomalies. However, identifying anomalies using monthly data is a less well studied problem.
At National Grid, we developed the RADAR (Revenue Anomaly Detection and Response) tool, which applies machine learning to monthly billing data to identify usage anomalies. This session focuses on demonstrating how we can find meter anomalies using less frequently (weekly or monthly) captured usage data. We will discuss how we can apply machine learning and statistical methods to capture trends, changes in patterns, and outliers in the customer usage data, as well as the implementation challenges associated with deploying such a model.
• Meter issues can lead to revenue loss for utility companies.
• Statistical and machine learning algorithms capture changes in usage patterns, thus assisting us in identifying meter issues or theft of service, even with monthly billing data
Members of the Custimer Analytics Community will receive a Microsoft Teams link prior to this meeting. Not a member of this community? UAI Utility Members can register for any of our communities by using our Request To Join link.