Leveraging Machine Learning for Phase Identification
January 21, 2020
Share this story on:
Inaccurate phase connectivity information may cause several operational inefficiencies – for example, unbalanced phases that lead to significant energy losses and sharply reduced asset lifetimes. Traditional approaches to phase identification require either manual intervention or costly signal injection. These methods are usually infrequently performed. As a consequence, the phase identification can quickly become out-of-date. Using robust machine learning techniques, Itron’s Strategic Analytics group has developed algorithms to accurately classify meters according to their phase using voltage information readily available from AMI meters.
Itron’s phase identification is offered as a service, minimizing upfront cost. In addition, pilot programs are available for a limited number of feeders to allow the opportunity to evaluate the service’s accuracy and benefit to your utility.
Paige Schaefer is a Product Marketing Manager in Itron’s Outcomes group for the strategy, planning and implementation of projects supporting marketing functions spanning electricity, water and gas business units. She interacts directly with sales, product and corporate marketing to identify new marketing opportunities, recommend actions and the coordination of targeted campaigns to increase brand awareness and market share where she works closely with the teams to develop content and strategies. In addition, she provides website support, event coordination and manages Itron’s Energy Forecasting Group (EFG), which supports end-use data development, the Statistical End-use Approach (SAE) and coordinates their annual meeting for discussing modeling and forecasting issues. Paige has a B.S. in Business Administration from San Diego State University with an emphasis in Marketing.
