Case Study: Portland General Electric improves energy loss detection with Amazon SageMaker

John Nichols Describes PG&E’s Efforts to Use AWS to Innovate & Save Money

Portland General Electric (PGE) faced a challenge as its on-premises data center limited its ability to innovate, scale, and quickly query data, preventing it from meeting its goals for delivering clean energy solutions. To overcome this, PGE chose to migrate to a hybrid cloud model using Amazon Web Services (AWS) and its services to increase operational efficiency.

By implementing a solution on AWS, including Amazon DynamoDB and Amazon SageMaker, PGE achieved significant results. The utility now delivers subsecond latency for most APIs, serving five times more traffic than its on-premises system. AWS enabled a significant improvement in energy loss detection and an algorithm that detects decay in wooden poles with 80% accuracy, all while scaling to support its 900,000 customers.

PG&E