Frequency and duration of outages are increasing in some parts of the US, causing utilities to struggle with effectively communicating with customers about restoration plans. To better predict the estimated time of restoration (ETR) and personalize the customer communication during outages Exelon has developed a data-driven predictive algorithm to generate customer specific ETRs that improves the speed of availability and accuracy of ETRs compared to existing methods. Accurate and effective communication of ETRs drastically reduces the number of customer calls and increases the overall satisfaction of customers.

Key Takeaways

• Viewing ETR from customer perspective rather than a utility perspective
• How historical and live outage data can be leveraged to predict accurate customer specific ETRs using advanced machine learning tools
• Impact of accurate ETRs on customer satisfaction
• Change management required at utilities to replace existing ETR tools with analytic tool