Forecast, Mobilize, and Restore: Inside Exelon's AI-Driven Storm Outage Suite

Mar 04, 2027
B214
Reliability and Resilience
Exelon is building an integrated AI suite that spans the full storm lifecycle—from outage prediction days before a storm, through crew mobilization planning, to real-time estimated time of restoration (ETR) during active events. In this panel, three product owners walk through each component: the Outage Prediction Model (OPM), which forecasts outage volumes and locations using weather, grid, and vegetation data; the Response Planning Tool (RPT), which converts those forecasts into actionable crew deployment plans; and POSEIDON, which generates real-time ETRs based on live damage, crew, and weather conditions. Panelists will discuss how each model was developed independently, the challenges of aligning them, and the ongoing work to converge them into a unified Outage AI Suite with a seamless user experience. 
Speakers
Michael Warren
Michael Warren, Principal Quantitative Engineer - Exelon
John Reid
John Reid, Principal Data Scientist - Exelon
Vladyslav Anderson
Vladyslav Anderson, Manager of Data Science, Infrastructure and Safety Analytics - Exelon
Valli Muthukaruppan
Valli Muthukaruppan, Principal Quantitative Engineer - Exelon
Chairperson
Ashley Stern
Ashley Stern, Director Project Management - Itron