Artificial intelligence and machine learning are being used at Con Edison to withstand the challenges of weather events and their increasing severity across their service territory. Utilizing historical weather, outage, and restoration data, Con Edison is using a tool that identifies similar past storms to help forecast the expected number of outages and customers affected based on upcoming weather forecasts.

Coupled with a user interface driven by the Con Edison team, the platform delivers a powerful tool that displays a projected restoration timeline based on the potential number of crews and includes expected key milestones such as 50%, 90%, and 100% restoration at various levels of geographical granularity.  This provides a repeatable process, leveraging data and technology in which Con Edison can better hone its mutual assistance requests to complement their internal crews and communicate more accurate restoration timelines to customers during major outage events. After the weather passes through the service territory and during restoration, actual numbers for the outages and restorations are used to adjust the projections, providing Con Edison with up-to-date information.  By leveraging this tool Con Edison can:

• Communicate data-driven projections with key stakeholders
• Ensure restoration is tracked with initial projections
• Facilitate increased field efficiency through better planning
• Ensure an effective number of FTE's are requested through mutual aid