North American Forecasting Model: Outage, Damage and Load Resilience


A Scalable “MEGAPROJECT” for National Resilience:

PI: Diego Cerrai (UConn)

Co-PIs:

Marina Astitha, E. Anagnostou, D. Song (UConn);

Jeffrey Freedman, Sukanta Basu (Albany)

The WISER North American Forecasting Model transforms weather data into actionable intelligence, allowing energy companies to shift from reactive to proactive grid management by integrating AI-driven predictions for:

  • Infrastructure damage

  • Customer outages

  • Energy load spikes

Industry Relevance

Storm outages increase financial risks & undermine reliability. Current prediction gaps lead to inefficient resource allocation.

Develop the North American Outage Prediction Model to predict system failures across the entire CONUS + Quebec, blending proprietary & public data

Project Goal


Phased Roadmap

The Proposal for Phase 1 is currently on a 24-month plan per its deliverables.

Phase 1: Local

(Years 1-2)

ISO-NE, NYISO, CAISO. Operational OPM Pilot

Phase 2A: Regional

(Years 3-4)

East Coast (FL to QC) & West Coast (CA to WA)

Phase 2B: National

(Years 5-6)

Full CONUS + Quebec Deployment


Technical Approach

Collaborative Teams: Weather (AI-enhanced forecasting, CONUS404, HRRR) & Outage (Integrate proprietary + public EAGLE-1 Data to fill gaps).

First modular, scalable OPM for all WISER members.

High ROI, attracts new utilities.

Synergies with Demand Models.

Research Impact