Human-Centric Analytics for Resilient and Modern Power System Control Rooms

Power system operators at control centers play a critical role in ensuring grid resilience and reliability during blackouts and other critical events. This paper introduces COGNIZANT, a next-generation human-machine interface platform that integrates advanced power system simulation with artificial intelligence-driven decision support and behavioral analytics. Developed jointly by Texas A&M University and the National Renewable Energy Laboratory (NREL), the platform addresses the critical knowledge gap created by retiring experienced operators while preparing the workforce for increasingly complex grid challenges. The platform combines PowerWorld Dynamic Simulator and GE's Advanced Energy Management System with large language models fine-tuned using Low-Rank Adaptation and retrievalaugmented generation to create an immersive training environment. It enables simulation of over 1,000 diverse scenarios including extreme weather events leveraging historical ERA5 reanalysis data from 1940-2025, cyber-physical threats, and compound multi-domain events. The system is designed to achieve sub-second alert generation and targets a 20 % improvement in operator response times through comprehensive training validated using eye-tracking, cognitive load assessment, and behavioral analysis. This joint research effort establishes a closed-loop framework where NREL's ARIES platform generates high-fidelity grid scenarios and Texas A&M's Smart Grid Center evaluates operator performance, bridging the gap between grid physics and human factors to accelerate deployment of operator-centered resilience solutions.