2026 Event Schedule

As utility infrastructure ages and grid complexity increases, traditional protection schemes and time-based maintenance practice are no longer sufficient for maintaining high levels of system reliability and resilience. In this technical session, SDG&E’s System Protection team presents a next-generation approach that integrates machine learning, high-resolution time-series data, and digital twin modeling to enable predictive grid operations and smarter protection strategies. Building on implementations of primary circuit digital twins, the methodology utilizes time-synchronized measurements along with machine learning algorithms to identify early indicators of asset degradation across critical components including transformers, circuit breakers, and relays. These insights are applied in real time to support predictive maintenance, accelerated fault localization, and strategic asset replacement, ultimately minimizing downtime and operational risk. The presentation will detail the end-to-end technical architecture, including the role of a unified operational data platform and the ML training developed for failure prediction and localization. The speakers will share the model validation techniques, integration with legacy protection schemes, and measurable outcomes such as reduced outage duration, improved asset utilization, and enhanced field crew efficiency. This session is designed for engineers, data scientists, system protection experts, and grid modernization leaders seeking to implement scalable and intelligent solutions that shift grid management from reactive to proactive.