Electrifying Big Issues Through Analytics
summary
Southern California Edison uses its sensor and enterprise data across 15 million customers, 1.4 million poles and 1 million tracked trees to run daily probabilistic models that predict storm damage and assign risk up to 21 days ahead, cutting response time and prepositioning crews. Technologies in production include Postgres, Shiny dashboards, Azure, random forests, XGBoost and deep learning plus computer vision on poles and drone imagery to flag faults, and a field-facing digital recommender of hazards that integrates into crew tailboards. Remaining work: reliably predicting slow visual degradation from time‑series images.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 6:58Most dangerous jobs in the US
- 14:57Proactive Storm Analytics
- 19:04Field Inspection Using Drones
- 25:48Bootstrapping
- 30:15Expanding Renewable Energy
- 33:52Virtual Reality
related talks
23mDataEngBytes 2023 - MEL-T3-01 - Rajath Vanikul
32mPredicting Grid Faults Before They Happen: AI for Weather-Driven Grid Resilience
27mPower of big data and data science in service of excellent customer experience
38mReal world data & analytics journeys
18mDatabricks drives predictive maintenance, grid optimization with analytics
10mPaco Nathan (Liber 118) Interview - OSCON 2014
38mData and Business
24mCreating a Data Platform for Software-Defined Electric Vehicles