| 1. Introduction to ML in Utilities | EIA-930 (regional demand), Synthetic SCADA snapshots | Simple regression and visualization of utility data trends |
| 2. Utility Data Foundations | Synthetic SCADA, AMI smart meter data, GIS feeder layouts | Data cleaning, resampling, joining telemetry and asset layers |
| 3. ML Fundamentals for Grid Applications | Synthetic load curves (AMI), SCADA readings | Regression (load prediction), classification (asset health labels), clustering (AMI load profiles) |
| 4. Load Forecasting and Demand Analytics | EIA-930 demand data, NOAA weather data | Time series forecasting (SARIMA), regression (weather vs. load) |
| 5. Predictive Maintenance for Grid Assets | SCADA telemetry (temperature, current), Synthetic EAM records | Classification (failure prediction), anomaly detection (early degradation signals) |
| 6. Outage Prediction and Reliability Analytics | NOAA storm data, PUC outage reports, GIS feeder maps | Classification (storm outage risk), geospatial joins for feeder exposure |
| 7. Grid Operations Optimization | SCADA voltage/reactive power data (synthetic IEEE feeders) | Reinforcement learning (voltage control), optimization for VAr dispatch |
| 8. Renewable Integration and DER Forecasting | NREL NSRDB solar irradiance, PV generation profiles | Regression (PV power modeling), time series forecasting (PV output) |
| 9. Customer Analytics and Demand Response | AMI smart meter data, synthetic DR event logs | Clustering (customer segments), classification (participation prediction) |
| 10. Computer Vision for Inspections | Synthetic drone imagery of lines/substations, NDVI vegetation maps | CNN-based defect detection, segmentation (vegetation encroachment) |
| 11. NLP for Maintenance and Compliance | Synthetic inspection logs, NERC CIP text | NLP: text classification (routine vs. failure), entity extraction (assets, failure modes) |
| 12. MLOps for Utilities | Predictive maintenance model outputs (Chapter 5) | Model versioning, automated retraining workflows (MLflow) |
| 13. Cybersecurity Analytics | CICIDS2017 network traffic dataset, synthetic SCADA logs | Anomaly detection (unsupervised), intrusion classification (supervised ML) |
| 14. Integrated Analytics Pipelines | SCADA, outage risk models (Ch. 6), maintenance scores (Ch. 5) | Orchestration (Prefect pipelines) combining outputs for operational dashboards |
| 15. AI Ethics and Governance | Predictive maintenance risk outputs (urban vs. rural segmentation) | Fairness audits (performance parity), explainability (SHAP analysis) |
| 16. Workflow Orchestration | Combined datasets (load, outage, maintenance) | Automated scheduling of multi-model analytics |
| 17. Large Language Models and Multimodal AI | Maintenance logs (NLP) + SCADA telemetry + drone imagery | Multimodal AI combining text, structured data, and image insights |
| 18. AI Roadmap for Utilities | Aggregate datasets from prior chapters | Scenario modeling of AI maturity impacts (cost savings, SAIDI/SAIFI improvement) |
| 19. Enterprise Integration (SCADA/GIS/EAM) | SCADA telemetry, GIS asset layers, EAM maintenance records | Unified data pipelines linking IT/OT for analytics consumption |
| 20. AI Platform Deployment | Predictive maintenance and outage prediction models | Real-time model deployment (API endpoints, streaming inference) |