AI for Renewable Energy & Utilities

Course outline and learning expectations

This is a tutor-led course. The outline shows what your tutor will cover; teaching materials and examinations are provided directly to enrolled students.

What you will learn

• Explain energy-access, grid, off-grid and mini-grid challenges relevant to Uganda and Africa
• Prepare and validate weather, generation, demand, payment, sensor and maintenance time-series data
• Build and compare baseline and machine-learning models for solar power generation forecasting
• Forecast electricity demand and evaluate performance through time-aware backtesting
• Formulate battery scheduling models using state-of-charge, efficiency, reliability and degradation constraints
• Develop fault and outage risk models and assess class imbalance, false alarms and operational consequences
• Evaluate PAYGo solar risk models using privacy, fairness, explainability and human-oversight principles
• Prototype maintenance forecasting for boreholes, pumps or related water-utility assets
• Build, document and present a reproducible energy or utility decision-support prototype

Modules

1. 1

   Energy access and utility systems in Uganda and Africa

   Analyse the operational and data challenges affecting grid, off-grid and mini-grid services in Uganda, East Africa and Africa

   Energy access pathwaysGrid, mini-grid and stand-alone systemsRenewable energy resourcesUtility data sourcesData quality and missing observationsAI opportunities, limitations and responsible use
2. 2

   Solar power generation forecasting

   Build and evaluate short-term solar generation forecasts from historical production and weather-related data

   Solar PV generation fundamentalsIrradiance, temperature and cloud variablesTime-series feature engineeringPersistence and statistical baselinesMachine-learning forecastingMAE, RMSE and forecast uncertainty
3. 3

   Energy demand forecasting

   Model household, commercial and mini-grid demand patterns for planning and operations

   Load profiles and seasonalityCalendar and weather featuresData aggregation and leakage preventionRegression and time-series modelsPeak-demand forecastingBacktesting and error analysis
4. 4

   AI-assisted battery-use optimisation

   Develop a practical battery scheduling model for off-grid and mini-grid operations while respecting technical constraints

   State of chargeCharging and discharging limitsRound-trip efficiencySolar-load balanceRule-based and optimisation baselinesCost, reliability and battery-degradation trade-offs
5. 5

   Fault and outage risk prediction

   Use operational data to estimate equipment-fault and outage risk without overstating model certainty

   Failure and outage labelsAsset and maintenance featuresClass imbalanceAnomaly detectionPrecision, recall and false-alarm costsRisk ranking, maintenance prioritisation and human review
6. 6

   PAYGo solar analytics and responsible credit-risk modelling

   Evaluate payment and service data for responsible PAYGo portfolio analysis with privacy, fairness and human oversight

   PAYGo business modelRepayment and usage featuresDefault-risk baselinesExplainability and bias testingConsent and data protectionHuman review and appropriate limits on automated decisions
7. 7

   Water-utility and pump maintenance forecasting

   Prototype condition-monitoring and maintenance models for boreholes, pumps and related water assets

   Sensor and maintenance-log dataFlow, pressure, runtime and energy featuresFailure-mode definitionAnomaly detection and remaining-useful-life conceptsMaintenance alertsData scarcity and field-validation limits
8. 8

   Capstone: renewable-energy forecasting prototype

   Build and present a reproducible decision-support prototype for a Ugandan or African energy or utility use case

   Problem definitionData audit and baselineModel developmentOperational constraintsEvaluation and uncertaintyDashboard or APIDocumentation, ethics review and presentation