KUL01: IslandEnergyAI: Federated Generative AI-based Electricity Forecasting with Data Security for Islands

Remote islands face major challenges in electricity management due to isolated micro-grids, unpredictable electricity demand, energy grid cybersecurity vulnerabilities, and limited access to centralized intelligence. These issues lead to energy waste, electricity distribution errors, blackouts, high operational costs, and sustainability issues. The proposed IslandEnergyAI system introduces a decentralized, privacy-preserving solution for secure electricity forecasting using Generative Artificial-Intelligence (AI) based forecasting models integrated with AES-HMAC cryptography (AES: Advanced Encryption Standard; HMAC: Hash-based Message Authentication Code) in a Federated-Machine-Learning environment. This design enables distributed edge-island-micro-grid nodes to forecast energy load patterns collaboratively. LSTM (Long Short-Term Memory) machine-learning-based Generative-AI supports distributed energy forecasting for island nodes, while AES-HMAC-cryptography ensures secure data communication. Expected outcomes include enhanced grid reliability, reduced energy waste, optimized load distribution, and strengthened cybersecurity. This sustainable approach reduces fossil fuel dependency and supports renewable energy adoption. The business direction focuses on offering the solution as a subscription-based digital service for island utilities, governments, and energy-focused startups, contributing to the global transition toward smart, green, and secure microgrid infrastructures.

We aim to develop a secure, AI-powered digital framework for island microgrids: a prototype energy forecasting dashboard that combines machine learning-based Generative AI with AES-HMAC cryptographic protection to enable accurate, privacy-preserving and sustainability preserving electricity forecasting.

SQL Database
Marine Science
Python
Deep Learning
Socio-economist