Research & Innovation Portfolio

Developing Artificial Intelligence, Industrial IoT, and Smart Energy solutions for predictive maintenance, real-time monitoring, and sustainable infrastructure.

Featured Project

AI-Driven Predictive Maintenance for Hydropower Plants

Overview

Hydropower plants often experience unexpected equipment failures that can lead to costly downtime and reduced energy production.

This project develops an AI-powered predictive maintenance framework that combines Industrial IoT, cloud computing, and machine learning to monitor hydraulic turbine health in real time and detect anomalies before failures occur.

Project Information

Problem Statement

Traditional maintenance strategies are often reactive and rely on manual inspections.

This can result in:

  • Unexpected equipment failures
  • Increased maintenance costs
  • Reduced operational reliability
  • Energy production losses

Proposed Solution

An intelligent monitoring system integrating:

  • 📡 IoT Sensors
  • ☁ AWS Cloud Infrastructure
  • 🧠 LSTM Deep Learning Models
  • 📊 Real-Time Monitoring Dashboard
  • 🚨 Automated Anomaly Detection

Realtime Supervision

Operator’s Dashboard at Kavumu plant

Real-time monitoring interface displaying:

  • Turbine vibration
  • Generator temperature
  • Blade position
  • Electrical parameters
  • Anomaly alerts

350 kW Capacity production

Hydropower Real-Time Monitoring Dashboard

Open Realtime Anomaly Detection

Turbine’s Vibration

Latest: —

Cloud Infrastructure: Amazon Web Services (AWS EC2)

Real-Time Data Processing and Deployment Architecture

LSTM-Based Real-Time Anomaly Detection for Hydraulic Turbine Vibration

Research Contributions

✔ AI-based predictive maintenance framework

✔ Real-time anomaly detection

✔ Cloud deployment architecture

✔ Industrial IoT integration

✔ Time-series analytics

✔ Hydropower monitoring system