Loading...
Back to Projects
Predictive Maintenance Monitoring System
An IoT-based predictive maintenance solution using XGBoost machine learning model to forecast equipment failures with 97.95% accuracy, monitoring multiple machine conditions in real-time.
Project Overview
A comprehensive predictive maintenance system designed to prevent unexpected machinery breakdowns in manufacturing environments. The system leverages IoT sensors and an XGBoost classifier to monitor critical parameters including temperature, pressure, motor power, vibration, and volume flow. By analyzing real-time sensor data through MQTT protocol, the system predicts five key machine conditions: cooler condition, internal pump leakage, hydraulic accumulator status, and overall stability, achieving an impressive 97.95% accuracy rate. This solution helps industries reduce maintenance costs, increase equipment lifespan, and prevent costly production outages.
Architecture Diagrams
Key Features
Real-time monitoring using IoT sensors (temperature, pressure, vibration, voltage, current, flow)
XGBoost classifier model with 97.95% overall accuracy
Predicts 5 critical machine conditions simultaneously
MQTT-based real-time data gathering and processing
Flask backend with machine learning model integration
Interactive dashboard for equipment health visualization
Preventive maintenance alerts and notifications
High scalability and customizability for various industries
Images
Technology Stack
Backend
PythonFlaskXGBoostMQTT Client
Frontend
ReactTypeScriptChart.js
Tools & Platforms
MQTT BrokerESP32IoT Sensorsscikit-learn