AI based Predictive Maintenance System
An intelligent monitoring solution for industrial HVAC-R systems using IoT and Machine Learning
Project Overview
The AI Predictive Maintenance System is designed to enhance the reliability and efficiency of industrial HVAC-R (Heating, Ventilation, Air Conditioning, and Refrigeration) systems through continuous monitoring and predictive analytics.
This project addresses the critical need for proactive maintenance in industrial settings, where unexpected equipment failures can lead to significant financial losses and operational disruptions.
Technical Stack
Key Features
Real-time Monitoring
Continuous tracking of temperature, humidity, vibration, current, voltage, and gas levels.
Figure 1: Dashboard showing real-time sensor data
AI/ML Predictions
Advanced machine learning models including LSTM Autoencoder and Random Forest for accurate fault detection and predictive maintenance.
Figure 2: Machine learning model architecture
Chatbot Interface
Natural language processing for intuitive system interaction and querying sensor data.
Figure 3: Chatbot interface for system interaction
Hardware Implementation
Custom ESP32-based hardware with multiple sensors for comprehensive environmental monitoring.
Figure 4: Hardware components and setup
System Architecture
End-to-end architecture showing data flow from sensors to cloud and mobile application.
Figure 5: System architecture diagram
Custom PCB Design
Optimized PCB layout for reliable sensor integration and data acquisition.
Figure 6: PCB layout design
Results & Impact
The implementation of our AI Predictive Maintenance System has demonstrated significant improvements in equipment reliability and maintenance efficiency. The system successfully integrates IoT sensors with advanced machine learning models to provide real-time monitoring and predictive insights.
90%
Prediction Accuracy
25%
Reduction in Downtime
30%
Cost Savings
The system's modular architecture allows for easy integration with existing industrial equipment, making it a versatile solution for various industrial applications beyond HVAC-R systems.
Achievements
- OpenAI Researcher Access Program: Our UG Final Year Project "AI-based Predictive Maintenance System for Industrial Machines" was accepted into the OpenAI Researcher Access Program, granting $2000 worth of credits.
- 1st Rank in Tantrapradarshini: Secured 1st Rank in the department-level project exhibition Tantrapradarshini, organized by the ENTC Department in association with IEEE Student Branch and Institutions Innovation Council.