Predictive Maintenance for Elevators

Imagine giving elevators a sixth sense—this system does just that. Using real-time sensors and ML, it predicts when maintenance is due before things go south. It’s smarter than your average maintenance guy and works 24/7 without a coffee break.Safety, efficiency, and no awkward breakdowns—what’s not to love?

To see how its done click here!

This project focuses on enhancing elevator reliability and safety through a smart predictive maintenance system. By integrating sensors and machine learning, the system ensures timely maintenance, preventing unexpected breakdowns and reducing downtime.

Key Features:

Data Collection and Integration:
- Real-time data collection using sensors such as accelerometers, IR, DHT (temperature and humidity), and sound sensors.
- Monitored key elevator components, including environmental conditions, steel rope wear, and brake speed, to gather actionable insights.
Predictive Model Development:
- Built a predictive maintenance model using the Random Forest algorithm, achieving a 95% accuracy rate for service date forecasts.
- Leveraged sensor data to identify early warning signs of component wear and predict maintenance needs.
System Integration:
- Developed a Flask server connected to an Arduino for seamless data transmission and processing.
- Built a real-time frontend application to display component health, environmental conditions, and predicted maintenance schedules.
Impact and Results:
- Reduced elevator downtime by 30% through proactive maintenance scheduling.
- Enhanced safety by identifying and addressing maintenance needs before failures occurred.
- Improved operational efficiency and service reliability in elevator systems.

This project demonstrates the potential of combining IoT sensors, machine learning, and web technologies to revolutionize maintenance strategies in industrial and commercial systems.