Smart Anomaly Detection and Prediction for Assembly Process Maintenance in Compliance with Industry 4.0

Area of Science:

• Manufacturing Engineering
• Industrial Internet of Things (IIoT)
• Machine Learning

Background:

• Assembly lines face significant risks from unexpected cessations, often caused by component failures like seized carrier bearings.
• Standard maintenance procedures are insufficient for resolving certain anomalies, leading to costly production halts.
• Carrier bearing seizure on assembly conveyors can halt the entire production process.

Purpose of the Study:

• To develop a predictive system for carrier bearing anomalies on assembly conveyors.
• To mitigate risks associated with unexpected production cessation in manufacturing.
• To integrate Industrial Internet of Things (IIoT) devices, neural networks, and sound analysis for anomaly prediction.

Main Methods:

• Deployment of Industrial Internet of Things (IIoT) sensors for data acquisition.
• Application of neural networks for pattern recognition and anomaly detection.
• Utilizing sound analysis to identify early signs of bearing degradation.

Main Results:

• A smart system capable of detecting and predicting arising anomalies was created and deployed.
• The integrated approach successfully reduced unexpected production cessations.
• The system demonstrated the benefits of combining IIoT, neural networks, and sound analysis.

Conclusions:

• The proposed unique approach effectively addresses the limitations of standard maintenance for bearing failures.
• The developed smart system enhances predictive maintenance capabilities in manufacturing.
• Significant reductions in unexpected production downtime were achieved through the implemented solution.