Predictive Maintenance for Injection Molding Machines Enabled by Cognitive Analytics for Industry 4.0.

Area of Science:

• Industrial Engineering
• Artificial Intelligence
• Data Science

Background:

• Industry 4.0 leverages big data and cognitive systems for enhanced manufacturing.
• Predictive maintenance is crucial for real-time decisions and early fault detection.
• Challenges include diverse industrial data streams and lack of labeled data.

Purpose of the Study:

• To introduce a cognitive analytics system for Industry 4.0 predictive maintenance.
• To present a methodology for real-time anomaly detection in industrial data.
• To apply the system to injection molding machines for fault detection.

Main Methods:

• Developed a cognitive analytics system with self- and autonomous-learning capabilities.
• Implemented ensemble prediction models combining supervised and unsupervised learners.
• Deployed models on a real-time monitoring system for incoming data streams.
• Incorporated a novel cognitive mechanism with real-time self-retraining and a double-oriented evaluation objective (data-driven and model-based).

Main Results:

• The system effectively detects faults in real-time industrial data streams.
• Demonstrated successful application on injection molding machines.
• The cognitive mechanism enables continuous improvement and adaptation of the predictive models.

Conclusions:

• The proposed cognitive system enhances predictive maintenance in Industry 4.0.
• Artificial intelligence capabilities significantly advance industrial maintenance activities.
• The system offers a robust solution for real-time anomaly detection and fault prediction.