Advancing smart city factories: enhancing industrial mechanical operations via deep learning techniques

Area of Science:

• Industrial Engineering
• Artificial Intelligence
• Environmental Science

Background:

• The need for robust anomaly detection systems in industry is critical for operational integrity.
• Existing methods often lack the precision and real-time capabilities required for modern industrial environments.

Purpose of the Study:

• To develop and evaluate a deep learning-based system for real-time anomaly detection and mitigation in industrial settings.
• To assess the system's performance in terms of accuracy, efficiency, and environmental impact.

Main Methods:

• Implementation of a Long-Short Term Memory (LSTM) deep learning model for anomaly detection.
• Integration of data acquisition and analytical processing for real-time surveillance.
• Development of autonomous capabilities for proposing or implementing remedial actions.

Main Results:

• The LSTM model achieved high performance metrics: 95% accuracy, 90% recall, and 92.5% F1 score.
• Demonstrated significant environmental benefits, including a 25% reduction in CO2 emissions and a 20% decrease in water usage.
• The developed system outperformed previous models in speed and precision.

Conclusions:

• Deep learning, specifically LSTM networks, offers a powerful solution for industrial anomaly detection.
• Automated systems are key to enhancing operational efficiency and driving sustainability in the industrial sector.
• The study validates the effectiveness of advanced AI in creating more efficient and environmentally conscious industrial operations.