Yong Ge, Hiu Hong Teo, and Dr. Lip Kean Moey, all researchers from SEGi University, have contributed impactful research to the field of intelligent manufacturing with their recent publication in the International Journal of Advanced Manufacturing Technology (Vol. 132, 2024), titled: “Monitoring Tool Wearing: A Hybrid Algorithm Integrating Residual Structures and Stacked BiLSTM.”
As digital transformation reshapes the manufacturing landscape, the need for accurate, efficient, and real-time monitoring of tool wear (TW) has become critical to sustaining productivity and quality control. Traditional monitoring systems rely heavily on manual feature engineering, which can hinder adaptability and efficiency.
In response, the SEGi research team developed a hybrid deep learning framework that leverages the strengths of residual neural networks and stacked bidirectional long short-term memory (SBiLSTM) models to predict tool wear more precisely and with greater generalisability.
Key Contributions of the Study
-
Multi-Scale Feature Extraction: Residual network structures extract adaptive, multi-scale local features from sensor fusion data, improving the depth and accuracy of signal representation.
-
Temporal Pattern Recognition: The SBiLSTM architecture captures time-series dependencies in tool wear signals, enhancing the prediction of wear progression in dynamic environments.
-
Post-Prediction Smoothing: A novel smoothing correction method enhances prediction stability and improves MAE and RMSE scores, demonstrating high-performance consistency.
-
Superior Benchmark Results: The hybrid model outperformed traditional machine learning and other deep learning models, validating its effectiveness in precision manufacturing contexts.
This research demonstrates the value of interdisciplinary integration between AI and mechanical systems, supporting the evolution of predictive maintenance in line with Industry 4.0. It also reinforces SEGi University’s commitment to driving innovation through real-world, application-based research.
This research supports the following United Nations Sustainable Development Goals (SDGs):
-
SDG 9: Industry, Innovation and Infrastructure – by enhancing the adoption of smart manufacturing technologies.
-
SDG 12: Responsible Consumption and Production – by improving machine efficiency, reducing downtime, and minimising material waste through predictive maintenance.