Applied artificial intelligence-based equipment condition monitoring in manufacturing industry

Advancements in the smart manufacturing sector have significantly increased the adoption of real-time equipment health monitoring systems. Digital twin technology has the potential to reduce rework costs, enhance machine tool uptime, and improve the dimensional accuracy of manufactured products. By utilizing digital twin technology, machining processes can now be monitored in real-time with operational and environmental variations. Digital twins provide a comprehensive view of equipment health, enabling more effective anomaly detection and fault diagnostics. The proposed study introduces an anomaly detection framework for machining, leveraging digital twin technology to facilitate real-time equipment health monitoring. The proposed framework integrates secure monitoring through blockchain technology combined with the Internet of Things to analyze production trends. Furthermore, it employs a dual-directional convolutional neural network approach to enable real-time vulnerability detection and optimize decision-making processes. Experimental simulations were performed to determine the validation of the presented approach. Based on experimental simulations, enhanced performance was registered in terms of Delay Efficiency (5.21 ms), Prediction Performance (Accuracy (92.25%), Sensitivity (93.25%), (Specificity (94.25%)), F-Measure (95.15%)), Energy Efficacy Analysis (1.18 mJ), Model Reliability (95.24%), and Stability Analysis (79%).