DIGITAL TWIN-ASSISTED QUALITY MONITORING FOR INTELLIGENT ADDITIVE MANUFACTURING PROCESSES

Abstract

Digital Twin technology has emerged as a transformative paradigm for realizing intelligent, adaptive, and autonomous manufacturing systems by establishing a real-time virtual representation of physical manufacturing assets. In additive manufacturing, maintaining consistent product quality remains a major challenge because of the highly dynamic interactions among process parameters, material characteristics, machine behavior, and environmental conditions. Conventional quality monitoring techniques primarily rely on offline inspection and post-process evaluation, making defect detection costly, time-consuming, and incapable of preventing quality degradation during production. The integration of Digital Twin technology with Industrial Internet of Things (IIoT), advanced sensing, artificial intelligence, and cloud computing provides an effective solution for continuous process monitoring, predictive quality assessment, and adaptive manufacturing control. This research proposes a Digital Twin-assisted quality monitoring framework for intelligent additive manufacturing processes that combines real-time sensor data acquisition, virtual process modeling, machine learning-based defect prediction, process simulation, and closed-loop feedback control to continuously evaluate manufacturing quality throughout the production lifecycle. The proposed framework synchronizes physical manufacturing equipment with its virtual counterpart using bidirectional data communication, enabling real-time visualization, process optimization, anomaly detection, and intelligent decision support. Multiple sensor streams, including thermal images, melt pool characteristics, acoustic emissions, vibration signals, environmental conditions, and machine operating parameters, are continuously integrated into the Digital Twin model for dynamic process analysis. Machine learning algorithms, including Random Forest, Support Vector Machine, Extreme Gradient Boosting, Artificial Neural Networks, and Long Short-Term Memory networks, are employed to predict quality deviations and manufacturing defects while optimization algorithms recommend corrective process adjustments. Experimental evaluation demonstrates that the proposed framework achieves a quality prediction accuracy exceeding 98%, reduces manufacturing defects by approximately 42%, improves process stability by 35%, decreases inspection time by 55%, and enhances overall production efficiency by nearly 38% compared with conventional quality monitoring approaches. The proposed Digital Twin-assisted framework provides a scalable, intelligent, and data-driven solution for achieving autonomous quality assurance in Industry 4.0- enabled additive manufacturing systems. Keywords: Digital Twin, Additive Manufacturing, Quality Monitoring, Machine Le