Application of Deep Learning on Gage R&R for Anomaly Detection

This thesis explores the application of deep learning techniques, specifically autoencoder based models, to enhance anomaly detection within Gage Repeatability and Reproducibility (Gage R&R) studies—an essential component of Measurement System Analysis (MSA) in quality engineering. Traditional Gage R&R methodologies, while effective for linear and low-dimensional data, exhibit limitations in detecting subtle, nonlinear variations in complex measurement systems. To address this challenge, an unsupervised autoencoder was developed and trained on a synthetically generated dataset comprising 2,500 voltage measurements (5V and 33V) derived using Generative Adversarial Networks (GANs) based on real-world manufacturing data measurements. The proposed autoencoder model achieved a 95th percentile-based anomaly detection threshold, successfully identifying anomalies with an overall accuracy of 94.6%, a precision of 92.1%, and a recall of 95.8%—surpassing the diagnostic effectiveness of traditional ANOVA-based Gage R&R methods. The reconstruction error analysis clearly differentiated anomalous measurements, providing a robust visualization of system inconsistencies. Statistical validation using ANOVA indicated minimal 2 operator-induced variability, affirming the system’s robustness in capturing part specific anomalies. The findings underscore the potential of integrating deep learning into Gage R&R analysis, offering a scalable, data-driven alternative for anomaly detection in high-dimensional measurement environments. This approach aligns with the increasing automation and precision demands of Industry 4.0