Encoding Time Series as Images for Anomaly Detection in Manufacturing Processes Using Convolutional Neural Networks and Grad-CAM

Abstract

This study aims to develop an artificial intelligence-based model for analyzing the condition and detecting anomalies by encoding time-series data from manufacturing processes as images. Deep learning has demonstrated the significance of data analysis and anomaly detection in the vision field, and Convolutional Neural Networks (CNN) models have shown exceptional performance and high applicability in image analysis. Based on this, our study intends to utilize image encoding techniques to perform anomaly detection on time-series data. Data such as force, vibration, and sound from equipment during the manufacturing process are collected and transformed into images using various methods, including Gramian Difference Angular Field, Gramian Summation Angular Field, Markov Transition Field, and Recurrence Plot (RP). The transformed image data is then trained and classified for equipment conditions using various CNN models. Finally, we adopt the RP image encoding method and ResNet50 model, which demonstrated the highest accuracy of 99.6%, and compare them to the top 5 models. Based on the high accuracy demonstrated by the top five models, our proposed approach has proven to have significant performance, exhibiting a high success rate of over 90% even when applied to actual data for CNC-machining process. Through this, we propose a process that utilizes the explainable AI Grad-CAM system to identify the feature layer area of the image and confirm the presence of anomalies. With the proposed process, workers can identify abnormal areas or segments of abnormal conditions in the transformed image graph. By providing evidence for state judgment, even inexperienced workers can easily check the condition of manufacturing equipment.