Joo Yong-han, Sunmoon University
Currently, various components of the smart factory are being actively applied in the manufacturing industry. In this flow, many excellent examples of acaricides applying various process data to big data analysis models have been published. However, the big data analysis model has difficulties in popularization due to constraints such as huge resource allocation, relatively long learning time, and fine-tuning by experts. In this study, we would like to propose a Functional Data Explorer that can analyze trace data of various processes and equipment. This technique can model the collected data based on the basis function, and returns a functional value as a result. Analysis was performed based on a semiconductor chamber sensor, and it was confirmed that there was a better effect in detecting anomalies based on basic statistics such as mean and deviation. It is expected that the convenience of real-time data analysis will be improved through the analysis technique of this study.
