Change detection of profile with jumps and its application to 3D printing

Abstract

Three-dimensional (3D) printing, or additive manufacturing, is widely accepted as a disruptive technology, and becomes increasingly popular in manufacturing industries in recent years. As a result, quality control of 3D printing is crucial to improve the quality of products. In this paper, motivated by a novel 3D printing application with fused deposition modeling, we propose a change detection procedure for monitoring profile data with jumps that represents regular structural changes at certain positions. Jumps along with phase variability make profile monitoring challenging because it is hard to combine information in different profiles properly. First, jumps detection procedures are developed by using jump regression analysis technique. After the jumps are detected, a novel piecewise profile registration procedure is suggested to eliminate phase variability. The key information on jumps, profile registration, and the registered profiles is then integrated into an exponentially weighted moving average charting scheme. We use simulation studies and a real-data analysis to show the efficiency of the proposed control chart.