The industrial transformation, driven by decreasing manufacturing times and increasing quality requirements, brings a variety of challenges to the production technology. Techniques such as high-speed laser beam cutting are of great importance in this field of study. It allows for the production of precise products with the highest possible flexibility. The complexity and speed of such a laser cutting process has high demands on signal processing, as a large number of parameters must be evaluated within short periods of time.

To enable the control of such a process, this project analyzes the response signals of a minimally invasive laser modulation. The goal of the project is to extract features from a raw diode signal and perform an automated classification. In previous analyses, it was found that anomalies in the signal are only detectable shortly before the cutting process is interrupted. In order to enable timely control, an analysis result must be available in a timely manner. In addition, it needs to be examined to what extent the signal analysis carried out allows conclusions to be drawn about the trend of the process. It should be determined which signal characteristics have the greatest significance for maintaining the process.