This paper addresses the application of artificial intelligence (AI) for optimizing parameters in the field of laser cutting. The goal of this concept is not only to reduce expensive parameter studies but also to shorten the required research time and to connect the evaluation with the process parameters to create a digital dataset.


For this purpose, an experimental setup has been developed which enables fully automatic experiments with various materials and material thicknesses. The results of these experiments are captured using high-resolution cameras and a 3D sensor, then evaluated and automatically stored in a database. The analysis of this data allows for the identification of optimal parameters for laser cutting.


Particular attention is paid to the implementation of Bayesian optimization in the field of AI. By creating this automation unit, it becomes possible to determine parameters for different materials with varying geometries in fully automatic mode. The parameters are automatically transferred to the laser cutting system, which performs the experiment. Subsequently, the sensor data is evaluated to assess the quality of the parameter combination.


A crucial element is the design of the cost function, which has been integrated into the optimization process. This function has been developed to enable efficient evaluation of parameter combinations and thus minimize the number of required experiments.


This system promises more efficient laser cutting parameter adjustments, boosting manufacturing quality and productivity. An efficient cost function helped identify optimal parameter combinations with fewer experiments.