Data-driven Exploration and Process Optimization for a Milling-boring Machine

The incorporation of cyber-physical systems to manufacturing enables the collection and storage of big amounts of machine data. These data, properly studied, provide useful information about the machine, its use and its state. In this work a use case about the utilization of machine data for maintenance and process optimization for a milling-boring machine is presented. First, some guidelines on data exploration and machine operation identification from raw machine data are introduced. Then, the results of this exploration are used to compute some descriptive statistics and to train a machine learning model. From this data analysis some conclusions about the relation of the spindle vibration with other machine variables are drawn.