SUMMARY

Computer numerical control (CNC) milling machines are vital pieces of equipment in modern machine shops that allow a facility to quickly and accurately produce components. The ability to autonomously perform real time identification of the cutting parameters can help the machine users. In a makerspace environment, it can act as a safety check that can alert users to a potentially unsafe cut. This built-in safety check can provide quality control and lower the barrier to entry by enabling amateur users to operate the complex manufacturing equipment (like the CNC machines) in makerspaces. Additionally, in a manufacturing plant, the operator can be notified if a toolpath is being performed outside the expected operating window. Machine learning has allowed researchers to analyze data and gain new insights from these machines that were previously impossible. Researchers have used advanced sensors and machine learning in manufacturing to gather the information that can provide insight into factors such as tool behavior or material properties. However, these approaches are limited to specific tool paths and require external sensors to be mounted to the machine. Real-world machining encompasses a wide range of different tool paths with varying machining parameters and external sensors add cost and complexity to a machine. It was demonstrated that applying machine learning strategies to data collected from built-in sensors on a CNC mill can be used to detect and identify characteristics of the cutting parameters being used for a range of tool paths. Spindle power and load sensor data were extracted and used to train different machine learning algorithms to estimate machining parameters.