AI-Driven ANFIS Modeling for Smart Drilling to Maximize the Utilization of the Cutting Tools—I4.0 Application

The aim of this research is to develop an ANFIS modeling for smart drilling to maximize the utilization of cutting tools. In this study, first, the tool wear behavior is investigated and clearly recognized. Second, the torque and thrust force signals are measured through a series of experiments at different drilling parameters conditions (flank wear, spindle speed, feed rate, and drill diameter to be used as an indicator of the drilling performance). The relationships between these independent parameters and the torque and force signals are statistically evaluated using (MANOVA) to determine the importance of each parameter in the response. Third, two adaptive neuro-fuzzy inference system (ANFIS) models are established to accurately predict the values of the drilling thrust force and torque at various drilling conditions and to accurately identify the thrust force and torque values at the maximum tool wear corresponding to the end of tool service life where it is necessary to change the tool to avoid tool deficiency and workpiece surface damage. The developed ANFIS models are verified through a series of verification tests. Finally, the Tool Condition Monitoring (TCM) architecture is proposed to be established in future as a smart way to maximize the utilization of the tool for smart drilling operation. This not only guarantees the minimum possible production cost related to the utilization of the tool but also guarantees the best possible dimensional and surface finish accuracy avoiding surface damage correlated with the use of worn tool exceeding the tool service life.