A Novel Computer Vision Based Machine Learning Approach For Online Tool Wear Monitoring In Machining

Abstract

With the increased scope of automated machining processes, one of the essential requirements is the reliable predictions of the tool life. It is crucial to monitor the condition of the cutting tool during the machining process to achieve high-quality machining and cost-effective production. This paper presents a computer vision technique for flank wear measurement and prediction using machine learning, specifically support vector machine (SVM) and boosted decision trees has been used. The proposed methodology for tool wear measurement is illustrated for the CNC machining experimentally. The direct method of tool wear measurement and prediction have been proposed. Flank wear measurement is carried out on PVD coated tool insert, and experiments are performed on an alloy steel workpiece under dry machining. For capturing images, a CMOS camera with a lens mounted on the machine is used. To avoid environmental effects on the images LED ring light is used. Captured tool insert images are provided to the image processing algorithm built-in MATLAB software. The measurement of flank wear is also carried out using a microscope. The prediction accuracy of SVM and optimized boosted tree models is 97% and 96%, respectively, proving prediction algorithms' effectiveness. The findings showcased that the proposed methodology can measure and predict the tool wear with higher accuracy. It has demonstrated the ability to increase cutting tool utilization with the improved surface finish of the machined component.

Innovative Approaches to Sustainable CNC Machining: A Machine Learning Perspective on Energy Optimization

“…The optimization of cutting parameters setting in turning processes is accomplished by employing a response surface approach and machine learning technology. In end milling operations, a machine learning methodology, specifically the Nelder-Mead simplex method, has been developed to optimize the machining variables [47,48].…”