Data-driven Model for CMM Probe Calibration to Enhance Efficiency and Sustainability

Abstract

Ensuring precise measurement is critical for manufacturing industries; and a coordinate measuring machine (CMM) is widely used for automated inspection of critical and complex components. But there is a lack of clarity regarding optimal timepoint for CMM probe re-calibration. Too frequent calibrations increase the cost and an uncalibrated CMM probe affects the inspection quality. When to get a CMM probe calibrated is still an open question. It is important to accurately determine the optimal probe re-calibration intervals to mitigate the adverse effects on the quality and the time. This paper presents a novel data-driven approach to predict the optimal timepoint to re-calibrate a CMM probe. The data is collected at various levels to develop, train, and validate the predictive model, which includes numerical and categorical data-frame such as probe usage metrics, probe calibration environmental condition, probe specifications, workpiece data, probe calibrated dimension data and calibration requirement. The research potentially aims to reduce the CMM downtime without compromising the product quality variations that arise from the anomalies/errors related to the probe re-calibration thereby enhancing the inspection efficiency as well as sustainability.