<![CDATA[Statistical quality control (SQC) is a fundamental approach in modern manufacturing for ensuring that production processes consistently conform to specified quality requirements. This study applies an integrated SQC framework comprising the Xbar-R control chart, one-way analysis of variance (ANOVA), and process capability analysis to evaluate the dimensional quality of mechanical components sourced from an open manufacturing dataset publicly available on Kaggle (Parts Manufacturing Industry Dataset). Three critical dimensions were examined: length, width, and height, each measured by 20 operators with five replicates per operator (n = 100 per dimension). Prior to control charting, data normality was verified through probability plots based on the Anderson-Darling statistic. The Xbar-R analysis revealed out-of-control signals for the length dimension at subgroups 3 and 5, and for width at subgroup 3, indicating the presence of assignable (special) causes of variation. One-way ANOVA demonstrated that inter-operator effects on width were statistically non-significant (F = 1.34, p = 0.185), whereas significant operator-to-operator differences were detected for length (F = 2.10, p = 0.012). Process capability indices showed that all three dimensions fell below the minimum acceptable threshold of Cp ≥ 1.33, with length being most critical (Cp = 0.84, Cpk = 0.71). The findings underscore the necessity of targeted corrective actions including measurement system standardisation, operator retraining, and process variability reduction. ]]>
Copyrights © 2026
