Implementasi grafik pengendali Multivariate Exponentially Weighted Moving Average (MEWMA) pada pengendalian kualitas proses produksi gula

Studi kasus PT PG Rajawali I Kabupaten Malang

Abstract

This study discusses the application of the Multivariate Exponentially Weighted Moving Average (MEWMA) control chart in monitoring the quality of white crystal sugar (GKP) production at PT PG Rajawali I, Malang Regency. GKP is one of the strategic commodities in Indonesia whose demand continues to increase, making quality control an essential aspect to ensure that products meet both national standards and consumer needs. In this research, quality control was carried out by analyzing three key quality characteristics, namely moisture content, grain specific gravity, and solution color (ICUMSA). The research employed a quantitative approach using secondary data obtained from the Quality Assurance division of PT PG Rajawali I during the milling period from May 16 to July 20, 2025. The analysis included a multivariate normality test, correlation analysis among variables, and the construction of MEWMA control charts with various weighting factors (λ). The results showed that all MEWMA charts indicated out-of-control signals, which reflects the presence of variation in the production process. The optimal weighting factor was identified at λ = 0.1, as it was the most sensitive in detecting small shifts in the process mean. These findings suggest that production stages such as evaporation, cooking, and centrifugation significantly affect sugar quality. Thus, the implementation of the MEWMA chart proved to be effective as an early-warning statistical tool to detect deviations, reduce product defects, and improve the efficiency of white crystal sugar production.