Analisis Tingkat Kecacatan Produk Biji Kopi Robusta Menggunakan Statistical Process Control (SPC) (Studi Kasus: PT. Harta Mulia, Blitar, Jawa Timur)

Abstract

Quality is one of the crucial aspects that determines the success rate of a product or service in competing in the market. This study aims to analyze the level of defects in robusta coffee bean products and analyze the application of Statistical Process Control (SPC) in improving the stability of the production process at PT Harta Mulia, Blitar, East Java. The method used is a survey with a quantitative and qualitative approach through observation, interviews, documentation, and analysis of production data for 30 days in January 2026. Analysis tools include check sheets, control maps (p-charts), Pareto Diagrams, and Fishbone Diagrams. The results showed that at the sorting stage, the total production was 1,190,055 grams with 309,651 grams (26%) rejected, while at the roasting stage, the total production was 880,404 grams with 19,933 grams (2.3%) rejected. The control map shows that the sorting process has not been statistically controlled because many points are outside the control limits of 0.260 (CL), 0.267 (UCL), 0.254 (LCL). In the roasting process, although the average defect is relatively low at 0.023(CL), there are still points outside the control limit of 0.025 (UCL) 0.020 (LCL) which indicates a specific variation. The Pareto Diagram analysis identified the dominant defects in the sorting of small beans (90%) and in roasting in the form of broken beans (83%). The Fishbone Diagram shows the main causes coming from the quality of raw materials, manual sorting methods without standardization, as well as the lack of calibration and control of roasting machine parameters. The application of SPC provides a quantitative overview of process variations and forms the basis for improvement recommendations, such as monitoring raw material standards, standardizing process parameters, and calibrating machines at regular intervals. The consistent implementation of statistical-based quality control is expected to minimize the rejection rate and improve product quality stability in a sustainable manner.