Perancangan Pebaikan Kualitas Produk Baut dan Sekrup Menggunakan Metode Six Sigma dan Data Mining di PT. A

Abstract

PT. A is a manufacturing company that manufactures bolts and screws in accordance with customer demand. This study aims to determine the type of dominant defect that causes a decrease in quality, the factors that cause disability and apply proposed improvements to reduce the level of disability. Improvements are made by using the Six Sigma method, namely the DMAIC stage and making a decision tree (decision tree) using the Data Mining application. Where for the DMAIC stage are: Define, Measure, Analyze, Improvement, and Control (DMAIC). At the define stage, it was found that the production process with the largest number of defects was in the process of heading with the number of defects exceeding the target of 600 PPM (parts per million) that had been determined by the factory during April - June 2017. The heading process was the most important problem focused on the research conducted. The results of identification of Critical to Quality (CTQ) for the characteristics of attribute defects in the production of bolts and screws in the heading process are head burry, head no center, head crack, and body scratch. The second stage is the measure stage, where the control used by statistical process control (SPC) for attribute data is using the P control map and is known to have a sigma 3.50 level with a DPMO value of 22,727. The third stage is the Analyze stage, analyzing the problems of the four types of defects using the Ishikawa diagram. The improvement stage resulted from the decision tree in data mining processing that is obtained by standardizing inspections with seven IF THEN RULE functions, and making maintenance activity plans in the heading process. After implementation, there is a control stage, the results of the analysis show that after implementation using the data mining method and the Six Sigma method, the sigma value has increased which was originally 3.50 to 3.67.