Perbaikan kualitas pada produk pig receiver menggunakan Metode Six Sigma dan Data Mining pada PT Farrel Internusa Pratama
P enelitian ini berfokus untuk memberikan usulan perbaikan pada proses produksi pig receiver di PT Farrel Internusa Pratama menggunakan pendekatan six-sigma dan data mining. PT Farrel Internusa Pratama merupakan perusahaan yang bergerak pada industri minyak dan gas, dengan produk utamanya yaitu pig launcher dan pig receiver. Pig receiver dipilih sebagai objek penelitian karena nilai rata-rata persentase produk cacat untuk pig rceiever adalah 7.46%, nilai ini melebihi persen toleransi produk cacat perusahaan yaitu 5%. Penelitian ini hanya berfokus pada proses welding yang menjadi penyumbang produk cacat dengan jumlah tertinggi. Pada tahap define dibuat diagram Supplier Input Process Output Costumer, dan identifikasi Critical to Quality (CTQ). Terdapat 5 jenis kecacatan yang menjadi Critical to Qualitu yaitu: pinhole, overlap, crack, undercut, dan distorsi. Selanjutnya pada tahap measure, dilakukan perhitungan stabilitas proses menggunakan peta kendali p dan u. Pada tahap ini juga, dilakukan perhitungan DPMO dan tingkat sigma dari proses produksi. Adapun nilai DPMO dan tingkat sigma untuk kegiatan produksi perushaan secara berturut-turut adalah 44600 dan 3.1996. Pada tahap analyze dilakukan analisa untuk mengetahui penyabab kecacatan. Adapun jenis kecacatan yang paling dominan menurut diagram pareto adalah crack, pinhole, dan distorsi. Serta berdasarkan alogritma data mining yang diterapkan yaitu decision tree C.45, diketahui bahwa pinhole merupakan jenis kecacatan yang paling berpengaruh pada produk reject. Metode 5 why’s analysis and 1 h digunakan pada tahap improve untuk merumuskan usulan perbaikan. Usulan perbaikan yang berhasil dirumuskan yaitu: pembuatan form pembersihan mesin, check sheet pengecekan kawat las, dan juga pemberian pelatihan terhadap operator welding. Adapun usulan perbaikan yang berhasil diimplentasikan adalah form pembersihan dan check sheet. Setelah melakukan implementasi, pada tahap control dilakukan perhitungan kembali terhadap stabilitas proses produksi dan nilai DPMO. Berdasarkan hasil perhitungan diketahui bahwa tingkat sigma kegiatan produksi setelah implementasi usulan perbaikan mengalami peningkatan sebesar 0.1812
T his study focuses on providing suggestions for improvements to the pig receiver production process at PT Farrell Internusa Pratama using a six-sigma approach and data mining. PT Farrell Internusa Pratama is a company engaged in the oil and gas industry, with pig launchers and pig receivers as their main products. Pig receiver was chosen as the object of research because of the average value of the percentage of defective products for pig receivers is 7.46%, this value exceeds the percent tolerance of the company's defective products, which is 5%. This research only focuses on the welding process which contributes the highest number of defective products. In the define stage, a Supplier Input Process Output Customer diagram is made, and the Critical to Quality (CTQ) is identified. There are 5 types of defects that are included as the critical to quality, namely: pinhole, overlap, crack, undercut, and distortion. Furthermore, at the measure stage, the process stability is calculated using the p and u control charts. As well as DPMO and sigma levels of the production. The DPMO value and sigma level for the company's production activities are 44600 and 3.1996. At the analyze stage, an analysis is carried out to determine the main causes of the defect. Cracks, pinholes, and distortions are the most dominant defects according to the pareto diagram. Based on the data mining algorithm applied, the C.45 decision tree, it is known that pinhole is the type of defect that has the most influence on reject products. The 5 why's analysis and 1 h methods are used at the improve stage to formulate improvement suggestions. Improvement suggestions that have been successfully formulated are machine cleaning forms, check sheets for welding wire checking, and also providing training to the welding operators. The proposed improvements that have been successfully implemented are the cleaning form and check sheet. After implementing, at the control stage, recalculation of the stability of the production process and the value of DPMO is carried out. Based on the calculation results, it is known that the level of production activity sigma after the implementation of the proposed improvement has increased by 0.1812.