Q Quality Improvement on Pipe Production Using Six SigmaMethod and Data Mining in PT Farrel Internusa PratamaPT. Farrel Internusa Pratama is an industrial company engaged inmanufacturing, especially construction of oil and gas pipelines. The product withthe highest number of defects is pipe products. Therefore the research is devoted topipe products. Based on the company's monthly production reports, it was foundthat the largest percentage of defects was during the welding process which reachedmore than 6%. The aim of this research is to improve the quality of pipe productsusing the Six Sigma method and Data Mining. The approach used DMAIC (Define,Measure, Analyze, Improve, and Control). At the Define stage there was a SIPOC(Supplier-Input-Process-Output-Customer) diagram and the determination of CTQ(Critical to Quality), there are 4 CTQs, namely porosity, hot crack, under cut,distortion. At the Measure stage, the DPMO value is obtained at 20600 and theconversion result to the sigma level is 3.54. At the Analyze stage, the Paretodiagram is to see 80% of product defects in the form of porosity and under cut.Identification of defects that have an effect on reject products is also carried outusing Classification and Regression Tree (CART), namely hot crack, the cause ofthe defect will be identified using Ishikawa Diagram and FMEA (Failure Mode andEffect Analyzes) to determine the priority of defects that require repair based on thehighest Risk Priority Number (RPN) value. the result is a priority for porositydefects with the root cause of failure in the form of welding conditions and humidpipes, as well as hot crack because it is the most influential defect for rejectproducts. At the Improve stage there is a QC PASS decision standardization withthe IF THEN Rule function of the CART and offers suggestions for improvement,namely the welding area cleaning form. The results of this study are an increase inDPMO to 1800 and a sigma level of 3.60 due to the implementation.