Metal frame for actuator manufacturing process improvement using data mining techniques

• Main Authors: Wimalin Laosiritaworn, Pongsak Holimchayachotikul
• Format: Journal
• Published: 2018
• Subjects: Biochemistry, Genetics and Molecular Biology; Chemistry; Materials Science; Mathematics; Physics and Astronomy
• Online Access: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=78249264467&origin=inward; http://cmuir.cmu.ac.th/jspui/handle/6653943832/50538
• Institution: Chiang Mai University

Hard disk drive manufacturing has recently played an important role in Thailand's economy, with the number of hard disk drives produced increasing rapidly. The case study company is a manufacturer of metal frames for actuators; one important part in hard the disk drive head. More than 300 computer numerical control (CNC) machines are used to fabricate the contour of the metal frames. During production, random sample are taken from the process so as to be inspected within the quality control (QC) department. If samples show a tendency to be out of specification, the machines that produced them have to be adjusted or even shutdown. Large amounts of data are produced during this procedure, and due to the large number of samples to be inspected, a queue forms in the QC department. If the machine producing the defect is inspected late, the damage caused might be large. This paper proposes the application of data mining tools in order to cluster the machines into groups. After that, the inspection order can be arranged so that the samples from the machines that have the highest tendency to produce a defect can be inspected early. In this study, actual data was used from the production process in the case study company to demonstrate the proposed method. The results suggest that the proposed method helps to detect faulty machines earlier hence reducing the number of defects found in the production line.