Hard Disk Drive failure mode prediction based on industrial standard using decision tree learning

Quality and reliability are extremely important for Hard Disk Drive (HDD) manufactory, which increasing upon the expansion capacity of HDD. The production process contains many tests in order to evaluate quality and efficiency which resulting high amount of complex data. Some HDDs will be randomly selected from each lot to test with various environments simulation. Due to complexity of tests and higher capacity of HDD manufacturer according to increasing demand from the consumer, for process test, it might take a long time. However, some failing HDD might be found at the customers test process or end users. This paper introduced the failure prediction using decision tree learning and the procedure of data collection and preparation. The main purpose of the study is to eliminate the simulation of the various environments in the tests which reducing time and increasing speed and efficiency of the overall analysis process prior to deliver the products to the customers. Moreover, Corrective Action: CA, the action that will solve and prevent the cause of the failure, can be verified faster. This study can improve quality and reliability of the production.