(414-2368-1-PB)PRODUCT QUALITY-BASED METHODOLOGY FOR MACHINE FAILURE ANALYSIS AND PREDICTION

摘要

The need for maintenance management has increased in order to efficiently manage maintenance activities that are vital to keep production running competitively. The problem of preventive maintenance management and control has been addressed considering a quality-integrated methodology. Failure and repair behavior of a machine are modeled statistically thorough distribution fitting, then a regression model is developed to predict the next failure according to product defects. The integration of the distribution of the time between failures and the regression model identifies the probability of a failure to occur at time t. The results of implementing the methodology on a production line that belongs to a local company are discussed. Further researches may address the problem at the component level to be able to predict and plan the failure of a component for more accurate maintenance planning and spare parts inventory management.