A two-stage equipment predictive maintenance framework for high-performance manufacturing systems

It has been a long interest from researchers to have an effective approach optimizing maintenance scheduling due to the large budgetary item factories spent on equipment maintenance. Since nowadays large scale of machinery log data is already collected and maintained in most manufacturing plants, it is feasible to extract useful information from this database and predict equipment failure utilizing intelligent and statistical techniques. In order to cope with the high complexity raised in predicting equipment failure, a two-stage equipment predictive maintenance framework based on a systematic integration of biological inspired algorithms and statistical analysis considering each advantages and disadvantages has been proposed and developed. Evaluation and development of the genetic algorithm, neural network, and multiple regression forecasting components in this framework for predicting equipment failure is presented. Through the case study on a wafer fabrication plant in a semiconductor company, the feasibility and effectiveness of the proposed system is demonstrated.