A simulation study of predictive maintenance policies and how they impact manufacturing systems

The success and effectiveness of modern lean manufacturing concepts requires robust and highly reliable machinery. In this thesis, we develop several simulation studies to compare the performance of a several manufacturing systems under different maintenance polices. The main focus of this work is to compare traditional time-based maintenance policies with degradation-based predictive maintenance policies that utilize real-time sensory information to assist in decisions regarding maintenance management and component replacement. The simulation studies developed in this thesis demonstrate the benefits of using sensor-based degradation models to predict failure.