Abstract
Condition-based maintenance (CBM) is a valuable tool to any industry looking to improve productivity, product quality and overall effectiveness of critical systems. Based on the University of South Carolina’s (USC) experience and research partnerships, USC’s vision of a smart predictive system involves a comprehensive integrated methodology that involves a two prong approach: i) CBM analysis involving gathering experimental data from mechanical systems on test stands; ii) theoretical analysis involving modeling and simulation. A theoretical framework based on component, subsystem, and integrated system models is presented to understand the physics of the failure modes. The two approaches provide complementary information that can be used to enhance CBM. The correlation of these two data fusion approaches will allow the implementation of predictive tools to capture the condition of a component, subsystem and system to maximize useful life and minimize cost and risk. In addition, the predictive tools can be used to inform the next generation of system design.