Systems and methods for detecting and predicting faults in an industrial process automation system use trend data to forecast alerts and allow action to be taken before a problem occurs. The systems and methods provide fault/failure predictions that improve over time as more empirical data is collected for a related set of system components. The systems and methods may identify relationships among the components of a process automation system; identify and collect changes to system configuration; identify and collect data to inform reliability and predictive models; develop a domain-specific predictive model for one or more components that allows for component-based failure or degradation prediction; develop a system-predictive model that leverages reliability and criticality relationships, component-based predictions and operating parameters to predict the health of a part of or the entire process automation system; deliver a prioritized alert system; and identify root-cause failures of a component.