Unplanned downtime in industrial machinery costs an estimated $50 billion USD per year in the manufacturing sector alone, due to broken parts, additional maintenance expenses, and lost productivity [1]. This highlights the need for more advanced condition monitoring approaches, which may be enabled through the use of models that can predict and prevent potential issues before they occur. While a model that represents the original, or baseline, state of the system is useful for fault detection and diagnostics, updating the model as faults occur is an even more desirable goal. Integrating faults into a systems model can provide us with detailed information such as the progression of a fault over time, its effect on other components, and how and when a failure in the system may occur. In this thesis, a combination of physics-based and data-driven modeling techniques are applied to the problem of updating a systems model as faults initiate, focusing on a class of faults that has received relatively little attention in literature. Faults can manifest either as changes in system parameter values, or as changes to the mathematical structure of the system, the latter of which we refer to as emerging dynamics. These faults introduce additional dynamics which are not present in the baseline system and may involve invalidation of assumed boundary conditions. While there are many established methods for updating model parameters over time [10, 43], considerably less work deals with integrating emerging dynamics into a model. This thesis work aims to address this gap through the development of a comprehensive framework for integrating emerging dynamics behaviours into a systems model to enable predictive condition monitoring analyses. Key framework components include methodologies which have been developed for the detection, identification, and modeling of emerging dynamics. Each involved addressing fundamental theoretical challenges and associated gaps in the literature. The effectiveness of the proposed framework is demonstrated using a test rig. A selected emerging dynamics fault is modeled in an experimental setup, detected and identified as it develops in a rotating machinery system, and finally, integrated into the baseline model.