The process of performing a Probabilistic Risk Assessment (PRA) is the systematic, scientific way to create a model that represents complex system behavior and associated risk. The outcomes from these PRA models provide system- and scenario-level information about a variety of outcomes. This information that is produced helps to answer questions targeted to ”what can go wrong” by identifying hazards that may lead to off-normal conditions or initiating events. As such, a scenario-based model for each upset condition that describes how the facility will respond to the off-normal condition is created. Further defining the model, the PRA analyst determines likelihoods for the probabilistic elements of the model, including the frequency of events and the failure probabilities of systems, structures, and components. The typical results of these models include lists of ways in which the system can fail and importance measures that provide insights into component-level behavior. Recently, there have been advances in reliability and risk analysis software tools (e.g., the EMRALD open source software produced by the Idaho National Laboratory) that include dynamic elements (both timing and phenomenology). However, obtaining importance measures for these types of dynamic models has been limited. In this presentation, we address the extension of traditional importance measures to the approach of dynamic modeling. We consider the challenge of estimating the importance of components and parameters in a dynamic simulation of complex systems using a computationally-efficient, data-driven methodology. As such, we are able to bridge the gap from traditional (i.e., based upon static models) importance measures into the realm of dynamic reliability and risk models. Our approach works by defining the importance measures in such a way that they draw from the detailed output of a dynamic simulation exploiting information that is typically hidden when focusing solely on the probabilistic-level data (e.g., component failure probabilities, system-level failure probability). By incorporating detailed observable information such as failure times and component operational characteristics, additional dimensions of decision making are made available to system designers and operators that allow a focus on the margin to failure instead of just probability of failure as a metric. The overall goal of our work is to create a scalable and implementable approach to enrich the insights from a time- and physics-informed simulation with explanations as to what are the drivers of the system behavior at a component level of behavior.