Bridging The Gap Between MBSE And Reliability Analysis

SUMMARY & CONCLUSIONSTraditionally, model-based systems engineering (MBSE) and reliability analysis are conducted independently within separate workflows. As system designs are evolving, reliability analysis quickly becomes outdated and misaligned with the system design, invalidating prior risk assessments. Reliability and safety analyses are typically conducted using specialized tools like Windchill Risk and Reliability (R&R) [1] or ReliaSoft [2]; however, the reliability system model must be recreated from the original design specifications. This process is error prone and leads to a disconnect between the system design and reliability analysis processes. Integrating reliability analysis into MBSE ensures that reliability considerations are embedded early in the design process and evolve continuously, in alignment, with the system design.More recently, Risk Analysis and Assessment Modeling Language (RAAML) [3], an extension of the Systems Modeling Language (SysML) [4], was developed to standardize how risk and reliability information is captured and integrated directly into system models. RAAML defines specialized stereotypes and profiles to add reliability-specific constructs to system models including parameters and distributions related to reliability analysis and reliability block diagrams (RBDs). While RAAML supports simple degradation functions such as Weibull and exponential models, it lacks mechanisms to reflect how degradation varies with operational conditions or to incorporate real-world operational data informing equipment-specific degradation behavior.The solution demonstrated in this paper introduces a reliability plug-in for Cameo [5], PRIMERA, which enables advanced reliability analysis for systems of systems (SoS) by supporting probabilistic, context-aware simulation. The plug-in builds on the RAAML standard, capturing reliability-related information according to the RAAML specification. Unlike RAAML, PRIMERA extends reliability simulation from simple distributions to full equipment-specific probabilistic models of degradation and repairs that can account for nuanced context-specific degradation and repair behaviors. It provides reusable reliability templates for individual equipment, which are automatically applied across components and aggregated at subsystem and system levels.We demonstrate the plug-in on a simple but illustrative example of a computer system [6].