Learning Framework For Maturing Architecture Design Decisions For Evolving Complex SoS

Abstract

Architecting a complex System-of-Systems (SoS) poses significant challenges due to the uncertainty and perceptions associated with understanding the implications of constituent system’s architecture design decisions at SoS level. Due to significant knowledge gaps, architects may find it difficult to uncover the ramifications of a specific decision on various Measures-of-Effectiveness (MOEs) and emergent behavior of the SoS. Subsequently, for complex SoS, learning cycles maybe experienced on the architecture design decisions. As the SoS evolves, these experiential learnings need to be factored into the uncertainty assessments of decisions and the impacted SoS MOEs, while evaluating and deciding on a specific decision. This paper proposes a knowledge based decision learning framework that factors the learning cycles experienced into the uncertainty associated with the decisions and impacted SoS MOEs. The framework takes into consideration, through an architectural knowledge base, multiple knowledge dimensions such as the attributes of the architecture design decision and the feedback loops experienced, in tandem with the complexity attributes and the knowledge gaps associated with the decision.