Towards a Theory of Complicatedness: Framework for Complex Systems Analysis and Design

Abstract

Global, dynamic, and competitive business environment has increased the complexity in product, service, operational processes and human side. Much engineering effort goes into reducing systems complexity. We argue that the real issue is reducing complicatedness. This is an important distinction. Complexity can be a desirable property of systems provided it is architected complexity that reduces complicatedness. Complexity and complicatedness are not synonyms. Complexity is an inherent property of systems; complicatedness is a derived function of complexity. We introduce the notion of complicatedness of complex systems, present equations for each and show they are separate and distinct properties. To make these ideas actionable, we present a design methodology to address complicatedness. We show examples and discuss how our equations reflect the fundamental behavior of complex systems and how our equations are consistent with our intuition and system design experience. We discuss validation experiments with global firms and address potential areas for further research. We close with a discussion of the implications for systems design engineers. As engineers, we believe our strongest contributions are to the analysis, design, and managerial practice of complex systems analysis and design. We illustrate the difference between complexity and complicatedness. Relative to a manual transmission, a car’s automatic transmission has more parts and more intricate linkages. It is more complex. To drivers, it is unquestionably less complicated, but to mechanics who have to fix it, it is more complicated. This illustrates a fundamental fact about systems; decision units act on systems to manage their behavior. Complexity is an inherent property of systems. Complicatedness is a derived property that characterizes an execution unit’s ability to manage a complex system. A system of complexity level, Ca, may present different degrees of complicatedness, K, to distinct execution units E and F; KE=KE(Ca) ≠KF=KF(Ca). We summarize relevant literature on systems complexity in Figure 1. Columns [1] to [15] are keyed to the references. Rows identify key areas of research results; e.g., metrics, complicatedness, etc. We make four observations about the locus of results. One, there is a dearth of quantitative frameworks or metrics. There is no research on complicatedness and complexity as distinct properties of systems. Two, research seems to cluster around engineering management and physical products. The focus is on modularization and interactions with a bias to linear systems and qualitative metrics. Three, there are efforts on methodologies and tools, but theory, foundations and software have a demonstrably lesser presence. Ferdinand’s work in software systems complexity is a happy exception [1]. It is analytical, rigorous and elegant. Three, services and enterprise solutions are barely addressed. This is a serious omission given the high proportion of services in industrialized economies. Fourth, although layering of abstract systems and reintegration have a long history; the literature is skewed to decomposition rather than integration.