Assessing the informative value of complexity metrics within Design Structure Matrices in early development phases of complex systems

Design Structure Matrices (DSMs) are a widely extended and useful tool for structuring and analyzing complex systems. Though this is a simple yet quite powerful tool, problems arise in using them, especially during early development phases due to uncertain information and frequent changes within the original system. In this paper, the informative value of selected complexity metrics of DSMs for early development phases is examined. We detect which and how complexity metrics are affected by certain degrees of uncertainty in relation to the fuzziness of the underlying information. Thereby we derive statements about the robustness and applicability of these complexity metrics in early development phases. Additionally we show how the complexity metrics are related to each other in order to simplify the analysis of complex systems. This can be crucial for project success as it already supports the detection and management of complex systems in the early phases.