Trade-offs among system architecture modularity criteria

Roozbeh Sanai, Kevin Otto, Kristin Wood, Katja Hölttä-Otto

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

1 Citation (Scopus)


Modularity is a key design research area of the Nordic countries, with a history of intensive work over the last twenty years. In the last decade, there has been considerable research in product modularity, measuring the level of modularity and various procedures for searching for ideal modular architectures. Different manual heuristics and computer clustering algorithms have been developed to search ideal modular architectures by optimizing a modularity metric. However, the different criteria can be in conflict and improving one criteria may drive another infeasible, without an appropriate compromising effect or insights into decoupling conflicts / contradictions. We pose here the research question as to how to visualize product architectural design criteria and trade-offs in the early conceptual or configurational design phase. We analyze correlations between different system architecture modularity criteria provided in the research field, namely the intra-cluster, extra-cluster costs, number of modules and the variance in the size of modules. We demonstrate that these criteria trade-off with each other, and therefore one cannot be improved without affecting the other. We also show that several of these metrics are directly correlated; for example, the variance in the size of modules can be controlled through the intra-cluster cost. Finally we observe that, although typically proposed agglomerative or divisive hierarchical clustering algorithms might able to obtain optimal architecture when only extra-cluster cost is of concern, such algorithms are not able to find optimal cluster when both extra-cluster and intra-cluster cost are matter of interest. Overall, as minimizing intra-cluster cost is in charge of proper sizing of modules, well-sized modules cannot be obtained through conventional algorithms such as K-means clustering or similar.

Original languageEnglish
Title of host publicationProceedings of NordDesign, NordDesign 2016
ISBN (Electronic)9781904670803
Publication statusPublished - 2016
MoE publication typeA4 Article in a conference publication
EventNordDesign Conference - NTNU, Espoo, Finland
Duration: 27 Aug 201429 Aug 2014
Conference number: 10


ConferenceNordDesign Conference
Abbreviated titleNorddesign
Internet address


  • Clustering algorithms
  • Product modularity
  • Trade-off analysis

Fingerprint Dive into the research topics of 'Trade-offs among system architecture modularity criteria'. Together they form a unique fingerprint.

Cite this