Metrics and methods for complexity analysis: Identifying driving factors of structural complexity in modular vehicle platforms

Abstract

This thesis explores how module and interface variation can be used to quantify architectural complexity in modular vehicle platforms at Volvo Group TTI. By combining structural complexity theory with entropy-based assessment of module variant complexity, a quantitative methodology for architectural complexity assessment was developed. The proposed metric evaluates architectural complexity based on module connectivity, module and interface variation, and interdependencies between modules. Shannon entropy was adapted to assess module variant complexity. In addition, a data-processing pipeline was developed to enable large-scale analysis of industrial platform data. The results show that the proposed metric responds consistently to structural changes in the platform architecture and captures important characteristics associated with architectural complexity such as connectivity and interdependency. The framework therefore enables systematic and data-driven comparisons of modular platforms. The proposed metric should be used in combination with other metrics and methods to fully capture architectural complexity and support balanced and robust design decisions. The metric is designed for hardware platforms. Future studies could improve measurement accuracy of module variant complexity by analysing the physical structure of the module variants in detail. Additionally, future work could develop a systematic method for selecting interface complexity values.