Premium
3.1.3 Applying quantitative methods for architecture design of embedded automotive systems
Author(s) -
Larses Ola
Publication year - 2005
Publication title -
incose international symposium
Language(s) - English
Resource type - Journals
ISSN - 2334-5837
DOI - 10.1002/j.2334-5837.2005.tb00678.x
Subject(s) - design structure matrix , automotive industry , reuse , computer science , modularity (biology) , modular design , systems engineering , architecture , balanced scorecard , dependency (uml) , abstraction , process (computing) , manufacturing engineering , software engineering , engineering , process management , art , biology , visual arts , genetics , operating system , aerospace engineering , waste management , philosophy , epistemology
Architecture design is often referred to as an art, performed in the conceptual stages of a design process. However, it can be supported by quantitative methods. At the highest level of abstraction a goal oriented approach is applicable. This can be realized by a keyfigure based methodology similar to the management directed Balanced Scorecard. The automotive industry requires that a system architecture is optimised not only for a single product, but for reuse over a range of products, and also for reuse over time with continuous improvements. To achieve these goals the product should be modular. Keyfigures are useful for managing trade‐offs but do not easily describe modularity. For this purpose DSM (Design/Dependency structure matrix) and cluster analysis is more useful. A combination of the methods is proposed for design of automotive control system architectures. The proposed method has been applied and evaluated in architectural work at Scania CV.