Hidden Costs of Decomposition: The Need for Fit between Technical and Organizational Architectures

George Washington University, Washington DC

Investigators

Abstract

Engineered systems are becoming increasingly complex, and as a result, they are more difficult to design. Effective design requires quickly bringing diverse expertise to bear on each iteration of the design. These two goals -- faster design cycles and better cross-disciplinary integration -- are enabled by model-based concurrent design approaches. These approaches bring multiple experts together into a single room to create a design in a series of short design sessions, and they are widely used in certain fields, such as spacecraft design. These approaches have many proven advantages, but also potentially serious drawbacks due to the level of decomposition they impose on the design problem. For example, many assumptions are made in order to decompose the design tasks into modules for each team member, and these assumptions may not be valid for all design problems. This award supports fundamental research to generate knowledge about how the "fit" between the technical decomposition (the problem being solved) and the organizational decomposition (people and tasks) impacts the quality of the design process. The results will enable design organizations to make better informed choices about how to deploy concurrent design approaches, such as when these tools apply and when they do not, and how to use them well even when the technical and organizational decompositions are slightly mismatched. The resultant gains in design efficiency would have implications for many sectors of the economy that depend on engineering design and innovation. In addition, this research involves several disciplines including aerospace and systems engineering, organization design, and qualitative research methods; this approach will foster multi-disciplinary research and engineering education. The objective of this research is to generate knowledge about how the lack of "fit" between a design organization's imposed decomposition and the problem's natural decomposition affect the speed and characteristics of the design process. The research will focus on understanding the nature of "fit," developing a framework for measuring it, identifying the types of problems that arise in the design process due to this lack of "fit," and explaining the mechanisms by which they arise. Current research in this area is limited by measurement challenges and limited empirical evidence. This research will address these limitations through an empirical study of the concurrent design team at the National Aeronautics and Space Administration's Jet Propulsion Laboratory. The research will be conducted through an in-depth observational process study of multiple concurrent design sessions. The design process will be observed, its key characteristics measured, and the technical and organizational decompositions will be recorded, so that the impact of the decompositions' "fit" on the design process characteristics can be evaluated and its drivers explained.