Application of Value Focused Thinking and Fuzzy Systems to Assess System Architecture

Abstract

Abstract Since a majority of resources are obligated during the design phase of a system lifecycle, critical assessment of candidate functional and system architectures is vital to identify optimal architectures before proceeding to subsequent lifecycle phases. Common challenges associated with generation and evaluation of system functional architectures include search of the expansive design space and assessment of key performance attributes that are particularly “fuzzy” and qualitative in early architecture development. Several assessment approaches have been presented in the literature to address the assessment challenge to include Quality Function Deployment (QFD), Analytical Hierarchy Process (AHP), Value-Focused Thinking (VFT), and fuzzy logic. In this research we combine the use of value functions and fuzzy assessment to assess a functional and system architecture. There are several benefits of a methodology that combines value-focused thinking and fuzzy assessment. A distinct advantage of the methodology presented is the explicit inclusion of the customer in the assessment process through validation of the TPM value functions Involving the customer in TPM value function development and validation ensures the customer has direct input regarding the TPMs and their associated value across the range of discourse The methodology presented is flexible enough to assess architectures early in the process when things are “fuzzy” as well as later when subsystem and component performance are well defined. The methodology can also be used to analyze and assess impacts of interface changes within the system architecture. . The methodology is domain independent and can be coupled with executable models linked to scenarios. The assessment methodology is applied to the architecture for a soldier knowledge acquisition system for which the key performance attributes are affordability, performance, flexibility, updateability, and availability.