Analyzing a Quality Function Deployment Matrix: An Expert System-based Approach to Identify Inconsistencies and Opportunities

Abstract

Identification of a need or a deficiency triggers conceptual system design. The first step in conceptual design is to analyze and translate the need or deficiency into specific qualitative and quantitative customer and design requirements. Design methods such as quality function deployment (QFD), parameter taxonomies and input/output matrices provide a useful framework for this translation. Well defined and unambiguous requirements enhance communications and can potentially reduce the number of 'detours' during subsequent design and development phases. However, imprecision and vagueness characterize the conceptual design phase. To accommodate imprecision, the QFD method and the concept selection methodology, initially proposed by Pugh, have been modified and extended by applying concepts from fuzzy set theory. The extended approach provides a rigorous yet graceful mechanism for dealing with imprecise requirements, priorities and correlations as prerequisites to concept selection. This technical paper presents an expert system-based extension to the fuzzy QFD methodology. Emphasis is on the: (a) identification of strategic market and product opportunities, (b) identification of applied research focus areas and (c) isolation of inconsistencies between customer articulation of functional requirements and the definition of system requirements and parameter target values. An expert system-based parser has been embedded within the fuzzy QFD tool to facilitate strategic product planning, early design decision-making and parameter target setting.