Modeling design quality

Abstract

Several qualitative and quantitative perspectives on design quality are presented. An analogy between testing and design verification is demonstrated, leading towards a quantitative metric for "design incorrectness" equivalent to the "defect level" in digital testing. The focus is on the design process and the probability for error insertion, rather than detecting existing stuck-line faults. We demonstrate that pursuing this avenue will lead to practical means of assessing and predicting the quality of designs, as well as designers, in real-life product development.<>