Managing Defect Reduction to Achieve Reliability Growth

The importance of reliability as well as a clear definition for a reliability requirement are presented. The concept of functional performance demonstrated early in product development with a very few hours of testing is presented as a more effective early warning sign that a program is behind schedule with respect to reliability goals. Utilizing the six types of defects encountered in a product development program as predictors of reliability is introduced. Types of failure modes and potential systemic causes of failure are discussed. The prediction relationship between manufacturing and design quality flaws with field failure mode experience is demonstrated. Systems dynamics modeling and empirical tracking of fix effectiveness can be used to predict the learning cycles required to launch a successful product. The concepts of defects per unit (DPU) (Defects in functional performance, product design quality and manufacturing process control); the Laplace estimate of proportion defective in a lot based on small sample size to predict the number of learning cycles required for each system program launch; and the rule of 10:4, where every 10 fixes of failure will lead to the discovery of four new causes of failure in the next set of prototypes, are introduced.