Learning to enhance reliability of electronic systems through effective modeling and risk assessment

Abstract

Now that electronic components have demonstrated high reliability, attention has centered upon enhancing the reliability of electronic systems. We introduce a modeling framework to support decision-making during electronic systems design with a view to enhancing operational reliability. We differentiate our work from those models that seek only to provide reliability predictions. Our premise is that modeling can be used to give a better understanding of the impact of engineering decisions on those factors affecting reliability. Through modeling, the decision-maker is encouraged to reflect upon the consequences of actions to learn how a design might be enhanced. The model formulation and data management processes are described for an assumed evolutionary design process. Bayesian approaches are used to combine data types and sources. Exploratory data analysis identifies those factors affecting operational reliability. Expert knowledge is elicited to assess how these factors might impact upon proposed designs. Statistical inference procedures are used to support an assessment of risks associated with design decisions. Applications to the design of electronic systems for aircraft illustrate the usefulness of the model. On-going research is being conducted to fully evaluate the proposed approach.