Towards establishing foundations for new classes of reliability problems concerning strongly innovative products

This paper proposes a conceptual framework to distinguish between different classes of reliability problems encountered in strongly innovative products. Next to the conventional (hardware and software) problems, new classes of failures have emerged with a wide range of often strongly related definitions, such as: soft failures, "No Fault Found" failures, "Fault Not Found" failures, "Cause Not Found" failures, nuisance failures. The fact that these new classes of failures do not have precise and orthogonal definitions, leads to difficulties in failure identification and classification. A list of dimensions is proposed to identify and classify failures in an unambiguous manner. Contribution of this research is two-fold: From the academic point of view, it encourages precise reasoning about the emerging failure classes in the general context of reliability problems, as well as forming grounds for consistent use of terminology within the community. From the industrial point of view, it potentially provides more accurate and easier detection of failures, hence facilitating more effective and efficient ways to handle them.