Reliability Assessment with Varying Operating Conditions

Abstract

Product reliability is a systems property that cannot be attributed to a single component. It is in many cases related to interactions between components, or to interactions between humans and the technical system. Product functionality includes both technical functions, like structural integrity and interactive functions, like ergonomics. Reliability assessments in the early stages of the development process are valuable, since design changes cost significantly less if made early. System reliability tests can only be made towards the end of the development process, but early estimates can be based on test data from component tests and function tests. Operating conditions often vary between component tests and system tests. Therefore, reliability assessments where data from one situation is used to predict reliability in a different situation must take this variation into account. We investigate how this can be done for both technical and interactive functions. The study is made in the context of system reliability for heavy trucks, where both technical functions and interactive functions affect product reliability. Two cases have been assembled from test data, one concerning a component on a truck cab, the other an interactive function of a truck. Two reliability estimation methods have been evaluated to investigate if the methods can be used for interactive functions as well as for technical functions. We conclude that a method for reliability estimation of interactive functions must be able to model increased uncertainty due to intrapersonal variation.