Quality and reliability monitoring in multi-stage processes

Abstract

In multi-stage processes, downstream variables are influenced by events in upstream stages. One example is the assembly process of a computing system, which relies on components produced at various locations over wide timeframes. The quality and reliability (Q&R) assurance process involves models and data pertaining to the components, assembly process itself, and product experience in the field. In this paper, we focus on the process of monitoring Q&R, diagnostics and root cause analysis based on the so-called timeslides: data orderings corresponding to passage of components, sub-assemblies and the product itself through various stages. Given that the product and component Q&R can be impacted by numerous stages, the problem acquires a combinatorial character and complexity. We discuss the design of search engines for selecting timeslides that merit investigation.