A system testability 'top-down' apportionment method

The authors present a top-down approach to specifying testability requirements which involves system-level analyses, tradeoffs, and allocation to lower hardware indenture levels. This approach provides guidelines, algorithms, and procedures for computing, assessing, and allocating testability within a new system design. Testability figures of merit (TFOM) used to describe and quantify testability, as applied to a given system, in precise and measurable engineering terms are presented. Testability allocation methods (TAM) to apportion system testability requirements cost effectively through lower levels of indenture to the replaceable unit level and generate subsystem-level requirements are also given. This allocation process starts with the overall testability requirements of a system and prescribes the distribution of these requirements among the various units constituting the system. The TAM problem is formulated as an optimization problem and solved using the augmented Lagrangian method. The complex interactions between the TFOMs and system performance (reliability, availability, maintainability, and life-cycle cost) are derived using analytical, heuristic, experimental, and historical data. A special case where BIT (built-in test) is the resource to be allocated leads to the top-down BIT prioritization.<
>