User:Jfritts-dsi/Testability

Testability, is a standard for defining the requirements of and analysing the capability of system diagnostics. As a requirement, Testability is used to define a contractal requirement that the supplier must meet when designing the diagnostic capability of a system. As a method of analysis, Testability is used by engineers involved in developing the diagnostics of a system to analysis where a system needs to be improved in its capabability to diagnose faults. Statistically, this involves calculating the detectable and isolatatable faults within a system. Additionally, when coupled with Failure Mode and Effect Analysis (FMEA) and Failure Mode, Effect and Criticality Analysis (FMECA) this process can be used to identify where the system design needs to change in order to improve the diagnostic capability of a system to meet the Cost, Maintainability, Availability and Safety requirements of a system.

FD/FI Calculations
Fault Detection and Fault Isolation (FD/FI) calculations involve taking the sum of the failure rates of all detected or isolated faults within a system and dividing by the aggregate or total failure rate of the entire system.

The Fault Detection statistic (FD) defines the percentage of the total possible system faults that will be detected by the diagnostics that have been designed for a system.

The Fault Isolation statistic is generally subdivided by the size of the group of suspected items that are indicted by the diagnostics.

Single Fault vs Multiple Fault Analysis
Single fault testability anaylsis assumes that one an only one fault can occur in a system at any given time. Multiple fault or common cause testability analysis assume that multiple faults may be diagnosed in a system. Traditional testability requirements are written in terms of single fault analysis. Generally, single fault analysis yields higher testability numbers than multiple fault analysis. A supplier attempting to meet their requirement using multiple fault anaylsis would generally be required to extend additional diagnostic engineering effort to meet the stated contractial testability requirement. However, single fault anaylsis would be incorrect in a system were multiple faults can occur prior to diagnosis of the faults. In an automatically tested system, single fault diagnosis may correctly diagnosis multiple faults if the faults are logged as they occur.

History
Testability was orgininally pioneered by Ralph De Paul, Jr. in the 1960's based upon diagnostic anaylsis methods conceived in the 1950's. Ralph De Paul began conceptializing an approach to system diagnosis after several friends were killed in the Korean war due to faulty equipment. At that time, Ralph De Paul found that existing systems were generally only developed to serve a function and not to diagnose faults within the system.

De Paul begin conceptionalize a logical approach to modeling a system which could be used to analysis the effectiveness of the diagnostic tools and capabilities of a system. His "Logic Modeling" allowed for all functions of a system under test to be mapped to the events, wether testable or not, that depend upon the proper operation of a system. This system of Logic Modeling eventually became known as Functional Dependency Modeling.

In 1974, the dependency modeling concept was incorporated into MIL-M-24100B, a military manual that documents a Logic Modeling process via the construction of a maintenance dependency chart as part of the development of Functionally Oriented Maintenance Manuals.

In 1985, William Keiner (US Navy) working for the US DOD developed MIL-STD-2165 as a standard for testability analysis. This standard, which includes the original logic modeling concept of MIL-M-24100B also document additional methods for analysis and documenting the diagnostic capability of a system. In 1995, MIL-HDBK-2165 replaced MIL-STD-2165 as the standard for Testability analysis.

In 1993, IEEE recognized Ralph De Paul by posthumously awarding him the John Slattery award for his contribution to diagnostic engineering.

In 2004, IEEE-STD-1522 documented the Design For Testability process.