Statistical measurment analysis of automated test systems

The innovation of automated test system technology necessitates continued process improvements to legacy test program software. These improvements to the test program software require verification and validation of their performance to support functional and diagnostic testing of units-under-test (UUT). Since measurement uncertainties in automated test systems (ATS) should not impact the pass/fail criteria when testing a UUT, statistical measurement analysis can be used to evaluate ATS measurement uncertainty. This paper discusses several statistical methods that can be applied to evaluate ATS measurement capability. Paired observations between two ATS were analyzed to evaluate preliminary performance measurement capability of the reconfigurable-transportable consolidated automated support system (RTCASS) to the consolidated automated support system (CASS). Utilizing the paired t-test a comparison was made with a 95% confidence interval. In addition, ANOVA gauge repeatability and reproducibility is demonstrated for evaluation of test program set performance between RTCASS and CASS, the UUT's and operators. A method for analysis of this system is discussed. Statistical tests were based on the evaluation of preliminary test program set (TPS) measurement data between RTCASS and CASS.