P
Hi - apologies if this is already on another thread.
Have been asked to perform GR&R on a test system that produces many hundreds of lines of results.
Interfacing to the unit under test is via connectors - so no mechanical set up is required.
The results are split between absolute, attribute and continuous
The appraiser has very little influence on the results - once started the the system runs automatically, the appraiser's only input is to confirm display patterns are produced on various monitors (i.e. is the display as per the test pattern (Y/N)?)
The majority of the remaining results are absolute - i.e. the system expects a specific value to be returned (i.e. hex / binary data) - test system automatically runs the test and captures the result - no appraiser involvement.
A minority of tests are continuous (voltage / current / time delay) - again the system captures the results with no appraiser involvement.
Are there any recommendations as to best method of GR&R?
Appreciate that 'ANOVA GR&R' (continous) and 'Attribute Agreement GR&R' (attribute / absolute?) can be performed - but with the appraiser having little influence what can be gained?
Thanks in advance for any feedback.
Paul
Have been asked to perform GR&R on a test system that produces many hundreds of lines of results.
Interfacing to the unit under test is via connectors - so no mechanical set up is required.
The results are split between absolute, attribute and continuous
The appraiser has very little influence on the results - once started the the system runs automatically, the appraiser's only input is to confirm display patterns are produced on various monitors (i.e. is the display as per the test pattern (Y/N)?)
The majority of the remaining results are absolute - i.e. the system expects a specific value to be returned (i.e. hex / binary data) - test system automatically runs the test and captures the result - no appraiser involvement.
A minority of tests are continuous (voltage / current / time delay) - again the system captures the results with no appraiser involvement.
Are there any recommendations as to best method of GR&R?
Appreciate that 'ANOVA GR&R' (continous) and 'Attribute Agreement GR&R' (attribute / absolute?) can be performed - but with the appraiser having little influence what can be gained?
Thanks in advance for any feedback.
Paul
. Should the GR&R not be done against the system as if it were the operator to determine the systems repeatability?