M
Hi. This is my first thread on Elsmar and I hope I have posted it in the correct area. My question is as follows:
I work in an FDA regulated industry and am responsible for validating our measurement equipment used to verify our assembly processes. I should note most of our measurement equipment is custom developed. In our validation efforts, we normally do Gage R&R and use AIAG recommendations for our specifications (less than 10% for our critical quality tools and less than 30% for the others). This works for us for most measurement systems.
There are instances where because of the design of the measurement equipment, the operator will absolutely have no affect on the outcome of the results. As such the reproducibility is 0%. We would like to be smarter with our test designs for these systems and not use multiple operators, multiple trials when we know going into it they will have no influence on the outcome of the results. The question then is what is the best way to do a repeatability study for this situation...and more importantly what specifications should be set. Because we are in a highly regulated industry, we will need to reference the source of the specifications (like we have for GR&R and AIAG) and justify them.
I have probably missed a few key details, so I will be happy to fill them in as others point them out to me. Thanks in advance for the help, and I know I am not alone in praising this forum.
I work in an FDA regulated industry and am responsible for validating our measurement equipment used to verify our assembly processes. I should note most of our measurement equipment is custom developed. In our validation efforts, we normally do Gage R&R and use AIAG recommendations for our specifications (less than 10% for our critical quality tools and less than 30% for the others). This works for us for most measurement systems.
There are instances where because of the design of the measurement equipment, the operator will absolutely have no affect on the outcome of the results. As such the reproducibility is 0%. We would like to be smarter with our test designs for these systems and not use multiple operators, multiple trials when we know going into it they will have no influence on the outcome of the results. The question then is what is the best way to do a repeatability study for this situation...and more importantly what specifications should be set. Because we are in a highly regulated industry, we will need to reference the source of the specifications (like we have for GR&R and AIAG) and justify them.
I have probably missed a few key details, so I will be happy to fill them in as others point them out to me. Thanks in advance for the help, and I know I am not alone in praising this forum.
