Calibration Software Validation - Are there any Testing Tools available?

[Andrews]

We have recently developed a inhouse software for the standards room / gauge room which takes care of calibration activities..

Now how should we validate this software to ensure that this software works as required by us ?

Are there any testing tools available?

 

[Al Dyer]

How about sending a sample of gages to a certified lab for their results and match them to your results? Doing it 1 time might not work, but if you can show over time that your system works you are in business. Even if you had bought a canned gage program you would probably still want to verify its results and equations.

Al...

 

[howste]

What kind of features does your software have? If all it is doing is storing records, then just verify that the records are retrievable, etc. If it is doing statistical calculations for you, verify through some other (independent) method that the results are correct, and you should have it. You can always download a trial copy of something like PQ Systems' GagePack and compare the results.

 

[Al Dyer]

Let's see,

How do we verify/validate that the equations and graphs in the software we buy are accurate and precise???

By taking the word of the salesman? NO
By matching against like software? NO
By testing against our own results? NO
By testing against an outside source like a RAB approved calibration lab? YES

I am not even the devil's advocate in this, it is common sense.

Al...

 

[howste]

I think we need to hear what the functions of the software are. I am assuming that if it is doing calculations, it is doing things like R&R studies, linearity plots, stability charts, and bias studies. These things are all simply mathematical functions! Are RAB approved calibration labs the only ones with the demonstrated ability to do math? I think not. I have never received a calibration certificate that includes any of the above statistical information. The fact of the matter is, I can verify function of software using completely fictional gage data by verifying the results using math!

 

[Andrews]

Hi everyone,

Our software not only stores the records of all calibration done but also cautions us when the due date for calibration is near.It also compares the actual size (that is entered ) against the specification (that is already entered in the database) and tells us if the gauge is OK or NOT OK.

 

[Atul Khandekar]

Andrews,

All these functions can be easily verified manually. Even if the software does some statistical calculations, IMHO, you can verify by doing them manually.

 

[howste]

Atul is right. Just put some data in and try out a variety of scenarios to verify that all functions perform as intended.

 

[Holly21cn - 2009]

I came across the same issue but our software is used to measure some paramters of raw material used in our product(Such as contour). The software is a programm designed by an engineer and no one can verify if the measurement is truely reflecting the parameter we expect and if the measurement is acurate. Engineering people claims that we shall just trust the capability and expertise of the engineer and shall not go through any verification. I don't agree. I need to find a effective way to verify. My idea is to ask the supplier to provide some samples with measurement datas and compare these data with our measurements, if the measurement system of supplier can be acceptable. The gage in our supplier site also needs to be designed purposely.

What is your recommendations?
Thanks a lot.
Holly