In order to test the working accuracy of New Instrument it is necessary you generate control charts, using 1 to possibly several months of data, using this data you find out whether the variability and target values associated with new instrument is within specification limits set by the manufacturing process under consideration.

In this regard you can also compare the historical control chart data if available from the other instrument, but these comparisons would make sense if only you are operating the new instrument at reasonably same instrumental specifications as the old one, and that the historical data is appropriately pooled so that every time instrumental settings for the old instrument has been changed the corresponding control chart data is appropriately pooled, then for comparison purpose you need to look only that subset of historical dataset that matches the current instrumental settings of the new instrument.

You can look into the theory and computation of control charts for more information and insights in this regard.

Dear researcher,

I have replied to your question in the attached file. I hope that to be useful to you.

Best regards.

Ramin Rostamkhani

Debopam Ghosh Thanks for your input! Will look into that material.

Take a look at the Measurement Systems Analysis information on aiag.org. There's good information there for both continuous and discrete measurements.

Mark is pointing you in the right direction. This is a classic MSA opportunity the only twist is that you are using it as a Go/No Go gage which will require an Attribute Agreement Analysis and achieve an >80% agreement level. Although, if you can get continuous (or variable) data from both test systems then it would be prudent to go ahead and perform a Gage R&R (10-parts X 2-systems X 3-replicates) to better understand the repeatability and reproducibility of the two systems in which the two systems are the two appraisers. This would give the best data in the shortest amount of time. The question you would need to answer is what level of error will you accept;

I recommend to perform a Gage R&R study by attributes, I recommend to use an statistical software.

You may apply measurement system analysis (MSA) as proposed by Gerardo Miguel Valenzuela Arellano

I recommend using Minitab for Gage R&R study.

If you need any help with this , I can support.