10 identical machines, can I conduct IOQ in one machine and do verification only to the remaining ones?

Hello,
I will appreciate your feedback for the following scenario;
An etcher laser is being replaced in 10 machines. All ten machine lasers will be replaced by a better one (different technology). An IOQ is completed for one machine replacement, but only a verification is made to the other nine. Is this reliable or statistically valid?

Thank you, Javier

Hello Javier,

For myself I would say this approach is acceptable, but I welcome others to input their views.

Like many things in manufacturing quality, the answer is a question of risk. Your example is a etcher laser. Forgive me if I am oversimplifying your situation. If the laser etches, for the sake of illustration, the part number on a manufactured part, and one machine prints 10 mm tall lettering and another prints 10.1 mm tall, I would accept that variability in the application as negligible risk. However if you are etching the lines on a measuring device, I expect the difference between 10 and 10.1 mm to be completely unacceptable. A less risky solution would be to assume the 10 machines are the same, but plan to measure a confirmation lot from each.

Thank you all! We understand this is the "Umbrella Approach", and I probably should expand my initial question. Since, this is a common practice, what type of documents & test are part of the verification of the remaining 9 machines to ensure traceability, variability (as mentioned above, etc is valid and acceptable? I believe that a change control, listing all 9 machines and referring the machine 1 (IOQ) will assure traceability, but what about tests? ANOVA?, AQLs? among all 10 machines? For instance, how we ensure that installation, operation and performance is ok, if I only conducted a verification?

Javier,

It is difficult to answer your questions explicitly without knowing your customer requirements.

Overall, process realization involves activities (and records from them) to ensure equipment is performing as expected. That can be first piece inspection, ensuring the equipment has been entered into the PM system and whatever else gets requested - there is n one set protocol. Depending on the output sensitivity, especially if there is an inspection component, ANOVA could be called for but I would not expect it in every case.

Jen Kirley said:

Javier,

It is difficult to answer your questions explicitly without knowing your customer requirements.

Overall, process realization involves activities (and records from them) to ensure equipment is performing as expected. That can be first piece inspection, ensuring the equipment has been entered into the PM system and whatever else gets requested - there is n one set protocol. Depending on the output sensitivity, especially if there is an inspection component, ANOVA could be called for but I would not expect it in every case.

Click to expand...

Thank you!

Javier Guan said:

Hello,
I will appreciate your feedback for the following scenario;
An etcher laser is being replaced in 10 machines. All ten machine lasers will be replaced by a better one (different technology). An IOQ is completed for one machine replacement, but only a verification is made to the other nine. Is this reliable or statistically valid?

Thank you, Javier

Click to expand...

Overall, this is established through appropriate risk assessment; considering the impact on the equipment, process and product; and comparative results from the performance characteristics.
hope this helps.,

....there are three considerations here.,
1. equipment compliance;- all thee equipment has to be independently commissioned; protocol could be the same, but the execution has to be established specifically and separately for each equipment

2. equipment performance - it's only pertinent to establish the performance with the integration/replacement of laser with the existing setup; hence protocol could be common, and again the data/results would be expected for each equipment. ( calibration or performance verification )

3. product/process specific requirements;- again the validation has to be re-established to determine the criteria for the acceptance ranges, nature and levels of defects etc. Depending upon the scope and recipe used for each of product, one could look into into certain approach, but performance verification would be expected for each of product/process. (note :- each product and process even when grouped under considerations of categorization, is required to be established at that group/category level; more often at each product level.)