Set up procedure for blister lines


We are a CDMO and we pack different products everyday. Our blister lines include a label printer for the blister backing, vision systems for checking lables and pockets and then sealing station, checkweigher etc. My question is when we do a line set-up before running the packaging line, how intensively should we be challenging the vision system for detecting empty pockets in the blister. During the qualification of the vision system we challenged empty pockets, multiple product in a single pocket, broken product and foreign product. But since our products change everyday we have to re-teach the vision system every time we setup the line. So my question is what all testing should we include during the setup to avoid faulty blisters and yet don’t spend hours setting up the line each day and still be complaint?

ps: currently we only challenge the vision system during the set up to detect empty blister pockets for each cavity for one forming cycle.

Let me know if any more info or clarification is required.

Thanks for all your helpful insights!

Ooh. This is a tough question. One that takes a “program” to correctly address. I think this needs to be a wholistic approach, which can greatly benefit from a risk analysis, and failure analysis (FMEA type evaluation).

Your goal should basically be to show anybody that you understand the failures (number, severity, impact, etc.) and have thought about them, understand the failure rate, have in place a failure response plan (for when you find defects) and many other things.

Do you have real time statistics/inspection? Do you do AQL sampling of finished products? Is the inspection manual/automated. Is it 100% inspection or a statistical inspection method?

If you do inspection just at start-up, then there is more risk. If there is a good continual monitoring/inspection, then there is less risk, and the start-up procedures might be less critical.

I think what is hard about packaging, is that constant adjustments are sometimes needed (based on differences in stock, wear of the belts/gears/rails, etc.

One questions I have: Do you have set-up sheets or SOPs for each type of product that you run? If so, where did those settings/instructions come from? Engienering studies and statistics, common practice?