Sampling Done Right: Through Automation
by Ryan Szporer
You don’t need to cut corners to save drastic amounts of time proofing. Once you’re organized and set up, all you really need to do is start automating your quality-control process.
In Part 1 of this series, the pharmaceutical and military sectors were discovered to have much in common in their shared pursuit of the highest quality standards, thanks in part to guidelines established by regulatory bodies within each industry.
Sampling is conversely as close to a universal as you can get in quality control. Guidelines are for all intents and purposes set internally as companies try to balance the needs for both thoroughness and cost-effectiveness, as they try to get a product out the door as soon as possible.
From the Printer to Production
If you outsource your printing, which is a fair assumption, there is an overwhelming, understandable desire to move to production as fast as possible once packaging components are completed and sent back to you.
That obviously doesn’t happen straight away. Companies know to check whatever is received, which takes time. However, when the shipment from the printer comprises literally a truckload of copies of carton packaging, it becomes unrealistic to manually verify them all, unless you want to spend days if not weeks on end guaranteeing the accuracy of every single one. That also costs valuable resources and manpower that could be devoted elsewhere during that same period.
That’s where sampling comes in, whereby the quality control department, using a formula, will determine the number of copies they need to check to reasonably conclude the whole shipment is good to go. So, for example, out of thousands, only 40 will be proofed, normally by eye.
There are of course inherent issues and risks of which to be wary. That’s just logical when an exceedingly large fraction of the shipment will go unchecked. That’s not to mention proofing fatigue, which, even with such a lower number set aside to be proofed, will still become a factor.
The Problem with Sampling
Based on the aforementioned formula, you may only “need” to pull 40 samples. Each of the 40 may be perfectly all right, but you may not cover each of the positions on the press sheet with those 40. A defect could be localized elsewhere, in a separate position that, just by mere chance, made it through this part of the process without being caught.
Clients should always be cognizant of the fact that the printers they deal with likely have other customers and complete many jobs daily. Gang runs, during which different products are printed on one sheet to reduce waste and save money are a regular occurrence. As a result, mix-ups become an unfortunate reality in the industry as well.
The main concern here isn’t that the carton of the wrong product altogether will make it all the way through the quality-control process. That’s unlikely, to say the least. However, what realistically can are product variants, where, instead of 500 ml ending up on the package, 250 will instead. Slight discrepancies like that are hard to catch when you’re looking in the wrong place.
Beyond the Printing Press
It’s an issue that extends beyond simple press sheets. For instance, it’s not uncommon for printers to deliver rolls comprising single lanes of samples to their clients. In their original form, the samples are printed across several lanes on large sheets and then cut before being delivered. During the sampling process, proofers might know to take samples from different rolls, but there’s rarely a way to know for sure if those rolls all came from one lane or not.
If so, what happens if a foreign particle only got introduced on the press sheet that was being printed at one specific time? An error like that would not be entirely dissimilar to a needle in a haystack, only you wouldn’t necessarily know to look for it. The nature of defects like this, and the differences between intermittent and continuous ones, will be examined in greater detail in a future post. Nevertheless, it’s abundantly clear that the packaging components in that lane would almost be rendered unusable. “Almost,” because they could very well end up being used and appear on store shelves due to ineffective sampling practices.
Proofers and quality control agents may have the very best intentions at heart and legitimately try to be as thorough as possible. It just isn’t possible, and risks, in certain processes, need to be taken so as not to miss deadlines. There will always be risks, but strict regulations, like military standards, combined with automation minimizes them. All the while, companies that abide by them ensure shipments are met on time, saving them valuable man-hours in the process.