Do production costs and product quality keep you up at night? Do you have nightmares about the words “product recall” appearing next to your company’s name in a front page headline?
If so, you may need to implement a sampling strategy that gives you better visibility into the root cause of production issues and helps you organize your enterprise quality management efforts.
“In-process sampling is valuable because collecting data throughout a manufacturing run allows you to monitor and ensure the process is operating in a desirable manner,” explains Steve Wise, vice president of Statistical Methods at InfinityQS, in an article he wrote on sampling strategy for Quality Digest. “Done properly, sampling provides an early detection point so operators can take corrective action before continuing a run of unacceptable product.”
What exactly should I be sampling?
There are two categories for determining sample size.
- Process parameters are measured for quality purposes. The goal should be to measure the parts that are crucial to the proper production of a specific product. For example, if you are distributing milk, you will want to focus on temperature versus the flow rate.
- Part measurement samples a process’s stability in terms of meeting the product’s intended target value. Examples of part measurements are diameter and thickness. Part measurements are taken for the sake of minimizing variability during production.
How should I determine sampling requirements? Sampling requirements refer to issues such as how often you should be collecting data and how many measurements are required to gauge an accurate reading for that particular sample.
To determine sampling requirements, first analyze how stable your current level of production currently is. A small amount of variation in terms of production quality will mandate fewer measurements be taken. If you are experiencing unstable production, then you might need to sample as often as every few hours to return to stability.
What is a proper sample size? Think about what you are measuring, and the number of specimens that will be required to gauge an accurate reading. For instance, there is nothing to gain from analyzing the mean of production between 3 cups of liquid from the same mixing vat if taken at the same time. However, one sample spaced out over a specific interval of time will produce an accurate reading.
How can I improve sample strategy?
There are three scenarios that will cause you to alter your production strategy. If problems are occurring late in the production process, you need to sample earlier. If you can’t detect any problems in your production, you can safely reduce the amount of samples. And if your results are producing no variation, you will want to check that the products don’t have a tolerance to your measurements.
Click here to read the article in full.