What is statistical process control?
In: Mathematics
Say you own a factory which cranks out 10,000 screwdrivers per day. Each of these 10,000 is manually checked for flaws, which costs $1 each.
Now you notice one thing: It is rare that just a single screwdriver has a flaw. Instead, flaws are replicated in most, if not all screwdrivers from a batch.
Therefore, you decide to change the procedure: Instead of checking all screwdrivers, the QC guys pick just a few of them at random. The idea is that if there are flawed screwdrivers in a batch, it is statistically very unlikely that none of them show up in the sample that they tested. If they do find a flaw, they can then do a more thorough testing for all screwdrivers that might be affected by the same flaw.
This kind of testing is very commonly used for parts which are mass produced. It doesn’t just save money, it also allows to make the testing procedure of the sample much more thorough, for example they can do destructive testing on metal parts or test for biological contamination in food products.
> SPC is a facet of quality management?
Generally yes, one would likely use SPC within a Quality Management framework.
Another way of seeing from a graphical / data point of view, is a process can have an “allowable” level of variation which is calculated using your standard deviation (in a basic sense) when your process data fits within this level of variation (which you can see on an SPC chart as Control Limits) in effect it is saying that you’re process is “Statistically In Control” – which in a simple sense means that your process is stable, with no obvious signs of trends, patterns, or outliers that could be leading to defects or defective items.
You can reference the Nelson rules to support this.
Latest Answers