Did you ever take the time to stop and watch the rain? I’m not talking about a sprinkle, I mean a real downpour or as we say in the South, a “gully washer”. Ok, I’ll translate that. A “gully washer” means that it rained so hard that the ditches were overflowing. This is a serious storm where the rain falls in sheets so thick that it’s hard to see through.
The next time you see a real downpour, take a few seconds and try to focus on a single drop of rain. It’s almost impossible. And if you were to see a single drop and track it as it hits an object, it would be hidden by other drops within a few seconds and you would lose focus on it.
Now let’s think about how the rain storm and Statistical Process Control (SPC) are related. Implementing SPC and attempting to capture too much information or trying to collect data too frequently can also create a “gully washer”. There can be so much information overload that what is collected is impossible to use properly.
Let’s say we create a strategy that collects data for one thousand characteristics. We attempt to monitor all of the characteristics and take action when their behavior changes. Chances are that any process changes would trigger alarms from many of the variables that we are tracking. Because of the interactions, there will also be instances where dozens of variables create an alarm at the same time. The operator would be overwhelmed trying to determine which variable to focus on.
We would likewise have cases where many of the variables would show an alarm, but the product that we are evaluating would remain stable. After a few of these “false alarms” the operator would begin to ignore the alarms altogether.
A process control system is much stronger if we focus on the key characteristics. Concentrating on the key characteristics will help to create a system that generates usable alarms that help us correct issues and improve the product.
Be aware, though, that even when we collect a reasonable number of characteristics, we can still create a downpour if we sample the data too rapidly. Process control charts are useful tools because they allow us to use a sampling of process data to create process behavior limits (control limits).
There are many reasons that collecting data too frequently will cause problems; however one key issue is that it will create process behavior limits that are too sensitive and generate false alarms. Every time we generate false alarms we tell the operator that the system doesn’t really work and therefore alarms don’t matter. (By the way, for those of you too young to catch it, I borrowed the title for this blog, “Who’ll Stop The Rain”, from a song by Creedence Clearwater Revival.) Their song also contains a verse that describes the heavy rain as “Clouds of mystery pouring confusion on the ground”.
Next time you are working on a sampling strategy be careful that you don’t inadvertently create a “gully washer” that pours out confusion upon the users!