9 hours 53 minutes
Hi, guys. Welcome back on Catherine McKeever and this is your lean six Sigma green belt. So they were going to pick up where we left off on our last module and talk through process stability, which is one of the core tenants of statistical process control.
With that, you're going to be able to understand the relationship between your organizational maturity and your process stability. So
in our last module, I mentioned that you don't have to do a domestic project to implement SPC. We're going to talk about where you need to be. In order for to use this tool effectively, you'll be able to understand special cause and common cause variation. So we've been hinting around at it. But this one will actually give you the definition,
and you will be able to differentiate between stability and capability.
They sound really from a similar, but they are, in fact, different concepts. So I just want to make sure that your crystal clear on the differences between them.
So with that, let's say that you are a super, super savvy green bell, and you understand that you can implement statistical process control without a process without a project When are you going to do it? So in order to use statistical process, control your processes.
I need to be relatively stable, at least the beginning, which means that you need to either be finishing up your level two and pushing into level three or firmly. And level three, which is your data driven decisions.
So level ones and twos, same people doing the same work every way. Groups of people doing the same work every time, data driven decisions. That's where we're going to start implementing metrics. We're going to start documenting and standardizing our processes, codifying that this is where you're gonna want a pert statistical process control.
So you're gonna wanna have those repeatable processes.
You understand? Of course, there's variation, but statistical process control is really helpful for determining what your next course of actions are. If you're looking at special cause or common cause variation, which will help you support the data driven decision making, it will give you a really good overview of how your processes function.
So with that process, stability is simply the consistency of your process outputs. It's how likely are you to get a similar result when you run the similar process or the same process. Eso When you're thinking about consistency, what I want you to think about is
constant mean, which is your average and constant variance.
So if we're gonna use the example of how long it takes to complete a process, let's talk about me commuting to work. And I say it takes me between 20 and 25 minutes. So I have a pretty constant mean my averages somewhere between 20 and 25.
Then we talk about very insults. Think I'm running a process in your running the same process because we're firmly in level two.
I get 100 errors and you get five errors. There's a wide variance there, so you're variances. Your standard deviation. It's how far away or your data points from your mean. So when we're thinking about process stability, we want consistency and both are mean and are variants.
And of course, smaller variance is better, because that means we're operating at a higher signal level
from our d PMO Sigma calculations earlier on.
So with that,
you know that everything that we can do can be done mathematically, or it could be done visually on when we're talking about measuring stability. There are two ways you can do it visually. On the first way on the left is a scatter plot. This should not be a surprise to you. It's the exact same graph we used for scatter plots and regression analysis.
But what this is telling us is the president
age when they were elected or the year they were born in the year they were elected, which will give us an age value. So when you're looking at a scatter plot to measure stability, your trend line is going to be equivalent to your average. So you're gonna follow your trendline,
and then your individual data points are going to show your variants by how far away from your trend line they are. So if you look on the far left when we started this whole presidential thing, everybody was basically the same age when they were elected. Those data points are right on your trend line.
However, when you look towards the right as you keep getting up to Georgia in 19 fifties,
you'll see that those data points are further away from your trendline, telling you that there is more variants in your process. So that's how you would read it. If you're looking at a scatter plot because we are looking at statistical process control. Of course, the other way is a control charts. So
you should recognize this. This is what we introduced our statistical process. Control with
what you are looking at here seem the type of thing you're dotted. Line is going to be your average, so your average isn't moving. This is good. Um, and your individual points the further away they are from the dotted line, the more variants you are seen.
This process that we're looking at is technically considered
out of control because it breaks shoe hearts. First rule, it has one data point outside of your upper control limit. So with that, as you are looking at your control chart, we're gonna have some idea of in control and out of control, as we have some shoe heart rules.
But in broad variation what you are going abroad
terms where you are looking for to determine process stability is how close to the average are your data points and how consistent is your average, or does it change over time? So with that
the next thing up is going to be our special cause and common cause variation. So special cause is a special event. These were the things that you were going to track down as a practitioner. These air what is interesting for a lack of a better term. These are things that are unexpected.
So if you understand that the you may hit a stop late and you may not hit a stoplight,
that's expected. If you are cruising into work and there's a giant traffic accident and you're hanging out on the shoulder for half hour, that is definitely special cause because it is unexpected. Another way to think of this is outside of the process. So if this is something that you can't control or isn't on your current state process map,
it's going to be special cause.
But really, when we get down to the very specifics of it, you may have heard the phrase signal and noise. It's very popular when we're talking about data, data analysis, big data. Those sorts of things signal indicates that there is something there,
so there's something that needs to be changed in your process, or there is a change happening.
Noise is just what happens. It's the static on the back of the radio. It goes back to that first assumption of statistical process control that all processes have variation. So if you look at my little picture common cause, all processes have some variation. But everybody's kind of hanging out in the same area.
It's because that's part of that normal inherent variation. Special cause way out of left field. We weren't expecting it. We weren't looking for it.
These are things that we really want to drill into and do root cause analysis on toe understand what caused this type of variation?
So with that process, capability and stability, they sound very similar, and they're not. It'll so process keep ability is whether or not your process performs within your customer's requirements. So, as you are looking at,
you're process capability. We're going to be looking at,
are lower level rose. So both the lower left and lower right. You notice where your upper control limit is in your lower control limit, where your upper specifications and your lower specifications all of these processes are functioning well within them, so that process is capable
If you look at your top row, your stability and capability. What you are looking at is the stability is that consistency. So on the left side you see your upper left. You have a lot of consistency. Your process does the exact same thing every single time.
But you are not functioning within your spec limits. So your process is not capable, but it is stable.
Um, your upper right? This is really about situation all around. Your process is not capable. So you look at anything that is outside your upper and lower spec limit. This is waste. Um, this is going to be reworked. These were going to be rejections. This is all bad,
and your process is all over the place, so you don't even know what to expect. At least on your upper left,
you have some sense of what you're gonna waste every time you run the process. So your ideal state
as of course, your lower left where you've got your process is very stable, same process every time and very capable. You have room for a little bit of that common cause variation because you've got wide margins between what you are actually doing and what your customers require.
So with that today, we went over process stability. You know that it's a measure of consistency, and consistency means mean and variance. You know how you're gonna identify it or monitor for it. So you're gonna look at your scatter plots of your control charts, and you're looking at your distance from your average on and you understand the difference between stability and capability.
So capability is whether or not you are functioning within your customer's requirements,
and stability is whether or not you are repeatable and consistent.
So with that, our next module, we're going to do even more on control charts, so I will see you guys there.