Organizational Performance Part 29: Process Characterization Application | Operational Excellence Quick Hits

Quick Hits share weekly tips and techniques on topics related to Operational Excellence. This week’s topic shares a process characterization example. We hope you enjoy the information presented!

, Organizational Performance Part 29: Process Characterization Application | Operational Excellence Quick Hits, Future State Engineering
, Organizational Performance Part 29: Process Characterization Application | Operational Excellence Quick Hits, Future State Engineering

Speaker 1: (00:05)
In last week’s session we talked about process characterization and the relationship of the independent variables and dependent variables of a process for our system. So this week we want to talk about an example of using process characterization to improve a process performance.

Speaker 1: (00:24)
So what I have here is a simulator that simulates a process, so it doesn’t matter what the process is. We have different variables in the process, so I have this funnel up here where I can change the position of the funnel. I have pins that are configured a certain way, and then I have these slots that are numbered. And what happens is when you run the process, a component goes through and falls into a slot. So I can just continue to run the process and take data on the process.

Speaker 1: (00:55)
What we want to understand is what are the variables in the process, what are our customer requirements, and how does the variables in a process affect our customer requirements? So what I’m going to do is I’m going to run samples of 1000, and then I’m going to take the data and look at what it shows me, and then understand what I need to do to improve the process performance and meet our customer specification.

Speaker 1: (01:23)
So what I’m showing here is, here’s my process characterization. And then I have different settings I can set for the process; I can measure performance. Here’s my customer needs and then I can type my capability. My independent variables are the things I can change in the process. So I can change this funnel width, so I can set the funnel width here. I can change the number of pins in the process, or I can change the position of the funnel. Our customer wants a certain voltage of this product, and they’ll accept a certain range of voltage in the products that they receive.

Speaker 1: (02:00)
So, first question is, of these different variables, how do they affect the outputs of the process of what the customers demand? So the customer wants a certain voltage, they are looking for a nominal voltage of 14.5, but will accept of voltage as low as 10 and as high as 19; the product that we shipped to them. So I said, okay, first thing is the funnel width; what’s the relationship with that funnel width to the voltage of where the heart’s going to fall when I produce it. So my funnel width is how wide it is here. And what effect does that have on the voltage? So I’m saying the relationship of that is about a 3 on a scale of 1 to 10. Also, what’s that funnel width on the variation of the voltage. And I’m saying that’s a high relationship’s. I’m going to make that an 8.

Speaker 1: (02:52)
And likewise, the number of pin rows on where it falls is a 3 and the variation of 7. Then the final position, I’m saying that really affects the voltage, but doesn’t have much effect on the variation. So if I set it up here and I run a thousand samples and I have a final with the 5, a pin row setting of 10, and my position of my funnel here is at 8 and I run it; let’s see what happens. So I run it and I get a product mix here with variation. And I see, I have an average of 13.05 and a standardization of 2.12. If I punch in my data here; so my width is 5, my number of rows is 10, my final position is 8, my mean from the process is 13.05, and the standard deviation is 2.12. So if I do the calculations, I see my lower control limit is 6.69. My upper control arm is 19.

Speaker 1: (04:00)
So that tells me my 6.69 is year and my 19.41 is here. So I’m saying, okay, all the data is going to fall within that, but my customers want between 10 and 19. So I type that in my Cpk of 0.48, which is not capable. And I see my Cp is 0.71. So it doesn’t even have the potential to be capable because I need the Cp to be above 1.33. And I also, I want my Cpk above 1.33, if I’m perfectly centered. I understand I’m not capable. So the question is, what am I going to change in the process to make it capable? Well, first of all, I see my variations too high cause the Cp tells me that I have too high a variation. So I need to reduce the variation. So when I look at this I say oh, variation, which variable has the most impact on variation?

Speaker 1: (04:58)
It’s my funnel width. So what I’m going to do is I’m going to change my funnel width here. I’m going to reduce that down to a 2. I’m going to do a process improvement, get that down to a 2. Now I’m going to run it again. And I see my standard deviation is improved. So I know the variation has reduced. So now my funnel width is 2, my number of rows is still 10, my final position goes to 10, and my mean goes to 12.47, and my standard deviation goes to 1.68. So my capabilities improved slightly to 0.49, but I’m still not capable as my Cp increased slightly from 0.71 to 0.89. So I still need to reduce the variability. So now I go back to my process characterization and said, oh, the next thing that affects the variability is the number of pin rows.

Speaker 1: (06:00)
So now I’m going to reduce the number of pin rows here and see what effect that has. So I’m going to reduce that to 3 and now I dropped the balls again. And what happens that reduces the variability, so I see my distribution get narrower. So now let’s punch in the data here. So my width is 2, my rows is 3, my position is still 10, and now my average is 12.52, my standard deviation is 1.10. Now I see my process capability has improved the 0.76. That’s still not capable, but I have the potential to be capable because my Cp is now above 1.33; it’s at 1.36. So now, what this tells me is I have the potential to be capable, but I’m not centered. So I need to center the process. So which variable affects the centering, which is not the variation, it’s the voltage itself.

Speaker 1: (07:02)
So the funnel position as a huge effect on the centering. So now I want to adjust the process to center it. So what I’m going to do is I’m going to move the…

Speaker 1: (07:14)
I need to increase the voltage because I’m, my process is at 12.52. I want to target it 14. So I’m going to move the position of the funnel two positions to the right. And I’m going to try running the process again. And now I’m more closer to my nominal and my variation is still small. So let’s see, now our width, what stayed the same at 2, our rows is at 3. I moved at two positions to the right, so now it’s at 12. Now my average is 14.79 and my standard deviation is 1.03. And so now I look at my capability and I’m capable. So now I see that there is a relationship between funnel position and variability because I see here and my Cpk and my Cp is now 1.36.

Speaker 1: (08:15)
So now it’s capable. So I systematically went through and said, okay, which independent variables affect the output? I said the funnel width was the first thing I changed; I got some improvement. Then I changed the number of pin rows and I got a much bigger improvement. So I’d probably want to change these and say, you know what? Pin rows probably has a bigger effect and funnel width has less of effect. Okay? So this is my ranking. Now I get a big jump in capability by saying, if I reduced the number of pin rows that affects the capabilities significantly. And now that I got up to have the potential of being capable at 1.36, I need to center it. So I center it and I say, okay, my funnel position affects that. So I adjusted the funnel two positions to the right, and now I got the process capable and producing all good parts. So this is an example of how we use process characterization, how we tie that to the process itself, and improve the process capability to satisfy customer needs.