S4 E9 - Bill Bellows - Unveiling the Spectrum of System Thinking Part 2

In the second episode of my two-part discussion with Bill Bellows, we explore the nuanced realms of quality management, variation, and integration, drawing from the rich legacy of W. Edwards Deming and Genichi Taguchi, and their profound impact on industries ranging from automotive to aerospace.

Bill Bellows shares enlightening perspectives on the historical collaboration between Ford and Mazda, illuminating the stark differences in their approaches to quality and production. He discusses the importance of minding the gap, not just in terms of parts but in the broader context of integration and systems thinking. This episode delves into the philosophical underpinnings of quality management, touching upon the concepts of red pen and blue pen companies, and the significance of aiming for systemic improvements rather than mere compliance with specifications.

The conversation extends beyond the specifics of automotive engineering, drawing parallels with software development, healthcare systems, and manufacturing processes. Bellows emphasizes the critical role of managing variation not just at the part level but as an integral part of a coherent system. This approach, he argues, is key to achieving superior quality, functionality, and customer satisfaction.

You can find Bill Bellows LinkedIn here:
https://www.linkedin.com/in/bill-bellows-218435/

Resources and Keywords:

1. "The Deming Dimension" by Henry Neave (1990) - A book described as a combination of "Out of the Crisis" and "The New Economics" by Deming.

2. Ford Batavia transmission story - A documentary by Ford about their comparison with Mazda transmissions. It can be found online by searching for "Ford Batavia transmission story".

3. "Learning to Lead, Lead to Learn" by Katie Anderson - A book about Dr. Yoshino, who worked at Toyota for four decades.

4. "The High Velocity Edge" by Stephen Spear - A book mentioned in relation to Toyota's approach.

5. "The Tyranny of Merit" by Michael Sandel - A book discussing issues with meritocracy.

6. Genichi Taguchi's work on the quality loss function.

7. Russell Ackoff's work on systems thinking.

8. Edgar Schein's work.

9. The Deming Institute podcast - Mentioned as a source of information on these topics.

10. American Supplier Institute - Mentioned as offering Taguchi methods courses.

11. The Anthropic website - Suggested as a source for more information on the Claude model family.

Transcript:

John Willis:

Cause I really want to get into, you know defense of my writing and then, but get way behind that, which is I, you, you said, you know, sort of Mazda looked at ideal values and, and they managed the integration. And you talked about how, you know, Ford minds the part where Mazda minds the gap. And this idea of the gap, I think is, is sort of fascinating.

You know, like, so, so, so let me take it back to in my book. I you know, from my research, not being an industrial engineer, [00:56:00] I was fascinated by the story of Mazda and and, and, and Ford, right? So Ford had purchased a, a portion of Mazda. You know, in somewhere in the late seventies or whatever. At some point there were these sort of conversations about power buyers or warranty, where the story that I'd heard was that people would basically like the, the big bulk buyers, the big distributor dealers would be saying, Hey, we only want the cars where the transmission came from the Japan ones.

Versus the Ohioans, right? But there's all variants of that, but there was definitely some type of perceived quality issue between some people knew some of the transmissions were made in the United States and some were made in Japan and I

Bill Bellows: would say the Henry Neve and a British statistician professor retired wrote a book 1990 called the Deming dimension, which [00:57:00] is a fascinating thing.

Transcribed it's almost a combination of out of the crisis and the new economics. A, it could be a brilliant first read before the new economics before out of the crisis. Anyway, he has 1 chapter on this transmission thing, which is probably the longest. I've, I've written some articles that reference that Henry's book.

Yeah, yeah, yeah. So this actually leads really well into the sort of

last part that I definitely wanted to cover and this will be a two parter if you still got time. I think something you said I thought was interesting. So you, I want to

talk about the Ford Mazda story and I'll sort of start off with what's in my book

and a couple of things we've had, but I thought was really interesting is.

Is easily the first written account of this anecdote and what I've heard is that yeah, as you mentioned Ford owns one Ford on as much as one third of Mazda back in the eighties. And this was likely at a time when us companies are interested in Japanese companies. And so buying part of them allows you to get inside and look for things.

But again, if you don't know what to look for, you're not going to find. What they're doing. And that's what you see in this [00:58:00] transmission anecdote. And so, so let, and also I remind our audience that Dr. Deming's documentary with the NBC documentary on Dr. Deming of Japan can, why can't we as June of 1980 by year end for it is reaching out to Dr.

Deming beginning in 1981. They're inviting Dr. Deming in for seminars. And in 1983, 8283 time frame, Ford discovered, the Ford warranty claims people's what I had heard and their warranty claim office, they discovered that there were three times as many complaints on the, on a transmission designed by Ford and made by Mazda than the same transmission, which was designed by Ford and made by Ford.

And so, you know, I mean, how could the same [00:59:00] design built by Mazda have fewer complaints. One third, fewer complaints, the same transmission made by, by Ford and Ford came out with a documentary, went out to their supply base and there's a number of different versions of it. You can find the, if, if, if our listeners go online and search for Ford Batavia transmission story.

Yeah, that's a 10 or 11 minute version. I, I've seen longer versions. Okay. So there's different versions of it. And and what you'll find in the, well, another piece is, and I think this is Henry's book, Borge went off and took 10 transmissions at random from both factories, at random, and they took them apart and collectively took over a million measurements on, on 20 transmissions, 10 from each.

And what comes out of the and Ford's documentary, which eventually went out to the Ford supply base [01:00:00] saying, look what we discovered and what you hear in there is that these are the Ford people talking, what they got out of it is that the Ford parts took up when they looked at the characteristics on dimensions and things, you know, diameter, surface roughness is they found that the Ford processes at best.

Were taking up to three quarters of the tolerance, 70% of the tolerance. Whereas Mazda was routinely using 28%. There's a 70 and a 28. Right. Okay. And what's

significant about the 70% was the 70 Ford using 70% of the tolerance was their best. Their best. Not their average. Their best. Mm-Hmm. . Whereas what they also say is kind of across the board, 28.

And I, this is not to say that 20 is the magic number, but what you hear in the video is [01:01:00] piece to piece consistency is the name of the game. And they say that, and they keep saying what we've learned, what we, and if we have piece to piece consistency as they do, then we'll be able to make the trans, trans, trans.

John Willis: Let me interject the point in my book where I think I was, I think it implied that I was saying that there was reduced. that reduced variation was the key. And what I really meant to express was in sort of my research, my version, I think I read something from Peterson. You know, there was a whole question of why did Peterson actually wind up calling Deming?

And I think there was a point of which, like Winston Churchill said, history will be kind to me because he wrote it. But, but where where the engineers couldn't explain, like, the tolerance difference or the the variation difference like, that, that was not a good enough explanation to the engineers.

Like, in their mind, the 70 percent of versus the 20%, [01:02:00] like, so what? It's still piece to piece consistency. And it wasn't until they could understand a deeper level of variation and economics related to variation that that's sort of what, I think, what Deming helped them understand. And what the piece that I missed is I didn't really understand sort of Taguchi and the sort of the, the, the conversation we had in the, in the original discussions that there was more to it than, so I looked at it as there was something, so I tied it back to pragmatism, right?

Like there, there's an economic, there is a point of improvement until there is a diminishing point of returns. Which is the sort of pragmatic argument like the,

Bill Bellows: that would be what I would call a red pen company construct.

John Willis: Oh, really? All right. Oh, boy. I'm getting banged through twice hard now. Okay. But, but no, there's no point of continuous improvement though.

Like you,

Bill Bellows: you literally, I have [01:03:00] seen so many incredible opportunities for improvement. The belief of diminishing returns. Is a and again is a red pen company construct. How do you what I would say to people is. I would say, if you don't look, you won't find. And if you do look, there's no, you just don't know, maybe not discover improvements that you just wouldn't believe.

No,

John Willis: I get that. I think that maybe diminishing returns is the wrong way to say it, but at the core of Shewitt's beliefs comes from C. I. Lewis's mind in the world order, which is about pragmatism, which comes to, you know, the whole the, the, the whole original, you know the Saunders, I guess the, the guy Pierce, C.

S. Pierce, who basically was trying to get the perfect pendulum measurement and decided there was a point of which There was no value of going any further, and I think that is sort of core to, to Shewart's belief, which I think [01:04:00] is part of you know, sort of understanding variation.

Bill Bellows: Well, I would say, I'm glad you brought that up. I would say,
first of all, I don't, I

John Willis: Let me say one more thing, but you hold that thought, because, I mean, the whole point of the name of the book, New Economics, is because what I think most people missed the point was Shewart's point of using statistical process control or all that was about an economic, which does tie in that there's a point that there's economics involved in variations.

Bill Bellows: Well,

I'd say red pen and blue pen companies each have their own economic model and they, and they each believe their economic model. I mean, they're each driven by an economic model.

John Willis: But a red pen. So hold on. So a red pen company would constantly want you to improve, improve, improve, improve, improve a red pen.

Yeah, I think so. [01:05:00] I think they would just, I mean, like in our world, we call it you know, how many nines of availability do you get? Right. And, and so a red pen company in our world would, some managers say, I want 10 nines and, you know, a blue pen would probably say, well, hey boss, wait a minute, 10 nines is going to cost us a billion dollars.

Like, could we just have 99. 99 that might only cost us a million dollars. And where I think, I think there is this maniacal, it must work, zero defects.

Bill Bellows: Well, let's bring it back to. Okay. We can talk about six sigma quality, which is 3. 4 defects per million. Yeah, yeah. Maybe it ought to be 12 sigma quality and 13 sigma quality.

Well, let's understand that. The definition of quality coming out of Six Sigma quality is Phil Crosby meeting requirements. And what they're talking about is if we shrink the amount of [01:06:00] variation in the distribution, then the tails of the distribution Which go to infinity, are going to be, you know, there's going to be less outside, less outside spec.

And so what's, what's the deal there? What they're focusing on is shrinking the variability to the point that there's fewer and fewer things outside of spec. What they're striving for is zero defects. That's got nothing to do with what Dr. Deming was talking about, nor Dr. Taguchi. What Taguchi's talking about, and what leads to, I mean, if meeting requirements is all you need to do, then the Ford and Mazda transmission would function the same.

They function differently because Mazda was paying attention to the, my belief. Now again, I'm surmising this because there's very little information in the, the Ford video about the Mazda transmission talks about piece to piece consistency. They don't talk about gaps, and I think they spend [01:07:00] so much time talking about the parts that I infer that they're talking about the parts and not the clearance.

So my explanation, which is not part of anything I've seen written, is that what Mazda was doing is minding the clearance. Between the bore, which is where the, you know, where the valve slides in minding the clearance between the valve and the bore. And to me, it's just, it's just what I'm not hearing in the conversation is what is the clearance between the parts?

What I'm hearing is the parts meet print. And likewise, a six sigma quality. It's all about through the parts meet print all to say that in a, in a Deming organization, we become aware of relationships. We, we become aware of

where does it make sense to mind the part, because there's situations where that is okay, where meeting requirements is all that, again, in baseball, there may be a batter coming up that can't hit the ball anywhere, so it doesn't matter where the ball is in the strike zone, if somebody else comes up, [01:08:00] and you've got to have it in a precise location away from, that's managing variation in the location, but relative, what I would also say is, in a red pen company, meeting print is all we have to do.

And if we ever achieve zero defects, then we're done. Once we eliminate waste, then we're done. Once we eliminate non value added, then we are done. And so, this reduce to zero mindset infers that once we get to zero, we are done. That's not what Deming is talking about. Relative to the economics, I would say, That

trying to make the gap more uniform, more and more and more and more uniform may not be a worthwhile adventure. So I would then say that just because I can improve the uniformity of the gap by having the distributions become narrower and narrower and to the point that the gap becomes more and more consistent, that may not [01:09:00] be worthwhile.

And so then I would say, just because I can continue to improve it doesn't mean the value. Is worth the expense. Yeah, I mean, that's

John Willis: my pragmatism argument that I mean, that's
Bill Bellows: what I agree with. But what I'm saying is that I don't think I think

I think you could just I think how you discover that that break even point.

I said, there's no to me. There's no end to the number of break even points. You can discover and that's why I don't think there's an end,

John Willis: but I think I go back to like, like, when you say like, okay, so, like, in the fourth scenario, they were you know, the, they were basically you know, if I'm saying this wrong, right, but managing the tolerance or what you said, managing the parts.

Yes. Mazda was managing the clearance or the gap, but I think isn't the point of why Mazda was doing it. is they saw an [01:10:00] economic value in doing that way. Yes. Or didn't understand. Absolutely. That's basically what I was trying to, and then, like, I wish I would have met you before I read the book because I would have known Taguchi May better, and then I would have explained it a little, you know, a different way.

But that was the point I was trying to make.

Bill Bellows: No, but my point is Red pen companies again. We can use that term. I think at this point where

John Willis: we can, we can clearly do

Bill Bellows: that. So let's continue to use that. So, so red pen companies, I would propose by no value and working on things that are good because they're good.

I've had people tell me within Boeing that I was told by the lean folks that once the holes met requirements on the low side, increasing the diameter of the hole to the target value, shifting the average to the target. Once I, once they're all within spec in the low end, they said, I was told by the [01:11:00] lean folks that shifting the average to the target doesn't add value.

And I said, yeah, if all you do is look at the whole, that's, yeah, but if you look at what goes into the whole and how the whole contributes to integration and function, then you draw a different conclusion.

John Willis: Well, it goes down to what good means too, right? Like, it's like good, like, it's again, is good.

Yeah, yeah, the good is. That's right.

Bill Bellows: And, and, but, but also let's take into account. that good is it is an end we are done as opposed to in a blue pen company, you can say, well, I can make the gap more and more uniforms. I, John, I don't know that we can, for the time being, let's put our money elsewhere. Right.

So that's what we did at Rocketdyne. Once we discovered the gaps that improved the design and fabrication of some hardware, then we use that thinking. On other engines, and then we look for what's the next [01:12:00] thing we can discover because if you think of the engine as being a bunch of gaps and a bunch of relationships.

Now, the question becomes, okay, we discover a relationship that we never paid attention to, like the clearance between the valve and the bar. Now, let's start to design all our transmissions that way. Okay. Now, let's look at what's the, what's the next functional thing we can focus on. And then what we do is and I think.

Okay. I mean, the reason we go from that one function clearance on to the next is, to your point, I think at some point in time we say, I don't think we can make a whole lot more money right now. I can be shown wrong, but I think there might be bigger money elsewhere and then again elsewhere. And I think there's so many other things to do that to come back to this, I mean, again, there could be some, Maybe somebody discovers new opportunities that are a worthwhile investment.

John Willis: You know, I, I often think, and I always wonder if like when I'm saying this, is it [01:13:00] Am I saying it right? Or are the people thinking like, well, why does he say that? But Stephen Spear, you know, Dr. Spear who wrote High Velocity Edge and, and you know, he he did his his dissertation was I always get the name wrong, but it was decoding the Toyota Production, DNA or har, it was one of the most downloaded Harvard Business reviews.

And, and again, i I, I do a little liter, a little bit of literary license on a quote that he has in it. But he said something to the order that Toyota was a community of scientists continually experimenting. And I wonder if a blue pen is, at the end of the day, a company that embraces. Scientific thinking or the theory of knowledge, which then like literally then goes back to epistemology, which goes back

Bill Bellows: to pragmatism.

Oh, yeah. No, these are people that are that the challenge becomes, back to the pragmatism is, is when do we move on to the next thing? And, and [01:14:00] but I would, and that's what I try to do at Rocketdyne is that we have these incredible breakthroughs And but but it's not to say there couldn't be a bigger breakthrough, but I think there could have been a an easier breakthrough elsewhere than continuing.

I think there's so many opportunities around the organization to, you know, mind the gap. And what we're talking about is shifting the focus from these are good because they meet requirements to how are they used and then you pay attention to how are things used. And then you start to see, oh my gosh, if we spent a little more on staplers, we could make it easier.

And those types of things. It's like

John Willis: you said, you know, the analogy you've used over and over, and even when I was rewatching the Deming Institute, which I'll put the link up,

which about the whole me,

Bill Bellows: [01:15:00]

We, we, you know, me and my spouse, our family conceives of it. We And it could be we get to a situation where the ones making the components, making the plan, planning the airfare, are thinking about the hotel. And it's done as a system where you go to work and one person does the airfare, somebody else does the hotel, somebody else does, you know, the other aspects of it.

And it comes together kind of fragmented. But at home, why does it not come together fragmented? Because a few of us have our minds together at home. [01:16:00] And then as soon as we go to work, We start to manage parts. And the point you're trying to make is why is it at home? We manage systems and relationships and at work.

We manage parts the same person. And that's what I think is so I think this is what I think is important. There's reason for hope is that when you show somebody how well they manage integration at home, then you have to believe that we all have that skill at home. The difference agrees. Then let's take that home skill set and take it in the work.

I've had people at work say, you know, build is focusing on the one line cutting to the line that's over our head. And I said, My three year, our three year old son knows how to do that. Don't tell me now again, it's, it's called covering within the lines. And he was just blowing smoke, trying to get me out of his office.

And I said, you know, bill, we're not ready for that. And I said, no, this is not. And this is, I think, important for our listeners is to have somebody [01:17:00] appreciate that what we do at home, we could do at work is a really big thing that we cut for the cut to the line at home, but also take into account the work system.

May not be designed for that. So it may take some time to redesign that system or procurement system. And I'm okay with that being a challenge, but I'm not okay with somebody saying. You're just flat out ignoring it. But also, I also

the long version of me and we, which is you go back

to like, you know, when you're building stuff at home, you are minding the gap.

Yes, because at home, that's right. And that's what I

was sharing with somebody, a number of people recently. Yeah, I had a lecture

last week at Cal State Long Beach and I was trying to explain to them. And I

think this is, I'm glad you brought this up is at home, whether we're planning the

vacation, planning the project in the garage, planning the project in the

backyard, whatever it is, we conceive of it as a whole.

appreciate that. I have to be patient in appreciating that the system has what the system is doing is is what Deming would say is exactly what it's supposed to do.

It is promoting a focus on parts. And as long as we've got that system, we've had issues. So then you have to begin to realize is okay. The personal transformation says this is possible. The organizational transformation is going to take some time. Now the challenge is can we work with others in the system to maybe do a small scale PDSA and, and, [01:18:00] and, you know, shed some light on that system.

But I think it's, it's appreciation of what there has to be a mutual appreciation. That's all I wanted to say is that the system is capable of more, but let's appreciate that. It's And make, take some time to jackhammer it. Well,

John Willis: and I think, you know, back to the home versus in the organization, I think again, to me, that the whole idea of the, the, the trip report is a great way to at least step out and see your system and have a, like you said, a shared mental model, or at least create a sort of something that can create a shared mental model so that you can see yourself outside of the,

Bill Bellows: the thing.

And I say. I've seen, I'm aware of people using the trip report in different ways, but not all using it associated with a Deming transformation, which is how I use it. So they may go off and, and, and get, you know, create the results of the four [01:19:00] quadrants and all that. And, and and there were people within Boeing running and doing those things and, and those of us that were using it with a Deming construct didn't quite know what they were doing, but they were getting mileage out of it as it Fine.

It's a, you know, it's, you know, freedom of speech. Go ahead and do what you want. But, but what I would say is, is, is for those of our listeners that are interested in a Deming transformation then associating that trip report with an explanation of the elements of the system of profound knowledge, and then how the elements work together as a system, I think that could become invaluable.

And even, again, what I was looking at is, you know, I was, a couple other things to say there. One is that this is not me telling people what a Deming organization looks like. And I don't even say Deming organization. And I could go into an audience, not say, they don't know who I am, I don't say anything about Deming.

I start talking about All Straw and Last Straw, or [01:20:00] Blue and Red, or Me or We. And out of it will come the contrast. And then I'll introduce Deming's work to explain the differences of the cultures in the, you know, the artifacts, the manifestations is, is shine, would reference them. But but again, I've seen people just take the trip report and have fun with it on its own.

I think there's a lot more we can do with it. Yeah,

John Willis: no, I, I love it. I, again, I have a real, this is great because just like everything else, I, you know, like I was trying to grasp what you were saying about Taguchi on your on the Deming podcast, and I was getting frustrated because, you know, I, those are great, that's a great series, but like, I was like, stop, I want you to go deeper.

And then I was like, okay, on my show, we're going to go deep on, on Taguchi, we're going to go deep on Arkaff, and now we're going to, and I have a real better appreciation for, you know, what you, what really You were doing with the red [01:21:00] pen, blue pen.

Bill Bellows: Well, let me, well, I think, and I really appreciate this time with you as well, because we get to explore in more depth.

I was, it's not uncommon when I present my interpretation of the Ford Mazda comparison, because I don't know the distributions. In fact, back in,

yeah, it was 1989. I came across, I first saw the Ford transmission video, whatever version I saw. I was in, in fact, I was in a Taguchi methods course For the very first time, actually, in 1987, I was in a Taguchi Methods course offered by the, the Dr. Taguchi's company, the American Supplier Institute, which still exists.

And and that's when I first saw the forward transmission video. And, and a lot of the focus was reducing variability, reducing variability, reducing variability. And it wasn't, It took some time to then realize it was more than that. It was, it was [01:22:00] managing the variation of the parts to achieve the clearance because if all you do is talk about the part, now you're back to life in a red pen company.

Well, the reason I bring it up is that I, I saw the video. I wrote a letter to the then VP of Ford. His name escapes me. He's a, he worked closely with Deming and and I wrote a letter to him saying, I've, I've recently discovered. The Ford

transmission story. I've recently driven a Ford Taurus that I had rented and I had some questions about that.

So I wrote a two paragraph thing to him and have back a handwritten, a typewritten response from him within a week with his signature on it. And again, his, he would have been the VP of quality. He would have been who Bill Shirkenbach reported to. And what I asked was, is there additional To Information on the transmission story, and he said there's no additional [01:23:00] information.

Yeah, so it's a lot of what I see in that. I infer if you listen, they're talking about the distributions separated. They're talking about making the valve on the high side, the bore, the bore on the high side, the valve on the low side. And what they're talking about is looking at things in isolation. I do not hear the word.

Glitch. You know, I don't hear any discussion that we have discovered that for that Mazda is managing the relationship between the parts. What I hear is how the parts are better, but that doesn't explain to me how they function better. So I offer that explanation as an, you know, I talk in terms of that.

They had, I mean, I tell you what we did at rocket time. We have a distribution for the, for the tube, which has an outer diameter. We have a distribution for the whole and literally what we did was, okay. was get them to a point that the clearance, the average [01:24:00] clearance was a particular value. And so what we knew is we knew there was distribution, there was variation in the holes, there was variation, there was variation in the ID of the hole, there was variation in the outer diameter of the tubes.

And once the variation was reduced enough and experiments were run, then it was a question of what's the clearance. So again, it is variation in both. Now we're trying to achieve a clearance. But the clearance depends upon the variation in both. I'm not looking at the clearance in isolation, I'm looking at, this is what we call the rocket on managing variation as a system.

That explanation I've not heard from anybody. But when I offer that explanation to people, They, they fall head over heels into the trap of Mazda change requirements. I said, no, Mazda didn't change the requirements. What Mazda did was change how they met the requirements. Yeah. Mazda change requirements. I said, no, the requirements are the same.

And the [01:25:00] reason I bring this up is for one senior audience of senior technical people, and it was so funny to go back and forth and find it like, Now

we get it. They're managing the clearance. I said, yes, they're managing the clearance. And so afterwards, one of these guys comes up to me and he says, he says, well, What we're doing is so close to that, so close.

And I said, here's, earlier before we started the recording, he was talking about close, and I said to him, here's how close we are. You're very close, but here's how close we are. I said, because we're standing pretty close together, I said, this is how close we are. I said, we're so close you can put an infinitely thin piece of glass between our noses.

I mean, we're almost touching noses. That's how close we are. I said, but the problem is the glass is infinitely wide. I said, we're as I'll give you a credit that we're close, but I was telling this guy is, but you weren't nowhere in the closeness. Were you focusing on the [01:26:00] clearance? All of your thinking was on the parts.

We really close, but still different. Yeah. Yeah. Close on the parts is not close on the gap. I, I

John Willis: think, you know, as you saw, we, I think this is great. I think we covered a lot of cool ground. The thing I, I, as we get near the end of these things, you know, the question comes up and, you know, I got to figure out a way to do this constructively that works for everybody that, like, I need to get more people, you know, cause I mean the real question that everybody's sort of like, I've gotten it already from some of the other podcasts we've done together.

Again, you're growing a pretty good fan base here. It was like, okay, like, These are brilliant ideas. Like, what do we do in software? And they like, I think I have some ideas about like some of the stuff in my book and, and, you know, stuff I think about a lot about like what we do and, you know, how do we mind the gap between, and we have all the, I mean, we in software delivery, we, we you know, one of the [01:27:00] things that, you know, sort of Gene Kim, you know, did with his original Phoenix project book, which was really try to isomorphically map, you know, sort of manufacturing paradigms or manufacturing economies to, you know, so ideation or sort of knowledge economies.

Right. And so, but we spend a lot of time describing the delivery of software, not so much the creativity side. That's a harder problem, like what we call ideation and how that's, you know, there's still a lot of debate on like, you know, Can you know, knowledge work, ideation and knowledge work be sort of a supply chain.

But there is a point, which when somebody, what we call a commits the software, right? And that's a point where we like, like every bank and every insurance company and Google and that is supply chain. I mean, it's, you know, there's a point of where you, we say, okay, here's the code. It's going into this process from here on in, it's going to have [01:28:00] a very mature set of integration points.

It's going to do this. And it is not. Dissimilar from sort of an assembly line of a car, right? You know, because all these parts and there's testing and there's sort of development and there's, you know, sort of delivery to a staged environment and, you know, then another set of tests. And so a lot of this stuff that you've been talking about when we talk about like Taguchi and the handoffs and the gaps and all that, it's all there in software delivery.

It's just be fun to try to figure out like exact examples and we need to get back to

Bill Bellows: it is well, a couple of things when

the [01:29:00]

And, and so we, we create the model, we pay attention, we look at how it's used and then we fold the use back into the system that that is beyond where companies are today. I mean, a value stream is not circular value streams that I've seen have a beginning in an end. And everything I saw at Rocketdyne within Boeing, value streams are beginning and end, not circular.

And so this whole idea that what comes around goes around, which to me in a blue pen company is the appreciation that whatever I put into the air, we all have to breathe. Whatever I send downstream comes back to me eventually. And so why am I, that's what I, you know, why at home do I focus on the line?

Because I have to be the one to integrate this and I want this to last a long time. Whereas in a red pen company is a. I treat you as an out [01:30:00] of state customer at my father's gas station. Not that we did, but we there were gas stations that saw Florida plates in New York and they say, Oh And they see as an out of towner, you're not going to come back here, and next thing you know, your car needs more work.

production viewed as a system, you know, ABC and D the inputs come in and

one thing I think our listeners

might appreciate something I didn't figure out till maybe 4 or 5 years ago in a

conversation with with our, our friend, Dick Steele. When Dr. Deming shared

big loop production viewed as a system, which could be software

viewed as a system.

Whereas if you're local, you didn't get that because, you know, what comes around haunts us. But if you have a circular mindset, which is an appreciation of the environment, you know, and you know, the generations after us, then you start to pay attention to these things. Well, the reason I bring that up is, Dr.

Deming did not meet Dr. Taguchi until the mid 50s, and Dr. Deming was there in 1960 when Dr. Taguchi was honored with the Deming Prize in Literature for his foundational work on the loss function, which is the appreciation that where you are, how you meet requirements affects integration, which is why at home we aim for the line, because if it's a little long and a little short, It's more work for me.

Well, if you have, if you don't have that model, [01:31:00] but your Dr. Deming in 1950 with this idea of what comes around, goes around that model absent to Taguchi's model. It's not going to help me figure out why those two transmissions function differently, because in that model, I'm still thinking meeting requirements.

They all meet requirements. They all meet requirements. But what what dawned on me was once I understand production viewed as a system coupled with a loss function, now I can make a lot more sense out of that, meaning that if all I have is what Deming knew in 1950, production viewed as a system, still not necessarily aware that how I meet requirements matters, then that's, you know, you may come to me in that model and say, Bill, these transmissions are failing.

You have more issues over here. And I'd say, well, John, it's got to be how the parts meet print. There's got to be something wrong with how they meet print. [01:32:00] Right. But once I become aware of Taguchi's work, now I'm thinking It's not a part issue. It's a relationship

John Willis: issue. Yeah, yeah, yeah. And you couple that with the circular thinking of the production as a system.

That's right.

Bill Bellows: And this is how I, and this is what I find exciting is, you know, as I look at Edgar Schein's work, as I look at the work of Michael Sandel, is a moral philosopher at Harvard, I think it reinforces my appreciation of what Deming's talking about. Which is why this transformation. Is an ongoing phenomenon.

But going back to your point, I think I think in any organization, if you looked at production viewed as a system one and then began to add to it the appreciation that how the requirements are met, whether you're baking something arranging airfares, I mean, If we, as we begin to look at the variation in the components and how the components, how that affects [01:33:00] integration.

Now we're on our way to managing interactions and inherently looking for where we can spend time to improve how things come together and there there's no, that's generic, how that translates into something specific is going to be different, which, and the beauty is. When you're talking all straw and last straw, there's no reference to manufacturing.

There's no reference. It's kind of as soon as we say red pen, blue pen, now people that software people would say, but we don't make pens. We don't make, yeah. So then keep it at all straw, last straw, keep it at me. Yeah.

John Willis: Yeah, I think, again, I would probably now I'm worried that the Zoom may blow up on the amount of time that I'll record, but but the I, the, the Red Pen, Blue Pen, I think does work for software, because we do have this sort of, there is a lot of fit, you know, it's not the same kind of fit, but, but like, you know, if you're sort of integrating multiple libraries and [01:34:00] different, there's a lot of fit, sort of constructs, so it, that, that one works, you know, probably just as well as the other two, so, but, but I do like the, I don't

Bill Bellows: know.

And I'm glad you, a couple of things you mentioned, one is, you may have heard me talk about is, Dr. Taguchi talks about quality, the quality loss function. Right. What I started talking about a few years ago is the integration loss function. Because I think integration, I don't, I've not yet come across what I think is a more universal term than integration.

If I say assembly, that's manufacturing. If we say welding, that's manufacturing. If we say mixing, that sounds like ingredients in a, in a, in the food service industry. Joining sounds like manufacturing. Yeah. Yeah. But integration to me is, so I talk about integration loss. I love

John Willis: that. Yeah. Yeah. Because that, that could be any, any [01:35:00] economic model or any sort of.

Yeah. That's your model. Yeah.

Bill Bellows: Yeah. And then you look at, you know, how much time do we spend and to what then you could say, what does integration mean in software? And there's also points of integration. That's right. Yeah,

John Willis: I think it goes back to when you said something earlier that I said earlier that I was in your video is.

You know, Mazda, you know, manage the integrations. I believe,

Bill Bellows: I believe, I think so. Again, I would, I would love, I would love John, if you and I could go deep into Toyota or Mazda and ask questions. And I think we would discover they're doing what we're talking about. And, and, and I don't know if you've heard this in, in a, in an institute podcast, that means to podcast, but I had the good fortune of flying home from Japan.

I was over there in a work visit. Boeing Rocketdyne was [01:36:00] developing a next generation upper stage engine. So the first stage gets the vehicle off the ground. The upper stage ignites in space and pushes the second stage out or the satellite out. So Rocketdyne was partnered with Mitsubishi on an engine called MB60 MBXX, a Mitsubishi Boeing 60, 000, 70, 000 pound thrust engine.

And I went over as on a on a contract with Mitsubishi to help them get ready for joint production of a Rocketdyne owned by Boeing Mitsubishi upper stage engine. And part of the visit was I was invited on the visit to look at how they thought about these types of issues. How did they manage variation period on the part level?

And what I found is they were doing no different than anybody else managing parts in isolation. Well, flying home from Nagoya, which is 14 hours. We got to [01:37:00] at Boeing at that time, you got to travel for overseas. You got to travel business class, not in the back. So sitting next to me. As a young engineer, Cal Poly San Luis Obispo.

And he works at the time he was working for Toyota at their new me plant. And so I did the trip report with him. We had 14 hours of conversation. I did the trip report. As soon as I found out he worked for Toyota is like, Oh, my gosh. So I explained the red pen, blue pen trip report, target thinking, tolerance, thinking paradigms, ABC and did all that stuff explained to him.

And he said, that's what we do. That's what we do. He says, I'm coming back from meeting with a supplier that wants to change the tolerance. And I had to explain to them, you have to hold this target or else it's going to screw up

integration. He used those words. What he explained to me that I can't find anywhere on the internet and I've looked many times is how does Toyota manage variation as a system?

And what he articulated Says that they do what [01:38:00] we talk about. They mind that they look at the variation. They align things to get the clearance. They do that

John Willis: So this woman Katie Anderson She wrote a book called learning to lead lead and learn and it's basically four decades of it might have been the guy you met Is yeah, you know, dr.

Yoshino and he was the John shook part. He was the Toyota side of the John Shook. He was the guy that literally was the training and the facilitator and the mentor to John Shook for NUMMI, but he had like four decades at at Toyota. He hired in 66, so like he literally worked with Ono and Shingo. I mean, like, and I, I got to ask him, I got to interview him on a couple of questions, right?

And, you know, this question comes up quite a bit about like, was Deming, some of the East Coast tend to say that Deming had no influence on Toyota, which is just nonsense. And and I asked him, what was Deming's impact? He said, he told me that it was, you know, Deming's [01:39:00] impact in the sixties at Toyota was incredible.

And he said that it was, we didn't understand data. And I think what he meant to say is we didn't understand all the things you're talking about. We didn't understand, like, how to look at data, how to understand data, how to use data, you know, I think all that. But the other thing I was going to mention, she ran a study trip, which was just, I went on it last year and we went to Toyota doesn't allow plant visits since the pandemic, but we went to a lot of tier one and tier two suppliers.

It was incredible to be able to ask questions and like, you, you can learn so much from their suppliers. You know you know, it was, there was a lot of sort of soft skill stuff, but but yeah, there's like like to me the answer is probably not at Toyota, you know, in the. The, the, the nerve center, but probably with a lot of the tier one, two, two suppliers that you could spend a lot of time [01:40:00]with over there.

You know, if you have the sort of energy and the timing and all that to

Bill Bellows: get it again, but, but I would also say is what you, the, the questions you ask are guided by a theory. So Rocketdyne had developed a supplier that was responsible for an integrated piece of hardware and the hardware. Included a main injector for a combustion chamber for a rocket engine.

It's about a quarter of a million dollar housing. And on the top surfaces are 628 holes, roughly three tenths of an inch in diameter. So that, that injector housing is made and machined in Minneapolis by this company, I don't know their name. And then sent to another Rocketdyne supplier near Disneyland.

Recognized Supplier Grade in Los Angeles, they make tubes that go into the holes. That's what they do. And those tools, those [01:41:00] tubes, actually have a tube within the tube. So think of a tube with a straw down the middle. And so through, you know, so it's called a coaxial injector. And so through those will flow liquid oxygen and liquid hydrogen.

And those tubes get brazed into the holes. And so this, this company makes the coaxial injectors, the other one does the housing. And Rocketdyne at one point did both, and then outsourced both. And then got the two people to receive the injector plate. Put the tubes into the holes with a thin piece of braised foil around the tubes.

So imagine they're taking the tube, putting a thin piece of foil around it, inserting it into the plate. Not that I've seen this done. This is my, this is what I'm told to say. I have no idea how much tooling is involved, but they put all the, all the injectors in, and [01:42:00] then they send it down the street to this house that braises them all together.

And they braise in one cycle. They, they get every fillet top and bottom the just say for our listeners, the integration is incredible, which at Rocketdyne, we never knew how to do because we never minded the gap. Yeah, but we were able to explain to the two people how to mind the gap. I mean, so Rocketdyne decided to outsource this idea was to go to them, which they at the time did not think was possible.

But, but our manufacturing geniuses went down there, explained to them, you focus on the clearance, not the parts, and you will be able to do this and held their hand. And they're doing it year upon year upon year. And so maybe 10 years ago, I'm mentoring some young guys, young engineers, helping them, trying to help them do more of this within Rocketdyne.[01:43:00]

And part of it was, was having them go visit this place and listen, find out what they're doing. How are they minding the gap? It wasn't quite, there were some aspects weren't clear to me. So I had them come up with a list of questions guided by this theory. That they're minding the gap, right? And the reason I shared that story is that if you go into Toyota with a bunch of questions about how they manage parts, You'll have different questions.

Yeah, totally. If you go in there with a, if you see the world through a red pen lens, you'll have red pen questions. If you see the world through a Deming lens, you'll have blue pen questions. And so the questions, and that's what part of what the system of profound knowledge is about. It helps us frame how we see things and how we question.

And they came, and they came up with some brilliant questions and some brilliant answers. That just where they were [01:44:00] doing. We knew they were minding the gap. We just weren't sure how they were doing it, and they wouldn't give us a lot of what they were doing as proprietary, but we learned enough about what they were doing that the other details were pretty straightforward because I just there were a few few aspects that weren't making, but how are they doing this?

Right? And they but again, guided by the questions. And this is why, if you and I were to go visit Toyota, Toyota. In Japan and dig in, we would go there with a bunch of questions that I don't find in a whole bunch of articles written in the Harvard Business Review or any of these other places. They're not asking those questions because they're not coming from those places.

No, and I

John Willis: think, you know, you like, like, and we probably definitely need to wrap up, but the, you know, like we, in one of the previous podcasts, we talked about how few people, even when you asked to Taguchi, you know, who uses his methods, he was like, nobody or like two companies or like,

Bill Bellows: Hey, like [01:45:00] Fuji, Toya Nippon, Denza and Toyota is John Willis: the famous tier one Toyota supplier, which is interesting, but, but

yeah, I mean that, that, the point is like.

And that would be a fascinating discovery process, but the point being is, if, if Taguchi says very few people, I mean, how many people who know Deming, know Ackoff, knows, you know, all these people who from our, even our

podcasts and my, you know, well read, you know, like people have read, you know, gotten degrees in a lot of things other than just software You know, who really like, man, this Taguchi guy is amazing.

And when, how come I never heard of him, you know, so, so, you know, I, I, you know, again, why I'm, I love having these conversations because I think you're spot on, I think there is a lot of there, there with this idea of sort of Deming's notion of, of, you know, as you said, the, you know, production uses a system or a sort of a cybernetic feedback loop, if you will a system that creates it.

A [01:46:00] feedback loop combined with the sort of the idea of minding the gap or to Taguchi ideas is about to Taguchi loss function. I think that is, you know, as as everything that I've learned to date in my today's my birthday 65 years happy

birthday. Yeah. Thank you. You know, I think that makes more sense than anything else.

I've listened to or heard. So
Bill Bellows: Well, Go ahead. I was gonna

John Willis: say, give your final thoughts. But like I said, I'm still where I think we're getting close to two hours, which is not a problem. But yeah, go ahead and give me your final.

Bill Bellows: I was going to say is I I never asked Dr. Deming a question when I met him the first time in 1990, it was all new to me. And so much of what he said, what I was trying to do is match what he was saying to what I knew from Taguchi. And that isn't, that isn't a common cause of special costs.

Taguchi doesn't say that way. Control charge. That isn't, I just walked out of there. And the only thing he said that struck me was [01:47:00] he didn't like competition. And again, not between companies within an organization. So, but I would say then that when I began to appreciate the new economics, when it came out in a system of profound knowledge.

That meant so much to me, but when I, you know, to understand the variation aspect, well, then I can look at Deming's work, Shewhart's work, Taguchi's work, to understand systems, even Deming turned to Aikoff to understand

systems. And there's many people out there talking about systems in different ways.

You know, Dr. Deming was not a psychologist, but if you go off and listen to the work of, you know, Gerald Suarez, who used to speak at In 2 In, and there's, you know, others whose work was in, you know, and that can bring a lot to the table relative to intrinsic and extrinsic motivation. There's you know, many who have written about that.

So what I. [01:48:00] find invaluable about Deming's work is that he created a framework from which we can go off and study all the different thinkers on systems and bring it back and all the different people on variation and bring it back and this psychology and bring it back and the epistemology and bring it back.

And, and, and so the framework to me is invaluable and to see how these things influence one another. So I don't look at just one of them, but I, but then I can go off and look at the work of Michael Sandel and his book, the tyranny of merit. And get that much more out of it relative to what's wrong with merit.

I mean, Deming did not like meritocracy and right. But having read and studied Michael Sandel's work. You know, then it's like all the richer my appreciation of what's wrong with meritocracy. It's like thinking I caused the grade by myself and I did everything by myself. And, you know, whereas, you know, Barack Obama back in his era said, you know, you didn't, you know, John Willis, you didn't create this company.

Not that you [01:49:00] didn't create, contribute to it. To it, what are we saying is you benefited from the Internet. You benefited from a national, you know, from a health care system, you know, from all these things, the transportation system, a mail system. And so, and that's what I see what Deming's talking about is that.

You know, everything each of us have comes from many different contributions. What we do is contribute to a greater system. And that's what Deming is talking about. He, he was very humble and appreciating. That he contributed to the success of Japan. It wasn't just him. And that's, to me, that's an all straw mindset.

That, you know, I brought a straw to the table, you brought a straw to the table, and what can we do to help others?

John Willis: Yeah, yeah. And that goes back to the core of Sodorow. Yeah, I mean, again, I think we Anybody who's stayed on this long listening is like a fan of System of Profound Knowledge, so But all right, my friend.

[01:50:00] Yeah, we'll try to regroup and you know, just dig a little deeper on some more of this. It's just, it's awesome.

Bill Bellows: We'll do it again. Always a pleasure.
John Willis: You have a good weekend and we'll be in touch. Thank you, John.