Apollo Guidance Computer History Project

Third conference

November 30, 2001

On the Factory Floor

DAVID MINDELL: I mean I'm curious to talk a little bit more about the part of the floor, that you have pictures of. How many women would be working on the Apollo program and how big was this area?

ED BLONDIN: Probably on the ropes, up to 12 on a shift. In the working, welding modules, could have been 30, but they were scattered.

DAVE HANLEY: They didn’t only do that. When the people were making ICs and other parts, we made sure the same astronauts would go visit them.

ELDON HALL: They visited in lots of places.

ED BLONDIN: We might have had half a dozen or so. We had a special little group who were module repair in a little room by themselves. There was a lady named Mary Tangey who ran that. These potted modules, you tested them before you potted them and then you’d test them afterwards, post-pot test. Quite often they would fail because the potting stressed them, and then what would you do with this thing that you had invested all this time and money in, and all these components that were so hard to come by? So we would send them to Mary and Mary’s girls. There might have been three of them. They used dental tools to chip away the foam part until they exposed components. They would get directions from the test engineers, who would guess – well, better than a guess - which component had probably died. And they would expose it and then send it back and the test people would verify that. Then they would take that component out, which was a difficult thing to do. They were welded in there with strips of ribbon. Then they would put a new one in which meant you had to weld ribbons together and retest it and send it back for potting. And mostly they passed the second time around. I referred to the fact earlier that when I was put in charge of Apollo in Milwaukee, we were behind schedule. So one of the first things I did was climb on an airplane and come out to Raytheon and figure out if they were doing anything different than we were, because they ran the same program.

In those days, there were all these people around that would say the contract says this or NASA says that, and you could never verify it. So I figured if I got out and walked along the floor of Raytheon, I could talk to people who really knew things. So I asked, "What do you do when these things fail?" "Well, send them over to Mary." Well, I got back to Milwaukee, and I found out that somebody at AC had said, "No, no, no. That can’t be done without ruining the module. Throw them away." And so we had this large quantity of these things with hard to get components and heat sinks. All this money invested in them and they were going in the trash. So I came back and I said, "Aha! Same space craft, same contractor. They're doing it in Waltham." Well, our guys had to come up with an improvement on that. So instead of the dental picks, they went out and got a machine that a dentist uses to abrade teeth. It was called an SS White machine, I think. It would sandblast them with ground up walnut shells. That worked fine at AC. So they came out and said--

JACK POUNDSTONE: Didn’t that generate a static charge, as I recall?

ED BLONDIN: Yeah, they said, put it in at Raytheon, so we did and with the integrated circuits in the Raytheon modules, we were generating a static charge and zapped the ICs. And Raytheon took great pleasure in telling AC, "You're not so smart. This thing you made us do is blowing our ACs." We couldn’t get on schedule because we were throwing all these modules away. I mean, this was not advanced technology, dental picks and things that would dissolve the potting compound. People were saying, "What effect is that going to have on the components? That solvent?" And how Raytheon got away with it, nobody at General Motors seemed to know.

HERB BRISS: Eventually we used a very small soldering iron to melt the foam.

ED BLONDIN: Well, if they had known that you were using hot soldering irons

HERB BRISS: That didn’t work too well. And the ladies decided they’d just pick away. The ladies decided. They did a lot of deciding.

JACK POUNDSTONE: Yes. They did.

ED BLONDIN: That got us back on schedule. That and a lot of other things.

DAVE BATES: You know Battin’s article, "A funny thing happened on the way to the moon?" One of the references in there was to the independence of the little old ladies in Waltham.

DAVID MINDELL: What other kinds of things did they do, if you remember other things?

HERB BRISS: I don’t know whether it contributed to the program, but they were high seniority ratings. They knew what they were doing. They were old enough to be my mother. I was in my 30s then, but they just felt confident they knew what they were doing. When we moved on to microwave stuff, they worked on the tool division.

DAVID MINDELL: They were watchmakers. They were making--

ED BLONDIN: Not the ones in my group.

ED DUGGAN: Like the Westinghouse experience, that they go back to in the industrial engineering field, where they tell how they lowered the lights. Westinghouse did a famous experiment, you're probably all familiar with it. And the more they dimmed the lights down, the more output came out, even though they couldn’t see what the hell they were doing. And it was basically the attention they were getting. In management, here you are in a highly visible and patriotic almost crusade to get to the moon before the Russians did.

HERB BRISS: The cream of the crop-- You guys are talking about Polaris, it wasn’t known then. There was a program or two before that. They just knew this was a great thing to do.

JACK POUNDSTONE: One of the interesting things about the women that I remember was that we had this policy of the girls who worked on ropes, that’s all they could do because they had got good at doing ropes and we didn’t want them to go off doing a bunch of other things and not be able to get them when you needed ropes. I remember when we got to that phrase of the program, we had three or four weeks to make a set of ropes.. We had 12 rope machines or something like that. I forget how many. As I recall, we paid those women to sit there and wait until the deck of cards or the tape came out. And they would be sitting there knitting for, you know, two or three weeks. And then a deck of cards would come out, or the program would come out and then they’d just go like hell.

DAVE HANLEY: On that subject, remember when they found a software problem at the very last minute and they had to redo a rope. And NASA went trying to find out how long it would take to redo the rope and so forth. And every rope you ever had had so many failures afterwards. And then it went into rework and so forth

ED BLONDIN: I was running manufacturing on the Apollo program at the time. For some reason or other, the ropes that had already been made had to be reprogrammed.

JACK POUNDSTONE: Software glitch, yes?

ED BLONDIN: There was some reason, and I don’t know what it was, and we were down to normal cycle on those things, which was, with all the rework, eight weeks, something like that. These ropes were required in less than three, I remember that. And none of them failed.

DAVE HANLEY: That's the point I want to make to these fellows. Those ropes were fabricated on the first time, and there were no failures and that’s the first time they ever had no failures.

ED BLONDIN: They just paid more attention.

DAVE HANLEY: And the gals used to fight and argue, you know, "Okay, you’ve had it for so many hours," and this sort of emphasizes the fact that people do identify with it, and it really worked.

ED BLONDIN: I remember it very well. Nobody would believe that it could be done. We had all kinds of people. I remember production control saying, "Well, we’ll put up charts that show what the schedule points are and hope the ladies see." They happily raced through that, and they did it perfectly the first time. We shipped right on time.

DAVE HANLEY: I think you were ahead of time.

ED DUGGAN: What you were referring to was called the Hawthorne effect. It came from Western Electric’s Hawthorne plant, outside of Chicago. A famous experiment was brought up in Industrial Psychology where they started off to see if increased lighting would increase production. So they needed a control group. They put the control group there, kept the lighting the same, stepped up the lighting on the test group and production increased on both. And they did it again, upped the lighting again, and it increased on both of them again. Then they started lowering the light on the control until they got it down where you could barely see, and production kept going up on both of them. The scientific conclusion from that was that if you give people attention, it’s more important than the lighting or anything else. If they think what they’re doing, for instance, is critical.

CLINE FRASIER: There’s one other thing about that set of experiments that doesn’t get reported in all the management literature and the management classes. But it is in the original papers from the early ‘30s, and that is the way people got paid in the big plant was there was incentive pay and that it was based on the output of large groups of people. So if they had big output, then they got even more. They put them in this special room, and they put production measurements on every single machine and they’d change them to individual periods.

BARD TURNER: There were time motion studies that that show at least in my experience in both the industrial one and also at the lab the importance of project-oriented groups rather than matrix. I have never seen a matrix organization work.

DAVE BATES: I got a paper I'll send you. It was done by my boss up at GE in Pittsfield, and it was put in just that way. It’s the people, not the organization. It’s only a one and a half page thing, but it succinctly points out the fact that you can put any organization you want in there and they’ll either be successful or they’ll fail. But the main thing is to get the people to work together and to determine what is the purpose and work as a team.

ED BLONDIN: That was another thing we had to do in management. We had to recognize when to break up a program team. And they said, "Well, why would you want to do that?" Well, whenever you introduced a new program coming out of engineering, it was much tougher to handle than a program that had been shaken down through the years. Much tougher. And with the program orientation, all the varsity players are on the old program, and to them it became second nature. Everybody was way, way, way up the learning curve. Then you put all the brand new, green people on this terribly difficult thing to build out of engineering where someone needed to be able to make a judgment that this is unbuildable and have it stick. To have an operator do that, she had to have, or he had to have, a tremendous amount of prestige. And the only ones who had that prestige were the ones in the old programs. So it didn’t take me long, as a manager of a plant with about 17 programs, to start taking the people out of the old programs.

HERB BRISS: Well, there's a certain amount of pride, too, they got out of working on a prototype program.

ED BLONDIN: Absolutely. Absolutely.

ED DUGGAN: There’s one other point, and it’s off the personal relation, which I think is paramount in any kind of team activity. But in a lot of these things we described there, we keep referring to materials that we’d change or that would cause us problems, or things that we’d do that would involve materials. Personally, I thought we had the world beat when we figured out we could weld two wires together. You know, but the material science aspect of this thing began to take on extraordinary sophistication and the level of sophistication moved up with every program I've been on since. Plated wire memory was a good, classic example of what moves through that collective era where there was a good portion of us learned. Where a lot of the development was now beyond the area of electronic or the packaging guy and the shop and everything else that was in those little cells that were doing the plating. And a good one was on the core rope when we got RTV in. Remember, if you over-welded that thing, the heat would expand it and bang go the wires. And this got to be a major kind of problem for a while. It got diagnosed. I remember I learned that you don’t drill a hole tight and then fill it up with potting compound and let it just pressurize itself because strange things happen and it expands. And respect for thermal control was another serious thing in electronics. Where does the heat go and how do you get it out of here? And then God bless those connectors, they were a major pain in the neck.

Next: Innovations in Management

Home: Conference 3