Apollo Guidance Computer History Project

First conference

July 27, 2001 

Integration Issues

DAVID MINDELL: While we have Joe [Gavin] here, it seems opportune to talk about the relationship between the Instrumentation Lab and Grumman, vis-à-vis the integration and the specification issues around the computer.

We've heard little pieces of that, but I would be curious to hear -- I mean, not least of which because system integration is a very interesting, somewhat emerging practice at this time. So I'm curious to hear about either in the trenches level or at the management level, what those relationships were like.

DAVE HOAG: Early on I was involved in negotiating interfaces, first with North American, and then with Grumman. And at North American, they had a team of maybe 20 year olds. It seemed to me that I could walk all over them. I had to tell them what the questions they should be asking me, and what did they really need to know so they could design their side to the guidance side. It worked out all right. 

Then came Grumman about a year later. And interfacing with Grumman, they had a sharp team. I met my metal there. There was one big problem there, and that was we already negotiated with North American some of our side of the interface, so it fit North American hardware and software. Now we hit Grumman, and they wanted to do something different. Like they should have had the same oxygen scrubber in the LM as in the command module, things like that. We had the standard hardware. (Hugh Blair-Smith adds: I suppose this refers to the LiOH canisters for scrubbing the carbon dioxide, where incomplete standardization made such a problem in Apollo 13.) 

But it was a strong team, a really strong team at that juncture. That's where I was working. Later on, other people got involved. I enjoyed that part of the stuff, but it was a struggle.

CLINE FRASIER: One impression was that Grumman was a little bit easier on the interfaces than was Marshall, with the interfaces of the booster.

DAVE HOAG: I don't know about that.

CLINE FRASIER: Remember? We had to do the interfaces from the guidance system down so you could control the booster on the launch? And it seemed to me that was somewhat more resistant to change and had to get sorted out.

DAVE HOAG: This is amazing. We had one standard computer go either place, and one standard inertial system go either place.

CLINE FRASIER: Part of that was because NASA insisted. Because Grumman wanted to have a different shaped display, they wanted to have different this and that. And I think it was Joe Shea who laid the law down on that. 

ELDON HALL: That was very important to the computer. So before that dictation came down, we were arguing with North American about the computer interfaces. Then we'd go to Grumman and argue with them, and they wanted a whole different set.

But NASA kept saying they wanted a common computer. And there were these two giants beating us over the head from both sides. It was impossible. If NASA hadn't taken a strong stand, we would be sitting here today without ever getting to the moon, because it was just completely impossible to work with the two space craft contractors.

FRED MARTIN: I can give you a bird's eye view from the trenches early on. I was the person who worked the CSM, so I don't know anything about the LM, except that we were absolutely sure that it would never work. (Laughter) That's all we knew about it.

I say that facetiously. I'm sure there were people in Draper who spent their time working the LM systems and were sure that it would work. But it was interesting. There was kind of a divide amongst the troops, you might say.

It was like if you're a Yankee fan or a Red Sox fan. You had your CSM folks, and you had your LM folks and they sort of, I don't want to say there was a rivalry. There was a lot of skepticism, let's say, about whether the LM would. It was sort of  a funny looking thing, but it worked.

JOE GAVIN: I can give you a brief view of it from where I saw it.

Of course, we were starting a year later because of the lunar orbit rendezvous decision. So we were always, always late. Let's face it, we were always late.

Now, we had the feeling, for better or worse, that NASA thought they were hiring a job shop to carry out their design of a LM. We very clearly understood, my god, we were going to design and build a LM that we had our name on, and we would argue every inch of the way, and we did argue every inch of the way.

DAVID MINDELL: Did NASA have a design that was viable? 

JOE GAVIN: Well, no. They had a notion of a design, but all we had was a notion of a design. But for the first two and a half years of the contract, we were defining what was then the design. We quickly became known as difficult. At the same time, the MIT group, under Milt Tragesor -- now Milt was a very aggressive guy. He would have been happy to have us ship the flight article to Milt, and he would install the guidance system and guarantee that everything would work.

We took the view that that isn't how it works. We're the integrator of this vehicle. We will do all of the testing to make sure that everything is compatible, and we will have the responsibility to say at the Cape that we're ready to go or we're not ready to go. And that led to a certain amount of debate, but it got straightened out. So there was a turf war in the beginning in practically all directions.

I think it was interesting how it worked out. I think that after perhaps by 1966, I think we all had our roles pretty well understood. From there on, it was a matter of grinding out the details.

CLINE FRASIER: From the standpoint of the whole program, people mentioned how amazing it was that we got all of these missions off. We didn't lose anybody in flight. I truly believe that the reason for that and the reason for the success is that we had all of these competing views. By design or accident, there was a structure so people could fight out the stuff and there was competition so that stuff got brought out that in today's environment would get swept under the rug and then cause problems.

I could go through, I could talk probably for days about those examples of those kinds of events. And the same thing you had between MIT and Grumman and MIT and North American, you had the same thing between MIT and General Motors, because General Motors had the contract to build the stuff MIT designed, and to get it all integrated together and make it work.

There was, I don't think it's much of an exaggeration to say that when something didn't work, or GM put together a computer, or Raytheon put together a computer and tested it and it didn't work, then there was a view from Cambridge that if those guys would have just built it like we told them to, it would work. And there was a view from the contractor that the people in the ivory tower, if they just turned us loose, we could adjust the design so it would really work.

You know, there are just lots of those, and the classic of that happened to be in, or one of the classics, was in the gyroscope world when we were having all kinds of trouble with gyroscope bearings. The folks at the Instrumentation Lab would say "If AC would just do what we tell them on how to do the bearings it would work." And the same story back. So at that point, I think my boss probably had the wisdom to say "Okay MIT, you go build so many sets, and AC, you're free to do anything you want to in the design of the bearing. Of course it still has to fit and function. You make any changes you want to, and then you build something and we'll see."

Well, as it turned out, neither group could build one that worked. But out of that set of experiments, we did get a design that did work. (Laughter)

But that created a healthy tension. And what it forced NASA into was when you couldn't get some technical conclusion, to bring in outside technical resources which sometimes threw the contractors together to decide that they could agree, or something. But it happened.

What I would do, I can still remember a couple of things. I couldn't figure out what was going on. I didn't know enough about what was going on in the bowels of the computer, or whatever. 

And NASA had this structure that I could make a phone call to Washington and kind of say I need the best people in the country to work on, this thing I can think of happens to be diode channeling. So we could then have a meeting with Stu Peck and Connie Zierdt at Bell Labs who were the best experts, and somebody from GM's research lab. They'd pull all of that information, and we'd get problems solved, and we could do all of that like that. (Snaps fingers) The problem could come up. You could go through all of this in a week. You just can't do that anymore.

JOE GAVIN: One of the things that George Lowe put in motion, which I think made it possible to meet the schedule we did, was set up a series of meetings, I think beginning in, '66, I'm not too sure. But he would take his leadership and bring them to Grumman for a meeting. We would go over all proposed changes.

The next day, that group would be out at North American, and somebody from Grumman would go with them. In two days, all of the pending changes throughout the spacecraft system would be settled right then. It was a very successful system.

JOHN TYLKO: Can I ask about the Tom Kelly book where he refers to the debate over the reliability of the guidance system that became a very tense meeting between MIT, and Grumman and Joe Shea. Several of you, I think, were participants in that meeting.

CLINE FRASIER: I was not, but it was an accurate portrayal of Joe Shea. I had been in meetings like that with Joe.

JOHN TYLKO: Could you guys describe that challenge? 

JOE GAVIN: Well, yes. The Kelly book, his description of the meeting is really quite accurate. We had, I should explain that Draper, Doc Draper was my thesis advisor. I have a preconceived tilt, you might say.

But in any event, at the time we did not have access to all the information about reliability that Draper had. We did have the requirement to build the abort guidance system, and we got a lot of information from would be suppliers. At a later date, and I really can't -- this is a rumor now I'm going to repeat, because I can't pin it down -- but I think that we were egged on by some of the middle level NASA people to pursue this question of reliability to force Draper to come up and put it on the table as to what they were going to do.

And so we went to this meeting in the neo-Aztec (?) building down there in Houston. I'll never forget the day. What happened, of course, was that Davey had a bunch of facts that we did not have. He wound up showing what they could do was perhaps not quite as good as what they had advertised, but it was a darn sight better than what we were saying they could probably do.

So what could I do at the end of the day, other than to say I'm sorry we caused all of this trouble, but at least we know where we stand. This is one of the many incidents where I had to eat crow. Kelly's account of that is pretty good. 

JOHN TYLKO: Any reflections, on the Draper side, of that meeting?

DAVE HOAG: Not for me. (Laughter) 

JOE GAVIN: He [Dave Hoag] presented the Draper case.

JOHN TYLKO: Do you remember the meeting? 

DAVE HOAG: Yes. I remember that, but my memory at my age is failing and I can't add any more details. I might think of them, but it would take me time.

JOE GAVIN: But we all survived the meeting, although at the time it was pretty humbling.

DAVE HOAG: It was a very dramatic forecast of how bad the liability would be, and we were incensed. ... (inaudible) We worked hard on that subject.

JOE GAVIN: They had a lot of information from Titan that we did not have. There were some personalities involved.

FRED MARTIN: I'd like to make  a comment on leadership. I think that one of the things that you could see in the Apollo program up and down the line is leadership. However from where I stood, I didn't have as big a view as some of these other folks did of what was going on amongst a lot of contractors, and some of the negotiations that were going on.

But from our standpoint and mine, I thought that the leadership in that program was personified in Chris Kraft, at least at the time that I knew him, and the way he was running that program. 

It was somebody  you looked up to, and had every confidence that the decisions he was making were correct. He was fair. He confronted people when he had to. He didn't let people get away with things. He didn't coddle the astronauts, for example, if they wanted things that he thought were out of line.

I know that at MIT, people had a real affinity for him. They really felt that they were delivering to Kraft. I was going to say this previously, they certainly wanted to please and provide the astronauts with everything that they really needed, but they sort of were working for Kraft. I mean, they really focused, and when you went to a meeting, you really looked to him. I think he really delivered on leadership.

Agenda for future meetings

site last updated 12-08-2002 by Alexander Brown