TAPE 1, SIDE 1

MR. COLLINS: Last time you ran through in detail the various guidance computers and guidance system contractors for the various missile systems that TRW handled at that time. What I was unclear about in these very early systems is the kinds of functions that the guidance computer was handling.

DR. BURNETT: Well, it varied. The earliest one I worked on was the ARMA guidance system, which was an all-inertial system that was to be used on the Thor and as a backup on the Atlas. The backup the radio inertial guidance system General Electric [GE] had, and the computer in that case, was a form of a digital computer that doesn't exist any more. The memory was acoustical delay lines. In essence, you have a bunch of bits going through a quartz delay line, and the programming trick there was to be sure to capture the bits at the right time when it came out the other end of the delay line, to do something with them in the arithmetic unit.

    The functions that had to be done on a computer like that were absolutely bare minimum, because it's so complicated to build and to implement. So about all that computer would do would be to make the inertial management unit work itself, you know, erect the IMU (inertial measuremet unit), and then implement the guidance equations, which were little more than, how do you steer? Where do you cut off? Where do you need to cut off the thrust so that you hit the intended target? A lot of work was put in to simplify the guidance equations to just the absolute bare minimum number of terms to do that. The nature of the guidance equations was such that there was no closed form solution so we usually wound up with a Taylor series expansion about the cutoff point. The issue was, how few terms did you use in that Taylor series to get the required accuracy? It was just bare bones guidance.

    Let me contrast that with what we did on the Minuteman program. By the time we got to Minuteman, we had memory to burn, relatively speaking. It was easier to come by. In fact, it was a disk memory, not too unlike what I have in a computer here. See, the big difference in the technology was that we had more memory. You know, quartz loop type of memory is difficult to build, difficult to control, and you don't want very much of it, you want the minimum amount you can have. By the time we got to Minuteman, we had disk type memory, which had been invented by what's now Rockwell, then it was North American, the Autonetics Division of North American. So we could put a general purpose computer together with a fair amount of memory. It was the equivalent of the magnetic drum memory, but it was still a lot of memory. So if you had enough memory, you can not only do the guidance equations neatly, you can also seriously consider doing all the autopilot equations too. So the big difference there was we decided we'd close not only the guidance loop through the guidance computer, we also had the guidance computer do the control loop for the whole missile. So we were able to eliminate, if you will, one complete subsystem on a ballistic missile by getting rid of the separate control system. That theme just progressed on and on through the ballistic missile program. That was a big technological change.

COLLINS: At this early point when you were working with either the Titan or the Minuteman, was there ever a consideration of a fully guided missile, or would that just be technology beyond what was required for the job?

BURNETT: No, we always considered then a fully guided missile, but we always thought of it as two subsystems. You have a guidance subsystem, be it either inertial or radio guidance system, and you have a control system that's separate. The control system keeps the attitude stable, because a ballistic missile is inherently unstable aerodynamically, and then the guidance system would tell it where to go. So the great simplification was, we combined the two into one subsystem, and eliminated a whole lot of hardware that you had to put on the missile.

COLLINS: I was thinking, fully guided to the target.

BURNETT: Well, we thought about it, but we didn't want to do it for a very simple reason. We were concerned about the vulnerability of the missile in guided flight, in terms of what vulnerabilities would we leave in the system that the enemy could exploit. If you had a fully guided radio guidance system, there's always the chance or the probability that the other guy could find out what the frequencies are, what the code was to get into the system, could build his own antennas and transmitters, code generators, and capture the missile, if you will, in powered flight or in flight, and deflect it to some place where he wanted it to go. So that's really the reason why radio guidance never went any place in the ballistic missile program. We finally convinced ourselves that that was too great a vulnerability.

    That's why we concentrated on inertial systems, because in an inertial system, unless you can change the gravity or rotation rate of the earth or something odd like that, there are no vulnerabilities you can grab hold of in the guidance system. We thought in the early days about, why don't we take the guidance system and put it in the re-entry vehicle, and then use that to steer out some of the re-entry vehicle error. We didn't do that for a couple of reasons. One of them was technological. You've got to have a 3 axis set of gyros controlling an inertial platform which has a 3 axis set of accelterometers. Further, the gyros can't drift very much. Usually if the gyros have a drift rate it's fairly constant, so the longer they're on and operating, the more they're going to drift off axis and hence degrade accuracy.

    Most ballistic missiles get done with powered flight in about three minutes, so our problem was to keep them from drifting too far in three minutes. That's a lot simpler problem than to keep gyros from drifting too far, in say, 30 minutes which is the approximate flight time. We have guidance systems or gyros today that are good enough that you wouldn't worry about the 30 minutes. But we weren't sure in the early ballistic missile days about that, so we wanted to use the gyros as short a time as possible. So after we did all the calculations, and figured out how accurate we could make one of these things, with the inertial state of the art, we decided, hey, that's good enough to kill the targets, so let's just forget about the the terminal guidance.

    Now, there have been some flight experiments done in the last few years, in what we call the ABRES program, which is the Advanced Ballistic Re-entry Systems Program, to actually do maneuvering guidance on the terminal end. We know how to do that. But in terms of what we're trying to do with strategic forces, the kinds of weapons we have on the missile, the kind of targets the Soviets have, there's no real need for terminal guidance. It could be done. We've done experiments on it. It would be a lot more expensive. So we just haven't pursued that.

COLLINS: Did I understand from our discussion correctly last time that ARMA both produced a computer and a guidance system?

BURNETT: Right. They produced this quartz delay line computer. In fact, I think there's one in the Smithsonian, if I remember right. They also produced the complete inertial mwasurement unit with the 3 axes of gyro control and 3 axes of accelerometers.

COLLINS: I'll have to check that.

BURNETT: I think I've seen one there. I know the original Burroughs computer that we put on the GE guidance system is in the Smithsonian. That was another very interesting machine. That's one that used a core memory, and the way we programmed it is, if you can think of a word as being 32 bits long, you've got 32 cores in a row, and you thread a wire through the cores. If you want it to be a zero, you go through the core; if you wanted a one you would go around the core. That's real programming! That's without any operating system. That's all done in machine language.

COLLINS: I guess we talked a little bit about the systems for Atlas and Thor, but not really too much about Titan, in your capacity as head of the guidance and control section.

BURNETT: I spent as much time on Titan as I did on any others. The Titan guidance contractor was Bell Labs. In fact, the project manager for the guidance system was a fellow by the name of Joe Shea, who's become famous through Apollo days. In fact, he's just recently retired from Raytheon and is an adjunct professor at MIT now. He was the project manager. It was a radio inertial guidance system. It had a radar on the ground. It had an inertial platform in the missile. That's before we knew an awful lot about either one. In fact, it was the original backup to the GE guidance system, which had the radar on the ground and a bit of a platform in the missile. It turned out, it was a very fine guidance system. Let's see, what was the backup inertial, was that an ARMA? No, the backup all-inertial system for it was made by A.C. Sparkplug, and the A.C. Sparkplug guidance system is the one that's evolved and evolved and is currently the guidance system that's used in the Titan program. In fact, that's evolved into the all-inertial system that's used on a bunch of the commercial airliners too.

COLLINS: This is the [Charles Stark] Draper influence.

BURNETT: Right. Yes, the Draper design was the one that A.C. Sparkplug put together, and it's an original Draperian design, with single degree of freedom gyros and single degree of freedom accelerometers, and as I recall, a beryllium platform, and what A.C. did was to build the thing, based on a complete Draper design.

COLLINS: We discussed Draper's role in the evolution of guidance systems fairly extensively last time, and you made a comment that I thought was very interesting. I don't know how much significance to attach to it. You simply said that in the end, Draper guidance system designs won out.

BURNETT: Oh--almost.

COLLINS: I guess I'm curious to have you elaborate on that a little bit.

BURNETT: Well, let's see. Let me go and talk about the Autonetics guidance system that was on the original Minuteman I. It had this general purpose digital computer in it, which we used for the control system. It had gas-bearing two degree of freedom gyros in it, which were quite an innovation at the time. This is what we'd call an absolutely non-Draperian gyroscope design.

COLLINS: What's the difference, if you can?

BURNETT: Well, for a platform that I need three axes of control, I can get by with two gyros instead of the three gyros if each gyro has 2 degrees of freedom. All of Draper's gyros were floated in fluorolube, which is an oily viscous stuff that you heat up and the mechanical parts actually float in this stuff, to eliminate friction. In the Autonetics design, they used a gas bearing. We use the aerodynamic force and a little propeller to keep the rotating mechanism essentially suspended in a gas cavity. So the original Autonetics guidance system had the two degree of freedom gyros from Autonetics, unique design, and they had a velocity meter, which was an accelerometer that had the force balancing element being a permanent magnet.

    In a Draperian accelerometer, called a PIGA, Pendulous Integrating Gyroscope Assembly, it is basically a peculiar form of a single degree of freedom gyro with a known weight on it, offset or pendulous. That was the original Minuteman I guidance system. Well, that worked pretty well. When we did some more systems studies in-house, we discovered that if we used the Autonetics platform, and the two degree of freedom gyros and their computer, and if we used the Draperian PIGA accelerometers, we'd get much better performance. So over violent objections from Autonetics, we made them do that. So the Minuteman II guidance system became the two degree of freedom gas bearing gyros plus the PIGAs. And that's also what we did on Minuteman III.

    Now comes along MX, and we wanted just the absolute best accuracy we could possibly get for MX. We again asked, what can you do with the Autonetics approach to the guidance system? We came to the conclusion, there's no way you can make that platform any more accurate. It has certain inherent flaws when you get the acceleration vector going through it, and although the gyros are very good, they're not quite as good as what you can do with the Draperian gyros. On the MX system, we torqued over and went into a completely Draperian system. By that time it had evolved the platform into a ball that's floated in fluorolube, that has three single degree of freedom gyros on it and three PIGAs on it, and we decided that that was the best performance that you could get, guidance-wise, so that's what's on the MX. So in the end, Autonetics lost out in the guidance battle and Draper won.

COLLINS: Because it provided superior performance?

BURNETT: It provided superior performance. The performance that we got out of the original Autonetics Minuteman I, II and III was comparable to what A.C. Sparkplug was getting out of their system. It was a lot cheaper. It cost less and it had a lot more reliability.

COLLINS: The Autonetics version.

BURNETT: The Autonetics version. We could turn an Autonetics guidance system on, it would run for a year and a half or two years before it would have a failure, and for that complicated set of electronics and gyros and stuff, that's very phemonenal performance.

COLLINS: Didn't Draper do the guidance for the Polaris system?

BURNETT: Yes, he did.

COLLINS: Did they have different reliability requirements than Minuteman?

BURNETT: Well, different requirements. In Minuteman, all we wanted to do was have those missiles, all thousand of them, at basically T minus 30 seconds and holding. So you had to have the guidance system up and operating and ready to go in 30 seconds. That was not the requirement on the Polaris system. The Polaris system, in the early days, you're talking about several hours before you'd eventually tell the skipper, "Hey, we're going to launch," because of the communcations problem with the submarine. Whether or not he launched quickly or not is kind of unimportant because the Soviets don't know where he is and they're not going to target him. Whereas they did know where the Minuteman silos were and they were going to target those, so you want to get those missiles out fast.

    So the Navy has plenty of time to turn the thing on, warm up the gyros, get the platform erected and aligned, get the guidance parameters in the computer. Well, the Navy has to know exactly where he is before he launches, and that's one of the sets of data you have to put in the guidance system. So if it took three or four hours to launch the missile, it didn't matter. In our case, we really wanted them out of the silos in about 30 to 45 seconds. So we had a different set of requirements in terms of reliability and the reaction time. Autonetics was the outfit that had the stuff that would do that. Now, Draper knew that, and they spent a lot of time with us, to know what Autonetics was doing. They were just simply able to improve on the Autonetics concept with some very bright people, to give us the inertial management unit we now have on the MX system.

COLLINS: What personal contact did you have with Draper during this period, if any?

BURNETT: Oh, quite a bit. I knew Stark quite well. Bob Duffy, who took over after Draper theoretically retired, was in the ballistic missile office and he was in charge of guidance. In fact, he and I worked very closely together when he was at Space Division or what's now Ballistic Systems Division, because he was the Air Force chief of guidance. So over the years I've gotten well acquainted with a lot of the fellows at Draper Lab. A bunch of them are retired now, but we did a lot of work back and forth on what should the guidance equations look like. They did a lot of very excellent theoretical work on guidance equations. What we tried to do was to, not simplify them particularly, but make them as suitable as possible for computation. There's a lot of trading back and forth. In fact, the government encouraged it.

COLLINS: One of my colleagues is doing a comparative study of Draper Labs and APL, to look at two postwar laboratories, and one of the things he's uncovered about Draper's work is that he very consciously sought to establish a very specific kind of way of talking about his gyroscopes, that was distinctive from the work that other people were doing.

BURNETT: Well, that's absolutely correct. He had his own jargon for talking about inertial guidance, and I can't remember all the terms, but he was always concerned about gyro drift rate. Now, most people would think of it in terms of degrees per hour, but he used the term "MERU," Mili-Earth-Rate Unit. He measured in terms of a thousandth of the rate of turning of the earth. Every one of the guidance terms he used were absolutely unique in that regard. You could always tell if it was a Draperian system by just the nomenclature. It was absolutely unique--but Doc was very proud of that. He just absolutely strove for perfection. Every time you'd see Doc, he'd get out his wallet and he'd pull out a piece of a strip chart, and he'd show you the latest drift data, from the latest model of gyroscope he'd just invented. You know, he was just absolutely striving for perfection. He was a zealot on that. That's why the system was so good.

COLLINS: How did this affect someone else's ability to communicate with him, your ability to communicate?

BURNETT: Oh, no real problem. There was a logic behind his system. I can't remember now, I can't quote you the details. I haven't thought about it in a long time. You know, 15 minutes with him, you're talking the same language. He was very familiar with what we were doing and terms we were using. He'd always correct us, of course, in his own way. No, that truly was an exceptional organization, from a guidance point of view, exceptionally good organization.

COLLINS: We also talked about your technical direction meetings. When you were head of the section of TRW, did you conduct the technical direction meetings for all the guidance and control, guidance systems?

BURNETT: I'd go around to ARMA or Autonetics or wherever and conduct a TD meeting. That's a thing of the past these days. But they used to be pretty hot and heavy. You had several forces at work in the meeting. You had the other corporation, and they had their position and what they wanted to do. You had the STL, TRW who not only knew what they wanted to do on the guidance system, but they also knew how it would fit into the rest of the weapon system. You had the Air Force Chief of Guidance, who really was responsible for the money. I mean, he's the guy that had to go get the money to pay for all this stuff, and he neither wanted TRW nor Autonetics nor ARMA to spend him into oblivion. Then you also had the contracting side of the house, called BMO (Ballistic Missile Office) who was there in the meeting to make sure that whatever was decided in the meeting was put into the contract immediately. I mean, when we had a technical direction meeting and you tell Autonetics, "Look, you're going to take that differential digital analyzer out of your guidance system and you're going to put in your general purpose computer that has that rotating disk memory in it," I mean, he's the guy that's got to put that in the contract, and Bob Duffy, who happened to be the chief of guidance at the time, is the guy who's got to be sure he's got the money to cover this. He also wants to be sure that TRW is not making some dumb mistake technically when they do that, because the contractor is just going to sit there and just argue and argue about it, which they did. But those are the kinds of issues you dealt with at a meeting like that.

COLLINS: So on the Air Force side, the person who worried about the contract, the language, the specifications, and the person who worried about fiscal matters were two different people?

BURNETT: Two different people, right. So you had a very interesting set of social dynamics, to get all those people in gear, because the last thing you wanted to do was have a big argument between the contract people, the financial people, and TRW, in front of the contractor. Not that that didn't occur occasionally.

COLLINS: So was the TRW role and your role specifically in these meetings to be the one who tried to seek harmony and resolution?

BURNETT: Right. Well, we were trying to get decisions that would make the whole weapon system go forward and be good enough to meet the specifications but not overdone. All kinds of issues would come up in those meetings. I don't know how I lived through some of them, now that I think about it.

COLLINS: Did most every problem that required some kind of action by the contractor get translated into a contract change? Or was it sometimes more informal than that, just "Yeah, I'll go do that and fix that," or was it something that was always fixed?

BURNETT: Well, there's a degree of informality, because these were cost plus fixed fee contracts. So the contractor knew he was going to get paid for what we told him to do, assuming the government didn't run out of money, so he never worried about that end of it. Where he would argue is that it's out of scope of the contract and therefore I need more profit, I need more fee, and that was mostly an argument that he carried on with the Air Force, not with us. We didn't get involved with the negotiations. We would get asked questions, or render an opinion as to whether or not what they were going to charge for that particular job was that reasonable or not. So we did that kind of cost analysis for the Air Force, but that's more what you might call engineering estimation than contract negotiation.

COLLINS: I guess I'm thinking more about the accountability issue. It seems like the contract was being used, or the writing changes into the contract was an issue of holding the contractor accountable for getting specific work done.

BURNETT: Well, partly. All the contracts had a clause in them which said that TRW was the technical director of the contract and they could issue technical directives to tell them what to do, content-wise. There was also a changes clause in the contract that says that when the government changes a contract, because of a technical directive or whatever, the contractor is entitled to sit down and negotiate the cost and the profit on that. So we were just playing a piece of the role of how the contract got done. Some of the things we did were to write the overall specification that went on the contract. In fact, that's one of the key things we did, because that spec that we put on a contract had to fit with the rest of the weapon system to achieve the objective. But there are a lot of things that go beyond what's in the spec that have to be done to make the thing work. You could put in a spec that the drift rate should be no more than so many degrees per hour. Then the question is, what is the actual hardware drift rate, how many degrees per hour is that? Well, you've got to go run some tests. Well, what do the tests look like? How do you instrument the tests? How do you reduce the data so that you can believe some numbers? We got into all kinds of battles in TD meetings over how you did the testing, how you did the data analysis, because you wanted to be sure that you knew what you were talking about numerically rather than--yeah, it looks like it doesn't drift any. We were pretty rigorous on how to do the testing.

COLLINS: So if I can extrapolate a little bit from what you're saying, the circumstances under which you would issue a technical direction order, you tried to back it up fairly substantially with some series of empirical information.

BURNETT: Oh yes. Absolutely. In fact, you also tried to be sure that the contractor understood exactly why you were doing it so he would go implement it correctly. Lots and lots of informal conversation on, what do you mean by this? What actually happened was that you started forming teams of individuals, regardless of what their organizational badge was, so there was a kind of informal team that would get together, say for example with Autonetics, and review what their current problems were. The team would have some Air Force people, have some TRW people on it. In the case of Minuteman it might even have some Boeing people on it, who wanted to be sure they understood how it worked because they had to put it together in a manufacturing environment. Those teams would come to conclusions, and from time to time you'd have to go change the contract to make things match. But we'd use the technical directive route to get the contract changed, make the matches to what Boeing was going to do, say, in the all ordnance destruct system that we put on the test missiles, because that had to interface with the guidance system.

COLLINS: Right. Within the framework of the contract relationship, how was all this testing activity defined and paid for?

BURNETT: Well, there was a task in the contract that said they had to do certain testing, and it was paid for through the contract. As a practical matter, what we would do, first of all, you'd have a specification, what the system ought to do, and then we'd have what we called preliminary design reviews and critical design reviews and finally hardware reviews. As part of the process, when you got to the hardware reviews, we'd also define a test specification. What test do you have to run to show that you've actually met the up front specification on the system? So that would be a separate document, which would be put on their contract, which they then had to cost out in terms of all the activity it took to do that testing operation. We could never write down perfectly in a document really what ought to be done in a testing operation, so you'd run a series of tests and we'd have a TD meeting and they'd come in and they'd present the test data, and we'd argue with them, and you'd decide, well, you know, there's some new tests that need to be run because we've overlooked something, and they'd agree, when it was over, what then ought to be done. So you'd say, OK, go run the new tests. That's when the contracting officer would have to change the contract, see to it they got paid for it. We'd have to go back and change the test specification to incorporate those new tests. But that was just part of how we did the business. We could do it very fast. You could get the contract changed in an afternoon.

COLLINS: One other thing that came out of my re-reading of the previous transcript, and I think it fits into our discussion here, was that at one point, and I'm not exactly sure when you would identify this point occurred, you felt you could essentially ask for "inventions to order," is the way that you phrased it in our previous discussion; that there was somehow some critical mass of creative activity and knowledge about the job, that you could fairly carefully plan out advances and expect to achieve them.

BURNETT: That's absolutely correct.

COLLINS: Does that tie in with how you were describing how these technical direction meetings worked?

BURNETT: The technical direction part of it wasn't the source of the invention. That was just a mechanism to get what you might call the contractual stuff out of the way fast, so that the engineers could go do their thing without spending an awful lot of time on the administrative trivia. What really happened was, as we had these many teams of really bright guys--let me back up a minute.

    You've got to understand some of the rest of the context in which this occurred. In other words, [Bernard] Schriever and [Simon] Ramo were in charge of this whole thing, and you might ask the question, OK, what did they do? They did a lot of things, but one of the things they did was, they'd go around and see the management of all these companies, like Autonetics and Boeing and Thiokol and Hercules and so forth. They'd beat on these guys to be sure that they had the brightest and the best people in the ballistic missile program that these companies had. That's one of the very key things they did.

    They'd also go around and listen to see where the TRW guys doing the technical direction were screwing up in the minds of the managers of these companies, the general managers, or where the Air Force was screwing up, you know, where there was a friction that we needed to get rid of, so that we could get on with the job. So Schriever and Ramo spent an awful lot of time on the process, to be sure that you had open communication, you had the best people available, and there weren't any obstacles in the way of getting things done. That's just absolutely key. Then when you had one of these technical teams involved with say the guidance system of Autonetics, you had a framework in which it was easy to do work, and it was very easy to do excellent work because you had, you know, the very best people on the job.

    So for Minuteman I and Minuteman II, by the time we got the flight test data from telemetry and instrumented impacts of Minuteman I missiles, we analyzed the data. We and Autonetics you could sort the error sources under various buckets. This is gyro drift, this is accelerometer scale factor, this is platform bending, this is re-entry vehicle dispersion, and so forth and so on. Pretty soon Autonetics was saying, "Geez, you know, we need better scale factor stability in accelerometers" or velocity meters, as they called them. So they knew what the problems were. So then you could sit down with them and say, "OK, this is what Draper's been doing in terms of scale factor stability using PIGAs, what do you think of that?"

    Well, they would go, buy some PIGAs and run some tests, convince themselves that the scale factor is a lot more stable. But the thing wouldn't fit in the system because of the way we'd implemented the velocity meter interfaces with the computer. So you suggest to them, "Well, look, why don't you see what you have to do to that interface to make the PIGAs work? Why don't you see what you have to do to the mechanical interface to fit them on the platform?" So pretty soon you had them redesigning the platform, redesigning the interface of the computer, and in no time at all you had a whole new platform that had PIGAs on it which we then proceeded to put on the Minuteman II system. It's that kind of atmosphere for invention that was easy to do, in other words, not much invention in putting PIGAs on in place of velocity meters, but that was kind of the idea.

TAPE 1, SIDE 2

COLLINS: The situation you also describe seems to indicate that one was asking for fairly incremental improvements, that we aren't talking about dramatic improvements in technological sophistication?

BURNETT: Well, that's not really true. That may have been the example I used, but that's probably a poor one, in terms of--

COLLINS: I guess part of my question is, what are the limits of this sort of thing? I mean, you can only make certain kinds of leaps, it seems like.

BURNETT: We didn't do anything like in the later gyroscopes or stuff of that nature. I guess, that's a good point, yes. Let's see, there were a bunch of system inventions and an awful lot of materials and process inventions. The materials and process inventions were probably the easiest to see in the solid rocket business.

    One of the things we needed to do was to steer the thrust vector some way. These were solids, and the traditional way of steering the solids that the Germans had come up with was to put a vane of carbon in the thrust, and tilt that vane and hence deflect the thrust. Several of us decided, if we got with it, we could take a solid rocket and we could move the nozzle. We were swiveling the nozzles with the liquids because it's fairly easy to do that with liquid plumbing. That had never been done with solid rockets. But we decided, hey, we had to do that, to be able to throw the thrust vector enough, because you've got to pitch the thing over pretty rapidly after it leaves the launch pad.

    In fact it was Bob Anderson who worked for me at the time, he's now retired, working with a bunch of guys at Thiokol, who went from the idea that we ought to be able to have movable rocket nozzles on solids, to actually inventing how to do it, in just a very few months. Plus they had the experiments, actually built some scale rockets. It turned out the key was, how do you make the throat of this thing? He had to make that out of a peculiar form of tungsten, and how do you get that whole mechanism incorporated into the aft end of this rocket body so that as you moved the nozzle you didn't cramp or change the shape of the aft end of the rocket where you had all the rubber insulation on the inside to keep it from burning through. That's the kind of thing that we did which I would call inventions. We invented a way to make the nozzle swivel, rather than have those carbon vanes in there that didn't last very long and were horribly unreliable. That was a real invention.

    There were some other inventions of a similar type. We needed a certain ISP out of the solid rockets to make a practical sized missile, and you couldn't do that with the currentpropellants at that time. What we wound up doing was making an aluminized propellant. You get very finely ground chunks of aluminum and you mix in with propellant, and the first idea is, it's just more mass you're throwing out the rear so you're going to go forward faster. It had never been done before either.

    Another thing that had never been done before that time was to solve the aerodynamic case heating problem. When you start taking that solid rocket and pushing it through the atmosphere at the speeds that we were going to go, the air friction's going to heat the case up and the case is going to get soft enough that it's going to expand and maybe explode. So how do you make a case that's strong enough to keep the internal pressure, but also won't weigh too much, and also won't get soft because of the air friction? Not a trivial problem. We thought of all kinds of things to put on the outside, including the kind of ablation material we were putting on nose cones at the time, but we came up with a great invention. It's called cork. It turns out cork from cork trees is one of the best things you can put on a solid rocket for the exterior insulation to keep the whole thing structurally sound at elevated temperatures. That was a real invention. Something like that just had never ever been done before.

    Another invention we had was, how do you stage this thing? You've got three stages. This may sound trivial to you. There was no way to shut off the thrust of stage 1, it just petered out. Then you wanted to start up the thrust of stage 2, and you've got to mechanically separate two planes. There was a lot of lore about staging from the Atlas program and from the Titan program, particularly from the Titan program. The Atlas, you know, kind of did a stage and a half. They weren't sure they could solve the staging problem so they just dropped two engines off. Left the third one there running. By Titan, we'd gotten smart enough to figure out how to do the staging, with liquids. We had to completely re-invent the whole staging process for solids, so that you didn't get too much of a tipoff error, or the aft stage when you dropped it off didn't bump into the nozzle of the next stage and break it off. But the whole trajectory dynamics and gas dynamics of how you did that, was really a very clever invention.

    So those are the kinds of things I considered to be inventions on schedule, because we knew when we were going to fly the missile, we knew when we had to know if we could stage it or not. When you started a program, you honestly didn't know how you were going to do it. You really didn't. You just knew you had to solve the problem. Within a few months, we got all these kinds of things solved. Of course, you could put massive hunks of brainpower on working these kinds of things because we had them available. But today's systems aren't done that way. If you don't know how to stage that missile before you sign the contract, the contract doesn't get signed. We had signed contracts, thousands of people working, factories starting to get tooled up, we're still trying to figure out how to solve some of the key technical problems. That's what I call technical fire fighting. There's a whole bunch of those.

COLLINS: Let's move on to Minuteman more explicitly. Were you involved with the setting up of the specifications for the guidance system on Minuteman?

BURNETT: Yes, that was one of my jobs. By the time Minuteman came along, it was decided by the company that I was going to spend most of my time working on that. So other people were working the Thor, Atlas and Titan. I helped write the original RFP [Request For Proposal] that went out for the Minuteman guidance system. At that time I was just working the guidance and control systems. There was a Minuteman program office that had gotten started a bit earlier. In fact Rube Mettler was the first head of that program office. They were the ones in charge of the overall weapons systems specs at the time. Fellow by the name of Bill Besserer was Rube's deputy, who actually was in charge of most of the specifications. Rube got promoted, and a fellow by the name of Bob Bennett came in and was the Minuteman program director for a while. He was the person that insisted I work on the guidance thing full time, which is what I did. I had a small crew of people that wrote the original guidance specifications for Minuteman.

COLLINS: What were the differences from the earlier work?

BURNETT: Well, more system integration. Instead of having separate control system and guidance system, make them one. Another big change was, how you integrated and tested the whole missile. It was my contention that there ought to be one electronics contractor on the missile, and that ought to be the guidance contractor. Guidance ought to have the equipment that moved the nozzles, and ought to have the cabling and connects through the raceways into the guidance system, from each one of the stages. That also enabled us to lay a complete guidance and control system out in a laboratory and test the whole thing. That assured the guidance system worked and the control system worked, rather than the way it had been on the other ICBM programs.

    So we did a lot of what I call integration, simplifications and changes along those lines. But on the Atlas, ARMA had a little package called the inertial guidance system. You know, Convair did the wiring up and down the missile. Rocketdyne put the actuators on the rocket engine. Convair was responsible to make the whole thing work if they could. Terribly inefficient. If you let one contractor do the whole job, it's just a lot easier. So we did a lot of things like that that were different. These are what I call lessons learned from programs like the previous ICBM programs.

    I don't know if you've ever looked at the rear end of a Titan. It looks like a plumber's nightmare. You wouldn't believe it, just to see all the wires and the plumbing in there. I took to Denver the person in charge of control systems at the time, and showed him the rear end of a Titan. I said, "Look, if the Minuteman turns out that way, you're fired." You know, you've got to make it a lot simpler. You've got to package that stuff up so that the hydraulics don't leak, and you can put it on and take it off easily, you can maintain it and you can test it. So he did. In fact, we even invented a different way to make hydraulic connections that didn't leak. So that's one of the early disaster sources from the early ballistic missile program that we eliminated in the Minuteman program. The hydraulic plumbing would spring a leak or burn through some place and you'd lose the ability to swivel the engine. Shortly thereafter you'd have a big explosion in the sky.

COLLINS: Right. I think you've alluded to some of them, what were some of the key technical challenges that you faced with the Minuteman.

BURNETT: Well, the first technical challenge was, how are we going to place these missiles in silos and get them out in an orderly fashion? It had never been done before. We had some very good gas dynamicists that looked at the problem. Later we were pretty sure that you can set the missile down in the middle of the silo and ignite it, it will come out all in one piece, and it won't bang against the walls. So we said, "Are you sure?" "Well, no, but the theory would indicate that that would happen that way." OK, what do we do? We go up to Edwards Air Force Base. We drill a bunch of circular cylinders in the ground. We get ourselves some three or four seconds of burning for a stage 1 missile. We tie a rope on it so it won't go too far, and proceed to run some tests, to see what the relationship was between the diameter of the missile and the diameter of the hole, so that you could actually fly this missile out of the hole successfully. We figured out empirically how to do that. That was a very big problem at the time.

    The next big problem was, how do you get enough ISP out of the rocket engines that you can make an efficient ballistic missile out of solids. That involved just an awful lot of grief, in terms of propellant mixes, how much aluminum do you put in, how much nitroglycerine do you mix with this whole stuff? What are the interior ballistics in a rocket engine so that you can burn this thing efficiently and have a reasonably stable thrust? You'd like the thrust to go up and be fairly flat and come down instead of going all over the place. How do you make these things reliable? Because solid rockets are notoriously bad in terms of having burn-throughs on them. We had the whole problem of, get the ISP up, but you also have to have an insulation system inside of the solid which will stand the interior ballistics, and hopefully it would burn through, ten seconds after you were through with the stage. Fantastic amount of effort went into that. ... Let's see, we were talking about propulsion problems. The toughest propulsion problem we had was on the third stage, because you had to be able to stop the thrust within about a millisecond of the predicted time to get the accuracy, and we worked like hell to make the thrust termination system on that work properly. That was a lot of invention, because that had just never ever been done before, to have a solid rocket that's going full bore, within a millisecond quit.

COLLINS: What's the technique on that?

BURNETT: Well, the technique was that we put four holes on the top end of the rocket motor, with explosive covers on then, so that when we wanted the thrust to terminate, all we had to do was to blow those four covers off, and the chamber pressure dropped so rapidly that the propellant would go out. Figuring out what size of holes and how fast and how accurately the ordnance would do all that was quite a detective story. Then these data must be plugged into the guidance equations, because as you might guess, there were some time biases in that, and you had to very carefully calibrate those out. That was the toughest propulsion problem we had.

COLLINS: Now, this was something you would have had to grapple with when you became Minuteman program manager?

BURNETT: Right. What I'm talking about now is with my Minuteman program manager's hat on. Then you had the little matter of making the re-entry vehicle work. By this time, the people in Washington had gotten reasonably sophisticated. They gave us requirements like the following. You've got to have less than one chance in 10,000 that you're going to drop a nuclear-armed re-entry vehicle more than about 10 CEP from the target. They were concerned that if we had a missile screwup it would go blow up New York City or San Francisco rather than Moscow. That's one that we really sweat bullets over until we finally figured out how to do it, and it's worked. We've never, ever had that problem. That was kind of the missile problem.

    Now, that was only part of the problem. We also had to design the whole launch control system and basing system at the same time. What the Air Force wanted to do was have unmanned silos with the people in the launch control capsule in some cases maybe 30 miles away from where the missile was. Remember, these are nuclear warheads we're going to put on the front end of the missiles. Never, ever in the history of the United States had anyone ever stored nuclear warheads except where people were. I think the theory was that the interaction of the brain waves and the neutrons made them safe, and that was the lore at the time.

    The key technical problem was, how can we assure people we can put those missiles out at the ends of wires, nobody could do anything to those wires that would cause those missiles to launch unless we wanted them to launch, and nobody could do anything to those wires to make those bombs go off unless we wanted them to go off? That took a fantastic amount of effort, to get those problems solved. There were various high level government committees set up to review what we'd done. The National Security Agency got involved to be sure that the codes weren't easily breakable that we used, and so forth and so on, and needless to say, that whole thing has worked out very well.

    We also had to do that launch control system so it would fit in the context of how SAC wanted to do it operationally, what they wanted their officers to do, how many people are you going to have in a capsule, are you going to have the two man rule or not in terms of controlling these things. We finally wound up with a configuration that has 50 missiles and five capsules all interconnected in a very clever way so that if any two of those capsules, if for whatever reason they decided to go Communist or didn't want to fight a war or whatever, the other three capsules can override them. So you go through all the logic problems of, how reliable are the officers? What can you do in case they defect in pairs or quadruples? That's a very interesting design on the launch control system, to be sure that you're absolutely sure that it's safe and you're absolutely sure that when you want it to work, it will work.

COLLINS: It seems that the order of your interactions with the contractors would have increased. As Minuteman manager, you had a pretty wide array of corporations to deal with.

BURNETT: That's true.

COLLINS: How did you keep tabs on everything that was going on? You've described some of the other mechanisms, the various review meetings and that kind of thing. I wondered whether any special things were involved.

BURNETT: Well, it turns out you travel a lot. Sam Phillips at that time was the SPO director. He was an old fighter pilot from WWII. He'd gotten himself the use of an Air Force T-39. It's a small jet airplane that's now out of service with the Air Force. We traveled just an awful lot.

COLLINS: You mean, together.

BURNETT: Together. He liked to fly, and so he'd get an instructor pilot and he and I'd get in the plane and we'd go around and visit every organizational element of the program. Literally. We'd have a formal series of meetings. By this time we're based at Norton Air Force Base. We'd have a formal series of scheduled meetings, with all the contractors, periodically, once a month, so-called TD [Technical Direction] meetings. About every other month you'd go around and you'd see everybody. You'd spend the day with them. You'd spend a day at Boeing. You'd spend a day at Hercules, a day at Thiokol, a day at Autonetics, a day at Sylvania. A day at SAC. What you're looking for were the problems that were impeding progress, so you can get somebody to solve the problems fast. Literally.

COLLINS: What would happen when you'd spend a day at a particular contractor? Would there be a series of progress reports, or how would that work? Would you be walking around the shop floor?

BURNETT: Well, some of both. You've got to think about a control room. At Norton we'd have a control room, and in the control room we'd have charts on the wall. We'd have a whole bunch of charts on say Autonetics. You'd have a series of milestone charts. You'd have a series of financial charts. You'd have a series of problem charts. There's somebody who's constantly in contact with Autonetics who updates those things on virtually a daily basis. So you could sit in this control room for half a day, and you could review where you are schedule-wise, cost-wise and problem-wise on the whole problem, every contractor.

    So you have that kind of data base, and you're doing that once a week. So when you go around to see the contractor, you know what to look for. Each one of these contractors also had a control room. In fact, it was wired to the one at Norton. So you could go sit in the Boeing control room, and you could see the source data from which the data at Norton was derived, and the program manager would sit there and he'd just run you through all the charts, tell you where it is financially, schedule-wise and problem-wise, and you'd have discussion. Do you need help on this problem or not? Is this other problem getting worked? What kinds of meetings do we need to resolve that issue? Do we need to make a decision? It's that kind of a drill. If you've got time after that, you go kick the hardware to be sure it's really there. Because that's where the payoff is.

COLLINS: This may be a difficult question to answer, but what are the characteristics of a good program manager, given what you've just described to me?

BURNETT: Haha. That's a good question. First of all you have to be a real leader. You have to be the kind of a person who commands respect and doesn't have to demand respect and can cause people to want to do things just because you suggested they ought to be done. That's fundamental.

    Secondly, he's got to be a guy who has fantastic intuition in terms of how to get problems solved, or what a probable solution to a problem is. By the time you went through Atlas, Titan and Minuteman, the Minuteman problems generically were exactly the same as they were on Atlas, Thor and Titan. They had different acronyms but they were the same kinds of problems. So you had to be able to invoke classical solutions to problems very quickly with not very much data.

    Thirdly, you had to be able to see to it that the programwas really planned out in considerable detail, and that that plan is realistic, that real people could actually go implement that plan, because if you didn't have a good plan baseline, you weren't going to get anything done. You might get random things done, but it wouldn't all fit together at the end.

    Fourthly, you had to have damned good financial control on it, because it would be all too easy for a program like that to spend all the money you could possibly put in it and still not get anything out.

    Fifthly, you also needed to be able to pick very able people to be in charge of each one of these things, like, who's the guidance officer, who's the propulsion officer, who's the re-entry vehicle officer, who's the launch control officer? Because you're going to depend on those people to get most of the work done. So it's the old business of plan, staff, organize, operate and feedback, in spades. Because on Minuteman we probably wound up at one time having a couple of hundred thousand people in the United States working on the program. You had to be in reasonable control of it, and your neck was darned well on the line, in terms of it also had to work on schedule, or else you're going to lose your credibility with the guys that gave you the contract to do it.

COLLINS: There was obviously a good deal of interest in Minuteman at the national level.

BURNETT: That's right.

COLLINS: What kinds of contacts did you have to maintain or get caught up in, outside of the normal Air Force ones?

BURNETT: Oh, gee, a lot. First of all, the PSAC, the President's Science Advisory Committee at that time paid a lot of attention to it. I probably spent, I don't know, six or seven days a year in Washington with them. You know, they wanted to know everything that's going on in the program.

COLLINS: Was this the committee set up by Lauritsen?

BURNETT: Oh no, that was another one. Hans Bethe was chairman of this one at the time. This met in the White House. These are the people that whisper to the President whether or not we were screwed up or not. Charlie Lauritsen headed the committee which I think was chartered by Schriever, to look after us, Charlie Lauritsen from Cal Tech, Pat Hyland from Hughes was on it, we had [Charles] Lindbergh, [Hendrik] Bode from Bell Labs. They spent a lot of time with us. I think I told you the story that Bode is the guy who told me about the instability of permanent magnets. I didn't know that. I thought I had a pretty good education at the time. He's the guy that pointed out that velocity meters probably wouldn't work beyond a certain accuracy because of the instability of the magnets.

COLLINS: So what was the nature of your contact with Hans Bethe's committee?

BURNETT: I was the guy with the charts to go brief them, in answer to that question. Sam would often come along. I mean, he would come along and talk about the programmatic details, the schedule, the contractors, the money, if they wanted to know. They weren't very interested in that. They were interested in, how you solve that nuclear safety problem. How are you going to stage this thing? What makes you think you're going to get the accuracy that you say you're going to get? Why do you think it's going to be as reliable as you say it is? Because all the other systems this big haven't been that reliable. The real questions. Those were the committees that we reported to.

    I spent a lot of time with DDR and E at the time, because, I'm not sure how much publicity we'd want on this, but eventually the Air Force got more and more control of the program. After Schriever left from out here, which was in the early sixties, by '65, the air staff was reviewing the budgets before we got the money. What the Air Force characteristically did then, and still does now, is, the SPO director would say, "OK, I need umpteen billion dollars to run the program next year." It goes up one level of command and they say, "We're going to take 10 percent out of this." It goes up another level of command, and "We're going to take another 10 percent out." It goes up to the air staff, they say, "Gee, maybe I'll take 30 percent out." My job was to go to DDR and E, which was very powerful at the time, and say, "Look, you're going to get the following story from the air staff on how much money they need, but let me tell you, that's not the right number. Here's the right number." Much to the Air Force's dismay, we'd almost always get the money. So I had to go do that kind of thing too.

COLLINS: Was this something that for example Sam Phillips counselled you to do?

BURNETT: Yes. Sam said, "I can't go down and talk to DDR and E, the Air Force would kill me. You go, Bob."

COLLINS: He realized this was a way to insure that the program got the funds that it required.

BURNETT: Right. We had a heck of a good team working the program. You could spend time doing things like this. This was almost comic relief, to go do things like that.

COLLINS: Well, I think T. Wilson's role as program manager at Boeing overlapped your role here at TRW, if I recall.

BURNETT: Not really. People don't understand that. Boeing was the assembly and test contractor, initially. Their job was to figure out how to assemble this missile and flight test it. The equivalent they got to build was the interstage structure, the stuff we had to cut at the staging time. They built the telemetry system that went on the bird. They outfitted the block houses, and they also ran what we called the MAB, the Missile Assembly Building. These solid rockets are really quite dangerous. They're just big hunks of explosives. Putting them together without having disasters is a very tricky business. They were responsible for that. That's all they did in the early program. They were never ever the prime. Now, later on they also got the role to manufacture this thing in the field. They were the installation and checkout contractor in the field, so the Corps of Engineers built the silos for us, did a miserable job, I might add, and then we asked Boeing to come in and fix the silos so we could use them, and install and check out all the equipment. They did a superb job. I tell you, I take my hat off to that outfit. If you want something integrated and manufactured, go see Boeing.

COLLINS: I thought they had some systems integration responsibility as part of that, is that correct?

BURNETT: Well, they eventually won the command and control job. They eventually won the ground electronics, to hook the whole thing together. But that was a separate competition.

COLLINS: Was T. already out of his role as Boeing's--

BURNETT: Well, T. Wilson was the original program manager. T. and I became very close friends over this whole thing. No, T. was still there when they won the command and control job. They had a very significant role, but they weren't deciding how the missile was going to be designed, aside from the details of how the interstage structure would go. They were deciding how we were going to flight test it, how we were going to instrument it, how we were going to handle it, how you're going to ship it. We couldn't have done the job without them, don't misunderstand me. They also ran a thing we called the production board. I mean, we were flight testing and high rates of production simultaneously and we had to have a very tight control over the production rate. We wanted to be sure that the peanut butter and the jelly came out even in the end, one stage, one, two and three stages for each missile instead of two one stages and one third stage. Sounds trivial but I mean, when you're trying to get a thousand things like that out in the field almost simultaneously, you had to be very good on production control, and they're superb at that.

COLLINS: I don't know whether you can comment on this, but the Minuteman program seemed to be fairly significant in Boeing's history. It seemed to provide some kind of critical experience for them.

BURNETT: It sure did.

COLLINS: How would you sort of characterize how this?

BURNETT: Ha, I'll give you Burnett's version. It's kind of interesting. I got onto this early on in the program. They wanted to do the launch control system, and we didn't think they were competent to do it because they didn't understand enough about electronics. When I say we, it was mostly me, me and the TRW guys. So Sam and I went and made a deal with them. The deal was, OK, they could be responsible, except that they had to hire a subcontractor to do the electronics, and TRW had to have the right to technically direct that subcontractor directly whether Boeing liked it or not.

    Well, they swallowed hard and they struck a deal. They said, "Okay, we'll do it." They held a competition. They picked RCA as a contractor. We did just what we said. For every engineer that RCA had on the program, Boeing had two, and Boeing learned how to do electronics. I mean, they turned that lemon into lemonade. They really learned how to do electronics. They set up their own plating shops, their own printed circuit board shops, and they used that as an opportunity to learn not only from RCA but also from us. Now, that was one thing they did. So I discovered that these guys are real learners. We had a bunch of what I think were very innnovative techniques for managing the program.

    I haven't told you the whole story, you know. You had to not only build equipment and make it work, you had to have the software and make it work, and we had to design the maintenance system simultaneously, and we had to design all the factories to manufacture all this stuff. We had to do the whole thing. We'd invented a thing which we called system requirements analysis, whereby you could, as I put it, turn a crank and you'd find out what kind of tech data you needed train people with, how many people you needed, how you're going to maintain it, what kind of stocks of what kind of parts you needed to maintain this thing with. We turned this into a grand art in terms of how to perform this process. Well, Boeing flooded us with people on that to help us, and what they were doing was figuring out what we'd done.

    It turns out, what they did was, they adapted this system requirements analysis process we'd come up with to their commercial airplane business. So the kind of system engineering they do on their commercial airplane business, they learned off the Minuteman program. They also use that training mechanism for a bunch of the key managers. T. decided, OK, it's time for him to go do something else, so he cycled through there a whole series of the key Boeing executives that run the place. A bunch of them are retired now. Tex Bouillon, who ran all the commercial airplane business, was the successor for T. to run the program for Boeing. Fantastic golfer and gambler, I've never seen anything like it in my life. I understand how he sold airplanes. It just went on that way. They're still doing it that way. They put some of their very best people on the program, not only to contribute, but also to learn how to pick up those techniques and plow them into the rest of Boeing.

TAPE 2, SIDE 1

COLLINS: If you could just give me a little more detail on the nature of this systems requirement process. I mean, what was new about this? Was it just that the size of the job required special kinds of approaches? Or there were new analytic techniques involved in defining work?

BURNETT: Well, these were new analytical techniques that go beyond say trajectory analysis and structural analysis, and what you might call the technical aspects of it. They got into the subject of how many people is it going to take to assemble a missile? What kind of ground support equipment do you have to have to develop and assemble a missile? It was a very structured orderly process that we came up with. It was the sort of a process where you didn't miss anything. The process might say, OK, you've got to have a piece of ground support equipment to check out the guidance system when it's in the vertical position in the silo. You have to have another piece of ground support equipment to check out the guidance system when it's horizontal in the missile assembly building. Now, then, some bright engineer had to sit down and say, OK, now, technically what's different, vertically, horizontally? Can I use the same piece of equipment? Can I use a modified piece of equipment for, say, the horizontal job? But you didn't miss anything. We never wound up thinking we'd missed equipment in the system that way.

COLLINS: So I guess it's really taking a look at the whole operation of doing a job which includes the systems you want to produce.

BURNETT: Right. It includes the people that are going to operate it and the people that are going to maintain it and the process by which you're going to operate it and maintain it. We thought, for instance, early in the program, that the guidance systems would not be as reliable as they were. I think the original requirements were, they had to be something like, mean time between failure of a thousand hours. We figured that a thousand hours with as many guidance systems as we could afford, you'd better have a float of so many and you'd better have a facility some place that can repair these things. None of that existed. So this process would define, how many guidance systems do you have to repair a month, and what kind of characteristics of equipment it's going to take to repair those guidance systems. Then we went and built the facility. In fact, it's still in operation. It's in Ohio.

COLLINS: How is this expressed concretely? I mean, in the case of worrying about the weapon system, you had you know specifications, configuration control.

BURNETT: Well, we had all of that too.

COLLINS: But when you're talking about the whole ball of wax, how did you express it and regard it as something you could use as a tool?

BURNETT: I don't have any examples here with me any more. It's big documents. You know, about twice 8 1/2 x 11 size, yeah, which would have pictures of the functions that needed to be done on one side of it, and a description of the functions that had to be done on the other side. Somebody had to put these things together draft-wise, and you'd have glorious meetings to go through all of these, with people from say the guidance shop, the propulsion shop, the re-entry vehicle shop and so forth, to be sure they were complete. This was the kind of thing that Boeing just loved to do. I mean, they became extremely clever at this brand of systems requirements analysis.

COLLINS: What you're saying dovetails with the little bit of research I've done on their work in what was called the Apollo TIE contract.

BURNETT: Well, you've got to remember the tie there. Sam Phillips was the SPO director on Minuteman. He didn't personally do the systems requirements analysis, but he knew what we were doing, because in our common conference room we usually had a lot of this stuff draped around the walls. Sam then went and became the Apollo program director, and realized what he was missing, because this pulling together through the systems requirements analysis just didn't take place there. But he remembered T. Wilson and Minuteman, and honest to God, that's how the connection got made. He called up T. and said, "T., you've got to do the job for me on Apollo," which Boeing did, and again did very well, much to the dismay of the NASA folks, I might add, because this was newthink to them. But that's where it came from.

    Boeing is also fantastic on what I'll call scheduling. This may sound trivial to you, but if you want a silo to become operational here, you know, you've got all these things you've got to do back here, to lead up to that, and they had some absolute masters of the art of scheduling. Programming is what we used to call it, but that has another connotation now having to do with computers. They had people who could take that problem and bedsheet it out so that for, say, a hundred silos or 150 silos, you could get them all flowed out in time, and the manpower levels to do the job just wouldn't ripple at all, which would be the most efficient way to do it. The equipment to do all this would be minimal, I mean, not try to put all one hundred in at once, you don't need a hundred cranes simultaneously. They're magnificent at that.

COLLINS: Is that just a question of extreme attention to detail? I mean, what sort of distinguishes that capability?

BURNETT: Well, guys like T. Wilson who realize that, a, it'simportant, and b, they can make money if they plan it very well, and c, it's worth putting some of your best people to make it come out right. They made a lot of money on those integration and checkout contracts in the field, because they could flow them out, they could minimize the cost, the equipment worked perfectly when they got done, and they made a bundle of money. I think that's the American way.

COLLINS: What other kinds of management organizational techniques are we to mention here?

BURNETT: Oh gosh. The Air Force came up with a whole series of--what did we call those things? Gad, it was a number. The Air Force finally took note of how we were managing that program, came in and audited the program, and put out a whole series of regulations, you know, "thou shalt run a program this way." 375 regs is what they called them. They've been cast aside now, but for a decade or so, in the Air Force the name of the game was to run a program by the 375 regulations. This was a system, a command set of regulations. They came in and they just captured everything we'd been doing. Because we had our own specifications, we had our own interface control drawings, we had our own interface control working groups, we had our own test documents, we had our own test working groups, and we had our own logistic scheme which revolved around the systems requirements analysis. I don't know, it just seems so straightforward and obvious now. Sam and I wound up teaching an SPO director school for many years after this, to pass on in some detail, you know, to new Air Force SPO directors how to do some of these things. We used to go back to Wright Patterson for days at a time and give a series of lectures at their system program officers school, which has now been converted to this defense management course that's down at Fort Belvoir.

COLLINS: Do you feel the approaches that were developed in the ballistic missile program are generic for tackling weapons system type jobs? Or do they just fit for particular kinds of weapons systems?

BURNETT: Well, the real thing we did, which may sound trite but it's true, was, we made a team out of all those associate contractors, so it wasn't Autonetics versus Boeing versus Thiokol, it was Autonetics plus Boeing plus Thiokol plus the rest of them working together as a team. The company badges disappeared and everybody wore a Minuteman badge. We had people who worried about the contracts and worried about the legality of what was taking place, but that wasn't the predominant thought. The predominant thought was how to get the job done, and how do you get the job done efficiently and very fast, because the emphasis was on speed. How do you get it done very fast? We weren't trying to decide how many vice presidents we could put in jail over this. We were trying to decide how many vice presidents could we get promoted because we did a hell of a good job.

    Now, what's happened more recently in DOD procurement is all the negative stuff. You know, if you were a real agnostic you'd say the whole purpose of DOD regs now is to figure out how many company vice presidents they can put in jail because of mischarging. I'd never heard of mischarging when I worked on the Minuteman program. The concept of mischarging, I've only learned recently, within the last ten years. We didn't worry about things like that. The deal was to get the job done, and we put our faith and trust in key guys in the government and in the companies, and a lot of what we did was by a hand shake. Now, when it was necessary to clean up a contract so somebody didn't get screwed financially, you know, that got done. But the emphasis was not on all this form that's currently going on in the procurement business. I think that's the root of the problem in the procurement business now.

    In those days Sam Phillips, for half the time I was with him, was a colonel. The other half the time he was a one star. The kinds of things he could do on his own authority, I don't even think the head of systems command today could do, literally. So what we've done is, we've de-balled, if you will, the system program directors that are responsible for these acquisitions, and we've made it very difficult for them to form teams with their contractors to get the work done. What we were able to do is form teams amongst the contractors, and the SPO director who was on the spot had the authority to do just an awful lot more than any SPO director today can do, that I know of. I don't know how you recapture that. It's not the paper process we're going through that was so important, it was the people process we were going through.

    Now, the Navy did almost exactly the same thing on their fleet ballistic missile program. If you go talk to Bob Wertheim, or some of the guys that were on that program, in fact, we talked to each other an awful lot. You know, we learned from each other the kinds of techniques we used to make these things work. Now, it may sound a little soft to you. There was a tough side to it. This I'm sure you don't want to publish, but at one juncture, we came to the conclusion that Autonetics management couldn't hack it. They just didn't have the management whammy to make the production go at the rate at which it needed to go. I'll tell you, we went and we got that management changed. We went to see the head of North American at that time and we said, "You've got the following problem, and if you don't believe us, go run your own investigation, and he, he, he and he just have to be replaced. They're just not good enough." Those kinds of things got done, too.

COLLINS: Would that ever happen in the case of a commercial customer, working with a corporation to go in and make that kind of declaration to corporate management?

BURNETT: It's more apt to happen with commercial customers than any place else today. The commercial world tends to operate thatway. We do a lot of work for say Ford Motor Company, a different part of the company, different part of TRW, and if our guys aren't hacking it with Ford, they come and they tell Joe Gorman, and by golly, Joe will change it. Now, you don't find much of that going on in the defense business any more today. You know, it's us versus them, and it's hard for the government to know well enough about how the company operates, to know who the screwups are, if you will. But in our day, in the ballistic missile program, we really knew who contributed and who didn't. We, with a vengeance, got rid of the guys that weren't really contributing. I mean, we didn't fire them. The company did something with them. But they got off the program.

COLLINS: I think for program managers it's pretty clear how they get evaluated, whether the product that they're responsible for works on time for the money. How does someone in your position now get evaluated? I mean, what are the kinds of things that the very top management looks for in someone like yourself, or the head of an operating division, to know that they've done a quality job?

BURNETT: Well, pretty simple. I don't think I have it in here, but I can tell you what's in it. We have what we call quantitative goals and qualitataive goals. In the quantitative goals, which we've gone over most carefully--well, they all get gone over carefully, but at least you can put numbers on the quantitative ones--in my case, or Ed Dunford and I happen to have the same set of goals, since we occupy the same office, the first goal is sales. What did you say you were going to have for sales at the end of the year, and what did it actually turn out to be? So a year before, let's see, a year ago last December, we told Joe Gorman how many million dollars of sales we'd have here when the books closed in December of '89. Now, we either did it or didn't do it. It turns out we were about 120 million bucks over that. That's good.

    Second parameter is, what are the profits? What were the planned profits and what did we actually achieve? Now, you've got to have some relevance to that. The industry standard on profits, after tax, in the aerospace industry is about 3 1/2 percent, so about 3 1/2 percent of the sales ought to turn in to be profits. Though there may be some special circumstance where the number ought to be higher or less than that, but that's very straightforward.

    The third parameter is, what's your return on assets employed? The company's got a lot of money invested in the place. Would they be better off investing it in space and defense or in T bills? Our goal last year was to have 15 percent ROAE, which would be the equivalent of, a buck you put in here, you get back a buck 15 for the year. We actually did 15.2.

    Then the last parameter is sales acquisitions. You know, is your backlog going up or going down in this crazy environment? Soyou tell them how many new contracts you're going to get and what the volume is, and at the end of the year you either did it or didn't do it. So we did very well on the financial end here last year.

COLLINS: Would those kinds of quantitative judgments be the basis on which they, someone in the late fifties, early sixties, would have been evaluated, sitting in a similar situation?

BURNETT: I don't think so. I'm sure they'd have been evaluated on profits, and I'm sure they'd have been evaluated either on sales or number of people that worked for them because it's almost comparable. You take the number of people we have here and multiply by 100,000 each and that's the number of sales we get characteristically. It's a little bit more than 100,000 now. We have 28,000 people and our sales are about 3.1 billion. We have some other criteria too. One overriding criteria is the highest standards of legal and ethical conduct. We have a whole program to go audit the place to see how we do on that.

    There are other sets of criteria. How well do we do in the outside world? The company wants us to be good citizens in our own communities. They want us to serve on government committees and boards. You know, a number of things like that, they want you to do. It's pretty easy to audit that.

    Then the other main category of goals has to do with what's commonly called total quality management. You know, are the customers happy? Is the product working well? Are the people involved, are you figuring out ways to manage the place more efficiently so that it takes fewer people to do the same work in the future than it did in the past? That's pretty easy to set up a series of, you know, two pages and we can write down all the goals. In fact, we had this conversation last month with Joe Gorman. I presented how I thought we'd done and he agreed with that. So, you know, he saw it about the same way we saw it. Yes, you can do that. It takes a lot of trust, and you've got to get yourself oriented to go do it.

COLLINS: I guess where I'm heading with this is, how this fits into your observation that the earlier environment was different. We can kind of point to it for the program manager level. For top management I'm curious how it was also different. If you can say anything on that.

BURNETT: Well, let's see. In the ballistic missile program, the top management had to have a key goal of working together with the other associate contractors. We'd never have gotten the job done if the key executives hadn't agreed, OK, we're going to work with each other. It doesn't matter if he's got a Thiokol badge or Boeing badge, we're going to treat him as a member of the team. That is uncommon in the industry since then, very uncommon. Along with that went the idea that, you know, if there's something we can do to help Autonetics solve a problem, we'll help them solve the problem, because it's in our mutual interest because if the total program is successful, we'll all make more money than if it fails. I'm not sure we're communicating on this one.

COLLINS: This is kind of too complex without getting into a lot of other things, so I think we can leave it at that. Just one final set of questions, and that is, you mentioned that you had fairly close contact with the fleet ballistic missile activity. What was the nature of your interchange there, and can you characterize the kinds of things that you learned from one another in this period?

BURNETT: Oh, we'd sit down from time to time and just review the whole program together. There were a few special circumstances. This gets almost into the classified area, but at one time, the DOD decided that these ballistic missiles had to be designed to be harder to nuclear effects. I don't know if promyt-gamma or EMP means anything to you but in the mid-sixties, we accidentally learned a lot more about some of the nuclear effects particularly on electronic equipment, and we were told by OSD at this time, in fact by Johnny Foster when he was in office: "hey, you gotta change your electronics so that they don't get upset by these transient nuclear phenomena". Well, none of us had ever done that before. He told the fleet ballistic missile office the same thing at the same time, so you call up old Bob Wurtheim and you say, "Robert, what are you doing about this?" He says, "Well, I don't know." "Why don't we get together and chat, see if we can find a way to go about it together?" So that's a focus point.

    Now, when you go with hardening the electronics, you're very quickly talking about almost a whole missile system. So you spent a lot of time talking about: OK, where do we get the EMP hardened semiconductor devices? Well, we didn't know. We had to buy a bunch of parts and run some test programs to see what the susceptibility levels were, and what manufacturer accidently made these things that were hard to neutrons, or not hard to neutrons, and you share all that kind of data. So we would know and he would know the same guys, vendors you could go to to buy the parts that had some chance of satisfying the requirements. Then pretty soon you could influence the microelectronics industry to build the stuff so that it was hard to nuclear effects in the first place. The extent to which we in the program office could cooperate on that, you could get more leverage on the industry. We also competed a lot too.

COLLINS: Did the nature of interservice rivalry affect your relationships?

BURNETT: Naw, not very much. Both organizations had the same mission in mind: let's get the job done. It's too important for the nation to screw around with. Let's really get it done in an outstanding fashion. One of the issues on the ballistic missiles is, how can you penetrate the other guy's ballistic missile defense system? We've done some work on penetration aids, as we call it, and we spent a lot of time with the Navy to see if they could use any of our pen-aids work on the Navy program, and they adopted some of the ideas. From time to time we both had guidance problems. I mean, even though they had the Draperian guidance system, they had peculiar anomalies on some of their test data, and you know, you'd sit down with each other and share the test data. There was a real competition for money, later on in the program. There was a competition to try to do better than the other guy in terms of flight test results and reliability and all that. But it was an honest competition. It wasn't a sick competition. A few of the contractors were the same. Some of the stage contractors were the same, let's see, Thiokol and Hercules I guess were in common. But they managed their program differently. They basically had a prime contractor called Lockheed to manage the program for them, whereas you know, we had an associate contractor structure and TRW was just the system engineer, technical director, which is different than the Navy scheme. Didn't make it right or wrong.

COLLINS: You know, one of your observations about the technical direction dimension in your approach was the attempt to develop efficiencies, to get a job done very quickly. I assume the same kinds of pressures were around for the fleet ballistic missile program.

BURNETT: I assume so. I've never been to any of their internal meetings like that, to get the feeling for it. But you know time is money. The real trick to saving money in procurement is to get the thing done fast. That's another thing Boeing learned off the Minuteman program, because we tried to do everything just as fast as we could. In their commercial airplane developments today, they learned a lot from Minuteman. They decide on what the schedule's going to be, because they want to hit some particular market window out here, and then they program an extra amount of money, pad it if you want to call it, which the program manager has at his discretion to do things in parallel to preserve that schedule. That's basically what we did on the Minuteman program. You can't do that any more because Congress won't let you put a pad in there. But somehow or other Sam had managed to do that. He's a master at that kind of thing. Money was never one of our problems. The issue was always, how do you get the technical thing to work fast enough?

COLLINS: Which I think sort of plays into your earlier thought about being able to do invention quickly and on schedule.

BURNETT: It may be a lot of adaptation, but it was clearly something that hadn't been done before in that way.

COLLINS: I think I've covered all the points I wanted to touch on. Do you want to add any final reflections?

BURNETT: It was a lot of fun. In the cool light of dawn, Ithink I had more job satisfaction out of doing the Minuteman program than anything else I've done since then. I think I made more friends in that time period too. We really became good friends, between all the key associate contractors, and we still keep in touch. It's not only just the contractors, it's also the Air Force people that worked on the program. Unfortunately a lot of them are starting to die off now, because none of us are getting any younger. But it was a lot of fun and the thing worked.

COLLINS: I think that's a good concluding line. Thank you very much.

Rev. 09/09/96