Oral-History:Brad Parkinson
About Brad Parkinson
Parkinson was a career Air Force Officer who also received a masters in Aeronautical and Astronautical Engineering from MIT and had significant scientific and research experience. He became program manager of what would become the Global Positioning System in November 1972, as an Air Force Colonel. The origins of the program went back to the 1958 realization that Sputnik was emitting a Doppler signal that could be used for ground location. All three services had come up with variations on the technology by the time Parkinson arrived—Transit, Roger Easton’s work at the Naval Research Laboratory, and the Air Force’s 621B program. Parkinson synthesized the technology of the three prior proposals (Transit’s orbit determination system, Roger Easton’s atomic clock technology, and 621B’s digital signal structure and concept of operation), made some improvements, and got all three services behind him for a joint GPS proposal. The program was accepted in 1973 and in operation by 1978. It involved 24 satellites in high-altitude, 12-hour orbits; the costs went down significantly, since the digital signal structure allowed GPS to take advantage of the digital electronics price revolution. The military built in inferior capability for civilian users, but differential correction systems researched by civilians made up for that inhibition soon enough. After 1978, Parkinson retired from the Air Force and went into private industry (Rockwell, Intermetrics), before taking a post at Stanford as a professor in charge of NASA’s Gravity Probe B project. He believed that GPS would continue to expand its usage, allowing for automatic planes, cars, tractors, etc.
About the Interview
BRAD PARKINSON: An Interview Conducted by Michael Geselowitz, Center for the History of Electrical Engineering, 2 November 1999
Interview #379 for the Center for the History of Electrical Engineering, The Institute of Electrical and Electronics Engineers, Inc.
Copyright Statement
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.
It is recommended that this oral history be cited as follows:
Brad Parkinson, an oral history conducted in 1999 by Michael Geselowitz, IEEE History Center, Piscataway, NJ, USA
Interview
Interview: Brad Parkinson
Interviewer: Michael Geselowitz
Date: 2 November 1999
Place: Georgetown Suites Hotel, Washington, D.C .
Parkinson's educational background and the origins of GPS]
Geselowitz:
Brad Parkinson was just about to tell me the origins of GPS.
Parkinson:
- Audio File
- MP3 Audio
(379 - parkinson - clip 1.mp3)
Obviously, there was an intersection between the concept and my background. My background is that I was a trained navigator. I had two years of formal education with the Naval Academy and an appreciation for what navigation and piloting is all about. I also had about five years experience in inertial guidance systems and inertial navigation systems, two years of which were formal education at MIT and the Draper Lab, then known as the Instrumentation Lab, and three years as an inertial guidance analyst. In addition, I had a fair amount of management experience, and my degree was in Aeronautics and Astronautics so I also had space experience. I had the right qualifications, if you look ahead to what GPS eventually became.
On the other hand, the alignment of the stars went like this. There were three competing ideas on how a second-generation positioning and navigation system might be created, using space satellites. The first system really was Transit. Transit was triggered by the original launch of the Russian spacecraft back in 1958. That was triggered because people recognized that a Russian satellite was going by and transmitting a tone that created a Doppler shift. If you use the laws of mechanics, the rotation of the Earth, and the Doppler shift, you could infer where the satellite was. Then a very bright person at APL (Applied Physics Lab at Johns Hopkins)—whose name I have unfortunately forgotten, but it is in that lecture that I gave—turned that on its ear and said, “Look, if we listen to the Doppler we can sort out where we are on the ground.” This background in part was the basis for the existing (in 1972) space-based navigation system called Transit. That was the first competing idea, and the Transit advocates wanted to evolve into the second generation. At the Navy’s Research Laboratory, NRL, there was another concept under Roger Easton that used space-based clocks. Eventually they put very precise clocks in space, and finally an atomic clock under my sponsorship. The third competing concept was something called 621B. That was an Air Force project that also has many of the attributes that you now see in GPS. It has probably never been given its due credit. I entered the picture when those three concepts were in a death struggle—none of them were going anywhere—and I was a young colonel in charge of advanced ballistic reentry system engineering, and I had about a $100 million yearly budget and I was happy as could be.
Geselowitz:
Did the Russians plan on using the Doppler effect to trace Sputnik when they launched it, or did people realize that after the fact? They just put a transponder in there and they were just doing one dimensional?
Parkinson:
I do not know whether they were doing it, and the realization that this would work was not tipped by the Russians. It was an independent realization by the people at the APL. Initially, of course, you try to track these things by using dish antennas and getting azimuth and elevation and sorting it out that way. But the sudden recognition that on a single pass you could determine the whole orbit just by listening to the Doppler led to Transit. Transit deliberately produced a Doppler signal from a known location, so a single pass would allow position determination if you were stationary. If you’re not stationary, you have to make corrections, and depending on how well you know your velocity, that can produce a significant error. Nonetheless, the fundamentals were already there, and I don’t think the U.S. knew how the Russians were tracking the new satellite. My guess is, given the state of their computer technology, that we may have been ahead of them in our ability to track the satellites. Of course satellite technology was in its infancy in terms of the reliability of electronics, in terms of orbit determination, and relative to the eventual uses of space. Of all the uses of space, communications, listening, and looking up and down, navigation was one of the things that popped on the radar screen. By 1959 or ’60 it was already well into its prototyping stage, then over the next six or eight years, Transit improved. It initially had reliability problems that were solved.
Parkinson:
To get back to the central theme then, my background had somehow been identified by a three-star general who was in charge of the Air Force’s space division, and he recognized that this navigation program was going to go nowhere (as he’d been told that by his superiors at the Department of Defense) until he found somebody who could pull together all these concepts and come up with one concept. Then he could tell the Pentagon, “There’s only one concept. This is it, and this it what we’re funding.” I was brought in to do that. My general background was just about exactly right. I initially spent perhaps three or four months digging deeply into the three existing concepts, technology, advantages and disadvantages. Meanwhile, I was getting enormous pressure from the Air Force, and in particular from the Aerospace Corporation contingent, to support the Air Force’s version of 621B. By this time I’d already formed an opinion 621B was not the optimum system—that some aspects of the competitors were good features.
I was appointed as the program manager in November 1972. I was still a young Air Force colonel with about fifteen years of service. I took the Air Force concept, 621B, and brought it before the Defense System Acquisition Review Council (DSARC), which is the top Department of Defense decision-making body, and we failed. We called it Black Thursday. It was sometime in August of 1973. I went into one of my hero’s offices, Dr. Malcolm Currie, who was the Deputy Director of Research and Engineering (DDR&E) for the Department of Defense. He was in control of all the development, research, and engineering for the DOD. He told me, “Brad, don’t worry about it. You can get this right. What I want you to do is go back and make your program a joint program.” I recognized that there was pressure on me from the Air Force to do it the Air Force’s way, so I came up with the idea that we would have a synthesis of all the ideas, and add some of our own, but we would not do it in Los Angeles. I called a meeting in the Pentagon over Labor Day, brought my best and brightest officers. By that time I had managed to recruit a bunch of AF officers, very bright masters and Ph.D. degrees.
Geselowitz:
I’m sorry, where was the Air Force program?
Parkinson:
The Air Force program was located at the Los Angeles Air Force Station in El Segundo, which was the headquarters of the Air Force Systems Command's Space and Missile Systems Organization. It was then known as SAMSO (Space and Missile Systems Organization), and is now named the Space Systems Division. I directed a three or four day gathering in the Pentagon over Labor Day weekend. Thus, over Labor Day weekend of 1973, the real synthesis that became GPS was created.
Many people take a lot of justifiable pride in the contributions they’ve made to GPS. It is not a one-person show. The three predecessors all made their contributions, and yet GPS is not solely any of them. It’s not to detract from them, but to recognize there was a balance. For example, Transit gave us orbit determination. They really knew how to do that, and we needed it, because GPS satellites have to know where they are. Roger Easton brought the atomic clock technology forward, and we needed very stable time because in essence GPS acts as a one-way radar. That is enabled by knowing precisely when the GPS signal was generated. The third contributions are the digital signal structure and the concept of operation. The Air Force was pretty close, and we took their signal structure and refined it further, added some features to it that weren’t there. All these ideas contributed to the final GPS of 1973 (which is still essentially unchanged).
Geselowitz:
So you had already had the idea earlier that the synthesis was needed. But you needed essentially the political cover of a joint effort, meaning having everyone in the room so no one can pressure anyone to take more from one side or the other to put it together the way you wanted to put it together. That, essentially, was Dr. Currie’s advice?
Parkinson:
That was his advice and direction. I knew Dr. Currie personally for a very strange reason. When he was appointed for the incoming administration in late ’72 or early ’73, he was relocating from Los Angeles to Washington, so he had to make a series of trips to get his family organized. But the DOD doesn’t allow you to make trips just for personal reasons; you always have to have an official purpose. So the purpose, he used was to visit SAMSO. Well, as near as I can tell, after he visited us a few times they ran out of things to tell him at the headquarters building, and at that point I think General Schultz said, “Well, let’s send him for the afternoon down to talk to Parkinson. He always has a good story to tell.”
So you have this very strange situation in which I, perhaps the most junior colonel in the whole Air Force, was holed up in a room with the number three leader in the whole DOD, who technically outranked all the generals, and we spent at least three or four hours together. I had a stack of view graphs that started out on one side of the table. He had a background in nuclear physics; he was a very, very bright guy. I went through it chapter and verse: why it was important, how you would do it, what I saw as pitfalls and risks. At the end of that meeting, he exited with a view that he wanted to do this.
It turned out that meeting was essential because the Air Force never fully backed this system. They wanted it their way, but they didn’t want to pay for it. It’s sort of analogous to asking the richest person in the neighborhood to pay for the whole high school. That’s how they viewed it. Here they were putting up a system, not just for DOD, but for all these civilians, and it was coming out of their Total Obligation Authority (TOA). They were not happy. On occasions I was “braced up” in the halls of the Pentagon as if I was still a plebe with the Naval Academy. Major Generals informed me that I was shortening my career by an enormous amount by persistently and skillfully selling this program. It was an interesting experience.
December 17, 1973, we again went through DSARC with the new synthesized concept that now brought all these ideas together. We invented a very innovative satellite and user equipment development program, because detractors in the Pentagon were saying is you can’t orbit satellites until you prove they work with user equipment. So you’re stymied in terms of your launch capability, because how do you prove it until you launch it? We set up a system of “satellites” at Yuma Proving Ground (in Arizona) that were really on the ground, and we called them pseudolites. As an aside, this concept is still being used. The system broadcasted satellite-like signals. We had a network of these; you need four to navigate. An airplane equipped with a receiver would fly through this network and we demonstrated that the user equipment would work. Then as satellites came along, we would turn off the pseudolites, so soon we could navigate using only satellites.
As you know, bureaucracies take great delight in saying no—they can’t necessarily say yes, but saying no is evidence of their power. You have to somehow defeat all these no-like arguments if you’re going to get anything through the Pentagon. One of the challenges that we had was to work our way through the morass of “no” logs that they tended to throw in our path. The use of pseudolites helped.
Also, instead of going to an Air Force base for testing, we went to an Army base, which also got me in trouble. I felt at the time that I had the Navy pretty interested and I had the Air Force pretty interested, but I hadn’t drawn the Army into a joint program. By going to Yuma Proving Ground, rather than say Holloman Air Force Base (in New Mexico) or Eglen Air Force Base (in Florida) or one of the other places that you might have conceived as a useful place to do your testing, suddenly we had the Army involved. They could identify with GPS. The Army now has many, many more pieces of user equipment than both the Navy and the Air Force, combined.
Having finally sold the program concept, the next step was to sort out exactly how we were going to do it. We pioneered a number of new ideas. For example I never delegated system engineering or total system integration to a contractor. It was done internally, in house, by me and the Program Office people. Who were those Air Force officers with masters and Ph.D.s I mentioned before? Fortunately they had a lot of program experience. By retaining core Systems Engineering, we were separately contracting the pieces: the satellite, the ground segment, the user segment, and the test program.
I almost lost my job at one point because the officer I worked for, General Schultz, did not initially grasp how we could do the Systems Engineering. (Incidentally, he was an outstanding boss for me.) It happened one day when I was trying to tell him what my concept was. He got upset because he couldn’t figure out how on Earth I could pull this off, how I could integrate all these pieces. I felt that it was essential. Unless I was at the center of the system engineering involved here, I didn’t think I could pull it off either, because I knew the contractors would quickly close you out of all the essential decisions. Making the trades would be left to them on whatever motivation they had. Our motivation was quite pure; we wanted a system that worked and worked well. The near-firing happened at a meeting in which I was standing up making a presentation up on the ninth floor at the Space Systems Division. I had about half a dozen of my people in the room with General Schultz. I could tell he was getting very angry with me because he could not understand how this was all coming together. At the same time the essence of his problem was a mystery. Finally, I got it. Fortunately, I had a backup chart that showed the interface relationships between user equipment, the space segment, and the ground segment. It showed that these interfaces were fundamentally defined by signal structure in space, not defined by physical things, because that’s how they interacted with one another. These interfaces I planned to manage directly. As soon as I showed this chart, General Schultz sat back in his chair, smiled, and nodded his head. That was the go ahead for us to contract for GPS in a relatively unusual way.
Geselowitz:
So there’s a schematic reproduced here.
Parkinson:
It was that chart, except it had a bar across it, and it showed that each of these interfaces was a thing that we could control. By controlling those and directly contracting rather than having integrating or general contractors across the whole thing, it allowed us to ensure that all those were done right. I’m very proud of what we did. Not necessarily what I did, but certainly what my people did in terms of defining the signal structures.
Geselowitz:
This was completely a DOD project? NASA was not involved?
Parkinson:
NASA was not involved. It was a DOD project, but from the get-go we also said that it was going to have two signals: a signal primarily dedicated to military use and a signal that was both a clear acquisition signal for the military but also available to civilians. I think that helped us a lot when we went to Congress to try to get the money. Phase I was $104 million, which doesn’t sound like much right now for six satellites, a ground segment, nine kinds of user equipment, and a test program. It sounds like a heck of a bargain, and it really was. We later added two more satellites, thanks to the Navy.
Technical challenges and satellite signal structure
Parkinson:
- Audio File
- MP3 Audio
(379 - parkinson - clip 2.mp3)
I should go back and talk about a couple of technical issues with which we wrestled. NRL, Roger Easton’s operation, had used an entirely different ranging signal, something called Side Tone Ranging (STR). STR, in our view, was trivially jammed. It had no inherent jam resistance by its very nature. It was a very clear and obvious signal. Furthermore, if you wanted to use multiple satellites you would have to transmit on fundamentally different frequencies. Instead we proposed a spread spectrum signal although spread spectrum was in it infancy. All satellite signals came down exactly on the same center frequency except for Doppler shifts. Because it was digital in nature, we could see that eventually digital receivers would be able to capture the whole digital bit stream and hence have no inter-channel bias shifts. When you get down to navigating farm tractors to better than a centimeter, as we are right now at Stanford, it turns out those inter-channel biases would have destroyed the accuracy.
So this signal structure, which the Air Force had fostered (not the particular one that we had but something along those lines) was essential, and in my opinion it was a success. If you look at the competing Russian navigation system that came along later, GLONAS, each satellite used a different frequency. The consequence is those inter frequency biases create a system that is inherently less accurate. There are ways to get around the problem to some degree, but if you look at the accuracy of the most precise surveying and real time kinematics receivers of today, you will see that GPS provides a far more stable, more accurate, and more precise signal than the GLONAS does. I think the decision was pretty sound, and I think has been vindicated over time, especially with the simplicity.
What we could see was that initially receivers were going to cost a great deal. I had a motto hung up in my program office, that I had invented for the program. The first part of it related to accuracy. It said, “The mission of this program office is, number one, to drop five bombs in the same hole.” What I meant by that is precision weapon delivery. Number two, “Build a cheap set that navigates.” Our goal was under $10,000.
Geselowitz:
Well, you basically succeeded.
Parkinson:
Well, yes. Of course, $10,000 today would probably be at least $30,000, so we beat it by several orders of magnitude. Over the last year I was President and CEO of Trimble Navigation, an interim assignment; I’m back at Stanford now. My guess is that Trimble is selling chip sets that represent a whole GPS receiver except for the antenna, for somewhere around $25. It’s amazing, because initially it took racks of equipment to do this job. Because we have selected a structure that was very amenable to digital signal processing, that price collapse was enabled.
Another issue we wrestled with is which satellite orbits to use. We did not want to be in geostationary or geosynchronous orbits. The reason was these alternatives would force us to deploy ground stations on the other side of the globe, whereas, by putting them in some orbit that periodically passed across the United States, you could update the knowledge of where they were and what time it was on the satellite, then store that information in the satellite and continue to broadcast as it went around the Earth. That is the fundamental way we ended up with twelve-hour orbits. We also wanted to be reasonably high because we didn’t want the orbits significantly disturbed by the atmospheric drag. At the same time, by going high you had more visibility, more coverage on the Earth. So with an Earth coverage antenna and suitable power densities on the Earth, you ended up with the ability of twenty-four satellites to provide very solid, total Earth coverage. Basically our original Labor Day proposal is still the current system. We had three rings of eight satellites, and they now have six rings of four nominally, but I’ve also seen studies that they may go back to three rings of eight. The reason we advocated three rings of eight is to allow replenishment so much easier. You can store one extra satellite in each ring and then just let it drift around to a spot where you need a replacement. With the six rings, they obviously would need twice as many spares on orbit.
NASA Advisory Council, dealing with technological debris
Geselowitz:
It’s just as well, because the low Earth service areas are going to get crowded with all these commercial outfits and their telecommunications satellites.
Parkinson:
Absolutely.
Geselowitz:
It’s going to be awfully full down there.
Parkinson:
I also chair the NASA Advisory Council, and one of the things I’m very concerned about is deliberate, man-made garbage. I won’t call it debris. I mean technology garbage that comes down very slowly unless you push it down. When you get above about six hundred nautical miles, it does not come down very fast. Many of these satellites are going to stay up there a long time. You can say, “Well, space is vast.” Yes, it’s vast, but you’re going around and around. There are some people including me who really worry about it. Also, everything you put up tends to not stay in one little piece. It tends to have stuff fall off it, particularly during the launch process. That adds the debris, and it’s those small sizes you can’t track. In the case of GPS, they kick it up into a higher orbit. I worry about that a little bit, too, frankly. It’s monotonic. It’s just accumulative. It just keeps happening.
Geselowitz:
So when they need to replace one, they let the extra one in that orbit drift and they pick the one where it is and kick it up?
Parkinson:
Yes. They take the one that they no longer need and they push it into a higher orbit.
Geselowitz:
Why don’t they push it down in the atmosphere and let it burn up?
Parkinson:
Because the energy requirement to get down to the atmosphere would be enormous. The delta V to get down—it would be a lot cheaper at these altitudes to put it into an escape trajectory.
Geselowitz:
That’s a concept. You could stick one of those alien communication plaques on the side and just send it off, and that way we have these probes all over the solar system.
Differential GPS system; commercial support and contracts, military applications
Parkinson:
- Audio File
- MP3 Audio
(379 - parkinson - clip 3.mp3)
To get back to the central chain of the history, we formed the program office. I was able to recruit some additional blue suit talent, and I also had strong support from the Aerospace Corporation, particularly two individuals named Phil Diamond and Walt Melton. We had relatively small Aerospace support offices, but they were effective and useful to us. By 1977 and ’78, we were launching GPS payloads. I’m certain it’s not a record in terms of getting things up, but if you consider we’d never prototyped the satellites, it’s pretty impressive. We first went on contract for them in June of 1974. So that was the absolute standing start. I think we launched the first payload in 1977 and the first satellite in February of 1978. In terms of start to finish, it was a pretty good record. The guy that gets credit for that is Dick Schwartz, who headed up the team at Rockwell building the satellites. It was a highly interactive process with me and my blue-suiters. The reviews were almost continuous. It was very much a team operation—now known as integrated product teams. We didn’t know that name, but that’s what we were doing. The other gentleman who needs a lot of credit was a fellow named Gaylord Green, who was a young major at the time. He finally retired as a colonel and actually ended up running the Joint Program Office after my watch, after a couple of other stints at various things. Those were some of the heroes that we had.
By 1978, the concept was established, and the test results rolled in over the next two years and confirmed our claim of ten meter. The DOD, in its great wisdom, elected to deliberately perturb the satellite timing in a way that makes the ranging accuracy for a civil user less than it can be. We argued (I say “we” because I served on an independent review team for DOD on this), we argued very vehemently that this is a grave error. This would (and did) speed up the advent of differential correction systems to eliminate those errors. We had already demonstrated by 1978 the differential GPS systems. The first differential GPS systems were already exhibiting two and a half meters of horizontal accuracy.
Geselowitz:
How does the differential work? What percentage is that?
Parkinson:
The fundamental idea behind a differential GPS is that you put a calibrating or reference receiver on a known location on the ground. For other receivers in that locale, it will perceive the errors to be approximately the same. You measure the errors, and then via some communication line, broadcast those errors to the users, who can apply them as corrections to each of the ranging signals. This capability can produce millimeters of accuracy for static users. The state of the art in that ranges all the way down. Dynamically it produces a few centimeters of accuracy. There’s now a concept called WAAS, wide area augmentation system, that has been demonstrated as a prototype at Stanford, also as part of my research program. It is a regional differential system that employs about two dozen reference stations across the whole United States. It works by not just calibrating the ranging error but breaking its sources down. Thus, the corrections are valid over a wider range. For example, the ephemeris, the clock, and the ionospheric are all corrected. The ephemeris is the location of the satellite, and that’s a three dimensional correction. You transmit to the user a correction that automatically compensates for geospatial decorrelations. By doing that you have allowed the user to apply the correction for his particular spot in the United States. We are routinely demonstrating three quarters of a meter horizontal positioning accuracy anywhere in the United States. The interesting thing is that the corrections come as part of the GPS signal. You don’t need another data link. In fact, they come directly from a geostationary satellite broadcasting on the GPS frequency. The handheld device that your dad has, as soon as the FAA’s Wide Area System (WAAS) is fully deployed will measure position to the width of your shoulders anywhere in the United States.
Geselowitz:
So this was a commercial response. In other words, DOD decided we’re going to degrade the civilian signal because we want to be able to have our tanks within one meter, but we don’t want anyone else to be within ten meters. Where was it worked out? In the private sector? You said you already knew about this, but who actually put up the units?
Parkinson:
Well, it turns out the first papers were written by me and one of my students based on a concept that I had outlined to him. The first prototype operation was at Stanford, and it was done under FAA research money.
Geselowitz:
So the DODs has one thing, and then the FAA says, “Okay, well, if that’s the way you’re going to play the game, then we’re going to develop differential because we need it for civilian aviation purposes.”
Parkinson:
Yes, and it’s really ironic, because the fear, of course, is that GPS might be used to guide hostile missiles against us or friendly nations. If you look at the sophistication of integrating a GPS system into a hostile missile and you hypothesize they can do that, then you deny that they could apply a differential correction, in most people’s minds that doesn’t compute—that if they’ve gone the first step of integration, they can certainly apply differential corrections as well. It’s just not sophisticated.
Geselowitz:
Compared to the initial problem of putting a GPS system inside of a missile.
Parkinson:
Right, integrating it in a missile. This argument has fallen on deaf ears in part because the bureaucratic process, again, does not lend itself to people saying yes. It lends itself to people saying no. If you go to Washington, I think they’re puzzled over who is supposed to say yes to turning this off. Jim Schlesinger, former Secretary of Defense, and I put on a presentation at U.S. Space Command about a month and a half ago, and my part was, “It’s time to turn off the selective availability. That it’s not doing anyone any good. It’s making the safety of the skies and boats less than it should be, and yet it’s not accomplishing anything in terms of denying accuracy to a sophisticated user.” So when I got done with that presentation, the four-star general in charge of Space Command said he agreed with me wholeheartedly. But the question is, who would have the prerogative of turning it off? I don’t know. I guess the President. The difficulty with bureaucracies is that people are reluctant to make decisions because the later process is to try to figure out who to blame if the decision goes wrong. I think they’re partially stymied by that.
GPS NAVSTAR name and marketing
Geselowitz:
Can we talk about how you got to Stanford? Now we’re at 1978, and so NAVSTAR is what the GPS was called initially?
Parkinson:
GPS NAVSTAR. I can tell you how those names came about.
Geselowitz:
Why don’t you tell me that, and then tell me what happened next?
Parkinson:
It was called the Defense Navigation Satellite System, DNSS, and similar names. At some point I worked with a one-star general by the name of Hank Stehling, who was the Director of Space in the Pentagon, and I needed his support. I was in his office one day and he said, “Brad, you know I’ve been thinking about this system that you’re advocating. I think it’s more than navigation. I think it’s positioning, and I think it’s around the whole world, and I think we should call it the Global Positioning System.” Being a good marketing person at the time, I said, “Excellent idea,” and that was done. How it got called NAVSTAR was more interesting. In the civil structure there was a gentleman named John Walsh; he was in charge of DOD’s strategic systems, and our program was under him. It was not clear that he was strongly in favor of the system in the beginning. I had a colleague, Colonel Brent Brentnall, who worked for him. At some point he was having a discussion with John Walsh. (This is a true story now. It’s going to get a little seamy, but I’m going to tell you anyway.) He was having a discussion in his office, and the hydraulic pressures of coffee worked their way, and they both went to the men’s room. They had been discussing GPS. While they were sitting there relieving this tension, John leaned over to Brent and said, “You know, Brent, you ought to pick a catchy name for that system you’re advocating. You ought to pick the name NAVSTAR.” Brent quickly picked up the phone, called me, and said, “I’ve got a second name for your system.” So from there on we were called GPS NAVSTAR. It’s kind of interesting the way things cycle through.
University and private industry employment
Geselowitz:
Okay, that’s how NAVSTAR got there, but then how did you leave NAVSTAR and go to Stanford?
Parkinson:
In 1978 I had been at the program for six years. The system had been launched. It had clearly met all the major goals we had for it. The Air Force decided it was time to move me. By that time, I was traveling back to Washington as much as twice a week, almost inevitably the Pentagon. Their reward for this, and it really was a reward in their view, was to assign me as the Air Force Colonel to the Secretary of Defense, which meant I would be in the Pentagon all the time instead of part of the time. I thought long and hard. I mean, this is a plum job if you want to be a general officer.
Geselowitz:
Are you from the West Coast originally?
Parkinson:
No, I’m from the Midwest. I’m from Minnesota. It wasn’t Washington so much as the fact that I really enjoy building things and leading concepts and being a program manager. Being a staff officer, carrying someone’s bag, as heady as the experience might have been, just didn’t appeal. I had never really had star fever. I had gotten promoted extremely early, which astounded me because I tend to be outspoken. I called them and said, “No, I’ve got twenty-two years in the service. I’m retiring.” So I left there. I took a year off as a professor at Colorado State University. In my previous background I’d also taught astronautics and computer science at the Air Academy. My Ph.D. was from Stanford in guidance control navigation, and I had two years as an instructor at the Air Force’s test pilot school. I was ready for a break. I went a year to Colorado State University, mostly working in solar energy, then I got tapped and asked to come to Rockwell as the Vice President. So I was Vice President at the group level, the Space Systems group.
Geselowitz:
This is where?
Parkinson:
Downey (Los Angeles). It was the group working on the shuttle. GPS was under it, although I wasn’t very involved with that. I was much more involved with the shuttle, and also advanced payloads and some small satellites that they had.I went to Downey as a Vice President. All told I’d been in the Los Angeles basin for about seven or eight years I got very tired of that. I got an offer to be a group general manager and vice president for a company called Intermetrics. I headed seven divisions for them, including a commercial division, of which I was president. It was a subsidiary called, of all things, Plantstar. It was industrial productivity monitoring. It was monitoring everything that goes on in a factory in terms of cyclic machine operations. It was micro processor-based. Unfortunately, it was at a time when interest rates on money were seventeen percent and no one was investing in capital.
I was there for four years, and towards the end I got permission from my boss, the president of the company, to consult with Stanford on a project called Gravity Probe B, which is a NASA project to test general relativity.
Geselowitz:
Where was Intermetrics headquartered?
Parkinson:
Cambridge, Massachusetts. It turns out there was a gentleman at Stanford named Dr. Bob Cannon, who had been a mentor of mine since the early days at Stanford. He kept trying to get me to come with him. As a matter of fact, he’d been a division chairman at Caltech in charge of four of their departments. When I got out of the Air Force he had gotten Caltech to give me an offer to be an associate professor there. After some agonizing I decided I couldn’t do that, primarily because I was an engineer and Caltech was a science school. I felt I would be a little bit of an alien. Cannon subsequently felt the same way and went back to Stanford, where he’d come from. Cannon got me first of all a research professorship, then after about three years there, I’d had a lot of publications and a lot of good students so I became a tenured professor and continued to run this program called Gravity Probe B.
Geselowitz:
Which department are you heading?
Parkinson:
I head the Stanford GPS program which is part of the Aeronautics and Astronautics department.
Geselowitz:
Is Gravity Probe B in that department, or is it an extra departmental kind of program?
Parkinson:
No, it’s actually in one of the independent labs, where I also have an appointment, the Hansen Experimental Physics Lab.
Geselowitz:
And that’s the equivalent of Draper Lab at MIT?
Parkinson:
Yes, to some degree, but it is a part of Stanford.
Geselowitz:
Where they spun them off because the students didn’t want military involvement?
Parkinson:
It was very similar, but it does medium-scale physics experiments mostly.[End of tape 1, side a]
Education at MIT
Parkinson:
I was a Course XVI (Aeronautical and Astronautical Engineering) major at MIT for my master’s degree. I also went to MIT determined to take all of the foundation courses in Course VI (Electrical Engineering and Computer Science). In 1959 I had an advisor at MIT (who will go unnamed to protect his reputation) and I went to him with my course plan. I think you needed ninety units to get a master’s degree, and I had two hundred and seventy units on this thing, and it included taking all these Course VI courses. He looked at me and said, “Well, what do you want to do that for?” In other words, “Why can’t you just take some nice Course XVI courses?” And I said, “I came here with the express purpose of taking Course VI and Course XVI.” We got in a drag-out fight over this, and he finally signed the paper and threw it at me. I’ll never forget. He says, “Well, Parkinson, if that’s what you want, take it.” So I took the EE courses. That background was invaluable when I got into some of the signal processing aspects of GPS. I understood correlation; I understood how it was done; I understood mathematically the whole thing. So to have had that background was invaluable. I was lucky that I was able to persuade this guy. My body was on the railroad tracks. I wasn’t moving. That train was going to have to run over me to make me not do that.
IRE and IEEE
Geselowitz:
That gets back to this question of what is aerospace and electronic systems. According to the IEEE records that I have access to, you didn’t join IEEE until 1978, which is when you left the service.
Parkinson:
No, I originally joined it in 1958.
Geselowitz:
1958 as a student?
Parkinson:
No, as an Air Force officer. I was in PGAC of the IRE (Institute of Radio Engineers). Occasionally my membership has lapsed because of failure to get something in the mail.
Geselowitz:
Because you’ve been traveling back and forth to the Pentagon.
Parkinson:
Or whatever. But if you go back in the records you will see, I’m pretty sure--
Geselowitz:
I don’t know if it had lapsed when they brought the computer system up. If someone joins, lapse, and rejoins they’re supposed to have both. I’ll have to go back and look again, but I didn’t see that. I’ll have to double check that.
Parkinson:
You ought to look, because I definitely was receiving PGAC transactions before I went to MIT. That was deliberate on my part.
Geselowitz:
PGAC was the Professional Group on…?
Parkinson:
Automatic Control. I’ve always loved control, guidance, and navigation. They just go together.
Geselowitz:
After the merger that became the Control System Society, I believe.
Parkinson:
It’s got two pieces. There’s still now a Robotics and Automation Society, and there is the Control System Society. The Control System Society is more applied “here’s how you do this”; the Automation group has gotten much more mathematical, in my opinion.
Geselowitz:
So you joined the IRE from the control point of view, really?
Parkinson:
Yes.
Education in Aeronautics and Astronautics
Geselowitz:
Then you went to MIT to do aeronautics because you were in the Air Force.
Parkinson:
Aeronautics and Astronautics.
Geselowitz:
Aeronautics and Astronautics. You were told by a professor of Aeronautics and Astronautics that Aeronauticians and Astronauticians don’t do basic electrical engineering?
Parkinson:
No.
Geselowitz:
So what were they doing?
Parkinson:
They’re doing Aerodynamics, fluid dynamics, structures, aero-elasticity, flutter, and they were doing the Draper Labs stuff. My master’s thesis was in the Draper Lab, and it was an adaptive auto pilot for an airplane.
Geselowitz:
Did it need electronics for that? They were thinking more from the mechanical and geophysical points of view?
Parkinson:
I don’t think that the process was as logical as you’re suggesting. I don’t know this for a fact, but my guess is that MIT was what they call a formula school in which they went through and counted how many noses were sitting in classrooms, and that determined how many professor slots they got. Stanford in particular does not do that at all. They strongly encourage cross-fertilization across departments, and will not tolerate any barrier to that happening. Control is taught by at least three or four different faculty, and it’s a common group across computer science, mechanical, and electrical and aero-astro engineering. So you’ve got four departments contributing professors and contributing a rich diversity of experiences. GPS has a very sophisticated control system embedded in it, which is the phase-locked loop that tracks the GPS signal. Usually these are third order loops, and in most cases they’re digitally controlled. Setting those parameters is a very fine balance between signal and noise and tracking ability to hang on. It almost came full circle there in terms of control theory.
Interdisciplinarity in IEEE societies, education, and GPS
Geselowitz:
What IEEE Society is your membership in now?
Parkinson:
I’m still a member of the IEEE as a whole.
Geselowitz:
Are you a member of the AES-S (Aerospace and Electronic Systems Society), that is sponsoring this interview?
Parkinson:
I don’t think I am.
Geselowitz:
You don’t get their magazines?
Parkinson:
No, because my bookshelf has gotten so crowded.
Geselowitz:
Stanford has a good library.
Parkinson:
I’m certain I’m still a member of the IEEE. I’m a member of the AIAA, and the Royal Institute of Navigation; both were kind enough to make me a Fellow. The Institute of Navigation did that too, just this last June. I was surprised because they just started their Fellow Program.
I should mention something else. GPS fits in several places in IEEE technologies. It fits in Aerospace Systems and is the reason for your being here. It certainly fits in both of the Control groups. It fits in a unique little conference IEEE sponsors called PLANS (Position, Location, and Navigation Systems). As a matter of fact, they give out an annual award called the Kirschner Award. The Kirschner Award is named for the head of APL at the time of the Transit satellites. They happened to give me the first one of those, too. The point is that both the technology and implementations of the GPS system is almost insidious—associated with many IEEE interest groups.
Geselowitz:
I’ve interviewed Cary Spitzer on avionics, Saj Durrani on space systems, and then Warren Cooper on microwaves from the older generation. I keep asking everyone the same question: Why is this a society? What you have to do in common with each other? You were able to go from aerospace to a division doing factory work, factory control. On the one hand the thread is the aerospace, in particular, even more than the aeronautics, and the aerospace angle pulling them together. But really they see that because of the problems involved with those systems and the fact that there was a lot of effort between NASA and the military, a lot of money was put into it after Sputnik was launched for the next twenty years, that’s where all these people have to be who are working on very complex electronic systems where different components or subsystems had to communicate to each other in real time in very complicated ways that the factory application wouldn’t necessarily see. We’re going to interview Henry Oman, an expert on power systems. They had to solve all special power systems. You’ve got six different systems, you’ve got to power them all, and they’re in one aircraft and it’s hurling through the air at however many miles an hour. There’s always problems we have to solve. Or like GPS, as seen in this diagram from your article.
Parkinson:
Yes, and in the case of GPS it’s even much more complex than that little diagram.
Geselowitz:
Still, just to take this one little diagram, you’ve got a control system with five monitor stations, three uplink stations, and a master control station. You’ve got the user segment. You’ve got the space segment, which is twenty-four satellites with atomic clocks transmitting coded RF signals. Yes there’s a Society concerned with control systems, but here you were the guys who really could take all these disparate pieces and build them in a system that really works.
Parkinson:
I think that’s accurate. But in many cases you see different pieces that are perhaps brought together. You could put paint on your house or you could put paint on your barn. The interaction of the structure of the house with the paint is negligible. You can say, “Well, when the paint wears out then the structure will collapse.” I don’t know about that. But I think what makes the system aspect so critical and interesting is that the interactions are in both directions. For example, the structure you put on a satellite, you put a nice strong structure to withstand launch. If you make it strong enough it will be heavy enough so you won’t launch it. Then you start to worry about the thermal aspects of where you hang things. So all of a sudden you see technology and the various engineering disciplines interacting through these systems, which have feedback paths that go in all directions. Whereas, if somebody gives me a spec for a black box, I can probably build a black box. All of the sudden you say, “Wait a minute. To your background, it’s got to be hardened.” That puts some constraints on us. It has to worry about EMPs on wires; that’s putting some constraints on us. Oh by the way, it has to do something functionally for whatever it is embedded in. As a matter of fact, it has to interact with this other box over here.
Geselowitz:
Coming back to the weight of the box, I think it was Saj Durrani who said at one point there was a big prize for each ounce that you saved. If you came up with a suggestion without reducing functionality that would save an ounce, you would get a $1,000 bonus or something like that.
Parkinson:
Yes, and it’s well worth it. I see this process as being about very bright people who have broad disciplinary backgrounds and can bring in engineers or specialists in a lot of areas, and ensure that the interactions of these various pieces—the technology, the boxes, the functionality that’s required—that interaction is done properly. The difficulty is, the more powerful our computers, the larger our systems become, the number of undesirable interactions grows as n factorial. In other words, each interaction could be good or bad. For example, one of the things going on GPS today, and I’m fighting this out at the national level, is the possibility to upgrade the satellites to greater power. You say, “Hold it! Even if I put on greater arrays and pump out greater power, I’ll roast the whole satellite with the heat of inefficiency. It doesn’t have the ability to reject that heat. So what do you do about that?” Well, it turns out you do two things. You add to the thermal radiators, but you also go in to the guys that are designing high power amplifiers now, because it’s their inefficiencies where the heat ends up. The power I spray out into space, I don’t worry about that. It’s the power that doesn’t go out into space. It turns out that the modern technology can improve that in terms of heat rejection per watt by a factor of about two. That’s an enormous improvement. Because what it says is I can go to twice the power out and maintain the same thermal heating in. Well, that’s an interaction, you see.
Geselowitz:
Because it also makes you wonder why you hadn’t done it before.
Parkinson:
The technology wasn’t there. It’s solid state electronics that now improve on the switching efficiencies and…
Geselowitz:
But that’s how you need to know what the other fields are doing.
Parkinson:
Absolutely.
Geselowitz:
Because they might not know you need that application, and you might not know that they— some lab somewhere else—has developed a new semiconductor.
Parkinson:
Yes. So the people that do this well are people that did something complex, be it airplane designers, spacecraft designers, maybe tank designers. The people that end up being able to do this well as a general process are people who have done it before and have been exposed to all these complexities. In many cases, you’ll find that people mentally can’t handle all of this. I know some engineers, they’re not into this. They say, “Just tell me what the specs are and I’ll go design my box.” They’re not used to interacting with all the rest of the team in terms of structure and heat and vibration resistance, functionally what it has to do, the timing and things. They’re not into that. What about the redundancy? What about what happens when this breaks and I’m not there with a screwdriver to replace that box? I think the thing that’s nifty about AES-S, is that it has people of like generalized skills, and those generalized skills, it turns out, can be applied to a lot of things. But they’re not even recognized. At MIT, Course XVI is catching on. They’ve now got a whole group working on what they call system engineering. I tried it at Stanford and I can’t get it through. The other engineering discipline says, “What is it you’re talking about?” They’re used to dealing with computers.
Geselowitz:
From what you said, though, Stanford should be encouraged by this kind of thing. Maybe if you keep talking to people eventually…
Parkinson:
We do. Virtually every dissertation in aero and astro ends up with a system-oriented demonstration proof of what it is they’re talking about. Be it control of farm tractors, control of space craft, whatever it is, we end up with demonstration proofs, which are system engineering. The students we turn out have that. But if we call it system engineering, the other disciplines say, “What’s that?” So we don’t call it that. It’ll come around. Recognize that the digital computer has accelerated this process. Now I can take a digital computer, put a lot of sensors with A-to-D converters on them, surround it with a lot of actuators to do something else, and I suddenly have this enormous engine of flexibility. Well, flexibility can get you in trouble or it can save the day—it can go either way. It’s fascinating to watch people’s mindsets and how they do this. It’s not a totally mechanical process. There’s still a lot of art in understanding where the problem is to be in attacking that. It’s an interesting piece of your society. There’s still a lot of room for engineering and creativity.
Predictions on the future of GPS; Y2K
Geselowitz:
I want to ask you where your thoughts on GPS were going. As a transition, I’m curious. You guys had, several months ago, what was billed as the Y2K dry run. It was really first hyped in the press and then apparently fizzled. Because we know in the ’70s, memory was expensive, it was heavy, so you didn’t want to waste a lot bits. We talked about a clock, you had to have a timer, you had to have some kind of date calendar. You chose a certain number of bits to count weeks, and it happened to roll over in 1999.
Parkinson:
It’s called the week number roll over or WNRO.
Geselowitz:
Just say a few words about that for the record, and then tell me what’s going to happen with GPS in the future.
Parkinson:
I got a call from the BBC, and they had heard about this horrible disaster that was about to befall them, and they wanted Professor Parkinson to hold forth on this subject. I started laughing and said, “Gee, you must be really hard up for news.” She stopped for a minute and she started laughing. She said, “Well, it is August in Europe.” I said, “Well, let me tell you. I don’t think anything is going to happen at all. I think that this is totally overblown as a disaster. Almost every manufacturer I know, for any of their modern equipment, has used the simulator and run right through WNRO, and I just don’t think it’s going to be a problem.” At which point they said, “Thank you very much, Professor Parkinson.” They didn’t want to hear that.
Geselowitz:
Not a good sound bite for the U.S.
Parkinson:
Not a good sound bite. It turns out, just this last week, I was hiking in the High Sierras. We’ve got a place up at about 7,200 feet, and you can get back in there in the hills and mountains and lose your way. I’ve got three or four hand-held GPSs. I pulled out an old Trimble. The batteries were totally dead. I hadn’t initialized this thing in probably three years. I stuffed the batteries in, walked out in the parking lot, turned it on, it took about two minutes to gather all the almanacs because it was totally out of date, and it navigated perfectly. It ran right through WNRO like it was not even there, which was my prediction. I think that we were constrained. As I remember it, we were constrained with ten bits. So what? It’s two to the tenth; it’s ten twenty-four?
Geselowitz:
Right, minus one. One thousand twenty three weeks, which is just under fifty-twenty years.
Parkinson:
Well, you always add the zero week or something, but it’s basically 1,000 and twenty-four and it turns over. I forget when the start time was for that; it was probably ’77 or sometime like that. When the thousand and twenty-fourth week ran up, it flipped over. But GPS handled it well. Incidentally, that’s not to pooh-pooh the efforts on Y2K. Because GPS is relative time-based, there’s a logic that says once you go beyond that mark, that whatever transit that might have occurred at that second is gone away. What the BBC person wanted me to talk about was planes crashing. I said, “I don’t think any planes are going to crash at all. I don’t think any planes are even going to know it happened.” They didn’t like to hear that at all.
Geselowitz:
Are you recommending that if I get on a plane on December 31 in the evening, it be a 777 that has GPS as a back up? Because we know that GPS is not going to be effected.
Parkinson:
That’s interesting. I don’t know all the systems the FAA has that are time-based. Certainly their radars aren’t. Radar just records what’s there. I guess I’d feel a little more confident if I had GPS. One of the things my students are doing, we also are using GPS to measure aircraft attitude, and one of my brilliant students showed that it was possible to put four antennas on the extremities of an airplane and measure yaw, pitch, and roll to a tenth of a degree at ten samples a second. It’s the second statement that’s startling. Ten hertz. So you can put that sensor in the middle of a control system, which we’ve demonstrated with both model airplanes and larger airplanes. But the point is that we now have taken GPS and used it for the automatic landing of 737s all the way down to touch down. Another of my students has taken that attitude and created a virtual reality based on your position and which way your airplane is pointing that looks exactly the way it looks out the cockpit. It’s like the Microsoft flight simulator except it’s real. I sat at the end of the runway in Juneau and I’m looking out, I’m right on top of the numbers, I can see the number 06 or whatever it was there. I look at my display, and it shows the 06. It’s obviously within two or three feet of where it is. It’s using that wide area system, the attitude, and this display. All of a sudden we’re coming back with system engineering. It’s integration of all those technologies and knowledges that are coming together. Back to your question, yes, I’d feel a lot better if I had that display and the GPS attitude and GPS position, because I know I can get on the ground.
Geselowitz:
Is that where you think GPS is going? Are we going to have complete take off flight and landing by GPS? I don’t know a lot about naval navigation, but what about complete piloting of ships by GPS?
Parkinson:
- Audio File
- MP3 Audio
(379 - parkinson - clip 4.mp3)
I was on a panel for advanced systems that was helping the FAA, and one of the things I proposed was totally autonomous cargo airplanes. They would not let that recommendation get into their final report. It was too threatening. It turned out that by the time I had written that, one of my students had already demonstrated the whole thing in a model airplane: complete take off, flying at two flight levels with nice square patterns, within about a meter of where he was supposed to be, and bringing it back and putting it down on the ground within a meter of where he had told the airplane to land. Completely hands off. There’s absolutely no reason that you can’t do that today with a real airplane. There are complexities, and the biggest thing that GPS has to do is get an established track record for robustness. By that I mean that it is always there, that it does not lie to you. The airplane people tend to call this integrity. What I see as far as GPS itself is concerned, if the government finally comes to its senses, is they’ll add additional signals and additional satellites, they’ll do this wide area thing, and the result will be a much more robust system. I also see this in probably half the farm tractors in the world, many of the pieces of construction equipment. Again, students of mine, have already demonstrated robotic tractors working on a rough field and holding within an inch of a straight line or a curved line. That is about a factor of three or four better than the best driver that this experimental farm could conjure up. If you think about dirty, dusty, terrible, boring jobs, driving tractors is going to go the way of auto painting. The reason your paint never sags anymore is it’s all painted the same way by a robot who is trained by a master painter. I am convinced that you’ll see robotic farm tractors, you’ll see robotic construction equipment, you’ll certainly see robotic airplanes. My guess is it will probably get down to the point where there’s one pilot flight engineer rather than two, and he’s mostly a back up. The trouble is, you’re going against a really tough union.
I’ll give you another example. When you land an airplane blind, you’ve got a pair of needles, and these needles are scurrying around, and they’re based on the old ILS glide slope localizer and glide slope technology. As this needle moves off or moves back, it takes a lot of training, because you’re moving. When I went to Juneau, I hadn’t flown on instruments in probably fifteen years, and my students pop me in the left seat. We take off, they flash this display, I come dead in on Juneau Airport. When I say dead in, I’m within five or ten feet of being right on the path I’m supposed to be on. I claimed a twelve-year-old could do it. The point is that it’s enabling in a lot of ways. One of the things it’s doing is making human paths more intuitive and useful in that and a hundred other ways. They’re using them in open-pit mines right now.
Geselowitz:
How about in the automobile? Is it ever more than just a gimmick? I mean, now expensive cars like Cadillac say, “Oh, we know where you are. We’ll send the police if you lock yourself out of your car,” or whatever. Do you think it’s actually that I’ll have a display and the traffic report.
Parkinson:
Yes, it’s going to become widespread. It will probably do it in a lot of dimensions that you don’t realize. It’s like your cell phone. You start out, a couple of pioneers have it, and you say, “Oh, it’s just a gimmick.” Then pretty soon you realize that if you’ve got a cell phone you’re a lot more efficient. If I’ve got something in my car and I’m going to somewhere that I don’t know, and I just punch in the address and it always gives me the most efficient way to get there. Initially everyone’s going to pooh-pooh it and say, “I know how to navigate streets.” After a while, it’s coupled into traffic reports, it steers you around the density, it shows you the short cuts. It’s going to come. It’s in Tokyo right now because no one knows where any of the street addresses are. I think that the Japanese are putting it into cars at the rate of 50,000 a month. Think beta sites. Think beta testing. Think we’re going to get it right and then ship it to the United States. That’s sort of what’s happened.
I think it’s interesting. To me it’s humbling, because I didn’t really volunteer for that job. I sat before General Schultz and I asked him if he was going to make me a program director. He said he couldn’t promise that, and then I said, “Then I’m not a volunteer.” It turned out that after I got fifteen feet outside of his office he called personnel and shipped me over there anyway. But meanwhile I stuck a stake in the sand saying that if I was going, I was going to be in charge. I think he kind of liked that, actually. It’s interesting how things come up. Bob Cannon has this law. He came into my office. He was my mentor.
Geselowitz:
Right. Is he still at Stanford or is he retired?
Parkinson:
He’s emeritus but he’s still at Stanford. He came into my office one day and he said, “Brad, Brad, I’ve figured out why everything happens.” So I sat down. This was going to be big. He says, “Yes. One thing leads to another.” If you think back to my taking essential Course VI courses (EE), and suddenly I happened to be placed in charge of GPS. I had all the right background, but I don’t claim that the underlying pieces that we synthesized—I didn’t invent all those. I had a strong hand in some of the innovations in the testing area and also the exact selection of how we were going to do things, but fundamental technologies were pieces that we stole. I think what was needed was someone to pull all that together in one package and tie it off with a bow and say, “That’s GPS.” I was lucky. I was lucky to be there.
Geselowitz:
We’re all lucky that you were. It’s a toy for some people and it’s incredibly useful for others, and soon it will be useful for everyone.
Parkinson:
How does your brother navigate? Do you know? Do you know what he’s using for reference?
Geselowitz:
Why don’t we end the formal interview. I’ll turn off the tape. I don’t think everyone wants to hear about my brother. But thank you very much. I appreciate it. It was great.
