The father of GPS

Bradford Parkinson, lead architect of the Global Positioning System

Of all the space-related inventions we rely on in our daily lives, the most transformative might be the constellation of 31 Global Positioning System satellites orbiting some 20,000 kilometers above Earth. That’s in no small part due to Bradford Parkinson, the Air Force officer who in 1973 led the creation of the original GPS architecture, a consolidation of the various satellite navigation projects in the works by the U.S. military. But to hear Parkinson tell it, there was some luck was involved because of the various “forks in the road” that could have led his career elsewhere. And of course, like all multibillion-dollar efforts, GPS was the product of years of work by thousands of individuals. I reached Parkinson by phone at his California home to hear about his path to the career-defining project, the lessons for future position, navigation and timing systems, and some of the exciting future applications that GPS could make possible. 

---

Q: How did you become interested in engineering?

A: If you look back, there are a lot of forks in the road. One was a teacher strike in Minneapolis. My dad pulled me out of the public high school and put me in an all-boys school that taught algebra to eighth graders. I thought I’d gone to heaven. I guess I was always destined to be an engineer. I got an RCA vacuum tube manual in high school, and building electronics stuff and controls became natural for me. I went to the Naval Academy in June 1953 and graduated with the class of ’57. Afterward, I chose to go into the Air Force because they had an advanced degree program, and they actually used the education. The Navy would maybe give you one tour using that education, then send you out to sea. After two years, the Air Force sent me to MIT.

Q: That was good timing, because you got a sense of what would be beneficial for the service and your career.

A: That’s true. I very much was at the beginning of the Space Age. MIT had transformed the name of the aeronautics department to “Aeronautics and Astronautics.” I had a whole series of courses on inertial navigation, how gyros and accelerometers work, how the system is put together, what the error equations are, the things you should look for. And fortunately, my next Air Force assignment was testing inertial navigation systems — so you could say it was kind of in my kit bag. My MIT associations were very valuable. Years later, when I was running GPS, Charles Stark Draper — someone I regarded as a semi-mentor — came to Los Angeles. My boss had me talk about GPS. Draper didn’t like anything with radio navigation. He said, “You don’t realize that in time we’ll have a whole inertial navigation system in this,” and pointed to his watch. He was before his time, but his vision was right.

The trouble is inertial navigation systems inherently are open-loop drifting things. As soon as you have any kind of bias or misalignment — which you always have, even if minuscule — it’ll drift off unbounded. So an inertial and a GPS are a natural and wonderful marriage, and the good GPS sets do that.

Q: What happened after your Ph.D. at Stanford?

A: Another fork. I was to be a professor at the Air Force Academy. Last minute, two guys show up from Edwards [Air Force Base in California]. The lieutenant colonel says, “We want you to be on the staff, teach academics, run the Simulation Division and go flying with the test pilots. Would you like that?” Would I ever. That was the pinnacle of the flying Air Force, and these guys were fearless. They let me actually do the flying, outfitted in a full pressure suit. We flew an F-104 up to over 90,000 feet. The adventures!

Q: After that, you taught at the Air Command and Staff College and the Air Force Academy.

A: I got assigned to the Department of Astronautics and Computer Science, teaching space mechanics. One day, into my office bursts academy classmate Rick Wills. The Air Force had been trying to put together a new version of the AC-130 gunship, including a digital fire control system that points the airplane so that its fixed-side firing gun hits the target. Digital gave a lot of flexibility, but also the flexibility to be wrong. Evidently, it was wrong most of the time. Rick said, “You’re the right guy to help fix this.”

Next thing I know, I’m heading for Wright-Patterson [Air Force Base in Ohio]. The key was either we get this fixed in time for the dry season in Laos, or rip everything out and put it back to the analog configuration.

The first AC-130 arrived in South Vietnam in 1967 for U.S. combat operations in Vietnam and Laos as part of the Vietnam War. — JC

We went down to Eglin [Air Force Base in Florida], and the test was in the Gulf of Mexico. The airplane shot the hell out of the raft, and the general says, “You’re good to go.” After that, I thought I’d go back to the academy and teach. But there were still some bugs in the system, so next thing you know, I’m in Ubon [Air Base in Thailand] flying combat missions every night, although technically I’m a professor.

Q: After the academy, you also taught at the Naval War College. Tell me how you got back into the field, so to speak.

A: I was at the Pentagon, working for Glenn Kent, head of studies and analysis. My folder came to Kent’s deputy, Bill Manlove, who asked me, “Do you like to build stuff or study stuff?” I said, “Building stuff.” And he replied, “I’ve got an assignment you’ll really like: chief engineer on the Advanced Ballistic Re-Entry System program.”

Established in 1963, ABRES was a Defense Department initiative to develop and test reentry vehicles, including the Mark 12 warhead for the Minuteman III ICBMs. — JC

Again, an abrupt path taken. I arrived in LA, and the general said, “Brad, I don’t know a lot about the technical part. You decide what we do.” For an engineer, how could there be a better deal than that? We were testing a lot of stuff: maneuvering reentry [vehicles] and using guidance systems in ballistic missiles. Then, as one of my friends said, “A hundred days later, Parkinson disappears.” Lt. Gen. Kenneth Schultz, commander of the Air Force’s Space and Missile Systems Organization, pulled me into his office one day. “We’ve got this little satellite navigation program called 621B,” he said. “I’m thinking you’re the right guy on that program.”

Q: That was November 1972. How did you go about getting what became GPS off the ground?

A: 621B was floundering in competition with a Navy program, Timation, and a Navy navigation system called Transit. The Naval Research Laboratory claimed they had invented GPS, but their system was two-dimensional, required every user to have an atomic clock and used a signal structure for ranging. It was passive ranging but required a different frequency for each satellite. It also was trivially jammed. The Air Force wanted to put up a demonstration of satellites in 24-hour inclined orbits over the western United States. In August 1973, I went before the Defense Systems Acquisition Review Council and briefed the program I had inherited. What happened is I failed like hell. So when I got called in to Malcolm Currie’s office, the No. 3 guy in the Pentagon, I thought my career was over. But he said, “You and I know that there’s better ways to do this. Take the best ideas, come in with a new proposal, and I think we’ll approve it.” So we spent Labor Day weekend hashing out the ideas to synthesize the new program, which became NAVSTAR/GPS. And what we chose was, based on an Air Force study, the hardest of 12 alternative techniques for navigation using satellites. We had to simultaneously passively range to four satellites. It implied you have quite a number up there to populate your constellation to ensure you had four. By December, I walked into that same meeting, and the council said, “Let’s go do it.” Four years later, we had satellites on orbit. Within five years, we had tested and built 12 different kinds of user equipment — emphasis on the word “we,” because those young Air Force officers and the contractors we picked, and the cadre of aerospace engineers that I retained, they made it happen.

The first GPS prototype satellite was launched in 1978, the first commercial receivers were marketed in 1984, and the 24-satellite constellation was declared officially operational in 1993. — JC

We could foresee it was going to be big. The first civilian receiver to lock up was built by students at the University of Leeds in England. That demonstrates we had freely given out the specs on the system and how the signal could be received.

Q: So in hindsight, that initial rejection was a blessing in disguise?

A: The beauty of what they were forcing me to do is show residual value. Instead of simply a demonstrator, we put up a piece of the operational constellation. It was no longer, “put something up and throw it away when you do the real stuff.” Instead, if everything worked out, we had the first quarter of the constellation on orbit — and if you added another 18 satellites, you had the whole constellation.

Q: What lessons does this experience offer for future private position, navigation and timing systems?

A: The people who use PNT are not all created equal — they have different needs; they have different sensitivity if it doesn’t work. They have different size pockets to spend, different demands in the time to field it, and whether five, 10 years from now, they’ll still be able to use it. Cell towers, lower-Earth satellites, optical cables all have a role potentially. You have to be careful about the high end. Category A is the highest end for dynamic users. For a farmer, that’s an accuracy of a couple inches dynamically. Category B is also three-dimensional, but on the order of a meter. A third category is perhaps 10-20 meters, conceivably worse and not necessarily three-dimensional. Another category is static. If I can process things statically, I can get down to millimeters with GPS. That’s the standard for surveyors. Against those categories, you have to measure the proposed solutions. I was chair of a review committee for FAA on eLoran.

He’s referring to the proposed enhancement to the U.S. Coast Guard’s Long-Range Navigation, LORAN, a ground-based radio PNT predecessor to GPS that is now dismantled. — JC

I strongly suggested that the Air Force, FAA and the government field eLoran as a backup and a deterrent to interfering with it. But eLoran is two-dimensional and probably only guaranteed to 10 or 20 meters if you apply the same rigor you do with GPS. If you’re talking safety of life, three dimensions, one meter stuff, there isn’t anything that could do that except for GNSS GPS or its equivalents by others.

Q: What do you see as the biggest threats to GPS reliability and integrity?

A: A great question. Physical vulnerability, yes; the Russians could, as an act of war, start shooting at GPS satellites. I don’t know how expensive, effective and reliable that would be. There are ways to make very inexpensive GPS satellites — in essence, proliferate the constellation and give them a targeting problem. That’s a war game answer to a war game problem. Master control segment, you could take out physically. Each satellite stores where it thinks it’s going to be in the future, and we update that every 12 hours or more frequently. If you fail, that ephemeris gradually gets ungood.

The real problem: jamming and spoofing. Nobody should be spoofed — nobody. There are enough checks and balances that you should be able to tell if someone’s attempting to give you a false position, but we may not have the means to ride through it. If we have an inertial, we can ride through such things for a little while. The elephant in the room is jamming, and the reason is that the GPS signal is tiny. What do you do about that? Back in 1973-74, I persuaded the avionics lab under my program to sponsor Collins Radio to demonstrate the ultimate jam-resistant receiver. If you add deep integration with inertials, null steering antennas and perhaps a more sophisticated signal itself, you can fly right near a 1-kilowatt jammer and never see it. The U.S. government has a regulation that forbids building antennas with more than three elements. In addition, they cannot be sold, installed or used in the U.S. Only this January has the government drafted language for the International Traffic and Arms Regulations that would abandon this stupid restriction.

He’s referring to the State Department’s January notice in the Federal Register that it plans to update the U.S. Munitions List to end restrictions on controlled reception pattern antennas, or CRPAs, for PNT. — JC.

It’s inexcusable that we don’t have this solution in place for civil users.

Q: Could we have predicted what GPS has become?

A: Of course not, but we could foresee quite a bit, long before we could do those things — before Reagan had guaranteed it, before Clinton turned off deliberate disturbances. And sure as hell before very large-scale integrated circuits drove the cost down. To give the credit where it’s really due, the engineers have taken ideas that in retrospect look pretty obvious and actually put them into systems that work. It’s the execution that really counts. A lot of engineers deserve credit for making those visions reality. By 1984, the cost of the receiver had gone down and the real-time kinematic was shown to be robust. That opened a whole new panoply of opportunities, taken advantage of by people who probably in 1978 had never heard of GPS.

Q: What do you see as the most unexpected application?

A: Back in ’78, I knew we could navigate airplanes. I didn’t think we could blind land them. But in ’92, we did 110 straight blind landings with a 737 loaned by United, sponsored by the FAA.

Q: Does that pave the way for autonomous cargo planes?

A: It’s totally feasible. We do it all the time today but on a very small scale: UAVs, unmanned aerial vehicles, now deliver things. It’s going to happen sooner or later. The bureaucracy has to screw up their courage and go down all the corner cases. Let’s start moving in that direction and see what happens.

Q: Switching gears, what guidance would you give on technical leadership, based on your experience?

A: Some program managers had a tendency to sweep problems under the rug. I was the opposite. I don’t want my boss surprised. If you’re running a program, you got to make decisions and get on with it. Kicking the can down [the road] doesn’t do much good if you have the right problem, if you don’t do anything about it. One of my friends calls it admiring problems. Sit around and admire that problem. Boy, that’s a great problem.

Q: You’ve been an AIAA member since the ’60s. What’s been the value for you?

A: The No. 1 value is the interfacing, the networking and listening to papers. Technical knowledge gets out, but then there’s the ability to interact with the people doing that. I can think of several cases in which I reached out at a meeting with a guy and said, “I’d like to collaborate with you on a paper,” and we did that. It gives an opportunity to not just network, but also to collaborate and push to state of the art, push the knowledge.

The internet has done more than books. Not only do I have the ability to print, now I have the ability to stick all this on the web, do searches, rapidly get technical papers or philosophical papers on any subject, and get knowledge and thought — good, bad, right or wrong — almost instantaneously.

The scary part now is AI. I don’t have a lot of experience in AI, but this morning at our Retired Active Men [group], I knew the speaker was going to talk about cars so I gave two nice poems about cars. Then I informed the membership that I had ChatGPT write those poems — understand what I’m saying? Virtually undetectable as not being from a pretty darn good poet. Where does that lead us in synthesizing new knowledge or at least amalgamating knowledge? It’s going to have an impact. I hope that we put checks and balances so that what happens is generally for the benefit of everyone, not malicious, but I don’t know how to flesh out that statement.