AN “ANSWER TO A MAIDEN’S PRAYER.”

The second half of senior Jet Propulsion Laboratory scientist William O’Neil’s career at NASA’s famous unmanned spaceflight center was almost entirely dominated by his management, as Project Director, of the Galileo mission to Jupiter – a post he held for 18 years. While it would be highly accurate to say that Galileo was a spectacular success – the craft spent eight years orbiting Jupiter – it would be no less true to say that probably no craft in the history of unmanned spaceflight has had as many problems and set-backs as Galileo – and lived to tell the tale. An initial grouping of hurdles involved the multiple delays in launching the craft. Among other reasons, these were because the powerful rocket which was supposed to have boosted the craft from low Earth orbit – the position where it was “dropped off” by the Space Shuttle – was never actually manufactured, necessitating multiple rethinks of the mission and trajectory. Later, with all that seemingly solved, the explosion of the Space Shuttle Challenger in 1986 delayed Galileo further.

A second serious Galileo problem came much later, many months after the probe left Earth, with the failure of the large umbrella-shaped high-gain antenna to open – a disastrous set-back that could have terminated the program entirely. It turned out that the various postponements of the mission resulted in the probe’s being trucked back and forth across the United States – and that as a result, some of the lubrication in the antenna’s complex spring-loaded opening mechanism was scraped away. The result was that Galileo was left without its primary means of communication with Earth.

-- Michael Benson

[This is the second part of a three part interview.]

Q: Anything after successfully landing two Viking probes on Mars in 1976 must have seemed like a bit of a let-down. What did you do after that?

A: After Viking I ran a department at JPL, the mission design department, for about three years, which was responsible for the mission design for all the projects that JPL was doing. But as a department manager, of course I wasn’t dedicated or focussed on any one of the projects, but only on the management of the people who had to do those projects…

Q: So you probably didn’t enjoy that as much.

A: Well, I did at first. Because it was a clear promotion, and in my view it was one of the best departments to be running. I mean, ‘mission design’ – what could be better for someone like me? And I was very happy with that. But then the then-manager of the Galileo project – this was early in the development of the project – called me up and essentially told me he had an offer I couldn’t refuse. [laughs] Which was to be the manager of science and mission design for Galileo. Ironically we thought at the time the mission was designed – how naïve we were! We re-designed it many times after that. But it turned out to be a wonderful assignment. But I never expected I was going to be working on the project for 18 years. Which it turned out to be.

Galileo was of course the greatest experience of my career. And when I took the job in 1980 I hadn’t the wildest dream that I’d ever be in charge of the project. And I was – shortly after launch I became project manager and went through the entire primary mission, including the first tour, which was essentially an eleven-orbit tour of encounters with the Jovan satellites.

Just a footnote: of course the Voyagers were the vanguard in terms of outer planets spacecraft. The Pioneers preceded the Voyagers to the outer planets, but they were very rudimentary, very primitive. The Voyagers were the first true outer planets spacecraft, they started the long-term development of what was called thermo-electric outer planets spacecraft. Because we knew we’d have to go to nuclear energy to do anything at that distance. The Pioneers had nuclear power, but like I said they were very rudimentary. So the point being that at least when it comes to the power supply, the path-finder was the Voyagers. They had similar power systems [to Galileo], using plutonium and so forth. And of course Galileo would be the first in orbit around an outer planet and the first to put a probe into the atmosphere, to measure the atmosphere of Jupiter. So the important events on Galileo were, of course there were a total of five re-programmings, and we were…

Q: Before launch? Re-programmings of the launch date, you mean.

A: Yeah. And the mission. Originally we were set to launch in January of ’82. Then it went to ’84. Then it went to ’85. Then there were two different versions of a ’85 launch. Then it went to ’86. And we were at the Kennedy Space Center with the spacecraft, and the newly-developed Centaur upper stage for the Shuttle was there, ready for flight at the end of January. When the Challenger happened.

Q: And that disaster resulted in further delay of Galileo. Which was supposed to go up with a shuttle. And which did, in the end, go up with a shuttle.

A: It did. That was the longest and most massive impact, because the others basically had to do with missed launch opportunities. The Jupiter launch opportunity is every 13 months. And in fact ironically both of those first mission designs were to have gone by Mars to get a gravity-assist, to help with the energy problem to get to Jupiter. And then in ‘85 that wasn’t available. So one of those missions, the first re-programming, actually split the orbiter and the atmospheric probe into two separate missions. Because the launch opportunity wasn’t nearly as good, and the mass of both the orbiter and the probe had increased. And the capability, the advertised capability of the shuttle and its IUS, which was being developed for this mission and other planetary missions, had decreased.

Q: IUS standing for?

A: What that actually stood for, ironically, was Interim Upper Stage. Implying that there would be another one, a better one.

Q: Upper stage being something in the shuttle payload bay which you could use to launch an inter-planetary space probe.

A: Which could be used to accelerate out of Earth orbit on the escape trajectory to wherever you were going. The bandits actually kept the acronym, kept the “I”, but changed it to “inertial.” They figured out a word they could use with the “I” which would allow them to keep the “IUS” acronym! [laughs]

Q: And you were stuck with it.

A: To make a long story short, in July of 1986 we had no way to get Galileo to Jupiter. We had figured out a way to do it, wherein we would build an upper stage, a third stage to the two stage IUS, and do a so-called ‘Delta-Vega’ trajectory, where you go for two years around the Sun, with a deep-space maneuver to change the trajectory and return to Earth, so as to use the gravity assist of the Earth to go out to Jupiter. And NASA approved that, and then not two weeks after they approved it the people at the Johnson Space Flight Center who are responsible for the Shuttle called us up and said ‘Well, you can’t do what we said, because the weight is too great to land.’

Q: To land the Shuttle in case of an aborted launch.

A: Exactly. So they pulled the rug right out from under us.

Q: What a mess, what a total mess.

A: It was a mess.

Q: Meanwhile hadn’t the spacecraft migrated across the country once or twice?

A: Not yet. Not yet. The antenna had made one trip – excuse me, two trips – and the antenna was obviously upset more than was intended. Now, ironically the antenna was built by Harris in Florida. It was trucked across the country from Harris to JPL for integration and testing with the spacecraft, then taken off the spacecraft, put back in the box, and shipped separately back down. By truck. So one round trip, two trips across the country.

Q: And it wasn’t a refrigerator truck, or anything… It was just sitting on the back of a truck, in the south-western deserts, for example.

A: No, that wasn’t the problem. Oh, but your statement is incorrect. It was in an environmentally-controlled van. The problem was vibration. Not the atmosphere.

So under these circumstances, the only thing we could do, the only way we could figure out how to get Galileo to Jupiter, was to split the mission into two pieces again. We had to have two launches, because we only had the small upper stage. The only way to get that mass to Jupiter was to split it into two pieces, two smaller pieces. And we had a meeting at NASA headquarters in the first Monday in August of ‘86 to present this plan. And by an absolute miracle, one of the principle designers for the satellite tour around Jupiter had been working at trying to figure out an elegant way to solve the problem of how to go about getting to Jupiter in the first place. And he was the first person who ever realized that you could go around the Sun and come past the Earth twice: a dual gravity assist. That fellow was Roger Diehl. And he was very recognized for this marvelous realization.

It was literally a brain-storm. He was literally laying awake in bed when this came to him, and he rushed in to work the next morning and tried it out on the computer, and it worked. And it was fabulous, because it was such an elegant solution. Because it required only the two-stage IUS, it didn’t require any more contraptions, or another stage, or anything like that. And it had more performance margins both at the launch and at Jupiter than anything we’d ever looked at.

Q: Performance margins in the sense of what?

A: The amount of propellant. That you had to have, and that you had to launch with.

Q: So you had more propellant, when you got to Jupiter?

A: Yeah, right.

Q: Interesting. So let me try something out on you. Could you say that one reason why Galileo is still operative now, after almost six years in orbit around Jupiter, is because of such an elegant trajectory design? That you would have run out of propellant earlier if you had done it any other way?

A: Oh, certainly.

Q: Very interesting. So there was a real trade-off there.

A: Well, but that wasn’t the reason we did it. It was serendipitous, I mean it was the only thing we could do and it turns out that that gave us more capability than we had imagined. The price, of course, was that the trip time went from three years, or five, to six. So the earlier trajectory design, the Delta-Vega, was five years, because it was two years around the sun and then three to make the Earth-to-Jupiter transit. So this was just a miraculous discovery.

Q: Was there a term for this type of trajectory?

A: Yes, Veega. And it’s well known now. It’s Veega – Venus-Earth-Earth Gravity Assist. And the Cassini project, the mission they’re flying to get to Saturn, is a derivative of Veega, that would logically not have resulted had we not had Diehl. So this invented the technique for how to get to the planets with massive spacecraft but relatively small launch vehicles compared with what we had been anticipating.

And it turned out to be more than two years before the first return to flight of the shuttle. We set November of 1989 as the launch period for Galileo, we had about a three week launch period. And – oh, the thing that almost killed it at first was the idea that you couldn’t go to Venus. Because the spacecraft was not designed to go closer to the sun than the Earth is. It was designed to go out, immediately. And in fact in past years these trajectory geniuses would go first to the spacecraft people and ask ‘Could we possibly fly to Venus?’ and they would say ‘Absolutely not. End of story.’ But on this particular occasion things were so desperate that, instead of going this way [illustrated with a horizontal slash of his finger], they went this way [a vertical upwards motion of the finger]. The guy in charge of this fellow Diehl called me – this guy was working for me – and said ‘We’ve got this great solution. There’s only one thing, it has to go to Venus.’ And I said ‘Absolutely, we’re interested in it.’ ‘Even though it goes to Venus?’ ‘Absolutely.’ And I called my boss, project manager John Casani, and said ‘We’ve got the answer! But we have to go to Venus. Can we go to Venus?’ And he said ‘Absolutely! We’ll make this thing work.’ And that very day, that morning – well, the first thing he told me was ‘send Diehl over here, I’m going to kiss him on both cheeks.’ I mean, this thing, it was a miracle, it was the best thing that had ever happened in John’s professional life. He thought he had a billion dollar bucket of bolts on his hands. He couldn’t fly it anyplace! And here was an elegant solution. In his term, it was an ‘answer to a maiden’s prayer.’ [laughs]

So he called in the thermal designers immediately and said ‘figure out how to make this thing go to Venus. You’ve got the day to do it.’ I remember sitting in John’s office late that afternoon, when this guy comes back with his report with a long face, and says ‘All the thermal experts looked at that. It’s impossible, we can’t do it.’ And John looked back at him and said ‘you can do it – go off and figure it out.’ End of story. And of course John got busy and sketched a way to do it. Which didn’t turn out to be the way we did it. But that’s John, and that’s how most of us are – tell us ‘no’ and we’ll figure it out for ourselves.

So there was the miracle solution to Galileo’s long-standing… forever-standing problem, how you get this huge space-craft to Jupiter. Which had to be huge, in order to function, we had eleven instruments on that space craft, and seven more on the probe, and an enormous job to do, and excuse me, but it was just as powerful as all get-out. That spacecraft is unequalled in its performance and in its return on investment. In spite of the fact that it was a billion dollar project.

Q: And despite that antenna problem later, it was still unequalled?

A: Yeah. And that even adds to the success and the victory of it, that we could overcome all that, and make it all work. JPL was the solution. It was the JPL infrastructure, and JPL research in communications theory, and the JPL Deep Space Network, which we built and operated and evolved with NASA, I mean all of that had to play to make this work. And it played fabulously.

Q: Well, earlier you said that Galileo was the most elegant interplanetary probe ever designed. More so than the Voyagers, huh?

A: Yeah. Much more. Right off the bat I can tell you that one of the main reasons is that it’s the only ‘dual-spin’ planetary spacecraft. You have the best of both worlds. Spacecraft are typically either spun, or three-axis stabilized. And the fields-and-particles [experiment] people want to be spinning, so they can survey the entire environment around the spacecraft – the whole sky, like airport radar. And the camera people, or remote sensors, they want to be stabilized, so they can look very steadily at some target. By doing dual spin – and this was an elegant new design by JPL for NASA – you can have one section spinning, the other section de-spun, acting like a three-axis stabilized probe. And this system has been beautiful in its execution. It was one of the toughest challenges, we had all kinds of development problems with the thing. Because imagine the problem of getting power and signal across that spinning interface. And a propulsion system. We actually ran plumbing through all that! So it was just fabulous.

Q: Ok, so let’s jump forward. You’ve gotten it out of the Shuttle, you did a burn, it’s heading towards Venus. You still don’t know that you have a problem, right? When did you decide to try to open that high-gain antenna?

A: We had to put shades on the spacecraft, to protect it from the heat of the sun at the distance of Venus. So we had one massive buff shade, as we called it, which shaded virtually everything, and then a shade at the tip of the folded antenna, to shade the antenna. And we were absolutely certain, I can swear to you that there was never any thermal problem. The spacecraft always maintained attitude properly, and we never had a temperature problem. But by design – by mission design and spacecraft design – the spacecraft was not to open its antenna until after the first Earth fly-by. So we could open it once we had come by the Earth for the first time, for the first gravity assist. So by that design, necessarily, we flew the mission from launch past Venus past Earth-1, in great euphoria, because we had no idea we had an antenna problem. I’ll tell you, in fact, that – I can’t really explain this, but Earth-1 was the most euphoric professional event in my life.

Q: Passing Earth for the first time with Galileo.

A: Yeah. Why wouldn’t that have been arriving at Jupiter, when everything worked perfectly?

Q: Yeah, I wonder – why?

A: I don’t know. Maybe it was age, or lack – I don’t know. [long pause] I think that it was because everything seemed to be perfect. We had been through all this, and in particular we had been to Venus, we had demonstrated that we had overcome that challenge, of taking a spacecraft that was never intended to go there, taking it there successfully and coming back – and the answer is, probably, that it looked like all the challenges were handled. That we had survived. It was just that everything was working. And in November I had been manager of the project for less than a year, so there were just a lot of ingredients to this. [thoughtful pause] It’s just that way, I can’t explain it.

Q: Ok, so you’ve flown past the Earth, you’re euphoric –

A: We took this marvelous movie of the Earth for a day, this beautiful Earth-rotation movie never before filmed from space…

So then comes the problem when the antenna didn’t open. We were devastated by that, but early on we thought we would be able to get it to open. And as we struggled with that for months, happily soon enough we figured we needed to have our other experts at the laboratory, in communications, to figure out a way to do the mission on the low gain antenna. And we had that solution in place before we had to declare a failure of the antenna. And that’s what was so powerful about Galileo, that at no time were we out of business. Before we lost hope on the antenna we had a very good solution as an alternative.

Q: So, just the abridged version of the solution for Galileo’s serious antenna problem was, you used the low-power antenna to send data home at a slow rate, and meanwhile you used the data tape recorder to download the spacecraft’s science findings temporarily. Which was originally in the spacecraft for what purpose? Just as a safety measure?

A: That tape recorder was there for one primary purpose, and that was to back up the Jupiter atmospheric probe data. Because recall the probe was also a ‘kamikaze’ mission. One shot – no second chance. So as we released the probe five months before Jupiter – the probe flew silent, and unguided, for five months – but then both vehicles arrived at Jupiter, the probe plunges into the atmosphere, the orbiter’s overhead to receive the data the probe is sending as it’s parachuting through the atmosphere…

Q: …and if at that moment you had had a problem…

A:…any problem, even if the station that was tracking the orbiter on the ground had had a problem, if you lost that real-time link you would have lost the probe mission. So it was essential to capture that probe data on-board. That was the reason there was a tape recorder.

Q: Ok, so your solution was to load the data onto the tape recorder and then feed it back to Earth over the low-gain antenna, bit by bit, from Jupiter.

A: No, you jumped a step, inappropriately.

Q: Sorry.

A: And, again, this came from our communications research area at JPL. And the idea was, yes to record the tape recorder full of images and other high-rate data at each encounter, and then to read that data out, in small bites – not with a “y”, small pieces! – into the central computer, and then with new software, based upon this communications research technology, compress the data by factors of ten to twenty to even fifty or a hundred to one, and then send the bits representing the compression to the ground. And the capability over the low gain antenna, raw, at the time was slightly less than ten bits per second… The low gain antenna was very broad beam, necessarily, because you want to be able to communicate in most any spacecraft attitude.

Q: And the primary reason to have a low-gain antenna is to communicate with the spacecraft if the tightly focussed high-gain loses its lock on Earth?

A: Yeah. And there was a huge effort, a 30 million dollar effort, using that technology and developing new software. We developed new software for the entire spacecraft. And completely re-loaded the central computers on Galileo, in flight, several times. We did it in stages. And then equally on the ground, the Deep Space Network spent about 30 to 40 million dollars improving their receivers, arraying antennas together, electronically ganging them... We actually got the bit-rate up to 160 bits per second, from what would have otherwise been 8 bits per second. So a factor of 20, on the ground.

Now, we couldn’t change the output power from the spacecraft. We had only that, and that was dispersed over this broad beam, so the power density in the beam was ten thousand times weaker than it was supposed to be. So the information that was in the broadcast beam when it was hitting these antenna farms, and the density of power there was ten thousand times lower than originally intended. That means you have to make the bit rate much smaller, because you need more time to make sure you have a one or a zero. So we got a factor of ten, basically, in the ground receiving system, and we got at least a factor of ten for the highest resolution data on the spacecraft, with the compression. So ten times ten is a hundred: we had a factor of a hundred. At a basic rate of ten bits per second, times a hundred, it’s a thousand. So we wound up with a kilobyte. An effective communications rate of a kilobyte, instead of a 134 kilobytes! But by carefully selecting the images and other data we took, and then very judiciously selecting the compression on the images, we were able to get what you know Galileo has produced.

Q: By the way, a few years ago, when I was talking to a group of planetary scientists, one point they made was that one reason that so much good hard science has been done from Galileo is actually because there wasn’t an overwhelming flood of data. So they had to really focus on what they had. Isn’t that interesting? There’s another unexpected positive side-effect of all these problems.

A: [Smiling] Yeah, because it’s a matter of being very frugal in planning what you’re doing, so that you plan for getting the very best information. And then you can fully exploit that very best, instead of shot-gunning it and then trying to wade through it all. It’s a much more focussed approach than might otherwise be used. That’s a very interesting point. I hadn’t heard that stated so precisely before.

[Interview concludes with Part 3]

back