[70] Planetary exploration is an ancient and sanctified pursuit, underway, if you count exploration of Earth, for at least a million years. In spite of all the years, we have hardly begun the task of exploring Earth. There are great areas of land that have only been sampled, and we have only the skimpiest knowledge of the 69 percent of the surface that is under water. It was only after we left Earth for orbit that we began to understand our planet better, recognizing it as one of the most varied and fortunately endowed objects in the solar system. During the last two decades we have learned much by traveling around it, walking around the home block for the first time, so to speak. Views from the vantage of near-Earth space have provided exciting new perspectives of the planet, clearly revealing the land masses and their encircling oceans, the continent-sized storm formations, the restless clouds, and the dynamic environment in which our lives are spent. Being alive, we care about life, and for this our magical and unique atmosphere is the key. Using sounding rockets and satellites we have plumbed and defined this thin planetary coating that shields us from high-speed particles, filters out lethal radiation, ameliorates insupportable temperature variations, and carries to us the water without which life would soon vanish.
In these same decades we have also gained insight into our planetary neighbors. Using increasing sophistication in our automated spacecraft, we have visited the Moon, Venus, Mercury, Mars, and those strange gas giants, Jupiter and Saturn. Close-up photography has made details of the Moon's face familiar, its scarred surface recording both great eruptions from within and violent bombardment from space. Cataclysmic history as preserved in scrambled detail on that pockmarked surface is being used by scientists to fill gaps in Earth's remote past, for similar scars on our own planet have been all but obliterated. Planetologists see the Moon as a Rosetta stone for our solar system, helping to unlock the cyphers of an enigmatic past.
[71] Men first began to explore Earth by looking at their immediate surroundings, gradually widening their travels to encompass larger areas. With our new exploring tools, we have a better way. Obviously, an unknown planet should not first be examined locally, with a restricted view, but should be seen in its entirety, from flyby and orbit. We should begin with reconnaissance of gross features: clouds, continents, polar caps, mountains. We should examine the atmosphere, if any, as knowledge of the atmosphere not only offers many answers about the nature of the planet and its history, but also prepares for the engineering of successful landings
The next phase should involve landings at preselected sites, chosen as the likeliest to provide a maximum of information. These are often at the boundaries between different types of terrain: at the edges of polar caps, the bottoms of large hills or valleys, and the rims of high plateaus affording distant observations. Finally-before the risky and costly landing of men-we might use automated roving or flying vehicles, remotely guided from Earth, affording the intelligence-gathering skills of our best substitutes for human eyes and human senses without risk to life. Then, if conditions warrant, we should send a combination of machines and the best multisensory decisionmaking resources we have-human beings-when the extra costs of life support and confident retrievability can be justified.
After a few missions are flown, some critical components are more or less taken for granted and put from our minds because they can be counted on. Persons coming into established projects or those observing projects underway may never realize how everything came together the first time or have a good perspective for judging the totality of the logic and the processes involved. No major technical undertaking is ever done from scratch-we have mentioned the tremendous contribution to space missions of missile programs-but developing appropriate plans and putting the necessary systems in place to support the first Mariner flights took a great deal of ingenuity and effort.
As mentioned repeatedly, the biggest initial hurdle to exploring other planets was our marginal ability to escape from Earth. The launch vehicle has always been a limiting factor, restricting the mass of spacecraft to a degree that has challenged designers to provide useful payloads. This constraint figured in making the Moon our first target; however, it would have been foolish to bypass the Moon, the nearest neighbor of Earth, to go first to far planets Yet there were compelling scientific reasons to obtain information [72] about several planets as quickly as possible; by studying them we obtain perspective about their similarities and dissimilarities to Earth, providing advances in knowledge of the solar system through the use of broadly based comparisons.
Overall plans for programs were developed in the Lunar and Planetary Programs Office and reviewed in a number of forums. Reviews were provided by the scientific community, often through the Space Science Board, set up for that purpose by the National Academy of Sciences. During the budgetary approval process, mission proposals passed through the NASA administration and then to the committees and bodies of Congress, providing repeated opportunities for us to examine and defend our rationale for exploring the Moon and the planets.
A key step was the selection of payloads, requiring not only scientific review of the entire mission, but also the allocation of priorities among individual experiments. I believe that the process NASA developed for payload selection has withstood the test of time. It required a subtle yet complex interaction among people, machines, budgets, and politics (after all, these were publicly funded programs). Also intrinsic was a tolerance for considering everything of relevance, from the engineering of a tiny subsystem to testing a theory about cosmic origin.
From a scientific viewpoint, the most difficult part of planning missions was choosing experiments that addressed the most fundamentally important questions. After lengthy discussions, order finally was provided for the process through the coalescing of views on major classes of questions for the Moon and the planets. Once these had been defined and accepted, it was possible to develop balanced experiment packages and to consider individual experiments in proper relationship to others. The simplifying approach also helped address experiment sequencing or priority questions; in some cases time-critical interactions with other experiments affected priorities for instrument selection.
Four classes of scientific experiments were initially defined to address major planetary questions; these are now logically explained by thinking about examining a planet "from the outside in."
The first class of experiments addresses a planet's environment as determined by external influences such as the Sun, especially radiation, particle fluxes, or varying energy fields and their effects. Planets in our solar system exist in an environment largely Sun dominated, although the environs of a planet can be modified by the presence of its own unique magnetic field.
[73] Field strength and orientation also provide insight into many characteristics related to the origin and nature of the body. The name given the first class of experiments was simple and descriptive: particles and fields.
The second class of experiments deals with planetary atmospheres: their compositions, densities, pressures, temperatures, clouds, or other peculiar features. Instruments for this class of experiments come in a variety of forms, ranging from direct sampling to remote sensing.
A third class of experiments known as body properties is broadly directed at the celestial body itself: external shape, mass, density, precise orbit, and rotation rate and direction. There is interest, too, in mass distribution and tectonic condition (whether the planet is volcanically active or quiescent), because these conditions can be related to planetary history and help in hypothesizing the conditions of evolution. Experiments in the third and fourth classes generally were labeled "planetology" experiments, a term coined to convey their relationship to the field of geology, but having a broader connotation.
The fourth class of experiments deals with the planetary surface: its texture, features, and composition. This includes topography-mountains, hills, valleys, craters, and other forms of nonuniform disturbance-and chemical composition and materials properties. It is of vital interest to determine whether the materials composing the planet are minerals like those on Earth or unique constituents. Physical measurements of surfaces-hard or soft, sandy or dusty, lava-like or deposited in other ways-are needed, as are measurements of other properties such as conductivity, temperature range, and magnetic susceptibility. Some of these characteristics call for in situ analysis, and a few require sample return to laboratories on Earth. Initially this class covered the broad questions related to the search for life; it was not until biology experiments like those carried years later by Viking that a special biology class was added to the four basic classes.
When all proposals for experiments were fitted into this framework, choosing balanced instrument payloads became easier. Looking back with the assurance provided by time and experience, I wonder why the simple process of defining major classes of experiments was so momentous; it does not seem very profound today. Perhaps it is because at the time no one had the interdisciplinary knowledge necessary to define the broad options of first-time missions. I remember listening to the individual scientists lobbying for their particular experiments, fearing that we might mistakenly succumb to "squeaky wheel" pressures and overlook a prime question that no one had [74] fully considered. Once the four classes had been tested and accepted by most of the experiment proposers as encompassing, however, our confidence grew, and the concept of using them to develop balanced payloads was not challenged.
While working in industry on the design, development, and production of flight hardware, I had become accustomed to a "projectized" organization with strong, almost military, hierarchical leadership. I naturally supposed that it would be necessary to organize the total NASA effort, including the scientific component, along those lines if we were to successfully conduct projects with such demanding hardware requirements, engineering interactions, and strict deadlines. To my dismay, wiser heads than mine decided that the scientists who participated in space missions should be allowed to remain in their laboratories and classrooms, retaining as much freedom and independence as possible. I had grave doubts that we could make successful teams for the difficult missions unless the generally undisciplined scientists could be brought together with the other project members under rigid controls. This view was proven wrong, for the system worked and schedules were met most of the time. Now I realize that if my management concept had been forced on academic investigators, it would have severely limited the long-term dividends of space science, because it would have compromised the "fresh-eyes" benefit of their participation.
The NASA Space Sciences Steering Committee was a particularly important element in the total process. This committee was developed by Homer E. Newell, who had been a successful administrator of scientific activities with the Naval Research Laboratory before joining NASA. The committee as prescribed embodied an exceptionally balanced blend of managerial engineering, and scientific viewpoints. It was chaired by Newell, who, as Associate Administrator, had the highest line responsibility within NASA for space science and applications programs. His alternate was the Chief Scientist, who, by organizational assignment, was necessarily concerned with the relationship between NASA and the scientific community. Other members of the steering committee were space science program office directors and their deputies-responsible for physics and astronomy programs, lunar and planetary programs, and bioscience programs. This body of eight had review responsibility for all scientific payload recommendations, with Newell as final selecting authority.
Within the steering committee a system of subcommittees was established, oriented along scientific discipline lines and chaired by NASA personnel [75], with members chosen for their specific scientific expertise. In the early 1960s seven of these subcommittees were appointed by Newell, covering the areas of particles and fields, solar physics, planetology, planetary atmospheres, biosciences, physics, and astronomy. Though the pyramid of committee and subcommittees may appear rather complicated and unwieldy, the various bodies interacted well and functioned smoothly, thanks primarily to the talented and dedicated professionals who served on them.
Whenever a major mission or series of missions was being planned, an announcement of flight opportunities (called an AFO) was made to the scientific community. This announcement outlined the nature of the mission, the types of experiments of general interest, and gave, to the extent possible, broad guidelines for proposed experiments. Proposals for specific experiments came from all quarters and were categorized and submitted to a subcommittee for review . Sometimes as many as 60 or 70 experiments were proposed for a mission that could accept only 5 or 6. Subcommittee responsibilities involved review to determine the scientific importance of each experiment, an assessment of its readiness to be integrated into the spacecraft, and an assessment of the competence of the investigative team. After subcommittee consideration, proposed experiments were placed into one of four categories and presented to the steering committee:
This sorting process afforded a thorough review, yet left final selection to the management team responsible for making the mission as worthwhile as possible. It afforded ample opportunity for inputs from all sources and, in general, withstood the fairness test quite well. There were a few cases in which our selection process was criticized, but it was broadly accepted by the scientific and engineering communities as a reasonable approach.
Although a full-fledged member of the Space Sciences Steering Committee, I was in the minority as an engineer, along with Jesse Mitchell, Director of Physics and Astronomy. It was always a serious matter to select a specific [76] set of instruments on a scientific basis and to determine how the meager amounts of weight, power, and physical space could best be allotted. This called for close collaboration between scientists and engineers in a manner that tried to account for all the variables. Though outnumbered six to two by scientists, I always felt that final decisions for payload selection gave ample consideration to the engineering judgments Jesse and I frequently brought into the overall process.
One thing that did puzzle me during the early payload discussions with scientists was an initial reluctance to fly camera equipment. Somehow pictures were not thought of as scientific, not as informative, for example, as a telemetered record of a varying voltage. A camera was not regarded as a scientific instrument at first, and to go after "picture postcards" was seen by some as an unscientific stunt.
This was a curious bias. Some of it may have been due to the fact that many of those hardworking scientists who were prepared for space science experiments had been accustomed to using only numerical data; thus, the fact that images represented a means of packaging data was not immediately obvious. In addition, some resistance might have been rooted in a wariness toward NASA and the motives of its administrators. Since NASA was a government agency dependent on popular and legislative support, some scientists may have suspected that we wanted photographs primarily as Barnumesque publicity attractions.
Charles P. Sonett, my deputy at the time, recalls one space science subcommittee meeting in 1962 involving more than a score of scientists gathered to consider the instruments to be flown on a future lunar mission, perhaps Surveyor. Most of the group, which included Nobel Prize winners, voted against flying TV cameras. Cameras had been supported only by Sonett (he was chairman of the meeting), and by Gerard Kuiper, a celebrated astronomer concerned with obtaining detailed images similar to those he was accustomed to seeing with telescopes. Sonett recalls that he reported this strongly biased view against flying cameras to me and I replied, "Fine, so let's fly them." I don't remember our dialogue in detail, but we did fly cameras on Rangers (and Surveyors and Lunar Orbiters as well), with broad agreement among scientists later that the cameras did much to enrich those spectacular missions.
In time the anti-image prejudice dimmed. It did not disappear overnight, but the startling effect of Mariner 4's shadowy images of craters on Mars and the torrent of images returned by the Surveyors did much to quiet the skeptics.
[77] I knew the issue was passé during an early Surveyor mission when I came upon Gene Shoemaker, a geologist of great foresight and conviction, surrounded by inquisitive colleagues from other scientific disciplines in a back room at JPL. They were prying from him interpretations of a newly acquired sea of prints, almost like zealots seeking meaning from a disciple reading the scriptures.
Nature imposed her own unappealable constraints on lunar and planetary programs through the geometry of the orbits of the Moon and the planets and their relationships to Earth and the Sun. Not the least annoying of the variables Mother Nature controlled was the weather, often a troubling factor at the time of launch and once even upon arrival at Mars. Mariner 9 had to orbit around Mars for quite a while waiting for a planetwide dust storm to subside. Weather here on Earth was more often a problem; for example, dense rainfall at Earth stations sometimes impaired the quality of returned data.
This snarl of planning variables was particularly challenging for Mars and Venus, because usable launch opportunities occurred only once about every 2 years. The fixed launch period scheduling problem caused the greatest consternation to scientists preparing their experiments. The entire schedule-planning, budgeting, development, and testing-had to be worked out backwards and events had to be time-phased so that they were completed when the launch period arrived. Launch opportunities typically lasted only about 1 month; when two launches were planned, they had to be made in rapid succession, often no more than 3 to 4 weeks apart. This in turn necessitated either dual launch pads or an extremely rapid turnaround and closely integrated use of launch facilities.
For NASA Headquarters managers a recurrent headache was the need to synchronize mission planning with congressional budget cycles: a desirable condition that seemed rarely to happen. Neither the actions of Congress nor the movements of the planets could be made to accommodate the other, and it was our task to do all the adapting that was needed.
As a practical matter, it was not reasonable to expect most scientific mission objectives to be accomplished with a single flight. Because of the unreliability of launch vehicles and the unrevealed problems of new spacecraft, it was difficult to know how many missions of a like kind should be planned, since there was no way of knowing which would succeed and which might fail. Scientists risked severe frustration by trying to perform experiments on flights for which the launching rockets and spacecraft were [78] themselves largely experimental. Unfortunately, multipurpose flights, for example, developmental and mission-oriented flights, were a way of life. Even with considerable flight experience, the reliability of large multistage vehicles was rarely greater than 70 percent and often far less.
Many debates were devoted to possible ways to plan programs that might achieve flight objectives. Some thought it desirable to plan a series of test flights solely to develop the vehicle and the spacecraft, and then add the scientific payload. Others argued that even test flights should reckon with the possibility of total success. Another often-asked question was, should a series of missions have identical payloads so that those on failed missions could be duplicated as soon as possible, or should they be varied? The choices could work hardship to the edge of cruelty on individual scientists, for those whose experiments were not chosen to fly might have to wait for several launches, perhaps 4 to 6 years, before their instruments could be incorporated. In some cases experiments that could have been significant never got a chance because of vehicle or spacecraft failures. These were worst cases, though. Looking back, I am amazed that final results seem logically sequenced, as if we had had better knowledge of what to expect than we actually did.
If all went well and the spacecraft arrived at its target planet, the hour of the scientist was at hand. Data would come to Pasadena for sorting and for preliminary calibration and computing. The data alone were not enough; one also needed to know the spacecraft location and attitude and the times at which the data were acquired-all essential to the scientists but not directly within their control. It was then incumbent on scientific investigators to study, analyze, and interpret their results. Typically, the results would be published in the leading journals of the disciplines concerned; frequently NASA would prepare a special publication describing the mission and its results; sometimes a symposium would be held in which individual scientists would compare data and defend their interpretations. After the long, tense years that had gone into planning and executing a mission, I was always delighted to observe the cooperative and coordinated way in which highly individualistic scientists contributed to the common store of human knowledge.
It wasn't all roses, of course. The amount of time and work, the competitive environment, and the chanciness of investing peak career years in an unpredictable venture meant that some were inevitably disappointed. A few [79] complained that the work of preparing proposals and then instruments and the difficulties in meeting strict schedules and complicated integration procedures were so overwhelming that their time was better devoted to laboratory work or other activities over which they had more control. And so it may have been-for them. But for those who braved the difficulties, waited for the opportunities, and sweated out risks of nonselection and mission problems to ultimately derive important new knowledge, it was a thoroughly worthwhile endeavor. Many scientists who worked hard on these investigations, and some who burned themselves out trying, feel that they received the greatest rewards of a professional lifetime for their efforts. Many, and perhaps most, felt that it was worth whatever it cost.
Quite apart from the planning and scientific processes of payload selection and development, there were many engineering and support functions to plan and prepare. Even with our missile background, there were few system test facilities that could simulate the space environment. Some vibration equipment useful for simulating aspects of the launch environment existed, but it was meager and limited in capability. Considering the high cost of a mission and the infrequency of planetary launch opportunities, it was important to check everything scrupulously while the spacecraft was still on the ground. In the moment of truth at liftoff, everyone concerned with a mission was prey to "for-want-of-a-nail" anxiety.
In the beginning, facilities were perilously jury-rigged or patched up. In 1960 spacecraft were assembled and tested at JPL in a building left over from previous missile work for the Army. Also used was a small building next door housing a makeshift shake table and small vacuum chamber. Within 2 years contracts were let and construction was begun on better facilities, including a realistic thermal vacuum testing simulator, but the new gear was not ready for use until five Rangers and two Mariners had been launched. Doubling up in the use of these limited facilities and extrapolating conditions well beyond known capabilities was a way of life until better facilities were completed.
Scheduling the use of available missile launch pads and blockhouses was continuously bothersome. Usually long lead times were required for preparation at the site, and uncertainties regularly arose about delivery and checkout of essential equipment. On top of these problems, uncertain weather conditions and unforeseen difficulties were likely to arise during intervening launches. As a practical matter, long-range scheduling always had [80] to be iterated in the shorter term because of difficulties and program delays. Following an exact timetable was rare, and it was no less worrisome when things seemed to be going well.
Factors of a different sort that had a gross effect on planning were the availability of manpower and the status of funding. Manpower and dollars were not only essential in and of themselves, but the rate at which they could be applied influenced the development schedule and the scope of effort. Juggling these factors was a major management challenge; to decide the best ways to spend either resource involved tradeoffs that ultimately affected mission results.
In the early days of 1960-1962, when Ranger and Mariner were under development, many engineering practices were outlined that later became almost standard. For example, three nonflying Rangers were built to validate the final model: a spacecraft mockup, a thermal control model, and a proof test model. The first was used to confirm mechanical aspects of the spacecraft, including fits and clearances, cabling harnesses, and layout of equipment. (Even gifted designers and engineers can benefit from insurance against momentary spells of inexplicable oversight.) The thermal control model was suspended in a small, early-model vacuum chamber and subjected to the vacuum and simulated solar heating it would encounter in space. The uncertainties were large in those days, and for Mariner 2, engineers underestimated the heat encountered on the path to Venus. This resulted in the spacecraft running a high fever at encounter and dying of it a few days later. The proof test model, known as the PTM, was as similar to the flight spacecraft as possible. It was subjected to vibration and other tests somewhat above the actual predicted levels in order to provide suitable margins against the unforeseen.
The process of building additional vehicles for test purposes evolved through many variations as time went on. The practice had the additional advantage of providing a spare spacecraft in case of trouble with the prime article, and it gave us a duplicate to study on Earth if telemetry reported puzzling misbehavior millions of miles away.
Considerable debate was devoted to questions of the best testing doctrine. One view was that testing should not be performed on hardware actually to be used in flight because the stresses of testing might wear things out and would obviously affect equipment life. Another viewpoint was that flight hardware should be designed with sufficient margins to withstand both [81] the rigors of testing and the mission itself, for this would eliminate weak links that could cause early failure.
An issue that to this day remains a matter of choice was highlighted by strong opposing positions taken by JPL and Goddard engineers when they were competing for planetary projects. The crux of the debate centered around the fact that spacecraft were going to operate "hands off" in space. Goddard test engineers believed the best way to wring out a spacecraft in the laboratory environment was to operate it through radio links, with no instrumentation connections, power, or other external connections that might cause or prevent a failure. JPL engineers, on the other hand, believed that they should try to exercise components individually and evaluate nonstandard conditions that might occur. This meant that there had to be synthesized inputs and special instrumentation connections to allow proper evaluation of subsystems or components.
One incident that occurred during the thick of competition for a Mars project involved Bill Stroud, one of the most outspoken Goddard engineers. He came to Headquarters one day predicting "doom and gloom" for JPL because of their approach, and offered to prove his point and help teach them how to do tests properly. For emphasis, he had with him a pair of goldplated diagonal wire-cutting pliers, called "dykes" by technicians. According to Bill, all we had to do was take the dykes to JPL when they were running a systems test, cut the many wires and cables they were using to support their simulations, and then we would find out that their spacecraft would not work. The inference that JPL cheated on their systems tests, plus the "know-it-all" impression his act created, went over like a lead balloon with JPL engineers when they heard what Bill had done. I recall being more amused than concerned, for both centers had proven their competence, and I was sure that either approach could be made to work. Of course, this episode did nothing to encourage commonality in testing techniques among centers, and I still do not know which philosophy is best, if there is a best.
As is often the case with such conflicts in judgment, tradeoffs were made and compromises struck. In time, though, the balance moved toward testing everything that flew, subjecting it to as nearly complete a lifetime simulation as possible This gave us confidence that the equipment was flight-ready, and I believe the principle was borne out by the successes that followed.
A particular bedevilment of those times arose from the requirement to sterilize everything that might land on the Moon or planets. No one objected [82] to the idea of avoiding seeding them with terrestrial microorganisms; the problem arose from the fact that the specified protocols were complex, rigid, and at war with the quest for reliability. A sure and first specified approach was heat sterilization, but many components would not survive being baked in a hot oven for long periods. A less obvious complication of sterilization developed from the fact that it had to begin at the subassembly stage and be maintained through final assembly and testing, thereby requiring that all testing equipment and facilities be of clean-room quality, including handling gear. In an environment where even the hoists are sterilized, and kept so, work does not advance with the speed and precision one might want.
I always took the position that a spacecraft should be built at the factory, checked out there to the fullest extent possible, and shipped to the Cape ready for launch. Most engineers tended to agree with this philosophy; however, special circumstances always seemed to dictate the need for a complete systems checkout at the Cape, requiring equipment for a thorough test of all finally assembled and adjusted hardware just before mating to the launch vehicle. The field checkout facilities were identical in many ways to the system test facilities at JPL; in some respects they were more complex because of the need to include launch vehicle and tracking elements. The apparent duplication between these facilities was not in fact real; they often performed complementary tasks on the same entities for different purposes. Checkout could indicate a need for replacement, and with the window inexorably approaching, return to the factory might be unthinkable.
The first spacecraft assembly and checkout building used for Rangers and Mariners at the Cape was Hanger AK. It had been built in the 1950s for Navaho missile preparations and had a low-bay area designed for a flying vehicle resting on a tricycle gear. It was a non-air-conditioned, metal building with small shop areas alongside the hanger portion and was unsuited to the peculiar needs of spacecraft. Returning to the same facility that I had been associated with almost 10 years before was a ghostly experience. Engineers from JPL and the Cape quickly defined modifications, and I obtained approval for a high-bay area addition with a 30-foot hook height that would allow the spacecraft to be assembled and enshrouded vertically. Also included was an air-conditioning system with filters that provided the cleanroom conditions needed for sterilization control.
When finished, it was one of the first clean facilities to be installed at the Cape, said to be cleaner by particulate count than most hospitals in the area. This appraisal led to its being named by Kurt Debus, Director of the [83] Kennedy Space Center, "Hank Levy's hospital." Hank Levy was then, and still is, JPL's principal resident in charge at the Cape, a man whose fingerprints have probably been launched on more lunar and planetary spacecraft than anyone else's
A significant amount of additional upgrading and modification went on, including changes to the spin-test facility, remote for safety reasons, to allow its use for fueling midcourse correction motors and to support the final ethylene oxide gas sterilization process. A new, designed-from-scratch systems test building was also begun. While prelaunch checkout facilities steadily became less ramshackle, it was 1964 before we could begin to treat our interplanetary travelers with the care they deserved.
At the same time, two other efforts of vital importance were being conducted. First, there was the establishment of a deep space network composed of radio tracking, telemetry, and command stations at different points around the globe, a control center from which it could be directed, and an Earth communications network to tie it together.
The Deep Space Instrumentation Facility (known universally as the DSIF) was a vital link in the chain. Obviously a good launch was not enough; mission success depended on good data return and analysis. The geographical position of the Earth stations, the communications frequencies to be used, the ground handling rates, and the priorities among spacecraft aloft were very real constraints that had to be factored into mission planning. The fact that two of the three tracking stations were on foreign soil, one of them subject to the vagaries of an unstable government, also led to occasional cases of heartburn.
Second, there was a flight operations facility at JPL, later known as the Space Flight Operations Facility (SFOF), with quarters and equipment for mission operations, including banks of computers for analyzing trajectories, acquiring and analyzing telemetry data, and generating commands to be sent to spacecraft. The SFOF necessarily had a very close relationship with the DSIF, and in fact they shared a common control center. In addition, there were launch operations facilities at Cape Canaveral for preflight testing of the systems for tracking and downrange support, plus all the diagnostic equipment needed to ensure that the launch phase was performed properly.
The man at JPL ultimately responsible for tracking, telemetry, and communications was Eberhardt Rechtin, a near-genius whose telecommunications achievements left his mark on space exploration. Rechtin had been a student of Bill Pickering in electrical engineering, graduating from CalTech [84] with a cum laude doctorate in 1946. During the time the tracking and data acquisition network was taking form, Rechtin was a key figure, forceful and enthusiastic, with a reputation for brilliant solutions to technical problems. He had an uncommon knack for grasping the unanticipated implications of large systems, foreseeing both problems and potentialities ahead of others.
Rechtin was joined in the network's formative period by a number of other good men, among them Walter Victor, Henry Rector, William Samson, and Robertson Stevens; some of them are still making important contributions to the field today. The team envisioned three Earth stations located so that one of the three would always have a spacecraft in view as the globe turned. Thus, information could be received and commands sent without interruption. The stations are about 120° apart in longitude, one in the Mojave desert not far from JPL, one in South Africa (later replaced by a facility in Spain), and one in Australia near Woomera. Each site has large steerable antennas that can be pointed accurately in space and are designed for maximum efficiency for receiving and transmitting signals. At the beginning the preferred frequencies were from about 890 to 960 megahertz; signals from this region in the radio frequency spectrum pass through Earth's ionosphere without much reflection. Each station can transmit commands and receive data, in addition to establishing one- and two-way Doppler links for determining positions and trajectories of remote spacecraft.
The mission command post was the SFOF. First-time visitors found it a dramatic place, a large, essentially windowless building on a hillside, with a well-guarded entrance and a set of big diesel-electric generators down below. In a large, dimly lit room with multiple wall displays, the controllers on duty "worked" the distant spacecraft, while dancing numbers on the displays continually reported changing measurements. In adjoining rooms other engineers were concerned with their specialized areas, such as trajectory computation, data collection and reduction, and spacecraft engineering conditions. To visitors it was a paradoxical place: everything progressed inexorably and yet nothing seemed to happen; distances were unnaturally distorted, with Spain and Australia brought next door and an unimaginably distant spacecraft giving its speed and course with extraordinary precision. Even time was skewed: when the spacecraft reported an event, a visitor was bemused to realize that its "now" had occurred before his "now"; even at the speed of light, signals took several minutes to travel to and from distant space.
[85] Of course the effect on visitors was the least of the concerns of those who designed the SFOF and obtained and programmed the computers that made it work. The man who deserves the most credit for this is Marshall Johnson, who came to JPL in 1957 as a computer engineer. Doing today what he managed to do some 20 years ago would be almost impossible, thanks not just to normal bureaucratic inertia but also to the encrustation of controls, loops, and reviews that sprout like rain-forest undergrowth. The computers had to be procured and installed; their software had to be written, checked, debugged, and specially adapted to the unique characteristics of the particular project; and then in a few months it all had to be done differently for another spacecraft on another mission. Marshall Johnson (and his staff) had the kind of wonderful competence that blossoms under tremendous pressure.
Transport of spacecraft hardware on Earth from JPL or from a West Coast factory to a Florida launch site was not simple; one does not simply nail a $50 million spacecraft in a crate and drop it off at the express office. Protection from contamination and from shock called for a controlled-atmosphere container traveling in a special air-suspension van on a route precharted to avoid low bridges and similar problems. Even then, there were the hazards of an occasional blowout, damage inflicted by irate snipers who didn't like "missiles," and the possibility of collision on the crowded freeways. It was something that schedule-minded managers learned by doing, worrying all the way.
The cost of disrupting human lives for unmanned spaceflights were far from negligible. There were questions of how personnel should be assigned to the assembly, checkout, and launch of almost-ready spacecraft and who should concentrate on developing the next one. The procedures continually evolved, but usually a large team had to spend the last six or more weeks before launch at the Cape. Families were split up or partially moved, with considerable hardship in either case.
Preparations for launch might begin with civilized 8-hour daytime duty periods, but as time shortened, working hours lengthened; there always seemed some critical milestone to be accomplished in the small hours of the night. To visitors, the preparations for launch seemed to go on in an informal but rather tense atmosphere. Foremen or supervisors were invisible among their subordinates-rolling up their sleeves, joining the workers, and doing what needed to be done. Many loved the excitement of the effort and were so caught up in it that they neglected their families; these were people [86] who worked hard by preference and played hard for compensation. I always felt remorse that the toll on personal lives was so severe, but I knew no one who would have traded the experience of a successful space mission for any other.