TRACKING, as the term indicates, means measuring the position of a moving object, natural or man-made. Optical tracking with sighting instruments is as old as astronomy, which is at least as old as written history. A child of modern science, radio tracking using radar, radio direction finders, Dovap, and other electronic schemes, emerged during the first quarter of the twentieth century. Radio interferometry, the technique employed in Vanguard's electronic tracking system, entered the picture in the 1940s. Like most of the electronic tracking techniques then in use, this one required the presence of a signal source, a transmitter, in the object being tracked. Employing two receiving points on the ground and comparing the phases of the signals each of them separately received from the airborne source, radio interferometry had the advantage of yielding highly accurate angles.

It achieved practical form in 1948 when engineers with Consolidated Vultee Aircraft Corporation (Convair) created for the Army the Azusa tracking system, using an interferometer.¹ Simultaneously with this development, NRL scientists were working with underwater sound interferometers. The two groups of experimenters were in close contact. They frequently exchanged ideas, and in the early l950s Milton Rosen and his NRL crew at White Sands built and field tested a tracking system, using the radio interferometer principle, for application to ballistic missile guidance for the Viking rocket.²

When in early 1955 Hagen's people, specifically Rosen and his colleagues, began drawing up their plans for launching an earth satellite, they devoted much thought to the problem of tracking so small an object. Most upper-air research scientists advised them to rely on optical tracking. It was the tried and true method. More to the point, Fred L. Whipple's camera observations of meteorites entering the earth's atmosphere had demonstrated in the late l940s that modern terrestrial optical instruments could "spot" an object weighing only a few kilograms and moving at a substantial distance.³

Rosen had doubts. At his request Richard Tousey of NRL checked Whipple's visibility computations. He confirmed these calculations, but found in them no answer to an important question: Granted optical instruments could see the satellite if they could find it in the first place-but could they find it? In Tousey's opinion their chances of doing so were only one in a million.⁴ There was the further consideration that optical tracking, although highly accurate, has limitations. The best of sighting instruments can pick up a satellite only when the sun is five degrees below the horizon-that is, at dusk and dawn-and even then only under certain weather conditions.⁵ Convinced from the beginning that NRL must look elsewhere for an adequate satellite-acquisition method, Rosen had asked John T. Mengel and his NRL Tracking and Guidance Branch to develop an electronic system for use in conjunction with an optical one.

For guidance purposes, the Azusa system had performed satisfactorily at White Sands and elsewhere. For tracking a satellite, however, it was out of the question since it required an airborne transmitter far too large for a small scientific payload weighing no more than thirty pounds, if that much. Refinements and modifications were indicated, and Mengel and his assistants shortly came up with an arrangement that, although based on the Viking radio interferometry techniques, required instead of a heavy transponder only a thirteen-ounce transmitter⁶ and employed different operating frequencies and antenna configurations. It was in essence a new system.

Mengel's name for the system, Minitrack, derived from the system's utilization within the satellite of "an oscillator of minimum size and weight … to illuminate pairs of antennas at a ground station which measures the angular positions of the satellites using phase-comparison techniques."⁷ As eventually developed, the thirteen-ounce Minitrack oscillator, quartz-crystal controlled and fully transistorized, had a ten-milliwatt output, operated on a fixed frequency of 108 megacycles, and had a predicted lifetime of ten to fourteen days.⁸

In a series of papers and speeches prepared over a two-year period beginning in 1956, Mengel and his colleagues-notably Roger L. Easton, his assistant at NRL, and Paul Herget, director of the Cincinnati Observatory and a consultant to Project Vanguard-assessed the role of tracking in the satellite program and described some of the characteristics of the Minitrack. In March of 1956 Mengel told a group of scientists and engineers:

The final realization of man's efforts to place a satellite in orbit about the earth will immediately pose a new series of problems: how to determine the precise orbit that it is following; and how to measure what is happening within the satellite from the vantage point of a ground station. The immensity of the first of these programs, how to prove that the satellite is in fact orbiting-the acquisition phase-can be realized by … an analogy … Let a jet plane pass overhead at 60.000 feet at the speed of sound, let the pilot eject a golf ball, and now let the plane vanish. The apparent size and speed of this golf ball will closely approximate the size and speed of a satellite 3 feet in diameter, at a height of 300 miles … The acquisition problem is to locate the object under these conditions, and the tracking problem is to measure its angular position and angular rate with sufficient accuracy to alert non-acquiring tracking stations, those trying to follow the satellite by optical means. as to the time and position of expected passage of the object.⁹

In a popular article,¹⁰ Mengel and Herget likened the antennas of Vanguard's Minitrack stations to human ears. "An individual," they pointed out, "locates the source of sound by virtue of the phase differences in the sound waves, arriving at different times at his two ears. Similarly the listening units of the minitrack system are pairs of receiving antennas, set a measured distance apart, which indicate the direction of the signal by phase differences in the radio waves.…" They described the reception pattern of a Minitrack station as a fan-shaped beam making an arc of "100 degrees north and south and 10 degrees east and west. In the long direction of the beam we have three pairs of antennas, spaced respectively 500, 64, and 12 feet apart; in the narrower east-west dimension, two pairs with spacing of 500 and 64 feet. The north-south and east-west [antenna arrays] give us two angles which [combine to indicate] the actual direction of the satellite."

The Minitrack stations, as they came into being, showed minor variations. Generally speaking, however, their major components were the fixed arrays for angle tracking, one fixed antenna array for telemetering reception, a rhombic communications antenna, a ground station electronics trailer, a telemetering trailer, a communications trailer, and associate power sources and maintenance units. These components required about twenty-three acres with a minimum gradient of the land less than one degree in the region of the angle tracking antenna arrays. Those in charge of choosing the sites took care to place each station where the adjacent terrain did not exceed an elevation angle of ten degrees for at least half a mile, twenty degrees for five miles. They also took care to select a site at least two miles from heavy electric power installations and at least five miles from airports or airways.¹¹

The inclusion of Minitrack in the satellite-launching proposal NRL submitted to the Stewart Committee had had direct bearing on the committee's decision in August 1955, to accept the Laboratory's proposal in preference to the Army-Orbiter proposal, which originally contained no provision for electronic tracking.¹² The NRL proposal also mentioned in passing the creation at the Laboratory of a less elaborate version of the Minitrack system. Developed by Roger Easton, this abbreviated system would come to be known officially as the "Mark II," unofficially as the "Jiffy" or "Poor Man's" Minitrack. Using radio-frequency phase-comparison techniques by means of hybrid junctions, the Mark II later became the nucleus of "Project Moonbeam," a program sponsored by Project Vanguard to encourage radio hams and their organizations to build their own stations and participate in tracking satellites.

There were two forms of the Mark II, known, respectively, as the "simple" and the "advanced." The amateurs who joined Project Moonbeam used only the simple form. Costing about $5,000 to erect as against twice that much for the advanced Mark II, this arrangement consisted of two matched antennas in an extended base array, a receiver, and oscillograph. Passage of the artificial earth satellite produced a pattern of reinforcement and cancellation, successively, of the signals received at each antenna. These were recorded as a pattern of peaks and nulls. A difficulty in interpreting these records, even where satisfactory time signals were recorded in an auxiliary channel, was in determining which null corresponded to the time of the passage of the satellite through the principal plane of the antenna array. In general these ambiguities were resolved at the Vanguard satellite computing centers by references to the data received at the prime Minitrack stations operated by professionals.¹³

As soon as Project Vanguard became official in the fall of 1955, the individuals charged with tracking began work on its implementation. In view of the technical hurdles to be cleared and the problems incident to coordinating the work of numerous military units, university laboratories, individual experts, private industries, and elements of the world scientific community, the progress of the electronic tracking group was remarkably smooth. Years later Mengel would be able to recall "a few personality clashes," but in his opinion these were "par for the course for a program that made use of some of the best astronomers in the country." In addition, the middle months of 1957 saw two slippages: the late arrival of construction material at one of the Minitrack installations and a one-month delay in completing the communications network set up to tie together all of the far-flung elements of the system.¹⁴

In April 1956, the technical panel's Working Group on Tracking and Computation (WGTC), stamped "approved" on the plans for the optical tracking system, as drawn up by the Smithsonian Astrophysical Observatory, that is, by Whipple and J. Allen Hynek. Simultaneously the group gave its blessing to Whipple's budgetary estimates, an action that soon thereafter had the effect of setting aside $3,380,610 for the optical tracking program.¹⁵ The plans prepared for the optical system envisaged the use of two separate but cooperating groups, one to acquire or "find" the satellite, the other to track it. Acquisition was the responsibility of the group known as Project Moonwatch. Whipple's initial announcement of his intention to use amateurs in this fashion brought expressions of skepticism from members of the technical panel. "Some of my colleagues," the feisty, personable SAO director would recall later, "were convinced that too few amateurs would volunteer, and that those who did would not always perform satisfactorily. Time, I'm happy to say, has proved these fears to be groundless. The amateurs joined up in droves all around the globe. They did a splendid job for Project Vanguard, and they have been doing an increasingly more effective one for the American space effort ever since."

The nucleus of the other phase of the optical system, the precision tracking phase, would consist of the twelve observation stations, set up around the world and operated by professionals. Each station was to have a high-precision camera and associated clock. It was assumed that reduced data obtained from these installations would permit the calculation of definitive orbits for use in correlating with satellite-borne and ground-based experiments, thus providing valuable scientific information. Early in the planning period staff members of the Smithsonian Astrophysical Observatory undertook to develop a list of possible camera locations. On visits to more than a score of countries, Hynek and other SAO scientists met with local scientists and government representatives to work out collaborative methods for setting up and operating the units. In the beginning thought was given to placing all of them at Minitrack stations. This arrangement would have saved money, but it turned out to be impractical because of the differing requirements of the two systems. "A camera station," Whipple has explained, "needs clear skies, whereas the principal need at a radio tracking station is a flat surface away from noise. The Minitrack station in Ecuador, for example, was ideally located for radio tracking purposes, but it stood in an area where the skies are overcast most of the time. Joining our camera stations with the Minitrack network would have spared us many logistic headaches, but generally speaking we just couldn't do it." In the end an optical and a Minitrack station were combined at only one point, Woomera, Australia. Of the remaining optical locations, two were in the continental United State-at Jupiter, Florida, and Organ Pass, New Mexico. The others were at Olifansfontein, Union of South Africa; Cadiz, Spain; Mitaka, Japan; Naini Tal, India; Arequipa, Peru; Shiraz, Iran; Curaçao, Netherlands West Indies; Villa Dolores, Argentina; and Haleakala, Maui, Hawaii.¹⁶

To cope with the unprecedented task of picking up a tiny man-made satellite orbiting at a great altitude, Whipple and Hynek supervised the development of the Baker-Nunn high-precision telescopic camera with an unusually large aperture.

The Baker-Nunn took its name from its principal creators: James G. Baker, consultant to the Perkin-Elmer Corporation of Norwalk, Connecticut, and Joseph Nunn of South Pasadena, California. Baker designed the camera, Nunn, its mechanical elements. Perkin-Elmer fabricated the optics, and the Boller and Chivens Company of South Pasadena built the camera proper.¹⁷

Although the performance of the cameras was destined to fulfill expectations,¹⁸ their production was plagued by setbacks. The October 1957 meeting of the technical panel at IGY headquarters in Washington found the scientists present discussing the lagging camera-delivery schedule with such heat that for once lean-faced, high-domed Fred Whipple lost his calm. Why so much fuss, he wondered, over the delays in the optical tracking program? Was the panel satisfied with the Vanguard launching schedule? His reference, of course, was to the inability of the NRL-Martin field crew to meet its flight-firing dates.

Panel chairman Porter took care of Whipple's query with his statement that the launching schedule was a Defense Department responsibility, not a panel responsibility. With a sharpness that the formal phraseology of the minutes fails to disguise, he added gratuitously that for Whipple's information the radio tracking program, also a Defense Department responsibility, was practically on schedule.

Whipple promptly changed the subject, or rather shifted it to different grounds. He described himself as miffed by a recent newspaper story charging that delays in the camera-delivery schedule were holding up the entire Vanguard project. Porter agreed that the newspaper story was in error. He offered to so inform the panel's parent body, the IGY committee, but the panel as a whole took no action on this suggestion and the matter was dropped.

Whipple reported that the first Baker-Nunn camera had been assembled and was under test at Pasadena. It would go to its station in a few weeks. A second camera was expected to arrive at Pasadena for assembling and tests by early December. Whipple anticipated that after that the "optics would be completed to allow for camera completion at about one-month intervals." In time he hoped to cut this to three weeks. Clock production was good. The same could be said for the station-building schedule. Basic materials had been shipped to six of the twelve stations, although recently the ship bearing material to Japan had caught fire-damage as yet undetermined.

Porter lost no time in getting to the heart of Whipple's summary. It was plain that the precision-camera network would not be in full operation until August 1958. At that point only four months would remain of the International Geophysical Year as then projected-the period during which Project Vanguard was committed to the launching of at least one satellite.

Porter concluded his gloomy observations by raising-and in effect answering-a pointed question. Should the camera program simply be canceled? In his opinion the answer depended on whether the satellite program did or did not continue beyond the termination of the IGY year. As to that, no one could yet say. A general discussion followed, at the end of which the sense of the panel found its way into the minutes. It was that the members viewed the slipping "delivery schedule of the cameras with grave concern." They had even considered their elimination, but had concluded that such a step would be inadvisable. Implicit in this action, as in so many of the actions of the panel scientists, was an abiding faith that Project Vanguard was only the beginning of a long American space program. Their final word on the camera-delivery problem was that the IGY committee should "adopt every means in its power to expedite the schedule."⁵⁵

In this respect, regrettably, the powers of the National Committee were limited by the severe technical problems involved in the production and testing of a highly advanced optical camera. Observations at the first completed station, Organ Pass, would not begin until November 1957. Not until June 1958-only a month ahead of the date so dolefully forecast by Porter-would the entire precision-camera network be in operation.²⁰

The delay would have been even greater had not Porter visited the plant of the fabricators of the camera optics, Perkin-Elmer, where he learned that the Perkin-Elmer people had underestimated the job, bid too low, and were reluctant about lavishing expensive overtime on an undertaking that was going to leave them with a loss. Having discovered the trouble, Porter persuaded the Smithsonian Institution to renegotiate its contract, raising the price to a point where Perkin-Elmer could break even. This was an intricate transaction, calling as it did for coordination with the National Science Foundation, the Smithsonian Astrophysical Observatory, and the American IGY Committee; but it was accomplished rapidly and had the effect of accelerating the camera-production program considerably.

Lest the delays in establishing the optical network take on more significance than they should, it is worthwhile anticipating a little to point out that its long-run contributions to the scientific satellite project were substantial. Nor do the Technical Panel minutes reflect all of the difficulties confronting the supervisors of the optical program. Actual release of the funds allotted to them was a function of the Smithsonian Institution, and in the 1950s the administrative procedures of the Smithsonian were not geared to a fast-moving project like Vanguard. As a result, Whipple and his associates in Cambridge had to spend an undue amount of time cajoling the firms with which they dealt into supplying them with material and services on credit while they waited for the creaking fiscal wheels in Washington to revolve. Project director Hagen sagely characterized the optical program as a "prudent" backup to NRL's radio tracking system.²¹ The program also engendered great interest in the project as a whole by its sponsorship of Project Moonwatch. This visual observing program-"visual" rather than "optical" since its members made no use of cameras-gave amateur astronomers around the world the opportunity of playing a useful role in the Vanguard tracking system. Organized and directed by the Smithsonian Astrophysical Observatory, Project Moonwatch received as much if not more attention from the press than any other aspect of Project Vanguard. Newspaper readers were intrigued by the picture of hundreds of small groups of enthusiastic star-gazers getting together in open fields or on lonely hilltops in an effort to "catch the satellite" with their binoculars and small telescopes.

Guided by instructions from SAO, the amateurs gathered preliminary orbital data on the satellites. These they transmitted to Cambridge, whence they were distributed in the form of ephemerides to the technicians orbital data on the satellites. These they transmitted to Cambridge, whence

As Whipple had anticipated, the announcement of the formation of Moonwatch in early 1956 brought an enormous response. Visual observation teams sprang up in North America. South America, Africa, Europe, and Asia, in the Middle East and at such remote specks on the map as Station C and Fletcher's Ice Island T-3 in the Arctic Basin. Before the Vanguard program terminated, 250 teams with approximately 8,000 members were functioning. Teams were organized by universities, high schools, government agencies, commercial organizations, private science clubs, and groups of laymen. The United States alone accounted for 126 groups.

The volunteers furnished their own equipment. As the program took form, however, SAO succeeded in obtaining some special equipment from army surplus. The observatory sent these items to some of the more effective groups.²² For a time, ironically, the success of Project Moonwatch was a source of embarrassment to the Smithsonian Astrophysical Observatory. The minutes of the technical panel find Whipple complaining that the unanticipated growth of the program was putting a severe strain on his small administrative staff.²³ So much for the penalties of glamor.

In projecting the electronic tracking program, during the early Vanguard days, the Naval Research Laboratory contemplated using only four Minitrack stations along with a prototype station, but by the time Vanguard was ready to issue its first full-scale report of progress in December 1955, the radio tracking experts were thinking in far more elaborate terms.²⁴ In its final form the Minitrack network would consist of fourteen ground installations.

One of them, the prototype Minitrack station at Blossom Point Proving Ground in Maryland, forty miles south of Washington, was a service-station for the other elements of the network. Here the electronics engineers, the foreign scientists, and other technicians chosen to operate the stations received the bulk of their training. Blossom Point also provided a center for the development of system tests and of procedures for calibrating the Minitrack antennas, principally by using ground-based cameras to photograph aircraft carrying Minitrack test transmitters and ground-controlled flashing lights against a background of stars.²⁵

Three stations, set up immediately downrange from Cape Canaveral on the islands of Grand Bahama, Antigua, and Grand Turk, functioned in connection with a fourth unit, a radar installation at Patrick Air Force Base, to keep tabs on the Vanguard vehicle during launch and shortly after the third-stage powered-flight phase of the launch sequence. Supplemented by a station in South Africa and another in Australia, the remaining elements, the so-called prime Minitrack stations, were strung out in a north-south line along the east coast of North America and the west coast of South America so as to form a "picket line" across the expected path of all satellites launched from Cape Canaveral. Mengel was confident that with this arrangement-this "fence of" stations located generally along the seventy-fifth meridian in the northern and southern hemisphere ²⁶ "we have a 90 percent chance of intercepting every pass of a satellite which is higher than 300 miles."²⁷ In accordance with a decision reached in December 1955, all prime Minitrack stations and some of the subsidiary units of the network included high-gain (TLM 18) antennas for gathering telemetered data from the satellite.²⁸

All branches of the military contributed to the erection, operation, and maintenance of the radio tracking system. Three Army agencies-the Corps of Engineers and its Army Map Service and the Signal Corps-took responsibility for most of the construction work and for setting up the communications network. The Bureau of Yards and Docks, a Navy unit, obtained use of the necessary lands for the Prototype Station at Blossom Point, and the Air Force arranged for the installation at PAFB and Grand Bahama Island of the two high-precision tracking radars, the XN-l and XN-2 models of the AN/FPS-l6.²⁹

With assistance from native scientists, Army men operated the five prime Minitrack stations in Latin America. They also operated the prime Minitrack station at Fort Stewart, Georgia. The Naval Electronic Laboratory took care of the prime station at San Diego, California, with NRL assuming responsibility for Blossom Point and for the vehicle-tracing units on Grand Bahama Island, Antigua, and Grand Turk. In Australia an agency of that country's Joint Service Staff, the Weapons Research Establishment, built and ran the prime Minitrack station at Woomera. In South Africa, the National Telecommunications Research Center discharged these functions at the prime Minitrack station in Esselen Park near Johannesburg.³⁰

Choosing sites for the Minitrack stations was also a cooperative effort. The Corps of Engineers supplied Mengel and his associates with a map study pinpointing potentially appropriate locations. The Department of State conducted negotiations for the lease of lands on foreign soil, and a Vanguard reconnaissance party under Commander Berg received substantial assistance from local representatives of the Inter-American Geodetic Survey during a tour of seventeen Latin American countries in the spring of 1956.

Using criteria previously drawn up at the Naval Laboratory, Berg's group chose six Latin American sites. They were Batista Field at Havana, Cuba; Páramo de Cotopaxi at Quito, Ecuador; Pampa de Ancon at Lima, Peru; Salar del Carmen at Antofagasta, Chile; Peldehune Military Reservation at Santiago, Chile; and Río Hata in the Republic of Panama. Shortly after the return of the Berg group in early May, NRL eliminated the Panamanian site, studies having indicated that a station at San Diego would be more useful. Indeed the day was not far off when the eyes of everyone connected with Project Vanguard would be fixed on this California station. Such was its position in the path of the critical orbit that it was always the first ground unit to receive and promulgate the glad tidings that what appeared to have been a satisfactorily launched satellite was actually in orbit.³¹

While Captain Berg's party and another Vanguard group scouted the world for real estate, tracking and telemetry specialists in Washington drew up specifications covering the material and services their electronic network would require. The Towson, Maryland, plant of Bendix Radio Division of Bendix Aviation Corporation built the Minitrack ground station assemblies, exclusive of the ground antenna arrays and the antenna feedlines. Installed for the most part in government-furnished commercial trailers, the ground stations consisted of ten major components, including rf receiver and rf power supply racks and operating consoles, rf phase measurement and phase measurement power supply racks, time standard racks precisely controlled by a crystal oscillator and analogue and digital recorder units.³²

As for the Minitrack antenna arrays: In May 1956 the Laboratory authorized two companies to develop these. Both produced prototypes in keeping with Vanguard requirements, so the contract went to the Technical Appliance Corporation of Sherburne, New York, as the lower bidder.^c33 Melpar, Inc., of Falls Church, Virginia, developed for the Vanguard vehicle a radar beacon (the AN /DPN-48), capable of furnishing tracking information on the vehicle during flight.³⁴ With minor exceptions, all of the companies involved in supplying hardware and services to Vanguard's combined radio tracking-telemetry network succeeded in meeting their by no means easy schedules. In August 1956, the finesse exhibited on NRL's side of the bargaining table moved the Chief of Naval Research, Admiral Rawson Bennett, to place on record a "comment upon the exceptionally competent manner in which Mr. J. T. Mengel … has conducted the precontract phase of the Minitrack Ground Station Units…"³⁵

Vanguard scientists were aware that the value of their radio tracking network would be directly proportionate to the speed with which they could convert the data acquired into usable form. For this reason they began working intensively in early fall 1955 on plans for a data-processing system.

To assist in this development Paul Herget of the Cincinnati Observatory joined the Vanguard project on a consulting basis in October. A universally respected astronomer, scholarly and firm-spoken, Herget impressed his Vanguard co-workers by the aplomb with which he performed the complex mathematical chores assigned to him. He worked closely with Joseph W. Siry of NRL, a lanky young mathematician, given to interlarding his remarks with thoughtful grunts like a doctor examining an intriguing symptom; and with tall, wispy-haired James J. Fleming, a genial data-reduction specialist with a striking resemblance to Alastair Sim of British movie fame.³⁶

As the data-reduction program took form, NRL brought in two more distinguished astronomers as consultants, Gerald M. Clemence and R. L. Duncome of the United States Naval Observatory, and Hagen set up a Working Group on Orbits, consisting of Siry, as chairman, Fleming, Herget, and the two Navy astronomers. Cooperating closely with the tracking people, the group established and supervised computational procedures, prepared an ephemeris, and extracted geodetic and geophysical information from orbital data.³⁷

Since the data-processing system called for the services of large-scale computers, the welcome mat was out at project headquarters when in September 1955 Cuthbert C. Hurd, then director of electronic data processing machines for International Business Machines, and other IBM representatives dropped in "to discuss the probable computer needs of the earth satellite program."³⁸

In a letter thanking the IBM experts for their visit NRL pointed out that "as now foreseen, a high-speed digital computer (similar in speed and storage capacity to the IBM 704) will be required for calculating the orbit of the satellite, [but]…computer plans are now in the formative stage, and it may be some time before detailed specifications can be formulated for bidding."³⁹

As a matter of fact, progress toward the bidding stage was reasonably rapid. In March 1956, the Office of Naval Research invited proposals "from several possible sources for the renting of computer facilities and the furnishing of mathematical and programming services to the NRL for … Project Vanguard." During April, IBM and two other companies responded. IBM's bid was substantially lower than the others, and the only one that fulfilled all Vanguard requirements. At a cost of $900,000, IBM was to supply six weeks' full-time operation of its 704 computer, in addition to a number of other services at no cost whatsoever to the government. Under the free items were orbit computations during the lifetime of the satellite or the lifetime of the Minitrack, whichever was shorter, for the first three successful satellites; the services of mathematicians for coding, programming, numerical analysis and related tasks beginning on the signing of the contract; one hundred hours of computing time to check programs; the establishment in the District of Columbia of a computing center to be made available on demand; a secondary backup center to be available for emergency use within five minutes of the need; necessary communications between the primary and secondary centers; and any rehearsals necessary for working out the routine of taking Minitrack data and computing the orbit.

In June the Navy and the big business machines company entered into the requisite contract. In July NRL announced that IBM was planning to create a computing facility for Vanguard in downtown Washington. In the spring of the following year IBM's proposed remodeling of the building leased for this purpose received the blessing of the District of Columbia Fine Arts Commission, and on 30 June 1957 the attractive Vanguard Computing Center at 615 Pennsylvania Avenue, NW., opened with appropriate fanfare. Later in the summer IBM provided Project Vanguard's data processing system with standby facilities by installing a transceiver connection between the Vanguard Computing Center in Washington and the company's Research Computing Center in Poughkeepsie, New York. Although this backup arrangement would prove unnecessary, the presence in upstate New York of a 704 computer, capable of handling Vanguard data should anything go wrong with the machines in the Washington Center, was comforting to the reliability-conscious managers of the scientific satellite project.⁴⁰

Preparation of the 704 for Vanguard use began at the Naval Research Laboratory where the Working Group on Orbits determined the formulations their projected launching schedule would require. The bulk of this work fell to Herget, assisted by Peter Musen of the Cincinnati Observatory. When the Laboratory's calculations were ready, IBM mathematicians and programmers translated them into computer language. In all they wrote some 40,000 discrete operations into the machine.

Throughout most of the Vanguard launching program-from December 1957 on, to be exact-the project utilized still another high-speed computer, an IBM 709 that the Air Force had installed in the vicinity of Patrick Air Force Base on Cape Canaveral. As, during each firing, the satellite-bearing Vanguard vehicle rose from its pad at nearby AFMTC, the 709 followed its early flight with calculations based on sightings made by the big tracking radar, the AN/FPS-l6. While the vehicle coasted up to the altitude where its third stage could power the payload into orbit, the preliminary information concerning the coasting phase traveled by teletype to the Computing Center in Washington.

This preliminary data indicated the speed and velocity of the vehicle at that moment. At the Washington Center technicians fed it into the 704. The machine combined it with previously prepared data on the anticipated performance of the third stage, and in this manner produced the first computations showing whether the satellite had a chance of orbiting and, if so, what the orbit would look like.

These initial computations came off the 704 while the third stage of the rocket was burning itself out above the Atlantic ten to twelve minutes after liftoff. From the Center the preliminary orbit predictions went by teletype to the Minitrack stations, enabling them to make accurate observations as the new satellite passed overhead. Each station recorded data about the satellite during the brief time it remained within range, then relayed the readings so obtained to Washington where the experts at the Center converted them to punched cards and fed the cards into the 704.

As the satellite continued to orbit the earth, the Minitrack stations reported further sightings to the Center. With these, the 704 continuously refined its preliminary orbit prediction. Eventually, usually between seven and nine hours after launching, the 704 was able to issue a definitive calculation, showing the shape of the orbit and the speed, position, and altitude of the satellite for every minute throughout the following week or ten days. Simultaneously the 704 determined the geographical locations from which the observers of the Smithsonian Astrophysical Observatory and the Project Moonwatch participants could make visual observations of the satellite.⁴¹

At Cape Canaveral the Vanguard technicians achieved a technological first by merging the 709 computer and the high-precision AN/FPS-l6 radar into an arrangement whereby the radar fed tracking data directly into the computer and the computer, in turn, drove the plotting boards in the Central Control room on the firing range. This radar-computer-controls linkage was a case of multiple applications of the same equipment. Not only did it yield data on orbit determination for transmission to Washington, it also provided the men at the range control rack with the information they needed, during the critical early phase of a launch, to make certain that their vehicle was performing safely and according to plan.

Driven by the computer, the displays in the Central Control room gave the range safety officer a second-by-second picture of the path the vehicle was following. If the rocket swerved from its appointed trajectory and appeared likely to fall on ships or land masses below, the safety officer could press the "panic button" and destroy the flying hardware before it became a hazard. The computer-activated displays performed a similar service for the man at the third-stage firing console. A mechanism in the vehicle was set up to ignite the third stage at a precalculated point in the trajectory. Utilizing data relayed by the computer from the tracking radar, the display boards at Central Control indicated whether this mechanism was functioning properly or not. If not, the operator at the third-stage console known as "Fire-When-Ready Gridley" could push his button, thus actuating ignition of the stage from the ground.⁴²

Installation of the radar-computer-controls system proceeded concomitantly with development for use at the Cape of a system for greatly accelerating the reduction to usable form of data telemetered from the Vanguard vehicle. A joint product of the Naval Laboratory and Radiation. Inc., of Melbourne, Florida, this time-saving arrangement brightened the scientific world's burgeoning list of acronyms by receiving the designation ARRF. for Automatic Recording and Reduction Facility. Housed first on a mobile trailer, later in the project's permanent hangar at the Florida missile center, ARRF provided the Vanguard scientists with final information on the performance of their birds in flight within seventy-two hours after liftoff.⁴³

Creation of the project's tracking and data-processing systems required about a year and a half of planning and labor. The Minitrack network became operative in October 1957, prior to which date Vanguard relied for tracking data on existing AFMTC facilities, supplemented by a prototype of the advanced version of the simplified Minitrack station, a Mark II that NRL had established at the Cape toward the end of the preceding year. ARRF began functioning in late fall 1957 and the radar-computer-controls system began functioning in December of that year. Meanwhile other phases of the scientific satellite program had moved ahead, although in the view of hard-pressed participants, dogged by the sense of urgency implicit in their commitments, progress was never totally satisfactory.