The Orbcomm-X mission originated in early 1990, not long after Orbital Sciences Corp.'s branch offices opened in Boulder.

In defining the technology, President Alan Parker had searched for appropriate portions of spectrum that would suit his business plan -- the plan being to make a satellite phone service as inexpensive as possible. As a data-only system, Orbcomm needed only slender patches of spectrum in the UHF/VHF bands, which would allow the company to rely on commercially available electronics -- radio parts, essentially. With the help of consultants, Parker considered several slices of bandwidth. Soon he also realized that interference in any band he selected would have to be well-documented so he could assure investors that it would operate clearly.

Choosing frequencies was perhaps the most significant risk in the project. At the time, most of what was known about radio interference within particular UHF/VHF bands from space was based almost entirely on theoretical assumptions. Although U.S. defense satellites had, in all probability, collected an enormous amount of data on space noise across the spectrum for decades, the Federal government prohibited public collection or publication of such data on the grounds that knowledge of these matters would jeopardize military defense. Whatever documentary evidence already existed remained strictly classified.

Boulder Division Director Jan King's engineers thought they knew better. According to their experience and their sources -- most likely, other satellite engineers -- bands under review by Parker should have been clear enough. If they were not totally free of interference, King believed, they would at least be quiet enough to support a simple data system like Orbcomm.

Parker couldn't be convinced. If the bands he had chosen proved noisy, Orbcomm would be doomed from the first launch. Likewise, if Orbital could not provide data that argued convincingly for specific spectrum selections with the International Telecommunications Union, the organization that allocated radio spectrum, then Orbcomm would never open for business. The need to choose the clearest set of frequencies weighted so heavily that Parker finally ordered the Boulder team to build him a small satellite. He wanted something simple but elegant that would collect information on interference levels from space. Engineers in Boulder went to work on a lightweight satellite the size of a briefcase, which they called Datasat.

In the meantime, other consultants hired by Parker began to consider alternate bands they thought might prove more efficient and less costly overall.

Indecision about which frequencies to employ finally stymied King's engineers. Rather than continually rebuild the little Datasat as Parker's consultants made other choices, they simply decided to put the unfinished satellite in cold storage -- an old Frigidaire -- and await more definite orders.

A few months later, in February 1991, Parker called with a firm set of frequencies and said he had scheduled a piggy-back ride with a remote-sensing satellite set for launch in July. With only five months to complete the instruments, the Boulder team revived Datasat, renamed it, and went to work days, nights, and weekends overhauling for an impending launch.

Orbcomm-X, as it was then officially known, presented a clear expression of classically conservative, awesomely inexpensive ideas, reflected most ostensibly by an unusually simple and efficient design. For one, it weighted only about 35 pounds.

Having built and launched more than a dozen satellites themselves, the Boulder team had learned to take any initial schedule, multiply the number of days to launch by 1.7, and fairly predict the actual launch date. A margin of a couple of weeks, in any case, they assumed, would probably bump the date to August or September of '91, giving them time to do an assortment of tests, particularly with the satellite software, which required a "burn-in" time of several hundred hours.

Instead, the launch held firm. By early July, the Boulder engineers found themselves scattered across the world, some working night-and-day in a clean room near the launch site in Kourou, French Guiana, and others at Orbital headquarters in Virginia debating with Parker and other executives the need to forgo an immediate launch so they could continue testing. In the end, rather than risk losing a relatively inexpensive ($200,000) piggy-back launch or risk damaging the satellite by flying it back home, the engineers crossed their fingers and agreed to launch Orbcomm-X without a final round of software tests.

More than their reputations were at stake. Orbital CEO David Thompson had unveiled the satellite at a public ceremony that spring among a group of dignitaries in Virginia. The state of Virginia had kicked in $250,000 to fund the project and representatives from the Environmental Protection Agency had offered equipment to carry out tests with the satellite. AT&T, Ford Motor Company, the American Automobile Association, and Allstate Insurance each had heard detailed presentations about the Orbcomm system and awaited the results of Orbcomm-X data. The FCC expected to see whether a successful launch would provide data that made Orbital eligible for an unusual "experimental" license status from the agency, recognition that would grant competitive benefits toward the launch of the first two satellites. Competitors, private investors, and Wall Street firms also expressed curiosity about the launch, suggesting that detailed questions would follow as data poured in from space.

"When Orbital's Orbcomm-X satellite goes up this summer, it will carry with it Virginia's hopes for achieving a $10 billion space industry within a decade," Thompson announced in a posh ceremony at the Center for Innovative Technology, a state-owned new technology think tank. As Orbital's first satellite, Orbcomm-X was heralded like the arrival of a new baby.

Within hours of launch on July 16, 1991, Orbcomm-X nailed its 480-mile circular, nearly polar orbit. But on its second revolution around the Earth, the satellite suddenly let out a loud squelch. One of the Boulder technicians heard the scream on his radio monitor while driving into work. He reached for his tuner and adjusted the volume. Nothing. Pure static.

Unlike many lost satellites, Orbcomm-X never left any sign of why it disappeared. Satellites rarely simply abandoned the Earth without comment. But the industry lived with ghosts of a few nightmarish blundersats for whom, once launched, Murphy's Law ruled. On those occasions when the cause defies investigation, an enduring torment attacks its engineers. The team at Boulder spent weeks testing models of Orbcomm-X in its lab searching for clues that would explain what happened. Every effort failed. They never again made contact with Orbcomm-X.

Fortunately, Parker arranged enough support among delegates from around the world to avoid a technical challenge from ITU. The Orb-X failure did not prevent him from winning the frequencies requested.

At the same time, Orbcomm's potential investors turned cold with news of the failure, a response that proved almost as disastrous as having lost the bid for spectrum. Parker made it known that he would never completely trust the Boulder crew again.

Gary Dorsey has written narrative non-fiction for more than a decade. He is the author of the critically acclaimed Congregation: The Journey Back to Church, a portrait of contemporary church life, and The Fullness of Wings: The Making of a New Daedalus, which chronicles the creation of a human-powered airplane. He currently works as a science writer at Johns Hopkins University and lives with his family in Catonsville, MD.

Reprinted with permission from Silicon Sky: How One Small Start-Up Went Over the Top to Beat the Big Boys into Satellite Heaven
by By Gary Dorsey.
Copyright 1999, by Gary Dorsey
Published by Perseus Books, a member of the Perseus Books Group
A Cornelia and Michael Bessie Book
Sloan Technology Series
Adapted by permission of the author and Perseus Books
May not be modified, reproduced, republished, uploaded, posted, transmitted, or distributed in any manner.
Available beginning April, 1999, from bookstores nationwide, online retailers, or by calling 800 386-5656