Inside an immaculate but otherwise nondescript industrial building just outside the Beltway in suburban Virginia, half a dozen engineers and technicians, wearing lab coats, shoe covers, and hairnets, hover beside a silver-grey 10-foot-diameter metal puck. For years, the bulk of Orbital Sciences Corporation's work has been designing, building, and launching commercial, scientific, and defense satellites-probably the closest thing there is to a reliable, profitable niche in the space business. Now the company is trying something new.

In 2012, this silver and gold puck bristling with guidance sensors and wing-like solar arrays-Orbital's cargo-carrying Cygnus Advanced Maneuvering Spacecraft-is scheduled to navigate itself to a rendezvous with the International Space Station.

Any company in the business of space must be prepared for extreme complexity, as technical, logistical, regulatory, political, operational, and management challenges collide. The up-front costs are tremendous; the returns are uncertain.

Tolerance for error is close to zero, yet the materials and engineering must push the bounds of what is currently possible. And though they seem innumerable, every contingency must be planned for. This isn't just rocket science; it's the business of rocket science.

Orbital has to complete a test mission to show that Cygnus, once released from a rocket about 150 miles above Earth, can navigate itself to the space station, which is orbiting at an altitude of 250 miles. It is an exercise akin to a car automatically merging onto a freeway and matching speeds with a particular car-but in three dimensions.

Once Cygnus has found the space station, it must hold a position about 30 feet away while astronauts complete the hookup using a grappling arm.

Because it's a test flight, the cargo will be inexpensive, but needed, items like t-shirts, underwear, and food. The astronauts on the Space Station will unload the cargo, and then repack it with trash. Cygnus is an expendable craft, meaning that when it re-enters the atmosphere, much of it will burn up any surviving pieces will fall into a remote part of the Pacific.

Before Cygnus can head out on its demonstration flight, the engineers and technicians must grind through tedious, meticulous tasks such as testing every possible combination of connections among 17,000 wires running through the device.

Nonetheless, there is real energy in the room: people at Orbital say that it's exciting and fun to build a spacecraft. But the company also has a lot on the line: a successful demonstration by Cygnus enables a $1.9 billion contract with NASA for eight additional missions.

Risking to Grow
Cygnus is new, but it draws on the technology and production practices that Orbital has been developing for nearly 30 years. It will be carried into orbit in the nosecone of Orbital's new Antares rocket (originally branded as the Taurus II), which will launch from a new launch pad developed by the Mid-Atlantic Regional Spaceport (MARS) and the Virginia Spaceflight Authority at NASA's Wallops Island facility on the Virginia shore.

Any one of these three projects-spacecraft, rocket, or launch pad-would be a major undertaking for any organization. Orbital's ability to bring all three together will define the company's future.

The company has been building rockets and spacecraft long enough to know that each one is a new gamble. "With craft manufacture where you're making a few a year-and these are very complex, very high-energy devices that operate near materials and physical limits-the chance of accurately predicting how long it will take to develop these systems is, I think, near zero," says Bill Claybaugh '83, senior director for human space systems at Orbital.

One thing is certain: it won't be cheap. Aerospace compensation can average $200,000 a year, and, Orbital has more than 600 employees contributing to Cygnus, the rocket, and the launch pad. From 2007 through the 3rd quarter of 2011, Orbital had $350 million in research and development costs, primarily attributed to the Antares and Cygnus programs.

The Commercial Space Industry
People who get into the space industry tend to be passionate about it. Claybaugh worked with NASA and invested venture capital in the industry before joining Orbital. To unwind after a hard week, he builds amateur rockets-not the cardboard, light-a-fuse-on-the-bottom sort, but the kind that starts as a block of aluminum on his lathe and eventually flies to 60,000 feet at four times the speed of sound. And he isn't an engineer at Orbital; he's a business guy, one of the few who wears a suit at a company where jeans and t-shirts are more common.

Claybaugh reports to Frank Culbertson, a former astronaut who went on three shuttle missions and was the only American in space on September 11, 2001 (from the windows of the International Space Station, Culbertson saw the smoke plume coming from Manhattan and the gash in the Pentagon). The CEO of Orbital is a NASA and Hughes Aircraft veteran with aeronautics degrees from Caltech and MIT and an MBA from Harvard. It's a company and industry of high performers.

Claybaugh suggests internal drive is a key factor for people who get into commercial space, since it's never been easy to make a buck.

"Total profitability of the entire American airline industry since 1903 is negative," he says. "And that's almost certainly true of space transportation as well. It's the kind of business that is unlikely to generate any significant profit, and will probably, in the long run, generate a negative profit.

"This is due to what is called the public benefit, private cost problem: the often very large benefit of transportation is in the increased economic activity that it allows, but that benefit is generally captured by government (through taxation) rather than by the transportation industry-the industry is a commodity business in which prices are inevitably driven to near-or below-cost."

Henry Hertzfeld, professor of space policy and international affairs at George Washington University, says, "The overall return from space is, I believe, positive, but it's so diffused through so many sectors there isn't any equivalent to a business ROI for government R and D." He adds, "A lot of things that we take for granted, people don't see as coming out of expenditures on the space program... They just see the shuttle accidents. They read about how expensive it is."

A great many daily activities are touched by the space program-some everyday technologies come directly from the space program; for others, development was sped up and stimulated by the business of getting into space. These include things like GPS systems, weather tracking, many noninvasive medical procedures, electronics and computer miniaturization, and clean-room systems, along with numerous materials and lubricants. And, of course, much of our communication and entertainment passes through satellites.

NASA's budget is around $18 billion annually. The Department of Defense spends more than that on space but the full figures aren't known. In difficult economic times, there's pressure on those budgets. Government agencies try to reduce costs by passing risks and uncertainty to private firms. Many aerospace startups have burned brightly and briefly before disappearing over the past forty years.

Orbital employs some 3,700 employees in six facilities around the country; it had revenue of $1.3 billion and net income of $47 million in 2010. The company has been around since 1982, making it much younger than the old guard of Boeing, Lockheed Martin, and General Dynamics (and also entirely space focused, unlike those behemoth aerospace/defense corporations).

On the other hand, Orbital represents demonstrated experience compared to a crop of recent startups led by Space Exploration Technologies (SpaceX), started in 2002 by PayPal co-founder Elon Musk, which is also building a cargo delivery system with a NASA contract.

Because space companies tend to depend on anchor customers-typically NASA or the DOD-small- and mid-sized companies can face challenges in trying to work with those massive organizations.

"One of the shocking things that all startups go through in dealing with the government is the day that they're holding some review and the company has four people and the government shows up with fourteen," says Claybaugh.

"Boeing and Lockheed know to send twenty even though the presentation only takes four because this is a contact sport. You need to be one-on-one with every one of those government guys to make sure they are getting what they want."

Another barrier to entry into the business is the bureaucracy involved. Here's an example: components for the communication system of the International Space Station are made by a Japanese company. To order a part, Orbital must go through the State Department, because such components are controlled by International Traffic in Arms Regulations (ITAR).

This means that the component arrives packed in a padlocked stainless steel case with vibration sensors both outside and in. If the component doesn't work, Orbital must obtain a license to export it back to the maker-and can't discuss what is wrong with the supplier. To put that in perspective, there are roughly 7,000 parts in Cygnus alone.

Engineering for Space
Setting aside a gaggle of largely unproven new companies, Orbital is the only space-focused startup to have succeeded since 1970 (SpaceX may be joining this very short list). Over its lifetime, Orbital has launched 200 satellites and 165 of its own launch vehicles with a 95% success rate; the challenges are such that failing only one time in twenty is remarkable.

And mistakes are very visible: one of Orbital's smaller rockets, the Taurus XL, had two consecutive failures, each destroying a payload of NASA science satellites with a combined value estimated in excess of $600 million.

Failure can come in many ways. Beyond the physics of lifting a large object out of the atmosphere, there are details like electromagnetic radiation in space, which can cause electronics to fry or reset in unpredictable ways. And tiny bits of debris litter space; they can hit a spacecraft with about the force of a 50 caliber bullet-which is why the craft is tested to see how it stands up to fired projectiles.

"Space is a very harsh environment. These machines are exceedingly complex. That's what makes them expensive," says Hertzfeld. "And we've learned a lot over the years, but there's always something new. The FAA certifies a type of plane, the 777, for example, and every flight is almost identical. I don't think there was ever a shuttle flight that was identical to the one preceding it. NASA certifies flights, not the vehicle. We practically rebuilt the thing each time."

That is why the work on the ground is so painstaking with any space project. "The entire point of this business is that there should be no surprises while you are in operation," says Claybaugh.

One bay in Orbital's manufacturing building is devoted to testing equipment-a $100 million capital investment. In one part of the facility, a commercial satellite is sealed in the Thermal-Vacuum Chamber, which, over several days, in a near-vacuum, reproduces the temperature extremes experienced by the sun-facing and dark-facing side of a craft.

In a separate part of the bay, there is a vibration test underway on another satellite. A cluster of engineers watch a small monitor set outside the massive blast door. The entire building rattles and roars. A Cygnus craft already being built in anticipation of a successful demonstration flight sits waiting for a turn in machines that do electromagnetic and UV radiation tests.

A New Rocket and a New Launch Pad
SpaceX took four and a half years and $300 million to develop the Falcon 9 rocket, which will be launching its cargo missions to the Space Station (a NASA study found it would have cost $1.4 billion to build the same rocket within NASA).

Because the up-front costs are so high, aerospace companies typically develop a new rocket only if they have a contract to do so. But Orbital is paying for Antares itself. This new, larger rocket will open up opportunities for the cargo missions and a wider range of satellite launches.

Asked when Orbital expected a financial payoff from the Antares, Claybaugh says, "Sometimes investments are strategic rather than purely economic. Orbital's investment in Antares can be best understood as a Real Option: Antares allows multiple future business opportunities that do not exist in the absence of that transportation capability."

Rather than starting from scratch, Orbital decided to assemble a rocket from components built largely by other companies. For example, the company bought mothballed engines that had been manufactured for the Soviet moon program.

While such recycling isn't unusual, it comes with its own risks: during an engine test last summer, a piece of corroded metal failed, leading to a burnout. To lower the risk of such an event in flight, Orbital developed a non-invasive test to detect corrosion. The company will also perform a full launch test of Antares before sending Cygnus aloft.

For all the technical challenges of the rocket and the spacecraft, the new liquid fuel launch pad at the Mid-Atlantic Regional Spaceport at Wallops Island has been the hardest project to keep on budget and on schedule. The initial plan for the launch pad called for the state of Virginia to be the owner, enabling it to provide launch services to other organizations. Additional costs have required investment by Orbital and a more complicated business arrangement.

Beyond the financing and ownership challenges, the pad is creating the need for new infrastructure for liquid-fueled rockets. "For the last 50 years there has been a liquid fuel competency developed in Florida around Cape Canaveral and the Kennedy Space Center," says Les Kovacs, the Wallops site project manager.

The area has gathered multiple layers of contractors, subcontractors, and sub-subcontractors with increasingly niche skills. "In Florida, if you need to get a hose cleaned to liquid oxygen capability, you get it back in half a day," Kovacs says.

"Here, because there is no in situ capability, we have to send it to New Jersey and it may be a four or five day turnaround. The schedule implications are much more severe."

Launching a rocket is a controlled explosion of extraordinary power. The sound alone is so powerful that it could bounce off the ground and knock the rocket off its path or destroy it all together; during a launch, nozzles on the launch mount shoot water into the thrust plume of the rocket to dampen the sound.

More water is dumped over the entire mount to keep it from being damaged. (Some 200,000 gallons will be used in each Antares launch, roughly the volume of 10 swimming pools.)

Even tucked into the nosecone of the rocket, Cygnus will experience much of the violence of liftoff. Back at the Orbital manufacturing building, the craft undergoes a special test to make sure it will stand up to these conditions: sound technicians who usually rig stadiums for rock concerts surround the spacecraft with a wall of speakers and bombard it with white noise that reproduces the shock and shuddering of liftoff.

Cygnus's test mission is expected to take place during 2012 (in addition to coordinating the completion of spacecraft, rocket, and launch pad, Orbital must find a launch date that fits into the space station's surprisingly crowded schedule). If things go well, Claybaugh says, "You are going to see at least one company in the business of transporting cargo to the space station. That's literally a new line of business that has not existed previously."