Putting dreams in orbit

Computers have become cheap and portable, cars are turning from steel gas-guzzlers to light, plastic structures propelled by fuel cells. Power stations are small and quick to build, in contrast to their behemoth predecessors. This book tells what happened when a group of clever engineers in Virginia tried to push satellite communications along a similar virtuous curve.

Their company, the Orbital Sciences Corporation, had already performed one space-industry miracle by developing the cheap, aircraft-launched Pegasus rocket launcher. It was technologically innovative and was also the first launcher developed without state support. Now Orbital's president David Thompson, who pushed the firm from a spare-bedroom start-up to quoted company status, wanted to be a communications provider to end users. To write about how it happened, the author of Silicon Sky , Gary Dorsey, was given substantial access to players at all levels. The result, which is part of the Sloan Foundation technology series, is a rich mix of colourful people attempting a near-impossible task.

Dorsey is not shy of giving us technical detail when it is needed. The new business, Orbcomm, aimed to use a large "constellation" of small satellites in low-earth orbit to deliver worldwide communications to small handsets. This meant tough engineering problems. For the satellite to read the signal from a handset on the ground is roughly, as one of the developers put it, like trying to hear a whisper at the other side of a football field during a noisy half-time, while one person shouts in one of your ears and someone else blows a foghorn in the other.

This and other acute technical challenges had to be met by a satellite of minute size and minimal power. It was, theoretically possible because of the growing scope for replacing hardware with software. But making it happen meant Silicon Valley-type effort for the engineers.

Alongside this technical tale there runs a business story. The initial aim was to supply telecommunications services and a global search-and-rescue system - with no more people lost in blizzards, floods, deserts or storms at sea. In practice, it turned out that the big bucks were in tracking freight containers, monitoring meters on gas pipelines, and other prosaic, but massive markets.

Silicon Sky is eloquent on the office politics of the high frontier. Arguments with the launcher side of the firm and with a parallel design department in Colorado ate away at morale. And, while the aim was to replace the big-iron mentality of the US space programme with a newer, faster-paced ethos, Orbital found that the US Air Force and the National Aeronautics and Space Administration turned out to be its biggest customers. It even bought the space interests of Fairchild, an old-economy space company with big Nasa business, causing trepidation among staff that their culture was in danger.

Dorsey also makes it clear that Orbcomm was no virtual company. There were real pieces of electrical, mechanical and electronic engineering as well as all the software, much of it made by contractors of maddening ineffectiveness. In the end, the technical and organisational problems nearly proved fatal. Delays all but destroyed the project's plausibility and, astoundingly, the first satellites launched were crippled by entering the South Atlantic anomaly, a well-known accident blackspot in earth orbit. Even this slow progress depended on wooing new investors (Teleglobe of Canada became a joint owner) and wearying negotiations with regulators.

Dorsey tells this story convincingly, though perhaps with too little context. For example, he barely mentions other organisations, such as Ball Aerospace in the US and the University of Surrey in the UK, with similar missions to make space simpler.

However, Orbcomm still exists after many twists and turns. It now has 35 satellites in orbit, and is in business when better-funded rivals such as Iridium have run out of road. But the initial business plan has gone, in favour of using the satellites to extend corporate data networks to remote locations - a far cry from the altruism of the original mission. Orbital Sciences, too, exists as a notable private-sector success in space, although it has diversified into military electronics and ground equipment. The fact, cited by Dorsey, that the Space Shuttle needs 1.5 million signatures on paper every time it is launched proves Thompson was right to think he could do better than Nasa.

Martin Ince is deputy editor, The THES .