Science: 1998: A real space odyssey

As it turns 40, Nasa is shedding its bureaucratic image and heading for the stars under a brave navigator.

Employees of America's National Aeronautics and Space Administration (Nasa) will celebrate the new millennium this month, 14 months before everyone else. The reason for this strange timing is that Deep Space 1, the first spacecraft in Nasa's New Millennium programme, is scheduled for launch on 25 October.

The significance of Deep Space 1 derives not from its monumental dimensions or vast cost, but, instead, from its modest size, innovative design and rapid development. It represents the new face of Nasa, an example of the reinvention of an ageing organisation that has just passed its 40th birthday.

The new source of inspiration is the agency's director, Dan Goldin. When he arrived in 1992, Dr Goldin inherited an agency that spent more than 50 per cent of its budget on sending shuttles soaring into orbit instead of investing in the cutting-edge research which the nation and space programme required. Critics began to label Nasa as a bureaucratic dinosaur, unable or unwilling to adapt to the needs and aspirations of 21st-century science and technology. It was a far cry from Nasa's heyday 30 years ago, when its Apollo missions put men on the Moon.

Faced with a declining budget and organisational inertia, Dr Goldin realised that, as it reached middle age, Nasa required a radical new philosophy if it was to survive and prosper in the next century. Determined to change the agency's direction by switching from operations to research and innovation, he called for a "smaller, faster, cheaper" approach to the design and development of future spacecraft.

The first fruits of this call to arms were the Discovery missions, relatively small-scale scientific craft whose launch and development budget was capped at $150m (pounds 94m) - peanuts, in the space business. Despite some problems, this programme has already borne fruit with the notable successes of the Mars Pathfinder, with its small automated rover, and the Lunar Prospector, which confirmed the existence of water and ice around the lunar poles.

The next stage in this reinvention process is the New Millennium programme. The eventual aim is to develop "smarter" fleets of microspacecraft - miniaturised, automated armadas of "intelligent" probes that can be sent to all corners of the solar system and beyond. Nasa envisages launching dozens or even hundreds of these microchip-sized spacecraft.

Deep Space 1 marks the first step along this road. Although the mission may return significant scientific data, its main raison d'etre is to test 12 new advanced technologies. These are crammed on to a spacecraft weighing less than half a ton and less than five feet high.

The most important innovation is a propulsion system that is 10 times more fuel-efficient than traditional chemical rocket engines. Electrical power of 2KW is provided by concentrating sunlight on to panels made up of gallium arsenide solar cells. This is then used to accelerate atoms of xenon gas, causing them to be fired into space at high velocity. Although such "ion propulsion" produces a tiny amount of thrust - about the same as the pressure of a sheet of paper on the palm of your hand - the continuous expulsion of gas has a telling effect. If used on a near-continuous basis for several months, it can eventually propel a spacecraft across the Solar System at more than 70,000mph, about seven times faster than the fastest chemical propulsion engines.

"This marks an exciting step in deep space exploration," said Jack Stocky, the engineer leading the ion-thruster programme. "After years of speculation about the potential of this form of propulsion, we are finally nearing the day when we can validate solar-electric propulsion as the system of choice for tomorrow's distant missions."

Another innovation that has attracted attention is the so-called "autonomous operations system", which has been likened to HAL9000, the renegade computer in the film 2001: A Space Odyssey. Deep Space 1's computer relies on an artificial intelligence program, called Remote Agent, which is intended to operate and control spacecraft with minimum human assistance.

"The goals of the Remote Agent development are twofold: to reduce the cost of exploration, and to extend exploration to realms of space where no ground-controlled craft could venture," explained Bob Rasmussen, a computer expert at Nasa's Jet Propulsion Laboratory.

One of its most significant features is the spacecraft's ability to monitor its health, schedule on-board activities and distribute resources such as electrical power, so replacing most of the human controllers back on Earth. The Remote Agent computer program can also diagnose failures and suggest solutions, while another part of the program has overall control of the spacecraft and can even reject flawed commands transmitted from the ground.

Linked to the Remote Agent is an automated navigation system which uses positions of asteroids and stars to calculate and correct the spacecraft's course. Other revolutionary technologies on board Deep Space 1 include a miniature camera and spectrometer, and a low-mass communication system that uses a high-frequency, solid-state amplifier to boost the radio signal.

If this groundbreaking hardware survives the first few months of flight, it will be employed to fly within a few miles of an asteroid on 28 July 1999. The option will then be open to climax the spacecraft's landmark mission with visits to two comets over the following two years.

Deep Space 1 will be the forerunner of an ambitious programme intended to develop cutting-edge technologies for the space exploration missions of the 21st century. The next in the series, Deep Space 2, is set for launch in January 1999. This time, the intention is to test technologies that will enable miniaturised craft to roam the solar system and send back data.

Deep Space 2 comprises a pair of 4.4lb prototype penetrators designed to pierce and probe the soil near the Martian south pole. They will ride piggyback to the Red Planet on board Nasa's next Martian lander, then be released shortly before entry into the thin atmosphere. The protective heat-shields on the dart-shaped penetrators will shatter as they strike the surface at more than 400mph, allowing the probes to separate into two sections as they bury themselves into the topsoil.

While the five-inch-diameter circular top section, which contains the batteries and a micro-telecommunications system, will remain on the surface, the pointed four-inch-long nose will be buried some six feet into the ground. Information on the water content and temperature of the subsurface layer will be relayed to the upper section along a cable that unravels during separation of the two segments.

Looking even further ahead, Deep Space 3 will attempt to pioneer a powerful new technique which could eventually enable astronomers to find Earth- like planets around other stars. Set for launch in 2002, it will involve precision-flying by multiple spacecraft, each equipped with its own optical telescope. These will be linked together so that they can achieve a spatial resolution far superior to any existing telescope.

Just as ambitious is the plan to drop a lander on to Comet Tempel 1 from the Deep Space 4 spacecraft as it flies alongside the comet's black, icy nucleus. After a seven-year odyssey through the Solar System, the ultimate challenge in 2010 will be to return to Earth a sample of the primordial material which makes up this fragile leftover from the birth of the Solar System.

Whether these high-risk ventures succeed or fail is not particularly significant. The most important legacy of the New Millennium programme will be its influence on global space-exploration strategies. The European Space Agency has already climbed onto the "smaller, faster, cheaper, more capable" bandwagon by planning its own series of "smart" missions, and others are sure to follow.

Peter Bond's new book, `Zero G', published by Cassell, is due next spring