Reaching for the stars is the ultimate challenge

Space Transport System flight 107 (STS-107) started with the launch of the Columbia space shuttle from Cape Kennedy on January 16, 2003.

It was the 28th flight of Columbia , which was the historic first of the shuttle spacecraft to enter Earth orbit. On board STS-107 were seven astronauts, six of them militarily trained (one an Israeli air force colonel) and one a female civilian scientist, Kalpana Chawla.

Eighty-one seconds into an otherwise textbook perfect launch, a 1.7lb piece of the foam that was insulating the cryogenic fuel in the main external fuel tank fell off and struck the forward edge of the left wing of the Columbia spacecraft. It weakened a reinforced carbon-carbon panel, part of the wing's thermal protection. Sixteen days later, on February 1, 2003, Columbia descended to Earth. During re-entry into the atmosphere, superheated air penetrated the damaged area of the panel and progressively melted the aluminium structure inside. The wing failed, the pilots lost control of the spacecraft and it disintegrated at 12,000mph at an altitude of 200,000ft over Dallas, Texas. There was no possibility that the crew could have survived the disaster.

There had been significant trouble on 13 previous shuttle flights associated with the thermal protection system, including the first, when 300 insulating tiles fell off the spacecraft. Six of the incidents involved foam that had been shed from the same area of the main fuel tank. After the third such incident, which took place on STS-50 in 1992, the evaluation of the incident concluded that the event was definitely outside the design requirements for the shuttle (ie, could well cause a serious accident) but was an acceptable risk to shuttle missions. The issue became one of maintenance, with repairs and strengthening of the foam at the critical place required between flights. In being so defined, the issue ceased being one of flight safety.

This mind-set, developed as part of the launch cycle of the shuttles, continued into STS-107. The separation of the insulating foam had been recorded by video cameras set up to monitor the launch, and noticed during a Nasa review of the launch the day afterwards. Months before, in February 2004, on flight STS-112 (the flights had got out of the planned sequence) a similar foam-debris strike had occurred. The mission pressed on to completion, regardless.

The same happened with STS-107. Knowing that there had been a strike on the left wing, some engineers called for examination of the area by crew members and/or imagery by satellites. to evaluate the damage The mission management denied these requests, in the belief that these actions were unnecessary because the problem would be fixed on return to Earth, and because an unscheduled space walk by the crew would interfere with the mission programme. In any case, the managers believed that nothing could be done, whatever was found. In fact, there was the possibility that the damaged area could be repaired in orbit (with "high risk") or that the crew could be rescued ("challenging but feasible").

All this came out in the report of the Columbia accident investigation board. It described the proximate causes of the Columbia disaster. The present book looks behind the report. Why did Nasa continue to launch spacecraft with a history of potentially catastrophic incidents? Why did the managers conclude that the foam debris strikes were not threats to the safety of the missions? What was the safety culture in Nasa and did the culture contribute to the occurrence of the accident? Did the external pressures of politics and budgets provide an environment that made the accident more likely?

The book is thus not really about the Columbia disaster at all. It is certainly not a nerdish examination of the design and operation of the spacecraft, although inevitably the engineering forms the background to the book's discussion. The book is about organising safety in large complex organisations and projects, as illustrated by the Columbia spacecraft disaster. It is about how society, organisations and individuals respond to risks and how, in some cases, the very measures taken to reduce risk (patching the vulnerable area of foam) contribute to it (by providing a reason to postpone action in flight to repair the damage).

These are not issues that are encountered only within Nasa. After the Old Bailey trial of Railtrack and Balfour Beatty for breach of health and safety regulations after the Hatfield rail crash of October 2000, Ian McAllister, the chairman of Railtrack's successor, Network Rail, said that new technology had "changed the mentality of the industry away from 'find and fix' to one of 'predict and prevent'". It seems that even after the lessons of STS-107 had been well rehearsed, they still remained to be learnt by Nasa, given the fact that in August this year the fuel tank of the first "return to space" shuttle flight, STS-114, shed potentially catastrophic foam debris.

The book is an edited collection of 18 chapters by 36 authors, virtually all of whom are academics in business management schools and the like. It will find a suitable place on the shelves not only of a specialist space library but also of the libraries of any organisation that has an interest in complex, innovative, risky technology. In the conclusion, the book's editors identify the factors that push organisations to and beyond their limits. Large size causes organisations to decentralise and fragment.

Complex technologies compel organisations to develop complex management structures. Problems evolve quicker than the organisation can develop solutions. Weak prioritisation of goals causes organisations to waste resources on unimportant activities while performing primary activities poorly. Some organisations are run in an environment that is incompatible with their goals - finance may be provided short term, whereas the goals are long term.

There are no heroics in the book, no sentiment, no glorification of the "right stuff". In fact, the Columbia disaster occurred at a time when the manned space programme had been making the transition from its origins as the kind of activity that appeals only to test pilots and groupies to include a growing civilian participation.

There remains one chance in a hundred that a space launch will end in disaster (consistent with this number, there have been two shuttle disasters in 120 flights) and it is the ambition of Nasa in designing the next generation of reusable spacecraft to reduce the probability to one in one thousand. The probability of being killed on a given trip in a car is about 1 in 10 million, in a civil passenger aircraft 1 in 50 million, in a train about the same. As the book shows, space travel is not yet ready to be certified as a means of public transportation. The organisation of space travel will be "at the limits" for years to come.

Paul Murdin is senior fellow at the Institute of Astronomy, Cambridge University, and former director of science, British National Space Centre.