"Building rockets is hard." Part of the problem is that space travel is in its infancy. Although humans have been launch-ing orbital vehicles for almost 50 years now ｭ about half the amount of time we have been flying airplanes ｭ contrast the numbers. Since Sputnik, humans have launched just over 4,500 rockets towards orbit (not counting suborbital flights and small sounding rockets). During the first 50 years of aviation, there were over one million aircraft built. Almost all of the rockets were used only once; most of the airplanes were used more often.

There is also the issue of performance. Airplanes slowly built their performance from the tens of miles per hour the Wright Brothers initially managed to the 4,520 mph that Major William J. Knight flew in the X-15A-2 research airplane during 1967. Aircraft designers and pilots would slightly push the envelope, stop and get comfortable with where they were, then push on. Orbital rockets, by contrast, must have all of their performance on the first (and often, only) flight. Physics dictates this - to reach orbit, without falling back to Earth, you have to exceed about 17,500 mph. If you cannot vary performance, then the only thing left to change is the amount of payload - the rocket designers began with small payloads and worked their way up.

Rockets, by their very nature, are complex and unforgiving vehicles. They must be as light as possible, yet attain out-standing performance to get to orbit. Mankind is, however, getting better at building them. In the early days as often as not the vehicle exploded on or near the launch pad; that seldom happens any longer. It was not that different from early airplanes, which tended to crash about as often as they flew. Aircraft seldom crash these days, but rockets still fail between two-and-five percent of the time. This is true of just about any launch vehicle - Atlas, Delta, Soyuz, Shuttle - regardless of what nation builds it or what basic configura-tion is used; they all fail about the same amount of the time. Building and launching rockets is still a very dangerous business, and will continue to be so for the foreseeable future while we gain experience at it. It is unlikely that launch-ing a space vehicle will ever be as routine an undertaking as commercial air travel - certainly not in the lifetime of any-body who reads this. The scientists and engineers continually work on better ways, but if we want to continue going into outer space, we must continue to accept the risks.

Part One of the report of the Columbia Accident Investiga-tion Board is organized into four chapters. In order to set the background for further discussion, Chapter 1 relates the history of the Space Shuttle Program before the Challenger accident. The events leading to the original approval of the Space Shuttle Program are recounted, as well as an examination of some of the promises made in order to gain that approval. In retrospect, many of these promises could never have been achieved. Chapter 2 documents the final flight of Columbia. As a straightforward record of the event, it con-tains no findings or recommendations. Chapter 3 reviews five analytical paths ｭ aerodynamic, thermodynamic, sensor data timeline, debris reconstruction, and imaging evidence ｭ to show that all five independently arrive at the same con-clusion. Chapter 4 describes the investigation into other pos-sible physical factors that might have contributed to the accident, but were subsequently dismissed as possible causes.