The Columbia Disaster Investigation

I hadn't bought a copy of The Atlantic magazine in years, until today. I had some time to kill today while my car was getting a much-needed brake and alignment job, so I walked to a local book store and headed over to the magazine rack to browse. The cover photo of this month's issue showed the space shuttle Columbia lifting off, and I immediately decided that it would be worth buying.

The article is not on-line. Well... actually, it is... it's just not free. You can purchase the Columbia article here. I'd write up a few little fair-use excerpts, but my daughter has absconded with my copy of the magazine, and that means I'll probably not see it again. ;-)

The reason I think it's worth mentioning this article here isn't that it's full of scientific details about the disaster. In fact, it's a little bit light on them, and there's a reason for that. What the investigation concluded, and Langewiesche documented, was that the Columbia disaster, like the Challenger disaster years ago, certainly was the result of a failure in a critical component of a highly complex physical system -- but the space shuttle is far, far more than just a complex physical system. Each flight of each shuttle is a complex engineering project. involving so many risks that physical redundancies, failsafe systems design, and contingency procedures -- though all vitally important -- are not the most difficult problem. The human side -- the management and communication issues -- are harder to get right. In both the Challenger and Columbia disasters, they failed.

The article mentioned email several times. It also mentioned Powerpoint -- and not favorably at all, by the way! At one critical meeting, the single most critical fact -- that there was insufficient information to judge the real consequences of the foam impact on the shuttle wing -- was lost in the sea of bullet items describing and analyzing what little was known. As for the email, what was described was more or less a typical case of crossing messages resulting in incomplete understanding. The article didn't mention any other forms of electronic collaboration going on within NASA, which makes me wonder... What if the culture of NASA -- which was more at fault for the breakdowns than any individual -- encouraged open collaboration, frequent posting of problems, ideas and comments in a central place where all interested eyes might see it? What if there had been a "mission blog"? Would the decision not to get satellite imagery of the shuttle's wing damage have been reversed? We can't know the answer to that, of course, but I still wonder.

By the way, the real reason I decided to buy the magazine and read the article was that I recognized the name of the author, William Langewiesche, who was the only journalist given unrestricted access to Ground Zero in New York City immediately following 9-11, and whose book "American Ground" about what went on at Ground Zero appeared on many of the Top 10 lists for 2002. I figured that his take on the Challenger investigation would be worthwhile. There's an interview with Langewiesche on the Atlantic's web site, and that is also worth reading.