Wednesday, December 31, 2008

Shuttle Missions are Dangerous Enough Without Management's Help

Another report on the Columbia accident was released yesterday. This one was authored and researched by astronauts, some of them close friends of fellow astronauts who died on Columbia.

What struck me of the report are the echos of Challenger. It's painful. From a New York Times article about the report:

The impact of the foam was obvious in videos taken at launching, and during the Columbia’s 16-day mission, NASA engineers pleaded with mission managers to examine the wing to see if the blow had caused serious damage. The managers, however, held firm to the then-common belief that foam strikes were relatively harmless and constituted a maintenance problem, not a fatal risk.

In a scathing report issued in August 2003, an investigative board later found that a “broken safety culture” at the National Aeronautics and Space Administration was largely responsible for the deaths. It criticized managers as complacent and too tightly focused on scheduling and budgetary pressures.

This means that little changed since the 1986 Challenger disaster when o-rings made by Thiokol failed, causing the Challenger to blow up. The o-rings failed because the Challenger had launched in freezing temperatures, and the o-rings were only rated to 40°F. From a Wikipedia writeup about the accident:

...At a teleconference which took place on the evening of January 27, Thiokol engineers and managers discussed the weather conditions with NASA managers from Kennedy Space Center and Marshall Space Flight Center. Several engineers—most notably Roger Boisjoly, who had voiced similar concerns previously—expressed their concern about the effect of the temperature on the resilience of the rubber O-rings that sealed the joints of the SRBs. Each SRB was constructed of six sections joined in three factory joints and three "field joints".

...Despite public perceptions that NASA always maintained a "fail-safe" approach, Thiokol management was influenced by demands from NASA managers that they show it was not safe to launch rather than prove conditions
were safe. It later emerged in the aftermath of the accident that NASA managers frequently evaded safety regulations in order to maintain the launch manifest (schedule).

That NASA management, almost 20 years after the Challenger accident, had not learned to listened to their engineers is despicable. But I think a deeper problem sits below. For reasons I don't understand, NASA promotes an image of its manned space programs "fail-safe". Oh sure, everybody "knows that they are dangerous"... But on the Challenger they were sending a teacher, Christa McAuliffe, into space! (Factoid: my elementary school was named after her.)

NASA should change two things, then:

1) It needs to make missions actually safer by listening to its own engineers, and those of its contractors.

2) It should stop promoting their "fail-safe" image and instead emphasize that astronauts are really heroes on dangerous missions operating under immense personal risk.

Flying on the shuttle is dangerous enough. Both shuttle disasters were avoidable (especially Challenger). Maybe NASA managers have been drinking too much of their own "fail-safe" kool-aid, believing their systems to be so robust they would withstand their bad decisions. Maybe by embracing the fact that these missions are so dangerous, the managers would actually get a clue and understand how much trust is being placed in their hands. Maybe, just maybe, they would take their responsibilities to the astronauts more seriously.

Monday, December 29, 2008

Congratulations SpaceX, Orbital Sciences Corp, and NASA

NASA has awarded contracts to SpaceX and Orbital Sciences Corp. contracts to deliver cargo to the International Space Station for the next dozen or so years. This is news, big news, huge news, for the geek that I am. From a Wall Street Journal article:

"When the U.S. space shuttle fleet is retired sometime after 2010, such so-called Commercial Orbital Transportation Services contracts temporarily will be the only way for U.S. spacecraft to reach the space station."

This is a huge risk. These companies do not have the traditional experience of building such systems for the government. But the news is even bigger than this.

The procurement will be via the NASA "Commercial Orbital Transportation Services" contract type. So what's the big news about this? It goes like this:

If the government wants to buy, lets say, a military base, it can't just buy one. It puts a team of government employees together to put together ideas for what the base should be like and then look for contractors willing to do it for a good price. This is how our government pretty much gets rockets today. It doesn't buy "rockets", it buys "rocket programs". (As an alternative metaphor, imagine if you couldn't buy clothes for yourself, but instead had to keep creating clothes patterns for what you wanted and shop the designs out to tailors for a good price. That would be an annoying world!)

If the government wants to buy an SUV, it just friggin' buys it. It doesn't need to assemble a team to put together ideas for what an SUV should be like! Sure, there might be some mods but for the most part the government gets in line with the rest of us and buys a friggin' SUV!

THIS is the real news. SpaceX and Orbital want to sell rockets commercially, more or less the way auto companies sell cars. So if our government wants a rocket, it should get in line with telecom companies and other governments and buy the next rocket slot.

This is a big gamble. NASA will take a sort-of hybrid approach, trying to be as hands-off as possible while still assuring success. It will check in on these companies at critical junctures and try to stay out in between. If it works, it would point the way to a new way to procure rockets, though. No more managing the contractors, no more having to create requirements for rockets, no more eating the inevitable cost overruns and the schedule slips. Just buy the next rocket.

If either SpaceX or Orbital succeed, then the entire experiment succeeds because it proves the concept. Even better, there are added dividends for our nation:

0) It proves that space companies can exist as truly private entities pursuing commercial interests. Contrast that to the current base of government contract-driven mega corporations.

1) Commercial companies should spur rocket innovation faster than Lockheed or Boeing or Northrop (or Energia or Arianspace, for the internationally minded) because they can try stuff out without waiting for government permission or government requests.

2) Commercial companies should reduce the overall cost of these systems to the point where more rocket launches should be possible, which should create more higher-paying jobs that are supported by the private industry.

3) By being a commercial company, SpaceX and Orbital will reduce the price not only for satellite operators and governments, but also for our fledgling Space Tourism industry.

All of this is great. Outstanding.

Lastly: I just really want to applaud Elon Musk for having the courage to gamble so much of his private fortune from his PayPal sale to create SpaceX. His achievement cannot be overstated: to privately develop a rocket from scratch. It hurts me a little to say this since I work at one of the government contract-driven megacorporations, but I think that his Silicon Valley shtick is going to rock our Aerospace world.