11 March 2010

Rambling Thoughts On Failure

Failure isn't good.

I don't mean nothing good can ever come from a failure. I certainly don't mean all failures are the same. I'm just pointing out that the word "fail" means something, and it's not a positive thing. Redefining failure as something good, no matter how well intentioned, inevitably means we end up talking about something else.

Now, there is real value in recasting a particular situation as something other than a failure. Call it a learning experience, an investment, an education - great! And there's much wisdom in finding goodness in a failure experience. But let's not treat failure itself as something greatly to be desired.

Nobody wants to fail. We want to succeed. Unfortunately, a certain amount of failure is inevitable. But the $1M question is "How much failure is enough?"

NASA's Faster Better Cheaper initiative had a 90% success rate for its first 7 years. Then in 1999, it had a 20% success rate, for a grand total of 62%. For unmanned space missions, the sweet spot is probably somewhere between 90 and 20% (preferably closer to 90, right?). But where? I don't have an answer on that. All I know is that if you can't tolerate failure, then you absolutely deserve every mediocre ounce of "success" you achieve.

4 comments:

Glen Alleman said...

Those early success were for small-missions.
Once the mission reached a larger scale, FBC failed to provide the oversight needed to address the "mission success" factors.
One size does not fit all, and when attempted FBC failed in it own mission.

Unknown said...

I'm not sure the Pathfinder mission to Mars (for example) quite constitutes a "small" mission, but if you insist, I won't quibble.

More to the point, I'm also not sure the fact that FBC kept missions small is an indictment of the approach. Smaller missions do indeed have a higher success rate, even on a per-attempt basis. So, one way to increase our success rate is to keep the size of projects small.

And it's OK to be small. A project's size doesn't determine its scientific worth, and bigger isn't automatically better. Small projects can be highly significant, both scientifically and militarily.

I think one of the things we can learn from FBC is that meaningful space missions don't have to cost so much, take so long, be so big or so complicated. The inevitability of bigness is a myth, as is the belief that bigness equals goodness.

Glen Alleman said...

yes it's OK to be small. But from the Process Governance POV, when the politics of FBC overtook the appropriateness of the Safety and Mission Assurance role, things started to unravel.

As a process improvement (Hammer trained) professional, the process fails when it is not properly applied, adjusted or assessed for unidentified failure modes.

Glen Alleman said...

" think one of the things we can learn from FBC is that meaningful space missions don't have to cost so much, take so long, be so big or so complicated. The inevitability of bigness is a myth, as is the belief that bigness equals goodness."

I would strongly conjecture that there are 100's of other dimensions to this issue beyond the simple statement above.

Bigness is an artifact of the desired capabilities. Reduce the capabilities, and the mission complexity "may" be reduced.

You're working the wrong end of the stick. Start with controlled the "Mission Capabilities" process, not the resulting project.

Shuttle is an example along with FSF. Control the appetite of the mission planners and you "may" get the desired outcomes of FIST.

FBC was an attempt to do that WITHOUT controlling the appetite of the mission planners. JPL finally learned, other Centers inside NASA have not. CLV was a recent example.

FCS is the current poster child for uncontrolled capabilities. SBIRS as well.