Saturday, June 12, 2010

The Big Oops

Gulf oil spill from space
My last post turned out to have some bogus calculations - adding/dropping zeroes, a rich source of careless error - which by one of those semi-happy accidents turned out to yield a correct final result - unless you looked at it too closely. H/T to commenter nqdp for wondering out loud why my figures didn't all quite make sense.

I've corrected the errors (or so I hope), while preserving the original text for posterity.

The stakes in this case were remarkably low; at worst I wasted a few hours of your collective time. In real world applications, of course, the cost of brain dead errors is higher, and the errors will hardly ever be neatly offsetting. The example that comes first to mind is the infamous feet/meters fail that did it for the Mars Climate Orbiter. There are of course many others.

Murphy is an Old Testament style prophet, not a gentle father confessor. He comes into your lab or shop to thunder a warning, not give absolution after an error has blown up in your face.


The image of the Gulf of Mexico is from NASA.

16 comments:

Mr. Blue said...

MAAATHH! YOU BASTARD!!!
(shakes fist)

Citizen Joe said...

Gulf of Mexico, now with Invigorate!

Anita said...

A bit off topic, but not much

Congratulations are in order.

Japan's Hayabusa space probe successfully returned home, Oz to be exact, a few hours ago.

The little capsule that could spent seven years journeying over a billion miles around the solar system, even made a brief pit stop on an asteroid.

Well done Traveler.

Rick said...

Hayabusa is a nice counterpoint, because it was a hugely complex mission, and I believe it looked like a goner more than once.

The time scale of space missions sneaks up on you. I remember reading some while back that Hayabusa was approaching Earth, but then forgot all about it again.

Anonymous said...

Sometimes when you make a mistake, you learn something important...unfortunately we often feel that the cost of that learning is far too high. Apollo 1, Challenger, Columbia, and numerous others...

Ferrell

Jean Remy said...

Murphy doesn't really warn you though. Murphy is a quiet little guy that visits your workshop and pokes around a little. And when you smile at him, proud of your accomplishments, he takes out a brick and smashes that one experiment you had that could absolutely not in any way possibly go wrong. Then he smiles at you gently and walks out without a word.

Anita said...

Murphy waits for someone to say, "What could possibly go wrong?" or "What's the worst that could happen?"

Murph doesn't smash anything; he finds that one diode buried under the mountain of bells and whistles and 'phsssst'.

As my spousal unit, the auto mechanic, says, "First, remove the license plate ....."

Rick said...

You're right - it is Murphy's disciples who give warning, usually in vain.

Now you have me thinking of Great Engineering Fails. There was reputedly, a class of Civil War river monitors that were heavier than water and sank immediately when launched.

UmbralRaptor said...

That sounds like an urban legend. There were some monitors with absurdly small amounts of freeboard, though.

Anita said...

My favorite Civil War "Ooops" are the twin canons which (eyes to God) would fire chain linked balls simultaneously.

Even I can figure out some of the ways Murphy could mess with that one, starting with the simultaneous thing.

Byron said...

Don't worry about the math. At one point, I became convinced that the Aurek's engine violated thermo, and that the temperature of the oxygen was higher after mixing than the post-fusion helium. I finally figured out that I had used kilograms instead of grams in mass of oxygen, and that it was three orders of magnitude off.
Anyway, so long as you recognize them, those sorts of errors are easily fixed.

Rick said...

Well, they had electric firing, but with Civil War electrical tech ... probably just as well that one stayed on the drawing board.

Kilograms v grams is a fast track to screwy results!

Thucydides said...

A lot of what seems to be "Murphy's law" may have more to do with pushing things past their design limits. (If you have a poor understanding of what these limits are, it is a lot easier to unknowingly slip past them!).

Another factor seem to be based on the human psyche: "It won't happen to me!"

This sort of optomism allows you to overlook things and become sloppy with procedures; you didn't check the tire pressure and nothing happened on your drive; soon you don't bother to check the tire pressure at all and are surprised by the weird noises under the car and the uneven wear pattern on the tread (if you are lucky).

Some organizations like the military, police and fire departments adhere to rigerous checklists of things to do, but complacency happens even there, the only saving grace is there is a chain of command and accountabilitywhen theings do go pear shaped (as well as a pool of trained and skilled personel to fix things and pass on lessons learned).

Of course if the organization is static for too long, then Pournelle's Iron Law kicks in:

"in any bureaucratic organization there will be two kinds of people: those who work to further the actual goals of the organization, and those who work for the organization itself. Examples in education would be teachers who work and sacrifice to teach children, vs. union representative who work to protect any teacher including the most incompetent. The Iron Law states that in all cases, the second type of person will always gain control of the organization, and will always write the rules under which the organization functions."

Under those circumstances, checklists and accountability become notional and Murphy happliy takes the driver's seat.

Rick said...

I'm reminded of a remark a chess player once made to me, that chess clubs tend to be run by the weakest players. The good players just want to play chess.

Though I'd regard it as a tendency, not an 'iron law.'

Albert said...

Bah, if I had a dime for each *epic* fail in simple math or physics like converting m3 to dm3 (I always fail that)....

I think that's the reason why most serious projects are checked multiple times by different people.

Even pros make idiotic mistakes. This makes me feel warm inside. :)

Btw, that Iron Law looks suspiciously like AV:T's Three Generations Law.

-Albert

VonMalcolm said...

Nothing to say (again): I am just 'bookmarking' as to be notified for future comments via Gmail!