Spirit

[DavRedf]

It seems the spirit is talking to NASA again. I told them not to upset it, it sulks if they do.
At the last report it had found evidence that there was water on Mars.
But what really got them thinking was the vodka bottles they found behind a rock.

Seriously, nice going guys keep it up.

David

 

[Rocketmaniac]

I believe that Spirit has been going for 360 days now. And Opportunity has been going for 346 days.

Both are like the "Energiser bunny" they keep going and going and going...... And I thought they were suppost to have a life-span of 90 days...... I mean I am not complaining, but how did the engineers predict 90 days and the rovers last for more than 4 times that amount?

 

[sandman]

but how did the engineers predict 90 days and the rovers last for more than 4 times that amount?

See, that was the trouble with a slide rule...you have to figure out where the decimal point goes.

 

[Habu]

Was that figured in Martian days...?

 

[illini]

Originally posted by Rocketmaniac
...but how did the engineers predict 90 days and the rovers last for more than 4 times that amount?

As an engineer, I've noticed that pi is a suitable multiplier to apply wherever guesses and estimates are concerned. Some evidence:

A old professor of mine used to take the tests he was about to give his students, then multiply the amount of time it took him by pi in order to estimate the time it would take the students. Worked every time.

I was doing a study once about 10 years ago where cost estimates were required. As the study manager went through the numbers (plucked from thin air in most cases), I noticed the trend that non-recurring engineering costs were always about pi times the acquisition costs. When I pointed this out to him, he stopped making up numbers and started just multiplying things by pi. Strangely enough, a large DoD acquisition program wound up budgeting procurement based on those numbers! This trend has held up repeatedly over several other studies I've been involved in.

Whenever waiting for funding to arrive from one of my customers, I always multiply their time estimate by pi. If that customer is the government, I multiply by pi squared. Works like a charm. A coincidence? I think not.

I could go on and on, but you get the idea. The answer, it seems, is always pi!

 

[DynaSoar]

Originally posted by Rocketmaniac
I believe that Spirit has been going for 360 days now. And Opportunity has been going for 346 days.

Both are like the "Energiser bunny" they keep going and going and going...... And I thought they were suppost to have a life-span of 90 days...... I mean I am not complaining, but how did the engineers predict 90 days and the rovers last for more than 4 times that amount?

They were asked to build it to a successful lifetime parameter and did so. They wouldn't promise it for more. If it happened to work better than they thought, that's fine, but they couldn't plan the design better because the design/build/test/redesign cycle for improvements is expensive. They were told to build a sufficiency, not a perfection. However, some of the factors were actually well in excess of sufficiency. They just weren't advertised as such.

I got that from a good source. My younger son was working on it at JPL and redesigned the transmission. It was much better and it was lighter. He was told not to expect a good reference when he left because he didn't work within the project parameters. Had he just got the design incorporated, everything would have been fine. But he went the official route and tried to present the design to the administration to get it approved.

Didn't bother him. He got a job at National Ignition Facility after that, before moving on to buckytube engineering. And he kept the design. He wants to use it on a Battlebot.

 

[cls]

pi is a suitable multiplier to apply wherever guesses and estimates are concerned

that's fine for engineering estimates. but for software development time estimates, what seems to work is to add 10% to the engineer's estimate, then double it, then move to the next higher unit (week, month, year).

for example, "I can knock out that code in a day" becomes two weeks and a day; a two week project takes 10 months; etc.

 

[tbzep]

Probably a prediction based on continual exposure to worst possible conditions. If they had predicted based on average conditions and ended up in a mess of bad weather, extra cold, higher than normal radiation, etc. they were guaranteed to be persecuted for overestimating their machinery.

Predicting the short end and having it exceed everyone else's expectations makes them look like they built superbots.