Slashdot Mirror
Probe Crash Due to Misdesigned Deceleration Sensor
squirrelhack writes "Seems as though the Genesis spacecraft was able to launch from earth, travel through space, avoid aliens, and cruise back into the atmosphere to be caught by stunt pilots waiting patiently with their helicopters. Alas, the brakes didn't work because a sensor was designed upside down.
Look on the bright side. The craft was not a complete loss, and it was the first probe to successfully test the Interplanetary Superhighway. (Article with pictures) Now that we know the IPSHwy works, we have the capability to launch cargo ANYWHERE in the solar system.
The primary limitation is the maximum weight we can get to the Earth/Moon Lagrange points. Once at the L-points, the cargo pretty much travels one gravity slingshot to the next with nearly no fuel expenditure. This could be a massive boon for sending Interplanetary mission cargo, especially when staging manned missions!
The only down side is that the IPSHwy is simply too slow for manned travel. Not too bad of a tradeoff, however, when you consider the amount of mass that can be more easily staged at Mars in advance! It's certainly reasonable that we could have a complete microsat network at Mars before a human ever sets foot there. Services that could be provided include:
- Mars GPS system
- Deep Space Network Uplink
- Satellite Radio Communicators for landing teams
- Detailed mapping and emergency surveillance of problem areas
In short, we could have a complete technological infrastructure on Mars before we risk anyone's life going there. It wouldn't have to be like the moon mission. We could go to stay.
Javascript + Nintendo DSi = DSiCade
Jonah Hex
Horror & SciFi Erotic Nudes
Seriously. Correct me if I'm wrong, but THEY're the ones who:
Thought we still use Imperial for SPACE WORK (Mars Climate Orbiter IIRC?)
Recently dropped a sat because it wasn't bolted down when they moved it.
Now this.
Can I get like a billion dollars to fail repeatedly? PLEASE?
-- IANAEG - I am not an elder god.
The sensors, which are estimated to be less than an inch (2.5 centimetres) wide, were apparently installed in a circuit board in the wrong orientation - rotated 180 from the correct direction. But the problem stemmed not from the installation but the design, by Lockheed Martin, based in Bethesda, Maryland.
I caught that too. What I don't understand is what was wrong with the design. Is the crash investigation team saying, "Yeah, the sensors were designed wrong, but, huh, check it out, they were installed backwards too, but that doesn't matter" or what? It seems like them (the sensors) being in backwards would be a big deal, but the article seems to imply that the design flaw was the only relevant mistake.
Anybody have any idea what the flaw was or why the sensors would still work when installed backwards?
... turning to the 3-D map, we see an unmistakable con
They installed the switch backwards.
For some reason, I'm reminded of the origins of Murphy's Law. I recall that too was the result of some sensors being installed backwards...
Tedious Bloggy Stuff - hooray?
Have you ever made a mistake that hurt so much you knew you'd never make that mistake again? And when it came around next time, you made so much effort to not make that mistake that you ended up making a completely different mistake?
Mistakes happen, as you say. As is commonly accepted my many software developers, software has bugs.
The parent notes that mistakes happen in even the most expensive projects. I think it's more likely to happen in complex (and therefore expensive) projects.
A while back, one of the main things I admired NASA for was the redundant design concept. You just have a backup path for everything.
But recently it looks like they kind of dropped this concept, at least partially. Probably as a cost-cutting measure. The success of the whole mission now depends on the reliability of several single components, like the sensor in discussion.
BTW, did you know that a Mars Rover has a single CPU that carries out all the computation? I found this puzzling. Today, you add redundance in every piece of equipment - even in web blades.
Don't try to use the force. Do or do not, there is no try.
What did we learn? Um... accelerometers only work in one direction... if you install them backwards, things don't happen right!
Yes. But the real lesson here is that when you are designing something of this sort, don't design it so that it only works one way round. Make sure that it works in both directions, with the output only enabled for the correct direction...
Call me old fashioned, but I like a dump to be as memorable as it is devastating - Bender
Yes, sucess requires failures, but not of this kind!! Imagine if in the early days of cars they had spent millions of dollars researching and designing the latest carburator, then installed it BACKWARD.
The carburator wouldn't work, it would be removed and replaced, and nobody would think anything untoward had happened.
The problem here is that there's no way to test something like this on, say, a half-dozen demo models before it goes out the door. Every single thing has to work right the first time, without ever going through a full test of all systems. The Mars lander, for example -- we'd have known that the legs bounce hard if we'd landed one before, but guess what? We only got one chance!
Considering this unique design parameter -- make it work without the ability to do a full-scale test -- I think NASA's done a heckuva job.
Stressed? Me? Of course not. Stress is what a rubber band feels before it breaks, silly.
This is the NASA equivalent of accidentally filling your car with diesel instead of gasoline.
I did that to a tractor once. Hey, nobody told me it was the one gas-powered tractor on the farm.
I live ze unknown. I love ze unknown. I am ze unknown.