Slashdot Mirror
One Last mission For Deep Space 1
Vertigo01 writes: "Looks like NASA has found a fitting end for Deep Space 1, they're going to fly her THROUGH the coma of a comet to try and take some pictures of the comet's core ... the kicker is that they're doing it with barely any fuel left, and a kludged-together science-camera to replace the toasted navigation system ... kind of a fitting end for her IMO."
This is just the sort of thing we used to expect from the JPL: "We've got fifteen bytes spare and a few milli-amps left in the batteries. We can probably take out the Death Star with that."
What was that old story? With a small amount of memory remaining after all the main programs had been entered, someone at JPL wrote a program to look for and identify previously unknown moons of Jupiter and send pictures back.
dave "wist"
Reminds me of that story about pens in space. NASA found out that pens did not work very well in zero gravity, so they decided to develop a pen that actually did work in zero gravity. After a long time and hundreds of thousands of dollars, they finally presented a pen working in zero gravity. The russians, facing the same problem, used a pencil.
Will work for bandwidth
I mean, you've got to give them credit for refusing to throw in the towel. Ideally, things like Deep 1 wouldn't malfunction in the first place, but at least NASA is trying to make the most of things.
http://nmp.jpl.nasa.gov/ds1/
Check out the monthly reports. They are quite fun to read, because they are written in a "layman" fashion. Especially the parts where they are putting together the "using science camera for navigation"-kludge. And rebooting a system half a solar system away and hoping it comes up again after an OS upgrade.
It's kinda sad that all the public focus is on the Mars missions, when there's stuff like DS1, Galileo, and NEAR that just keep on going..
If for $12 million they can show that the special effects in the $140 million movie "Armageddon" were crap, I say it's a great value.
I didn't know McGyver worked at NASA.
It was a good, clean and funny TV show that has made me want to become an engineer.
Oh yeah, we've really been programming things lately that don't fail. OK, maybe I'm a little harsh -- I'm certain I couldn't do any better. But I am curious as to how much of the $12MM would have to be allocated to other projects if we chose not to include this last flyby. What I mean is, those people that are getting paid would probably still have jobs, but the costs would be allocated to other projects. So how much of this cost is truly related to this final song?
GreyPoopon
--
Why is it I can write insightful comments but can't come up with a clever signature?
I'm sure there are plenty of ways to justify the $12m bill - the ground operation will already be fairly immense, and keeping it running costs money (admin staff, office space, hefty electricity bills etc). Mainly, I expect it goes to consultants and contractors, and on purchasing hardware from "military approved" vendors (i.e. the expensive ones).
In this country (UK), a post-doctoral space scientist at a top academic institution probably earns around 20,000 pounds PA if they're lucky (that's about $30,000, I think). If you consider that they'll be employed for maybe three years doing the data analysis and planning the next comet missions, a team of 20 scientists would account for $1.8m.
So I can see (almost) where they get the figure from. And it's probably quite easy to convince the funding bodies (is that the US public? I don't know how NASA do things) that comets are already lining up to take aim at the Earth, and we must learn more about them so we can work out a defence mechanism... sounds insane, but that's how a lot of science gets funded nowadays.
Me? I'd rather they built some hospitals in Africa.
These sigs are more interesting tha
The space pen's development was funded by a private company.
Mea navis aericumbens anguillis abundat
For goodness sake, will people stop posting this trolling story? As has been said before this is misleading.
For the first few missions, the Soviets did use pencils. Then the Soviets went to Fisher (the American company that made the pens) and bought several cases. The reason is that pencils produce a lot of graphite dust. When you are locked in a room the size of a telephone booth for a week, you don't want graphite dust floating around, getting into your lungs, eyes and your equipment.
Slashdot monitor for your Mozilla sidebar or Active Desktop.
First of all, the small mass of the satellite could only slightly alter the orbit of the comet. There is no way that it could impart enough energy to take the comet out of solar orbit and send it towards a distant star.
Secondly, since comets are made of "dirty ice", imparting enough energy to take the comet out of solar orbit via a single impact would almost certainly vaporize the comet rather than moving it.
Thirdly, even if you could nudge a comet out of solar orbit, since the average speed of a comet is on the order of 1/6000 the speed of light or less (about 100,000 mph), it would take about 25,000 years for the comet to get to the nearest star (Alpha Centauri A), let alone "distant stars".
So no, it wouldn't be a cool way to contact alien civilizations.
In order to communicate the probe you need to rent time on the Deep Space Network. This network is currently running at capacity, so getting time on it is rather expensive.
But an even bigger expense is the mission software. Modifications to the programming of the probe need to be codded. Then the code has to be proved to be mathematically perfect. You cannot afford to compile it, upload it, and get a message back saying "stack overflow, press any key to continue". The software must be proven to be 100% bug free before it goes up.
It takes a lot of people to manage a space mission correctly. Cut corners, and your mission fails because of something stupid (e.g. metric vs imperial).
Slashdot monitor for your Mozilla sidebar or Active Desktop.
Hi! I've just sent my baby for conformal coating today. It's the flight spare model of the power supply units to power the ccds of two cameras and it took me about 2 years to design completely (as a flight unit). The experiment is called Osiris, also here, the satellite is ESA's Rosetta and its target is some comet named Wirtanen. I'm quite happy to see my first piece of flight hardware already being integrated with the full satellite and hope the Osiris will give us some nice pictures one day.
Cheers,
Alejandro
Instituto Nacional de Técnica Aeroespacial - INTA
One of the important things to remember is that just like Apollo 13, these guys are where they are for coming up with innovative fixes to tough problems. This is just another great rehearsal for a situation that could just as easily come up with human life at stake. This is why these guys are kept on the project long after the system gets put into cruise mode. It's just another case of "I've done so much with so little for so long that now I'm attempting the impossible with nothing." You have to push the boundries to find where they are in practice. Also, real problems are far more challenging than anything they might have considered in simulation.
The costs stated are the continuing costs of tracking the satellite versus turning a deaf ear and ignoring it. Deep Space Net time isn't free, they have to allocate it and maintain the dishes used.
When this is done, continue past the nucleus into the comp and try to get a closeup picture or two of the coma before being blown to smithereens by the particles.
It's the continuing cost of tracking that add up to the $12 million versus just turning it off.
Then the code has to be proved to be mathematically perfect.
Does the proof itself need to be prooved to be correct, or is that taken on faith?
The problem NASA had with the Metric vs Imerial calculations was due to rounding errors in the conversion equations (ie: only going out x amount of decimals points; where x wasn't large enough). The error introduced by the lack of precision wasn't due to a single conversion but due to multiple back and forth conversions (probably in the order more than a hundred). It was not a single incident of "oops, I meant five meters, not five feet."
This doesn't justify it, but I don't think a lot of people actually know what the real problem was. It was a precision error, not a Metric vs Imperial error.
All editorial writers ever do is come down from the hill after the battle is over and shoot the wounded.
This mission is somewhat similar to that of the Giotto probe in 1986. Here is the link to the ESA site with more information about Giotto. But where Giotto was a dedicated mission, designed to take pictures and collect data of a comet core, the DS1 comet mission is "just" a great bonus mission.
[--- PGP key and more on http://www.root42.de ---]
A comet gives off gas which is fairly tenious and gets blown away more or less directly away from the sun. Its the dust and grit that is the risk, and that gets left behind in the vicinity of the orbital path, in the same region of space where DS1 will be. (same stuff also causes meteor showers when earth orbits accross the dirt trail)
The important thing from DS1's point of view is to keep the relative motion between the coment and the probe as small as possible, both to maximise encounter time and to make it easier to 'aim' the probe and its cameras at the comet. (this also saves fuel, which is a heavy, scarce and precious resourse in outer space)
In effect, the two objects are on almost on a parallel path, at slightly different speeds, not a perpendicular intersection as one would think.
Its like two veichicles on a slowly curving highway, one slowly overtaking the other. If the comet is an open dumpster truck in the slow lane, you will be showered with garbage for miles before you eventually pass it out! (even though you are only 'alongside' it for a few seconds)
-- We don't understand software, and sometimes we don't understand hardware, but we can *see* the blinking lights
"There's a very real chance that none of this is going to work"
Yeah...there's a chance that something nasa does (think polar lander) isn't gonna work...what is he, some kind of rocket scientist? =)
Juiced? Or Not?
It's taken on faith--you have to assume some axioms.
Axioms aren't in question; it's whether all of the steps of the proof are correct. It seems to me that it's just as easy to make a mistake in a proof as it is in a program.
There's also probably some magic bean counter stuff going on as well. The project likely gets assigned its "share" of a lot of things like the cost of the building, electricity, phones, the coffee machine and the salaries of people who would be doing the same job regardless: janitors, security, support staff, etc.
Weird, but it's a common accounting practice.
One line blog. I hear that they're called Twitters now.
If you make a mistake in a proof, then you will go over both the proof and the program, which means that you will discover that the proof is faulty, and correct the proof. If you have no proof at all, then you can't say with 100% certainty whether or not the program works (unless it's "Hello world").
Formal proofs of programs increase the probability of noticing a mistake since you're essentially implementing the program twice, but they don't guarantee 100% certainty of correctness, since there is always the possibility of an error in the proof. Computer-system theory is littered with published papers containing incorrect proofs.
It all boils down to redundancy, similar to N-version programming, where you implement N (N>=3) version of a program (implementation techniques should differ as much as possible).
Triple-modular redundancy also has difficulty when applied to software systems because systems have some parts are easy and some that are hard, and the implementors of all three systems are most likely to make most of their mistakes in the harder parts.
If only software systems were as trivial to build as bridges and airplanes!
{now if you'll pardon me, I have a flight to catch...}
The point of the Deep Space series of missions, of which the Mars Polar Lander was #2 and went AWOL, was to test new tech for next to nothing (in NASA terms).
Do any of you realise that DS1, apart from being 8 revs away from the greatest Trek ever, was powered by an ion engine? You know, like Star Wars?
Plus, when the nav system went tits up, they were able to retask other optical instruments to allow for autonomous piloting.
DS1 wasn't even supposed to make it this far. IIRC, it was expected to have a 3-month primary mission to test the equipment. Then, if there was enough gas in the tank and the thing still worked, they were going to find something else for it.
An asteroid flyby and now a comet encounter...not bad for $12 mil!
P.S. I'm a bit biased on this one - I watched the launch and have read every one of the oddball logs posted by Dr. Raymond.
C'mon, NASA, where's DS3?
GTRacer
- Wants to be first at something
Defending IP by destroying access to it? That makes sense, RIAA/MPAA. Go to the corner until you can play nice!
This is offtopic, but I wish you luck on that flight...
Cancelled.
Dan Bernstein. ;-)
Mind you, Patrick Moore was on fine form, that night, and was able to turn some fuzzy low-res photots of two vents into something dramatic & well-worth the watch.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
I always thought that the Deep Space series was for labeling probes testing experimental technology. The Deep Space 2 probes on Mars Polar Lander were part of that because they were designed to survive impact after being just DROPPED from MPL during descent, penetrate the surface, take samples, and beam back data. MPL itself was not really part of the Deep Space series.
;).
The whole concept that MPL and the other Mars probes (and maybe... probably... the deep space probes) fall under has generally been refered to (at least in the press) as "better, cheaper, faster". Prevailing wisdom seems to suggest the first two parts of that are mutually exclusive given the Mars failures. But there are some big successes there too, and perhaps even a 50% casualty rate is better than quadrupling the cost. Hopefully they'll build in more redundancy
Regardless, I'm just parsing semantics...
- StaticLimit