Slashdot Mirror
NASA Preparing Manned Hubble Service Mission
danimrich writes "According to an article at Space.com,
'NASA's new Administrator Mike Griffin told reporters today [April 29] that he informed key members of Congress Thursday evening that he would direct engineers at Goddard Spaceflight center to start preparing for a space shuttle servicing mission to the Hubble Space Telescope on the assumption that one ultimately will go forward.'"
That would depend entirely on your location. If you car was in the back yard, in the middle of a densly populated area, just down the street from <insert major chain store name here>, probably not. It would be very cost effective to walk to the store, buy the parts, then fix the car.
OTOH, if your car is located at a research camp, on the icecap at one of the poles, far enough away from 'civilization' that the _only_ way to bring in those spare parts is to fly a ski equipped C-130 3000 miles to deliver the parts, you will rethink the whole thing. The cost of transportation far exceeds the cost of the equipment being transported, by a couple of orders of magnitude. If the C-130 is going to be sent anyways, it may well be more efficient to just load up a new car in the back, and deliver that.
If one goes on the assumption there is budget for a shuttle trip, then the real question _should be_, what is the appropriate payload to carry? Should it be carrying spare parts for the existing old hubble, or should it be carrying a brand new telescope of some kind. In either case, the 500 million launch budget will be used.
In the case of hubble, pork politics, and budget line items get in the way. It's really silly, because the arguement to decide if the old one is fixed, or a new one is launched, has nothing to do with final cost, and everything to do with 'which budget does it come from?'. Launching a new modern replacement would entail creating a new mission line item in the budget, a process that's not likely to happen. Fixing the old one would shift funds into an existing line item, a process that may well be able to be pushed thru. The amount of funds in each case doesn't even enter the equation, it's all about what can be achieved politically.
Dropping 500 million into an existing line item is possible, but creating a new line item instead, with a value of 300 million, not gonna happen. That's how the 'efficiency' of a beaurocracy works, in particular one that's designed to entice voters with financial mumbo-jumbo. Joe congress-critter knows it's cheaper to fix an old car, than to buy a new one, so it's _gotta_ be cheaper to fix hubble than to launch a new telescope.
The real problem with a system that works this way, it's so damn full of pork. When you sit back and ask 'wheres the beef?', you'll discover, the politicians live an a diet of pork. The congress critters have become so adept at slicing and dicing pork for serving to the constituents, dont think they even remember how to throw some beef on the grille and serve up a steak.
I've posted about this topic before (here: http://science.slashdot.org/comments.pl?sid=146007 &cid=12230905, and here: http://science.slashdot.org/comments.pl?sid=146007 &cid=12232506)
There are several important factors in deciding between them. Lets look at the pros and cons.
Cost:
1. Shuttle servicing will cost about $300M to fly the mission plus ~$1.5B-2B to keep the shuttle program and staff going for an extra 4-6 months. Total cost then is conservatively $2.3B.
2. Robotic Deorbit Only is estimated to cost about $850M, for development, launch, and operation of the vehicle.
3. Robotic servicing is expected to cost $1.4B for dev, launch, and operation through splashdown.
However(!) if we take option 1 or 2, we'll have to fly a 'robotic proving' mission around 2015 or so to enable missions to Moon and Mars. This could cost anywhere from $500-800M (likely closer to 800 if it's to be at all ambitious). So lets look at the total score-card:
Shuttle: $2.3B + $800M = $3.1 Billion
Deorbit: $850M + $800M = $1.65 Billion
Robotic: $1.4B - ~250M already spent = $1.15 Billion
So that was cost. Now lets look at education:
Doing another shuttle servicing mission will teach us very little. Sure, we'd learn some EVA techniques, management techniques, things like that. But nothing significant. That's why we'd need to launch a robotic proving mission in 2015.
Robotic Deorbit would teach us a lot about autonomous rendezvous (since my last post it's apparent that we need to work a little harder on that; DART bumped into its target, I hear). Bear in mind that craft had no forward-link commanding from the ground... it was entirely autonomous. It cost only $100-something million to dev, launch and "operate". These are lessons we need to learn to go to the Moon, and Mars.
Robotic Servicing would teach us a lot about the autonomous rendezvous and proximity operations (see above) since it's the same problem here as the robotic deorbit. It will also teach us a HUGE amount about ground-to-space tele-robotic operations. So much so that if it works we could be confident enough not to need an expensive proving mission later on. We'll be doing complex robotic tasks on things that were designed for humans (on space-station, everything's designed to be robot-friendly). We'll be pushing the envelope of our knowledge.
Don't let that put you off though. We're pushing the envelope on the ground here, right now. We've pushed it so far now that most tasks on the Hubble robotic mission will be trivial. We aim to push it far enough that ALL tasks will be trivial (or at most 'complex') by the time we launch. We have a robust capacity to re-plan and re-approach a problem on orbit. We have the advantage of time (see next pro/con) on our side. And we have contingency in case some more critical item fails before we launch. I believe that up to 30 days before launch we have the ability to re-manifest the cargo. Don't quote me on that figure though.
Now lets look at perhaps the most important feature of each mission: The quality of the result:
Some say a shuttle servicing mission will do a better job at servicing Hubble. This used to be the case. In looking at the robotic mission we had to give up some things. The STIS failed last summer, as some of you may remember. The robotics guys evaluated that task, and decided it would be too difficult. Many bolts in hard-to-reach places, etc. So that was dropped. However, I've recently heard on the wind that a Shuttle mission will only have a few days of EVA available between tile inspection and prep for landing. The shuttle mission will be forced to leave things out too, and the result is that the priorities we identified for the robotic mission are pretty much the same priorities we'd have for the sh