NASA - Robotic Repair Of Hubble 'Promising'

mykepredko writes "According to CNN, using a robot to repair/update the Hubble observatory is much more feasible than NASA originally believed. According to the article, the desires for keeping Hubble operational, while keeping shuttle astronauts safe seems to be the impeus for suggesting robotic repair of the satellite. The article goes on to discuss 'Robonaut' and 'Ranger robot', two machines which can approximate the capabilities of a space-suited astronaut. I'm wondering if these robots could be used for the ISS assembly/maintenance, minimizing crew EVAs while maximizing assembly time and hopefully reducing costs."

My question

The question is why didn't the design hubble to be repaired in this way in the first place? The cost launching the space shuttle is around 375 million dollars [psu.edu].. Probably more for a space walk..

I don't accept that you can't design a repair bot for under that launch cost?

Simon

Re:My question

Cost probably wasn't their only constraint. Although the costs involved in developing such a system would be huge, an even tighter constraint would be time -- NASA (well, "the western world") seems to operate on the premise of "Sooner is better"; features may be minimised or cut completely in order to provide a quicker release date. This is fast becoming the trend in software engineering, with 'big players' such as Microsoft starting to cut features and release bug laden products in order to "please" their client with a quick release.

Re:My question

This is understandable, when you understand NASA politics and funding - In NASA, getting stuff done as soon as possible is ideal because if you take another four years, you could end up with an unfunded hunk of half-working metal instead.

NASA is great, but its a bit difficult to run an agency with 20-year projects when everything changes every 4.

Re:My question

The orbit of the ISS was altered after the russians got onboard the project so that they could reach it from their launccomplex - even if it meant that the shuttles could carry somwhat less up there. I don't know if the original planned orbit would have put the ISS in a better position in regard to the issue at hand, ie making possible a shuttleflight that could reach both Hubble and the ISS (which even had a different name back then since it wasn't international)

Re:My question

NASA has plenty of launch vehicles which are cheaper and simpler than the shuttle to get the robot into orbit

Re:My question

While $375 million is nothing to shake a stick at, it's worth noting that the Hubble [wikipedia.org] was launched in April 1990 at a cost of $2 billion US. Robotics, communications, and short-term automated decision-making have progressed signficantly in the last 14 years to make this feasible.

Re:My question

"The cost launching the space shuttle is around 375 million dollars [psu.edu]"

It's not. The average cost of a shuttle flight is actually more like $1,000,000,000. However, pricing shuttle flights is complicated because that's almost entirely due to fixed costs of running the shuttle side of NASA: the variable cost of flying another shuttle once those fixed costs are covered for the year is about $200,000,000.

Re:My question

They DID design the hubble to be repaired. By the space shuttle, no less.

Remember, back in the 1970's when it was designed there was this "Really Great" new technology called the "Space Shuttle" that was supposed to make the cost of getting things into orbit downright cheap. With 100 launches a year, completely reusable, and safe!

The dimensions of the cargo bay on the shuttle were more or less dictated by the hubble.

Re:My question

Because, if they had designed it to be robotically repaired, they would have had to make the design for robotic capabilities of 14 years ago. At that time robotics were so bulky/heavy/difficult to precisely control that sending a robot to repair such a delicate instrument in such a hostile environment was, literally, unthinkable.

Conversely, in another 20 years after the next-gen telescope has been in orbit for 15 years, someone on slashdot will inevitable ask why didn't we design it today to be able to be serviced by the super high power remote laser dohicky from the front lawn of the White House.

Of course this question will immediately be followed by does it run Linux, and In Soviet Russia...

Other uses

Aside from being used to reduce costs by limiting the number of astronaut EVA's on the ISS, I wonder if these robots could be sent to the moon/mars? I am not exactly sure of their entire ability, but surely semi-intelligent robots could be built on their framework that could be subsequently used to build bases and buildings on celestial bodies? It would be much cheaper than sending a human crew to do it, that's for sure.

Re:Other uses

I doubt that these android robots can even stand upright on the surface of any planet. If they were designed for operation in zero gravity their framework is probably to weak to support the full weight.

The article does not mention that, but I think these robots will be remotely controlled from the earth. This is the simplest way to do it in earth orbit. Why bother with a complicated computer intelligence, when you can use a human operator? Of course, this wouldn't work for mars because of the time lag.

Also I wonder what kind of a power supply do they use? If they are really the size of a man, they don't have plenty of room for bulky fuel cells, which means that their autonomous time must be pretty short.

Hubble

The article goes on to discuss 'Robonaut' and 'Ranger robot', two machines which can approximate the capabilities of a space-suited astronaut.

Am I the only one who thought of the Power Rangers when reading this? NASA seems to be mixing work with play...

It would be nice.

It would be nice to have a robotic ( or any ) rescue and refit of Hubble but NASA management will ( I predict ) be against it.

Sad to see NASA go down the tubes by playing it safe.

NASA is not a commercial airline and no one should expect it to have the saftey record of one. I sure as hell don't.

Fix Hubble, then get us back on the moon (just for the hell of it) before I die. OK NASA.

I believe the money would be better spent

on world peace and saving kittens. I mean, stars and stuff don't really do anything, who cares if we can see them or not.

Robots in space?

Robots in space? What's next, a computer that can beat a human in chess?

To Infinity -- And Beyond!

maximizing [ISS] assembly time and hopefully reducing costs

How about if we stop sending rockets and shuttles to the ISS altogether? That would drive costs to zero and drive assembly time to infinity, exceeding both of your stated goals.

Reducing costs?

man, I love the advancements of NASA, but honestly, how does maintaining such an old telescope with expensive maintenance...reduce cost? Especially when cheaper alternatives might be more feasible...or when money is put into reasearching newer, cheaper technology.. Nostalgia shouldn't be an issue here.

Re:Reducing costs?

"how does maintaining such an old telescope with expensive maintenance...reduce cost?"

It doesn't. When Hubble was designed, NASA were claiming that the shuttle would fly fifty times a year and launch payloads for $250 a pound, so repair made sense. Now that it actually flies four or five times a year and payload costs $25,000 a pound, it doesn't make much sense... launching new Hubbles every few years on expendable boosters would probably have been a lot cheaper.

Robotics are the best option in any case

Of course robotic repair makes sense. In fact, it's a better option than the shuttle in any case. There ins't a lot you can't get a robot to do to hubble for the $1.2B a shuttle flight would cost you. NASA's made a science out of trying to prove manned spaceflight makes some kind of sense, and it just doesn't.

Begs the question

This all begs the question - why wait until theres no alternative before coming up with the robot idea? If it saves money anyway, and reduces the need for the shuttle, surely they should have looking into this long ago as an alternative to humans on both hubble and the space station.

Why indeed?

Why send a machine to do a human's work? Or was it the other way around? ;)

This is sad.

How are we supposed to send humans to the Moon and Mars if we are afraid to send them into Low Earth Orbit?

There is evidence that it is actually safer [marssociety.org] to send astronauts to the Hubble than it is to send them to the International Space Station.

I am sure a robot could do the job, but where does it leave humans in the long run if we don't take risks ourselves. Will we leave exploration of the universe to the Von Neumann Machines [wikipedia.org] and maroon ourselves on Earth?

Re:This is sad.

Good point - many people forget that (all, not just space) exploration does (and should!) entail risks.

Sorry, but I have to be nit-picky:

There is evidence that it is actually safer to send astronauts to the Hubble than it is to send them to the International Space Station.

That doesn't matter when the issue you're having is surviving takeoff or landing.

Re:This is sad.

Will we leave exploration of the universe to the Von Neumann Machines and maroon ourselves on Earth?

Does it matter whether the universe is settled by biological von Neumann machines like us, or by mechanical von Neumann machines like our robots, as long as it actually does get settled by somebody? I for one wish our von Neumann successors the very best of luck in their explorations.

Ah, robots used to aid space exploration!

If the robot used to repair Hubble is anything like the ones used to poke around on Mars, does this mean that it will probably take it a month to tighten up a screw?

Consequences.

Assume this works. At least two further questions then need to be asked.

Firstly, if we have built robots that can do anything in space that humans can, what is the point of ISS? Why have a human who requires air, food, sleep, sanitary facilties if Robonaut can do the same thing.

Secondly, are there consequences for the James Webb telescope? This is going to lurk out at L2 and is currently going to be inaccessible for repair or, more significantly, refuel. It is currently being designed with a finite life because of a finite supply of coolant for the IR sensor. Surely the same technology that can repair Hubble can refuel Webb. And Webb is probably being designed with fastenings suitable only for earthside maintainance. Perhaps they should design fasteners to be undone in orbit, even if they don't have the technology to undo those fasteners now. By the time Webb starts running low, about 2016, they probably will have the technology. Wingnuts instead of welds - then Robbie can fix it.

Re:Consequences.

Firstly, if we have built robots that can do anything in space that humans can, what is the point of ISS? Why have a human who requires air, food, sleep, sanitary facilties if Robonaut can do the same thing.

There is very little point to ISS. It's a make-work project for NASA and the Russian space program. About the only thing we have learned from the ISS is that putting humans in LEO for extended periods is a waste of money at present launch costs.

He hasn't made many good decisions, but ending the US commitment to the ISS in 2010 beyond "core complete" is one of Shrub's correct ones. The money could be better spent going to Mars, on unmanned planetary probes, on untold research projects (fusion, a big atom smasher, nanotube research...). Heck, deficit reduction would probably be a more useful thing to do with the money, cause, boy, you guys need it.

robotic exploration, automated

Seven years ago, I remember reading an article in Popular Mechanics [popularmechanics.com] (the article's long gone, unfortunately) about an idea for a completely autonomous robotic system to explore and develop space.

The plan was to construct a simple network of small mining robots that ran on tracks that they themselves laid down. Minerals mined would initially go to the construction of more tracks, track-riding robots, micro-smelters, and power sources (solar or otherwise). In this way, you could construct a self-sufficient mining operation with minimal initial investment that would grow at an almost exponential rate, given sufficient local resources. Land on an asteroid, send minerals and metals out of it a year or two later - avoid the gravity well entirely.

At the time, though, it was just an idea and we didn't have the tech to pull it off. You need some relatively sophisticated AI decision techniques to deal with the nitty-gritty details of such an operation, as we're finding from even such comparatively simple things as the mars rovers today, and it's hard to reproduce the robot-critters on the spot. It's for reasons like the first, though, that I originally got interested in CS and majored in it, and I think we're getting close. Depending on this Hubble work and similar projects, robotics may have finally caught up too.

Instead of worrying about how to get the materials into orbit to build in space, we should start using what's already there. Here's to hoping.

Whats wrong with the shuttle?

I just don't get this whole shuttle thing Nasa are going on about at the moment. It's been flying for decades with respectably few accidents, but now its suddenly too dangerous to go anywhere other then ISS.

At any rate if the only danger is that the heat proof tiles get damaged then why on earth don't they just pack enough supplies to let them hang around in orbit long enough to be rescued?

It just seems really stupid to waste the shuttles just because they're so image conscious that they have to avoid losing astronaughts at all cost, I mean they may as well not go anywhere near space if that's going to be their attitude

Re:Whats wrong with the shuttle?

I just don't get this whole shuttle thing Nasa are going on about at the moment. It's been flying for decades with respectably few accidents, but now its suddenly too dangerous to go anywhere other then ISS.

Actualy, the accident rate on the Shuttle isn't too bad, considering. However, the CAIB inquiry was by far the most in-depth study of the Shuttle, even including the Challenger inquiry. And it revealed lots and lots of potential failure situations that could lead to loss of craft and crew. Once those potential failure scenarios are known about the past safety record doesn't matter. They could happen - that they haven't up until now is luck.

At any rate if the only danger is that the heat proof tiles get damaged then why on earth don't they just pack enough supplies to let them hang around in orbit long enough to be rescued?

It's not the only danger. Firstly it's not just the tiles - there are a lot more components to the thermal protection system on the Shuttle. The component damaged on Columbia was one of the reinforced carbon-carbon wing leading edge panels. Secondly, longevity on orbit is a tradeoff between payload capacity and supplies. You take more supplies, you take less payload. Plus there are some systems that will degrade or run out on orbit and can't be replenished in orbit - thruster fuel is one, if I recall rightly. And thirdly, there's always the possibility that damage to the thermal system might be combined with another fault. Some of the Shuttle's abort modes (like TAL (Transoceanic Abort Landing) and AOA (Abort Once Around)) are required for things like life support problems, and have almost the same heating as a normal reentry. In those situations they can't wait on orbit.

Plus, of course, what happens if they do have to be rescued? It takes a long time to prep a shuttle. In the case of Columbia Atlantis was being prepped and perhaps could have been prepped for a rescue mission in time - but it would have required triple shifts and no problems turning up, plus the assumption that the same thing wouldn't happen on launch. Plus you can't really keep a Shuttle on the pad "ready to go" - again, systems degrade.

It just seems really stupid to waste the shuttles just because they're so image conscious that they have to avoid losing astronaughts at all cost, I mean they may as well not go anywhere near space if that's going to be their attitude

It's all tradeoffs. Nasa's attitude doesn't really matter in this circumstance; it's what the American people - and, let's face it, mainly Congress et al - think that counts, and Nasa are desperate not to have another disaster. Nasa like manned spaceflight, and want to do more of it - they want to get the funding and be allowed to do it, not forced into doing only robotic exploration for the next 50 years,

Re:Whats wrong with the shuttle?

...the accident rate on the Shuttle isn't too bad, considering.
If two out of five 747s exploded, would you call that a bad accident rate? Even considering?

How many Soyuz have we lost in the past thirty years?

Re:Whats wrong with the shuttle?

...the accident rate on the Shuttle isn't too bad, considering.

If two out of five 747s exploded, would you call that a bad accident rate? Even considering?

747 is a bad example. The Comet is a better example. The high losses of Comets was down to one factor; no one knew any better. First versions of anything have high losses.

Considering that the Shuttle was so fundamentally different from ever other spacecraft before it, and how few of those there were, losing two craft in over a hundred flights isn't that bad. The only reason why there were no losses in other US manned craft was down to low flight numbers, and Apollo (somwhere near 15 manned flights including Skylab and Apollo-Soyuz, IIRC - someone will correct me) came close on a couple of occasions.

How many Soyuz have we lost in the past thirty years?

About the same as shuttles, IIRC - two. Over less flights. (Soyuz isn't up to 100 yet, unless I'm misremembering). I think there have been some unmanned losses, and Proton failures (which would have lead to a Soyuz loss if it had been a Soyuz mission) - again, someone else will likely know the exact figures.

Why would the robots be automated ?

What's wrong with manually controlled repair machines ? they would not have to automate anything, it would be dead cheap and it could be launched now.

Assembly Time

minimizing crew EVAs while maximizing assembly time

Wouldn't it make more sense to minimise assembly time so that it gets done faster?

And the little robots names are...

Huey, Dewey, and Louie and you can teach them to play poker after the repairs are done. CLUE=Silent Running.

Robotic repairs of ISS

"I'm wondering if these robots could be used for the ISS assembly/maintenance, minimizing crew EVAs while maximizing assembly time and hopefully reducing costs."

Robotics repairs of ISS is already in the plans, in order to minimize crew EVA time. The SSRMS (Space Station Remote Manipulator System aka Canadarm2) is scheduled to receive a "hand", the SPDM (Special Purpose Dexterous Manipulator). SPDM is designed to be able to carry replacement parts to most external locations of ISS and swap them in place of a defective part. The robotic system is controlled remotely by the crew from inside ISS. The ETA for the launch of SPDM was 2005. That might have changed now that the shuttles are grounded.

For more info on SPDM from the Canadian Space Agency: Dextre (Special Purpose Dexterous Manipulator) [space.gc.ca]

Hubble Future

I remember reading that the reason they wanted to stop service on the Hubble was that they were going to replace it with a much better telescope (which one would assume would be cheaper and better). This being the case I don't see what the fuss is about them not wanting to mess with the hubble anymore. It's obsolete. At least let them keep it in orbit for a few decades while until it becomes feasible to bring it back to Earth. That'd be costly enough but to keep it operational is extreme.

Re:Hubble Future

Hubble is far from obsolete . . . even if the James Webb Telescope was launched today, Hubble can do things James Webb can't and vice versa. One is not a drop in replacement for the other. Hubble's optics and sensors are optimized for shorter wavelength light than the James Webb telescope, so the two are looking at different part of the spectrum. The News Hour has an article here

One key difference between the two telescopes is that the new one will have better instruments for seeing infrared light, which has a longer wavelength and is seen at the far reaches of the universe. Meanwhile, Hubble is better at detecting the shorter wavelengths of light that can be seen with the human eye. Because of these differences between the two telescopes, the NASA panel recommended that the two telescopes' operations overlap so scientists can study both types of images from certain objects.

The Robonaut

Here [nasa.gov] is more information and pictures about the NASA/DARPA Robonaut, a cool humanoid robot proposed to repair Hubble.

I am not a robot engineer, but this look like state-of-the-art to me...

2 ideas

1. Send up a gyro "robot" that simply finds the hubble and grapples it or drills holes and mounts istelf to the side, thereby supplying gyros. Make it so that it has a replaceable uint, that can be sent up on another launch to replace any bad gyros.

2. Move the hubble to the ISS, put it on a long arm that will let it move around but is still attached to the ISS. Possibly use the above #1 idea to attach a rocket pack to get it to the ISS. Then servicing would be a lot easier.

Comments please...

Much less costly.

I think they are beginning to think with their brains for once. A robot doesnt need the same amount of life support system that humans do. you could send the robot up with the russian rockets instead of with the Shuttle/firecracker. Maybe even pick up the US rocket program wich was also better than the shuttle. The cost of the launch itself will then be reduced significantly. Then again maybe its a goal in itself to have humans up there. Much better PR but damn what an expensive PR campaign.

SSL

I worked at the Univrsity of maryland's Space Systems Lab, where the Ranger Robot has been developed. They basically have been working on this robot for the last 13 years, and they have a working prototype in the lab. The prototype is designed to work underwater, since the SSL has a Neutral Bouancy facility, so the difference between the prototype and the actual ranger is that it has no rocket thrusters, only fans. But other than that, the robot is fully functional, and very capable. Back when I workedthere I helped design a test where it would put together parts of the James Webb Space telescope, and it worked perfectly. If any robot is qualified to fix hubble, ranger is.

I will be proven right

Don't you hate how, when things like this are published, and you make some prediction about it and turn out to be right, no one believes you?

Well, thanks to the magic of the interweb, I can preserve my prophecy for future reference!

Thus:

THIS WILL NEVER WORK.

Leave the bot on-station

Build a bot FOR the Hubble, maybe even with its own solar charging station. After that, you just launch the repair parts needed at any particular time in a supply rocket...

NASA giving up

The major premise that the Space Shuttle was built upon was how we could go into space and do all these wonderful things with it.

I seem to remember that being a big repair/utility truck was one of those premises.

Which leads me to wonder just exactly how much one could trust NASA in terms of defining a moon base or a mission to mars scenario?

These are not the droids you're looking for.

An R2 unit is the obvious solution.

Siegfried & Roy

It seems the Cadre's plans are proceeding apace. Bright Sky's ahead, Tova.

Finally a Use for the ISS?

Is there finally a real use for the International Space Station, namely as a control platform for remote cybernetic 'robots'? A person could strap on a control mechanism allowing full freedom of movement of all limbs, without interference from gravitational force at the earths surface.

another Hi-Tech job lost...

Sigh, another high-paying professional job bites the dust.

Wrap It To Go

One of the robots has in interesting concept.
It is designed to work within a space suit.

Why is that interesting?
Because one of the big problems with making equipment work in space is that the environment is so different. You can't just take an industrial robotic arm and bolt it to a remotely-controlled rocket. The oil and grease will evaporate into vacuum and freeze if it is in shadow. Cold shadows can make plastic rigid and fragile, while unfiltered UV breaks down the chemicals.

But by putting a robot inside a spacesuit, the robot is in an Earth environment.

So an industrial robot arm could be wrapped inside a custom spacesuit. It might not last for months, as it would require more heating and cooling than a space-rated arm, but I suspect the spacesuit could be created much more cheaply and quicker.

I've met Robonaut

No, really. While I was interning at JSC, we got a tour of the robotics lab and got to see Robonaut in action. There's a picture floating around out there somewhere where I'm shaking Robonaut's hand. My eyes are closed and I have a really goofy look on my face. Robonaut is bigger than a person, though not by a lot, and has extra degrees of freedom. It can do things like swivel its wrists in a full circle.

While Robonaut is Really Cool, it's not quite as cool as the article implies. It has extremely limited autonomy...they're still teaching it to do things like tell the difference between a wrench and a screwdriver, since machine vision is not a trivial problem, though when I was there they had gotten it to the stage where it was capable of following commands like "Robonaut, get wrench."

It does much better when piloted by a human. The operator puts on a helmet and Power Glove-looking ensemble, and Robonaut will mimic the operator's actions. The operator has to move slowly, however, because Robonaut can't move all that quickly, and if there are too many intervening actions, the program will miss them and it will take the shortest distance between the start condition and end condition, even if there were intervening movements.

I saw it tie a knot in a rope, under operator guidance, and it was able to take a pen from a programmer's hand, hold it correctly (not clenched in a fist, but held between thumb and forefinger) and write with it.

It doesn't have legs. And the head looks like Boba Fett's helmet. When I asked, the explanation I got was that it's designed to look like a Roman Centurion's helmet, but when the designer told me that, he got a really shifty look on his face, so I know the truth.

Anyway...Robonaut linkage [nasa.gov].

-Carolyn

Email I got from savethehubble.org

Below is the text of an email I got from savethehubble.org [savethehubb.org]. I agree with it completely. While I'm in general a big fan of robotic technologies and exploration, it really frustrates me that we've become a country full of risk-averse [chron.com] cowards, unwilling to further our knowledge if it involves even a chance of sacrifice. Never mind the fact that those who would actually be risking their lives are completely willing to take that chance. In any case, if politics keeps us from sending up a manned repair mission, I hope we'll be able to send up a robotic one.

April 24th is National Astronomy Day, and a good time to make waves on Hubble's behalf.

No logic can support the notion that while the Space Shuttle is safe enough for multiple flights to the Space Station over the next decade, it is not safe enough for even one flight to Hubble. It is disingenuous to announce bold plans for a risky manned flight to Mars while at the same time retreating from a flight to Hubble just a few hundred miles away. NASA's leadership should either defend the risk of the loss of life as justifiable given the overall benefits to mankind, or it should retreat from manned missions altogether. We can ill afford to spend another decade funding manned projects such as the Space Station and the trip to Mars, only to have them shelved when NASA realizes it has no appetite for the inherent risk. If the shuttle can not be made safe enough at any cost, then abandon it and the Space Station, and spend more resources developing a robotic solution to fix Hubble, and to launch future scientific missions. The impact of Hubble on society and the enlightening new discovery of water on Mars make it clear that for the foreseeable future there is much more to be gained, in terms of science and political capital, from robotic initiatives (Hubble is an optical robot after all) than from projects that require NASA to make the environment safe enough for a man. Let's get back to manned flights when either we as a people have decided to accept the inevitable loss of life, or at such time as we have designed a space ship that is capable of traveling at near the speed of light. Only then will the benefits outweigh the risks.

Mixed-up on your optimization

minimizing crew EVAs while maximizing assembly time

Why on earth would you want to maximize assembly time? ;)

This could be highly optimal

I like this as a concept. Hell, send a whole armada of these repair bots, and just use something on the lines of VR to repair things. I mean, all this tech from scifi movies might just be something feasible for the purpose, for a fraction of the cost - and the running theory I have is that you only have to send up the fleet once.

The only problem I forsee is that of supplies. Fuel, equipment, parts, that sort of thing. I mean, if you stage everything at another substation out in orbit, that's great, but occasionally you need to send a part up there. Suppose you could send the shuttle or a Soyuz or something up and stock the yard. Hell, if the space plane project ever gets completed, just send it up, pop it out, re-enter - easy peasy. That, and is my idea really that feasible? I'm almost certain that once in a great while you'll have to bring the equipment down from orbit to give it a once over maintenance - or send someone up for the same purpose, thus cutting that cost in half. Remember, much of this stuff that I just came up with will spend the majority of its life in orbit, much like the sattelites littering our exosphere and higher.

If this is the first generation...

Wouldn't it be cool if this would work and then further development of the robonaut could continue? We could have an entire army of these things to build a proper space station in orbit completely ready for people to move in.

Then the robots could be continue to be useful for routine maintenance/repairs around the space station without exposing humans to the dangers of space.

Simple

At the end of the day, seeing into outer space is not absolutely necessary for keeping the nation/world running in the state it currently is.

Although not 100% necessary, it sure is convenient to have an excuse to have a large chunk of your military force in the region with a large chuink of the world's oil supplies. That and we just have to prove we're the biggest, baddest SOB's on the block.

Re:Addendum

The free market needs a customer. NASA is that customer. NASA does diddly squat itself. They contracted out the design and construction of the shuttle and ISS; NASA just drives the truck and puts the tinker toy together on-site.

NASA has more than one supplier and they periodically re-bid the work -- how much more "free market" do you want? Do you think Boeing and Lockheed and the rest wouldn't fall all over themselves to build you a space station of your own if you wanted one and could pay for it?