Time now for the space station to deliver
The launch of a space shuttle is an awesome experience.
First, you see it - an incredibly bright light from its engines. Then you hear it - a low-pitched rumble and a deafening crackling.
But the overriding sensation that stays with you is one that you feel - a vibration that goes right through your body.
With launch to the ISS on Friday, we are now really in the end days of this extraordinary vehicle.
There are just before the orbiter is retired and America has to turn to a new concept - a new vision - for human spaceflight.
A led by former Lockheed Martin chairman and chief executive Norm Augustine is in the process of giving President Obama its assessment of the different options for getting future US astronauts into space.
The new White House incumbent must then decide on a way forward, taking into account the advice of his top scientist John Holdren and new US space agency chief Charles Bolden.
The current Nasa roadmap calls for the creation of two replacement rockets known as and . The first would launch a crewship, Orion. The second, much bigger rocket would be used to lift heavy cargo.
This "space architecture", which goes under the name , would aim to get humans beyond low-Earth orbit again, to go back to the Moon by 2020 and then to Mars someday.
The Augustine committee has held a number of its meetings in public and its discussions have given some remarkable insights into the difficulties that now face the roadmap.
For one thing, it is billions of dollars short of the funding it needs for full implementation.
It also contains a strange paradox: it produces a new crew-carrying capability at just about the same time as the US is currently scheduled to abandon the International Space Station (ISS) in 2016 by ditching it in the Pacific Ocean.
In other words, Nasa would be able to launch its astronauts into space but they wouldn't immediately have anywhere to go.
No-one I speak to in the space business seriously believes that will happen. Indeed, most seem to think that if there is one sure-fire outcome of the review process then it will see the US sign up to an extension of ISS operations, in the first instance to 2020 and then perhaps .
The reasoning that is given: what is the point of spending $100bn to build an enormous science lab in sky only to dump it just as it becomes fully capable?
It is thought that the cumulative lab time devoted to microgravity research in orbit prior to the ISS programme is about six months if you take all short-duration flights together.
Even in its assembly phase, the ISS has surpassed that, and in truth its science endeavours haven't really got going yet.
The station has only just gone from a three-person crew to a six-person crew. That doubling actually translates to a four-fold increase in astronaut time available to science activity.
Friday's shuttle launch also marks something of a watershed. Discovery's mission will be fitting out the station with the last of its major research facilities.
Among the final delivery boxes is the Materials Science Research Rack, which incorporates the European-built . This is really the start of materials science on the ISS.
I've about the advantages of studying the melting and solidification of alloys and other metals in the absence of gravity.
The types of fine-scale structures that form when molten samples cool in weightless conditions can look very different to how they would on Earth, and this can fundamentally change the properties of those materials.
The intention is not to mass-manufacture in orbit but to better understand processes and systems such that when the lessons are applied back on Earth in the large-scale industrial setting, huge benefits accrue.
I spoke with Olivier Minister just prior to the launch of Discovery. Olivier coordinates the materials science experiments on the ISS for the European Space Agency, and I asked him to give me an example of the sort of return that could come from the aluminium research that will initiate MSL operations:
"If we can have a full understanding and control of the casting process of an engine block, for example, we could conceptually direct the solidification process such that the properties are what they should be where they should be.
Ìý
"In the bulk of the engine block, it doesn't matter too much but in the location of the cylinder lining, you would want to have specific microstructures and properties to limit the wearing out of the engine block.
Ìý
"And in terms of a production process: if it takes only 30 seconds instead of 40 seconds to fill the mould of the engine block and solidify it, then at the end of the year on several hundreds of thousands of engine blocks, that's a massive economy."
If the US does indeed extend its commitment to the ISS, it won't of course ease the financial headache it now faces. Money spent on the station is money that cannot be spent on other activities such as developing new rockets.
And if the Augustine committee buys into the ISS, it will narrow somewhat the options the President can pick up.
Mark Uhran, Nasa's associate administrator for the space station, is in no doubt however where American commitment should lie, and he spoke bullishly last weekend about the need to maintain the orbiting platform:
"It is clearly a benefit-cost decision that has to be periodically revisited over the next few years. It's expensive to continue to operate and maintain a spacecraft of this magnitude in orbit, and the benefits have to be worth that cost.
Ìý
"We're confident that once we ramp up this R&D programme, we will have ample justification to continue it. How many years it continues remains to be seen in the success and productivity of the research programme."