Funding cutbacks and an arguably anti-science fiscal policy haven't stopped exciting new projects emerging from NASA's Jet Propulsion Laboratory (JPL) in La Cañada Flintridge, California. We spoke to Dr Michael Sievers from JPL about one such project, cutely named the Optical Testbed and Integration on ISS eXperiment (OpTIIX). …
When a NASA engineer describes a computer as "high performance"...
He probably talks about "good enough for the job". Until they stopped flying shuttles into space, they used a sophisticated AP-101, 480 kHz processor for their flight-control computer.
By now, the specialists collected lots of experience with adaptive optics on the ground, so they'll probably build something that they tried on the ground before. Or maybe they want to compete with that other space telescope and really aim for a 2025 launch date and a few billion in extra cost.
Re: When a NASA engineer describes a computer as "high performance"...
And by "good enough for the job" you mean work consistently and reliably in the harshest of conditions. Even getting a 386 to do exactly what you expect it to do while ignoring the effects of cosmic rays et al is what I'd call high performance.
"When a NASA engineer describes a computer as "high performance" or flight software as "sophisticated", you know they mean it."
my mother thinks operating a VCR is sophisticated.. i guess its all in the eye of the beholder :P
Paris, because she thinks anyone rubbing his goatee and making 'hmmm' sounds while contemplating what position to choose is sophisticated too
Really there are only three bits of software which need to be exhaustively tested on Earth. One is the comms link, so you can always send it commands. One is the reflash code, so you can always reprogram it. And the final one is the hardware protection logic, so that even if the software goes tits-up, it can't drive sensors off the end of their travel or burn out motors/actuators.
With all that in place, they could stop wasting time on pointless unit testing, and get on with actually doing stuff. Unit testing is fine if your requirements never change, and the way you're going to do it never changes, and no-one thinks of a better way of doing something. Chances of this, on a mission with all this new stuff? Near zero as makes no difference.
My opinion? Exhaustively test the critical bits which would break the system, and for the rest just do enough system testing that you have fair confidence it'll work. And then get on with the actual mission.
Re: Testing hell
Erm, how do you know the hardware will solve the problem you are tackling? Reflashing software will not make up for actuators/signal processors that don't have the required control bandwidth.
Did you program any other space projects?
Didn't Fobos-Grunt have some issue with untested situations? Testing and correcting is a big part of NASA's cost plans...
Foolish beauricratic waste of money?
Re: Did you program any other space projects?
It's because they have a very "manned mission" mindset - you can't upload a new crew if you kill them with a bug.
The reason Apollo worked - in spite of the technology at the time - was that everything was tested and rehearsed to a ludicrous degree. But this attitude has continued as religous doctrine in everything they do.
I was involved with one of the original instruments on Hubble. Everything on it had to be "man rated", that made sense as far as not using toxic or hazardous propellants that an astronaut could come into contact with, but it extended to everything. We were restricted in comms chips to very low bandwidth designs that were man rated (ie apollo era) that severely restricted the science data.
One report put the build cost of Hubble at 300% of what it would have cost if had been made to launch on a rocket rather than the shuttle.
Re: Testing hell
But unit testing doesn't check that - system testing checks that. Unit testing checks that under these particular cases on the last month of August in the third phase of the moon, we add one instead of subtracting one.
So what I said. If you're unit testing, that will typically eat somewhere around 30% of your time, maybe more. Wouldn't you be better doing more thorough system testing with that time? Or working with the system to figure out better ways of using it?
Adapting adaptive optics
> OpTIIX is exciting because if the "adaptive optics" experiment is successful, it'll pave the way for much larger, free-flying telescopes in the future
Uhhh, don't almost all ground-based professional telescope already use AO? Haven't they developed it for many years in order to reduce the effect of atmospheric turbulence on their cameras. Why would you need to put this as an "experiment" on a space-based telescope that was above the atmosphere, if you only want to look at stars through it?
However, if you wanted a space-based telescope that was looking downwards through the atmosphere, to spy on people then I can see the benefit - and where the funding would come from.
Now, I appreciate that vibration from all the machinery inside the ISS makes it a no-no for anything that needs precision pointing, but the simple solution to that is (as has been done since the beginning of the space era) is to NOT put your telescopes on a platform that wobbles it's way around the planet. Hubble, JWT and all the other observation satellites managed this decades ago, so it's mystifying why we suddenly need a telescope on what is probably the least suitable orbital location.
I can see that some of the other attributes of this experiment: kit telescopes, are useful. But AO on an extra-atmospheric 'scope? What's it really meant for?
Re: Adapting adaptive optics
Err paranoid much?
Did it every occur to you that the required adaptations for orbital mechanics may be completely different to those compensating for atmospheric interference?
Did it occur to you that this gets around the size requirements for getting big mirrors into orbit?
Did you notice you yourself said "reduce" atmospheric interference - not eliminate.
Did it occur that disregarding AO you can get better images in space.
Did it.....sheesh enough already.
Re: Adapting adaptive optics
We used to use "active optics" for correcting mechanical effects and "adaptive" for correcting atmosphere - but now people use adaptive for both.
Adaptive optics doesn't really help for looking down into the atmosphere - but isn't as necessary. Standing at the bottom of the atmosphere looking up - any small change in refractive index turns into a big angle distance and so the image of a distant object moves a long way. Standing at the top of the atmosphere looking down - any movement of an atmosphere near the object has very little effect
In el'reg terms it's like putting your eye up close to a crinkled bathroom window and looking out vs pressing a body upto the glass and looking at it from a long way off.
Doing something useful with the overpriced moneypit ISS.
Putting people in space for years on end isn't advancing colonisation one iota, and they'll never build a moon base whilst Fujistu, EDS and HP etc are draining the government pots with their overpriced support contracts and we rely on breathing oxygen.
As for bench testing, remembe how much money it took to build Hubble and launch before someone noticed the mirror was wrong BIG OOPS!
So this is about colonization?
There is a lot to do before colonization can be planned. Whether it is colonization of a station, an asteroid, or Mars. They are all many years and technological hurdles away...
- Apple's spamtastic iBeacon retail alerts launch with Frisco FAIL
- Submerged Navy submarine successfully launches drone from missile tubes
- Cache in the Attic El Reg's contraptions confessional no.2: Tablet PC, CRT screen and more
- Developer unleashes bowel-shaking KILLER APP for Google Glass
- Pix Astroboffins spot HOT, YOUNG GIANT where she doesn't belong