Slashdot Mirror
Robot Helps NASA Refocus On Hubble
The ailing Hubble telescope keeps refusing to die; jdoire points out this story at the Washington Post which reads in part "Largely because of the Canadian robot named 'Dextre,' NASA has gone in less than a year from virtually writing off the Hubble to embracing a mission that will cost between $1 billion and $1.6 billion and approach in complexity the hardest jobs the agency has ever undertaken." (We last mentioned Dextre back in August.)
Actually, it would be Dextre in REAL english. Just like in actual english, it's centre, not center.
Robot Helps NASA Refocus on Hubble
Written-Off Mission to Extend Telescope's Life Is Revived Because of 'Dextre'
By Guy Gugliotta
Washington Post Staff Writer
Friday, November 12, 2004; Page A03
The promotional video shows a multi-jointed titanium handyman untwisting knobs and disconnecting an electrical cable with slow-motion aplomb, displaying fine motor skills that the voice-over assures will enable it to install "new batteries, gyroscopes and scientific instruments" aboard the aging Hubble Space Telescope.
But the video is only a teaser. In April, when NASA's Goddard Space Flight Center in Greenbelt showed the whole sequence to headquarters VIPs, what had first seemed an elusive dream -- a robotic mission to service Hubble and extend its life by five years or more -- suddenly became real.
"I remember coming to look at this stuff and asking, 'Is that an [animation]?' And somebody said, 'No, it's really happening,' " recalled Edward J. Weiler, who was NASA's associate administrator for space science at the time and is now Goddard's director. "I didn't think robots could do this kind of stuff."
It is by no means a sure thing. Yet largely because of the Canadian robot named "Dextre," NASA has gone in less than a year from virtually writing off the Hubble to embracing a mission that will cost between $1 billion and $1.6 billion and approach in complexity the hardest jobs the agency has ever undertaken.
"Almost as difficult as landing on Mars successfully twice," Weiler called it. Servicing the Hubble, like the nine-month tour de force that has kept two rovers tooling around the Martian countryside, will demand a host of technical tasks and tricks that have never been tried.
To do it, the United States must develop its first-ever robotic docking vehicle, fill a bag with tools that, in many cases, have not been invented, and use the robot repairman to unscrew j-hooks, open and shut doors and "drawers," disconnect and attach electric connectors, and rig jumper cables.
By the end of 2007, NASA hopes to put into orbit its Hubble Robotic Vehicle of four components: a de-orbit module designed to dock with Hubble; a grappling arm to seize the telescope during docking and serve as a repair platform; an ejection module to carry spare parts and tools; and Dextre.
The jobs, in descending order of importance, are to change Hubble's batteries; install new gyroscopes; swap an old camera for a new, more sophisticated one; install a new spectrograph; and, if possible, replace a telescope pointing device and repair another spectrograph.
"There's nothing easy about it. It's all firsts," said Goddard's Preston M. Burch, Hubble's program manager. "And some of the things we're thinking about make people nervous." The fundamental tenet for a servicing mission, he noted, is the same one that doctors espouse: "Above all, do no harm."
In the past, shuttle astronauts had the job of servicing Hubble, missions that required a few days of spacewalks lasting six hours each. Dextre "can work 24-7," Weiler said -- a fortunate feature, because robots are not as supple as humans. "Watching it is like watching grass grow," Weiler said.
Burch hopes to complete the mission in a month. Some of it will be done by the robot working on its own, but most will be handled by ground controllers manipulating the robot's two arms -- like playing a video game.
"Astronauts are keen to do this," Burch said, and they will probably get the call because of their experience and knowledge of the perils inherent in handling large objects in space -- where something pushed or pulled does not slow down until it is checked.
"Hey, if they ask me, I would be very happy to do this," said Michael Massimino, an astronaut who serviced the Hubble in 2002 and has joysticked Dextre in the lab. "It's an interesting and challenging project -- it's cool, really cool."
Dextre, so nicknamed by the Canadian Space Agency, was developed by MD Robotics, of Brampton, Ontar
I don't read your sig, why do you read mine?
Interesting article! However, I am a bit surprised the article doesn't mention that a replacement to the Hubble is already planned: The James Webb telescope. The only thing that one doesn't have and the Hubble does is a UV viewer (which can't be done on earth either due to the ozone layer). But apart from that it is a replacement for Hubble.
The new telescope will be mostly infrared and infrared seems to also be blocked by the Earth's atmosphere (althrough not as badly as UV). The new telescope seems to be aimed at catching what the Hubble couldn't catch, as the Hubble seems to be only near-infrared.
I thought that I heard that James Webb telescope is primarily infrared. If so, its not at all the same thing as Hubble, although it has virtues of its own, it doesnt give us the same visible light capabilities as Hubble. I heard Webb does have some visible capability but it doesnt sound like it has as much as Hubble. It has an infrared viewer, which is quite different and will allow us to see through clouds of dust to say the centers of galaxies better, but it may not provide the same quality visible light photographs we are used to with Hubble. It seems like something like Hubble and Webb would compliment each other, since they each photograph different parts of the spectrum better.
I am an astronomer, and I've worked on AO systems, and I can tell you that I'd rather have Hubble than all the AO-corrected Kecks in the world. AO sounds like a good idea, but in the end the data you get out is hard to calibrate, and unreliable. The problem is that the properties of the atmospheric turbulence keep changing, making it hard for the AO to keep up. The best AO-systems available today achieve maybe 70% of the performance of a diffraction limited system such as a telescope in space. But the remaining 30% of the light goes into a big "halo" that has all sorts of complicated image structure in it.
Then there is the fact that the field of view that you get with an AO system is much, much smaller that you'd get with Hubble. And then there are issues with higher thermal background, etc. A while back HST published a light-curve of an eclipsing extra-solar planet - something like that could never be done from the ground (i.e. with the same precision).
The University of Arizona's $120 million Large Binocular Telescope is the world's most powerful optical telescope, with images about 10 times as sharp as the Hubble's.
No, not really. LBT will not produce sharp images in the visible, at least not with any AO system that one could build today. In the near-Infrared LBT will still be subject to all the disadvantages inherent in AO systems, and in addition will have the problems associated with interferometry, since it is actually two telescopes cobbled together to act as one. LBT will, if it ever works, and press-releases notwithstanding, not be quite the Hubble-killer it's sometimes made out to be.
Human genome = 3 billion base pairs = 6 GBit. Windows + Office = 20 Gbit. Which is more impressive?
How do we justify a $1.6B repair for a device that originally cost $1.5B. Seems we could design and launch a much improved model for the same amount.
Current Hubble instruments:
Initial James Webb Telescope instruments:
The only real overlap is in Near-Infrared. It's important that the Hubble be saved, as the Webb telescope has virtually no non-IR capabilities.
You can never go home again... but I guess you can shop there.