Panel says keep options open for Hubble repair
BY WILLIAM HARWOOD
STORY WRITTEN FOR CBS NEWS "SPACE PLACE" & USED WITH PERMISSION
Posted: July 13, 2004
Editor's Note: Portions of this story were written by the author for Astronomy Now magazine.
In an interim report, a National Academy of Sciences panel today recommended that NASA leave open the option of launching a shuttle mission to repair and upgrade the Hubble Space Telescope while engineers assess the viability of a complex robotic servicing flight.
Saying recommendations in the wake of the Columbia disaster did not preclude a shuttle visit, the panel urged NASA to commit to a servicing mission, whether robotic or manned, "that accomplishes the objectives of the originally planned (shuttle) mission," including installation of two already-built scientific instruments and replacement of aging batteries and gyroscopes.
The NAS Committee on the Assessment of Options for Extending the Life of the Hubble Space Telescope also raised questions about the complexity of a proposed robotic mission.
"Because of inherent uncertainties in the early stages of development of a robotic mission to the Hubble Space Telescope, as well as the uncertain current status of the shuttle return-to-flight program, the key technical decision points for committing to a specific service scenario are at least a year in the future," the NAS committee wrote in one of three findings.
Then, in a recommendation, the panel said "at the same time that NASA is vigorously pursuing development of robotic servicing capabilities, and until the agency has completed a more comprehensive examination of the engineering and technology issues, including risk assessments related to both robotic and human servicing options, NASA should take no actions that would preclude a space shuttle servicing mission to the Hubble Space Telescope."
NASA Administrator Sean O'Keefe stunned the astronomical community - and touched a nerve with the American public - when he announced in January that he had decided to cancel a fourth and final planned shuttle servicing mission to Hubble.
The decision was revealed just two days after President Bush announced his new moon-Mars initiative, calling for establishment of permanent lunar bases and eventual manned flights to Mars.
As part of that initiative, the president ordered NASA to retire the space shuttle by 2010, or as soon as space station assembly is complete, and to develop a new manned spacecraft that could carry astronauts to and from low-Earth orbit and, in some form, on to the moon.
O'Keefe said the timing of the Hubble decision was coincidental and that he was motivated solely by concerns about shuttle safety and by his commitment to implementing the recommendations of the Columbia Accident Investigation Board.
Hubble Servicing Mission 4, or SM-4, was the only flight left on the shuttle schedule that was not bound for the space station, where the crew of a crippled orbiter could attempt repairs or await rescue. The Columbia accident board told NASA last year that it had to develop stand-alone repair techniques for non-station missions.
The cost and complexity of that effort, coupled with the Bush administration's directive to complete station assembly as soon as possible, left O'Keefe with little choice, according to astronaut John Grunsfeld, NASA's chief scientist and a veteran of two Hubble repair missions.
"When you look at what it would take to support this one single-use unique mission ... in light of retiring the shuttle, it kind of pushed the decision over the edge and allowed the administrator to make that decision," he said.
Even so, newspaper editorials lambasted O'Keefe and Sen. Barbara Mikulski, a Democrat whose district includes the Space Telescope Science Institute in Baltimore and the Goddard Space Flight Center in Greenbelt, Md., demanded an independent assessment of the risk associated with a shuttle servicing mission. O'Keefe then asked the National Academy of Sciences to look into the matter.
But time is short and the pressure to make a decision, one way or the other is mounting. Left on its own, Hubble will lose its gyroscopic pointing stability in 2007 or 2008, bringing science operations to an end. Battery power is expected to drop below the minimum threshold needed to sustain life shortly thereafter.
Addressing O'Keefe, the NAS committee issued an interim report today "because you and your NASA colleagues have made clear to the committee that there is some urgency in issuing any recommendations related to Hubble." The aim of the interim report "is to provide useful guidance to NASA that can be utilized during the time that the committee (as well as NASA) continues to investigate the servicing options in greater detail."
The panel promised to deliver a final report by late summer or early fall. But Mikulski was elated with the interim recommendations.
"The preliminary assessment of the National Academy of Sciences is enormously encouraging," she said in a statement. "Hubble is the greatest scientific instrument since Galileo's telescope. It should not be abandoned without a thorough, independent and rigorous review by the experts. That's what they panel is doing, I thank them for their service."
Responding to the NAS committee, O'Keefe said in a statement, "We agree with the Committee's view that the Hubble Space Telescope is arguably the most important telescope in history." But the only "commitment" he made was to "exploring ways to safely extend the useful scientific life of Hubble."
"The challenges of a robotic mission are under examination and we'll continue our exhaustive and aggressive efforts to assess innovative servicing options," O'Keefe said. "In parallel with the Committee's ongoing research and deliberations, NASA will evaluate proposals we expect to receive shortly."
In his only reference to the committee's recommendation about doing nothing to preclude a shuttle visit, the administrator said "along the way, we'll keep options open to assure the best possible outcome."
Engineers at NASA's Goddard Space Flight Center in Greenbelt, Md., are hard at work developing a daring plan to repair and upgrade Hubble using a remotely-controlled robot instead of astronauts.
Equipped with a sophisticated robot arm, stereo television eyes and the equivalent of smart bomb targeting technology, a 20,000-pound two-piece spacecraft would be launched by a heavy lift booster in late 2007.
After attaching itself to the aft end of Hubble, the spacecraft's robot arm would run cabling to tap into the telescope's solar array power system and install a new camera equipped with six new gyroscopes to upgrade the observatory's guidance system.
The arm also would be used to install a new spectrograph, the most sensitive such instrument ever launched. The lower section of the servicing spacecraft, carrying the robot arm and the discarded instruments, then would be jettisoned to fall back into the atmosphere.
The smaller upper section, carrying six new batteries and a propulsion system, would remain attached for the duration of Hubble's life. Engineers believe the batteries and gyroscopes could extend the observatory's life by five years or more, keeping Hubble scientifically productive until its replacement - the James Web Space Telescope - is launched in 2012.
When the new equipment finally fails, or when funding runs out as NASA transitions to JWST operations, the propulsion system would be used to drive Hubble safely into the atmosphere over the Pacific Ocean where its wreckage would pose no danger to the public.
"There are enormous challenges to be faced and no guarantee of success," O'Keefe told astronomers at the summer meeting of the American Astronomical Society. "We are not yet at a point where we have a firm alternative, but we're getting pretty close."
He then announced an official "call for proposals" from contractors to develop and submit design concepts for the propulsion module. Those proposals are due by the end of this week.
"In essence, we seek capabilities that highly dextrous robots assisted by humans on the ground could bring to this mission," he said. "What we are looking for is not autonomous robotics, but tele-robotics. If this mission goes forward, people will still be servicing Hubble."
How much the proposed robotic mission might cost is not yet known, but it is believed to be nearing $1 billion. Whether NASA can find the money to pay for it is an open question. But Hubble managers believe the mission is well within their technical capabilities.
"We think it's very doable," said Preston Burch, Hubble program manager at Goddard. "We're doing this on a very compressed time schedule but we're looking at pulling this off. Our goal is to launch this in December of '07 in order to give ourselves ample margin."
When Columbia was launched on Jan. 16, 2003, Servicing Mission 4 was scheduled for takeoff the following November. The objectives of the flight were to upgrade the telescope's scientific capabilities and to repair its aging subsystems. The two new science instruments, both with built-in corrective optics to compensate for a flaw in Hubble's primary mirror, were expected to boost the observatory's data output 44 times above what it was 10 years ago.
Spacewalking astronauts planned to install:
From an operational standpoint, the most serious issues are the batteries and gyros.
"Batteries are the limit on total spacecraft life," Burch said. "Gyros are the limiting factor on science, the usefulness of the observatory."
When Hubble was launched in 1990, its six state-of-the-art batteries, charged during the daylight portion of each orbit, provided around 540 amp hours of capacity to keep the telescope warm and to run its instruments, computers and communications systems during orbital darkness.
The batteries currently have about 300 amp hours of capacity and they are declining at an average of about 6.3 amp hours per battery per year.
"In order to get through an orbital night period, we need 40 amp hours total for the whole system," Burch said. "But that means you would come out of the orbital night period with nothing, so you need some reserve. It's sort of like flying an airplane. You wouldn't fill the tanks with just enough gas to get there. You'd want extra.
"So our benchmark that we've set for ourselves is we would like a minimum of 110 amp hours. (That) would give us one orbit to cope with any kind of a major failure on the system and entry into a safe mode or something like that."
If the observed battery degradation continues at its current rate, Hubble will reach that 110 amp hour point in late 2008 or 2009. Once power is lost, Hubble cannot be resurrected. Within days of a total power loss, low temperatures would cause titanium fittings to unbond and the optical system would lose its critical alignment.
Engineers are developing new techniques for recharging the batteries and reducing Hubble's power requirements, such as not operating instruments simultaneously. That will reduce the flow of secondary data to the ground and ease the load on the telescope's power-draining radio transmitters.
Hubble's gyroscopes are another pressing concern. Three of the six gyroscopes must be operational to conduct science. Currently, Hubble is running on gyros 1, 2 and 4. Gyros 3 and 5 are failed and gyro 6 is on standby as a spare. Based on how gyros have failed in the past, there is a 50-50 chance Hubble will lose two more gyroscopes by early to mid 2006.
Engineers at Goddard are developing complex computer techniques to continue science operations using just two operational gyroscopes in concert with Hubble's magnetic sensing system, fixed-head start trackers and a fine guidance sensor. While some observations would suffer, the observatory should be able to accomplish about 70 percent of its current science agenda.
"We figure the two-gyro science mode will give us an extra 12 to 18 months of science," Burch said. "That takes us out to about mid 2007. So we expect around mid 2007 is when we're going to go off the air in terms of science."
Given all of Hubble's ailments, and its obvious scientific value, it's not surprising O'Keefe's sudden decision to cancel SM-4 triggered a wave of protest and depression.
"People here are brushing off their resumes," an astronomer at the Space Telescope Science Institute said the day the decision was announced. "It was like walking around a funeral home."
NASA already planned to launch a robotic module to safely drive Hubble into the atmosphere at the end of its life. It didn't take long for the agency to begin studying the prospects for a robotic servicing mission.
While the design is far from being finalized, the HST Robotic Vehicle, or HRV, taking shape on the drawing boards at Goddard would feature two primary components: a permanently attached Deorbit Module and a disposable Ejection Module. The new batteries and charge controllers would be housed in the Deorbit Module, along with fuel tanks, a small rocket engine and other electronic gear.
The larger Ejection Module would carry additional fuel for the rendezvous, along with the two new instruments and six gyroscopes. It also would be equipped with a 32-foot-long robot arm similar to one being developed for an experimental satellite servicing spacecraft known as Orbital Express.
Also on board: a special purpose dexterous manipulator, or SPDM (pronounced spih'-dum) built by MD Robotics of Toronto. This is a robotic "hand" developed for use on the international space station that is capable of small-scale manipulation.
The HRV would be launched atop an Atlas 5 or Delta 4 heavy-lift booster. After a two- to 12-day orbital rendezvous, the HRV's robot arm would lock onto one of Hubble's two shuttle grapple fixtures and reposition the repair craft so the forward face of the drum-shaped Deorbit Module was just under the observatory's aft bulkhead. The Deorbit Module then would lock onto three fittings normally used to secure the telescope on a shuttle servicing platform.
Hubble servicing would take place in deliberate stages over the next 30 days or so. The first objective would be to wire the new batteries into Hubble's electrical system.
"We would hook the batteries into Hubble electrical system through the external umbilical, which is on the bottom," Burch said. "But unfortunately, that's a one-way street for electrons. It allows electrons in, but there's no way to get power from Hubble's solar arrays to the batteries through that connection."
The solution? Use the Ejection Module's robot arm and SPDM to run wiring harnesses from the Deorbit Module up the side of the telescope to so-called diode boxes at the base of each solar array. Once plugged into unused, redundant connectors, the cables would carry power directly to the charge controllers in the Deorbit Module. Power from the new batteries, in turn, would flow into Hubble through the umbilical connection on its aft bulkhead.
Engineers came up with an especially clever way to equip Hubble with new gyros.
"We said, why don't we mount the gyros in the Wide Field Camera 3?" Burch recalled. "Because the Wide Field Camera 3 has got a very stiff structure and it's relatively easy to install ... There are only three bolts that have to be turned to get (the old instrument) out. And that camera latches right into the focal plane assembly."
Getting data from the gyros to Hubble's pointing system is another matter. To accomplish that, the SPDM must loosen six bolts, open a door on electronics bay 1, remove a terminator plug from an unused test port on a computer and plug in a cable from the gyros.
"That task, we have demonstrated at MD Robotics up in Toronto," Burch said. "We took a high fidelity mechanical mockup of bay 1 and we used the SPDM robot up there, the test bed (version) of it, to open the bay and pull the terminator plug off and put the new connector on there. And everybody who has seen that video, they become instant converts. It's one of those things you have to see to believe."
The robot arm is equipped with stereoscopic cameras and what Burch described as smart-bomb technology.
"On board, it's got a CAD (computer-aided design) drawing of the area that it's working on," he said. "And what it does is it takes the imagery and does a compare to the CAD drawing. When the two get overlaid, you've basically got the arm registered, if you will. It's pretty neat."
To install the Cosmic Origins Spectrograph, the SPDM must open two large doors and pull an unused coffin-sized instrument called COSTAR out of its slot. The arm then must pull COS from its storage location in the Ejection Module and maneuver it into position in Hubble's optical path.
Once repairs are complete, the Ejection Module, carrying the robot arm and the old science instruments will detach from the Deorbit Module and fall back into the atmosphere. The 4,000-pound Deorbit Module will remain in place until it performs its final service: pushing Hubble itself back into the atmosphere when its science days are finally over.
"The robotic servicing development effort at Goddard was officially initiated in 2004 and is a very recent undertaking," the NAS committee wrote. "While considerable advances have been made in just a few months, there has been little time for NASA to evaluate and understand the technical and schedule limitations of robotic servicing.
"The committee was gratified by your assurance that the robotic efforts will be adequately supported by the required resources in a timely manner. During the next year the robotic servicing mission project will have to achieve key milestones (including a critical design review in the summer of 2005) that will clarify the feasibility of a robotic servicing mission. Substantial resources will be required in Fiscal Year 2005 to accomplish this.
"The committee finds the proposed robotic mission to be highly complex due to the inherent difficulties with supervised autonomy in the presence of time delays; the integration of vision and force feedback in six-degree-of-freedom assembly and disassembly tasks with high-degree-of-freedom, dexterous manipulators; and the coordinated control of the high-inertia HRV with a long-reach robotic arm grappling with a high-inertia payload. Robotic emplacement of a deorbit module and replacement of instruments and subsystems on Hubble will require a rendezvous with a non-cooperative vehicle together with a human in a telerobotic loop that has a substantial (on the order of 2-second) time delay."
The NAS committee's final report is expected in August or September, around the same time Goddard engineers plan to brief O'Keefe on the technical feasibility and cost of a robotic repair mission.
The NAS panel strongly urged O'Keefe to keep an open mind.
"The committee was cognizant and most appreciative of your extensive discussions with us related to the ownership that you, and NASA, have for the shuttle return-to-flight and for astronaut safety in the nation's civil space program," the panel wrote. "You stressed that total elimination of risk in crewed space flight is 'impossible' and that you and NASA are 'not risk averse.'
"From information it has received, including the risk information to date, the committee concludes that there would be little additional investment in time and resources required over the next year for NASA to keep open an option for a human servicing mission to Hubble."
But O'Keefe's opposition to a shuttle servicing flight is so entrenched, it would appear the NAS panel would have to strongly endorse such a mission to give Mikulski and other lawmakers the ammunition needed to force NASA's hand.
In theory, NASA could award a contract for design and construction of the Deorbit Module as early as September or October. The Ejection Module would be built in-house at Goddard.
"It's exciting, new work and with the servicing mission on hold, a lot of folks were wondering where's my future and a few folks were starting to wander away," Burch said. "Now we've got people lined up at the door who want to come work on this mission, because they know A, it's fun and exciting and they're going to learn a lot and B, this is really going to be a big part of NASA's future."