While astronauts carry out the complicated task of rendezvousing with and grappling the bus-sized Hubble Space Telescope during the fourth and final servicing mission this fall, a small hybrid computer will execute the same exacting maneuvers in a simulation to test whether computers can do the work of humans.
As the real docking unfolds, algorithms installed on the SpaceCube computer will pretend that they are autonomously controlling the Shuttle’s docking with the observatory.
“The astronauts will be controlling the real docking, but SpaceCube’s docking algorithms will be running in parallel behind the scenes,” said Thomas Flatley, a Goddard technologist who is using Internal Research and Development (IRAD) funds to develop science applications for the technology. “If the algorithms work well, we may be able to use them to do automatic computer-controlled dockings in the future.”
Developed by Goddard technologists John Godfrey, Dave Petrick, and others on the Hubble Space Telescope Relative Navigation System, SpaceCube is a small, hybrid computer system that provides 15 to 25 times the processing power of a typical Rad750 flight processor used on most command, data, and handling flight computers. Making it even more attractive is SpaceCube’s use of commercial-grade components, which are significantly less expensive than those that have been “radiation-hardened” for space flight.
The Hubble servicing mission is the first space-based demonstration of the technology, Flatley said. Should it perform as Flatley expects, its potential to revolutionize NASA’s command, data, and handling needs is enormous, he added.
For future missions, the technology’s most important contribution could be in on-board data reduction and product generation.
Future NASA missions are expected to generate enormous quantities of raw data that will easily overwhelm onboard data-storage systems. “If data processing could be performed on the spacecraft itself, then only the final data products would have to be stored,” Flatley said. “This could allow an instrument to collect 10 to 50 times more data without increasing onboard storage capabilities.”
Developed by Goddard technologists, SpaceCube is a small, hybrid computer system that provides 15 to 25 times the processing power of a typical Rad750 flight processor. It will be demonstrated for the first time in space during the Hubble servicing mission.
To help realize that goal, Flatley is using IRAD funds to develop a variety of computer algorithms that would give the spacecraft’s data-handling system the ability to process memory-intensive raw data into less weighty images and other products.
Future Flight Opportunities
SpaceCube also has potential to dramatically decrease the cost of spacecraft missions.
Today, spacecraft engineers overcome the ill effects of radiation on electronic components, such as computer processors, through the use of “radiation-hardened” hardware. Although these specially built components work well in space, they are very expensive to produce. One application-specific integrated circuit, for example, can cost as much as $1 million to produce.
SpaceCube, however, is equipped with radiation-tolerant circuits, which means that while they can withstand the harsh space environment, they are nonetheless prone to radiation-induced “upsets.” However, by employing special mitigation software techniques, SpaceCube can find and computationally correct these errors. Flatley will get his chance to demonstrate the capability in space next year when SpaceCube flies as part of a Naval Research Laboratory experiment on NASA’s Express Logistics Carrier, an attached payload on the International Space Station.
Despite the technology’s huge potential, Flatley conceded that it might be unsuitable for some spacecraft applications. Computer systems needed for risk-intolerant applications, such as providing oxygen inside human habitats on the Moon, for example, may always require true radiation-hardened devices and/or “super redundancy,” he said. “Our goal is to see how far we can go.”
Goddard technologists win new work, secure follow-on funding to mature new technologies, formulate concepts, and validate new instrument concepts in flight demonstrations — successes that benefit Goddard and the scientific community as a whole.