|by Jim Oberg<p >The space bus infrastructure becomes more sharply definedNASA has just announced new plans for a familiar type of robot spacecraft. Of intimate interest to robot builders is NASAs new philosophy of asking for help from the outside world, and of creating a long-range plan in which a wide variety of space robots will be inserted into an integrated space expansion strategy leading ultimately to human interplanetary flight.
At a two-day symposium in Galveston, Texas, a parade of NASA officials described their vision to several hundred industry and university guests, myself among them. The presentations impressed me they were well thought out, well coordinated, and candid in their confession that NASA didnt have enough in-house smarts to make it all happen.
The opportunities for outside robotics teams have never looked wider or more genuine. NASA people at the seminar showed a willingness to listen, and to struggle honestly with tough questions (I had a few). I found the atmosphere immensely encouraging.
<p >SMART SPACE TUGTake the one particular robot in question for this article the smart space tug. While it lacks the anthropomorphic appendages of arms and hands, it has a robot brain with a robots perceptual and planning functions. And its services in space are urgently needed.
The fundamental mission of the Space Shuttle was always to assemble a large space station by bringing large modules into orbit and gentle attaching them together. By the end of 2010 that mission will have been accomplished and smaller rockets can now handle the task of crew transport and supply deliveries. The shuttle, which costs a lot more than smaller spacecraft to operate but in turn delivers special services no longer required, can be retired.
But NASA planners realize that on occasion big packages will still need to be delivered to the station, especially as its lifetime has officially been extended until 2020 and probably will be further extended in years to come, perhaps indefinitely. New modules for a growing crew complement will be needed, and new science research laboratories, and eventually new solar power arrays. And prototype plasma propulsion systems for interplanetary voyages have already been scheduled for launch in coming years. Ultimately, large components of spacecraft to carry humans into interplanetary space will be assembled in orbit, and that process needs special robot guidance.
Getting such packages into orbit on expendable rockets is relatively simple. But getting the package delivered to the doorstep of the space station is a highly complex challenge.
<p >AUTOMATED ROBOTIC DOCKINGFor this type of mission, NASAs new space strategy looks to develop an Automated Rendezvous and Docking (AR&D) system, installed on an automated orbital tugboat. The idea is not new, since the Russians have used remote-controlled unmanned supply drones for more than thirty years, and both Europe and Japan have recently built bigger ones. A few years ago, Russian space engineers proposed a remote-control space tug called Parom, but couldnt get funding. What NASA now intends to pioneer is a new level of on-board navigation and guidance smarts to make the vehicle almost entirely self-sufficient for carrying out its mission.
NASAs passion for buzzwords and acronyms can often seem comical, but this approach actually functions well as a means to unambiguously isolate different functional parts of a bigger, complex process. For example, the following semantic nit-picking can actually help in the piece-by-piece understanding of the big picture for the as-yet unnamed space robots primary duty, Automated Rendezvous and Docking (AR&D).
In NASAs dictionary, the term automated (as in automation) deals with human versus computer control, this way: Control or execution of a system or process without human or commanding. Function [is] performed via onboard and/or ground software interaction.
The related term autonomous (autonomy) deals with ground versus onboard functions. NASA explains it this way:: Capable of operating independent of external communication, or control (i.e. commands from mission control on Earth). [It] can involve crew and software in nominal and contingency operations.
<p >HUMANS VS. ROBOTS Under these definitions, NASA explains: The level of automation or autonomy is the split between onboard crew, onboard flight computer, and humans ground. This is typically a range instead of strictly human/computer or ground/onboard with levels varying on the application. For envisioned future human exploration endeavors this entire range will be required.
Its the treatment of the wide range of capabilities and how they evolve to more and more powerful capabilities that sets this new NASA space robotics strategy apart from previous hobby shop demonstrations. This time, the missions are to be interlocking building blocks. Officials at the Galveston seminar described flagship missions to be launched in the nearest possible future four or five years from now to try out the most advanced space robotics that industry and academia have to offer.
Their plan is to demonstrate an Automated/ Autonomous Rendezvous and Docking by creating a spacecraft to rendezvous, operating independently and without other back-up. This task, explained Craig Tooley of NASAs Goddard Spaceflight Center in Maryland, requires advances in sensors, realtime on-orbit positioning and flight control, among other technology, which will be :critical to the ultimate success of capabilities such as propellant storage and refueling, and complex operations.
<p >FLAGSHIP SPACE BUS This multi-purpose vehicle would be used across several other Flagship missions. It would be the bus for specialized experiments with ion drive and space refueling. By the second or third test it would fly to a target, attach itself, and tug that target to a different target usually but not always the space station. This would require basic control capabilities plus a lot of push to change orbital paths, and the ability to be refueled and reused for a series of different missions. To carry out these missions to robot needs to be able to see, using hardware and software: This so-called Vision Processing Unit will in NASAs words enable the ability to process images for relative navigation information as well as host other important GN&C [guidance, navigation, and control] software functions. It will collect images in visible and infrared light, create its own illumination with spotlights and laser beams, and then the really hard part correctly interpret its targets relative position and orientation even without any help from the target or from Earth.
By enhancing the degree of autonomy of these space tugboats, NASA will be developing vehicles that can operate more efficiently in orbits close to Earth. Even better, they will be able to operate in regions so far from Earth that ground control (or even occasional ground advice) becomes undependable due to the time lag induced by the speed of light.
<p >CONCLUSIONThis project is just one of many parallel developments in a new and controversial space strategy that says that really interesting future space missions with and without astronauts will be more feasible if NASA enhances its repertory of space operations capabilities rather than relying on incremental improvements of old tricks. The approach is an opportunity for more participation by a wide cast of new players with far-out ideas, and developing the eye-brain combination of a future robot space tug is one such opportunity for academic and smallindustry teams to contribute.
NASA, www.jpl.nasa.gov For more information, please see our source guide on page 89.
Words by Jim Oberg