Geosynchronous orbit is characterized by high radiation, so assembly and construction of the “Sun Disk” power satellites will be done almost entirely via semi-autonomous “Dragonflys” and telerobotically operated “Stevedores”. There will be several other special purpose machines/robots which will participate, but for the purposes of this study these will well represent the two classes of “robots”. The term robots will be used here in a broad sense to encompass a wide variety of devices that interact with men and women and perform work under direct, remote, infrequent or no control and/or supervision. Such robots are common now and becoming more common. The use of robots is also intended to nearly eliminate the exposure of hundreds of space walking astronauts to ionizing solar flares and other galactic radiation - a major health risk and cost saving. Murphy (Murphy’s Law) will also guarantee that some direct, but small amount of on-site astronaut assistance will be needed, such as in control and communications final assembly and checkout.
“Stevedores“ , weighing about 1000 kg are intended to perform the heavy
work of unpacking freight, minor assembly and positioning of equipment
and fastening various pieces together. Several actual space qualified
and tested examples of this class of telerobot currently exist, for
example, the Manipulator Flight Demonstration (MFD) which flew August 17-19
aboard space shuttle Discovery ( “Robotic-Arm Trials Lay Groundwork for
Japan’s Module”, Space News, September 1-7, 1997, page 9) That six
jointed fine-arm attachment successfully performed all scheduled
functions, including loosening and tightening bolts, opening and shutting
hatches, attaching and detaching MFD’s Orbital Replacement unit, performing
remote visual inspections, (using a camera on its “wrist”) It was also
tele-operated from ground controls at Johnson Space Center via file transfer.
Stevedores will be capable of refueling themselves from a mobile service
module. The mobile service module will also provide check out and return
for various common and special purpose tools and actuators that have hardware
interfaces to their arms and other appendages. Stevedores will normally
“check-out”, carry and use different several types of tools, such
as unpacking tools and socket wrenches and that are effectively extensions
of their arms.
The Remote Manipulator System’s planned functions include maneuvering, collecting and storing payload and experimental equipment. The cost of Toshiba’s prototype was $50 million. In several years 200 such devices could be available commercially for $5 million each including refuelable thrusters on each “Stevedore”. (Toshiba is already planning the commercial use of these maintenance telerobots for LEO satellite service, in-orbit refueling, etc., and in fact has begun test flights – same Space News 9/1/97 article.) NASA also has a non-commercial telerobot Ranger that has many attractive similarities to those that will eventually be needed. A few links among thousands on such telerobotic work and equipment include:
NASA AMES
Telerobotics
NASA JSC’s
"
Deneb
Robotics
Univ.
of Washington
Univ of Toronto
The shape of the satellite will be twelve sided. Straight edges will eliminate the waste of wrong shaped end pieces that were trimmed from rectangular stock, simplify construction, and create twelve useful control vertices where vibration damping weights and control arcjets for handling might be positioned. Reinforcing of the disk against anticipated typical stresses like stretching, vibration, potential tears from micrometeorite hits, etc., may require some additional backing prior to the application of electrical collection wires and buses which will add weight and possibly some strength. UltraFlex’ inorganic backing may be of low enough thermal dimensional response to ameliorate, but not eliminate, stretching problems over the several kilometers traversal distance.
Dragonflys will be assigned the task of joining the thousands of solar
“blankets” which compose the SPS. These solar blankets have a heat
sensitive bead on all sides which is insensitive to the operational 60°
C temperature of the disk surface, but will melt briefly when the Dragonflys
apply temporary additional heat to join the blankets. The joints after
bonding will be stronger than the PV blanket or fabric. The bonding material
must be durable, flexible and radiation tolerant, such as a low temperature
composite fiberglass.
Other possible alternatives include lock stitching and zipping. The
inorganic base of UltraFlex’ backing materials may open the door to research
the possibility of replacing this fabric backing material with a similar
material derived from inorganic lunar regolith.)
Bonding, once initiated by the Dragonflys, proceeds at about 1 cm per second. Independent of set-up times, other assembly, etc., the bonding process alone (for 10 x 500 meter blankets) will consume about 1000 hours per SPS. Ten Dragonflys could complete it in 100 hours, perhaps faster than blankets could be received and deployed by the Stevedores, were it not for interruptions including communications, tangles, reinitializing Dragonfly bonding runs, etc.,.
They are called Dragonflys because they will have large “wings” that collect solar power enabling them to be quite mobile and move independently along fabric edges, although they will cooperate with the Stevedores which will provide transportation and placement on the blanket edges to be joined. The Stevedores will also clip devices like clothes pins on the corners of the blankets which will maintain the blankets in rough position until the Dragonflys have completed their joinery of each seam. Dragonflys will have a rudimentary communications capability consisting of a tiny local transmitter that can barely reach the service module communications center during typical solar wind interference. During the early prototype work, Dragonflys will no doubt experience problems which will require further intervention by Stevedores and design modifications by their manufacturer. A few of them may be lost in space since their primary locomotion of gripping the blanket edges and moving along the edges will not be foolproof. The six legs of the Dragonfly will be busy gripping and feeding the blanket edges to the heating element in the head of the Dragonfly as it moves. The blanket edges the Dragonfly has already joined will serve to prevent the still soft seam from parting until it cools and gains strength. This is one of the key processes to be mastered in assembling the SPSS.
A recent book “STIQUITO, Advanced Experiments with a Simple and Inexpensive Robot” by James M. Conrad and Jonathan W. Wills, ISBN: 0818674083 Paperback, 328 pages, Published by IEEE Computer Society December 1997 is a valuable primer for this class of robot.
Command, control and communication of the Stevedore robots will
be handled via a telerobotic link to earth. A typical Stevedore will carry
several end effectors, many of which are tested and others are under development
in the Ranger program, for example.
Communications
A communications and service module will oversee and provide many support capabilities at each SPS construction site. It will also contain refueling tanks and stations for the Stevedores to refuel themselves. They will handle up to a hundred TV channels at each of four construction sites. These will share twenty transponders at a common laser linked communications satellite, making two to four hundred TV channels available for use during construction. These will “float” as dynamic multiplexing handles channels that vary in usage characteristics from a fast moving operation, such as the busy unpacking bay, to the placid flow of a panel bonding monitor, which has low video dynamics and low channel demand. (A fast-moving professional basketball game, for example, demands much more bandwidth, than a “talking head” news program.
All these channels will be dynamically “packed”or compressed real-time
with MPEG-2 or a better later compression standard. MPEG-2 will greatly
reduce the weight and cost of real-time video communications with and control
of telerobots with 90% lighter weight transponders. The channels
will be received, decompressed, demultiplexed and distributed to various
operators and operations supervisors at the primary Telemetry, Tracking
and control (TTC) center directly below the SPS construction site (plus
or minus 5 degrees). The TTC center will maintain constant communications
with another Construction, Command and Operations (CCO) center which will
house a working staff of 1,500. Costs for both centers combined will be
$ 1.5 Billion per year during the construction process, with four SPSs
under construction simultaneously. In the currently soft Asian communications
satellite market, twenty transponders at GSO can be bought for $50 million.