DESIGN FOR ON-ORBIT SERVICING

Stephen J. Leete NASA Goddard Space Flight Center

ABSTRACT Servicing to date has been in low earth orbit from the , and in the near future will also be based from the ISS. Future servicing should occur at much more energetic orbits, such as HEO, Earth-Moon L1, Sun-Earth L1 or L2. Mass for systems at these orbits will be at a high premium. These systems will also typically include large, highly flexible components (gossamer structures, sun shades, sails, solar arrays, mirrors) which can sustain only limited loads once deployed. However, the current EVA interface requirements and reference designs impose considerable mass and cost penalties on . This is partly because of high inadvertent EVA loads (kick loads, PFR ingress loads), and massive EVA-friendly connectors and mechanisms. A new generation of requirements and design solutions is needed to enable the designer of new systems to incorporate requirements related to servicing. An assessment of the current situation and a roadmap of the technology development needed to enable future ambitious serviceable systems is presented.

HISTORY OF SERVICING The servicing of spacecraft on-orbit has been one of the highlights of the space activities. The three major categories of this are servicing of space stations, such as , Salyut, , and the International (ISS), engineering development of servicing capabilities, and servicing of . These activities are summarized in Table 1, History of Extravehicular Activities, and have been taken from an excellent book by Portree and Treviño for all activities up to April, 1997 [Ref 1], and various NASA web sites for the time period after that [Ref 2].

Vehicle Year(s) Activities, Accomplishments Voskhod 1965 First EVA Gemini 1965-1966 10 EVAs, 1st successful EVA, demonstration of EVA tasks 1969 Crew transfer Apollo 1969-1972 Crew transfer, lunar surface exploration & repairs, Apollo 13 safe return Skylab 1973-1974 Deploy sunshades & solar array, repair & service science instruments Salyut 1977-1986 Install solar arrays, remove 10-m antenna, repair propellant system leak, demonstrate welding tool, erect trusses, maintain science experiments. Space Shuttle 1983 Demonstrate MMU, MFR on RMS, hydrazine transfer. Space Shuttle 1984 Solar Max Mission recovery and repair, Palapa-B and WESTAR-IV recovery and return to Earth. Space Shuttle 1985 Retrieve and repair, release Leasat-3/SYNCOM-IV ; test EASE & ACCESS truss assembly. Space Shuttle 1986 Challenger disaster Mir 1987-1982 Repair various Mir hardware, repair and install solar arrays, reconfigure Mir modules, install boom, test Soviet MMU (SPK), install and tend science experiments, install new propulsion module. Space Shuttle 1991 Contingency GRO antenna deployment, tested EVA hardware Space Shuttle 1992 Retrieve Intelsat-VI, install new PKM & release; erect ASEM truss, test Crew Propulsive Device

Page 1 of 12 Vehicle Year(s) Activities, Accomplishments Space Shuttle 1993 Gain general EVA experience, safe an appendage on EURECA payload for landing Space Shuttle 1993 Service HST (SM1): replace solar arrays, install WF/PC-2 & COSTAR with optical prescription corrections, gyros, magnetometers, and a 386 co-processor, reboost Mir 1993-1996 Install solar drive units, deploy Papana truss, extensive photography of Mir exterior, reconfigure modules, deploy and install solar arrays, install 2nd Strela boom, install and maintain science instruments, filmed a Pepsi commercial. American astronauts performed an EVA while docked to Mir. Space Shuttle 1994-1996 Test SAFR crew rescue device, improve suit performance in extreme cold, practiced various EVA tasks Mir 1997 Joint EVAs between cosmonaut and astronaut, test Orlan-M suits, inspect damage to due to severe impact in June, internal EVA to reconnect power cables from Spektr solar arrays Space Shuttle 1997 Service HST (SM2): install STIS and NICMOS axial instruments, replace FGS, install solid state recorder, replace reaction wheel, replace data interface unit, replace magnetometers, and unplanned installation of blanket patches ISS 1998 Joining of Unity and modules, first use of Space Vision System Space Shuttle 1999 Service HST (SM3A): replace gyros (3), computer, FGS, SSR/tape recorder, s-band transmitter, and install new outer blanket layers (2) Space Shuttle 1999 ORU Transfer Device and Russian Strela crane installation, Chandra deploy ISS 2000 Zvesda Module joins ISS, complete Strala and OTD installation, add Z-1 Truss & Pressurized Mating Adapter #3, re-locate S-Band Antenna, install Space to Ground Antenna, install EVA Tool Stowage Devices, Soyuz brings first expedition crew, added solar arrays (including EVA intervention to complete solar array deployment) ISS 2001 Destiny lab module installation (Curbeam sprayed with ammonia, cleaned with hydrazine brush), re-locate PMA-3 on Unity, install External Stowage Platform on Destiny, Space Station Remote Manipulator System installation, UHF antenna install/deploy, Joint Airlock Module installation using SSRMS, and install Russian Piers airlock. Table 1: History of Extravehicular Activities

SPACE STATIONS If a satellite is to be serviced by humans, it must not present a hazard to those humans. It must not present unacceptable electrical hazards such as high-current connector mate/demate, thermal hazards such as extremely high or low touch temperatures, or cutting & entrapment hazards such as sharp edges that could cut a spacesuit or holes in which gloved fingers could get stuck. The spacecraft must not be a hazard when subjected to inadvertent EVA loads, also called bump or kick loads. There are provisions for allowing some hazards, as long as the crew will be adequately briefed and trained in avoiding them. This would then be mitigated by more care in other EVA accommodations, more detailed training, or other requirements on the servicer tools. However, the designer may decide to allow some or all of the satellite to be damaged to the point of loss of function.

Page 2 of 12 For a space station, damage to the operation of the space hardware may also present a hazard to the occupants of the space station, so nearly all damage becomes hazardous either directly or indirectly. A space station is likely to be serviced many times over its life, sometimes on short notice, so having to brief a crew on so-called "no-touch" and "no-damage" zones can be overly risky and costly. The solution to this situation adopted for the ISS is to have a very low threshold of tolerance for potential hazards on ISS hardware, and to make the hardware very robust with regard to inadvertent astronaut damage. These practices are comparable to more mundane guidelines for ensuring safe work-sites on the ground, such as clear labels on chemicals, wearing steel-toed shoes and hardhats, etc. These choices make good sense for the ISS. Being in low earth orbit, launch costs to bring up major components are high but not unreasonable. Adding weight once that will increase the reliability, safety, ease of servicing, and other important characteristic at the cost per pound of the Space Shuttle or other Russian launch vehicle is a good decision. If servicing of the ISS is too time-consuming, costly, or dangerous, its entire mission is compromised - as was found to be the case for an earlier version of Space Station Freedom. [Ref 3, 4]

SPACECRAFT SERVICING DESIGN - MMS This approach can also be taken with free-flyer . They can be rugged, with many design decisions made to favor ease of servicing and simplicity. The satellites which exemplify this approach are the Multimission Modular Spacecraft (MMS) design. During the years prior to the Challenger disaster in 1986, NASA had a broad vision of satellite servicing to take place from the Space Shuttle or space station. This would be in just about any inclination low earth orbit. Even post-Challenger, the vision was that there would be space tugs (orbital maneuvering vehicles) that would bring high orbit spacecraft back down to STS orbits for servicing. For various reasons, this vision has not yet come to fruition. A graphic of the MMS design is shown in Figure 1. [Ref 5]. With the post-Challenger decision not to launch shuttles from the west coast, most of the Earth observing satellites which are in polar orbits, such as the Landsat series, were no longer candidates for servicing. The Solar Maximum Mission (SMM) launched in 1980, Landsat 4 and Landsat 5 launched in 1982 and 1984, the Upper Atmosphere Research Satellite (UARS) launched in 1991, and the Extreme Ultraviolet Explorer (EUVE) launched in 1992. All of these were built for the NASA Goddard Space Flight Center. Additional missions using the MMS design may have been built for other users. The MMS design was for a system of serviceable and re-useable satellites and satellite components. This modular design was also intended to make satellite design, assembly (integration) and test faster and less expensive. It was the implementation of a vision of regular access to space, modular and interchangeable spacecraft components, easy integration of new technology, the establishment of design standards, and other concepts.

Page 3 of 12

Figure 1. MMS Components

The hope was that the design, construction and servicing of satellites could be made to resemble more the current situation with, to use a current analogy, personal computers. There are a large number of these in the field, strong demand for them being kept up to date with current technology (thus a rapid turn-over, frequent up-grades), and a multiplicity of providers of components and chassis. Vendors of systems manufacture some components themselves, and purchase the rest on the open market from vendors who manufacture to standards. There is intense competition at all levels (with the possible exception of the operating system software), and customers benefit from the resulting low prices. The benefits to NASA, as well as other purchasers of satellites such as industry and military, from having a similar set of circumstances would be large. In addition to lower costs, the speed at which a new spacecraft could be procured, developed and put in orbit would be driven down. NASA could focus on developing the payload, then pick a spacecraft bus off the assembly line, integrate and launch. If, in addition, the standard design included features which made it serviceable, even more operational benefits could be achieved. The training and shuttle interface workload needed for preparing to service a satellite would be greatly reduced if most of the satellite

Page 4 of 12 was nearly identical to a satellite that has been serviced previously. The only unique elements would be on the payload side. The concept makes a lot of sense if servicing is going to be common, the number of satellites involved is large, and the weight, cost and other technical parameters of the design are comparable to those of a point-design satellite that can satisfy the same mission requirements. Since the above conditions have not been met, no MMS satellite has been launched by NASA since 1992.

SPACECRAFT SERVICING - HST The Hubble Space Telescope is the last spacecraft to be designed and built for servicing. The path to its current design was somewhat irregular. The original design had many ORU's with mechanized connector mates and other ORU features favorable to servicing. In 1982, as a cost-cutting measure, the number of electronics units designed as ORUs was dramatically decreased. In 1985, money was restored to fund converting boxes back into ORU type by adding adapter plates to some, Lockheed also installed servicing features such as instructional decals, added handrails, connector backshells, etc. Lockheed also developed the HST Portable Foot Restraint, connector tools and 7/16 & 5/16 sockets, This process continued from 1985 to 1989, shortly before launch in April, 1990. [Ref 6] The following summary was included in a presentation by astronaut Bruce McCandless in a presentation at the fourth Satellite Servicing Working Group meeting in June, 1989. "The HST has 51 different types of Orbital Replaceable units in three classes. Almost everything not welded to the vehicle or located inside the forward light shield is replaceable. Notable exceptions include all wire harnesses, magnetic torquer solenoids, and magnetometers. Class I ORU's are fully "EVA-rated" with wing-tab ganged electrical connectors and mechanical interfaces optimized for EVA use with standardized sockets. Class II ORU's have substantial provisions for EVA compatibility, but in general require use of one or more of the 72 HST-unique tools. Class III ORU's include everything else that is assessed as potentially replaceable on-orbit." [Ref 7] Note that magnetometers were serviced on-orbit, despite not being designed for servicing. Since the time that a number of HST-unique tools were made prior to HST deployment, additional tools have been made for each of the servicing missions. These have included the Pistol Grip Tool that was developed for the second HST servicing mission, which is now the standard power tool used on the space shuttle and ISS. The activities carried out in the servicing of the HST were included in the summary of EVA activities in Table 1. These maintenance and upgrade activities have accomplished many things. They have made it possible to have state of the art instruments take advantage of this unique resource. The telescope continues to make discoveries worthy of front-page news articles on a regular basis 11 years after its launch. It has been expensive to perform this maintenance. Replacing HST with a new satellite and new instruments would have probably been more expensive (some would be much more definite in this statement). The servicing that has been performed on HST falls into five major categories. These include 1) direct replacement with identical or nearly identical units, 2) replacement with a significantly upgraded unit that includes new technology, 3) installation of additional hardware that performs new functions not included in the original design to enhance the

Page 5 of 12 functionality, 4) retrofit and repair via addition of hardware or replacement of units with hardware having entirely different functions, and 5) improvised repairs of problems not anticipated prior to a servicing mission. These are summarized in Table 2 below. [Ref 8]

Category Units Replacement RSU, ECU, FGS, RWA, DIU, SADE, Fuses, SSAT, PCU, (Batteries)* Upgrades 486 Advanced Computer, Solid State Recorder, STIS, NICMOS, ACS, (COS, WFC3) Additions 386 Co-Processor, OCE-K, VIK, BIK, SA-3, NCS, ASLR (ASCS) Retrofit Repairs COSTAR, WF/PC-II, SA-II, GHRS-K, NOBL Improvised Repairs Magnetometer Cover, Blanket patches Table 2, Categories of Servicing of HST * Items in italics are planned for SM4

The COSTAR corrected the optical prescription for the other three original axial instruments. WF/PC-II provided a radial instrument which could take large-format images with a corrected optical prescription. The second solar arrays (SA-II) corrected the thermal snap problem. All these were installed on the first servicing mission in 1993. Also worthy of special note is the NCS, or NICMOS Cooling System. With this system a key instrument on HST will be brought back into operation. The NICMOS is an infrared instrument that played a key role in the recent observation of a distant supernova which has helped solidify the theory of dark energy in cosmology. This instrument was initially cooled by a block of solid nitrogen until the cryogen sublimated. The NCS will add a closed-loop refrigeration system that will cool the detectors back down to operational temperatures. It includes a cryocooler, a control electronics box, an external radiator to be mounted to external handrails, and interconnecting electrical cables and flexible heat pipes. It will be installed on SM3B in 2002, and will be followed on SM4 with the Aft Shroud Cooling System which will use a similar radiator to provide cooling to the axial instruments to allow more instruments to operate simultaneously and extend their useful life. [Ref 9]

SPACECRAFT SERVICING - AXAF Another major example of a spacecraft designed for servicing is the Advanced X- Ray Astronomical Facility, another of the NASA great observatories with HST, Gamma Ray Observatory (now de-orbited) and SIRTF (split into two missions, not yet launched). The AXAF was intended as a first-generation design, but without the fits and starts that HST was subjected to. In 1989, Bruce McCandless summarized its design. "AXAF … [was] planned [for] Space Station Freedom servicing with Shuttle backup." "[It would feature] module removal and replacement, [with] medium-sized systems modules to be repaired on-orbit in shirtsleeve environment, [and] large scientific instrument modules non-repairable on- orbit." [Ref 7] The plan in 1989 was to perform routine servicing at 5 and 10 years from launch, including science instrument replacement. Emergency servicing one year after discovery

Page 6 of 12 of the need was also planned, and three years after for a mission to restore science capability. The spacecraft was shaped as a semi-cylinder that acted as a cradle for the telescope. The system included 6 Class I ORU's which were planned to be replaced on a schedule, eight Class II ORU's whose failure would put the system one failure away from loss of mission. While most of the ORU's were individually packaged, the command and data management system was mounted in a multimission modular spacecraft (MMS) module. There were 23 contingency replaceable unit categories, with all but the power electronics module packaged individually. The design of the ORU's and ability of astronauts to replace them using standard techniques was demonstrated at the MSFC neutral buoyancy simulator in a series of several runs. [Ref 3, 10, 11] Unfortunately, funding profiles (the yearly funding for a number of years) for AXAF caused a similar reconsideration as happened for HST in 1982. In the case of AXAF, this decision also put AXAF into a high, elliptical orbit quite in accessible from the space shuttle. This decision was not ever reversed, and the added expense of making a unit an ORU was cut permanently. Technology development for making AXAF serviceable ended in 1991. [Ref 6]

SPACE SCIENCE MISSIONS The NASA Space Science Enterprise has been identified as a likely customer for satellite on-orbit assembly and servicing. There are a number of missions for which large optics, gossamer structures, multiple satellites in tightly coupled formation orbits, and very long life-times. These goals are stated in NASA strategic plans [Ref 12, 13]. These missions will be extremely challenging to meet using current disposable spacecraft design approaches. It will be possible to meet some of the objectives using faster, better, cheaper techniques. However, NASA management has clearly identified the human and robotic servicing of these missions as part of an overall plan to have an integrated approach bridging the human exploration and science missions of NASA such that the two efforts complement and enable each other. In this scenario, the enhancement of humans ability to get to an operate in Earth's neighborhood beyond LEO is a prime example of something that will enhance both missions. A case has been made for servicing satellites in the Earth-Moon first libration point (E-M L1), which is a halo orbit about a point on a line between the Earth and Moon, 327 km (about 50 Earth radii) from Earth and 57 km from the Moon. The Earth-Moon libration points are all energetically close to escape from the Earth's influence. The L1 and L2 halo orbits in particular are coupled very closely, with near-zero delta-V needed to transfer into Sun-Earth libration point orbits, S-E L1 and L2. [Ref 14] The Solar and Heliospheric Observatory [1 from our paper] (SOHO) and Advanced Composition Explorer [2 from our paper] (ACE) spacecraft are currently operating in halo orbits about the Sun-Earth L1 location, and the Microwave Anisotropy Probe [3 from our paper] (MAP) is at the Sun-Earth L2 location. Genesis is en route to the Sun- Earth L1 at the time of this writing, and Triana was designed for operation at Sun-Earth L1. Future observatories that require cryogenic cooling, particularly the Next Generation Space Telescope (NGST), will probably operate at L2. These locations are the intended home for several more, as listed in Table X. [Ref 14]

Page 7 of 12

Candidate NASA Proposed Structure Size Observing Wavelength Programs Sites Regime Astronomical Search for Origins Terrestrial Planet Finder S-E L2 15 m IR Life Finder S-E L2 40 m IR, Visible, UV Planet Imager S-E L2 Not specified Visible, IR, UV Structure & Evolution of the Universe Space Interferometry 1 AU 10 m Visible Mission Laser Interferometer 1 AU Laser Interferometry Space Antenna Constellation X S-E L2 8 – 10 m X-ray Early Universe Observer S-E L2 300 m X-ray Submillimeter Probe S-E L2 30 m FIR – sub-millimeter Sun-Earth Connection .5 AU, high Solar Polar Imager 100 m solar sail inclination Space Weather Sentinel .95 AU sub L1 100 m solar sail Geosynchronous SAR GEO 30 m RF Geosynchronous LIDAR GEO 100 m Visible Various missions S-E L1, L2 > 2.5 m IR, Visible, UV Table 3. Candidate Programs for In-Space Assembly and Servicing at Libration Points

CURRENT EVA REQUIREMENTS The central document currently guiding designers of spacecraft hardware intended for EVA servicing is the EVA Hardware Generic Design Requirements Document, published by the Johnson Space Center as JSC 26626A. This document was revised in 1995 to be compatible with ISS requirements. In the author's opinion, this is an excellent strategy for the current situation. There would be advantages to keeping this commonality, especially in terms of crew training. Additional current requirements documents for servicing are listed in Table 4. [Ref 15]

Document Reference Document Title NASA STD 3000 Man-Systems Integrated Standards (MSIS) NSTS 07700 vol XIV Space Shuttle Systems Payload Accommodations NSTS 07700 vol XIV, System Description and Design Data Extravehicular Appendix 7 Activities, Space Shuttle Systems Payload Accommodations SSP 30256 Extravehicular Activity Standard Interface Control Document SSP 30550 Robotic Systems Integration Standards SSP 30558 Fracture Control Requirements for Space Station SSP 41000 System Specifications for the International Space Station SSP 50005 International Space Station Flight Crew Interaction Standard SSP 50006 International Space Station Internal and External Decals and Placards Table 4. EVA Requirements Documents

Page 8 of 12 CURRENT EVA TOOLS Several improvement have been made recently to EVA hardware which can have applications to servicing at highly energetic orbits. These are tools with the capacity to transfer much of the weight burden from the spacecraft being serviced to the servicer. If this hardware is transferred once to a servicing site, and re-used multiple times with little additional propulsion use, significant saving could result. A sample of these newer devices were tested on-orbit in 1995 on STS-69 by astronauts Michael Gernhardt and James Voss. These include crew aids that can attach to simple, low-profile features on a spacecraft to either permanently or temporarily convert a spartan worksite into an area rich in EVA interfaces. [Ref 4]

EVA Hardware Features / Capabilities On-Orbit Installed Worksite Interface (WIF) Fastens to a male dovetail fitting, provides mounting feature for an APFR Body Restraint Tether (BRT) Provides an easily adjusted rigid link between an astronaut's torso and a dog-bone handrail at a worksite Rigid Tether Assembly with Push Lock Tether Tool Grasps a equipment tether loop for no-hands equipment translation Multi-Use Tether (MUT) Similar to BRT and RTA, holds an ORU's dog-bone handle Dog-Bone Handrail Lighter Articulating Portable Foot Restraint (APFR) Load limiter allows lower load rqts on serviced s/c On-Orbit Installed Handrail Provides installable crew translation path Extravehicular Mobility Unit Helmet Lights Provides good worksite illumination Table 5. EVA Hardware Recently Demonstrated for ISS

TECHNOLOGY DEVELOPMENT FOR SERVICING If humans will be participating in the servicing of satellites that will operate at obits beyond LEO, whether the work is done in LEO or in a higher energy orbit, there are some design implications and a need for technology development. Existing designs should be considered a fledgling effort, a first shot, a proof of concept. Development has continued during HST servicing and ISS development. However, both of these are examples of servicing from a large base of operations (the shuttle or ISS) in low Earth orbit. For servicing at E-M L1 or other remote locations, the servicer will probably be smaller. The satellites to be serviced are of a wide range of sizes. Some will have large optical elements, although the total mass of these may not be that large. Mass will be at a premium. The systems will likely be relatively fragile compared to HST or ISS. One servicing concept places a relatively permanent facility at E-M L1. This could allow the burden of massive servicing features such as handrails and such to be put on the servicer, and to include only minimal features on each spacecraft that needs servicing.

SECOND GENERATION MMS: One concept that should be pursued is the development of a new standard spacecraft bus designed to the same principles as the MMS. For the purposes of this paper, call it the Second Generation MMS, or SGMMS. No serviceable satellite has been designed since AXAF, and that version was never built except as mock-ups for neutral buoyancy testing.

Page 9 of 12 It has been 10 years since AXAF servicing was cancelled, and very little, if any, work on free-flyer satellites for servicing has been done since. If MMS was being designed today, it would take advantage of work done on micro-satellite development with highly miniaturized components. The additional guidelines are that more modern EVA tools & techniques, as well as robotics, can be considered in the SGMMS development. In addition, the requirement for minimizing mass will be a far stronger design driver than it was for the MMS. Lessons learned from the MMS experience should be researched and implemented, before those who worked on them have retired. The other programs for using a standard spacecraft bus to support multiple missions could be studied, and an assessment made of what it would take to add servicing features to them. These include current designs of the various spacecraft manufacturers, such as Boeing, Lockheed, Loral, Orbital, etc. A path more likely to succeed is to start with a fresh sheet of paper and design a light-weight, serviceable satellite bus that could be used for the missions in question. Designs that allow for either human or robotic servicing would be optimal. One option would be an updated MMS design, with better weight and packaging characteristics. The emphasis in spacecraft design is to provide simple, light-weight and flexible design features. The ability to mate to a servicer such as the shuttle, ISS or free-flyer servicer should be built in. An RMS grapple fixture, or alternate feature that serves the same purpose should be included. The electrical design of a SGMMS is also critical, for it directly impacts the complexity of the interconnections between electrical units. A good model for this is the Small Explorer (SMEX) design. Most boxes are connected only by a 28V power source and a 1773 (or 1553) bus. This also has the benefit of robust, standardized interfaces which simplifies testing of individual ORU's, simplifies electrical integration, and minimizes the weight and complexity of the interconnecting harness. It also reduced the probability of problems with inserting a new version of an ORU into an existing spacecraft already on-orbit.

ADVANCED ORBITAL REPLACEMENT UNITS An integral part of an advanced spacecraft design is to have advanced Orbital Replacement Units, or ORUs. An advanced ORU would have reduced electrical interfaces, light-weight installation hardware Note that the Science Instrument Command & Data Handling (SI C&DH) ORU on HST is probably the best example of an ORU interface. A drive mechanism slides the box into its connectors, and then four key-hole bolts lock the box down. The ORU is a tray with five individual boxes, interconnected with a large number of wires using standard "D-type" connectors. All the EVA interfaces are on the tray, including EVA handle and tether loop. This ORU has not yet been changed out on-orbit. [Ref 6] The worst ORU on HST is the Power Conditioning Unit (PCU), which is going to be replaced on STS-109, SM-3B. This is a large box with 36 circular connectors that do not have wing-tabs. Also, there is no way to dead-face all the connectors to within the low- current limits set by JSC for EVA safety. Changeout of the PCU is a race against time, with several EVA actions needed for preventing thermal and other types of damage to various other units on the spacecraft whose power must be shut off, followed by the

Page 10 of 12 difficult demating of the 36 connectors, changing out the box that is held down by 10 keyhole bolts, mating 36 connectors and returning the spacecraft to a nominal configuration. It has been a major challenge to accomplish all this in a single EVA day of 6 hours, so that the spacecraft will be left overnight in a state in which it could survive if it had to be released due to a shuttle emergency. [Ref 16] The MMS was designed to have large modules that could be replaced either by an astronaut in a spacesuit or by a simple robot. The experience on HST involves many smaller ORUs, such as rate sensor units (gyros), tape recorders, and the computer, as well as larger ORUs such as the 800-pound science instruments. A small number of wing-tab connectors has been a simple matter. A large number of standard aerospace circular connectors such as on the DIU or PCU has proven to result in a very difficult procedure. Changeout of hardware not designed for servicing has also been a challenge, including the S-Band Single Access Transmitter (SSAT), which included SMA-type connectors, was very difficult and required special tools and extensive training. The entire process of deciding how to bundle electronics into ORUs should be revisited in light of the on-orbit experience and advances in electronics design and miniaturization.

ADVANCED SERVICING TECHNOLOGY If servicing of highly sensitive telescopes is going to occur at much higher altitudes, there are other issues which must be addressed. Spacesuits have been improved recently for their performance in cold conditions, but the thermal environment at Earth-Moon L1 will be even more severe. The current suits provide internal cooling by venting to the ambient vacuum. This may produce a contamination source that is unacceptable to some advanced systems. Humans must also be able to get to the worksite, and be protected from the radiation environment once there. The propulsion system of the servicer must somehow not contaminate the observatory. Mechanical loads from human servicers must be kept to very low levels. A next generation of the Space Vision System will be needed, so that a human or robot can have full knowledge of the relative position and orientation of hardware being linked together. Such a system should have small features that interact with a laser or similar system, rather than have the very large targets that need visible light illumination used in the Space Vision System. Another solution is to implement cameras that have full view of interfacing structures. [Ref 17] One possible solution to many of these problems may be a pod, similar to the one depicted in the movie "2001 A Space Odyssey" but with much better manipulators. This is also the model put forth by Werner von Braun in at least one of his films produced with Disney back in the 1950's. Such a device could put the human where he can see the worksite well, and operate robotic devices without a time delay. [Ref 8] The original type of solar arrays on the HST serve as an example, and to some extent a prototype of light-weight, flimsy space structures that are deployed on orbit and remain deployed during servicing. To protect these deployed structures, several precautions are taken regarding operation of the space shuttle rockets, the structural rigidity of the carrier that HST is mounted to, venting of the EVA airlock, and other EVA actions. These are all carefully analyzed, and the results documented in the flight rules. [Ref 18]

Page 11 of 12 CONCLUSION Satellite servicing may face a bright future within a few years. A NASA vision exists of expanding human servicing beyond low earth orbit. This will require technology development in areas that have been largely dormant for several years, including EVA tools, techniques and requirements. The design of free-flyer serviceable spacecraft is to some extent a lost art, and will need to be revived. Education is needed for design engineers in the history of this field, and it must be brought up to date with current and future space technology.

REFERENCES Ref 1: Walking to Olympus: An EVA Chronology, David S. F. Portree and Robert C. Treviño, Monographs in Aerospace History Series #7, October, 1997. Available at: http://spaceflight.nasa.gov/spacenews/factsheets/pdfs/EVACron.pdf Ref 2: http://spaceflight.nasa.gov/shuttle/archives/ Ref 3: On-Orbit Servicing of Space Systems, Donald M. Waltz. Krieger Publishing Company, Malabar, Florida, 1993 Ref 4: Supplement to On-Orbit Servicing of Space Systems, Donald M. Waltz. Krieger Publishing Company, Malabar, Florida, 1998 Ref 5: Multimission Modular Spacecraft (MMS) External Interface Specification and User's Guide, S- 700-11, NASA Goddard Space Flight Center, 9/15/81 Ref 6: Conversation with Ron Sheffield, Lockheed Martin Missiles and Space, October 11, 2001 Ref 7: B. McCandless, "Shuttle Supported Satellite Servicing (S4), Satellite Servicing Workshop IV presentation package, June 21-23, 1989 Ref 8: Conversation with Rud Moe, NASA GSFC, October 15, 2001 Ref 9: http://hubble.gsfc.nasa.gov/servicing-missions/ Ref 10: "AXAF Servicing Design Concepts", Charles F. Lillie, Satellite Servicing Workshop IV presentation package, June 21-23, 1989 Ref 11: Design for On-Orbit Spacecraft Servicing, AIAA Proposed Guide G-042, American Institute of Aeronautics and Astronautics, Washington, DC, 1991 Ref 12: National Aeronautics and Space Administration, Strategic Plan 2000, Washington, DC, 2000. Ref 13: National Aeronautics and Space Administration, the space science enterprise strategic plan, Washington, DC, November 2000. Ref 14: "Site Selection and Deployment Scenarios for Servicing of Deep-Space Observatories", H. Willenberg, M. Fruhwirth, S. Potter, S. Leete and R. Moe [publication pending in proceedings of IEEE Conference, 2002] Ref 15: JSC 26626A, Extravehicular Activity (EVA) Hardware Generic Design Requirements Document, EVA and Crew Equipment Projects Office, NASA Johnson Space Center, 1995 Ref 16: JSC-48024-109, EVA Checklist, STS-109 Flight Supplement, Mission Operations Directorate, EVA, Robotics, and Crew Systems Operations Division, NASA Johnson Space Center, 2001 Ref 17: Conversation with Jim Newman, NASA Johnson Space Center, 10/10/01 Ref 18: NSTS-18308-STS-109 Space Shuttle Operational Flight Rules Annex, Flight STS-109, Mission Operations Directorate, 2001. Biographical Information: Stephen J. Leete is an Aerospace Engineer at the NASA Goddard Space Flight Center in Greenbelt, MD. He is currently the EVA Systems Engineer for the Hubble Space Telescope Development Project, supporting the servicing missions. Mr. Leete has been the Instrument Systems Engineer for the CIRS instrument on the Cassini mission to Saturn and the Differential Microwave Radiometer on the Cosmic Background Explorer. He holds an MS from George Washington University and a BSE from Princeton University.

Page 12 of 12